A Complete Affordable Control System for Remote Astronomical Observing Accompanied by an Intelligent Controller

This paper comprises two studies; the first one provides an advanced and low-cost implementation for a remote astronomical platform applicable to small and medium-sized telescopes. It has been carried out for the 14-inch observatory, which includes a Celestron German Equatorial (CGE) telescope at the Kottamia astronomical observatory (KAO) in Egypt. This integrated control system is based on embedded systems, internet of things (IoT) technology, row packets communication procedure, and the Transmission Control Protocol (TCP) based on the Internet Protocol (IP). Using this platform, remote astronomers could control the whole system, observe, receive images and view them efficiently and safely without any human physical intervention. The proposed design has been achieved without dependence on commercial control kits or software. Indeed, many previous studies have focused on this field; however, their area of interest was limited or non-affordable. The excellence in this practical research is revealed and compared with others in terms of cost, inclusiveness, and communication speed. The other contribution of this research is to enhance the performance of the telescope pointing and tracking to be adapted with remote action. It has been achieved based on the mathematical model of the telescope where two fractional controllers have been applied, tilt integral-derivative (TID) and integral derivative-tilted (ID-T) controllers. After that, they have been optimized using a recent optimization algorithm called the peafowl optimization algorithm (POA) and compared with one of the well-known algorithms, particle swarm optimization (PSO). Simulation results under the MATLAB/SIMULINK environment reveal that modified ID-T-based POA has minimized the pointing error sharply. Moreover, compared with previous studies, it has significantly improved the telescope system characteristics represented in overshoot, settling, and rising periods.


21
A stronomers and astrophysicists from ancient to now are 22 interested in sky objects observation, such as stars, plan- 23 ets, and galaxies translating observed images into discov-24 eries [1] and data that analyze the universe's construction 25 and our location inside it [2]. The astronomical observa-26 tion is performed inside an observatory that comprises a 27 The associate editor coordinating the review of this manuscript and approving it for publication was R. K. Tripathy . telescope, a dome, an image detector, and a weather sta-28 tion [3]. The optical telescopes gather emitted light from 29 visible sky objects then concentrate it using mirrors [4]. 30 Thus, optical observatories are consistently located at distant 31 sites to offer better seeing, less atmospheric water vapor, 32 and reduced light pollution [5]. Therefore, the attitude of 33 developing telescopes to run remotely has grown exponen- 34 tially to enable astronomers to perform observation from 35 anywhere, avoiding traveling efforts toward the telescope's 36 physical site. 37

VOLUME 10, 2022
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ from north to south. The mathematical model of the tele-93 scope is based on a set of dynamic equations with nonlinear 94 coupled differential equations that contain different terms for 95 moments of inertia, a centrifugal and Coriolis, and a frictional 96 term. It also includes a gravity term; however, it tends to zero 97 in case of accurate telescope balance [20]. 98 Studies have been interested in how to force the telescope 99 to accelerate its motors at a suitable velocity for reaching 100 the selected target in the desired position by controlling the 101 telescope's joint torque. Thus, several controllers have been 102 applied to keep the telescope pointing in reference trajectory 103 with minimum error using different optimization techniques 104 based on artificial intelligence. Based on previous research, 105 the main conclusion is that the fuzzy logic control (FLC) 106 strategies have provided moderate performance. However, 107 FLC has flaws in terms of the rise time (R s ) and the settling 108 time (T s ), which are not accepted in the case of robotic 109 operation. 110 As shown in Table 1, FLC has been applied and optimized 111 by a genetic algorithm (GA); even though the R s was suit-112 able, it produced an unacceptable overshoot (O sh ) [20]. The 113 response with O sh is an impermissible issue due to the use of 114 a fixed charge-coupled device camera (CCD) on the telescope 115 while tracking a sky object [21]. Moreover, although FLC 116 that PSO drives have provided better performance according 117 to previous techniques followed by proportional-integral-118 derivative (PID) controller tuned from FLC, they are not 119 sufficient as the R s was inappropriate in case of remote 120 operation [22]. Furthermore, in the face of system uncer-121 tainties, conventional proportional-derivative (PD) and PID 122 controllers have some difficulty as their defects of high O sh 123 and R s are inappropriate.

125
Since the design of small and medium-sized observatories 126 does not support remote operation [14], the modification 127 requires extra hardware and software to overcome the absence 128 of the human attendance [23]. The dome is represented the 129 main obstacle toward remote operation as a reason for its 130 manual operation in rotation and shutters handling [24]. 131 In addition, the telescope itself always requires a set up 132 before any observation using its hand control (HC), and 133 the focus which has a great impact in image quality [25] 134 is adjusted under watch without any physical limitations. 135 Beginning with the hardware, it should be a robust control 136 system, including a suitable controller, various sensors, and a 137 to overcome these restrictions is to write a complete code that 158 manages the telescope and instruments. 159 Following modification of the telescope efficiency dur-160 ing pointing, the major challenge is selecting an efficient 161 controller and extracting optimal gains using advanced opti-  that requires a telescope with high-performance [6]. More-191 over, these flaws prevent the anticipated advantage of remote 192 telescope from being realized, so it has been decided to use 193 a sophisticated controller to achieve the most satisfactory 194 outcomes. The authors intend to modify the CGE controller 195 in the future; therefore, the theoretical study is necessary to 196 decide which controller will provide the best response. 198 This practical study includes creating a complete control sys-199 tem for the telescope, the dome, the camera, the focus system, 200 and associated devices. It covers all aspects of observation 201 with self-protection against natural hazards, faults, and inter-202 ruptions. Furthermore, no commercial kits or software have 203 been used, and as a result, the cost was overly small compared 204 to commercial systems and previous research. Moreover, 205 remote access is designed based on the Transmission Con-206 trol Protocol and the Internet Protocol (TCP-IP) client-server 207 socket, which perfectly offers numerous utilities that enhance 208 the performance and security of accessing and transferring 209 data between the server and clients [27].

210
The another contribution in this paper is involved in apply-211 ing the fractional TID and ID-T controllers. FO controllers 212 such as TID and ID-T have proliferated due to their flexi-213 bility in designing and stability, particularly with dynamical 214 systems [28]. They include extra tuning parameters compared 215 with conventional PID controller, which reduces processing 216 in optimum parameters demonstration [29]. FO controllers 217 are distinguished with more expanded freedom in tuning for 218 system stabilizing, so researchers have a great interest in 219 such controllers [30]. Moreover, the TID family is poten-220 tially one of the best FO controllers as it has a robustness 221 in estimating the parameters of a system with a closed-loop 222 enhancing the system response [31]. As well as, gains tuning 223 is performed using different optimization techniques such as 224 POA and PSO. The POA technique replicates the peafowls 225 swarm behavior in their foraging, chasing, and courtship, 226 where a triple group of peafowls (i.e., peafowl cubs, pea-227 cocks, and peahens) emulate the behaviors of dynamic swarm 228 and hierarchy in food searching. It is considered an efficient 229 exploratory searching operator and adaptive searching as 230 well [32]. Simulation results and comparison diagrams imply 231 that the response characteristics of the pointing position have 232 been enhanced by minimizing the times of overshoot, rising, 233 and settling.

234
The outlines starts with the introduction and followed by 235 the construction and operation of the 14-inch observatory. 236 After that, the automation process is revealed separately for 237 each device and remote operation preparation. The section 238 after that summarizes the overall control system, while the 239 graphical user interface section comes after. Subsequently, 240 the CGE telescope model and proposed controller are pre-241 sented. The next section is the results with discussion and 242 finally, a conclusion is provided.

244
This section provides a detailed description of 245 the 14-inch observatory before any modification when 246 VOLUME 10, 2022

278
The astronomical dome is the enclosure of the telescope 279 and attached devices inside the observatory; it is necessary 280 to protect them from the horrible weather conditions [36]. 281 The 14-inch telescope dome is a hemisphere steel Ash-dome 282 shown in Fig. 2 with approximately 5 meters in diameter 283 and located upon a concrete wall. Dome rotation is driven 284 via a three-phase, 1370 rpm, and 0.37 KW induction motor. 285 This motor is fixed on the concrete wall and driven through 286 a variable frequency drive to control the rotation speed. The 287 mechanical motion is transferred from the motor to the dome 288 through a circular iron gear fixed with the dome. The dome 289 often includes an upper and a bottom shutters to form a 290 dome slit providing a viewing aperture for the telescope 291 during observation, even during windy conditions. The upper 292 shutter is of the ''up and over'' and driven with a 200-watt 293 DC motor. Meanwhile, the bottom takes some cranking to 294 laterally raise and lower the shutter through a 600-watt single-295 phase induction motor. The observer uses a selector switch to 296 rotate the dome left and right and push buttons to control the 297 shutters [37]. The CCD camera is widely used in astronomy; it is an 300 integrated circuit consisting of a certain number of coupled 301 capacitors [18]. The utilized CCD camera is SBIG version 302 STX-16803 with 16 megapixels and a resolution of 4096 x 303 4096 pixels. It has two chips for imaging and guiding to 304 enable imaging and self-guiding simultaneously. Cooling 305 techniques are based on intelligent fans and connect with a 306 water cooling technique [38]. There is an SBIG FW7-STX 307 filter wheel equipped with the camera via the inter-integrated 308 circuit I 2 C protocol. It comprises 7-position, 50-millimeter 309 square filters with an optical encoder for position centering.
Digital interface is available within two different protocols, ries to be operated remotely without direct human control [7].

333
The remote control system which has to be designed to safely 334 control the telescope, the dome, the CCD camera, and auxil-  should be aligned for accurate pointing and tracking [39]; 344 otherwise, the coordinate values will not be meaningful [40].  step is achieved by pointing the telescope to two bright 351 reference stars; however, the observed stars will be shifted 352 from the center of the image frame. Therefore, the user has to 353 center them manually and press the Align button on the HC.

354
By using these steps, the telescope will calculate the pointing 355 error and store it [41]. In order to overcome this problem, 356 the key lies in the data transferred between the telescope 357 mount and the HC; extracting these data packets will be the 358 solution. In order to obtain these commands that responsible   a custom code between them for spying on serial packets for 364 commands and responses. Table 2 reveals the home position 365 acquired packets for commands and responses among HC and 366 the mount. After analyzing them, it has appeared that packets 367 1 and 2 are sent to move the telescope in two directions; 368 RA and DEC, then packets 3 and 4 are continually sent with 369 a ''00'' response until the exchange to ''ff.'' This response 370 indicates that the telescope is at the home position parking. 371 Command format is as following: 0 × 50 <message Length> 372 <destination Id> <message Id> <data1> <data2> <data3> 373 <response Bytes>, where <data13> are the message data 374 bytes and If there is empty data byte it must be represented 375 by zero, while <responseBytes> is the echo back's byte num-376 ber [42]. For instant if the hexadecimal PC command is to get 377 the HC version, the packet will be in this form; (0 × 50 0 × 01 378 0 × 11 0xfe 0 × 00 0 × 00 0 × 00 0 × 02) and the response 379 from the HC will be (0 × 04 0 × 03 0 × 23) which is 04.03 in 380 decimal.

381
On the other hand, Table 3 displays acquired packets for 382 the alignment process. These packets set the telescope coordi-383 nates to the last alignment correction values. The CGE written 384 software relies on a communication protocol regarding the 385 HC that instructs how to send commands to the telescope 386 and get a response; it does not include the setting up com-387 mands. Talking to the mount is based on RS-232 serial com-388 mands in row hexadecimal packets covering all HC operating 389 aspects [40] and has been executed by connecting a serial 390 cable between the HC and the computer. Furthermore, the 391 user could send raw commands to the telescope through a 392 custom graphical window. Eventually, the mentioned packets 393 will enable the observer to point the telescope efficiently 394 and read its position in a coordinate system based on motors 395 positions named setting circles.

396
The CGE telescope mount receives and sends coordinates 397 in two different coordinate systems; one of them is through 398 VOLUME 10, 2022 coordinates [43]. Julian day (JD), and the telescope longitude (L) [43].

433
The telescope is permanently located at its dome at the 434 Kottamia observatory, meaning that the alignment process 435 is performed once by an expert [7]. Meanwhile, during ished. The modified correction procedure will convert these 446 axis position coordinates into the horizontal coordinates.

447
By applying a set of transformation equations, the alignment 448 correction is achieved, and the accuracy of pointing will be 449 enhanced sharply [44]. The Mentioned stored coordinates are 450 performed in Eqs. 9 and 10 to calculate parameters l2, m2, 451 n3, L1, M 1, and N 1, after that the same two equations will 452 be applied for the second star to calculate l2, m2, n2, L2, M 2, 453 and N 2. By making use of all calculated values, l3, m3, n3, 454 L3, M 3, and N 3 will be the output from Eqs. 11 -14 then T 455 is estimated as in Eq. 15.
Now that, suppose the the remote observer will point the 471 telescope to a sky object with equatorial coordinates of δ and 472 α at time t, the parameters L, M , and N will be obtained 473 from Eq. 16, then Eq. 17 will give the output of l, m, and n. 474 Finally, the horizontal coordinates of the acquired sky object 475 are calculated from Eqs. 18 and 19.
Eventually, the authors integrated the data of HC commu-481 nication protocol with the extracted packets using reverse 482 engineering to create a custom computer program for the tele-483 scope. This interface access all aspects of mount functions, 484 including home position setting, last alignment process, and 485 remote pointing correction.  [46].

497
The proposed integrated control system has been designed and sensors [47]. Moreover, it is efficient in practical imple-503 mentation, particularly in robotic projects [48]. This modified feeding exhibited the most challenging issue. The electric 524 source is fixed at the concrete wall that carries the dome; 525 meanwhile, the shutters are rotated with the dome. So that the 526 observer has to connect its cables manually to the nearest out-527 let in the wall to power the dome slit; after that, he opens the 528 shutters completely, then plugs the cables off to start observa-529 tion and repeats this sequence after finishing the observation. 530 As a result, the dome automation starts with designing an 531 electric mechanism that feeds the shutters automatically.

532
Standard solutions for supplying rotated objects with elec-533 tric power from a fixed electric source are represented in two 534 techniques. The first is a complete solar cell system with a 535 battery mounted above the dome from outside with wireless 536 communication for signals [49]. Another solution is a set of 537 circular electric conductor bars named slip rings mechanism 538 with carbon brushes that transfer power to the shutters where 539 signals are transferred through wires [50]. These two tradi-540 tional solutions are not affordable, need excessive mainte-541 nance, and are risky. So that our modified technique takes 542 part of the second solution based on the electric conductor as 543 indicated in Fig. 4; however, it is inexpensive and prevents 544 any risk probability. It comprises two platforms; one is fixed 545 and has the electric source, while the other is rotated with the 546 dome body. Fig. 4(a) depicts the power platform construc-547 tion where the fixed one includes a spring at its base point 548 for smooth contact and flexibility. It also contains a limit 549 switch to determine the dome's home position. Meanwhile, 550 the rotated one is routed directly to the shutters unit and 551 isolated perfectly.

552
The technique concept is shown in Fig. 4(b); at the dome's 553 home position determined by a limit switch, the two platforms 554 are aligned together and contacted. In case of the need to 555 handle the shutters manually, the observer can use one bush 556 button, which will rotate the dome to its home position to 557 transfer power to the shutters. After the open process is 558 completed, the power is switched off automatically, and this 559 procedure is repeated when the observation is finished. After 560 that and based on the Arduino controller, a control system has 561 been implemented for the remote operation of dome rotation 562 and shutters units. Moreover, there is a wireless HC for 563 in-site observation and attended maintenance. A sample from 564 control units designed and implemented for the observatory 565 appears in Fig. 5 relative to the shutters. Fig. 5(a) shows the 566 VOLUME 10, 2022  [52]. Auspiciously, our telescope mounts support base is 583 established at the center of the dome; subsequently, the ϕ is 584 equal to the telescope azimuth (φ), referring to Eq. 21 [53].

601
STX-16803-SBIG CCD camera is used in our developed 602 observing platform; it has a USB and Ethernet connection. 603 Besides, the SBIG FW7-STX filter wheel is equipped with 604 the camera via the I 2 C protocol. The hypertext transfer pro-605 tocol (HTTP) allows the STX camera to be accessed across 606 various PC platforms without any special driver requirements 607 using a commands in [54]. Based on this document, the 608 authors have created a GUI interface that provides all avail-609 able aspects controlling the STX. The observer can take a 610 scientific or guiding image and get information about the 611 camera and filter wheel. The acquired images are compressed 612 and sent to the remote observer. In addition, he can change 613 the image header parameters, such as the observer and object 614 names.  Fig. 6 depicts a schematic diagram for the overall system 618 of the 14-inch observatory. The remote observer commu-619 nicates the dome server via a peer-to-peer (P2P) socket-620 based TCP-IP protocol. The P2P is a network communication 621 architecture currently considered an efficient technique for 622 allowing data transmission between custom software [55]. 623 A router with a subscriber identification module (SIM) pro-624 vides the observatory with the internet. This SIM includes a 625 real IP address to enable access from all over the world. The 626 dome server is a personal computer (PC) installed inside the 627 dome and connected directly with all the observatory devices 628 through three communication protocols, RS-232 protocol for 629 the telescope and USB for both dome units master and focus 630 system. Meanwhile, the SBIG and monitoring cameras are 631 connected via the ETHERNET protocol. The monitoring 632 cameras consist of two movable web cameras with infrared 633 (IR) to quickly check the dome and devices inside the obser-634 vatory as it is necessary for remote observation [52]. 635 It is evident that occupying all PC ports reduces the effi-636 ciency of operation; thus, the master unit handles all dome 637 The interaction design of GUI in this research relies heav-677 ily on written computer programs to avoid any restrictions 678 related to the individual software of devices. Designing 679 and implementing a custom GUI provides access methods 680 for modification, keyboard shortcuts definition, and design 681 interaction window for user [2]. The GUI application of 682 the 14-inch CGE observatory is a single window covering 683 all requirements for remote astronomical observing. It also 684 exhibits visibility for the whole system with a fast display 685 for the status of attached devices. Furthermore, it has an out-686 standing efficiency in making appropriate decisions in case of 687 risk, such as closing the shutters automatically during lousy 688 weather. Moreover, each motion in the observatory has a 689 physical limit and time restriction, and if a motion exceeds its 690 normal time, the system will detect a fault. This time limit is 691 essential to protect motors and mechanical parts from damage 692 in case of the failure of the limit switches. The GUI includes 693 several modes of operation, such as remote and onsite, beside 694 an engineering mode in order to overcome technical faults 695 remotely. In the engineer mode, the remote technician could 696 set and check critical parameters in case of technical tests, 697 debugging and fault diagnosis. The remote technician has a 698 significant facility for violating system limits and directly 699 commanding the ports.

700
The GUI structure is eleven tabs, as shown in Fig. 7, 701 which displays a screenshot from the GUI; each tab displays 702 current information about the instrument besides a facility for 703 controlling it. Primarily, this figure focuses on two essential 704 tabs, a telescope pointing window and a soft buttons for the 705 dome. The telescope tab depicts the local and the world time; 706 moreover, it includes a catalog for some sky objects where the 707 observer has a facility to select from them or enter another 708 object's coordinates. After selecting, all transformation and 709 information of the object will be displayed before point-710 ing. The soft buttons tab shows necessary information about 711 the dome position and direction, shutters status and current 712 weather station information. Furthermore, It displays a soft 713 buttons with a full capability to control the dome rotation, 714 shutters, switching devices and lights besides an emergency 715 stop button.

719
The dynamical model of the telescope is represented in some 720 differential equations with a nonlinear coupled type that 721 includes a varying inertia term (M ), Coriolis and centrifu-722 gal term (C), and gravity term (G). The (G) term tends to 723 VOLUME 10, 2022 .
The 14-inch telescope model is a coupled system; therefore, The linearized model comprises the nonlinear telescope 751 model and the compensator is depicted in Fig. 8, where 752 U 1 and U 2 are the controller outputs that will feed the com-753 pensator; meanwhile, the output of the compensator reveals 754 to the input joint torques. The state variables will be returned 755 to the compensator then the output positions will be the 756 actual feedback signals to the controllers for RA and DEC, 757 respectively.

759
FO concept is based on the expansion of integer order inte-760 gration and differentiation to a fractional order operator. 761 FO calculus-based controllers such as TID are utilized fre-762 quently due to their improved capacity to reject disturbances 763 and higher sensitivity to changes in model parameters [60]. 764 TID controller is similar to the PID structure; however, the 765 proportional component is replaced by a tilted component 766 1/s 1/n to produce a transfer function described in Eq. 25 [61]. 767 Block diagram of the TID controller is shown in Fig. 9 where 773 the coefficients take a direction of feed-forward to ensure ease 774 tuning as well as optimal response approach beside better 775 sensitivity to parametric variations of the system [62]. The 776 input parameter r(t) refers to a reference set value, while 777    Table 5; however, all mathematical details of PSO are  most significant aspects of this type of swarm, in which the 800 POA algorithm is built.

801
Establishing a mathematical model efficiently for a 802 peafowl swarm requires dividing it into three groups, the 803 first is the adult peacock, the second is adult peahens, and 804 the third group is peafowl cubs. In real-world optimization 805 issues, every peafowl is rated according to its fitness value, 806 and initially, there are five solutions, starting with peacock 807 one and ending with peacock five, respectively. Furthermore, 808 the first remainder, 30% of individuals, is classified as adult 809 peahens, while the rest are classified as peafowl cubs. Imme-810 diately after finding a food source by peacocks, they will 811 rotate around it to show off and attract peahens as a way to 812 mate. The show is represented in tail spreading with dancing 813 behavior, flapping feathers producing attractive sounds, and 814 rotating around peahens that have been approached.

815
Converting peacock behaviors into an appropriate math-816 ematical model is primarily based on their rotation aspects, 817 leading to their fitness estimation during iterations. Peacocks 818 with a higher fitness mark are more likely to circle the food 819 source with a smaller circle radius, whereas those with a 820 lower fitness value are more likely to spin in site with a 821 broader circle radius. During the whole seeking process, 822 peahens use an approaching mechanism and an adaptive 823 searching to dynamically change their behaviors at various 824 phases. The peacock's fitness rating indicates the likelihood 825 of attracting a peahen. Peafowl cubs serve as seeking agents, 826 randomly exploring for the best-quality food source in the 827 searching region in a Levy flight distribution, in addition to 828 approaching male peacocks with good food sources with a 829 high fitness value, one of the assumed five peacocks. As a 830 result, the four peacocks will go toward the selected peacock 831 in a random path within 90 degrees of the line between 832 peacock one and the other, as shown in Fig. 11.    The mathematical model that describes behaviors of peahens 878 and the approaching mechanism is represented in  Interaction behaviors among peacocks start as peacock one 901 has the best food source. The four remaining peacocks will be 902 motivated to reach it in a random direction within 90 degrees 903 of the line between peacock one and the others as shown in 904 Fig. 11 with a mathematical relation as      Table 6 depicts. 2: Set initial rotation radius (R S0 ) from Eq. 34 using LB and UB in Table 7. 3: Initialize position of peafowl population. 4: Assign roles based on fitness function in Eq. 45. 5: FOR i := 1 to i max , 6: update µ, γ , and σ via Eqs. 36, 39, and 40 respectively. 7: Evaluate X Pc1 , X Pc2 , X Pc3 , X Pc4 , and X Pc5 from Eqs. 27 -32. 8: FOR k := 1 to 5, 9: Inspect and correct X Pci to be within a correct range. 10: IF X Pci fitness is better than X Pci (t). 11: X Pci (t + 1) = X Pci . 12: END IF. 13: END FOR. 14: FOR j := 1 to number of peahens, 15: Calculate X Ph using Eq. 35. 16: Inspect and correct X Ph to be within a correct range. 17: IF X Ph fitness is better than X Ph (t) fitness, 18: X Ph (t + 1) = X Ph . 19: END IF. 20: END FOR. 21: FOR j := 1 to number of peafowl cub, 22: Estimate X SPc using Eq. 37. 23: Calculate X PcC using Eq. 38. 24: Inspect and correct X PcC to be within a correct range. 25: IF X PcC fitness is better than X PcC (t) fitness, 26: X PcC (t + 1) = X PcC . 27: END IF. 28: END FOR. 29: Evaluate X Pc2 , X Pc3 , X Pc4 , and X Pc5 from Eqs. 41 -44 30: FOR i := 2 to 5, 31: inspect and correct X Pci to be within a correct range. 32: IF X Pci fitness is better than X Pci (t + 1) fitness, 33: X Pci (t + 1) = X Pci . 34: END IF. 35: END FOR. 36: Based on fitness reassign roles. 37: END FOR. and DEC with achieving the best performance characteristics. 929 Therefore, and because of its benefits in reducing settling 930 time, the integral time absolute error (ITAE) is used as a 931 performance criterion; meanwhile, the integral time square 932 error (ITSE) based control structure generates a large output 933 signal in response to an unexpected change in the reference 934 signal [66]. Furthermore, the ITAE produces less overshoot 935 and oscillations than the integral-of-the-absolute error (IAE) 936 and integral-of-the-square error (ISE). The applied ITAE 937 objective function in Eq. 46 is to achieve the optimal con-938 troller parameters for optimizing the fundamental telescope 939 characteristics like maximum O sh , R s and T s [67].   Fig. 13 represented in the telescope, which is considered the 957 core of observation attached by the SBIG camera, the filter 958 wheel, and the focus motor. The figure also reveals the dome 959 rotation unit; it comprises a controller cabinet provided by 960 bush buttons and an emergency stop for in-site operation. The 961 controller box, which is responsible for the dome rotation and 962 transfers power to the shutters unit, is fixed on the left of the 963 control cabinet. Moreover, the box on the right side comprises 964 a three-phase inverter drive to adjust the rotation speed. All 965 mentioned units are connected together in a wire form; mean-966 while, they communicate with the server PC wireless.

967
A sky's brightest star named Pollux in Fig. 14 has been 968 observed remotely using the modified CGE control system; 969 however, it has a shift, as shown in Fig. 14(a). By applying 970 the procedure of remote pointing correction, the star has been 971 approximately centered in the middle of the frame, as repre-972 sented in Fig. 14(a).

973
Adjusting the focus of sky objects is essential to pro-974 vide an adequate resolution; therefore, Fig. 15 confirms 975 how the observer can focus celestial images remotely. 976 Fig. 15(a) shows an observed frame with unfocused stars 977 that appear stretched. Meanwhile, Fig. 15(b) reveals better 978 resolution-focused stars with less size and higher quality.

979
Concerning the issue of dome automation cost, including 980 control unit, software, and weather monitoring, the study 981    Tables 8 and 9 show the 999 coefficients optimized values of the TID and ID-T controllers, 1000 which have been obtained based on POA, PSO, and classical 1001 tuning. These values have resulted in the best performance of 1002 the telescope system by minimizing the times of settling and 1003 rising beside removing oscillations, as Table 10 reveals. The 1004 proposed ID-T controller optimized by POA has produced 1005 values as 0.1099 and 0.2050 seconds for R s and T s in the RA 1006 direction. The ID-T tuned by PSO has followed it with a little 1007 difference in the longer time for rising and settling. Although 1008 the TID output has small values as 0.0614 and 0.1796 seconds 1009 for R s and T s in the same direction, it is inconvenient due to 1010 the produced O sh as of 2.7568 and 5.4947 seconds in POA 1011 and PSO, respectively. Meanwhile, classical tuning for both 1012 controllers has produced poor performance, particularly in R s 1013 which reached 2.0512 and 1.3137 seconds in TID and ID-T. 1014 Fig. 17 depicts the performance investigation of the tele-1015 scope position and velocity responses in the RA direction 1016 with a reference angle position is 50 degrees. It is observed 1017 from Fig. 17(a) that the characteristics achieved by the ID-T 1018 controller are better than the TID. On the other hand, the TID 1019 controller performance has produced an O sh by executing 1020 both algorithms POA and PSO; however, the O sh is less 1021 in the case of POA. Regarding classical tuning for the two 1022 controllers, although the O sh is zero, the consumed time to 1023 track the desired position with stability is long.

1024
Simulation results concerning the variations of the velocity 1025 in the arms of RA are revealed in Fig. 17(b). The figure illus-1026 trates that the velocity generally tends zero immediately after 1027 reaching the desired position. The consumed time in the case 1028 of the TID-based PSO and POA is short; however, it generates 1029 some oscillations due to the O sh produced before stability. 1030  In contrast, the ID-T optimized via POA has tended to zero 1031 fast to achieve a stable performance without any oscillations,    which produces zero O sh persistently, as shown in Fig. 19. 1043 In conclusion, the proposed ID-T controller-based POA has 1044 exhibited an appropriate enhancement for the 14-inch CGE 1045 telescope.

1051
This study furnishes a practical and low-cost complete control 1052 system for real-time remote astronomical observation; it is 1053 applicable for small and medium-size telescopes and has been scope stability significantly in terms of overshoot, rising, and 1093 settling times. This hybrid technique represented in ID-T 1094 optimized POA has achieved stability more than twice the 1095 best of previous studies, particularly in FLC-PSO. Simulation 1096 results refer to the poorness of the TID controller in dimin-1097 ishing the overshoot and prove how the POA produces more 1098 accurate results than PSO in our case study.

1099
For future work, it is intended to enhance the telescope 1100 pointing efficiency through advanced image processing tech-1101 niques involving artificial intelligence keeping the observed 1102 sky object inside the frame center. It is based on a continuous 1103 comparison between the observed and standard images in the 1104 catalogs for the same target. The decision toward the best 1105 pointing is based on the distribution of the stars surrounding 1106 the observed target. In addition, simulation results and output 1107 performance of the applied controller are pretty encouraging 1108 to be applied experimentally in the future to adapt the remote 1109 operation to introduce an intelligent product for observing 1110 remotely. 1111 compact remote telescope system,'' in Proc. Int. Conf. Complex, Intell.