Early warning method and device for NOx exceeding standard of gas turbine based on LSTM deep learning

A technology of deep learning and gas turbine, applied in the direction of instrumentation, adaptive control, control/regulation system, etc., can solve problems such as large prediction error, control influence, large detection data value, etc., achieve high accuracy and avoid omission or error , the effect of collecting a large amount of data

Pending Publication Date: 2019-12-27
杭州华电江东热电有限公司
View PDF7 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Its shortcoming is that the selected characteristics are parameters that have obvious correlation with the NOx content of the output measuring point, and it is necessary to select key influencing factor characteristics for the input of the mo...

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
  • Early warning method and device for NOx exceeding standard of gas turbine based on LSTM deep learning
  • Early warning method and device for NOx exceeding standard of gas turbine based on LSTM deep learning
  • Early warning method and device for NOx exceeding standard of gas turbine based on LSTM deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] Such as Figure 1-3 As shown, a gas turbine NOx excessive early warning method based on LSTM deep learning is characterized in that it includes the following steps:

[0055] Step 1: Obtain the historical operation data of the gas turbine and preprocess the historical data;

[0056] Step 2: Divide the model training set and test set according to the preprocessed historical data, and based on the training set, construct an LSTM prediction model with NOx emissions as the output;

[0057] Step 3: Use the test set data to input the gas turbine NOx emission model, compare whether the deviation between the actual value of NOx emission and the predicted value is within a reasonable range, and judge whether the model is valid; if it is invalid, proceed to step 4; if it is valid, save the model and end ;

[0058] Step 4: Adjust the hyperparameters such as the number of network layers and learning rate of LSTM, and re-use the training data for modeling and prediction after adjus...

Embodiment 2

[0096] A gas turbine NOx excess warning device based on LSTM deep learning that installs the above method, including a memory that is electrically connected and stores a program that implements the above method, a processor, I / O equipment, and an alarm device, and the I / O equipment is connected to the installation of the gas turbine. Computers and / or networks of monitoring software to access and obtain real-time measurement point data.

[0097] The processor is connected to the hand-held user terminal through wireless transmission. Remote monitoring and early warning through handheld devices.

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

The invention belongs to the field of power plant safety control system, and particularly relates to a early warning method and a device for NOx exceeding standard of a gas turbine based on LSTM deeplearning. This patent lies in applying the LSTM prediction model to the NOx prediction of the gas turbine, as the NOx emissions need to be affected by a variety of factors, mainly factors such as flue gas temperature, differences in fuel gas, and oxygen content, therefore, it usually takes a while for these factors through a series of reactions to pass through related equipment and finally lead to NOx emissions. Through real-time model calculation, if it is found that the NOx content value reaches 50ppm, pollution prediction can be issued in advance, and emission reduction measures can be formulated in advance, so as to achieve intelligent monitoring of NOx emissions and early warning in advance. The LSTM method is used to process the raw data to adaptively find out the characteristic parameters associated with NOx emissions and avoid omissions or errors caused by manual selection of features by traditional machine learning methods. The invention has a large amount of collected data,small analysis errors, and high accuracy of early warning results.

Description

technical field [0001] The invention belongs to the field of power plant safety control systems, and in particular relates to a gas turbine NOx excessive early warning method and device based on LSTM deep learning. Background technique [0002] In the prior art, methods for detecting abnormalities of gas turbines can be roughly divided into two categories. One is the method of using a mechanism model to establish a physical model based on the thermodynamic properties and thermodynamic principles of the gas, and use the model to calculate various KPI indicators of the gas turbine, and compare them with the measured values. If there is a large deviation between the measured value and the theoretical value, it is considered that the gas turbine is abnormal. The main problem of the mechanism model is that when using the principles of physics to establish the analysis model, there are a lot of assumptions and simplified conditions, which are not suitable for complex systems in r...

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
IPC IPC(8): G05B13/04
CPCG05B13/042
Inventor 俞利锋李勇辉方继辉
Owner 杭州华电江东热电有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap