A Machine Learning-Integrated Model for Cost-Optimized Economic Load Dispatch

Economic load dispatch (ELD) is essential for efficient power system operations, where the objective is to minimize generation costs while adhering to operational constraints. This paper presents a hybrid optimization approach that combines the Bat Algorithm (BA) with linear regression to improve the accuracy of cost estimation in ELD problems. The BA, inspired by the echolocation behavior of bats, is employed to explore the solution space, while linear regression is integrated to model and predict cost functions based on power generation parameters. The proposed hybrid approach is tested on two benchmark systems: the IEEE 30 and IEEE 14 bus test systems. In addition to optimizing generation costs, the integration of linear regression aids in refining the relationship between power generation and cost, enhancing convergence speed and solution precision. Simulation results demonstrate the effectiveness of the BA with linear regression, providing superior performance in terms of cost minimization and computational efficiency compared to traditional methods. This approach offers a promising alternative for solving nonlinear, constrained ELD problems with improved predictive accuracy.