粒子群优化BP-PID的矿井提升机调速系统

传统PID控制器在矿井提升机变频调速系统应用中,由于控制参数固定且不易整定,导致电机转速超调大、电磁转矩和转子磁链脉动大,进而出现矿井提升机调速系统控制效果差的问题。针对这一问题,文中提出一种改进粒子群优化BP神经网络PID控制器的算法。由于BP神经网络算法存在收敛速度慢和极易陷入局部最优的缺点,现将粒子群算法收敛速度快和全局最优特性与神经网络结合,并通过设计神经网络收敛系数进一步加快收敛速度。仿真结果表明,粒子群优化的神经网络控制效果比神经网络好,且效果明显优于传统PID控制器;相较于神经网络PID控制器,矿井提升机转速调节系统稳速调节速度明显提高;与传统PID控制器相比,电机电磁转矩和转子磁链脉动明显降低,具有较强的稳定性和鲁棒性。

Particle Swarm Optimization BP-PID of Rotor Variable Frequency Speed in Mine Hoisting System

The control parameters of traditional PID controller applied in the mine hoist frequency conversion speed control system is fixed and difficult to be set, which aeaks leads to large speed overshoot and ripples of the electromagnetic torque and rotor flux linkage. In order to improve the performance of the system, an improved particle swarm optimization BP neural network PID controller algorithm of hoist rotor frequency conversion speed regulation system is proposed in the present study. The application of particle swarm algorithm which has fast convergence speed and global optimal characteristics in neural network can overcome the disadvantages of slow convergence speed and easy to fall into local optimum of BP neural network. In addition, the neural network convergence coefficient is designed to further accelerate the convergence speed. The simulation results show that the neural network control effect of particle swarm optimization is better than the neural network, and the effect is obviously better than the traditional PID controller. Compared with the neural network PID controller, the steady speed adjustment speed of the mine hoist speed regulation system of particle swarm optimization is obviously improved. Compared with the traditional PID controller, the ripple of the electromagnetic torque and rotor flux of the motor are significantly reduced, which has strong stability and robustness.