Abstract:
Power Economic Dispatch (PED) is one of the essential steps in operation and planning of a power system. It is online function and is carried out after every fifteen minutes or on request in control centres. It is generation allocation problem which can be defined as the determination of optimal generation schedule of machines subject to satisfaction of equality and in-equality constraints. PED is non-convex problem in nature because of Valve Point Loading Effects (VPLEs), Multiple Fuel Options (MFOs) and Prohibited Operating Zones (POZs). However, mostly it is addressed as a convex optimization problem solved by conventional techniques e.g. equal incremental cost criterion, Newton's Method (NM), Lambda Iteration Method (UM), Non-Linear Programming (NLP), Linear Programming (LP) and Dynamic Programming (DP) etc. But these conventional techniques lack the ability of solving such complex PED problems. Various Artificial Intelligence (AI) tools have been developed as the potential solution methodologies for solving such non-convex PED problems i.e. Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Fuzzy Logic, Artificial Neural Network (ANN), Simulated Annealing (SA) and Tabu Search (TS) etc. These techniques are sometimes used in hybridization to further optimize search time and results. Evolutionary Algorithms (EAs), their variants and hybrid approaches are being developed for PED solution. This research presents the implementation of a variant of Differential Evolution (DE) named Stud Differential Evolution (SDE) on convex and non-convex PED. DE is from the class of EAs. SDE is an improved version of traditional DE that includes both genetic algorithm and evolutionary strategies; as a result it is capable of providing better solutions compared to classic DE. In the present work, SDE has been developed for convex and non-convex PED due to MFOs and VPLEs. The proposed SDE model has been programmed in C++ environment and tested on 10-machines standard test systems available in literature for multiple fuel PED problems. The proposed SDE's effectiveness has been validated by comparing the simulation results with those of the other algorithms reported in literature.
