Design On Control System Of Scale And Clean Semi Coke Produce And Research On Prediction Model Of Charring Coke

The low metamorphic coals are abundant in North Shaanxi and have a high output.The main way to process local produced coal is carbonization at lower temperature withsemi coke and so on as the products. In pace with the escalate of industry on carbonizechalk, the technology on scale and clean turn into the latest developing and improvingdirection on produce of carbonization at lower temperature with semi coke. Thedeveloping of carbonization at lower temperature technology not only improve thebeneficial result but also have benefit of the rational use on resources utilize.Put the composite produce of semi coke project from one large-scale cokingenterprises in northern Shaanxi as research object. Choose the CENTUM CS3000ofYOKOGAWA as control system by general requirements of system was analysis. Thehardware and software of CS3000system was studied with the FCS、NODE UNIT andconfiguration as emphasis.Six work part and one work area by study its technology process was divided. Thecontrol points was designed according to the system partition. The hardware equipmentof control system by analyze control points in total was selected. A complete andfeasible install plan of control system by design rule of hardware was putted forward.The configurations of seven sections were used to simulate the whole system.Existing research achievement of coker was referred, the main equipment ofproduce of carbonization at lower temperature with semi coke was studied. The modelfor prediction of productivity based on the Fuzzy Neural Network, which had the abilityof fuzzy processing and self-learning characters, was selected. Then a self-organizingalgorithm is applied to optimize the model, which had the superiority of clusteringhandle capability. The error were analyzed after the training and testing simulation for the two models for prediction of coke productity by MATLAB. Simulation result showthat the prediction model of coke for productivity based on had self-organizingalgorithm for fuzzy neural networks is better than the normal. Prediction model ofcharring coke for productivity based on self-organizing algorithm for fuzzy neuralnetworks is trusted.