Design And Development Of Trend Monitoring And Analysis System For Beneficiation Production Index

In mineral processing enterprises,it is necessary to measure production status through production indicators.In the production process,we want to ensure that the index trend is in the optimal state,keep the trend stable,and try to avoid abnormal or fault conditions.Therefore,it is particularly important to monitor and analyze the index trend of mineral processing enterprises.Monitoring and analyzing the trend of indicators can enable the enterprise management personnel to grasp the production status in real time,find problems before the occurrence of abnormalities or faults,make corresponding decisions in time,try to avoid the loss of beneficiation production,guide the efficient operation of production by means of information technology,and improve the efficiency and efficiency of production.At present,most of the ore dressing index monitoring systems achieve the collection of index data,present the collected real-time data to enterprise engineers and some other functions,and realize the index monitoring.However,the existing indicator monitoring system lacks effective data analysis methods.Engineers’judgment on whether the production indicators are abnormal,the future trend of the production indicators,and whether the current production is in the optimal state mainly depends on their own experience.In addition,the analysis and tracing of the abnormal indicators are also completely dependent on the experience of processing experts,so the monitoring system cannot provide corresponding support to users to make decisions.In view of the above problems,the research of the trend monitoring and analysis approach is carried out and the related system is developed.The main work of this paper is as follows:(1)Combined with the production process,the paper focus on analyzing the production indicators,determining the target of this paper,and analyzing the demand of the trend monitoring and analysis system of production indicators.The trend monitoring and analysis system of production indicators needs to extract features from data through similarity measurement,combine feature library with data evaluation model,and carry out functional requirements such as condition analysis,and meet performance requirements such as system stability.(2)According to the production demand of the concentrator,the single dimension index and multi-dimension index data characteristic analysis methods based on the similarity detection algorithm are designed respectively.Because of the actual production and analysis demand of the concentrator,the algorithm is improved in combination with the practical application background,and the similarity measurement algorithm suitable for multi-dimension index is proposed.The experimental results of the improved algorithms and the original algorithm are compared,The best of them,made by PCA-DTW algorithm,is obtained.This paper designs a real-time trend monitoring and evaluation method based on similarity detection algorithm and short-term memory network,and uses the actual data of the concentrator for experiment,compares the effect of the two real-time trend monitoring and evaluation methods,and the system uses a more accurate method for real-time trend monitoring and analysis.(3)The main work of also including designing the software architecture,function module,interface and other aspects of the trend monitoring and analysis system of production indicators,and developing the system in combination with the actual application background and actual operation environment of the beneficiation plant.The trend monitoring and analysis system of production index includes index selection module,data selection module to be analyzed,data characteristic analysis module,characteristic sample management module,real-time trend monitoring and evaluation module,and index abnormal mode analysis module.In the data feature analysis module,the similarity detection method based on similarity measurement algorithm is used to extract the similar features of the data,mining the potential relationship and rules between the data,so that the data can play its maximum value;the feature sample library is the basis for real-time trend evaluation;the real-time trend monitoring and evaluation method will collect the real-time data through the LSTM model,and the real-time data will be Operating condition classification,identify whether there is any abnormality,provide trend monitoring and evaluation tools for enterprise production management personnel;index abnormal mode analysis module provides enterprise management personnel with the analysis of indicators exceeding the quantity value.The above modules constitute a complete index trend monitoring and analysis system,so as to assist enterprise managers to make production decisions.(4)The monitoring and analysis system of ore dressing production index trend is applied and verified on the spot.The actual application results show that the system completely realizes the monitoring and analysis of the whole process production index including data characteristic analysis and sample library construction,meets the actual demand of the monitoring and analysis function of ore dressing plant index trend,and enables the enterprise management personnel to better carry out the monitoring and analysis according to the system information Production adjustment,so that the enterprise production efficiency,safe operation,reduce losses.