Research On Method For Feed Physical Properties Testing And Crusher Screen Breakage Recognition Using Linear Array Optical Imaging Technology

With the rapid development of modern feed industry, the market for the processing quality pellet feed product requirements increasingly stringent, the means testing of feed processing in the production is particularly important. Physical properties of pellets is an important parameter of the quality of feed processing, In our country, the physical characteristics of the pellet feed mainly adopt artificial sampling inspection,such as uniformity, density and size of the particles, etc., these methods are tedious, time-consuming, low level of automation, it has been unable to meet the development needs of modern feed industry. Therefore, we use linear optical imaging techniques to build the system which can detect the physical parameters of pellet feed and identified the screen breakage in the crusher, proposed the uniformity, density and granularity of the pellet feed and screen breakage recognition methods, to achieve real-time automatic detection of the physical parameters of feed particle and the crusher’s screen breakage fast automatic recognition. The results for the pellets of the basic physical parameters of real-time online detection and automatic identification of screen breakage of crusher provides a new way of thinking and methods. The main contents are as follows:1. Based on linear optical imaging technology to build the physical parameters of the pellets and crusher screen breakage recognition dynamic detection system. System mainly consists of vibrating feeder, linear CCD camera, linear light source, computer, etc., on the feed module, transport module, image acquisition module, serial communication module and other components for the selection, and adjust the key parameters of each component.. We developed a virtual instrument by Labview which include bonded particulate segmentation, image stitching, particle feature extraction, balance serial data extraction, data storage and dynamic real-time processing module software.2. The use of a detection system constructed to extract the long diameter, short diameter, perimeter, area, and other parameters of pellet feed, the establishment of the particle size, uniformity and density detection methods to achieve particle feed size distribution, uniformity, density, etc. dynamic detection of physical parameters. Compared with manual measurement, the detection system of the mean absolute error feed particle size distribution range and the maximum relative deviation of 0.63% and 2.62%, mean absolute percentage error and the maximum relative deviation for length uniformity were 3.92% and 4.86%, the average density of the absolute percentage error and the maximum relative deviation were 0.43% and 3.23%. The results show that the system for detecting particles feed samples with good accuracy and repeatability.3. To corn and wheat grain raw materials as experimental samples to determine the damaged and undamaged screen crushing material mean area and the coefficient of variation of the area threshold. In the screen aperture were 2mm, 4mm, 6mm conditions, the mean area of the particles is the distinguishing index, crushed corn sieve breakage average recognition rate was 91.25%, 82.50%, 96.25%, crushed wheat screen breakage average recognition rates were 93.75%, 86.25%, 78.75%; coefficient of variation in particle area judging indexes, crushed corn average recognition rate was 93.75%, 98.75%, 97.5%, crushed wheat screen breakage average recognition rate of 93.75%, respectively, 87.5%,96.25%. The results showed that the coefficient of variation as determined by area to identify the effect of discriminating index screen breakage better than average recognition performance area.Through the development linear optical imaging system of physical properties testing of pellet feed and crusher screen breakage recognition, the establishment of pellets granulometry, uniformity and density of the detection method, the detection accuracy can be used in actual production. At the same time, the constructed system can achieve automatic recognition of broken crusher screen automatic identification of high recognition accuracy. The results for the realization of pellet feed physical characteristic parameters of real-time online detection and crusher screen damage automatic identification provides a new idea and method.