Self Tunned Fuzzy-PID controller for LED in Daylight Harvesting Environment for Smart Buildings

Abstract

An aggregate of 19% energy is manifested in urban buildings worldwide, due to lighting. Daylight harvesting (DH) is an innovative technique which promises to reduce energy consumption by 20-60% in smart buildings (SBs). Since variations in daylight are nonlinear and weather sensitive, hence there is an urgent need of a control strategy which adapts to these variations swiftly and provides stable operation. This dissertation work presents efficient DH technique, incorporating optimum target setpoint (TSP) illumination calculations for different applications in SB. Optimum TSP calculations form basis of adaptive Fuzzy-PID controller. The controller promises self tunning of LEDs with nonlinearly varying daylight pattern. The Fuzzy controller tracks daylight changes and performs continuous tunning of control parmeters Kp, Ki & Kd. Takagi-Sugeno Fuzzy inference system is incorporated to tune nonlinearly varying daylight with linear PID controllers. The PID controller receives its inputs from fuzzy system and drives optimum PWM pulses to DCDC buck converter to regulate the output current. Output current from the converter is driven into LED to achieve TSP illumination. The simulation are performed in MATLABSimulink environment.