Daylight perceptive lighting and data fusion with artificial neutral network

Abstract The basic concept of daylight perceptive is that the artificial light automatically dims to the appropriate level to keep constant indoor illumination when sufficient daylight is not available. Energy efficient character of daylight perceptive lighting was analyzed and a method to evaluate energy saving was present; according to the presentation, the more energy is saved if the longer Tlast—duration of indoor illumination being is less than threshold. A daylight perceptive controller was developed and the output power of LED light source under different PWM was measured, according to the measurement, work current and output power of LED is proportional to duty cycle. With two sensors, a data fusion method was put forward to exclude interference and improve lighting comfort. With more than two sensors, data fusion based on neutral network was discussed.