Economic Load Dispatch of Generating Units with Multiple Fuel Options Using PSO

This paper presents a method to solve the optimal generation and dispatch of electrical power with multiple fuel options at different power levels. In general, the cost function for each generator is considered by a single quadratic function of power, for the optimal generation and dispatch problem. However, it is more realistic to represent the generation cost function for fossil fired plants by non smooth cost function i.e. the total generation cost function has non differentiable points such as segmented piece-wise quadratic function and valve point loading. Some generating units utilize multiple fuels sources (such as gas, coal, oil etc) are faced with the problem of determining which is the most economical fuel to be selected. This problem has been solved by Hota & Das [17] considering equal incremental cost. In this method the program has to be run for a number of times for fuel combinations for different units and then to select the best one out of all combinations. This is a complicated process. Also if some other type of nonlinear cost function is considered along with this, the method will not be applicable. In view of the above mentioned problems of economic load dispatch (ELD) of multiple fuel generation, a more general heuristic method known as Particle Swarm Optimization (PSO) is considered in this paper to solve the problem in which either type of nonlinearity in cost function can be considered. For solution of this problem dynamic PSO is proposed. It is dynamic in the sense that the velocity bound or limit is updated in each iteration. This proposed method is applicable to the generating units that can use multiple fuels through valve at different generation levels as well as other problems which results in multiple intersecting continuous and discontinuous cost curves for any unit. The advantage of the method is that the method is applicable to both continuous and discontinuous cost curves of systems. Also it does not require selection of unit and type of fuel to be used after computation for a number of combination which is a cumbersome task as used in ref [17]. The complete method is explained and validated by taking an example. The simulation is carried out using MATLAB software. It has been shown that the method is simple, direct, and practical. It can be used for real time implementation and operation.