Based on Improved Artificial Neural Network Sewage Monitoring Alarm System Method

Sewage discharge has become a key issue affecting the quality of water environment, and how to effectively monitor manage sewage behavior factor avoid pollution improve quality. However, current domestic monitoring system is not perfect, resulting in lack effective enterprise by regulatory authorities, which provides an opportunity for enterprises steal discharge. In background treatment plant, comprehensive design alarm carried out based on idea physical information fusion. The adopts four-layer architecture, divided into four parts: perception communication, fusion processing, push, execution. part, neural network intelligent algorithm used predict dissolved oxygen, oxygen delivery adjusted according predicted value achieve accurate aeration optimize effluent push execution parts adopt multiparameter realize smooth operation equipment ensure security. A new optimal control strategy proposed. Through large number experiments historical data, intake index tank under condition outlet are obtained as samples. According sample training, BP optimized particle swarm optimization adopted prediction different inlet conditions. accomplished lower machine upper machine. treatment, each process section collects status strict accordance with order facilities. Then communication between computer sensor designed. PLC core, programming through STEP7, uses PID oxygen. PC developed C language, so user login, real-time data display, over-limit fault alarm, report query, management, etc. integrates MATLAB platform mixed quantity. improved artificial sensitive excellent performance. It detection monitoring.