Research On LED Driver And Monitoring System

With the aggravation of the energy crisis, energy conservation has become an important issue, and lighting as huge power consumption industry is faced with deep reform. In this background, LED will gradually become the mainstream of modern lighting due to its energy conservation and environmental protection, high photosynthetic efficiency, long life, strong controllability, etc. This article takes LED as research object, mainly carried on the research to LED driver and monitoring system, design a set of LED lamps and lanterns system composed of adjustable LED driver and ZigBee wireless sensor network, the whole system can realize the function of adjustable LED light and real-time monitoring, extending the application of the LED. The main research contents are as follows:Firstly, this paper analyzes the topology of LED driver, comparing the advantages and disadvantages of conventional dimming method, proposed the scheme of quasi-resonant flyback converters with PWM dimming, and combined with the technology of APFC and primary feedback, finally designed a 45 W LED driver based on quasi resonant mode. Test results indicate that the conversion efficiency is greater than 0.9, the power factor is greater than 0.99 and adjustable optical depth reaches 0.8 when the input voltage in the range of 198-264 V AC.Secondly, in view of the demand for LED driver real-time monitoring, this paper carried research on the hardware circuit of LED monitoring system, developed a set of LED driver voltage, current, temperature collectable monitor hardware circuit which based on CC2530, and combined with ZigBee wireless sensor technology, design a set of LED real-time data monitoring system, this system can realize the function of data real-time acquisition, display and save.Finally, this article carried research on the LED junction temperature prediction. Using BP neural network deducting LED driver electric parameter to LED junction thermal parameters. Constructing the 45 W lamps lighting system test platform, made the input voltage, current and junction temperature measurement experiment, established the input and output network model of the corresponding system by BP neural network identification algorithm. The relationship between junction temperature and input voltage, current can be concluded through simulation, then compared to LED junction temperature and life relation curve, can quickly forecast residual life of LED.