A computational tool for dispatching units in pool-based electricity markets

In this paper a new approach for solving mixed-integer nonlinear programming problems is presented. The approach is named probabilistic search. The main idea is to search the feasible solution area probabilistically instead of randomly. The presented approach is used to determine the dispatch value of each generator in a pool-based electricity market. In this approach the ON and OFF units are determined using the probabilistic search and the generation value of each unit is determined using the linear programming. A software was developed based on the proposed approach for unit commitment. The software is used in the Iranian electricity power pool now.