论文部分内容阅读
提出一种双模糊神经网络复合控制策略(D-FNNC)以解决永磁直线同步电动机(PMLSM)垂直提升系统因失步而不能稳定、安全运行难题。这种复合控制结构通过改变供电电压或电机运行速度实现失步预防。针对PMLSM的非线性,时变性,易受扰动且无精确数学模型特性,采用模糊控制器FC;利用神经网络控制器(NNC),克服FC存在稳态误差且无自适应能力问题;通过一种复合控制策略(CCS):一个随功角稳定余量增加而逐渐趋于1的非线性函数将两者结合,以提高系统动态性能,保证稳态精度,并提高系统自适应能力。仿真结果表明,该新型控制策略能有效预防失步发生,使系统时刻保持稳定运行。
A double fuzzy neural network compound control strategy (D-FNNC) is proposed to solve the problem that the PMLSM vertical lifting system can not be stably and safely operated due to out-of-step. This compound control structure prevents out-of-step prevention by changing the supply voltage or motor speed. Aiming at the nonlinear, time-varying, easily disturbed and imprecise mathematical model of PMLSM, the fuzzy controller FC is adopted. The neural network controller (NNC) is used to overcome the steady-state error of FC without self-adaptive ability. Compound Control Strategy (CCS): A non-linear function that gradually approaches unity as the power margin increases, to combine the two to improve system dynamics, ensure steady-state accuracy, and improve system self-adaptability. The simulation results show that the new control strategy can effectively prevent the occurrence of out-of-step and keep the system running steadily.