Short-term hydro thermal generation scheduling using modified invasive weed optimisation algorithm

This paper presents a radical methodology to find the solution of optimal short-term hydrothermal generation scheduling problem using modified invasive weed optimisation algorithm (MIWOA). The hydrothermal system is highly non-linear and non-convex due to its constraints, making the optimal scheduling difficult. In this paper, a modified invasive weed optimisation algorithm with parallel probabilistic search method is applied for better convergence of the optimisation process and simultaneously, the implicit fundamentals of conventional invasive weed optimisation algorithm are preserved. The proposed MIWOA is applied to a test system with cascaded hydro reservoirs and a complementary thermal plant. The results are compared with those obtained by the dynamic programming, non-linear programming, classical evolutionary programming (CEP), fast evolutionary programming (FEP), improved fast evolutionary programming (IFEP), differential evolution (DE) and improved particle swarm optimisation (IPSO) approaches. The simulation results reveal that the proposed MIWOA performs better in finding the optimal cost than the other techniques compared.

Keywords: short-term hydrothermal generation, scheduling, modified invasive weed optimisation algorithm, MIWOA, prohibited operating zones, valve-point loading, cascaded hydro reservoirs, thermal power plants, simulation