路灯智能识别控制系统的设计与实现

发布时间：2018-05-19 19:16

本文选题：PWM调制恒流驱动 + 运动目标识别　；参考：《中南林业科技大学》2016年硕士论文

【摘要】：近年来,LED灯以其低功耗、寿命长、安全、绿色无污染等众多优势快速占据了照明市场,然而现有路灯控制方式局限于分时段控制、难于统一管理、亮度不能自动调节及排查故障依赖于人工巡视,直接影响着LED路灯的应用与发展。设计并实现一种基于视觉感知技术的LED路灯无线自组网的智能识别路灯控制系统,具有重要的实际意义与应用价值。论文主要工作与取得的成果如下。(1)根据控制系统的功能和性能需求以及主控CPU的性能及功能,选择了合适的主控处理器,采用PWM调制恒流驱动实现路灯亮度的调节,完成了功能模块外围接口电路和供电电源设计,系统核心板结构的设计与接口板设计。(2)通过软件开发流程引出该设计系统对运动目标识别的功能性需求,选择合适的嵌入式操作系统；根据视频图像采集驱动V4L2及运动目标识别软件motion,设计并实现了运动目标识别系统的整体框架及系统流程。(3)在分析和比较几种常用运动目标识别算法优势与不足的基础上,提出了一种将中值背景模型算法与motion算法有机结合的优化算法,MATLAB仿真实验结果表明,该优化算法在识别精度、环境适应性及算法复杂度上都有一定的优势,该算法已经被应用到产品中。(4)完成了智能路灯识别控制系统的软件流程设计、视频采集功能与motion软件识别功能的测试,测试表明：该系统能够根据环境亮度以及路况识别,实现路灯开关和亮度调节的智能控制。实际运行表明,该智能识别路灯控制系统提高了识别精度,能够根据时间、环境亮度及路况信息实现LED路灯亮度的自动调节,在一定程度上实现了“智能学习”,达到了深度节能、便于监控和统一管理、降低运营成本的目标。
[Abstract]:In recent years, LED lamp has occupied the lighting market rapidly because of its advantages of low power consumption, long life, safety, green and no pollution. However, the existing control mode of street lamp is limited in time control, so it is difficult to unify management. Brightness can not be automatically adjusted and troubleshooting depends on manual inspection, which directly affects the application and development of LED street lamp. The design and implementation of an intelligent recognition street lamp control system based on visual perception technology for LED street lamp wireless ad hoc network has important practical significance and application value. The main work and achievements of this paper are as follows: 1) according to the function and requirement of the control system and the performance and function of the main control CPU, the suitable main control processor is selected, and the adjustment of the luminance of the street lamp is realized by using the PWM modulation constant current driver. The peripheral interface circuit and power supply design of the function module are completed. The design of the system core board structure and the interface board design. Select the appropriate embedded operating system; According to the video image acquisition driver V4L2 and the moving target recognition software motion. this paper designs and realizes the whole frame of the moving target recognition system and the system flow. On the basis of analyzing and comparing the advantages and disadvantages of several commonly used moving target recognition algorithms. This paper presents an optimization algorithm which combines median background model algorithm with motion algorithm. The simulation results show that the algorithm has some advantages in recognition accuracy, environment adaptability and algorithm complexity. The algorithm has been applied to the product. It has completed the software flow design of the intelligent street lamp recognition control system, the testing of the video capture function and the motion software recognition function. The test results show that the system can recognize the system according to the environment brightness and the road condition. The intelligent control of street lamp switch and luminance adjustment is realized. The actual operation shows that the intelligent recognition street lamp control system can improve the recognition accuracy, can realize the automatic adjustment of LED street lamp brightness according to time, environment brightness and road condition information, and realizes "intelligent learning" to a certain extent. Achieve deep energy saving, easy to monitor and unified management, reduce operating costs.
【学位授予单位】：中南林业科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM923.34;TP273

【相似文献】