【摘要】 运用k-均值聚类法计算RBF神经网络隐含层神经元激活函数的中心及归一化参数,提出并运用了在线自适应RBF权矢量修正算法进行摆式客车通过曲线时离心加速度的多步预测。经过对实测信号的仿真分析证明:RBF神经网络可以以满意的精度对摆式客车离心加速度进行多步预测,预测的时间间隔大,有效解决了由于各种因素造成的滞后补偿问题,并且可以完全替代用于触发倾摆和判定曲线方向的陀螺仪的作用。更多还原

【Abstract】 k-means clustering algorithm was used to compute the center and normalizing parameters of the active function of the hidden unit of the RBF neural networks’ hidden layer,and the online self-adaptive algorithm with adjusting the RBF neural weight vector was put forward to predict the acceleration in multi-step.The data detected was simulated and the conclusion was that the RBF neural network can predict the centrifugal acceleration of the tilting train in the multi-step and that the lag result from all kinds reason may be compensated,and it can replace the gyroscope and its tiger function of the tilting and justify the curve direction completely.更多还原