Research On Intelligent Classification System Of Micro Powder Based On Microscopic Image Technology

Abrasive powder is a necessary material for grinding,grinding and other finishing processes,and its quality directly affects the precision of product finishing.At present,the production of abrasive micro powder has already entered the automation line stage,often using the overflow wet classification equipment to produce.The control mode of this kind of equipment is based on the time based open loop control method.The product quality and efficiency depend on the experience of the expert,and with the environment temperature,humidity,air pressure and so on The quality of the products produced by the same process will fluctuate greatly.In order to solve these problems,a micro powder Intelligent Grading System Based on microscopic image technology has been built in this paper,and the key technologies have been studied deeply and systematically.In this paper,the structure and function of the traditional overflow wet classification equipment are improved,and the micro powder intelligent classification system based on microscopic image technology is designed,and the overall scheme of the system is given.On the basis of the original system,the micropowder characteristic parameter detection platform and automatic control module are added.Among them,the micropowder feature parameter detection platform includes microscopic image acquisition module,auto focusing module,image analysis module and motion control module.In this paper,the autofocus technology of microscopic image is studied,and the autofocus of the micropowder characteristic parameter detection platform is realized.In order to solve the problem that the anti-interference ability of the definition evaluation function of the traditional square gradient method is low,a definition evaluation function based on the improved square gradient is proposed,which solves the problem of the poor anti-interference ability of the definition evaluation function by the square gradient method.In this paper,the focus search algorithm is studied,and an improved mountain search method is proposed,which reduces the influence of the local maximum in the auto focus process and improves the accuracy and efficiency of auto focus.In this paper,the micro image detection technology of abrasive micro powder is studied,and the micro powder image is analyzed and processed.The automatic calculation and statistics of the characteristics parameters such as the circumference,area,particle size,roundness and ellipticity of the abrasive particles are realized.It provides a reliable basis for the design and adjustment of the technological parameters of the grinding micro powder grading equipment.The control technology of intelligent classification system for micro powder is studied in this paper.STM32F103ZET6 is chosen as the master chip of the controller.In the VS2010 development environment,the Modbus protocol host software of the upper computer is developed through C# language.Under the MDK5 development environment,the slave protocol software of Modbus protocol of STM32 controller is developed through C language,and communication with host computer is realized.In this paper,the micro powder characteristic parameter detection platform is set up.The software and hardware of the micro powder characteristic parameter detection platform are designed in detail.The experimental results of the micro powder characteristic parameters measured by the micro powder characteristic parameter detection platform and the laser particle size analyzer are compared,and the correctness of the method is verified.