Gas pipeline data management system and method based on artificial intelligence

Abstract

The invention discloses a gas pipeline data management system and method based on artificial intelligence, and relates to the technical field of data management. The system comprises a data acquisition and processing module, an abnormal point identification and path construction module, a dominant point analysis and deviation calculation module, a real-time data analysis and anomaly detection module and a historical data comparison and maintenance suggestion module. The data acquisition and processing module is responsible for acquiring sensor gas pipeline data, processing the data to form a data unit and extracting features; the abnormal point recognition and path construction module analyzes data features, recognizes abnormal points and constructs an abnormal point path; the dominant point analysis and deviation calculation module analyzes abnormal point path features, determines abnormal dominant points and calculates deviation indexes; the real-time data analysis and anomaly detection module obtains and analyzes real-time data, and identifies and constructs an abnormal point path; and the historical data comparison and maintenance suggestion module compares the real-time abnormal point path data with the historical abnormal point path data and outputs notification information.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a gas pipeline data management system and method based on artificial intelligence. Background Art

[0002] With the acceleration of urbanization and the popularization of gas applications, the gas pipeline system, as an important part of urban infrastructure, is directly related to the safety and stability of residents' lives and the smooth progress of industrial production. However, the gas pipeline system is complex and large, widely distributed, and has long been affected by various factors such as the environment, material aging, and improper construction. It is prone to leakage, blockage, abnormal pressure and other faults. If these faults are not discovered and handled in time, they will not only cause economic losses, but may also cause serious safety accidents.

[0003] Although the existing gas pipeline data management system can meet the needs to a certain extent, it still has shortcomings. The existing gas pipeline data management methods mostly rely on manual inspections and regular testing, which have problems such as low efficiency, high cost, and slow response, making it difficult to achieve all-weather, real-time and effective monitoring and management of the gas pipeline system. Especially when dealing with abnormal points in gas pipelines, they can often only be analyzed based on single-point data, ignoring the possible correlation between abnormal points in different locations, thus affecting the accuracy of fault location and processing efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a gas pipeline data management system and method based on artificial intelligence to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A gas pipeline data management method based on artificial intelligence, the method comprising the following steps:

[0007] Step S100. Obtain the gas pipeline data of all sensor devices in the monitoring area, record the location data of the sensor devices, correspond the gas pipeline data with the location data of the sensor devices, and form several data units; analyze the gas pipeline data of each data unit to obtain corresponding data features;

[0008] Step S200. According to the data characteristics of the data unit, the trend of the data characteristics changing over time is analyzed to identify the abnormal points of the gas pipeline; the location data of all the abnormal points of the gas pipeline are obtained, and the abnormal point path is constructed based on the location data of the abnormal points of the gas pipeline;

[0009] Step S300. Based on the abnormal point path, analyze the correlation between the data characteristics and position data of the abnormal points of different gas pipelines, and obtain the abnormal dominant point of the abnormal point path according to the correlation; analyze the deviation relationship between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path, so as to obtain the corresponding deviation index;

[0010] Step S400. Acquire the real-time gas pipeline data of the monitoring area, analyze the real-time gas pipeline data at every selected time period, thereby obtaining the corresponding real-time data features, and identifying the real-time gas pipeline abnormal points; construct a real-time abnormal point path based on the real-time gas pipeline abnormal points; calculate the deviation index between the real-time gas pipeline abnormal points on the real-time abnormal point path, and output corresponding prompt information based on the deviation index.

[0011] Furthermore, step S100 includes:

[0012] S101. Obtain a gas pipeline distribution plan in the monitoring area, mark the location of the sensor equipment on the gas pipeline distribution plan and number the sensor equipment, and each sensor equipment number is unique; record the location data of the sensor equipment according to the gas pipeline distribution plan, and obtain the corresponding gas pipeline data according to the sensor equipment number;

[0013] S102. Obtain the gas pipeline data of the selected time period in the monitoring area in chronological order, match the gas pipeline data of each sensor device with the corresponding location data, thereby forming a number of data units, and each data unit includes the location data of a sensor device and the corresponding gas pipeline data; perform time series analysis on each data unit to obtain corresponding data features.

[0014] Further, step S200 includes:

[0015] S201. Obtain data features of the data unit and preprocess the features of the data unit; for each data unit, construct a feature vector V according to the preprocessed data features, and the feature vector V = [v1, v2, ..., vn], wherein v1 represents the first feature data of the data unit, v2 represents the second feature data of the data unit, and so on, vn represents the nth feature data of the data unit, and n represents the data feature number of the data unit; for the feature vector V of each data unit, principal component analysis is performed to obtain the data feature representation Vr after dimensionality reduction;

[0016] S202. According to the data characteristic representation Vr of each data unit, a trend curve L of each characteristic data corresponding to the data characteristic representation Vr over time is drawn in chronological order, and according to the set standard characteristic data trend curve range LQ, the trend abnormal point of each data unit is identified, and according to the position data of the sensor device corresponding to the data unit, the position where the abnormal point of the gas pipeline occurs is obtained;

[0017] S203. Obtain the location data of all abnormal points of the gas pipelines, mark the abnormal points of the gas pipelines on the gas pipeline distribution plan, and connect the abnormal points of the gas pipelines with adjacent location data to form a number of abnormal point paths.

[0018] Furthermore, principal component analysis is performed on the eigenvector V of each data unit. The specific analysis process is as follows:

[0019] Standardize the eigenvectors V of all data units so that the mean of each feature is 0 and the variance is 1, and record the standardized eigenvector as X; calculate the covariance matrix C of the standardized data features, and C = [1 / (m-1)]X T X, where m represents the number of data units; perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ and the corresponding eigenvector W, and C = WλW T ; Get all eigenvalues ​​and arrange them in descending order, and record the corresponding eigenvectors; select the first k eigenvalues ​​and their corresponding eigenvectors, combine these k eigenvectors to form a principal component matrix Wk, and Wk = [w1, w2, ..., wk], where w1 represents the eigenvector corresponding to the first eigenvalue, w2 represents the eigenvector corresponding to the second eigenvalue, and so on, wk represents the eigenvector corresponding to the kth eigenvalue; use the principal component matrix Wk to perform a linear transformation on the standardized eigenvector X to obtain the reduced-dimensional data feature representation Vr, and Vr = XWk.

[0020] Furthermore, step S300 includes:

[0021] S301. For each abnormal point path, obtain the corresponding gas pipeline abnormal point data feature representation Vr and the corresponding gas pipeline abnormal point location data; obtain the corresponding trend curve L according to the gas pipeline abnormal point data feature representation Vr, calculate the area S enclosed by the trend curve L and the coordinate axis, and associate the area S with the location coordinates of the gas pipeline abnormal point to form an associated data pair;

[0022] S302.According to the associated data pair, the area S corresponding to the abnormal points of different gas pipelines on the current abnormal point path is obtained, and the size relationship of the area S is compared. The abnormal point of the gas pipeline with the largest area S is taken as the abnormal dominant point of the current abnormal point path, and the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is calculated according to the position coordinates of other gas pipeline abnormal points on the current abnormal point path; the abnormal dominant point on the current abnormal point path is taken as the center of the circle, and the maximum value of the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is taken as the radius, and the enclosed area is the abnormal dominant point radiation area;

[0023] S303. Obtain the associated data pair between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point, and calculate the deviation index P between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point according to the associated data pair, and the specific calculation formula is: P = [(1 / H)Σ h∈[1,H] (S1-Sh)] / S0, where S1 represents the area enclosed by the trend curve L of the abnormal dominant point and the coordinate axis, Sh represents the area enclosed by the trend curve L of the hth other gas pipeline abnormal point in the current abnormal point path and the coordinate axis, and S0 represents the difference between the maximum area and the minimum area enclosed by the trend curve L and the coordinate axis corresponding to the set standard characteristic data trend curve range interval LQ; the deviation index P of each abnormal point path is associated with the radiation area area Sf of the abnormal dominant point to form a historical abnormal point path data pair;

[0024] S304. Obtain historical abnormal point path data pairs of different abnormal point paths, combine with historical gas pipeline maintenance records, obtain time periods with maintenance records, and match historical abnormal point path data pairs of corresponding time periods with corresponding maintenance records.

[0025] Furthermore, step S400 includes:

[0026] S401. Acquire real-time gas pipeline data in the monitoring area, analyze the real-time gas pipeline data at selected time intervals, and obtain corresponding real-time data features; analyze the real-time data features in the same way as step S200, thereby identifying real-time gas pipeline abnormal points, and construct real-time abnormal point paths according to the real-time gas pipeline abnormal points;

[0027] S402. For each real-time abnormal point path, the corresponding real-time dominant abnormal point and the real-time abnormal dominant point radiation area Sf0 are obtained in the same analysis method as step S300; based on the real-time dominant abnormal point and the real-time abnormal dominant point radiation area, the real-time deviation index P0 of the current abnormal point path is calculated, thereby obtaining the real-time abnormal point path data pair;

[0028] S403. Compare the real-time abnormal point path data pair with the historical abnormal point path data pair. If the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is greater than or equal to the similarity threshold, the real-time abnormal point path data pair, the real-time abnormal dominant point and the corresponding historical maintenance record are output to relevant personnel; if the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is less than the similarity threshold, only the real-time abnormal point path data pair and the real-time abnormal dominant point are output, and the relevant personnel perform corresponding maintenance processing.

[0029] A gas pipeline data management system based on artificial intelligence, the system includes: a data acquisition and processing module, an abnormal point identification and path construction module, a dominant point analysis and deviation calculation module, a real-time data analysis and abnormality detection module, and a historical data comparison and maintenance suggestion module;

[0030] The data acquisition and processing module obtains the gas pipeline data and its location data of all sensor devices in the monitoring area, records and connects the data to form a data unit; performs time series analysis on the data unit to extract data features; the abnormal point identification and path construction module analyzes the trend changes of data features and identifies abnormal points in the gas pipeline; constructs abnormal point paths based on the location data of abnormal points in the gas pipeline, and marks all abnormal points in the gas pipeline; the dominant point analysis and deviation calculation module analyzes the correlation between the data features and the location data on the abnormal point path to obtain the abnormal dominant point; calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points to form the abnormal dominant point radiation area; the real-time data analysis and anomaly detection module obtains and analyzes real-time gas pipeline data, identifies real-time abnormal points and constructs real-time abnormal point paths, calculates the deviation index of the real-time abnormal point path, and compares it with historical data; the historical data comparison and maintenance suggestion module compares the real-time abnormal point path data pair with the historical abnormal point path data pair, and outputs the corresponding notification information based on the similarity threshold.

[0031] Further, the data acquisition and processing module includes a data acquisition unit, a data recording unit and a data timing analysis unit;

[0032] The data acquisition unit obtains the gas pipeline data and sensor location data from the sensor equipment in the monitoring area, marks the sensor equipment location and its number according to the gas pipeline distribution plan, and records the location data; the data recording unit records the location data of each sensor equipment according to the sensor number and the gas pipeline distribution plan, and corresponds the gas pipeline data with the sensor location data to form a data unit; the data timing analysis unit performs timing analysis on each data unit, extracts and records data features.

[0033] Further, the outlier identification and path construction module includes a data feature analysis unit and an outlier annotation and path construction unit;

[0034] The data feature analysis unit performs principal component analysis on the data features to obtain the data feature representation after dimensionality reduction; draws a trend curve of the feature data changing over time to identify trend anomalies; the anomaly marking and path construction unit marks the gas pipeline anomalies according to the identified anomaly location data, and connects the adjacent location data to form an anomaly path;

[0035] The dominant point analysis and deviation calculation module includes an abnormal point data feature extraction unit, a dominant point determination unit and a deviation calculation unit;

[0036] The abnormal point data feature extraction unit obtains the data feature representation of each gas pipeline abnormal point in the abnormal point path, and calculates the area enclosed by its trend curve and the coordinate axis; the dominant point determination unit compares the areas of different gas pipeline abnormal points to determine the abnormal dominant point; calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points, and generates the abnormal dominant point radiation area; the deviation calculation unit calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points, and associates it with the historical abnormal point path data.

[0037] Furthermore, the real-time data analysis and anomaly detection module includes a real-time data acquisition and feature analysis unit, a real-time anomaly point path construction unit, and a real-time dominant point and deviation calculation unit;

[0038] The real-time data acquisition and feature analysis unit acquires real-time gas pipeline data, analyzes it at selected time intervals, extracts real-time data features, and identifies abnormal points in the real-time gas pipeline; the real-time abnormal point path construction unit constructs a real-time abnormal point path based on the abnormal points in the real-time gas pipeline; the real-time dominant point and deviation calculation unit performs real-time dominant abnormal point analysis, calculates the deviation index of the real-time abnormal dominant point, and generates real-time abnormal point path data pairs;

[0039] The historical data comparison and maintenance suggestion module includes a historical data comparison unit and a maintenance record association and suggestion unit;

[0040] The historical data comparison unit compares the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair; based on the similarity threshold, a comparison result is generated; the maintenance record association and suggestion unit outputs the real-time abnormal point path data pair and the historical maintenance record according to the similarity result.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] Artificial intelligence algorithms are used to analyze gas pipeline data in real time, thereby achieving real-time monitoring around the clock. This avoids the problems of low efficiency, high cost and slow response of relying on manual inspections and regular inspections in traditional methods, and significantly improves the response speed and management efficiency of the system. By analyzing the time series characteristics of data units and principal component analysis, the changing trend of multi-point data can be comprehensively considered. This method can not only identify single-point anomalies, but also reveal the potential correlation between anomalies at different locations, improving the accuracy and comprehensiveness of anomaly detection. Constructing anomaly point paths based on the location data of anomalies can effectively visualize the relationship between anomalies, thereby helping to identify abnormal patterns and potential system vulnerabilities. This path analysis method is more comprehensive than traditional single-point anomaly detection and helps to quickly locate the source of the problem. By calculating the deviation index of the dominant point of anomalies, the severity and impact range of the anomaly can be quantified. This method can accurately evaluate the relative importance of anomalies and help determine the dominant point of anomalies for priority processing, improving the accuracy of fault location and maintenance. Comparing the real-time anomaly point path data with the historical anomaly point path data can effectively identify the changing trend of the anomaly pattern. This comparative analysis not only helps to understand the relationship between current anomalies and historical anomalies, but also can optimize maintenance strategies in combination with historical maintenance records. Outputting prompt information based on deviation index and similarity threshold helps to respond and handle potential problems in a timely manner. The system can automatically feedback important information to relevant personnel, thereby improving the timeliness and accuracy of problem handling and reducing the need for human intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0044] Figure 1 It is a schematic diagram of a gas pipeline data management system module based on artificial intelligence in the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1 , the present invention provides a technical solution:

[0047] A gas pipeline data management system based on artificial intelligence, the system includes: a data acquisition and processing module, an abnormal point identification and path construction module, a dominant point analysis and deviation calculation module, a real-time data analysis and abnormality detection module, and a historical data comparison and maintenance suggestion module;

[0048] The data acquisition and processing module obtains the gas pipeline data and its location data of all sensor devices in the monitoring area, records and connects the data to form a data unit; performs time series analysis on the data unit to extract data features; the abnormal point identification and path construction module analyzes the trend changes of data features and identifies abnormal points in the gas pipeline; constructs abnormal point paths based on the location data of abnormal points in the gas pipeline, and marks all abnormal points in the gas pipeline; the dominant point analysis and deviation calculation module analyzes the correlation between the data features and the location data on the abnormal point path to obtain the abnormal dominant point; calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points to form the abnormal dominant point radiation area; the real-time data analysis and anomaly detection module obtains and analyzes real-time gas pipeline data, identifies real-time abnormal points and constructs real-time abnormal point paths, calculates the deviation index of the real-time abnormal point path, and compares it with historical data; the historical data comparison and maintenance suggestion module compares the real-time abnormal point path data pair with the historical abnormal point path data pair, and outputs the corresponding notification information based on the similarity threshold.

[0049] The data acquisition and processing module includes a data acquisition unit, a data recording unit and a data timing analysis unit;

[0050] The data acquisition unit obtains the gas pipeline data and sensor location data from the sensor equipment in the monitoring area, marks the sensor equipment location and its number according to the gas pipeline distribution plan, and records the location data; the data recording unit records the location data of each sensor equipment according to the sensor number and the gas pipeline distribution plan, and corresponds the gas pipeline data with the sensor location data to form a data unit; the data timing analysis unit performs timing analysis on each data unit, extracts and records data features.

[0051] The outlier identification and path construction module includes a data feature analysis unit and an outlier annotation and path construction unit;

[0052] The data feature analysis unit performs principal component analysis on the data features to obtain the data feature representation after dimensionality reduction; draws a trend curve of the feature data changing over time to identify trend anomalies; the anomaly marking and path construction unit marks the gas pipeline anomalies according to the identified anomaly location data, and connects the adjacent location data to form an anomaly path;

[0053] The dominant point analysis and deviation calculation module includes an abnormal point data feature extraction unit, a dominant point determination unit and a deviation calculation unit;

[0054] The abnormal point data feature extraction unit obtains the data feature representation of each gas pipeline abnormal point in the abnormal point path, and calculates the area enclosed by its trend curve and the coordinate axis; the dominant point determination unit compares the areas of different gas pipeline abnormal points to determine the abnormal dominant point; calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points, and generates the abnormal dominant point radiation area; the deviation calculation unit calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points, and associates it with the historical abnormal point path data.

[0055] The real-time data analysis and anomaly detection module includes a real-time data acquisition and feature analysis unit, a real-time anomaly point path construction unit, and a real-time dominant point and deviation calculation unit;

[0056] The real-time data acquisition and feature analysis unit acquires real-time gas pipeline data, analyzes it at selected time intervals, extracts real-time data features, and identifies abnormal points in the real-time gas pipeline; the real-time abnormal point path construction unit constructs a real-time abnormal point path based on the abnormal points in the real-time gas pipeline; the real-time dominant point and deviation calculation unit performs real-time dominant abnormal point analysis, calculates the deviation index of the real-time abnormal dominant point, and generates real-time abnormal point path data pairs;

[0057] The historical data comparison and maintenance suggestion module includes a historical data comparison unit and a maintenance record association and suggestion unit;

[0058] The historical data comparison unit compares the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair; based on the similarity threshold, a comparison result is generated; the maintenance record association and suggestion unit outputs the real-time abnormal point path data pair and the historical maintenance record according to the similarity result.

[0059] A gas pipeline data management method based on artificial intelligence, the method comprising the following steps:

[0060] Step S100. Obtain the gas pipeline data of all sensor devices in the monitoring area, record the location data of the sensor devices, correspond the gas pipeline data with the location data of the sensor devices, and form several data units; analyze the gas pipeline data of each data unit to obtain corresponding data features;

[0061] Step S200. According to the data characteristics of the data unit, the trend of the data characteristics changing over time is analyzed to identify the abnormal points of the gas pipeline; the location data of all the abnormal points of the gas pipeline are obtained, and the abnormal point path is constructed based on the location data of the abnormal points of the gas pipeline;

[0062] Step S300. Based on the abnormal point path, analyze the correlation between the data characteristics and position data of the abnormal points of different gas pipelines, and obtain the abnormal dominant point of the abnormal point path according to the correlation; analyze the deviation relationship between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path, so as to obtain the corresponding deviation index;

[0063] Step S400. Acquire the real-time gas pipeline data of the monitoring area, analyze the real-time gas pipeline data at every selected time period, thereby obtaining the corresponding real-time data features, and identifying the real-time gas pipeline abnormal points; construct a real-time abnormal point path based on the real-time gas pipeline abnormal points; calculate the deviation index between the real-time gas pipeline abnormal points on the real-time abnormal point path, and output corresponding prompt information based on the deviation index.

[0064] Step S100 includes:

[0065] S101. Obtain a gas pipeline distribution plan in the monitoring area, mark the location of the sensor equipment on the gas pipeline distribution plan and number the sensor equipment, and each sensor equipment number is unique; record the location data of the sensor equipment according to the gas pipeline distribution plan, and obtain the corresponding gas pipeline data according to the sensor equipment number;

[0066] S102. Obtain the gas pipeline data of the selected time period in the monitoring area in chronological order, match the gas pipeline data of each sensor device with the corresponding location data, thereby forming a number of data units, and each data unit includes the location data of a sensor device and the corresponding gas pipeline data; perform time series analysis on each data unit to obtain corresponding data features.

[0067] In this embodiment, the process of obtaining the data characteristics of the data unit is as follows:

[0068] Arrange the data units into a time series format and obtain the corresponding statistical features, including basic statistics such as mean, standard deviation, maximum and minimum values; fit the trend line of the data unit by calculating the difference and slope of consecutive time points, and obtain the trend coefficient and intercept as the trend feature of the data unit; use Fourier transform or seasonal decomposition method (such as STL) to extract the amplitude, phase and frequency of the periodic component as periodic features; however, for gas pipeline data in practical applications, people often pay more attention to low-frequency seasonal changes or long-term trend changes rather than high-frequency small fluctuations. Therefore, when processing such data, the complete Fourier transform results may not be needed, but some simplified methods are used to estimate the periodic features, among which the simplified method for periodic feature extraction is:

[0069] Seasonal indicators: If there is a clear seasonal pattern in the data (such as annual, quarterly, or monthly changes), calculate seasonal indicators. For example, for annual data, you can calculate the mean or median for each season (such as spring, summer, fall, winter) and use these seasonal averages as features.

[0070] Cyclic Rate of Change: Calculates the rate of change of data at the same point in time within adjacent cycles, which can help identify cyclical growth or decline trends.

[0071] Seasonal decomposition: Use seasonal decomposition methods (such as STL decomposition) to decompose time series data into trend, seasonal and random components. The seasonal component can be directly used as a periodic feature, while the trend component can be used to analyze long-term changes.

[0072] Step S200 includes:

[0073] S201. Obtain data features of the data unit and preprocess the features of the data unit; for each data unit, construct a feature vector V according to the preprocessed data features, and the feature vector V = [v1, v2, ..., vn], wherein v1 represents the first feature data of the data unit, v2 represents the second feature data of the data unit, and so on, vn represents the nth feature data of the data unit, and n represents the data feature number of the data unit; for the feature vector V of each data unit, principal component analysis is performed to obtain the data feature representation Vr after dimensionality reduction;

[0074] S202. According to the data characteristic representation Vr of each data unit, a trend curve L of each characteristic data corresponding to the data characteristic representation Vr over time is drawn in chronological order, and according to the set standard characteristic data trend curve range LQ, the trend abnormal point of each data unit is identified, and according to the position data of the sensor device corresponding to the data unit, the position where the abnormal point of the gas pipeline occurs is obtained;

[0075] S203. Obtain the location data of all abnormal points of the gas pipelines, mark the abnormal points of the gas pipelines on the gas pipeline distribution plan, and connect the abnormal points of the gas pipelines with adjacent location data to form a number of abnormal point paths.

[0076] For each data unit’s eigenvector V, principal component analysis is performed. The specific analysis process is as follows:

[0077] Standardize the eigenvectors V of all data units so that the mean of each feature is 0 and the variance is 1, and record the standardized eigenvector as X; calculate the covariance matrix C of the standardized data features, and C = [1 / (m-1)]X TX, where m represents the number of data units; perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ and the corresponding eigenvector W, and C = WλW T ; Get all eigenvalues ​​and arrange them in descending order, and record the corresponding eigenvectors; select the first k eigenvalues ​​and their corresponding eigenvectors, combine these k eigenvectors to form a principal component matrix Wk, and Wk = [w1, w2, ..., wk], where w1 represents the eigenvector corresponding to the first eigenvalue, w2 represents the eigenvector corresponding to the second eigenvalue, and so on, wk represents the eigenvector corresponding to the kth eigenvalue; use the principal component matrix Wk to perform a linear transformation on the standardized eigenvector X to obtain the reduced-dimensional data feature representation Vr, and Vr = XWk.

[0078] Step S300 includes:

[0079] S301. For each abnormal point path, obtain the corresponding gas pipeline abnormal point data feature representation Vr and the corresponding gas pipeline abnormal point location data; obtain the corresponding trend curve L according to the gas pipeline abnormal point data feature representation Vr, calculate the area S enclosed by the trend curve L and the coordinate axis, and associate the area S with the location coordinates of the gas pipeline abnormal point to form an associated data pair;

[0080] S302.According to the associated data pair, the area S corresponding to the abnormal points of different gas pipelines on the current abnormal point path is obtained, and the size relationship of the area S is compared. The abnormal point of the gas pipeline with the largest area S is taken as the abnormal dominant point of the current abnormal point path, and the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is calculated according to the position coordinates of other gas pipeline abnormal points on the current abnormal point path; the abnormal dominant point on the current abnormal point path is taken as the center of the circle, and the maximum value of the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is taken as the radius, and the enclosed area is the abnormal dominant point radiation area;

[0081] S303. Obtain the associated data pair between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point, and calculate the deviation index P between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point according to the associated data pair, and the specific calculation formula is: P = [(1 / H)Σ h∈[1,H](S1-Sh)] / S0, where S1 represents the area enclosed by the trend curve L of the abnormal dominant point and the coordinate axis, Sh represents the area enclosed by the trend curve L of the hth other gas pipeline abnormal point in the current abnormal point path and the coordinate axis, and S0 represents the difference between the maximum area and the minimum area enclosed by the trend curve L and the coordinate axis corresponding to the set standard characteristic data trend curve range interval LQ; the deviation index P of each abnormal point path is associated with the radiation area area Sf of the abnormal dominant point to form a historical abnormal point path data pair;

[0082] S304. Obtain historical abnormal point path data pairs of different abnormal point paths, combine with historical gas pipeline maintenance records, obtain time periods with maintenance records, and match historical abnormal point path data pairs of corresponding time periods with corresponding maintenance records.

[0083] Step S400 includes:

[0084] S401. Acquire real-time gas pipeline data in the monitoring area, analyze the real-time gas pipeline data at selected time intervals, and obtain corresponding real-time data features; analyze the real-time data features in the same way as step S200, thereby identifying real-time gas pipeline abnormal points, and construct real-time abnormal point paths according to the real-time gas pipeline abnormal points;

[0085] S402. For each real-time abnormal point path, the corresponding real-time dominant abnormal point and the real-time abnormal dominant point radiation area Sf0 are obtained in the same analysis method as step S300; based on the real-time dominant abnormal point and the real-time abnormal dominant point radiation area, the real-time deviation index P0 of the current abnormal point path is calculated, thereby obtaining the real-time abnormal point path data pair;

[0086] S403. Compare the real-time abnormal point path data pair with the historical abnormal point path data pair. If the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is greater than or equal to the similarity threshold, the real-time abnormal point path data pair, the real-time abnormal dominant point and the corresponding historical maintenance record are output to relevant personnel; if the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is less than the similarity threshold, only the real-time abnormal point path data pair and the real-time abnormal dominant point are output, and the relevant personnel perform corresponding maintenance processing.

[0087] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0088] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A gas pipeline data management method based on artificial intelligence, characterized in that: The method comprises the following steps: Step S100. Obtain the gas pipeline data of all sensor devices in the monitoring area, record the location data of the sensor devices, correspond the gas pipeline data with the location data of the sensor devices, and form several data units; analyze the gas pipeline data of each data unit to obtain corresponding data features; Step S200. According to the data characteristics of the data unit, the trend of the data characteristics changing over time is analyzed to identify the abnormal points of the gas pipeline; the location data of all the abnormal points of the gas pipeline are obtained, and the abnormal point path is constructed based on the location data of the abnormal points of the gas pipeline; Step S300. Based on the abnormal point path, analyze the correlation between the data characteristics and position data of the abnormal points of different gas pipelines, and obtain the abnormal dominant point of the abnormal point path according to the correlation; analyze the deviation relationship between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path, so as to obtain the corresponding deviation index; Step S400. Acquire the real-time gas pipeline data of the monitoring area, analyze the real-time gas pipeline data at every selected time period, thereby obtaining the corresponding real-time data features, and identifying the real-time gas pipeline abnormal points; construct a real-time abnormal point path based on the real-time gas pipeline abnormal points; calculate the deviation index between the real-time gas pipeline abnormal points on the real-time abnormal point path, and output corresponding prompt information based on the deviation index.

2. The method for gas pipeline data management based on artificial intelligence according to claim 1, characterized in that: The step S100 includes: S101. Obtain a gas pipeline distribution plan in the monitoring area, mark the location of the sensor equipment on the gas pipeline distribution plan and number the sensor equipment, and each sensor equipment number is unique; record the location data of the sensor equipment according to the gas pipeline distribution plan, and obtain the corresponding gas pipeline data according to the sensor equipment number; S102. Obtain the gas pipeline data of the selected time period in the monitoring area in chronological order, match the gas pipeline data of each sensor device with the corresponding location data, thereby forming a number of data units, and each data unit includes the location data of a sensor device and the corresponding gas pipeline data; perform time series analysis on each data unit to obtain corresponding data features.

3. The method for gas pipeline data management based on artificial intelligence according to claim 1, characterized in that: The step S200 includes: S201. Obtain data features of the data unit and preprocess the features of the data unit; for each data unit, construct a feature vector V according to the preprocessed data features, and the feature vector V = [v1, v2, ..., vn], wherein v1 represents the first feature data of the data unit, v2 represents the second feature data of the data unit, and so on, vn represents the nth feature data of the data unit, and n represents the data feature number of the data unit; for the feature vector V of each data unit, principal component analysis is performed to obtain the data feature representation Vr after dimensionality reduction; S202. According to the data characteristic representation Vr of each data unit, a trend curve L of each characteristic data corresponding to the data characteristic representation Vr over time is drawn in chronological order, and according to the set standard characteristic data trend curve range LQ, the trend abnormal point of each data unit is identified, and according to the position data of the sensor device corresponding to the data unit, the position where the abnormal point of the gas pipeline occurs is obtained; S203. Obtain the location data of all abnormal points of the gas pipelines, mark the abnormal points of the gas pipelines on the gas pipeline distribution plan, and connect the abnormal points of the gas pipelines with adjacent location data to form a number of abnormal point paths.

4. The method for gas pipeline data management based on artificial intelligence according to claim 3 is characterized in that: The principal component analysis is performed on the feature vector V of each data unit, and the specific analysis process is as follows: Standardize the eigenvectors V of all data units so that the mean of each feature is 0 and the variance is 1, and record the standardized eigenvector as X; calculate the covariance matrix C of the standardized data features, and C = [1 / (m-1)]X T X, where m represents the number of data units; perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ and the corresponding eigenvector W, and C = WλW T ; Get all eigenvalues ​​and arrange them in descending order, and record the corresponding eigenvectors; select the first k eigenvalues ​​and their corresponding eigenvectors, combine these k eigenvectors to form a principal component matrix Wk, and Wk = [w1, w2, ..., wk], where w1 represents the eigenvector corresponding to the first eigenvalue, w2 represents the eigenvector corresponding to the second eigenvalue, and so on, wk represents the eigenvector corresponding to the kth eigenvalue; use the principal component matrix Wk to perform a linear transformation on the standardized eigenvector X to obtain the reduced-dimensional data feature representation Vr, and Vr = XWk.

5. The method for gas pipeline data management based on artificial intelligence according to claim 3 is characterized in that: The step S300 includes: S301. For each abnormal point path, obtain the corresponding gas pipeline abnormal point data feature representation Vr and the corresponding gas pipeline abnormal point location data; obtain the corresponding trend curve L according to the gas pipeline abnormal point data feature representation Vr, calculate the area S enclosed by the trend curve L and the coordinate axis, and associate the area S with the location coordinates of the gas pipeline abnormal point to form an associated data pair; S302.According to the associated data pair, the area S corresponding to the abnormal points of different gas pipelines on the current abnormal point path is obtained, and the size relationship of the area S is compared. The abnormal point of the gas pipeline with the largest area S is taken as the abnormal dominant point of the current abnormal point path, and the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is calculated according to the position coordinates of other gas pipeline abnormal points on the current abnormal point path; the abnormal dominant point on the current abnormal point path is taken as the center of the circle, and the maximum value of the distance between the abnormal dominant point and other gas pipeline abnormal points on the abnormal point path is taken as the radius, and the enclosed area is the abnormal dominant point radiation area; S303. Obtain the associated data pair between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point, and calculate the deviation index P between the abnormal dominant point and other gas pipeline abnormal points on the path of the abnormal point according to the associated data pair, and the specific calculation formula is: P = [(1 / H)Σ h∈[1,H] (S1-Sh)] / S0, where S1 represents the area enclosed by the trend curve L of the abnormal dominant point and the coordinate axis, Sh represents the area enclosed by the trend curve L of the hth other gas pipeline abnormal point in the current abnormal point path and the coordinate axis, and S0 represents the difference between the maximum area and the minimum area enclosed by the trend curve L and the coordinate axis corresponding to the set standard characteristic data trend curve range interval LQ; the deviation index P of each abnormal point path is associated with the radiation area area Sf of the abnormal dominant point to form a historical abnormal point path data pair; S304. Obtain historical abnormal point path data pairs of different abnormal point paths, combine with historical gas pipeline maintenance records, obtain time periods with maintenance records, and match historical abnormal point path data pairs of corresponding time periods with corresponding maintenance records.

6. The method for gas pipeline data management based on artificial intelligence according to claim 1, characterized in that: The step S400 includes: S401. Acquire real-time gas pipeline data in the monitoring area, analyze the real-time gas pipeline data at selected time intervals, and obtain corresponding real-time data features; analyze the real-time data features in the same way as step S200, thereby identifying real-time gas pipeline abnormal points, and construct real-time abnormal point paths according to the real-time gas pipeline abnormal points; S402. For each real-time abnormal point path, the corresponding real-time dominant abnormal point and the real-time abnormal dominant point radiation area Sf0 are obtained in the same analysis method as step S300; based on the real-time dominant abnormal point and the real-time abnormal dominant point radiation area, the real-time deviation index P0 of the current abnormal point path is calculated, thereby obtaining the real-time abnormal point path data pair; S403. Compare the real-time abnormal point path data pair with the historical abnormal point path data pair. If the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is greater than or equal to the similarity threshold, the real-time abnormal point path data pair, the real-time abnormal dominant point and the corresponding historical maintenance record are output to relevant personnel; if the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair is less than the similarity threshold, only the real-time abnormal point path data pair and the real-time abnormal dominant point are output, and the relevant personnel perform corresponding maintenance processing.

7. A gas pipeline data management system based on artificial intelligence, applied to a gas pipeline data management method based on artificial intelligence as claimed in any one of claims 1 to 6, characterized in that: The system includes: a data acquisition and processing module, an abnormal point identification and path construction module, a dominant point analysis and deviation calculation module, a real-time data analysis and abnormality detection module, and a historical data comparison and maintenance suggestion module; The data acquisition and processing module obtains the gas pipeline data and position data of all sensor devices in the monitoring area, records and connects the data to form a data unit; performs time series analysis on the data unit to extract data features; the abnormal point identification and path construction module analyzes the trend changes of data features and identifies the abnormal points of the gas pipeline; constructs the abnormal point path based on the position data of the abnormal points of the gas pipeline, and marks all the abnormal points of the gas pipeline; the dominant point analysis and deviation calculation module analyzes the correlation between the data features and the position data on the abnormal point path to obtain the abnormal dominant point; calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points to form the abnormal dominant point radiation area; the real-time data analysis and anomaly detection module obtains and analyzes real-time gas pipeline data, identifies real-time abnormal points and constructs real-time abnormal point paths, calculates the deviation index of the real-time abnormal point path, and compares it with historical data; the historical data comparison and maintenance suggestion module compares the real-time abnormal point path data pair with the historical abnormal point path data pair, and outputs the corresponding notification information based on the similarity threshold.

8. The artificial intelligence-based gas pipeline data management system according to claim 7, characterized in that: The data acquisition and processing module includes a data acquisition unit, a data recording unit and a data timing analysis unit; The data acquisition unit acquires gas pipeline data and sensor location data from sensor devices in the monitoring area, marks the sensor device location and its number according to the gas pipeline distribution plan, and records the location data; the data recording unit records the location data of each sensor device according to the sensor number and the gas pipeline distribution plan, and corresponds the gas pipeline data with the sensor location data to form a data unit; the data timing analysis unit performs timing analysis on each data unit, extracts and records data features.

9. The artificial intelligence-based gas pipeline data management system according to claim 7, characterized in that: The outlier identification and path construction module includes a data feature analysis unit and an outlier marking and path construction unit; The data feature analysis unit performs principal component analysis on the data features to obtain a data feature representation after dimensionality reduction; Draw a trend curve of the characteristic data changing over time and identify the trend abnormal points; the abnormal point marking and path construction unit marks the abnormal points of the gas pipeline according to the identified abnormal point position data, and connects the adjacent position data to form an abnormal point path; The dominant point analysis and deviation calculation module includes an abnormal point data feature extraction unit, a dominant point determination unit and a deviation calculation unit; The abnormal point data feature extraction unit obtains the data feature representation of each gas pipeline abnormal point in the abnormal point path, and calculates the area enclosed by its trend curve and the coordinate axis; the dominant point determination unit compares the areas of abnormal points of different gas pipelines to determine the abnormal dominant point; Calculate the deviation index between the abnormal dominant point and other abnormal points of the gas pipeline, and generate the radiation area of ​​the abnormal dominant point; The deviation calculation unit calculates the deviation index between the abnormal dominant point and other gas pipeline abnormal points, and associates it with the historical abnormal point path data.

10. The artificial intelligence-based gas pipeline data management system according to claim 7, characterized in that: The real-time data analysis and anomaly detection module includes a real-time data acquisition and feature analysis unit, a real-time anomaly point path construction unit, and a real-time dominant point and deviation calculation unit; The real-time data acquisition and feature analysis unit acquires real-time gas pipeline data, analyzes it at selected time intervals, extracts real-time data features, and identifies abnormal points in the real-time gas pipeline; the real-time abnormal point path construction unit constructs a real-time abnormal point path based on the abnormal points in the real-time gas pipeline; the real-time dominant point and deviation calculation unit performs real-time dominant abnormal point analysis, calculates the deviation index of the real-time abnormal dominant point, and generates a real-time abnormal point path data pair; The historical data comparison and maintenance suggestion module includes a historical data comparison unit and a maintenance record association and suggestion unit; The historical data comparison unit compares the similarity between the real-time abnormal point path data pair and the historical abnormal point path data pair; based on the similarity threshold, generates a comparison result; the maintenance record association and suggestion unit outputs the real-time abnormal point path data pair and the historical maintenance record according to the similarity result.

Images