The Research And Development Of Coal Slime Flotation Medicine Control System

Flotation is an important a high energy-consumption link of coal preparation. During the coal slime flotation process, there are many variables which changes greatly. And with more and stricter requirements on the quality of the market, slime flotation process automation level has attracted more and more attention. Therefore, the flotation automation is an inevitable trend of development. However, there are some problems in the process of realization of coal flotation automation, which has become a stumbling block. First,the field lacks operational parameters detecting device and detection means; secondly, digital image processing technology has not been applied to the flotation; What’s more, although the control theory of the flotation process has been deeply studied, further research is needed to use the control strategy in production practice. Therefore, this paper carry out the research from the three aspects such as the coal slime flotation process variable detection, digital image processing and the establishment of flotation dosing control system.(1)The coal slime flotation dosing control expert system is designed based on the study of coal slime flotation process. The control architecture is determined by disturbance variable, which contains the flow rate, the concentration of the float, the bubble size, the gray level of the bubble and the ash content of the tailings, and operating variables like the adding amount of foaming agent and collecting agent, and control variable of ash and bubble size of tailings. The basic function of the system and the structure of the software in the system is defined by the field environment and the operation habit of the workers.(2)A coal slime flotation tailings ash soft measurement model based on the combination of image processing and neural network algorithm is put forward. On the basis of extracting the gray histogram of the tailings image, the BP neural network algorithm, ELM algorithm and I-ELM algorithm are carried out to predict the ash content. The experimental results show that I-ELM has the advantages such as high accuracy and short training time, and it is suitable for the detection of the ash content in tailings.(3)Bubble image acquisition and bubble image processing interface is developed in virtue of the visual interface development kit of Matlab. Firstly, the collected bubble image is filtered and enhanced, then the bubble image is segmented by watershed algorithm with sign, the extreme bubble is removed and the average size of the bubble is obtained. Meanwhile, based on the gray histogram, the gray features are extracted, and the correlation analysis of extracted feature and fine coal ash content is done.The image acquisition interface can set the parameters of industrial camera, and store the image, what’s more, by which the time interval of the collected image according to the actual situation of the field can be set; the image processing interface can transform, filter and enhance the image. The bubble image can be segmented based on image processing to get the size of bubble. And parameters such as Gray mean, variance, smoothness, skewness, energy and entropy can be extracted based on the gray histogram.(4)Floatation dosing control expert system is designed according to the design method of the expert system. Firstly, according to the input of the bubble size, gray level, tailings ash content and other main features, a flotation reagent adding rule library based on fuzzy inference is estabilished. The dosage of crude flotation reagent is adjusted by fixed value control method, the dosage compensation of reagent is obtained through fuzzy inference method according to the flotation state characteristic quantity, which completes the compensation of crude adjustement. The inference strategy is estabilished and the design of inference engine is completed. The Kingview interface is designed which is used as the human machine interface of expert system, the data base is designed as the comprehensive databse of expert system, which can realize management, inquire and storage.(5)The hardware and software platform of the Coal flotation automatic control system is established on the basis of expert system and image analysis; the simulation operation of slime flotation dosing control system is realized; using communication protocol to realize the seamless connection of the control system internal data. Kingview and Matlab communication protocol is established on the basis of OPC protocol, which allows the data transferring between Matlab and the host computer. Through the DDE technology, the combination of Kingview and inference engine has been realized; use the SQL access function of King view to visit the database, and the operation results show that data transmission is stable, which achieves data sharing and the expert system reasoning correctly.