A Simplified Model MEMS Gyroscope Zero Bias Temperature Drift Calibration Method Based on BP Neural Network

To deal with the effects of temperature drift and static bias errors on MEMS gyroscope, this paper proposes a new error calibration method, which is based on BP neural network technology to ensure the calibration accuracy, and improves the calculation and storage efficiency in practical applications through the B-spline method in computer graphics. In this paper, the most common MPU6050 chip on the market is used as the experimental object. Compared with the traditional method, the design can effectively improve the calibration accuracy. In terms of the completeness of the research, the existing methods generally lack consideration of the feasibility of the actual application environment. The optimal algorithm application steps from error calibration to model optimization are comprehensively studied. The final experiments show that B-splines can improve the computing and storage efficiency while ensuring the error effect of BP neural network, and increase the feasibility of the model to be applied in resource-constrained MEMS application scenarios.