Waste water treatment monitoring method and system based on fuzzy neural network

Abstract

The invention provides a waste water treatment monitoring method based on a fuzzy neural network. According to the method, global search optimization is performed on the weight value and the threshold value of the neural network by applying a neural network structure algorithm and a real number encoding-based genetic algorithm on the basis of inspecting the influences of parameters such as the HRT (Hydraulic Retention Time) of an A/O (Anaerobic/Oxic) waste water treatment system, the DO (Dissolved Oxygen) in an aerobic pond and the reflux ratio r of mixed liquid on the COD (Chemical Oxygen Demand) and ammonia nitrogen of effluent water, so that an optimal BP-ANN (Back-Propagation Artificial Neural Network) soft measurement model and a GA-ANN (Genetic Algorithm-Artificial Neural Network) soft measurement model are built respectively. An intelligent dissolved oxygen optimization control system scheme based on the soft measurement models is built on the basis, data communication between the MATLAB (Matrix Laboratory) and MCGS (Microwave Command Guidance System) is realized in combination with an OPC (OLE for Process Control) technology, and an intelligent dissolved oxygen optimal control system is applied to dissolved oxygen optimal control in a paper-making waste water A/O technical treating process, so that energy saving and consumption reduction in paper-making waste water A/O treating engineering is realized. By adopting the waste water treatment monitoring method, the stability of the water quality of waste water discharged in the paper making industry can be increased greatly, accurate control is realized, and the chemical drug dosage and power consumption in a waste water treatment process are reduced.

Description

technical field [0001] The invention belongs to the technical field of comprehensive and stable treatment of papermaking waste water by using an intelligent control system. Background technique [0002] With the increasing pressure of national, provincial and municipal environmental protection policies, and the need of various paper-making enterprises to reduce production costs, while the water consumption per ton of paper is declining, the concentration of drainage is increasing, the quality of drainage water is deteriorating, and the fluctuation of water quality is also relatively high. large, which has brought severe pressure to sewage treatment. In addition, due to the different varieties produced by each paper mill, the chemical raw materials used are also different, and the production schedules are also arranged, resulting in large fluctuations in the influent water quality of the centralized sewage treatment plant, influent water flow, influent COD and The amount of ...

AI Technical Summary

Problems solved by technology

However, the gradient training algorithm used is a local search algorithm, which is easy to fall into a local optimum; and has the disadvantages of high inspection accuracy but low prediction accuracy, and long network training time.

Images