Implementation Of Motion Control Module And Study On Crack Identification Algorithm In The Solar Cell Detection System

Image detection technology was used in measurement for a few decades, but it has been got more attention because of the benefits of multi-disciplinary. Based on optical, the technology combines with high-speed computing, image processing, and laser-electronic, to build measurement detection system. It's widely used in aeronautical remote sensing, micro-size measurement and other related fields in which image processing researches. And industrial detection based on image processing catches images of measured object to extract the information of problem solution. The target of the study on the image detection is that, how to measure and process correctly.Image detection technology extensive applied in the industry is quite mature, and as a computer vision technology of cross-application in the industrial detection, it mainly refers to identify the image in order to obtain some explanation of objects and reduce inconvenience caused by manual detection. At present, the technology has been applied in many occasions, such as ultrasonic and X-ray NDT image processing system, surface of components defect detection, IC weld point locating etc. Not only we can analyze the internal structure of components, predict reliability and failure of the parts function, but also using this technology handle problems of chemical textile flaws and fiber loss.Image processing technology can be totally applied in industrial on-line detection, takes the responsible for monitoring, quality testing and so on. It is quite suitable for solar cell cracks detection, which is traditional done through observing by human's eyes, required no touch.Solar cell array is the only energy supply equipment of satellite, and the solar cells are the mainly components and the most important part of the array. In order to ensure that satellite works well, the array should maintain a high percentage of solar cells working properly to produce electricity. As the photo-electric conversion efficiency is not high actually, we must hold superiority in number to provide enough power output, need to monitor performance of cell's work. Detection in assembly and pre-lauch ensures that the cells are not damaged. And the cell is composed of glass cover sheets and the silicon or other material chips under the cover sheets, usually its production, transportation, installation are very careful, the damage mainly comes to the cover sheets and injury events to the silicon chips will happen rarely. If cover sheets has been broken on the ground, we can replace them conveniently, but when the satellites in the orbit of near-Earth, the solar cells will be fully exposed to the complex environment of cosmic space, which must lead the solar cell's service life reduced. And replacement can't easily be done, that will shorten cycle of satellite life, to make enormous losses. Therefore, the detection of pre-using is quite significant. All of above statements elicit the introduction of the solar cells detection system which replace the pure manual work for cell defects detection.The satellite solar cell defect detection obtains the images of cells first by operations of image processing technology, to extract the information of features to identify the defects, and submit a report for manual verification. According to the technical requirement of use, coupled with solar cell array structural characteristics and detection standards, this article introduces every module's design of the detection system clearly, special on implementation of the motion control and improvements of defects identification of image processing module.To show the whole impression of the system clearly from outside to the inside, this article began to explain the detection system with its function.The first part is the overall design of the system, which consists of three-dimensional rails for the movement and photography camera, to achieve automatic detection of the mechanical part. Because a separate cell is small, and it's very easy to be broken, these chips'detection has a quite stringent and accuracy requirement.No matter what movement rails, electrical circuit boards or its control procedures, all considerate the requirement of precision and safety. System is supported with suitable equipment and ancillary programs, to capture images and protect the cells array.Since then, the system need to be designed and the corresponding software modules should be written, which is divided into two main parts of image processing and motion control. The prior image processing has been a overall process design and basic processing operations, including image acquisition, peak binarization, removing of border and gate lines, cracks extraction, etc. Rail movement control and laser ranging are the major part of the motion control.The improvement of image processing focuses on the identification about cracks and noise of background, and now the image only has the segment cracks divided by the gate line and some noise group. Using morphology thining method to extract the skeleton of contours, the system takes the differences between skeletons and original contours as the feature, which is coupled with previous features to form a feature vector. Finally, to identify the segments of cracks from the interference noises uses the classifier based on support vector machine to make prediction.The design and implementation of the motion control obey the need of system design with discipline of simple efficient and facilitative maintenance, depending on precision of hardware. For example, there is a need for parsing and using the AutoCAD files, extracting the interest information which the motion control requests, considering the errors in the conversion. To the special application, transmission mode of laser ranging sensor adopts the normal one called stop-wait mode. The rail control is completely resolved in accordance with map information extracted from the AutoCAD file. The motion makes the snake-like trace move with details of continuous movement in vertical direction and moving the interval distance which is computed on-line in the horizontal direction between columns. At the same time getting the end of the information of chips position, the peripheral equipment reaches the corresponding location.