A method of state estimation with variable weights iteration based on piecewise linear weight function

Due to the traditional weighted least square state estimation (WLS) methods cannot resist gross errors, the state estimation method with selecting weiht iteration is proposed based on piecewise linear weight function. The M-estimation is converted to WLS algorithm by using of equivalent weight factor function, while the resistance to gross errors and estimate calculation can be accomplished simultaneously in iterative process. The corresponding weighted values are modified according to measurement residuals and linear weight function after each iteration, which can decrease the weight of gross errors and weaken its influence on the subsequent iterative process. The boundary coefficient of linear weight function is automatically calculated by historical multi-section measurements of state estimation, and it makes weight area more reasonable. Calculation results of practical power grid show that the proposed method can effectively reduce or eliminate the gross error influence of measurement system, and have strong practical value for online state estimation techniques.