Photoelectric sensor design of solar auto-tracking system.

Solar power is the ideal use of solar energy. It is better to adopt tracking mode to track the sun automatically, and the solar receiver can get more solar energy for tracking system lets incident sunlight keep parallel to collector. The auto-tracking modes include program tracking mode, photoelectric tracking mode and hybrid tracking mode. Considering that program tracking mode has accumulative error although it can be all-weather tracking, and photoelectric sensor tracking has better tracking sensitivity and higher tracking precision only in the sunny day, solar auto-tracking system generally adopts hybrid tracking mode which is a combination of program tracking mode and photoelectric sensor tracking mode; for photoelectric sensor tracking mode eliminates accumulated error of program tracking mode, the system can achieve the goal of all-weather high-precision tracking. In this mode, photoelectric tracking mode uses light intensity detection information of photoelectric sensor to adjust the tracking system's position, so the design of photoelectric sensor's structure can directly affect solar auto-tracking system's tracking accuracy and running stability. At present, the most widely used types of photoelectric sensor are baffled type, pyramidal type, tube type, etc. Baffled type photoelectric sensor is vertically arranged on the base; making use of barrier, when there exists the included angle between barrier and incident light, light sensors on both sides will receive different light intensity, and then sun position can be determined. It has lower price and convenient installation with simple structure, but its motor rotates frequently and the system has lower tracking accuracy, for it is easily affected by external lights from different directions. Pyramidal type photoelectric sensor makes full use of cosine principle, and the light exposure to silicon photocell of the pyramid side, is equal to the area multiplied by incident light angle cosine. Even though its accuracy is higher than baffled type's, it is still influenced by environmental astigmatism. Tube type has higher tracking precision than others, but the solar incident angle and the height of the tube are conditioned by each other so that it causes the mutual influence between system's tracking stability and tracking precision. Photoelectric sensor, whose structural design has a great influence on the system's tracking accuracy and operational stability, is the detecting element of photoelectric tracking. In order to solve existing sensors' problems, such as single form, low tracking accuracy, bad stability, this paper designs a kind of photoelectric sensor, which is a combination of pyramidal type photoelectric sensor and box type photoelectric sensor, and can track accurately. The pyramid type photoelectric sensor is outside the box type and roughly tracks, while the box type accurately tracks. The concave-mirror, inside the box type, has reflection effect so that it can effectively reduce the box height of the sensor, and the box type and the pyramid type photoelectric sensor can be disassembled to work alone. LM324 comparator, as the core of the system, controls the motor to drive actuator to track sun in the circuit of the sensor. This paper then builds controller models and simulates by MATLAB/Simulink, and the fuzzy PID (proportional integration differentiation) control can let the tracking system return to steady state under disturbance. This control method can achieve optimal control, which means not only the simulation result is intuitional and believable, but also the method can greatly save the cost. The system can quickly return to steady state even in a case of interference, therefore, it has realized tracking feature without error, and also has proved that the fuzzy PID control method can maintain certain performance of the system, and the system has great robustness. Tracking system has better performance with the fuzzy PID controller. In addition, it is simple and continent with Simulink to establish the model and then simulate, and the simulation results are intuitively credible. Simulink can realize the different systems' real simulation better, and then get better simulation results. [ABSTRACT FROM AUTHOR]