Distributed Parametric Model Predictive Control of Street Lighting

Designing a distributed Model Predictive Control (MPC) system for street lighting is characterized by optimization computational complexity, communication burden, lack of tailored reference setpoint of each lamp, etc. This paper proposes a distributed-MPC approach for street lighting with an adaptive dimming profile. The street is first divided into zones and each zone may contain one or more lamps. The adaptive dimming profile is generated within a simulation environment based on the local data of weather, pedestrian, traffic, and original dimming profile. Each zone has its own MPC controller which calculates optimal lighting control actions, and it may have its own reference setpoint. These zones, and thus the controllers, communicate with each other mutually and iteratively toward a joint lighting energy efficiency goal of the whole street. Only essential data is transmitted to the central controller. To verify the validity and effectiveness of the proposed approach, a one year simulation is conducted using MATLAB/Simulink environment. Simulation results show that the proposed distributed-MPC algorithm can save up to 34.85% of the operating costs. The results also showed that the proposed algorithm can achieve a good performance of the whole system with limited burden of communications.