选矿生产指标可视化监控平台研究

Research on a mineral production index visual monitoring platform

针对选矿工业过程流程长、工序多、生产指标多的特点,结合数据可视化及可视分析技术,开发一种选矿生产指标可视化监控组态平台。该平台包括生产工艺可视化组态设计环境、生产指标监控和可视及可视分析3个工具。组态设计环境支持以组态方式绘制工艺流程图,并能通过可视界面自定义生产工序的输入输出指标、触发事件、约束规则、工序状态属性、提示信息等;支持集成专家知识、经验和规则以实现基于知识的生产指标监控;提供算法配置接口,方便集成指标监控算法。其组态出的每一个工序都可以复用和扩展,可以构建选矿行业的基础工序单元组件库,形成选矿行业的知识积累。基于可视的工艺流程,利用指标监控工具为指标配置可视化方案,实现指标可视化监控。可视化方案由可视及可视分析工具提供,包括实时数据、历史数据及其统计特性、多指标综合对比分析、指标关联关系分析、多视图等可视方案。此外,为了提升工序指标监控效率和减轻操作人员监控强度,系统提供因子分析、Pearson相关分析、互信息、信息熵等分析手段,以辅助人们提取出工序关键监控指标,从而实现对监控生产指标的约简。由于该平台以组态方式提供,使其可以快速应用于其他流程行业,实现生产指标可视化监控的组态化。最后,该系统作为选矿生产执行系统的一部分成功应用到某选矿厂生产过程中,取得了良好的应用效果。

According to the characteristics of the mineral processing industry of long production line,multi workshops,multi production indices,combined with the visualization and visual analysis technology,a configuration platform of visual monitoring system for mineral production index is proposed.The platform is composed of three tools:visual configuration design environment,production index monitoring and visualization&visual analysis.The configuration design environment supports the process flow chart by configuration,and can customize the input and output indices,trigger events,constraint rules,process status attributes,prompt information,etc;supporting to integrate expert knowledge,experience and rules to achieve knowledge-based production indices monitoring;providing an algorithm configuration interface to facilitate integration of the index monitoring algorithms.The configuration of each process can be reused and extended,we can build the base of the mineral processing process unit module library to form mineral processing industry knowledge accumulation.The index monitoring tool is used to configure the visualization scheme for the indices to realize the visual monitoring of the index based on the visual process flow.Visual solutions are provided by visual and visual analysis tools,including real-time data,historical data and its statistical characteristics,multi-index comprehensive comparative analysis,correlation analysis of indices,multi-view and so on.In addition,in order to improve the monitoring efficiency of process indices and reduce the intensity of operator monitoring,the system provides analysis methods such as factor analysis,Pearson analysis,mutual information and information entropy to assist people to extract the key monitoring indices of process to realize the reduction of monitoring production index.Finally,as part of MES,the system has been successfully applied to a concentrator mineral production process and achieved good effect.