Hydrothermal generation scheduling package: A genetic based approach

Abstract

Novel solution algorithms and results based on a genetic algorithm for solving the hydrothermal generation scheduling (HTGS) problem are presented. This is a nonlinear, combinational optimisation problem which aims to minimise the total fuel costs of a power system while satisfying various local and coupling constraints. This results in a complete and efficient HTGS software package for system operation planning needs. In the thermal unit commitment subproblem, the difficult minimal uptime/downtime constraints are embedded and satisfied throughout the proposed encoding and decoding algorithms. Therefore, the global optimum of the problem can be approached with rather high probability. In the hydroelectric scheduling subproblems, complete solution algorithms and encoding/decoding techniques are proposed for solving different types of hydro plants involving hydraulically independent plants (HIPs), hydraulically coupled plants (HCPs), and pump-storage (P/S) plants. In the proposed approach, the hydraulically coupled plants which are located on the same river are solved concurrently. The difficult water balance constraints caused by hydraulic coupling are embedded and satisfied throughout the proposed encoding and decoding algorithm. The software package is applied with great success to the actual Taipower system, which consists of 34 thermal units, two HIPs, three HCPs, and four P/S units.