Hybrid RSM–RFR Approach for Economic Load Dispatch of a Thermal Power System Considering Transmission Losses

This paper provides a thorough exploration of Economic Load Dispatch (ELD) issue of a thermal power system with a clear consideration of transmission losses, and the goal of generating sustainable, feasible, and economical generation scheduling. The study follows a systematic methodology beginning with problem formulation, literature review, generator cost modeling, and constraint handling, and concludes with the development and performance evaluation of a hybrid optimization–prediction framework based on Response Surface Methodology (RSM) and Random Forest Regression (RFR). The ELD problem is stated by taking into account realistic system properties like quadratic fuel cost functions, operating limits of the generators, power balance requirements, and transmission losses represented by B-coefficient approach. This model shows clearly the non-linear and constrained character of the dispatch problem and explains a necessity of strong optimization methods, which can be used to address realistic operating conditions. Response Surface Methodology is employed as the primary optimization tool for obtaining optimal dispatch solutions and demonstrates strong performance in minimizing total fuel generation cost while satisfying equality and inequality constraints. The results of optimization report a stable convergence, good transmission loss management, and good practical feasibility of operation of the multi-generator thermal power system. The Random Forest Regression-based dispatch predictor is then developed using the optimized RSM solutions as training data. The trained RFR model successfully captures the non-linear relationship between system load demand and optimal generator outputs, enabling fast dispatch estimation for new and unseen operating conditions. Although minor deviations from RSM solutions are observed due to the predictive nature of machine learning, the application of generator limit correction and power balance adjustment ensures that all predicted solutions remain physically feasible and practically acceptable. Comparative analysis between RSM optimization and the hybrid RSM–RFR framework demonstrates that the proposed hybrid model achieves near-optimal economic performance with a significant reduction in computational effort. The effectiveness of the proposed methodology is supported by the high prediction accuracy, better convergence behavior, and stable transmission loss treatment. This study successfully demonstrates that the conventional PSO–ANN framework can be effectively replaced by a hybrid RSM–RFR approach for solving the Economic Load Dispatch problem of a thermal power system considering transmission losses. The Response Surface Methodology offers good optimization of the total fuel cost minimization whereas the random forest Regression offers quick and precise prediction of optimal values of dispatch of the generators. The coefficient of high is.  determination (R² = 0.9975), the effectiveness of the proposed methodology is confirmed by stable convergence behavior, and viable transmission loss handling. Thus the hybrid RSM RFR method offers a trade-off, real-world and computationally efficient solution to Economic Load Dispatch problems and is very adaptable in real-time power systems tasks where quick, precise, along with reliable dispatch choices must be obtained.. The proposed framework can also be extended in future work for renewable energy integrated systems, microgrid dispatch optimization, and multi-objective economic–environmental dispatch problems.

[1] Wood AJ, Wollenberg BF, Sheble GB. Power Generation, Operation and Control. 3rd ed. Hoboken, NJ, USA: John Wiley & Sons; 2014. [2] Kundur P. Power System Stability and Control. New York, NY, USA: McGraw-Hill; 1994. [3] Saadat H. Power System Analysis. 3rd ed. New York, NY, USA: McGraw-Hill; 2011. [4] Grainger JJ, Stevenson WD. Power System Analysis. New York, NY, USA: McGraw-Hill; 1994. [5] El-Kib AA. Economic dispatch of power generation. IEEE Power Engineering Review. 1999;19(8):17–20. [6] Gaing ZL. Particle swarm optimization to solving the economic dispatch considering generator constraints. IEEE Transactions on Power Systems. 2003;18(3):1187–1195. [7] Park JB, Lee KS, Shin JR, Lee KY. A particle swarm optimization for economic dispatch with non-smooth cost functions. IEEE Transactions on Power Systems. 2005;20(1):34–42. [8] Walters DC, Sheble GB. Genetic algorithm solution of economic dispatch with valve point loading. IEEE Transactions on Power Systems. 1993;8(3):1325–1332. [9] Chen PH, Chang HC. Large-scale economic dispatch by genetic algorithm. IEEE Transactions on Power Systems. 1995;10(4):1919–1926. [10] Nanda J, Hari L, Kothari ML. Application of genetic algorithm to economic load dispatch with line flow constraints. International Journal of Electrical Power & Energy Systems. 2002;24(9):723–729.