Research Of Linear Motor Mover Position Measurement Method Based Image Entropy

Recent years, with computer control related technologies to flourish, the demand of high-precision equipment and highly integrated chip, IC system core, high precision numerical control systems and ultra-precision machining industry have become core technologies of a country. Ultra- precision processing equipment is inseparable from the high-precision servo control system, the permanent magnet synchronous linear motor (PMSLM) avoid necessary transmission links when rotating motor directly drive the workbench, which can directly drive the load and there are the advantages of high speed and high accuracy, which got more and more adoption in the linear feed servo control system. Precise control of the PMSLM depends on the real-time and precise position feedback of motor mover, the traditional mover position measuring method of the PMSLM have many issues such as high cost, complex environmentally sensitive, complex installation etc, which limiting the promotion and development of the PMSLM.Combining the digital image measure algorithms’advantages of good real-time, high-speed, high-precision and non-contact, this research introduced digital image measure technology into the mover position measurement of the PMSLM, proposed an fence image screening method based on the image entropy and calculated the amount of displacement between two fences by the camera captured by the improved digital image correlation function and the simplified Particle Swarm Optimization (PSO) algorithm, and then achieved accurate and real-time measurement of the PMSLM mover position.The main researches of this subject around the PMSLM mover position image measurement algorithms include the following aspects:1. Introduced the currently main mover position measurement methods of PMSLM and analyzed their advantages and disadvantages, described the image measurement technique is widely used in many industries especially in industrial inspection and measurement field etc, which lead to this research’s background and main content.2. Described the currently relevant calibration methods of the image measurement system, including traditional calibration methods, self-calibration as well as the improved CCD camera calibration method, analyzed the above calibration methods’ application field and merits and faults, and then proposed the calibration method of this paper’s image measurement system, introduced the composition of the system and selection of camera lens, finally completed calibration by experiments.3. Studied an improved digital image correlation measurement algorithm combing PSO search algorithm, which achieved image sub-pixel precision measurement. Firstly, this research obtained that digital image phase correlation algorithm is only applicable to micro-distance, whole pixel displacement measurement. So this research selected digital image correlation algorithm, proposed an improved correlation function and the concept of sampling calculation sub-area, and then obtained the image’s sub-pixel displacement value through PSO quickly search algorithm. Finally, performed experimental verification which took the ordinary image as sources image, experimental results show that the measurement accuracy can be up to 0.05 pixels.4. According to the inherent characteristics of the linear motor mover movement, constructed fence striped image as the camera shooting’s source image. Depth study of the impact of digital image different characteristics to the accuracy and robustness of image measurement algorithm, obtained that the more irregular of digital image gray value produces the better measurement results. Analyzed the relationship between Pnatt entropy and image pixels correlation of the fence image, proposed that taking image entropy as the feature of the fence image selection, finally, this research obtained the camera shooting’s source image which has the better non-periodic and robustness by selection.5. Finally, by comparing experimental verification, could known that the image measurement algorithm this research proposed combing the fence image selection based on image entropy has better accuracy, robustness and real-time.So this research carried out the study of PMSLM mover position measurement method based on image entropy could provide a good program and new ideas for the mover position measurement of the PMSLM.