BP混沌混合神经网络在光纤陀螺温度漂移预测中的应用

Application of chaos-BP combined artificial neural network in predicting FOG temperature drift

提出使用BP混沌混合神经网络建立FOG温度漂移模型的方法。该方法在BP算法中采用了改进型Logistic-Map映射生成的混沌变量,能够避免陷入局部最小,可迅速达到全局最优。应用该方法分析某型FOG温度漂移实测数据,结果表明其具有良好的预测效果。

The FOG’s temperature drift model was established using a chaos and BP combined artificial neural network(ANN),in which an improved Logistic-Map mapping was employed to produce a chaos variant.Due to this variant,the chaos-BP algorithm converges globally and had no local minimum.The temperature drift’s model of A certain FOG was analyzed using this algorithm,and the prediction of independent tested data was verified.