System and method for AI controlling waste-water treatment by neural network and back-propagation algorithm

摘要

PURPOSE: A system and method for controlling a sewage and waste water processing artificial intelligence by a neural net and a back-propagation algorithm is provided to draw a correlation using conventional data for considering an operating factor according to seasonal variations, suggest a control method for increasing a processing efficiency through an SRT(Sludge Retention Time) control, and control an operation according to seasonal variations. CONSTITUTION: If a program is started and an initialization is executed(201), conventional data are collected(202), the data are filtered(203), a measured characteristic of flowed water, a state of a responding tank, and a characteristic of outflow water are received(205), and the current state is checked(204). Control output calculations(207,210,211) of the checked current value and an SP input value(206) are executed, and the results are controlled and output(209,212) as an analogue and a digital signal and stored in a memory of a PLC(Programmable Logic Controller)(218). If the efficiency is proper(214), the result is changed into a physical value(217) and stored according to data dates(221) and the initial value is stored(219). Also, the result is displayed on a computer screen through a two way communication using a TCP/IP through an RS-232C communication unit(220).