Optimal Planning of Residential Microgrids Based on Multiple Demand Response Programs Using ABC Algorithm

The smart grid has revolutionized the conventional electricity grid with the proposition of demand-side management (DSM). A DSM program enables the user to schedule its energy consumption in compliance with any pricing signal. This scheduling helps the grid operator to reduce the peak load demand and jointly benefits the user to reduce its electricity costs. Despite that, while doing so, it jeopardizes the user’s comfort. In the present paper, the authors have investigated the impact of communal DSM programs on the consumption patterns of users, including single as well as multiple households. The objective is to simultaneously minimize the electricity costs and user discomfort to make a win-win situation for both the grid operator and the user. Therefore, a multi-objective optimization problem (MOOP) has been formed to simultaneously minimize the daily electricity cost, peak to average ratio (PAR) of load demand, user discomfort, environmental emission, and total net present cost (TNPC). In order to evaluate the best scheduling method, sizing scenarios for a residential microgrid in a Southern Pakistani metropolis surrounded by rural areas are presented in this paper. The originality of this article comes from a comparison of the techno-economic and environmental performance of several sizing options for a residential load powered by renewable energy. The artificial bee colony (ABC) algorithm has been selected to solve the MOOP. The DSM programs are based upon different pricing signals, including real-time electricity pricing (RTEP), critical peak pricing (CPP), time of use (TOU), and day-ahead pricing (DAP) pricing. The results of the proposed ABC algorithm are compared with GA and standard algorithms, and they reveal the effectiveness of the proposed method. When demand response is used, the suggested optimization technique shows that the SH spring with PV/WG/grid-connected microgrid is the most investable-reliable sizing option with a minimal TNPC of $\$ $ 1405.18 for DAP tariff with SH spring. Additionally, with a reduction in emissions of 6699 kg/yr, DAP tariff with SH spring shows that PV/WG/DG/grid-connected system has the greatest impact on the environment. For DAP tariff with SH spring, optimal sizes of PV, WG and converter are 26.7 kW, 30 kW and 6.67 kW, respectively.


I. INTRODUCTION
Since its inception, the generation capacity of a power grid is always aimed to meet the peak load demand fully from the user side instead of the average demand. That demands which peak-load power plants such as diesel generators and gas turbines fulfill at the maximum cost of operation. Over time the existing power grid is dealing with abundant challenges comprising of increasing load demand, old infrastructure, and absence of interaction and security issues. In this perspective, the smart grid has reshaped the operation of the conventional grid with the concept of demand-side management (DSM) [1], [2]. DSM has been presented to reshape the time pattern and extent of load demand by scheduling it. DSM further enhances the operational capabilities of a smart grid in several domains that include infrastructure construction, electricity market management and control of decentralized energy [3]. The main target of DSM is to schedule the load in strong correlation with low priced generation.
The published literature presents various methods which have been applied for DSM in smart grid paradigm. In general, DSM is about modifying the users' energy consumption so that to achieve a win-win situation for both user and utility grid [4]. Several DSM methods have been presented in the literature that includes peak cutting, shifting the load to the time when demand is low, the changeable shape of load, growth of load strategically, strategic conservation, and valley filling for this purpose [5]. Moreover, the communication infrastructure between the utility and user can also be managed by DSM. Also, it permits the incorporation of distributed energy resources (DERs) to enhance energy usage profile.

II. RELATED WORK
In [6], distributed DSM algorithms are applied that are built upon the game theory setup and proximal decomposition for minimizing energy payment by using energy storage devices and appliance scheduling. In [7], the working of home energy management controller (HEMC) is developed for scheduling of energy consumption based on heuristic algorithms including genetic algorithm (GA), binary particle swarm optimization (BPSO) and ant colony optimization (ACO). The objectives were to reduce electricity bill and PAR with integration of renewable energy sources (RES). In [8], cost of electricity and PAR are lessened by employing intelligent programmable communication thermostat by the use of GA to manage electricity load during the limitation of comfort. In [9], a distributed generation (DG) planning model is presented which considers DSM and system reorganization at the same time to reduce the total cost over the planning horizon. For energy optimization, authors in [10], present HEMS design and classify the domestic appliances. The purpose of the aimed design is to lower the cost of electricity and to address the degree of uncertainty associated with various types of loads by using fractional programming approach. For the optimization of energy at domestic area, authors in [11], present HEMS model in which DSM methods are applied in cooperation with time-differentiated rates, load priority and DGs using GA.
In [12], renewable energy sources (RESs) are incorporated in day ahead scheduling of micro-sources for minimization of the generation and startup expenses of the RESs by VOLUME 10, 2022 applying a combined differential evolution (DE) and Harmony search algorithm (HSA). In [13], a household load scheduling with incorporation of day ahead costing plan is demonstrated by using a hybrid teacher learning and genetic algorithm (TLGA). The authors in [14], propose a HEMS built using binary particle swarm optimization (BPSO) for reduction in the expenses of electricity with minimum consumer discomfort. The authors in [15], have used integration projection evolutionary algorithm (IPEA) to schedule the appliances at the time when electricity pricing is low. In [16], a non-cooperative game theory based DSM method is proposed to schedule the energy consumption while using storage appliances. In [17], a residential demand response with RES is proposed. In [18], a Ladson generalized bender algorithm (LGBA) is used to improve the energy usage profiles of multi-households with minimum discomfort level. The authors in [19], have used GA to execute demand response program using RES to minimize customer's electricity charges and peak load. In [20], authors have demonstrated energy management scheduling within a domestic area using two horizon algorithm (THA) to minimize electricity expense with reduction in computational time. In [21], a stochastic programming model is discussed to optimally schedule DERs. The objective of this work is to minimize energy expense and CO 2 emissions. In [22], a model is introduced to address the supply-demand balance constrained grid with a goal to get minimized generation expenses, CO 2 emissions and utility losses.
In [23], a multi-objective PSO algorithm is proposed to minimize dynamic economic and emission dispatch problem with consideration of DSM. In [24], a multi-objective PSO algorithm is proposed to minimize dynamic economic and emission dispatch problem with consideration of DSM. In [25], authors have introduced a design built on optimal financial options for the management of microgrid using GA. In [26], authors have proposed a multi-time scale optimization (MSO) for scheduling the energy utilization of various appliances. In [27], GA for DSM in domestic, commercial, industrial areas. In [28] and [29], both dynamic programming (DP) and integer linear programming (ILP) techniques are applied to reduce PAR and electricity expenses. However, these techniques are ineffective concerning the computational time. In [30], an energy management model with different types of appliances is presented for cost minimization and PAR. The authors in [31], have proposed a HEMS for load shifting to reduce in PAR and electricity expenses by using dynamic programming. In [32], economic analysis of DSM programs on unit commitment is discussed to schedule a load profile.
In [33], an improved PSO and shuffled frog-leaping algorithm (SFLA) based energy consumption and forecasting model is presented to implement the DSM plan. The authors in [34] have proposed the scheduling of generation units integrated with DSM programs to reduce electricity costs. Imperialist competitive algorithm (ICA) is employed for this problem. In [35], ABC algorithm is proposed to schedule residential loads subjected to cost and time constraints, however PAR and user comfort is neglected. In [36], electricity price and demand forecasting schemes have been presented to reduce peak load. Authors have used a three-part forecast model which includes a new flexible packet wavelength transformation, a multi input multi output (MIMO) model and ABC algorithm for a stable prediction of price and load. In [37], a scheduling technique for the equipment used in rice industry is proposed by using three optimization algorithms including DE, PSO and ABC. The objectives were to minimize feeder load and cost of electricity. Results prove that ABC algorithm gives better results than other algorithms. In [38], an ABC algorithm based on improvedglobal-best-guided-approach (IGBGA) and adaptive-limitstrategy (ALS) is applied in to reduce total electricity cost and total energy consumption with integration of PV generation. Authors in [39], have minimized the total investment, operating, and outage costs considering DSM by using ABC algorithm. In [40], cost saving of commercial area by shifting loads is achieved by ABC algorithm while using time of use (TOU) tariff. In [41], combined ABC-GA algorithms are used to solve OPF and operation problems, respectively. In [42], different load profiles are used to minimize system operation cost and losses by using TOU as DR program with ABC algorithm.
The authors proposed DSM in [43] by including VESS which is a common heating/cooling system used as a microgrid in residential buildings with multi-objective optimization problem. In [44], DSM as PODR is proposed to solve multi-objective optimization problem. In [45], MINLP is used for two heating/cooling based grid-connected residential microgrid system with multi-objective optimization. NSGA-II is then used for searching Pareto front. The optimal scheduling is handled using AHP approach. In [46], dayahead forecasting with DSM and EED is proposed with two versions of PSO to obtain DSM based load control plan and EED based power supply plan. In [47], MOGA is employed for solving multi-objective DSM problem. In [48], C&CG is used for minimization of multi-objective DSM problem. The optimal solution from non-dominated Pareto solutions is sorted by the fuzzy decision-making approach. The RERs uncertainty, the stochastic loads and energy prices are modelled using Monte Carlo method.
The majority of evolutionary algorithms are stochastic meta-heuristic procedures that draw their inspiration from nature [49]. Some of the often used and most appropriate algorithms in the context of smart grids include PSO [50], GA [51], ABC [52], GWO [53], TLBO [54], FF [55], CS [56], WO [57], and MPA [58]. PSO is inspired by bird-flock and is developed by Kennedy and Eberhart in 1995. GA is inspired by genetics and is developed by AS Fraser in 1957. ABC is inspired by honey-bee and is developed by Karaboga  It has been observed that the solution of the DSM problem with ABC algorithm is limited to certain objectives including minimization of electricity cost and load scheduling only. Table 1 shows synopsis of the application of ABC algorithm applied to DSM problems. A complete multi-objective DSM problem solved with the ABC algorithm has been seen to be missing in the literature. In the current paper, the authors have extended the scope of the DSM problem by reforming it as a multi-objective DSM and proposed the ABC algorithm for the solution of the optimization problem. ABC algorithm is selected due to its limit cycle ability which reduces the chance of local optimization, and therefore increases its diversity. Key contributions of this paper include: • The DSM of a single household (SH) and multi household (MH) is proposed to simultaneously minimize the multiple objectives. Two test cases have been created and are being looked into.
• In the first case, five objectives, including electricity cost, PAR, user discomfort, TNPC and environmental emissions have been simultaneously minimized by using the ABC algorithm, and results are compared with standard algorithms. The consideration of user discomfort ensures that average waiting of appliances is reduced to increase the luxury of user.
• In second case, three objectives, including electricity cost, PAR, user discomfort have also been minimized using the ABC algorithm. In this case, the results are compared with GA because existing literature lacks such a test case.
• Four types of tariffs including real time electricity pricing (RTEP), critical peak pricing (CPP), time of use (TOU), and day-ahead pricing (DAP) have considered while solving the optimization problem. The existing literature lacks detailed analysis with four tariffs which are investigated in this paper.
• To examine the role of electrical power generation in sustainable development and analysis of the TNPC associated with the rise in emissions is carried out. • A multi-objective, renewable energy-based method to sizing DERs of proposed area in Pakistan's south is proposed.
• Depending on the availability of power conversion sources, two operating sizing options (TNPC and emission) with thorough analyses are presented with different energy sources.
• The economic and environmental impacts are studied for producing power in remote or grid-connected microgrids in underdeveloped nations like Pakistan, as well as in cities surrounded by rural areas.
• How system sizing is affected by four DR programs (such RTEP, CPP, TOU, and DAP) is investigated. Additionally, the effects of DR programs on the share of VOLUME 10, 2022 renewable energy (RERs) and overall net present cost (TNPC) is analyzed.
• Comparing several time-based DR programs for the scheduling problem is studied. The findings of this study will be very helpful for designing tariffs since they need to make sure that the tariff they choose is as effective as feasible.
• Modifying the daily load curve using RTEP, CPP, TOU, and DAP. Rest of the paper is as follows. Section 2 shows the proposed multi-objective DSM of a SH and MH. It comprises of formulation of the problem. In section 3, ABC algorithm is explained. Whereas, test cases and results are presented in section 4. The conclusions are drawn in section 5.

III. DEMAND-SIDE MANAGEMENT
DSM is shown in Fig. 1, which envisions a single household (SH) comprising of different appliances as shown in Table 2. The ultimate objectives are to simultaneously minimize the electricity cost, PAR, user discomfort and environmental emissions by scheduling the load in the presence of a PV system. These objectives are subject to constraints of grid capacity limitations, time of operation limitations and user discomfort limitations. The optimal values of objectives are minimized by using the ABC algorithm based on two types of tariffs including RTEP and CPP in two test cases.

A. PROBLEM FORMULATION 1) ELECTRICITY COST
The first objective is to minimize the electricity cost by scheduling the residential load during low-cost hours. For a single household, Y = {y 1 , y 2 , y 3 . . . . . . y N } such that y 1 , y 2 , y 3 . . . . . . y N denotes each appliance through the time range t ∈ T {1, 2, 3 . . . ., N}. Each time slot constitutes one hour and the total time range is 24h (T= 24), considering a single day. The total energy utilization of all appliances in a day can be mathematically represented as shown in Eq. (1) [59]: where, E C,TL represents the total energy utilization of all appliances in a day which is the sum of energy utilization of all appliances over a time period of 24h, and its unit is in kWh. Similarly, E (yq,t) represents the power rating of each appliance in kW. The appliances are categorized into three groups. Each appliance is classified based on user preference, energy utilization and time of operation. Assume that Y n represents a set of appliances, and Y n = {E a U R a U S a }. Where, E a represents elastic appliances, R a represents frequently operated appliances and S a represents shift able appliances.

a: ELASTIC APPLIANCES
The energy utilization profile and time period of these appliances can be flexibly adjusted by DSM that is why these are considered as flexible appliances. These include air conditioner (AC), water heater, refrigerator and water dispenser. The total energy consumed in 24h by E a can be calculated as shown in Eq. (2) [59]: where, c Ea,TL is the total energy consumed in kWh in a day, u t Ea is the power rating of elastic appliances in kW and δ(t) is the ON-OFF state. The total cost of elastic appliances in a day can be calculated as shown in Eq. (3) [59]: where, e TL Ea represents the total cost of elastic appliances in a day in cents/h, E(t) represents electricity pricing signal and δ(t) is the ON-OFF state.

b: FREQUENTLY OPERATED APPLIANCES
The energy utilization profile of frequently operated appliances cannot be modified by DSM that is why these are called fixed appliances. These include oven, dishwasher, water pump and vacuum pump. The total energy consumed in 24h by R a can be calculated as shown in Eq. (4) [59]: where, c Ra , TL is the total energy consumed in kWh in a day, L t Ra is the power rating of frequently operated appliances in kW and δ(t) is the ON-OFF state. The total cost of frequently operated appliances in a day can be calculated as shown in Eq. (5) [59]: where, e TL Ra represents the total cost of frequently operated appliances in a day in cents/h, E(t) represents electricity pricing signal and δ(t) is the ON-OFF state.

c: SHIFT ABLE APPLIANCES
Time of operation of shift able appliances is changeable by DSM with any time interval without affecting their performance. The limitation with these appliances is that when they are switched ON their duration of functioning has to be completed. These include cloth dryer and washing machine. The total energy consumed by in 24h by S a is shown in Eq. (6) [59]: where, c Sa , TL is the total energy consumed in kWh in a day, v t Sa is the power rating of shift able appliances in kW and δ(t) is the ON-OFF state. The total cost of shift able appliances in a day can be calculated as shown in Eq. (7) [59]: where, e TL Sa represents the total cost of shift able appliances in a day in cents/h, E(t) represents electricity pricing signal and δ(t) is the ON-OFF state. The total energy utilization of appliances during a period of 24 hours is given as shown in Eq. (8) [59]: where, c TL is the sum of energy consumed in kWh by all types of appliances. The total cost per day of R a , S a and E a appliances is calculated as shown in Eq. (9) [59]: where, e TL represents the total cost of all appliances in a day in cents/h. which is the sum of cost of elastic, frequently operated and shift-able appliances. The purchased cost of electricity from grid is [41]: where, T GP and P GP represent the tariff and purchased power from grid, respectively. Selling electricity to grid is not considered in this paper. Hence, excess electricity may be available and can be analyzed in future reseach. The total electricity cost per day is calculated as shown [59]: where, Cost T is the total electricity cost per day, E(t) is the electricity pricing signal, g yi denotes the energy utilization of the appliances in kWh. The first objective can be mathematically represented as shown [59]:

2) PAR
The second objective is to minimize PAR by scheduling the residential load during low-cost hours. PAR is the ratio of the maximum combined load used during a specific time interval to the average of the combined load. Grid stability is damaged when PAR is high. It also increases the electricity cost of the user. Simultaneously when PAR is low, the stability of the grid is improved and electricity cost is minimized as well.
The second objective can me mathematically represented as shown [59]: min PAR = L peak L avg (15) where, L peak and L avg indicate the maximum combined load and average load in kW in 24h, respectively, and c T (t) denotes the total hourly energy utilization of appliances.

3) USER DISCOMFORT
The third objective is to minimize user discomfort. During load scheduling the energy utilization patterns of R a could not be shifted. On the other hand, energy utilization patterns of S a and E a are changeable to run during off-peak hours. However, it causes inconvenience for the user, and therefore average waiting time of appliances is reduced to minimize user discomfort. To estimate the waiting time of appliances, startup time instant a α and closing time instant b β , is assumed such that (a α < b β ). The waiting time of the appliances is calculated as shown [59]: where, W represents the waiting time in h and T r is the time of request of an appliance. The average waiting time of the appliances is calculated as shown [59]: where, W avg is the average waiting time of all appliances and Y N is the set of appliances. The third objective can be mathematically represented as shown [59]:

4) ENVIRONMENTAL EMISSIONS
The fourth objective is to minimize the environmental emissions including CO 2 , NO x and SO 2 . The emissions are taken into account to cater concerns of environmental protection VOLUME 10, 2022 and climate change. These emissions are measured in kg/h and are calculated as shown [60]: where, F E is the amount of environmental emissions in kg/h, P gi is grid's power in kW. a i , b i , c i d i and e i are emission coefficients. Table 2 shows the values of these coefficients. The fourth objective can be mathematically represented as shown [60]: The above-mentioned objectives are subject to certain constraints as stated below. The first constraint is about the limitations of grid capacity. The total energy utilization of the appliances during time interval t ∈ T should be less than equal to C g . Therefore, the total energy utilization is limited as shown: where, c TL is the sum of energy consumed in kWh by all types of appliances and C g is the grid's maximum capacity to supply power. The second constraint is about the limitations of energy consumption of E a and S a . The third constraint is about limitations of user discomfort. The W avg of E a and S a is restricted to be less than 5h [59].

IV. PROPOSED ABC ALGORITHM
ABC algorithm was at first introduced in 2003 by Karobaga [52]. It is an algorithm which was founded on the foraging behaviors of honey bees. At the moment it has been applied to different research problems by several researchers [61], [62], [63]. The employee, spectator, and scout bees are the three sorts of bees that the ABC algorithm is based on. The method is divided into three phases: the worker bee phase, the observer bee phase, and the scout bee phase.
• In employee bees' phase, the employee bees search for the food sources and store this food source information in memory. Each food source denotes a solution of the optimization problem. The employee bees pass on this food source information to the onlooker bees.
• In onlooker bees' phase, the onlooker bees stay at the hive and evaluate the food source information brought by the employee bees. They check the nectar amount of the food sources and decide to whether accept or reject them. This is normally done by monitoring the waggle dance of the employee bees. The onlooker bees also store the food source information and respective decisions in memory. Based on decision, they direct the employee bees for next iteration of search.
• A worker bee is labelled a scout bee if she provides the same food source information for a predetermined number of cycles (known as limit cycles) without producing improved results. The scout bee is told to look for fresh food sources in new, arbitrary locations. The likelihood of local optimization is decreased by this search procedure, and algorithm variety is increased.
In the current paper, a food source represents a solution in the form of a scheduled load. Similarly, the nectar amount represents electricity cost, PAR, user discomfort and environmental emissions. The flow chart of ABC algorithm is shown in Fig. 2 and the step by step procedure is as follows: • In first step, parameters of the ABC algorithm are initialized. These parameters include the number of employee bees, number of onlooker bees, number of limit cycles, initial food sources (population) and max number of iterations (MI).
• The worker bees look for food sources in this step. Each worker bee finds a food source, measures the nectar content, remembers its location, and transmits the information to a watcher bee.
• In this step, the observer bees assess the data on the food sources (including the amount of nectar brought by the worker bees). The spectator bees memorise all the data acquired from the nearby worker bees and choose the finest food source from it. The observer bees then lead the worker bees to look for more food sources. The worker bees keep looking for new food sources and gathering information about them. Once more, spectator bees assess the data and repeat the process of deciding on the greatest food source. The iterative process begins in this manner and is continued till the MI. If any observer bee notices repeated repeating of food source information by any employee bee for a predetermined number of cycles, that employee bee is classified as a scout bee during the evaluation process. This scout bee is instructed to search for food sources in different, illogical locations. Each observer bee compares the information about the food source with the knowledge about the nearby food source and changes this information in memory by [42]: where, R new shows updated solution, R oldm shows an outdated fix at an arbitrary food source location m and R oldn shows the dated remedy at the location of the food supply n. u is the random number between [−2, 2]. The probability R R of the fitness of the food sources is 116570 VOLUME 10, 2022 calculated by [64]:

8) FALL MH
The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.04 and 2.30, respectively. The unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. For SH summer load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 10.10 kW (at 08:00 hour), 11.60 kW (at 03:00, 04:00) and 11.60 kW (at 04:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.15 and 2.15, respectively. The unscheduled load shows minimum PAR. While PAR is same with GA as compared to ABC. For SH summer load with DAP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 10.10 kW (at 08:00 hour), 11.60 kW (at 02:00) and 11.00 kW (at 08:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.15 and 2.04, respectively. The unscheduled VOLUME 10, 2022  load shows minimum PAR. While PAR is higher with GA as compared to ABC. Fig. 103 shows the hourly load scheduling of fall SH with four tariffs by using GA and proposed ABC algorithm. For SH fall load with RTEP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 18.17 kW (at 13:00 hour), 17.95 kW (at 02:00) and 20.45 kW (at 01:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.60 and 2.88, respectively. The unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. For SH fall load with CPP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 18.17 kW (at 13:00 hour), 14.55 kW (at 02:00, 05:00) and 16.05 kW (at 02:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.05 and 2.26, respectively. The GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. For SH fall load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 18.17 kW (at 13:00 hour), 17.45 kW (at 06:00) and 18.95 kW (at 04:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.46 and 2.67, respectively. The GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. ABC shows highest PAR in this case. For SH fall load with DAP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 18.17 kW (at 13:00 hour), 12.05 kW (at 20:00) and 13.90 kW (at 07:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 1.70 and 1.96, respectively. The GA scheduled load shows minimum PAR. While PAR is lower with GA as  compared to ABC scheduled and unscheduled load. Unscheduled load shows highest PAR in this case. Fig. 104 shows the hourly load scheduling of winter MH with four tariffs by using GA and proposed ABC algorithm. For MH winter load with RTEP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 628.50 kW (at 13:00 hour), 578.50 kW (at 10:00) and 567.50 kW (at 19:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.34, 2.59 and 2.65, respectively. The unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. For MH winter load with CPP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 628.50 kW (at 13:00 hour), 447.50 kW (at 03:00) and 639.50 kW (at 17:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.34, 2.12 and 2.86, respectively. The GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. ABC scheduled load shows highest PAR in this case. For MH winter load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 628.50 kW (at 13:00 hour), 520.00 kW (at 01:00) and 542.50 kW (at 19:00 and 21:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.34, 2.36 and 2.46, respectively. The unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. ABC scheduled load shows highest PAR in this case. For MH winter load with DAP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 628.50 kW (at 13:00 hour), 497.50 kW (at 04:00) and 517.50 kW (at 23:00 hour), respectively. The PAR for VOLUME 10, 2022   unscheduled, GA scheduled and ABC scheduled load are 2.34, 2.23 and 2.32, respectively. The GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. Unscheduled load shows highest PAR in this case. Fig. 105 shows the hourly load scheduling of spring MH with four tariffs by using GA and proposed ABC algorithm. For MH spring load with RTEP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 390.00 kW (at 01:00, 11:00 hour), 319.00 kW (at 02:00) and 325.00 kW (at 03:00, 05:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.77, 1.97 and 1.91, respectively. The unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. For MH spring load with CPP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 390.00 kW (at 01:00, 11:00 hour), 350.00 kW (at 01:00) and 350.00 kW (at 03:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.77, 2.08 and 2.06, respectively. The ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. ABC scheduled load shows lowest PAR in this case. For MH spring load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 390.00 kW (at 01:00, 11:00 hour), 295.00 kW (at 07:00) and 315.00 kW (at 01:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.77, 1.82 and 1.85, respectively. The unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. ABC scheduled load shows highest PAR in this case. For MH spring load with DAP tariff, the unscheduled,  GA scheduled and ABC scheduled peak loads are 390.00 kW (at 01:00, 11:00 hour), 344.00 kW (at 03:00) and 300.00 kW (at 02:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.77, 2.02 and 1.78, respectively. The unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. Fig. 106 shows the hourly load scheduling of summer MH with four tariffs by using GA and proposed ABC algorithm. For MH summer load with RTEP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 505.00 kW (at 08:00 hour), 545.00 kW (at 02:00) and 505.00 kW (at 01:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.24 and 2.12, respectively. The unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. For MH summer load with CPP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 505.00 kW (at 08:00 hour), 580.00 kW (at 04:00) and 505.00 kW (at 04:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.46 and 2.12, respectively. The unscheduled load shows minimum  PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. For MH summer load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 505.00 kW (at 08:00 hour), 530.00 kW (at 02:00) and 400.00 kW (at 05:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.35 and 1.73, respectively. The ABC scheduled shows minimum PAR. While PAR is higher with GA as compared to ABC. ABC scheduled load shows lowest PAR in this case. For MH summer load with DAP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 505.00 kW (at 08:00 hour), 570.00 kW (at 02:00) and 545.00 kW (at 01:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 1.87, 2.56 and 2.35, respectively. The unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. Fig. 107 shows the hourly load scheduling of fall MH with four tariffs by using GA and proposed ABC algorithm. For MH fall load with RTEP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 908.50 kW (at 13:00 hour), 842.50 kW (at 07:00) and 714.50 kW  (at 17:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.80 and 2.35, respectively. ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC and unscheduled load. ABC scheduled load shows lowest PAR in this case. For MH fall load with CPP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 908.50 kW (at 13:00 hour), 580.00 kW (at 04:00) and 642.50 kW (at 21:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 1.89 and 2.10, respectively. GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows lowest PAR in this case. For MH fall load with TOU tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 908.50 kW (at 13:00 hour), 639.50 kW (at 17:00) and 622.50 kW (at 03:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.13 and 2.03, respectively. ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. ABC scheduled load shows lowest PAR in this case. For MH fall load with DAP tariff, the unscheduled, GA scheduled and ABC scheduled peak loads are 908.50 kW (at 13:00 hour), 725.00 kW (at 04:00)  and 672.00 kW (at 16:00 hour), respectively. The PAR for unscheduled, GA scheduled and ABC scheduled load are 2.56, 2.27 and 2.19, respectively. ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled. ABC scheduled load shows lowest PAR in this case. Table 7 shows the comparison of GA and proposed ABC algorithm. For SH winter load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. Average waiting time (AWT) is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. Average daily cost (ADC) is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH winter load with CPP tariff, PAR is minimized by 13.7% with the proposed ABC scheduling. While PAR is higher with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH winter load with TOU tariff, PAR is not minimized with the proposed ABC scheduling. While PAR is same with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared VOLUME 10, 2022   to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH winter load with DAP tariff, PAR is minimized by 2.1% with the proposed ABC scheduling. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH spring load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. AWT is not minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. Unscheduled load shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH spring load with CPP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. AWT is not minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. Unscheduled load shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC  shows lowest ADC. For SH spring load with TOU tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. AWT is not minimized with the proposed ABC scheduling. While AWT is same with GA as compared to ABC. Unscheduled load shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH spring load with DAP tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. AWT is not minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. Unscheduled load shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC. ABC shows lowest ADC. For SH summer load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH summer load with TOU tariff, the unscheduled load shows minimum PAR. While PAR is same with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. ABC shows lowest AWT. ADC is  minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH summer load with DAP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH fall load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. AWT is minimized with the proposed ABC scheduling. While AWT is same with GA as compared to ABC. GA and ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH fall load with CPP tariff, GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. AWT is not minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH fall load with TOU tariff, GA scheduled load shows minimum  PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. ABC shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For SH fall load with DAP tariff, GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. Unscheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC. ABC shows lowest ADC.
It is surprising to note two cases that for SH spring load with DAP tariff; ADC is lower with GA as compared to ABC. For SH fall load with DAP tariff, ADC is again lower with GA as compared to ABC. Fig. 108 shows the comparison of SH TNPC vs emission with four tariffs by using proposed ABC algorithm. Table 8 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH RTEP winter load. Table 9    shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH RTEP summer load. Table 10 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH RTEP spring load. Table 11 shows the optimal sizing based on minimization of emission and TNPC for gridconnected DERs by using proposed ABC algorithm with SH RTEP fall load. Table 12 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH CPP winter load. Table 13 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH CPP summer load. Table 14 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH CPP spring load. Table 15 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH CPP fall load. Table 16 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH TOU winter load. Table 17 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH TOU summer load. Table 18 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH TOU spring load. Table 19 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH TOU fall load. Table 20 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH DAP winter load.   Table 21 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH DAP summer load. Table 22 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH DAP spring load. Table 23 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with SH DAP fall load. Table 24 shows the comparison of GA and proposed ABC algorithm. For MH winter load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. Average waiting time (AWT) is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. ABC shows lowest AWT. Average daily cost (ADC) is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH winter load with CPP tariff, GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. ABC scheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC.  GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH winter load with TOU tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. ABC scheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH winter load with DAP tariff, GA scheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. Unscheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH spring load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. AWT is not minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. Unscheduled load  shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC and unscheduled load. ABC shows lowest ADC. For MH spring load with CPP tariff, ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. ABC scheduled load shows lowest PAR in this case. AWT is not minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC. Unscheduled load shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC and unscheduled load. GA shows lowest ADC. For MH spring load with TOU tariff, the unscheduled load shows minimum PAR. While PAR is lower with GA as compared to ABC. ABC scheduled load shows highest PAR in this case. AWT is not minimized with the proposed ABC scheduling. While AWT is lower with GA as compared to ABC and unscheduled load. GA shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC and unscheduled load. GA shows lowest ADC. For MH spring load with DAP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. AWT is minimized with  the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH summer load with RTEP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. Unscheduled load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH summer load with CPP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC. GA shows lowest ADC. For MH summer load with TOU tariff, The ABC scheduled shows minimum PAR. While PAR is higher with GA as compared to ABC. ABC scheduled  load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC. GA shows lowest ADC. For MH summer load with DAP tariff, the unscheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows highest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is lower with GA as compared to ABC. GA shows lowest ADC. For MH fall load with RTEP tariff, ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC and unscheduled load. ABC scheduled load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC and unscheduled load. ABC shows lowest ADC. For MH fall load with CPP tariff, GA scheduled load shows  minimum PAR. While PAR is lower with GA as compared to ABC scheduled and unscheduled load. GA scheduled load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH fall load with TOU tariff, ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC. ABC scheduled load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC. For MH fall load with DAP tariff, ABC scheduled load shows minimum PAR. While PAR is higher with GA as compared to ABC scheduled. ABC scheduled load shows lowest PAR in this case. AWT is minimized with the proposed ABC scheduling. While AWT is higher with GA as compared to ABC and unscheduled load. ABC shows lowest AWT. ADC is minimized with the proposed ABC scheduling. While ADC is higher with GA as compared to ABC. ABC shows lowest ADC.  It is surprising to note that for MH spring load with RTEP tariff, ADC is higher with GA as compared to ABC and unscheduled load. For MH spring load with CPP tariff, ADC is lower with GA as compared to ABC and unscheduled load. GA shows lowest ADC. For MH spring load with TOU tariff, ADC is lower with GA as compared to ABC and unscheduled load. GA shows lowest ADC. For MH summer load with DAP tariff, ADC is lower with GA as compared to ABC. GA shows lowest ADC. For MH fall load with RTEP tariff, ADC is higher with GA as compared to ABC and unscheduled load. Fig. 109 shows the comparison of MH TNPC vs emission with four tariffs by using proposed ABC algorithm. Table 25 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH RTEP winter load. Table 26 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH RTEP summer load. Table 27 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH RTEP spring load. Table 28 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH RTEP fall load.   Table 29 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH CPP winter load. Table 30 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH CPP summer load. Table 31 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH CPP spring load. Table 32 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH CPP fall load. Table 33 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH TOU winter load. Table 34 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH TOU summer load. Table 35 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH TOU spring load. Table 36 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH TOU fall load. Table 37 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH DAP winter load. Table 38 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH DAP summer load. Table 39 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH DAP spring load. Table 40 shows the optimal sizing based on minimization of emission and TNPC for grid-connected DERs by using proposed ABC algorithm with MH DAP fall load.

VI. CONCLUSION
In this paper a DSM has been proposed to simultaneously minimize electricity cost, PAR, user discomfort, TNPC and environmental emissions using ABC algorithm for a SH and MH with RTEP, CPP, TOU and DAP tariffs. Two test cases were formed and investigated. In the first test case, three objectives, including electricity cost, PAR and user discomfort were simultaneously minimized by using ABC algorithm, and results were compared with standard heuristic algorithms including WDO, HSA, GA and GHSA. It has been observed that proposed ABC algorithm outperforms these algorithms. In the second test case, five objectives, including electricity cost, PAR, user discomfort, TNPC and environmental emissions were simultaneously minimized by using ABC algorithm, and results were compared with GA. It has been observed that in terms of electricity cost both algorithms have major differences except some cases where almost same performance is observed. For PAR and user discomfort, ABC algorithm outperforms GA except multiple cases where GA performance is better. As of TNPC and environmental emissions are concerned, ABC algorithm is applied for the optimal sizing of grid-connected DERs with four tariffs such as RTEP, CPP, TOU and DAP. There is a trade-off between TNPC and emissions for optimal sizing of grid-connected DERs. Following points can be concluded from this research article: • Out of thirty-two case studies, thirty case studies shows minimum average daily cost (ADC) with ABC algorithm. Only two case of single home (SH) are not minimized with ABC. This shows better performance of ABC over GA.
• Residential loads' peak hours, which are at night, are reduced when demand response (DR) programs alter the load curve, but daytime demand is increased. These operations lower the value of maximum loads while raising the load factor. These plans also boost PV and WG capacity, which is especially useful for systems without any storage facility integration in the morning and afternoon.
• By implementing DR programs, grid power consumption is reduced while renewable energy shares of PV and WG increase. Environmental emissions are also reduced.
• Through DR initiatives, PV and WG capacity are raised, and the maximum amount of grid electricity purchased is decreased.
• According to the results, it is preferable to manage consumer participation levels before system establishment and system sizing to manage the hybrid system's configuration, the anticipated system cost, emissions, and revenues in the event that a grid connection is available.
• Developers can meet the required system configuration and preferred DR participation percentages based on the intended priorities in system planning, such as cost and emissions.
permanently with the Department of Electrical Engineering, University of Engineering and Technology Taxila, Pakistan. His eight funding proposals are under review. Apart from book chapters, he has published more than 28 publications in journals and conferences in last three years. He has worked in Pakistani industry and academia for more than 13 years. He has been actively involved in a variety of professional activities including planning conferences and workshops, overseeing the academic and administrative responsibilities of his department, and getting their Electrical Engineering Undergraduate Program accredited by Pakistan Engineering Council. His research interests include microgrid, smart grid, energy economics, control applications in power systems, renewable resources integration, power electronics applications to power systems, and distributed generation. He has been serving as a Reviewer for many journals, ASAD WAQAR received the degree in electrical engineering from UET Taxila, in 2002, the master's degree in electrical power engineering from RWTH Aachen, Germany, in 2011, and the Ph.D. degree in electrical engineering from the Huazhong University of Science and Technology, China, in 2016. He worked at different industries for several years. Currently, he is working as an Professor with the Department of Electrical Engineering, Bahria University, Islamabad, Pakistan. He has successfully supervised more than 30 master's thesis students. Currently, he is supervising three Ph.D. students at Bahria University. He has published research articles in many reputed international journals. His research interests include smart grids, microgrid operation and control, power quality, power electronics, network reinforcement planning, demand side management, and big data analysis in power systems. He serves as an Active Reviewer for Applied Energy, International Transactions on Electrical Energy Systems, IEEE ACCESS, Renewable and Sustainable Energy Reviews, and International Journal for Engineering Science and Technology.