A Computationally Efficient Solution Algorithm for Least Absolute Value State Estimation Problem

Abstract

The most common methods used to solve the state estimation problem are the Weighted Least Squares (WLS) and the Least Absolute Value (LAV) algorithms. WLS is computationally efficient but it is not robust to outliers. On the other hand, LAV is robust to outliers as long as they are not present in leverage points but it is computationally demanding. This paper presents a new LAV-based algorithm that is fast and robust. The LAV problem is formulated as an unconstrained non-linear optimization problem that can be solved using gradient-based approaches. The motivation is to combine the desirable bad data rejection properties of LAV with the computational efficiency of WLS. The proposed algorithm is compared with the traditional WLS and LAV algorithms in terms of computational time and robustness using test cases of various sizes, from 30 to over 13000 buses. The proposed LAV algorithm is shown to be faster than the traditional LAV while possessing the same bad data handling properties.