摘  要：針對一般的粒子群優化（PSO）學習算法中存在的容易陷入局部最優和搜索精度不高的缺點，對改進型PSO算法進行研究。由于慣性權重系數ω對算法是否會陷入局部最優起到關鍵的作用，因此，通過改變慣性權重ω的選擇，對慣性權重系數采取線性減小的方法，引入改進型的PSO算法。采用改進的PSO算法對PID控制器進行參數優化并把得到的最優參數應用于控制系統中進行仿真。仿真實驗結果表明：改進型PSO算法不會陷入局部最優，能得到全局最優的PID控制器的參數，并使得控制系統的性能指標達到最優，控制系統具有較好的魯棒性。
關鍵詞：粒子群優化算法；控制器；參數優化；性能指標；魯棒性
中圖分類號：TP273；TP214；N945.13；TM930.12      文獻標志碼：A       文章編號：1674-5124（2013）05-0096-03
Research on PID controller based on improved PSO algorithm
ZHANG Yan-hong
（School of Electronic Information & Electric Engineering，Changzhou Institute of Technology，
Changzhou 213002，China）
Abstract: For the shortcomings of easily falling into local optimum and lower search precision in common particle swarm optimization（PSO） algorithm， an improved PSO algorithm was studied. The inertia weight coefficient ω plays a key role on whether PSO algorithm falling into local optimum or not， so it is linearly reduced by the improved PSO algorithm. The parameters of PID controller are also optimized by the improved PSO algorithm and the optimized parameters are used in the control system. The simulation results show that the improved PSO algorithm does not fall into local optimal and the optimal parameters of the PID controller can be obtained， which makes the performance index of the control system optimal with better robustness.
Key words: PSO； controller； parameter optimization； performance index； robustness