基于大脑情感学习模型的转台伺服系统设计

Design of Brain Emotional Learning Model Based Turntable Servo System

提出了基于大脑情感学习(Brain Emotional Learning,BEL)模型的高精度转台伺服系统智能控制方案。BEL模型是一种模拟哺乳动物大脑情感学习过程的仿生计算模型。设计了融合系统跟踪误差、控制输入等信息的BEL智能控制结构,通过选取不同的感官输入信号可获得不同的控制结构,采用联想学习方法在线学习BEL模型内部的节点权值来调节控制器参数,从而实现转台伺服系统的自适应跟踪控制。仿真和实验结果均表明,BEL智能控制器学习能力强,能抑制摩擦等非线性干扰因素,在实时控制系统中表现出较好的稳定性和较高的跟踪性能。

Brain emotional learning （BEL） model based intelligent controller for a turntable servo system with high precision is presented. BEL model is a bio-computational model inspired by the emotional learning mechanism in mammalians brain. BEL based intelligent Controller is designed by fusing the information of system tracking error, control variables and etc. Different control structure can be designed by selecting different sensory input signals, and controller parameters are adjusted by online learning the nodes weights in BEL model based on an associative learning method, and then the adaptive tracking control of the turntable servo system is realized. Simulation and experimental results show that the BEL based intelligent controller has strong learning ability, and can restrain the nonlinear factors such as friction, and exhibits good stability and high tracking precision in application of the real-time control system.