【作者】 孙旭东； 王卫红；

【Author】 SUN Xu-dong, WANG Wei-hong (Automatic Control Department, Beijing University of Aeronautics and Astronautics, Beijing 100083, China)

【机构】 北京航空航天大学自动控制系；

【摘要】 单神经元PID以其计算量小,结构简单的优点被广泛应用。对基于单个神经元的直接PID控制算法进行改进,改进算法中采用不同的学习速率和权值调整方法,更好的体现了PID控制器中比例、积分、微分环节的作用。通过对典型的转台动态伺服系统仿真表明,改进算法比一般的单神经元PID算法控制效果好,抗干扰性好。更多还原

【Abstract】 Single neuron PID is extensively used because of its less computational cost and simple architecture. The single neuron PID algorithm was improved. The improved algorithm applies different learning rates and weight-adjusting methods, which show the effects of proportional, integral and derivative parts. The simulation on the servo system indicates it is more efficient than common single neuron PID algorithm and has better anti-disturbance ability.更多还原