基因优化网络技术在磁悬浮轴承控制中的研究

针对普通BP算法收敛速度慢，容易陷入局部极小值以及网络初始状态对网络性能的影响等问题，建立了单自由度磁悬浮控制系统的数学模型，设计了一种基因算法优化神经网络的智能控制方法，该方法提出用基因算法离线优化网络。并设计了一种双链异或交叉基因算子，改善了基因算法的效率，在较大程度上改善了单自由度磁悬浮系统的性能。

Research on Neural Network Optimized by Gene Algorithm of the Magnetic Levitation System

With the aim at the problems of normal BP algorithm with a slow constringence rate, easily falling into local minimum values and effect of network initial state upon network performances,etc,the mathematical model of one dimensional magnetic levitation control system is established, and the intelligent control method of the optimal neural network by gene algorithm is designed in this paper. Also,this method advances the off line optimized network using gene algotithm. A kind of double chain XOR or crossover gene operator is designed so that the gene algorithm efficiency is improved as well as the performance of the one dimensional magnetic levitation system is improved to great extent.