论文部分内容阅读
核电厂稳压器的压力控制系统具有非线性、时变性等特点。在分数阶理论的基础上提出分数阶PIλDμ控制算法(FOPID),实现了比例-积分-微分控制(PID控制)的由点到面的控制,同时引入短反馈算法,使控制更加精确。相较于传统PID控制多了2个参数的自由度,并采用一种间接算法(Oustaloup算法)实现其控制器的设计;并提出了一种生物启发式算法——鸡群算法(CSO),解决了分数阶PI~λD~μ参数的整定问题。为了避免局部最优和早熟现象,又提出了一种改进的CSO算法(ICSO),算法中更改了小鸡的位置更新公式,并引入混沌初始化和随机权重,提高算法的精度,从而获得全局最优。仿真结果表明:鸡群算法优化后的FOPID具有更加优良的性能指标和抗扰动能力。
Nuclear power plant pressure regulator pressure control system with nonlinear, time-varying characteristics. Based on the fractional theory, a fractional order PIλDμ control algorithm (FOPID) is proposed to realize the point-to-surface control of proportional-integral-differential control (PID control) and introduce a short feedback algorithm to make the control more accurate. Compared with the traditional PID control, it has more than two degrees of freedom, and uses an indirect algorithm (Oustaloup algorithm) to realize the design of its controller. A biological heuristic algorithm named flock algorithm (CSO) Solve the fractional PI ~ λD ~ μ parameter setting problem. In order to avoid local optimum and precocity, we propose an improved CSO algorithm (ICSO), which changes the position updating formula of chickens and introduces chaos initialization and random weights to improve the precision of the algorithm and obtain the global optimal excellent. Simulation results show that FOPID optimized by flock algorithm has better performance and anti-disturbance ability.