District Energy Optimization Based on MLP Simulation

Energy optimization is a substantial procedure for building management systems, especially for huge energy demand buildings, like factories, public transport facilities, etc. The aim of this paper is to present a simulation based decision support that models the energy balance and suggests the optimal energy-related actions based on specified optimization criteria. The energy modelling consists of the facilities energy demand and Renewable energy generation forecasting. A neural network is exploited to simulate the load of the district, while physical models are deployed to simulate wind power and Photo-voltaic energy power using weather data. Subsequently, the energy optimization is achieved through a multi-layer perceptron. Finally, the proposed tool is evaluated against real-life data from a Port, while the experimental results show the efficiency of the proposed system.