Application of computer intelligence to the unit commitment problem

This paper presents an-application of intelligent techniques for solving the unit commitment problem using Simulated Annealing, Tabu Search, and Neural based Simulated Annealing and Neural based Tabu Search. The objective of this work is to schedule the units of a power system in such a way that the overall generation is able to meet the demand while also minimizing the overall production cost. Simulated annealing is a powerful technique for solving combinatorial optimisation problems. It has the ability of escaping local minima by incorporating a probability function in accepting or rejecting new solutions. Tabu Search is a powerful optimisation procedure that has been successfully applied to a number of combinatorial optimisation problems. It has the ability to avoid entrapment in local minima by employing a flexible memory system. The neural network combines good solution quality for Simulated Annealing and Tabu Search with rapid convergence for artificial neural network. By doing so, it gives the optimum solution rapidly and efficiently. The Neyveli Thermal Power Station (NTPS) Unit II in India has been considered as a case study and extensive studies have also been performed for different power systems consisting of 10, 26, 34 generating units. The data collected has been used for implementation in the above methods. Numerical results are shown comparing the cost solutions and computation time obtained by using the intelligent techniques with the conventional method like Dynamic Programming and Legrangian Relaxation to reach proper unit commitment.