Using genetic algorithm for optimizing fuzzy logic controller for mode-based control algorithms of building automation systems

Abstract

The growing integration of renewable energy into building energy systems causes increasing complexity of energy conversion and distribution systems. This development creates the need for appropriate control algorithms implemented in building automation systems. We previously introduced the MODI-method to support structured development of mode-based control algorithms and allow simulation-based testing in early phases of the planning process. However, the control design concerns different aspects, such as efficiency and system lifetime. It is therefore challenging to determine the conditions of the transitions between operating modes. Furthermore, the control design process lacks an optimization approach for the generated control algorithms. In this paper, we investigate the application of a fuzzy logic controller to generate conditions for mode transition of control algorithms and transfer approximate human knowledge into the control design. We perform optimization based on genetic algorithm to improve the performance of the control system, considering several aspects. The case study presents structured development of a mode-based control algorithm for a cooling supply system and the functionality of a fuzzy logic controller implemented into the control algorithm. The optimization of the fuzzy logic controller is performed using genetic algorithm. As a result, the optimized parameters of the fuzzy logic controller are gathered, leading to improvement of the performance of the system.