论文部分内容阅读
矿井提升机制动系统运行的可靠性直接关系到井下工作人员的安全,为了提高其故障诊断的准确度,提出了一种改进粒子群算法优化RBF神经网络的诊断方法,建立了以提升机制动系统的故障特征参数为输入,并以制动系统的几种主要故障类型为输出的故障诊断模型。鉴于遗传算法具有较强全局收敛性,将遗传算法中的交叉、变异的思想引入到粒子群算法中,并用其优化RBF神经网络的隐含层参数。仿真结果表明,改进后的粒子群诊断算法改善了提升机制动系统故障诊断的速度和精度。
In order to improve the accuracy of fault diagnosis, the reliability of mine hoist braking system is directly related to the safety of mine workers. A new improved particle swarm optimization RBF neural network diagnosis method is proposed, The fault characteristic parameters are input, and the fault diagnosis model output by several main fault types of brake system. Due to the strong global convergence of genetic algorithm, the idea of crossover and mutation in genetic algorithm is introduced into PSO, and the hidden layer parameters of RBF neural network are optimized. The simulation results show that the improved particle swarm optimization algorithm can improve the speed and accuracy of the fault diagnosis of hoist braking system.