Method and system for wastewater treatment based on dissolved oxygen control by fuzzy neural network

Patent metadata

Field	Value
Publication number	US-9747544-B2
Application number	US-201113985482-A
Country	US
Kind code	B2
Filing date	Sep 22, 2011
Priority date	Feb 14, 2011
Publication date	Aug 29, 2017
Grant date	Aug 29, 2017

Abstract

A method and system for wastewater treatment based on dissolved oxygen control by a fuzzy neural network, the method for wastewater treatment comprising the following steps: (1) measuring art inlet water flow rate, an ORP value in an anaerobic tank, a DO value in an aerobic tank, an inlet water COD value, and an actual outlet water COD value; (2) collecting the measured sample data and sending them via a computer to a COD fuzzy neural network predictive model, so as to establish an outlet water COD predicted value, (3) comparing the outlet COD predicted value with the outlet water COD set value, so as to obtain an error and an error change rate, and using them as two input variables to adjust a suitable dissolved oxygen concentration. Accordingly, the on-line prediction and real-time control of dissolved oxygen wastewater treatment are achieved. The accurate control of dissolved oxygen concentration by the present method for wastewater treatment can achieve a saving in energy consumption while ensuring stable running of the sewage treatment system, and the outlet water quality meets the national emission standards.

First claim

What is claimed is:

1. A method of wastewater treatment based on control of dissolved oxygen using a fuzzy neural network, the method comprising:
measuring an inflow flowrate, an oxidation-reduction potential (ORP) value in an anaerobic tank corresponding to real-time aeration quantity, a dissolved oxygen (DO) value in an aerobic tank corresponding to real-time aeration quantity, an influent chemical oxygen demand (COD) value, and an actual effluent COD value in an anaerobic/oxic (A/O) wastewater treatment process;
collecting the measured data, sending the data via a computer to a COD fuzzy neural network predictive model, and computing as physical quantities, so as to establish an effluent COD predicted value;
obtaining an error and an error change rate of the effluent COD value;
comparing the effluent COD predicted value with an effluent COD set value;
determining an effluent COD error or change amount and an effluent COD error change rate;
inputting the effluent COD error or change amount and the effluent COD error change rate as two input variables to a DO fuzzy neural network control model, and using the DO fuzzy neural network control model to determine a correction amount of aeration quantity, thus obtaining a corrected real-time aeration quantity;
controlling an air blower to obtain a suitable dissolved oxygen concentration by a control system according to the corrected real-time aeration quantity;
inputting the corrected aeration quantity as an input of the COD fuzzy neural network model;
obtaining a second effluent COD predicted value using the COD fuzzy neural network predictive model using the corrected aeration quantity; and
repeating the same steps recited above in subsequent cycles, thus providing a method of wastewater treatment based on control of dissolved oxygen using a fuzzy neural network using an on-line prediction and real-time control of dissolved oxygen in the wastewater treatment method.

2. The method according to claim 1 , wherein the COD fuzzy neural network predictive model includes an input layer, a hidden layer, and an output layer, wherein the hidden layer is further divided into three layers: a fuzzification input layer, a rules layer, and a fuzzification output layer that comprises fuzzification, fuzzy inference, and defuzzification.

3. The method according to claim 1 , wherein the architecture of the COD fuzzy neural network predictive model comprises the following five layers:
Layer 1 is an input layer, which has five nodes for five input variables: an inflow flow rate, an influent COD value, an ORP value in the anaerobic tank, an aeration quantity, a DO value in the aerobic tank, and the actual effluent COD value;
Layer 2 is a fuzzification input layer, wherein the second layer calculates membership corresponding to each input variable (nodes: 5×11);
Layer 3 is a rules layer with 11 nodes that provide rules that are used to achieve a simple multiplier;
Layer 4 is a fuzzification output layer with 11 nodes, wherein the fourth layer calculates a fitness value of a fuzzy rule; and
Layer 5 is an output layer with 1 node, wherein the output node is the effluent COD predicted value.

4. The method according to claim 1 , wherein the DO fuzzy neural network control model includes five layers: an input layer, a hidden layer, and an output layer, wherein the hidden layer is further divided into three layers: a fuzzification input layer, a rules layer, and a fuzzification output layer that can comprises fuzzification, fuzzy inference, and defuzzification, and wherein a grid partition is used within the rules layer to classify input data and make rules.

5. The method according to claim 4 , wherein the architecture of the DO fuzzy neural network control model comprises five layers:
Layer 1 is the input layer, wherein there are two nodes in the input layer and input variables for the input layer are the effluent COD error or change amount and the effluent COD error change rate;
Layer 2 is the fuzzification input layer, wherein the second layer calculates membership corresponding to each input variable, and wherein the input variables are subdivided into seven reference fuzzy sets in fourteen nodes;
Layer 3 is the rules layer with 49 nodes, wherein there are two input vectors and for each input vector seven MFs are needed;
Layer 4 is the fuzzification output layer with 49 nodes, wherein the fourth layer calculates a fitness value of a fuzzy rule; and
Layer 5 is the output layer with 1 node, wherein the output node is a correction amount of aeration quantity.

6. The method according to claim 1 , wherein characteristics of the wastewater treatment method are described as follow: the influent COD value is 600˜2000 mg/l, the ORP value in the anaerobic tank is −200˜0 mv, and the DO value in the aerobic tank is 0.2˜4.5 mg/l.

7. The method according to claim 6 , wherein the COD fuzzy neural network predictive model and DO fuzzy neural network control model are embedded in a monitoring and control system.

8. The method according to claim 7 , wherein the DO fuzzy neural network control model is used to control the dissolved oxygen as follows: when influent loading increase, the air supply is increased, and when the influent loading decreases, the air supply is decreased.

9. The method according to claim 6 , wherein the method further comprises the following steps:
based on a TCP/IP and serial data interface (R232/485), real-time control of the method is achieved using the computer and a two-way communication tool;
after achieving the real-time control, a comparative analysis is made using a computer for process efficiency of the wastewater treatment method; and
optionally, the comparative analysis is saved to a computer file.

10. The method according to claim 1 , wherein in the method the inflow flowrate, the influent COD value, the ORP value in the anaerobic tank, the DO value in the aerobic tank, and the actual effluent COD value are detected by sensors and signals detected by the sensors are sent to the COD fuzzy neural network predictive model through an analog/digital (A/D) converter module of ADAM4017+ and ADAM4520, so that the effluent COD predicted value is obtained, and the digital signals are changed into analog signals by ADAM4024 to control the speed of a water pump and the air blower.

Assignees

• Wan Jinquan
• Huang Mingzhi
• Ma Yongwen
• Wang Yan
• Univ South China Tech

Inventors

• Wan Jinquan
• Huang Mingzhi
• Ma Yongwen
• Wang Yan

Classifications

• G06N3/043 Primary

  based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS] · CPC title

• C02F2209/003

  Downstream control, i.e. outlet monitoring, e.g. to check the treating agents, such as halogens or ozone, leaving the process · CPC title

• C02F2209/04

  Oxidation reduction potential [ORP] · CPC title

• C02F2209/40

  Liquid flow rate · CPC title

• G06N3/0436 Primary

  Physics · mapped topic