An Improved Stochastic Linear Cross-entropy Method for Non-convex Economic Dispatch

Economic Dispatch (ED) is an important issue in the modern power system operation. This paper proposes an improved stochastic linear cross-entropy algorithm, namely ISCE for solving the ED problem, which is a non-convex, non-linear and non-differential problem subject to a number of equality and inequality constraints. To overcome a drawback of the cross-entropy method (CE) which may easily fall into a local optimum and to enhance the solution diversity, smoothing parameters in CE are modified to become self-adaptive which makes ISCE simpler and more flexible, as well as more effective. A 40-unit ED problem with valve point and a 24-unit combined heat and power economic dispatch problem (CHPED) are investigated. The experimental results confirm that ISCE is a powerful optimization technique in ED problems in terms of solution quality in comparison with some existing methods.