论文部分内容阅读
闭环回路中输出噪声通过反馈与输入强相关,造成闭环系统的辨识难,甚至不可辨识的问题,并且在控制器参数的优化过程中,往往只强调性能指标或质量指标之一。为此,对于闭环稳定且可辨识系统,采用前向通道模型隔离输出噪声对输入的干扰,模拟闭环过程对象的动态特性,利用已知的闭环稳定PI/PID控制器参数和闭环系统输入输出数据,提取闭环过程信息,并用于确定前向通道模型参数优化的约束条件,充分结合PSO算法的全局寻优能力和SQP的局部搜索能力,优化辨识获得对象的前向通道模型。在此基础上利用NSGA-Ⅱ进行前向通道的基于质量指标和性能指标的多目标整定,采用相似度从优化得到的等效解中获得权衡最优解,最后将权衡最优解用于真实过程验证其优化的有效性。通过仿真研究,证明了该方法的有效性和合理性。
The output noise in the closed-loop loop is strongly correlated with the input through feedback, which makes the identification of the closed-loop system difficult or even unidentifiable. In the process of optimization of the controller parameters, only one of the performance indexes or quality indexes is often emphasized. For this reason, for the closed-loop stable and recognizable system, the forward channel model is used to isolate the input noise from the input noise and simulate the dynamic characteristics of the closed-loop process. Using the known closed loop stable PI / PID controller parameters and the closed-loop system input and output data , The closed-loop process information is extracted and used to determine the constraints of the parameters optimization of the forward channel model. The forward channel model of the object is optimized by combining the global optimization ability of PSO algorithm and the local search capability of SQP. Based on this, NSGA-Ⅱ is used to conduct multi-objective tuning of forward channel based on quality index and performance index. The similarity is used to obtain the optimal solution from the optimized equivalent solution. Finally, the optimal trade-off solution is used for real Process to verify the effectiveness of its optimization. The simulation results show that the method is effective and reasonable.