Electricity Generation Cost Optimization Based on Lagrange Function and Local Search

This paper proposes a method based on the Lagrange optimization function and local search technique for minimizing the total cost of two power systems. The first system comprises ten multiple fuel thermal units (MFTUs) while the second system combines the first system with renewable energies, solar and wind power. The proposed method has advantages over its conventional method without a local search technique, called the conventional Lagrange function-based method (CLM), such as having the same parameters and exploiting other search spaces after getting convergence. The proposed method is more effective than CLM for the first system with the last case of load demand. In addition, the proposed method has better costs than previous algorithms, such as the Hierarchical numerical method (HNUM), Hopfield neural network, Adaptive Hopfield neural networks (AHNN) and modified Lagrange neural network (MLANN). Especially, the proposed method can find smaller costs than them, up to $6.78, corresponding to 1.4% for Case 1, and up to $2.43, corresponding to 0.4% for Case 4. Only the proposed method is tested on the second test system. The simulation results indicate that the method is very efficient for the problem with solar and wind energies and multiple fuel thermal units.