Optimizing power consumption in street lighting systems with smart systems

Street lighting is needed at dusk in areas of high socio-economic and cultural importance (roads, gardens, crowded sidewalks). In order to ensure the safety of pedestrians or vehicles in these areas, the environment should be illuminated with an appropriate level of illumination. At the same time, considering that street lamps are almost everywhere, unnecessary use of electrical energy will lead to inefficient use of electrical energy. In this thesis, it is aimed to make street lamps smart and to use them actively at appropriate light levels when necessary. The power values at which the luminaires should operate were determined according to the positions of the pedestrians to the poles. These values are compared according to the normal time power consumed by luminaires at different distances. In order to increase the electrical energy saving rate in LED type street luminaires, the luminaire power levels were analyzed according to the luminous intensity that should be around the luminaires. This analysis is detailed with Artificial Neural Network (ANN) methods. The values were optimized by training the system with the received data.