Total optimization of a smart community by multi-population differential evolutionary particle swarm optimization

This paper proposes total optimization of a smart community (SC, Smart City) by multi-population Differential Evolutionary Particle Swarm Optimization (DEEPSO). In this paper, energy supply models such as electric power utility, natural gas utility, drinking water plant, and waste water treatment plant, and energy consumption models such as industry, building, residence, and railroad are utilized. Using the SC model, energy costs, actual electric power load at high load hours, and the amount of CO2 emission of the whole SC is minimized. Various evolutionary computation techniques such as particle swarm optimization (PSO) and differential evolution (DE) have been applied to the model. However, there is room for improving solution quality. Therefore, this paper proposes a multi-population DEEPSO using both migration and the best searching point so far among all of sub-swarms in order to improve solution quality. The proposed method is applied to a model of Toyama city, which is a moderately-sized city in Japan. Optimal operation by the proposed method is compared with that by the original DEEPSO with a single population.