摘要
针对数字控制DC-DC转换器用传统PID控制时响应速度慢、超调量大、抗干扰能力不强等缺点,提出了基于遗传算法和CMAC神经网络算法的PID控制方法及模型。通过设计系统数学模型、CMAC-PID控制器模型及仿真模型等,对所建的模型进行了仿真验证,用遗传算法在线整定PID参数,实时自适应调节;前馈型CMAC神经网络用来抑制DC-DC转换器的非线性干扰,确保系统的控制精度和响应速度。解决了传统PID控制中存在的超调量大、抗干扰能力弱等缺点。仿真结果与实验数据对比表明,模型具有超调量小、抗干扰能力强等优点。
Due to slow transient response, large overshoot and weak anti - jamming ability of digitally controlled DC- DC converter, the conventional PID control method is very difficult to achieve the accuracy control, a control method based on genetic algorithm and CMAC neural network algorithm is proposed. By designing the system mathe- matical model, CMAC -PID Controller model and simulation model, the P1D parameters are optimized online with genetic algorithm (GA) to achieve superior feedback control, CMAC neural network feed -forward nonlinear control is used to suppress disturbances for achieving control accuracy and response speed, the large overshoot and weak anti -jamming ability in the conventional PID control are overcome,. The results of simulation show that the proposed control method has small overshoot and strong anti -jamming ability.
出处
《计算机仿真》
CSCD
北大核心
2016年第2期418-422,共5页
Computer Simulation
关键词
转换器
遗传算法
神经网络
控制
仿真
Converter
Genetic algorithm
Neural network
Control
Simulation