摘要
针对RBF网络学习算法工作量大和类别数需预先确定的问题,通过引入具有自适应机制的的遗传算法,结合梯度下降法交互运算,提出了自适应遗传算法和径向基函数网络相结合的旋转机械故障诊断方法。通过对样本模式聚类和故障状态的分析,并利用自适应遗传算法优化相关的RBF网络,有效地解决了隐节点数和各参数的取值问题。应用结果表明,RBF网络和自适应遗传算法相结合提高了全局寻优效率。
In allusion to the problem of workload and need to be a pre-determined number for RBF network learning algorithm. Through introducing self-adaptive mechanism of genetic algorithm and combined with gradient descent method ,put forward self-adaptive genetic algorithm and radial basis function networks a combination of rotating machinery fault diagnosis methods. Through analysis sample model clustering and the state of fault and use adaptive genetic algorithm to optimization related to the RBF network,can effectively solve the hidden nodes and the problem parameters. Application results show that RBF network and the combination of adaptive genetic algorithm improve the efficiency of the global optimization.
出处
《煤矿机械》
北大核心
2010年第1期241-244,共4页
Coal Mine Machinery
