基于BP神经网络的闭环太阳能自动逐日系统

本文针对目前主流的光电追踪式太阳能难以检验是否达到最大正对面积的缺陷,设计一种在视日运动轨迹追踪基础上改进的太阳能自动逐日系统.该系统以89C52单片机作为控制核心,通过GPS模块接受地理信息,初始化电池板在该地区的倾斜角和方位角.当选择粗调节模式时,单片机通过设定的太阳运动轨迹算法,控制两个自由度上的步进电机的旋转;当选择人工精调节模式时,观测者通过观察电池板上的阴影长度,手动调节,同时单片机把调节信息通过RS232串口通讯传递到计算机,计算机利用数据库的样本数据,采用有导师学习的BP神经网络算法,在Matlab中计算并反馈之后时刻的电池板的旋转角度;当选择无人精调节模式时,系统定时自行从按照上述算法调整机械部分旋转.该系统特别适合大面积电池板的系统.该系统具有测试精度高的特点,达到了设计要求.

Design of closed-loop automatic solar energy tracking system based on BP neural network

This paper shows the designing of an advanced solar energy tracking system based on the solar trajectory tracking. This system aims at overcoming the drawback of current solar energy tracking systems which could not accurately examine whether to get the maximum surface area of the solar cell to the sun rays. This system, with the 89C52 MCU as the control core, receives the geographic information by the GPRS module as to initialize the inclination angle and the azimuth angle of the solar cell in this region. When selecting the coarse adjustment mode, the micro controller controls the two degrees ~ff freedom on the rotation of the step motor by setting the solar trajectory algorithm; When selecting the artificial accurate regulating mode, the observers could manually controls the system by observing the lengths of the shadow on the panel and simultaneously the micro controller could transfer the regulating information to the computer through the serial communication via RS232 port. By mining the sample data from the database and using learning algorithm of the BP neural network the computer could calculate and feedback the rotation angle of the following moment to the solar cell plate; when selecting the automatically accurate adjustment mode, the system could time to automatically adjust the rotation of the mechanical part according to the above algorithm. This system is especially suitable for those solar cells with large areas of batter plates. The system is highly accurate according the design requirement and the design details are shown.