MEMS Gyroscope Temperature Drift Compensation for SVR Optimized by Artificial Bee Colony

Abstract

AbstractAiming at the problem that the output of MEMS gyroscope is susceptible to environmental temperature, resulting in temperature drift and reduced measurement accuracy, an artificial bee colony algorithm ABC optimized support vector machine regression SVR temperature drift compensation method is proposed. The instrument is placed in a high and low temperature box to collect the output data in a variable temperature environment, smooth it and use it as sample data, and then map the data from the low-dimensional space to the high-dimensional space, and use the support vector machine regression of the radial basis kernel function to establish the model is used for linear fitting, in which the penalty parameters and the kernel function parameters are optimized by the artificial bee colony algorithm. Finally, the actual MEMS gyro temperature drift data is used to verify the proposed compensation method, and compared with the least square segmentation method and BP neural network method, the mean square error is reduced by 74.9%, 11.9%, the variance is reduced by 74.9%, 3.1%, and the compensation effect is greatly improved.KeywordsMEMSGyroscopeArtificial bee colonySupport vector machinesTemperature drift compensation