Multi-Area Load Dispatch with Emission and Fuel Cost Constraints Using Deep Recurrent Lstm Under Stochastic Pv and Wind Power Integration

Abstract

An interconnected power system must employ multi-area economic load dispatch (MAELD) to keep generation dispatch efficient, satisfy load demands, and stay within technological limits. The goal of MAELD's operational constraints is to reduce fuel costs and emissions from multi-area power production facilities. Consequently, optimization solutions that successfully tackle MAELD problems are highly sought after. By merging an alternative model with the Deep Recurrent Neural Network (DRNN) one, this paper suggests a fresh approach to addressing MAELD problems. This paper provides a detailed account of the various optimization methodologies used to address the issue of load dispatch when non-conventional energy sources are present. Four domains utilizing 3-, 13-, and 40-unit systems were used to test the suggested DRNN and LSTM approach. Comparing this method to various optimization algorithms as firefly, Salp Swarm, Squirrel search, Particle Swarm, and Gross Hopper, results produced from the MATLAB/Simulink environment show that it delivers superior trade-off solutions without breaching limitations.