A combination of MADM and genetic algorithm for optimal DG allocation in power systems

Distributed Generation (DG) can help in reducing the cost of electricity to the customer, relieve network congestion, provide environmentally friendly energy close to load centers as well as promote system technical characteristics such as loss reduction, voltage profile enhancement, reserve mitigation and many other factors. Furthermore, its capacity is also scalable and it can provide voltage support at distribution level. The planning studies include penetration level and placement evaluation which are influenced directly by DG type. Most of the previous publications in this field chose one or two preferred parameter as basic objective and implement the optimizations in systems. But due to small size of DGs output, placement according to one or two of just technical parameters usually leads to more theoretical results and with incorporation of less DG resources. Furthermore, optimization of one parameter might degrade another system attribute. In this paper a multi-objective placement and penetration level of Distributed Generators (DGs) is examined, concerning both technical and economical parameters of power system using Genetic Algorithm (GA) combined with Multi-Attribute Decision Making (MADM) method. In fact, by using GA best plans for system with incorporation of DG are determined. For approaching such aim, 4 technical parameters of system, including total losses, buses voltage profile, lines capacity limits and total reactive power flow, are consider with appropriate priorities applied to each objective. In the next step, Analytic Hierarchy Process (AHP) along with Data Envelopment Analysis (DEA) is used as a multi attribute decision making technique to form a decision making framework for selecting the best capacity and place of DG units. The attributes are defined as technical and economical parameters. The technical parameters are the voltages on the buses, the reactive power and losses in the transmission lines and the economical parameters are the emissions, congestion and capital cost. The proposed approach is illustrated by case studies on IEEE 30 bus distribution system which demonstrate significant improvement in optimization through this procedure.