Wastewater Quality Prediction System Based On Soft Sensor

Sewage treament is a major research project which has great significance in environment protection recently. Whether the sewage is qualified to be descharged used to be based on the comparision of water qualifty parameters detected by some sewage treatment factories with national standards. here are many water quality parameters, most of which are using sensors to measure directly with a processor, such as pH, temperature and other parameters. But there is still some data can not be obtained in this way, they can only be in the laboratory through the water samples cultured for a period of time to get, which obviously causes a time delay in sewage discharge. This paper focuses on how to measure or detect through the indirect way to improve the performance of the sewage treatment.Since the BOD parameter test takes a long time to get the results, this paper used an indirect BOD measured way to get the real-time monitoring of the sewage discharge. Traditional laboratory test is also one of the indirect modeling ways which measures sewage discharge by mechanism features. Based on these, we built and optimized the BOD model using nural network apprach with data which could be quickly acquired to improve the accuracy of the model system.This paper used the standard genetic algorithm and improved genetic algorithm to get data optimized respectively, which in turn determines the validity of the algorithm.Indirect sensor research is a new way to substitue the traditional physical sentor. The new indirect sensor we designed in this paper which conbined by both advantages of virtual instruments LABBIEW and MATLAB could replace the conventional laboratory equipments and have more applications in the future.