遗传算法优化的BP神经网络压电陶瓷蠕变预测

Prediction model of the creep of piezoceramic based on BP neural network optimized by genetic algorithm

针对压电陶瓷驱动器的蠕变误差随时间呈现非线性变化,会严重影响其定位精度的问题,提出遗传算法优化BP神经网络的压电陶瓷蠕变预测算法。采用遗传算法优化了BP神经网络的权值和阈值,构建了基于遗传算法的BP神经网络(GA-BP算法)的蠕变预测模型。用GA-BP算法对压电陶瓷蠕变进行了预测仿真,并将结果与实测数据进行了对比。结果表明,获得的蠕变预测结果与实验数据的最大绝对误差均不超过0.2μm,最大蠕变误差均小于1.5%,最大均方误差仅为0.004 6,因此,GA-BP预测模型可作为预测压电陶瓷蠕变误差的一种有效手段。

Aiming at the creep errors of the piezoelectric ceramics showing nonlinear change with the time,which affects the positioning accuracy of the piezoelectric ceramics badly,a creep prediction approach of back propagation neural network optimized by genetic algorithm is proposed for the piezoelectric ceramics.The weight and threshold of the BP neural network were optimized by genetic algorithm,and a creep prediction model was constructed based on BP neural network.The creep of piezoelectric ceramics was predicted by the model of GA-BP neural network.The results of BP neural network were compared with the measured data.The results show that,using this prediction mode,the maximal absolute error is below 0.2 μm,and the maximal creep error is below 1.5％.The maximal mean square error is 0.004 6.Thus,the GA-BP neural network prediction model can be applied into the creep prediction of the piezoelectric ceramics.