Evaluation of Meta-heuristic Optimization Methods for Home Energy Management Applications

Home energy management (HEM) requires optimization techniques to solve multi-variable and multi-objective problems. The optimal use of energy, the occupants comfort, the reduction of peak power and energy cost are objectives with dissimilar variables behaviors. Their solutions increase in complexity with the number of variables which would be a challenge if the real-time response is needed. Meta-heuristics optimization techniques offer great potential for the solution of such complex optimization problems, however, their main inconvenient is that a non negligible number of iterations must be executed which is reflected in a heavy computation loops and high resources utilization. In this paper, three meta-heuristic optimization algorithms are studied and evaluated focusing on HEM applications. As the better feasible option among them, Particle Swarm Optimization (PSO) method have been selected and applied to the comfort control in the residential environment. To achieve the real-time execution of the computational burden of the MPC-PSO implementation, the advantage of VLSI parallelism is used. The FPGA in the loop co-simulation results, using real data of an occupied house, demonstrate the potential of the implemented algorithm and future multi-objective target.