A heating pipeline leak detection method based on deep belief network information fusion
A deep belief network and pipeline leakage technology, which is applied in the field of intelligent leakage detection of heating main pipelines, can solve the problems of heating pipeline leakage, low pipeline leakage efficiency, and low precision, so as to enhance prediction capabilities, update calculations quickly, and improve prediction accuracy Effect
Inactive Publication Date: 2021-01-15
DALIAN UNIV OF TECH +1
View PDF10 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] According to the problems of low efficiency and low precision of pipeline leakage in the prior art proposed above, a heating pipeline leakage detection method based on deep belief network information fusion is provided
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0045] Such as Figure 1-2 As shown, the present invention provides a heating pipeline leak detection method based on deep belief network information fusion, comprising the following steps:
[0046] S1: Collect the original data of the heating pipeline leakage experiment as the initial training sample. The original data is the characteristic matrix of the pressure signal and the flow signal extracted from the heating pipeline pressure signal and flow signal in the normal state and the leaking state, respectively. The characteristic matrix is composed of the mean , RMS, kurtosis and skewness;
[0047] The feature matrix corresponding to the original data is used as input information, and input into two deep belief networks respectively. The pipeline state classification results output by the two models include normal state and leakage state represented by 0 and 1 respectively, and the information fusion of the two models is carried out: The deep belief network combined with ...
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
Login to View More Abstract
The invention provides a heating pipeline leakage detecting method based on depth confidence network information fusion. The method comprises the steps that 1, a depth confidence network combining a double-layer limitation Boltzmann machine and a least square support vector classifying machine based on the Gauss-Cos mixed kernel function is selected, and an information fusion detecting model for heating pipeline leakage is set up; 2, new data information is collected to test the correction rate of the information fusion detecting model; 3, when the correction rate is over 95%, the informationfusion detecting model meets the requirement, and the step 6 is executed; or else, the step 4 or the step 5 is executed; 4, Gibbs sample iteration is utilized for updating and improving the information fusion detecting model until characteristic matrixes corresponding to all original data are all selected; 5, weight bias matrix model parameters of the depth confidence network are updated, updatingof the information fusion detecting model is completed, and the step 2 or the step 3 is returned; and 6, the characteristic matrix of a pipeline to be detected is input into the information fusion detecting model, and the pipeline state classification result is output. The method solves the problems that a method for pipeline leakage is low in efficiency and low in precision.
Description
technical field [0001] The invention relates to the field of intelligent leakage detection of heating main pipelines, in particular to a heating pipeline leakage detection method based on deep belief network information fusion. Background technique [0002] With the gradual popularization of central heating across the country, the laying of heating pipelines has gradually increased. For heating pipelines, the integrity of the pipelines provides sufficient protection for heating. What followed was that heating safety was gradually paid attention to, but due to the increase in the total number of pipelines, safety accidents also gradually increased, posing a huge threat to people's personal safety and property safety. There are many factors for the occurrence of safety accidents, not limited to natural factors. The aging of the pipeline itself, inattentive road construction, human error in operation, etc. will all cause possible safety problems. Once a safety accident occurs, ...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): F17D5/02
CPCF17D5/02
Inventor 李琦杜晓东李萌谢梦琦
Owner DALIAN UNIV OF TECH