A new calibration method for MEMS accelerometers with genetic algorithm

In this paper, we present a new method to calibrate the misalignment error and zero offset of MEMS accelerometer, which applies the Genetic Algorithm (GA) to process measured data and get the error model parameters. This method can effectively eliminate the error caused by the assembly deviation between the sensitive sensor unit and the sensor package shell. Results show that the calibrated output is far more accurate than the raw data obtained by only used factory calibration. The mean squared error (MSE) before calibration is 1.754×10 −3 g 2 which reduces to 2.828×10 −5 g 2 . Furthermore, the proposed procedure shows more advantages than the BP Neural Network method. The genetic algorithm behaves convenient and suitable for the calibration problem of MEMS accelerometer and reduces effect of the misalignment error and zero offset.