基于SICPSO算法的PI^λD^μ控制器参数整定

Tuning of Fractional PI^λD^μ Controller Based on the SICPSO Algorithm

分数阶PID控制器具有5个可调参数KP、KI、KD、λ、μ,能够取得更好的控制效果。目前针对PIλDμ控制系统的参数整定与优化方法已成为分数阶控制领域的研究热点,但由于参数整定变得更加复杂,整定结果有待进一步提高。利用改进Tent映射的良好遍历性生成较优初始种群,以此增强PSO算法初期的全局搜索性能;应用参数ω、c2自适应的PSO算法侧重实现算法后期的局部优化功能以及融入混沌序列作为粒子分量更新的随机数,形成混合算法SICPSO,分别对典型整数阶和分数阶被控对象进行仿真和分析。针对传统ITAE误差准则作为目标函数的不足,在误差准则内引入输出值变化率加权项。结果表明,混合算法SICPSO寻优能力强,改进的目标函数有效,能够较好地提高系统的性能指标。

The fractional order PID controller can achieve better control effect with five adjustable parameters KP,KI,KD,λ,μ.Now the tuning of fractional PID controller has become a hot research topic in the field of fractional order control.But as the parameters setting becomes more complicated,the result remains to be further improved.In this paper,in order to enhance the initial global search function of the PSO,the application of ergodic of improved Tent map generates the optimal initial population.And parameters adaptive PSO is used to mainly realize the later local optimization function with chaotic sequence as random numbers of the particle weight updating formulas.The new hybrid intelligent algorithm is named synthetic improved chaos particle swarm optimization.The simulation and analysis is carried on with the SICPSO applied to the integer and fractional order controlled plant.For the shortcomings of the traditional ITAE criterion as the objective function,the weighted items of output value change rate are introduced in the new error criterion.The simulation results illustrate the improvement of dynamic performance and the effectiveness of the new objective function.