Research And Development Of Assistant Decision-making System For Coal-fired Boiler Production Process Optimization

Coal-fired boiler process industry is a kind of process industry of thermal power plant,with the production process quite complicatedly.A large amount of data produced in the process of industrial production is collected into computer and stored in the storage medium as historical data.In addition,valuable information that is more conducive to industrial production can be obtained through data mining.Therefore,this paper optimized and auxiliary analyzed the production process of coal-fired boilers and designed a computational model for discovering flow rules of coal-fired boilers.The main contents of the computing model include data preprocessing,timing discovery,association chain discovery and modeling operation.Spark framework is utilized to parallelize the approaches of each part,thus improving the efficiency of data mining.Finally,an auxiliary decision-making system for production process optimization of coal-fired boilers based on B/S network structure is designed and deployed in a thermal power plant in Jinan.In the boiler production process,the historical data is produced by each production link.The link refers to the key attribute parameters in the flow of production.We regard each attribute parameter as a link.Firstly,the low quality data in the computer are integrated and the accuracy of the data is checked.Secondly,in addition to preprocessing work,it is necessary to find the timing relationship between each link and find the order between links.Among them,the algorithm based on statistical extreme value is adopted to realize the discovery and adjustment of Time Series of data.By setting the reference link and using extreme value to calculate the Time distance between other links and this link,the Time Series before and after the link can be obtained.Then,using the preprocessed data,the data of a certain heating season is processed by link clustering.We adopt the classical k-means clustering algorithm to cluster each link separately and the optimal K value is obtained by contour coefficient method.And then finally,through the above processing,also needs to carry on the data mining to the data.In this paper,Apriori algorithm is employed to calculate the association rules between clusters for the data after clustering.The calculation results are only association rules between clusters,which need to be converted into binomial association rules between links.The relationship betweentwo links is obtained through calculation,and the final correlation chain is obtained by using the sequence of time series discovery.Association chain is a chain composed of links which is directly related to each other.Finding the chain with the strongest association strength is through the association rules,and the like,the other association chains are mined and discovered.By the processing of historical data,high quality data and link correlation chain are obtained,and sufficient data preparation is made for generating model.In this paper,a prediction model based on process data of coal-fired boilers is designed.The flexible neural tree algorithm is adopted in the model,which is a model superior to the neural network.This model has higher efficiency and accuracy than the common artificial neural network model in solving classification and prediction problems.By modeling to obtain the trend function of the data,so as to predict and simulate the key parameters in the boiler production process,we can get the regularity knowledge implied in the data.Experimental results show that the model can well predict the parameter values of each link.At the same time,this system uses the B/S-based network structure mode,and finally the optimization auxiliary decision-making system applied to the production of coal-fired boilers is getting.The system is deployed in a thermal power plant in Jinan,and the key parameters of the boiler are adjusted.Through the above design,the production process of coal-fired boilers is optimized,the energy saving and emission reduction in the production process are realized,and the safety of boiler production is improved.