Cumulative Error Elimination for Long-stroke PMLSM Mover Measurement Based on BP Neural Network Model

In order to solve the problem of cumulative error of long-stroke linear electromotor displacement measurement based on machine vision, a method of eliminating cumulative error of long-stroke displacement measurement based on BP neural network is proposed in this paper. Firstly, according to the motion characteristics of linear motor and the principle of machine vision measurement, the displacement platform of linear motor is built, and one-dimensional speckle image is selected as the target image. Then, according to the causes of cumulative errors, a BP neural network model is established to map the relationship between the measured values and the real values. Finally, the maximum absolute error is set up to ensure the stability of the model and the problem of long-stroke measurement is solved by periodic segment measurement. The simulation results show that the proposed method can effectively reduce the cumulative error of the measurement of the long linear motor motor, and has high robustness and real-time performance.