Design And Development Of Software Platform For Ball Mill Load Monitoring

Ball mill is the key equipment in mineral grinding processing. The ball mill load including the volume of the ball, ores, as well as water is an important parameter in the process of ore grinding, which is connected with the grinding efficiency and energy consumption. Its monitoring methods becomes a difficult problem in ball mill system automation. Most of existing applications are obtaining the value of load indirectly by soft sensor methods through the vibration and grinding noise while ball mill running. The main realization of ball mill load monitoring modes are through packaging those methods into embed instrumentations or developing software toolkits. However, those modes are excessively structural, method fixed, with lower rate of reusablity. In order to help the researchers foucs on soft sensor algorithms development and research those methods on a general platform by experiments, it is necessary to do the research on ball mill load monitoring software platform with modular structure and functions of soft sensor algorithms experiments,with which the researchers. Supported by "863Project" Research Topic—"Research on semi-autogenous/autogenous ball mill monitoring technology", ball mill load monitoring system is designed based on virtual instruments. The primary purpose of the paper is to design and develop a ball mill load monitoring software paltform. The detailed works have been summarized as follows:Constructed the general framwork of ball mill load monitoring system based on virtual instrument. Biuld up the PC-DAQ hardware platform and combine all parts organically. We organize the structure of ball mill load monitoring software platform with the purpose of clarity and reusability.In order to achieve ease of use and scalability, we make the software function modularized according to the requirement of ball mill load soft sensor. The software is divided into five modulars:signal acquisition modular for obtaining the monitoring variable; data storage modular for storaging data; pre-processing modular for signal processing; offline modeling modular for soft sensing model trainning; online monitoring modular for ball mill load soft-sensing.Within LabVIEW development environment, software developing is done in accordance with the design. We use DAQmx and DataSocket technology in signal acquisition modular, use LabSQL toolkit to achive database operation in data storage modular. Those processing methods are packaged into subVIs in signal pre-processing modular, and managed by using tree structure and menu. RBF and LS-SVM algorithms are packaged into subVIs with call interface in offline modeling modular. By using realtime data samples, online monitoring modular call the soft-sensing model to gain the ball mill load.System testing and related experimental verification are made on the ball mill load monitoring software platform. Test case is designed for each modular. Data acquistion experiments and offline modeling experiments are designed for the whole software platform verification. The results prove the effectiveness of the design and the feasibility of the developments.