Hybrid technique for leveraging unit commitment in smart grids: minimizing operating costs and carbon dioxide emissions

Smart Grid Systems, consisting of interconnected energy sources and consumers, often face challenges in managing energy resources. This manuscript presents a novel approach to address these issues by combining two advanced techniques Beluga Whale Optimization (BWO) and Tree Hierarchical Deep Convolutional Neural Network (THDCNN). The goal is to minimize total operating costs under various constraints, providing committed units and economical load dispatch for each operational hour. BWO optimizes cost, CO₂ emissions, and unit losses, while THDCNN predicts the optimal solution. The BWO-THDCNN approach is evaluated using statistical methods to compare cost and carbon dioxide emission reductions, unit losses, and load demand against existing methods like Color Harmony Algorithm (CHA), Gannet Optimization Algorithm (GOA), and Heap-Based Optimizer (HBO). Carbon dioxide emission reductions are 4.8% for CHA, 3.2% for GOA, 3.2% for HBO, and 5.8% for the proposed approach. The proposed method achieves the lowest cost at 1,007,525.23$. The Beluga Whale Optimization-Tree Hierarchical Deep Convolutional Neural Network method offers better carbon dioxide emission reduction and cost efficiency than existing methods. The novelty lies in integrating BWO with THDCNN to solve the UG problem in the smart grid.

Figure depicts the BWO-THDCNN technique for unit commitment in a smart grid. Inputs include smart grid data, unit demand, and operational constraints.The hybrid approach integrates BWO and THDCNNmethods to optimize scheduling. Outputs consist of committed units, cost reduction, high power delivery, and Carbon dioxide (CO₂) emissionreduction. The approach is compared with existing processesto demonstrate its effectiveness in minimizing costs and emissions while ensuring a reliable power supply.