Population-based bisection for solving non-convex economic dispatch with loss

The Combined Heat and Power Economic Dispatch (CHPED) problem poses non-convex optimization challenges due to heat-power coupling, valve-point loading effects, and transmission losses modelled by Kron’s formula. Traditional methods, including evolutionary algorithms with penalty functions, often rely on random searches that may struggle with solution accuracy and time complexity. This paper introduces a novel hybrid approach, integrating population-based exploration with a recursive bisection method for a more rational and efficient search. Unlike conventional evolutionary algorithms, the population-based bisection offers a structured approach to solving the non-linear equation system in Kron’s formula, resulting in better handling of transmission losses. Additionally, the proposed method aligns with conventional power system analysis practices by incorporating a slack bus (reference bus) to ensure compatibility with standard load flow calculations. Benchmark tests demonstrate significant improvements in solution accuracy, convergence speed, and computational efficiency compared to traditional methods. This hybrid framework offers a more reliable and system-compatible solution for optimizing CHPED problems in modern power systems.