Unlock instant, AI-driven research and patent intelligence for your innovation.

A Water Quality Index Prediction Method Based on State Pool Network

A prediction method and state technology, applied in neural learning methods, biological neural network models, neural architectures, etc., can solve the problems of time correlation without integration of water quality indicators, small prediction scale, and large prediction error.

Active Publication Date: 2021-06-08
ZHEJIANG UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although these methods can predict water quality indicators to a certain extent, because these methods do not incorporate the temporal correlation of water quality indicators, the existing methods either have large prediction errors or small prediction scales.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Water Quality Index Prediction Method Based on State Pool Network
  • A Water Quality Index Prediction Method Based on State Pool Network
  • A Water Quality Index Prediction Method Based on State Pool Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] Below in conjunction with experiment the present invention will be further described:

[0047] In order to verify the feasibility of the method proposed in this invention, we use: ammonia nitrogen, dissolved oxygen, CODmn, total phosphorus, total nitrogen, five kinds of water quality data as input data, and two water quality data of ammonia nitrogen and dissolved oxygen as output data to train the state Pool network model. After training, save the structural parameters of the state pool network, and use new data to test the state pool network to verify the effectiveness of the network. If the effect is good, train a new network model, replace the output data with water quality level and retrain

[0048] To collect water quality data, the source of this data is the water quality data set of a certain reservoir from 2012 to 2015, and the collection time is 12 o'clock every day. The time range of the training set is a two-year data set from May 1, 2012 to April 30, 2014....

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A water quality index prediction method based on state pool network, the steps are as follows: 1) determine the input and output data of water quality parameters; 2) design the state pool network structure for detecting abnormal water quality; 3) before carrying out state pool network training, Considering the errors and abnormalities caused by the unstable water body in the measurement of water quality, first screen the effective data and eliminate the abnormal parts; 4) Use the least square method to adjust the output weight of the network; 5) State pool network training; 6) Water Quality Prediction Using State Pool Networks. Save the network structure parameters of the state pool trained in step (5), and use the test data set to test and evaluate the network effect. The root mean square error is also used to evaluate the water quality prediction data to obtain the RMS value.

Description

technical field [0001] The invention relates to the application of water quality prediction by using a recursive neural network. The water quality grade can be determined by detecting the indicators in the water body, such as the content of ammonia nitrogen, dissolved oxygen, chlorophyll a, suspended solids, total phosphorus and total nitrogen. The present invention proposes a method for predicting water quality parameters by using a recursive neural network according to the corresponding relationship of water body index changes in time and space. The method has a certain application prospect for the prediction and monitoring of river water body grades. Background technique [0002] Traditional water quality testing equipment is often distributed in water treatment plants and some fixed points of rivers. In order to test the quality of water bodies, it is generally necessary to collect water body samples, and then perform chemical analysis on the samples to obtain the resul...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06N3/04G06N3/08
CPCG06N3/04G06N3/08
Inventor 程振波朱天奇肖刚黄初冬周华康唐文庆高晶莹
Owner ZHEJIANG UNIV OF TECH