Design And Implementation Of Expert System For Heavy Medium Coal Preparation Based On Measured Parameters

As the industry 4.0 coming,heavy medium coal preparation automation develops into a more intelligent direction with the features of strong coupling of related knowledge and being hard to establish the precise model between parameters.a main direction of coal preparation development of the Thirteenth Five Year Plan is to achieve intelligent by using expert system principles,constructing expert knowledge base and knowledge rules to build expert consultation system.The way of this intelligent control has been gradually infiltrated into the field of mineral processing.In Taixi Coal Preparation Plant of Shenhua Ningxia Coal industry group Co.Ltd,the research group broke the bottleneck of the development of automatic control of heavy medium coal preparation based on image processing technology of rapid analysis.Relying on the technology,we have builded the heavy medium cyclone expert consultation system,and completed the feedforward of heavy medium coal preparation control.However,since the establishment of expert knowledge base was not mature enough,the system can not form a complete closed-loop control.It is urgent to integrate the knowledge of heavy medium coal preparation and improve the knowledge base to achieve the overall closed-loop control system.The paper mainly studies how to improve knowledge base and then constructs heavy medium coal preparation expert consultation system.Focusing on improvement of knowledge base and demand of constructing heavy coal medium preparation expert system,the paper firstly describes the heavy medium process in Taixi Coal Preparation Plant and introduces the technology of rapid analysis of coal quality based on image processing,analyzes the advantages and disadvantages of the existing h.m.cyclones consultation system.Secondly,the paper analyzed the knowledge sources and storage of the main operation parameters of heavy medium coal preparation process,collected and builded the limit control knowledge base of operation parameters,combined the heavy medium coal preparation key knowledge with the existing model to forecast the sorting effect,ensured the coal ash constant and Ep value as small as possible,realized the optimization of some operating parameters by dichotomy and repeated calculation method,the original collection and knowledge extraction of fuzzy operation experience data by manual collection method,explore to use self-learning to obtain the optimal interval of operating parameters in the requirements of the coal ash and the corresponding adjustment of the operation,and then expand the knowledge base.Based on the characteristics of heavy medium coal preparation,combined with the expression style of produciton rules,the group designed the storage structure of the knowledge about heavy medium coal preparation,compiled the corresponding coal preparation rules by using CLIPS combined with the parameters and fuzzy limit control experience and knowledge.Finally,the group used the three layer model of the.NET framework to develop the expert system,using SQL Server 2008 to build the knowledge base,object oriented language C# to develop the main program.The expert system contains heavy medium coal preparation knowledge base management subsystem,heavy medium coal preparation for and diagnosis subsystem and user login and management subsystem.The prediction and diagnosis subsystem can call KXXQX.dll files to draw the optional curve according to the measured coal quality information,can also according to the operation parameters for parameters limit control,separation effect prediction,heavy medium separation of coal parameters,and get some adjustment operation suggestions through the matching between the existing data facts and CLIPS rules.The knowledge management subsystem realized the functions of addition deletion quering and modification of knowledge,the accumulation of fuzzy data experience and the optimal interval self-learning of fuzzy knowledge.The user login and management subsystem realized the authentication of user login information as well as the user information registration.