Enhancing flexibility of combined energy system by renewable energy sources using hybrid LOA-MVGAN approach

Renewable energy adoption is often limited by "heat-set" operation mode of Combined Heat and Power (CHP) units during winter, which reduces system flexibility. Installing small-scale Integrated Energy System (IES) with CHP units can improve flexibility and address this challenge. This manuscript presents a hybrid approach to enhance flexibility in operation of IES with renewable generation. The proposed approach combines the Lyrebird Optimization Algorithm (LOA) and Multi-view Graph Attention Networks (MVGAN), which is termed as LOA-MVGAN approach. The proposed technique's main goals are to increase system's operating flexibility, reduce generation costs, and guarantee the system's operational economy. Flexible load activities are optimized by the application of LOA approach. The MVGAN algorithm is utilized to predict electrical load demand. The proposed approach is assessed and contrasted on MATLAB platform with other existing methods. Compared to existing methods as Salp Swarm Algorithm (SSA), Cuckoo Search Algorithm (CSA), and Heap-based optimizer (HBO), proposed technique yields better results. The LOA-MVGAN method achieves the cost of 1.2 $, which is lower than that of existing methods. This cost reduction aligns with Sustainable Development Goal (SDG) 7 (Affordable and Clean Energy) by promoting economic efficiency in renewable energy systems, making clean energy more accessible and financially viable.