【摘要】 为了提高快速伺服刀架的控制性能,减小跟踪误差,实现正弦网格表面精密加工,提出了基于RBF和BP神经网络的自适应PID控制策略。由仿真结果可以看出,采用基于神经网络的自适应控制算法,使跟踪误差的最大值降低为1.37μm,跟踪误差的绝对均值降低为0.52μm。这两项指标相对与传统PID控制分别降低了28%和40%。更多还原

【Abstract】 To improve the control performance of FTS(fast tool servo),reduce the following error and achieve the fabrication of sinusoidal grid surface,an adaptive PID control method was brought out based on RBF(Riverbank filtration) and BP(Backpropagation) neural networks.This method can improve the tracing accuracy of FTS and then reduce the shape error of the fabricated surface.From the simulation results,it can be seen that by adopting this control algorithm,the maximum tracing error is 1.37μm,and the average tracking error is 0.52μm,each reduced by 28% and 40%.更多还原