The effects of weightage values with two objective functions in iPSO for electro-hydraulic actuator system

• Main Authors: Chai, Mau Shern, Ghazali, Rozaimi, Chong, Shin Horng, Chong, Chee Soon, Ghani, Muhamad Fadli, Md. Sam, Yahaya, Has, Zulfatman
• Format: Article
• Language: English
• Published: Penerbit Akademia Baru 2021
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://eprints.utm.my/id/eprint/97683/1/YahayaMdSam2021_TheEffectsofWeightageValueswithTwoObjectiveFunctions.pdf; http://eprints.utm.my/id/eprint/97683/; http://dx.doi.org/10.37934/arfmts.81.2.98109
• Institution: Universiti Teknologi Malaysia

In this paper, the Proportional-Integral-Derivative (PID) controller with improved Particle Swarm Optimization (iPSO) algorithm is proposed for the positioning control of nonlinear Electro-Hydraulic Actuator (EHA) system. PID controller is chosen to control the EHA system due to its popularity in industrial applications. The PID controller parameters will be tuned by using the iPSO algorithm to get the lowest overshoot percentage and steady-state error. The conventional PSO algorithm has only one objective function to get the optimum parameters. However, this is not enough to increase the control performance of the EHA system. Therefore, an improved Particle Swarm Optimization (iPSO) that includes the mean error and overshoot percentage as the two objective functions is proposed in this paper. The most popular method in PSO that included two objective functions is Linear Weight Summation (LWS). In this method, the two objective functions are combined with certain weightage into one equation to give the best control performance. This paper focuses on determining the suitable weightage between these two objective functions so that the EHA system can produce the best control performance with less overshoot and less error. Overshoot percentage and steady-state error are used to indicate the best control performance. The results showed that EHA system can perform better by using suitable weightage between the mean error and overshoot percentage.