SINERGI Vol. No. June 2021: 169-176 http://publikasi. id/index. php/sinergi http://doi. org/10. 22441/sinergi. OPTIMAL DESIGN PSS-PID CONTROL ON SINGLE MACHINE INFINITE BUS USING ANT COLONY OPTIMIZATION Nasrun Kadir1. Muhammad Ruswandi Djalal2* Department of Mechanical Engineering. Mechanical Engineering. Politeknik Negeri Ujung Pandang. Indonesia Department of Mechanical Engineering. Energy Engineering. Politeknik Negeri Ujung Pandang. Indonesia Abstract Optimization of the controller in a generator can improve system The right parameter optimization is needed to get the optimal performance from the controller. The application of the artificial intelligence method as a parameter optimization method is proposed in this study. By using the smart method based on Ant Colony, the optimal PSS-PID parameters are obtained. With optimal tuning, the system gets optimal Single Machine Infinite Bus (SMIB) system frequency and rotor angle response, indicated by the minimum overshot system response. The SMIB system's stability will be tested. A case study of adding and reducing loads is used, with the proposed control method PSS-PID being optimized using Ant Colony. Based on the analysis using the proposed PSS-PID control, we get the minimum overshoot for the frequency response and rotor angle of the SMIB system. When the load changes at 20 seconds, using the PSS-PID control scheme, the minimum overshoot is -4. 316e-06 to 7. 598e-05 pu with a settling time of For the rotor angle overshoot response, using the PSS-PID control scheme, the minimum overshoot is -0. 01113 to -0. pu with a settling time of 22. Keywords: Ant Colony. Overshoot. PID. PSS. Settling Time. SMIB. Article History: Received: June 22, 2020 Revised: September 5, 2020 Accepted: October 16, 2020 Published: February 10, 2021 Corresponding Author: Muhammad Ruswandi Djalal Department of Mechanical Engineering. Energy Engineering. Politeknik Negeri Ujung Pandang. Indonesia Email: wandi@poliupg. This is an open access article under the CC BY-NC license INTRODUCTION In the dynamic stability analysis, because the governor response is slow, the torque is With the slow governor response compared to the excitation system response, the excitation system regulates the controller. In its application, the addition of an excitation amplifier circuit is not optimal to stabilize the system, especially Low Frequency Oscillation 0. 2 - 2. 0 Hz . Low frequencies can turn into oscillations between areas, so additional controllers such as Power System Stabilizer and PID are needed. PSS is additional control equipment on generator excitation that is used to provide additional attenuation to generator excitation . PID is an additional control that functions to reduce local or global oscillations in the generator and predetermined variable values . It requires precisely adjusting the PSS and PID to reduce oscillation and properly stabilize the system optimally to get maximum results. Controller parameters are optimized using smart methods . The Ant Colony Optimization (ACO) algorithm is a Swarm Intelligence group that can solve optimization problems. ACO adopts group or insect swarming behavior . In several previous studies, the ACO method was also used for the optimization of PID controllers . Several other studies related to the application of performance, such as firefly method . , bat algorithm . , 13, . , flower pollination . , imperialist competitive . , and cuckoo search . Optimal parameter tuning of the controller will be very influential in stabilizing the system. However, the range of equipment parameters is Kadir & M. Djalal. Optimal Design PSS-PID Control on Single Machine Infinite Bus A SINERGI Vol. No. June 2021: 169-176 very diverse and broad, so to get the parameter values quickly, an optimization method using ACO is used. The response value is known by analyzing the overshoot and settling time values, while the objective function minimizes Integral Time Absolute Error (ITAE) . The case study used in this research is the Single Machine Infinite Bus (SMIB) system. SMIB is an electric power subsystem consisting of one or more generators that are connected to an infinite bus . Then analyze the simulation results by comparing the simulation system results without control. SMIB with PID. SMIB with PSS, and the proposed method of SMIB with PSS-PID with ACO. In this study, the authors implemented an Ant Colony-based intelligent method to solve the optimization problem of determining the PSS-PID parameters on the SMIB system. METHOD There are several steps in conducting The research begins with the study of literature, analysis, and conclusions. Data Analysis The case study used is a single machine infinite bus (SMIB) system regarding previous research modeling . Some stages of analysis are, modelling the synchronous machine system, modelling the PSS-PID using Simulink, entering the system data, designing the ACO, plotting the proposed method results, and comparing it with the other control schemes. Machine Model The synchronous machine model is modeled in linear modelling, as shown in Figure 1 . Eq' AE AEO = Time . = Rotor mechanical angle . ad-mechanica. = Engine damping coefficient = Stator field flux = Change in engine angle = Momentum = Speed Exciter Model Excitation equipment in the system is used to control the generator output variable . This study's excitation modeling refers to the Ie modeling shown in Figure 2 . Figure 2. Excitation Model Where: KA = Amplifier gain constant KF = Filter gain constant KE = Exciter gain constant TA = Amplifier response time TE = Exciter response time TF = Filter response time AEU2 = Change of an engine control signal Governor Model Governor is a controller used to regulate the mechanical torque value of Tm, which is input from the generator . Governor modeling is shown in Figure 3. Figure 3. Governor Model Figure 1. Synchronous Linear Machine Model Where: K1-K6 = Constant EFD = Magnetic field flux = Turbine mechanical torque (N. = Rotor electric torque (N. = Acceleration of torque (N. Where: = Constant gain = 1/RG = Constant governor time = Governor's groove constant iGSC = Governor Speed Changer Turbine Model The turbine model used is a steam turbine power plant model . The model is shown in Figure 4. Kadir & M. Djalal. Optimal Design PSS-PID Control on Single Machine Infinite Bus A p-ISSN: 1410-2331 e-ISSN: 2460-1217 Where: AEY = Change in valve height Tw = Steam turbine response time Tga = Steam turbine regulator response time Kga = Strengthening steam turbine regulator = Steam turbine regulator constant AEU1 = Feedback control signal changes Figure 4. Turbine Model Single Machine Infinite Bus (SMIB) Model is shown as the Single Machine Infinite Bus (SMIB) model in Figure 5 . Figure 5. SMIB Model Power System Stabilizer Power System Stabilizer is a device used to optimize system stability. in this study, the modeling is shown in Figure 6. AEAi K PSS Gain Tw s 1 Tw s Blok Washout 1 TA s 1 TB s 1 TC s 1 TD s Blok lead-lag VS max VS min Saturation Figure 6. PSS Modeling Where: Kpss = Gain T1-T4 = Fase Lag/Lead = Washout = Limiter Ant Colony Optimization (ACO) Ant Colony Optimization (ACO) is an algorithm that uses artificial ants to find the optimal value of an optimization problem. ACO was discovered by Marco Dorigo in 1990 . ACO adopts the ant behavior in finding the shortest path between the nest and food. Ants provide information to each other using pheromones on the path that is traversed. Pathways that contain large pheromones attract Every time they go on a tour, the ants leave pheromones on every path they take. The pheromone levels in the ants' paths will increase, while the pheromone levels in the trails where ants don't pass will experience evaporation . The shortest path will often be traversed by ants so that the pheromone levels in that pathway increase and attract other ants to pass the path . The ants tour according to a probabilistic function that takes two things into account. First, it calculates the visited nodes whose accumulation is calculated as the distance from the tour. Second, the Ants detect the Kadir & M. Djalal. Optimal Design PSS-PID Control on Single Machine Infinite Bus A SINERGI Vol. No. June 2021: 169-176 pheromones that other Ants leave behind. The ant then selects city j on the candidate list following the transition rule in . P is the Ant's movement from i to j. E argmax uEaJlk {[Eiu . ] } ,if q Cq0 j = EJ , the other E Table 1. Ant Colony Parameter Parameters Number of Ants . Max Iteration (I. Pheromone () Beta () Values PiJk = [Eiu . u ( t ) ] Eu [Eil . l ( t ) ] . lEaJ lk Where: E = Pheromones = the distance between two nodes q = A random variable . q0 = Parameters that can be set at intervals . J = A list of candidates Ant Colony Optimization Parameter Some parameters used in the ACO method in this thesis are as follows, the number of ants = 10, the maximum iteration = 50, and the Pheromone Resistance . = 0. Tuning PSS-PID Control with Ant Colony The objective function used in this study is to optimize the Integral Time Absolute Error. ITAE = E t AEA . ) dt After entering several parameters in Table 1, the Ant Colony algorithm is run to optimize the PSS-PID value. The exact value will greatly affect the SMIB response performance designed in this The Ant Colony algorithm requires a calculation process to get the optimal value. Figure 7 shows a graph of PID convergence optimization using the Ant Colony algorithm. Convergence is the value of the fitness function, which describes the optimal criteria for an optimization problem. Figure 7 displays the convergence graph of PSS-PID value optimization using Ant Colony. The Ant Colony algorithm does not take a long time to perform the optimization process. In the 10th iteration, the algorithm has found the optimal PSS and PID values with a minimum fitness value of 2. The ACO results obtained a fitness function value of 2. 724e-07, with iteration 50 times convergent in the 10th iteration, the nbest value is the best Ant Colony known as the PID parameter optimization results, namely Kp. Ki. Kd. Kpss. T1. T2. T3. T4. Table 2 shows the limits and values of the PSS-PID parameter optimization results set by the Ant Colony. Where A is a speed that is tested for stability. The PSS-PID parameters tuned by ACO are Kp. Ki. Kd. Kpss. T1. T2. T3. T4. The algorithm design is mentioned in Table 1, to run the Ant Colony optimization. The Ant Colony algorithm was created using the MATLAB-Simulink software. Table 2. Limits and ACO Optimization Results Parameter Kpss Limits Lower Upper Results As a comparison, several control schemes are used, namely the combination of PID. PSS and PSS-PID. In principle, the Ant Colony algorithm looks for food sources based on pheromones, which then in groups will follow the trail with the largest pheromone . With this principle, the algorithm will look for the most optimal parameters to be filled in the PID parameters, so that optimal control is obtained at the SMIB. The flow chart in this study is shown in Figure 8. Figure 7. Convergence Graph Kadir & M. Djalal. Optimal Design PSS-PID Control on Single Machine Infinite Bus A p-ISSN: 1410-2331 e-ISSN: 2460-1217 Start Literature study and data collection of the SMIB system Defines the parameters to be optimized and the objective function to be used condition. Pe> Pm, the electric torque and mechanical torque are not balanced. This condition causes the electric frequency . to change as well. During this instability, the rotor rotational speed . O) becomes out of sync. this condition, the frequency response graph drops before returning to a steady state. The control system function is then required to return to steady state. The characteristics of the overshoot response in this condition are shown in Table 3. Table 3. Frequency Deviation Modeling System Deviation Output variable plotting SMIB SMIB-PID SMIB-PSS SMIB-PSS-PID Performance Check Yes System Analysis Conclusion End Figure 8. Research Flowchart RESULTS AND DISCUSSION The system analysis is performed, namely the analysis of the system frequency and SMIB rotor angle. The analysis was carried out using several control methods, such as the uncontrolled system. SMIB-PID. SMIB-PSS, and PSS-PID SMIB. PSS and PID parameters are simulated using the Ant Colony Optimization The SMIB system is given a disturbance in the form of load changes to test the system's stability. SMIB Frequency Response The first analysis begins by reviewing the frequency stability response of the SMIB system. The simulation results are shown in Figure 9. Figure 9 display the simulation results of the SMIB frequency response with several control From the SMIB system's simulation results, a load change disturbance of 0. 01 pu is given in the second, then the load shedding that occurs in the 20th second is 0. 005 pu. At the first load change, the load increase increases the condition where the electric power is not the same as the mechanical power (P. In this Overshoot . 0002409 & 0. 0002069 & 0. 0001748 & 1. 0001533 & 7. Table 3 shows the overshoot system's characteristics when the load changes at the 1st second in the form of additional loads. The uncontrolled SMIB system has an overshoot of 0. 0002409 to 0. 0001873 pu with a turnaround time of 15. With PID control, the SMIB system gets an overshoot of -0. 0002069 to 000131 pu with a settling time of 11. With PSS control, the SMIB system obtained an overshoot of -0. 0001748 to 8,743e-05 pu with a settling time of 2,334s. With the proposed method using PSS-PID, the smallest overshoot is obtained from -0. 0001533 to 7. 662e-05 pu with a settling time of 3. Then in the next load change in the form of a load reduction which causes the electric power (P. to change. In this condition, the electric power is not the same as the mechanical power (P. Pe