A method for predicting buried pipeline failure based on spatiotemporal Bayesian

Abstract

The present invention provides a method for predicting failure of buried pipelines based on spatiotemporal Bayesian, including: collecting spatiotemporal data of buried pipeline failures, organizing the spatiotemporal data to obtain a failure event set; analyzing the failure event set by expert scoring and combining fuzzy set theory to obtain a priori probability of each failure event; training a Bayesian model by combining the spatiotemporal data with the corresponding prior probability to obtain a spatiotemporal Bayesian model; predicting the failure of buried pipelines based on the spatiotemporal Bayesian model to obtain failure prediction results, and visualizing the failure prediction results to obtain visual prediction results; and selecting a maintenance method to maintain the buried pipeline based on the visual prediction results. This method introduces a model that combines fuzzy set theory with an oil and gas pipeline integrity management effectiveness evaluation index, solving the problems of insufficient pipeline failure data and lack of external corrosion data.

Description

Technical Field

[0001] The present invention belongs to the technical field of pipeline failure prediction, and in particular relates to a buried pipeline failure prediction method based on spatiotemporal Bayesian. Background Art

[0002] Using buried pipelines to transport natural gas and oil to various places is a common mode of transportation. However, the failure of pipelines will pose a major threat to the transportation of oil and gas and the property of pipeline maintenance, and even serious consequences. Ensuring the safe operation of pipelines is the top priority for protecting national property and people's safety. Therefore, failure prediction of buried pipelines is an essential step.

[0003] At present, there are many methods for failure prediction of buried pipelines. Bayesian methods are widely used by researchers to deal with risks and reliability issues related to buried infrastructure. In the current existing technology, it is common to apply spatiotemporal Bayesian to failure prediction of buried pipelines. Due to insufficient and unreliable records of actual failure times and inaccurate measurements of related variables, data quality and model accuracy have become the main concerns. At present, most studies have no good solution to the lack of data types and quantities, and the lack of data will have a huge impact on failure prediction. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a buried pipeline failure prediction method based on spatiotemporal Bayesian to solve the above problems.

[0005] To achieve the above object, the present invention provides a method for predicting buried pipeline failure based on spatiotemporal Bayesian, comprising the following steps:

[0006] Collecting spatiotemporal data of buried pipeline failures, organizing the spatiotemporal data, and obtaining an input event set;

[0007] Analyzing the input event set by expert scoring and combining fuzzy set theory to obtain a priori probability of each input event;

[0008] The spatiotemporal Bayesian model is trained by combining the spatiotemporal data with corresponding prior probabilities to obtain a spatiotemporal Bayesian model;

[0009] Failure prediction of the buried pipeline is performed based on the spatiotemporal Bayesian model to obtain a failure prediction result, and the failure prediction result is visualized to obtain a visual prediction result;

[0010] A maintenance method is selected based on the visual prediction result to maintain the buried pipeline.

[0011] Preferably, the method for collecting spatiotemporal data of buried pipeline failure includes:

[0012] Collect buried pipeline failure data and organize the failed pipeline data;

[0013] Collect topographic map data of buried pipelines and sort out the pipeline location information;

[0014] The failed pipeline data is combined with pipeline location information to obtain spatiotemporal data of buried pipeline failure.

[0015] Preferably, the method for analyzing the input event set includes:

[0016] Experts evaluate input events based on the oil and gas pipeline integrity management effectiveness evaluation indicators and obtain expert scores;

[0017] and summarizing the expert scores to obtain an aggregated fuzziness of buried pipeline failure;

[0018] Defuzzifying the aggregated fuzziness to obtain a fuzzy interval;

[0019] Calculating a fuzzy probability score for each buried pipeline input event based on the fuzzy interval;

[0020] The fuzzy probability score is converted into a fuzzy failure probability.

[0021] Preferably, the expression for obtaining the aggregation ambiguity is:

[0022]

[0023] In the formula, n represents the number of basic events, m represents the number of experts, and w j represents the weight factor of each expert, A ij Represents the value of the horizontal coordinate of the membership function, M i Represents the aggregate fuzziness of elementary events.

[0024] Preferably, the expression for obtaining the fuzzy probability score is:

[0025]

[0026] In the formula, a represents the minimum value when the horizontal coordinate of the membership function is 0, b represents the maximum value when the horizontal coordinate of the membership function is 0, c represents the minimum value when the horizontal coordinate of the membership function is 1, and d represents the maximum value when the horizontal coordinate of the membership function is 1.

[0027] Preferably, the expression for converting the fuzzy probability score into the fuzzy failure probability is:

[0028]

[0029] Where FP represents the fuzzy failure probability and FPS represents the fuzzy probability score.

[0030] Preferably, the method for predicting failure of a buried pipeline comprises:

[0031] Input the spatiotemporal data of the buried pipeline to be measured, convert the spatiotemporal data of the buried pipeline to be measured into a raster file and input it into QGIS for analysis to obtain raster grid data;

[0032] CPT is used to quantify the relationship between related nodes in the spatiotemporal Bayesian model. Based on the grid data, the spatiotemporal Bayesian model is used to perform Bayesian reasoning calculations on each grid cell to generate a grid map.

[0033] A target variable state probability is determined based on the grid graph.

[0034] Preferably, the buried pipeline failure prediction method further includes updating the spatiotemporal Bayesian model according to the number of events occurring within a given time period, and the method for updating the spatiotemporal Bayesian model includes:

[0035] The prior probability is corrected based on the underlying events that occurred within a given time period, and the corrected prior probability is input into the spatiotemporal Bayesian model.

[0036] Preferably, the expression of the corrected prior probability is:

[0037]

[0038] In the formula, t represents the total number of experiences of key events; a represents the number of experiences of key events; and s represents the number of key events that occurred during the entire recording period.

[0039] Compared with the prior art, the present invention has the following advantages and technical effects:

[0040] (1) The spatiotemporal Bayesian plus fuzzy set theory is used to realize the updating of failure probability, and the probability of failure occurring over time is adjusted by recording the number of occurrences of basic events in a given time period of the buried pipeline.

[0041] (2) The problem of insufficient historical failure data and inadequate records of buried pipelines was solved by combining fuzzy set theory with oil and gas pipeline integrity management effectiveness evaluation indicators.

[0042] (3) The oil and gas pipeline integrity management effectiveness evaluation index is combined with spatiotemporal Bayesian to provide important data for the accurate assessment of failure probability when data is insufficient.

[0043] (4) Spatiotemporal Bayesian also uses GIS to solve the problem of difficulty in visualizing traditional Bayesian prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0045] Figure 1 The figure is a flow chart of the prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0047] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0048] Example 1

[0049] like Figure 1 As shown, the present invention proposes a buried pipeline failure prediction method based on spatiotemporal Bayesian, comprising the following steps:

[0050] S1: Data collection based on spatiotemporal Bayesian model.

[0051] S11: Collect a large amount of buried pipeline failure data and organize the failure events as input events.

[0052] S12: Collect buried pipeline topographic map data and organize pipeline location information.

[0053] S2: Based on the expert judgment data and combined with fuzzy set theory, the prior probability of each input event is obtained.

[0054] S21: The oil and gas pipeline integrity management effectiveness evaluation indicators are collected as expert judgment data, and the failure probability is expressed using 7 fuzzy languages, including: very low (VL), low (L), slightly low (ML), medium (M), slightly high (MH), high (H), and very high (VH).

[0055] S22: Summarize the expert scores. The calculation method is as follows:

[0056]

[0057] Where n is the number of base events, m is the number of experts, and w j is the weight factor of each expert, A ij are the values ​​of the horizontal coordinates of the membership function, they all belong to the inflection points of the membership function, M iis the aggregate fuzziness of the elementary events.

[0058] S23: Defuzzify the fuzzy numbers to obtain the representative probability values ​​of the basic events. Use the left-right type fuzzy number sorting method. The left-right function is described as follows:

[0059]

[0060] S24: The fuzzy probability score (FPS) of each event is obtained by the following formula:

[0061]

[0062]

[0063] Among them, a, b, c, and d are the values ​​of the membership function when the horizontal coordinates are 0 and 1.

[0064] S25: Convert the fuzzy probability score (FPS) to the fuzzy failure probability (FP), the formula is as follows:

[0065]

[0066] S3: Train the Bayesian model based on historical failure data.

[0067] S31: Collect historical failure data of buried pipelines.

[0068] S32: Establish a Bayesian model and use historical data to train the model.

[0069] S4: Input relevant data, perform failure prediction on buried pipelines, and visualize them.

[0070] S41: Input relevant data, download as raster files and analyze in QGIS.

[0071] S42: quantify the relationship between relevant nodes in the Bayesian model using CPT. Using the input grid data, perform Bayesian inference calculations on each grid cell.

[0072] S43: The model generates a grid plot where each cell reflects the probability of the state of the target variable determined by Bayesian inference.

[0073] S44: The predicted failure probability is classified into high, medium and low.

[0074] S5: Select maintenance method based on prediction results.

[0075] S6: Update the failure probability based on the number of times the event occurs within a given time period.

[0076] S61: Determine the basic events that occurred within a given time period, adjust the prior probability, and the corrected probability P can be calculated using the following equation:

[0077]

[0078] Where t is the total number of experiences of key events; a is the number of experiences of key events; and s is the number of key events that occurred during the entire recording period.

[0079] S62: Input the corrected probability into the spatiotemporal Bayesian model, and use the updated probability to obtain the dynamic probability of the top event.

[0080] S63: Select a maintenance method according to the obtained dynamic probability.

[0081] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for predicting buried pipeline failure based on spatiotemporal Bayesian, characterized in that: The following steps are involved: Collecting spatiotemporal data of buried pipeline failures, organizing the spatiotemporal data, and obtaining an input event set; Analyzing the input event set by expert scoring and combining fuzzy set theory to obtain a priori probability of each input event; The spatiotemporal Bayesian model is trained by combining the spatiotemporal data with corresponding prior probabilities to obtain a spatiotemporal Bayesian model; Failure prediction of the buried pipeline is performed based on the spatiotemporal Bayesian model to obtain a failure prediction result, and the failure prediction result is visualized to obtain a visual prediction result; The method for predicting failure of a buried pipeline comprises: Input the spatiotemporal data of the buried pipeline to be measured, convert the spatiotemporal data of the buried pipeline to be measured into a raster file and input it into QGIS for analysis to obtain raster grid data; CPT is used to quantify the relationship between related nodes in the spatiotemporal Bayesian model. Based on the grid data, the spatiotemporal Bayesian model is used to perform Bayesian reasoning calculations on each grid cell to generate a grid map. Determining a target variable state probability based on the grid graph; The buried pipeline failure prediction method further includes updating the spatiotemporal Bayesian model according to the number of events occurring within a given time period. The method for updating the spatiotemporal Bayesian model includes: The failure probability is corrected based on the basic events that occurred in a given time period, and the corrected failure probability is input into the spatiotemporal Bayesian model; The expression for the probability of corrected failure is: In the formula, t represents the total number of experiences of key events; a represents the number of experiences of key events; s represents the number of key events that occurred during the entire recording period; A maintenance method is selected based on the visual prediction result to maintain the buried pipeline.

2. The buried pipeline failure prediction method based on spatiotemporal Bayesian according to claim 1 is characterized in that: The method for collecting spatiotemporal data of buried pipeline failure includes: Collect buried pipeline failure data and organize the failed pipeline data; Collect topographic map data of buried pipelines and sort out the pipeline location information; The failed pipeline data is combined with pipeline location information to obtain spatiotemporal data of buried pipeline failure.

3. The buried pipeline failure prediction method based on spatiotemporal Bayesian according to claim 1 is characterized in that: The method for analyzing the input event set includes: Experts evaluate input events based on the oil and gas pipeline integrity management effectiveness evaluation indicators and obtain expert scores; and summarizing the expert scores to obtain an aggregated fuzziness of buried pipeline failure; Defuzzifying the aggregated fuzziness to obtain a fuzzy interval; Calculating a fuzzy probability score for each buried pipeline input event based on the fuzzy interval; The fuzzy probability score is converted into a fuzzy failure probability.

4. The method for predicting buried pipeline failure based on spatiotemporal Bayesian according to claim 3 is characterized in that: The expression for obtaining the aggregate fuzziness is: In the formula, n represents the number of basic events, m represents the number of experts, and w j represents the weight factor of each expert, A ij Represents the value of the horizontal coordinate of the membership function, M i Represents the aggregate fuzziness of elementary events.

5. The method for predicting buried pipeline failure based on spatiotemporal Bayesian according to claim 3 is characterized in that: The expression for converting the fuzzy probability score into the fuzzy failure probability is: Where FP represents the fuzzy failure probability and FPS represents the fuzzy probability score.

Images