齿槽效应是影响高速磁浮列车悬浮控制系统稳定性的重要因素,分析了悬浮间隙传感器齿槽效应产生机理,提出采用组合预测的方法建立传感器齿槽特性逆模型进行齿槽误差补偿,根据传感器齿槽特性分别建立RBF神经网络和LS-SVM齿槽补偿系统模型,通过齿槽位置检测线圈提供的补偿参考信号对传感器进行齿槽误差补偿。仿真结果表明组合模型能够较好地拟合齿槽逆特性,组合补偿模型的输出不受齿槽位置的影响,全量程最大误差为0.09 mm,在工作间隙范围内误差小于0.06 mm,且组合模型的补偿误差优于单一RBF或LS-SVM补偿效果,该方法可以有效地消除齿槽效应并提高传感器的检测精度,补偿后的传感器能够满足高速磁浮车悬浮控制系统运行要求。
The slot-effect is an important factor affecting the stability of the high-speed maglev train suspension control system. The generation mechanism of the slot-effect of the suspension gap sensor is analyzed and a combination prediction method is proposed to solve the slot effect problem. In this method,the combined model of the slot-effect inverse characteristic is designed to compensate the slot error.RBF neural network and LS-SVM slot compensation models are established according to the slot-effect characteristic of the gap sensor.The slot position detection coils are adopted in the probe of the gap sensor to provide the reference signal and detect the relative position in a tooth-groove period. The combined model compensates the slot error of the gap sensor according to the relative position signal. The simulation results show that the combined model can fit the inverse slot-effect characteristic well. The output of the combined compensator is independent of the tooth-groove position. The simulation studies show that this compensator can provide correct gap data with the error less than 0. 09 mm in full scale and less than 0. 06 mm in the normal work gap usually. The compensation precision of the combined model is superior to those only using single RBF or LS-SVM methods. The proposed method can effectively eliminate the slot-effect and enhance the detection precision of the sensor; with this method the compensated output of the gap sensor can meet the requirement of levitation control system of high-speed maglev train perfectly.