Research on Charging Pile Measurement Error Estimation Based on Alternating Linear Regression

With the widespread adoption of electric vehicles, the accuracy of charging piles has become a focus, prompting scholars to propose numerous algorithms. This paper presents a method that utilizes only simple power usage data, transforming a nonlinear model incorporating fixed losses and line losses into two linear regression models through an alternating linear regression approach. This method not only estimates the measurement errors of charging piles but also assesses the fixed losses and line losses at the charging station. Simulation experiments reveal that with properly selected initial values, the variation in pile coefficients is constrained within 1%, demonstrating the algorithm's high accuracy and hand practical value.