Research On Intelligent Early Warning Based On Fermentation Process Control

For exception handling in industrial control system only does alarm,early warningresearch less before the failure occurs. The precursor information before failure occurs isneglected.Based on PLC fermentation process control environment,we propose fuzzy decisionscale adjustment LS-SVR early-warning model,we studied intelligent early warningtechnology in industrial control system and its application.The current industrial control systems expands unceasingly.Early warning research andimplementation on technical feasibility is higher and higher.Fault problem is a process ofslow device status changes or emergency happens suddenly.Early warning technologyanalyses system state in order to avoid accident and improve safety,predict potential hiddentrouble in time.Early warning plays a key role in the real-time control and safe operation.This paper mainly introduces the following content:1. First of all, based on biological fermentation process temperature, pressure and othercontrol and characteristics of equipment, we give intelligent warning system theoryframework. Mainly includes the process of heuristic knowledge and process data, heuristicknowledge is given by the expert experience in equipment operation knowledge to guide thesystem status warning; Process data processing based on historical data from different process,equipment, components, then regression algorithm training and predict the system state.2. Secondly, on the basis of the theory of early warning system framework, then researchfor effective intelligent prediction algorithm, based on early warning system for real-time,completeness, and correctness, multi-scale LS-SVR forecasting fuzzy decision algorithm isproposed.Then algorithm accuracy is validated.The feasibility of algorithm is verified.3.Combined with biological fermentation technology environment, we design fermentation intelligent early warning monitoring system solutions, including upper unit statedesign, intelligent algorithm, PLC control equipment, etc. Then it realizes the real-timemonitoring of all biological control data.Based on a biological fermentation process control of a factory in ShanDong province,designed fuzzy decision multi-scale LS-SVR prediction algorithm of early warning system,and implement the solution of hardware and software structures. Early warning systemprovides certain reference value for certain applications of specific areas, has certain practicalsignificance. Paper finally has carried on the summary and prospect are put forward.