论文部分内容阅读
针对变结构、变时滞被控对象,将粒子群优化(PSO)算法与广义最小方差相结合,采用实时自校正过程对其进行控制,提出基于 PSO 自校正控制器算法.该算法应用隐式辨识方式,可减少辨识计算量,通过跟踪误差来改变辨识精度.以工业上典型的一阶、二阶和三阶系统的结构变化并伴随着有时滞突变的复杂被控对象进行仿真,并和基于最小二乘的传统自校正控制方法比较得知,在运用 PSO 自校正控制器的控制下,系统输出量与期望输出之间的方差趋于更小,控制跟随性和鲁棒性均较好.仿真结果表明该自校正控制器的有效性与应用价值.
Aiming at the variable structure and variable time-delay controlled objects, the particle swarm optimization (PSO) algorithm is combined with the generalized minimum variance to control it with the real-time self-tuning process, and the PSO self-tuning controller algorithm is proposed. The identification method can reduce the computational complexity and change the identification accuracy by tracking errors.The simulation is performed on the structural changes of typical first-, second-, and third-order systems in industry and along with complicated controlled objects with time- Compared with traditional self-tuning control method based on least squares, we know that under the control of PSO self-tuning controller, the variance between the system output and the expected output tends to be smaller, and the control followability and robustness are better The simulation results show the effectiveness and application value of the self - tuning controller.