An efficient predictor-corrector interior point algorithm for security-constrained economic dispatch

Abstract

This paper deals with the application of an advanced interior point method to the security-constrained economic dispatch (SCED) problems through successive linear programming. The nonlinear SCED problem is linearized, and then solved by a predictor-corrector interior point method. Besides describing the basic algorithm, the paper focuses on several important issues that are critical to its efficient implementation, including the adjustment of barrier parameter, the determination of initial point, and so on. Computational experiments are conducted to evaluate their impact on the performance of the algorithm. Some suggestions, such as using the feasibility condition to adjust the way of computing barrier parameter /spl mu/ and customizing initial point by adopting a relative small threshold, are proposed to reduce the overall iterations required by the algorithm. The computational results on power systems of 236 to 2124 buses have shown that these suggestions are very effective, improving the performance of the algorithm by a factor of 2. Comparison with a pure primal-dual interior point method is also provided to demonstrate the superiority of the proposed predictor-corrector method.