Experimental Validation of an AI-embedded FPGA-based Real-Time Smart Energy Management System Using Multi-Objective Reptile Search Algorithm and Gorilla Troops Optimizer

This paper proposes an AI-embedded FPGA-based Smart Energy Management System (SEMS) that ensures intelligent, secure, consistent, and synchronous energy management in an isolated microgrid. The proposed techno-economic SEMS comprises two levels of control to achieve optimal management and operation for an isolated microgrid. The first level adopts the use of the FPGA as a central controller, which is characterized by its high processing speed and small settling time. The second level aims to develop a coordinated operation strategy based on the optimal operation and management of an isolated microgrid in order to optimize the coordinated use of backup sources. An efficient multi-objective optimization problem for optimal operation and management of the microgrid is formulated. Two multi-objective optimization algorithms namely, Gorilla Troops Optimizer (GTO) and Reptile Search Algorithm (RSA) are applied to solve the optimization problem. The three main ob-jectives considered in this study are to minimize the operating costs, the loss of power supply probability (LPSP), and the surplus power consumed by the dummy load. The results prove the superiority of the RSA algorithm in achieving the goals of the objective functions. Within 100 min of the experimental testing, it achieves the lowest operating cost 166.2423 $. The cost savings reach about 6.467 % when using the RSA, while it is 6.0363 % when using the GTO. The developed SEMS reduces the wasted power in the dummy load. In addition, it achieves the lowest value of LPSP about zero, which is considered the best value as it ensures continuous supply.