Oil gas pipe network SCADA pipe network abnormity determination method and system

Abstract

The invention relates to the technical field of oil and gas pipe network governance analysis, and discloses an oil and gas pipe network SCADA pipe network abnormity determination method and system, and the method comprises the steps: obtaining the real-time data of each line node of an oil and gas pipe network, and carrying out the abnormity determination; performing security assessment and auditing on the SCADA pipe network, generating an abnormal determination period every time one vulnerability plan is completed, and recording as a normal operation period when vulnerability information does not exist; extracting an abnormal measurement result of the SCADA pipe network, processing the abnormal measurement result, diagnosing an abnormal reason, and recording the abnormal reason as a fault period; storing data of each period in a big data platform in a distributed manner, setting a weight value for each type of information in the big data platform, establishing a neural network data analysis model, updating the weight value of problem data according to an exception reason whenever a new fault period is recorded, optimizing an exception measurement rule, and obtaining a new fault period; therefore, a better SCADA pipe network abnormity determination effect can be achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas pipeline network management and analysis, and in particular to an oil and gas pipeline network SCADA pipeline network abnormality determination method and system. Background Art

[0002] Traditionally, the methods for monitoring and managing oil and gas pipeline networks mainly rely on SCADA (Supervisory Control and Data Acquisition) systems, which are responsible for real-time data collection, remote control and safety monitoring. However, with the expansion of the scale of oil and gas pipeline networks and the increase in operational complexity, the challenges faced by pipeline management systems are also increasing. Traditional SCADA systems are usually based on real-time data collection and monitoring, but when SCADA systems have version problems, they cannot effectively establish analysis models.

[0003] Therefore, there is an urgent need to invent an oil and gas pipeline SCADA pipeline anomaly determination method to solve the problem that the SCADA system version problem cannot be eliminated and it is difficult to efficiently process complex data when the network scale expands and the operation complexity increases. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for measuring anomalies in an oil and gas pipeline SCADA network, aiming to solve the problem of being unable to eliminate SCADA system version problems and the difficulty in efficiently processing complex data when the network scale expands and the operation complexity increases by optimizing the method of anomaly measurement through big data analysis.

[0005] On the one hand, an embodiment of the present invention provides a method for determining abnormalities in a SCADA pipeline network of an oil and gas pipeline network, including: step S1: obtaining real-time data of each line node of the oil and gas pipeline network and performing abnormality determination; Step S2: Conduct security assessment and audit on the SCADA network to obtain vulnerability information; Step S3: Generate a vulnerability maintenance plan based on the vulnerability information. Whenever a vulnerability plan is completed, an abnormal measurement cycle is generated. When the vulnerability information does not exist, it is recorded as a normal operation cycle. Step S4: extracting the abnormal measurement results of the SCADA pipe network, processing the abnormal measurement results, diagnosing the abnormal cause, correcting it, and recording the corrected cause as a fault cycle; Step S5: Distribute and store the data of each period in the big data platform, set a weight value for each type of information in the big data platform, and establish a neural network data analysis model; Step S6: Whenever a new fault cycle is recorded, the weight value of the problem data is updated according to the abnormal cause, and the abnormality determination rule is optimized.

[0006] Preferably, in step S1, the real-time data includes real-time operation data and real-time environmental data of the node, and the abnormality measurement includes a variety of operation abnormalities measured based on the relationship between the real-time operation data and a variety of preset alarm thresholds and a variety of environmental abnormalities set based on the environmental data and the concentration of combustible gas volume around the oil and gas pipeline, and the safety distance between the oil and gas pipeline and the ignition source.

[0007] Preferably, in step S2, the security assessment and audit of the SCADA network management includes: using vulnerability scanning tools and penetration testing techniques to assess the SCADA network; reviewing the security policies, configuration files and permission settings of the SCADA network; auditing the communication and network environment of the SCADA network; checking the configuration of network equipment and firewalls, network traffic and the security of communication protocols; assessing the physical security measures of the SCADA network, including physical access control of equipment, security of computer rooms and monitoring systems; auditing the log records and event response capabilities of the SCADA network, checking the integrity and accuracy of the logs, and assessing the detection and response capabilities of events.

[0008] Preferably, in step S3, a vulnerability maintenance plan with a time period and vulnerability level is generated according to the fluctuation range of the impact of the vulnerability information on the SCADA network and the type of error data, so as to ensure the security and stability of the SCADA network.

[0009] Preferably, in step S4, the result of the abnormal measurement is processed, and the cause of the abnormality is diagnosed and corrected, including: when an operational abnormality occurs, obtaining the abnormal data of the node, adjusting one or more dependent variable data of the abnormal data so that the data is lower than the alarm threshold, and checking the cause of the abnormality of the abnormal data, and performing corrective processing; when an environmental abnormality occurs, the severity of the environment of the node is judged as soon as possible, when the severity exceeds the standard, an emergency response is implemented and human intervention is immediately carried out on site, and when the severity does not exceed the standard, the node is automatically operated for humidity control or ventilation treatment.

[0010] Preferably, in step S5, the data of each line node in each normal operation cycle is distributedly stored in the big data platform by generating index nodes with different line nodes through data mapping; the data of each line node in each abnormal measurement cycle is distributedly stored in the big data platform by generating index nodes with vulnerability information through data mapping; and the index nodes are distributedly stored in the big data platform by generating index nodes with abnormal causes for each fault cycle.

[0011] Preferably, in step S5, multiple additional fields are added to the data in each index to store weight values, and the data is processed according to the correlation of each weight value to generate a multi-dimensional database, a data analysis model is established for the database of each dimension, and the anomaly determination method is optimized according to the analysis results.

[0012] Preferably, in step S5, the multiple additional fields are used to store multiple weighting rules, the first weighting rule selects the first data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the first data in sequence according to the importance of the feature, and calculates accordingly, and adds the first field to each data using the first weighting rule; the second weighting rule selects the second data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the second data in sequence according to the importance of the feature, and calculates accordingly, and adds the second field to each data using the second weighting rule; the third weighting rule selects the third data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the third data in sequence according to the importance of the feature, and calculates accordingly, and adds the third field to each data using the third weighting rule; wherein, the nth weighting rule selects the nth data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the nth data in sequence according to the importance of the feature, and calculates accordingly, and adds the nth field to each data using the nth weighting rule, and generates a multi-dimensional database after sorting.

[0013] Preferably, in step S6, the weight value of the problem data is updated, wherein only the problem weight rule with the problem data as the key value is updated, and the gradient weight method is used to sort the corresponding problem weight rules of all the updated data in sequence according to the importance of the features, and the alarm threshold of the data in abnormal measurement is optimized according to the key value assignment object in the problem weight rule, and one or more dependent variable data related to the problem data are optimized one by one according to the problem weight rule.

[0014] On the other hand, an oil and gas pipeline network SCADA pipeline network anomaly detection system is also provided in an embodiment of the present invention, which is applicable to an oil and gas pipeline network SCADA pipeline network anomaly detection method as described in the above-mentioned embodiments of the present invention, and includes: Monitoring module, which obtains real-time data of each line node of the oil and gas network management and performs abnormality detection; Security module, which conducts security assessment and audit on SCADA network and obtains vulnerability information; A security maintenance module generates a vulnerability maintenance plan according to the vulnerability information, generates an abnormal measurement cycle each time the vulnerability plan is completed, and records the absence of vulnerability information as a normal operation cycle; The operation and maintenance module extracts the abnormal measurement results of the SCADA pipe network, processes the abnormal measurement results, diagnoses the abnormal cause, corrects it, and records the corrected cause as a fault cycle; The data storage module stores the data of each period in a distributed manner on the big data platform, sets a weight value for each type of information in the big data platform, and establishes a neural network data analysis model; The data processing module updates the weight value of the problem data according to the abnormal cause every time a new fault cycle is recorded, and optimizes the abnormality determination rules.

[0015] Compared with the prior art, the oil and gas pipeline network SCADA pipeline network anomaly detection method and system according to the embodiment of the present invention have the following beneficial effects: 1. By acquiring real-time data of each line node of the oil and gas network and performing abnormal detection, abnormal conditions in the pipeline network, such as failures, leaks, etc., can be discovered in a timely manner. This helps to warn and take measures in advance to prevent accidents and ensure the normal operation of the pipeline network.

[0016] 2. By conducting security assessments and audits on the SCADA network and obtaining vulnerability information, security vulnerabilities in the network can be discovered in a timely manner, such as weaknesses, attack surfaces, etc. This helps to strengthen the security of the network and take corresponding repair and protection measures to protect the network from malicious attacks.

[0017] 3. Generate a vulnerability maintenance plan based on vulnerability information and repair the vulnerabilities according to the plan. This helps to systematically manage and maintain the safety of the pipeline network, ensure timely repair of vulnerabilities and reduce potential risks.

[0018] 4. Extract the results of abnormal measurements, process and diagnose them, find out the causes of the abnormalities and make corrections, which will help to promptly resolve faults that occur during pipeline network operation and ensure the stable operation of the pipeline network.

[0019] 5. Distributing and storing data from each period on a big data platform can provide a more efficient way of storing and managing data. By setting weights for each type of information on the big data platform and establishing a neural network data analysis model, data can be better utilized for analysis and prediction to discover hidden patterns and trends.

[0020] 6. Every time a new fault cycle is recorded, the weight value of the problem data is updated according to the abnormal cause, and the abnormality determination rules are optimized. This helps to continuously optimize and improve the accuracy and reliability of anomaly detection and improve the system's adaptive ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flowchart of a method for determining abnormality of a SCADA pipeline network of an oil and gas pipeline network according to an embodiment of the present invention; Figure 2 It is a structural block diagram of an oil and gas pipeline network SCADA pipeline network anomaly detection system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0023] like Figure 1 As shown, a method for determining abnormality of a SCADA pipeline network in an oil and gas pipeline network according to an embodiment of the present invention includes: Step S1: Acquire real-time data of each line node of the oil and gas network management and perform abnormality detection; Step S2: Conduct security assessment and audit on the SCADA network to obtain vulnerability information; Step S3: Generate a vulnerability maintenance plan based on the vulnerability information. Whenever a vulnerability plan is completed, an abnormal measurement cycle is generated. When the vulnerability information does not exist, it is recorded as a normal operation cycle. Step S4: extracting the abnormal measurement results of the SCADA pipe network, processing the abnormal measurement results, diagnosing the abnormal cause, correcting it, and recording the corrected cause as a fault cycle; Step S5: Distribute and store the data of each period in the big data platform, set a weight value for each type of information in the big data platform, and establish a neural network data analysis model; Step S6: updating the weights in real time every time the data is updated; updating the weights of the problem data according to the abnormal causes every time a new fault cycle is recorded, and optimizing the abnormal determination rules; Step S7: Whenever a new abnormal measurement cycle is recorded, the vulnerability information is analyzed according to each data, and the information in the big data platform is maintained or corrected according to the analysis results.

[0024] In some embodiments of the present application, in step S1, the real-time data includes real-time operation data and real-time environmental data of the node, and the abnormality measurement includes a variety of operation anomalies measured based on the relationship between the real-time operation data and a variety of preset alarm thresholds and a variety of environmental anomalies set based on the environmental data and the concentration of combustible gas volume around the oil and gas pipeline, and the safety distance between the oil and gas pipeline and the ignition source.

[0025] In some embodiments of the present application, in step S2, the security assessment and audit of the SCADA network management includes: using vulnerability scanning tools and penetration testing techniques to assess the SCADA network; reviewing the security policies, configuration files, and permission settings of the SCADA network; auditing the communication and network environment of the SCADA network; checking the configuration of network equipment and firewalls, network traffic, and the security of communication protocols; assessing the physical security measures of the SCADA network, including physical access control of equipment, security of computer rooms, and monitoring systems; auditing the logging and event response capabilities of the SCADA network, checking the integrity and accuracy of logs, and assessing event detection and response capabilities.

[0026] In some embodiments of the present application, in step S3, based on the fluctuation range of the impact of the vulnerability information on the SCADA network and the type of error data, a vulnerability maintenance plan with a time period and vulnerability level is generated to ensure the security and stability of the SCADA network.

[0027] In some embodiments of the present application, in step S4, the results of the abnormal measurement are processed, and the cause of the abnormality is diagnosed and corrected, including: when an operational abnormality occurs, obtaining the abnormal data of the node, adjusting one or more dependent variable data of the abnormal data so that the data is lower than the alarm threshold, and checking the cause of the abnormality of the abnormal data, and performing corrective processing; when an environmental abnormality occurs, the severity of the environment of the node is judged as soon as possible, when the severity exceeds the standard, an emergency response is implemented and human intervention is immediately carried out on site, and when the severity does not exceed the standard, the node is automatically operated for humidity control or ventilation.

[0028] In some embodiments of the present application, in step S5, the data of each line node in each normal operation cycle is distributedly stored in the big data platform by generating index nodes with different line nodes through data mapping; the data of each line node in each abnormal measurement cycle is distributedly stored in the big data platform by generating index nodes with vulnerability information through data mapping; and the index nodes for each fault cycle are distributedly stored in the big data platform by generating index nodes with abnormal causes.

[0029] In some embodiments of the present application, in step S5, multiple additional fields are added to the data in each index for storing weight values, and the data is processed according to the correlation of each weight value to generate a multi-dimensional database, a data analysis model is established for the database of each dimension, and the anomaly determination method is optimized based on the analysis results.

[0030] In some embodiments of the present application, in step S5, the multiple additional fields are used to store multiple weighting rules, the first weighting rule selects the first data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the first data in sequence according to the importance of the feature, and calculates accordingly, and adds the first field to each data using the first weighting rule; the second weighting rule selects the second data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the second data in sequence according to the importance of the feature, and calculates accordingly, and adds the second field to each data using the second weighting rule; the third weighting rule selects the third data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the third data in sequence according to the importance of the feature, and calculates accordingly, and adds the third field to each data using the third weighting rule; wherein, the nth weighting rule selects the nth data in the real-time operation data as the key assignment, and uses the gradient weight method to sort the dependent variable data of the nth data in sequence according to the importance of the feature, and calculates accordingly, and adds the nth field to each data using the nth weighting rule, and generates a multi-dimensional database after sorting.

[0031] Specifically, according to the calculation rules of the neural network, when the nth data and its dependent variable data are in a proportional relationship: assuming that the nth data, i.e. the independent variable, is X, and the dependent variable is Y, and there is a proportional relationship between them; use the following formula to calculate the gradient weight: a) Forward propagation: During the forward propagation process, the activation value A_i of each layer is calculated: A_i=f(\sum_{j=1}^{n}W_{ij}\cdot A_{i-1}+b_i) Among them, Wi represents the connection weight, A_{i-1} represents the activation value of the previous layer, bi represents the bias term, and f represents the activation function.

[0032] b) Backward Propagation: During back-propagation, the gradient of each layer is calculated: \frac{\partial L}{\partial A_i}=\frac{\partial L}{\partial A_{i+1}}\cdot\frac{\partial A_{i+1}}{\partial A_i} Among them, L represents the loss function and A_{i+1} represents the activation value of the next layer.

[0033] According to the proportional relationship, \frac{\partial L}{\partial A_i} can be expressed as: \frac{\partial L}{\partial A_i}=k\cdot\frac{\partial L}{\partial Y}\cdot\frac{\partial Y}{\partial A_i} Among them, k is the proportional coefficient, \frac{\partial L}{\partial Y} represents the partial derivative of the loss function with respect to the dependent variable Y, and \frac{\partial Y}{\partial A_i} represents the partial derivative of the dependent variable Y with respect to the activation value A_i.

[0034] According to the calculation rules of the neural network, when the nth data is in direct proportion to its dependent variable data: Specifically, assuming that the nth data, i.e., the independent variable, is X, and the dependent variable is Y, and there is an inverse relationship between them; the following formula is used to calculate the gradient weight: a) Forward propagation: During the forward propagation process, the activation value A_i of each layer is calculated: A_i=f(\sum_{j=1}^{n}W_{ij}\cdot A_{i-1}+b_i) Among them, Wi represents the connection weight, A_{i-1} represents the activation value of the previous layer, bi represents the bias term, and f represents the activation function.

[0035] b) Backward Propagation: During back-propagation, the gradient of each layer is calculated: \frac{\partial L}{\partial A_i}=\frac{\partial L}{\partial A_{i+1}}\cdot\frac{\partial A_{i+1}}{\partial A_i} Among them, L represents the loss function and A_{i+1} represents the activation value of the next layer.

[0036] According to the inverse relationship, \frac{\partial L}{\partial A_i} can be expressed as: \frac{\partial L}{\partial A_i}=-k\cdot\frac{\partial L}{\partialY}\cdot\frac{\partial Y}{\partial A_i} Among them, k is the proportional coefficient, \frac{\partial L}{\partial Y} represents the partial derivative of the loss function with respect to the dependent variable Y, and \frac{\partial Y}{\partial A_i} represents the partial derivative of the dependent variable Y with respect to the activation value A_i.

[0037] It can be understood that a weight sharing technique is used to set the gradient weight so that a main data has a greater impact on other data; in some cases, it may be desirable to give higher weights to certain specific data samples so that more attention is paid to these samples when training the model, which is very useful when dealing with unbalanced data sets or paying attention to the importance of specific samples. A weighted loss function is then used, in which the loss value of each sample is multiplied by a corresponding weight. In this way, by adjusting the weights, the sensitivity of the model to different samples can be controlled.

[0038] Specifically, the weight value update method is: Use the gradient descent algorithm to update the connection weights Wij: W_{ij}^{new}=W_{ij}^{old}-\eta\cdot\frac{\partial L}{\partial W_{ij}} Among them, \eta represents the learning rate, and \frac{\partial L}{\partial W_{ij}} represents the partial derivative of the connection weight.

[0039] Specifically, in the gradient calculation process, the gradient of each sample can be multiplied by the corresponding weight. In this way, the influence of the gradient will be greater on samples with higher weights, thereby affecting the update of model parameters more. This method can make the model pay more attention to samples with higher weights and make more targeted adjustments to them.

[0040] In some embodiments of the present application, in step S6, the weight value of the problem data is updated, wherein only the problem weight rule that focuses on the problem data is updated, and the gradient weight method is used to sort the corresponding problem weight rules of all updated data in sequence according to the importance of the features, and the alarm threshold of the data in abnormal measurement is optimized according to the focus assignment objects in the problem weight rules, and one or more dependent variable data related to the problem data are optimized one by one according to the problem weight rules.

[0041] In some embodiments of the present application, in step S7, the fluctuation range, error data type and correction parameters are obtained according to the vulnerability maintenance plan, and the historical storage data is compared and analyzed to see whether there are errors. If there are errors, the information in the big data platform is corrected; if there are no errors, the information in the big data platform is maintained.

[0042] Specifically, obtain relevant information on the fluctuation range, error data type and correction parameters from the vulnerability maintenance plan. This information may include the expected data change range, possible error types and corresponding correction parameters; obtain historical storage data from the big data platform and analyze it. Data analysis tools can be used or custom scripts can be written to process the data. The purpose of the analysis is to determine whether there are errors in the historical data; compare and analyze the fluctuation range and error data type in the vulnerability maintenance plan with the historical data. For each data point, check whether its value is within the expected fluctuation range and whether it matches the expected error data type; if errors are found in the historical data, correct the information in the big data platform according to the correction parameters in the vulnerability maintenance plan; this may involve correcting erroneous values ​​in the data, deleting erroneous data or repairing the data by other means; if no errors are found in the historical data, keep the information in the big data platform unchanged without making any modifications.

[0043] On the other hand, see Figure 2 As shown, an oil and gas pipeline network SCADA pipeline network anomaly detection system is also provided in an embodiment of the present invention, which is applicable to an oil and gas pipeline network SCADA pipeline network anomaly detection method as in the above-mentioned embodiments of the invention, including: Monitoring module, which obtains real-time data of each line node of the oil and gas network management and performs abnormality detection; Security module, which conducts security assessment and audit on SCADA network and obtains vulnerability information; The security maintenance module generates a vulnerability maintenance plan based on vulnerability information. Each time a vulnerability plan is completed, an abnormal measurement cycle is generated. When vulnerability information does not exist, it is recorded as a normal operation cycle. The operation and maintenance module extracts the abnormal measurement results of the SCADA pipe network, processes the abnormal measurement results, diagnoses the abnormal causes, makes corrections, and records the corrections as a fault cycle; The data storage module stores the data of each period in a distributed manner on the big data platform, sets a weight value for each type of information in the big data platform, and establishes a neural network data analysis model; The data processing module updates the weights instantly every time the data is updated; every time a new fault cycle is recorded, the weight values ​​of the problem data are updated according to the abnormal cause, and the abnormality determination rules are optimized; The data processing module also includes analyzing the vulnerability information according to each data each time a new abnormal measurement cycle is recorded, and maintaining or correcting the information in the big data platform according to the analysis results.

[0044] It can be understood that an oil and gas pipeline network SCADA pipeline network anomaly measurement system in each of the above-mentioned invention embodiments is applicable to an oil and gas pipeline network SCADA pipeline network anomaly measurement method in each of the above-mentioned invention embodiments. Therefore, an oil and gas pipeline network SCADA pipeline network anomaly measurement system in each of the above-mentioned invention embodiments and an oil and gas pipeline network SCADA pipeline network anomaly measurement method have the same beneficial effects, and therefore will not be repeated.

[0045] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0046] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. 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 method for measuring abnormality of an oil and gas pipeline network SCADA pipeline network, characterized in that: include: Step S1: Acquire real-time data of each line node of the oil and gas network management and perform abnormality detection; Step S2: Conduct security assessment and audit on the SCADA network to obtain vulnerability information; Step S3: Generate a vulnerability maintenance plan based on the vulnerability information. Whenever a vulnerability maintenance plan is completed, an abnormal measurement cycle is generated. When the vulnerability information does not exist, it is recorded as a normal operation cycle. Step S4: extracting the abnormal measurement results of the SCADA pipe network, processing the abnormal measurement results, diagnosing the abnormal cause, correcting it, and recording the corrected cause as a fault cycle; Step S5: Distribute and store the data of each period in the big data platform, set a weight value for each type of information in the big data platform, and establish a neural network data analysis model; Step S6: Whenever a new fault cycle is recorded, the weight value of the problem data is updated according to the abnormal cause, and the abnormality determination rule is optimized.

2. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 1, characterized in that: In step S1, the real-time data includes the real-time operation data and real-time environmental data of the node, and the abnormality measurement includes a variety of operation abnormalities measured based on the relationship between the real-time operation data and a variety of preset alarm thresholds, and a variety of environmental abnormalities set based on the environmental data and the concentration of combustible gas volume around the oil and gas pipeline, and the safety distance between the oil and gas pipeline and the ignition source.

3. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 1, characterized in that: In step S2, the security assessment and audit of the SCADA network management includes: using vulnerability scanning tools and penetration testing technology to assess the SCADA network; reviewing the security policies, configuration files and permission settings of the SCADA network; auditing the communication and network environment of the SCADA network; checking the configuration of network equipment and firewalls, network traffic and the security of communication protocols; assessing the physical security measures of the SCADA network, including physical access control of equipment, security of computer rooms and monitoring systems; auditing the log records and event response capabilities of the SCADA network, checking the integrity and accuracy of the logs, and assessing the detection and response capabilities of events.

4. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 3, characterized in that: The step S3 also includes: generating a vulnerability maintenance plan with a time period and vulnerability level according to the impact fluctuation range of the vulnerability information on the SCADA network and the type of error data, so as to ensure the security and stability of the SCADA network.

5. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 1, characterized in that: In the step S4, the result of the abnormal measurement is processed, and the cause of the abnormality is diagnosed and corrected, including: when an operational abnormality occurs, the abnormal data of the node is obtained, one or more dependent variable data of the abnormal data is adjusted to make the abnormal data lower than the alarm threshold, and the abnormal cause of the abnormal data is checked and corrected; when an environmental abnormality occurs, the severity of the environment of the node is judged as soon as possible, and when the severity exceeds the standard, an emergency response is implemented and manual on-site intervention is carried out; when the severity does not exceed the standard, the node is automatically operated for humidity control or ventilation treatment.

6. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 5, characterized in that: The step S5 also includes: generating index nodes with different line nodes through data mapping for data of each line node in each normal operation cycle and distributing and storing them on the big data platform; generating index nodes with vulnerability information for data of each line node in each abnormal measurement cycle through data mapping for distributed storage on the big data platform; generating index nodes with abnormal causes for each fault cycle and distributing and storing them on the big data platform.

7. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 6, characterized in that: In step S5, a plurality of additional fields are added to the data in each index for storing weight values, and the data is processed according to the relevance of each weight value to generate a multi-dimensional database, a data analysis model is established for the database of each dimension, and the abnormality determination method is optimized according to the analysis results.

8. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 7, characterized in that: In step S5, the multiple additional fields are used to store multiple weighting rules, the first weighting rule selects the first data in the real-time operation data as the key value assignment, and uses the gradient weight method to sort the dependent variable data of the first data in sequence according to the importance of the feature, and calculates accordingly, and adds the first field to each data using the first weighting rule; the second weighting rule selects the second data in the real-time operation data as the key value assignment, and uses the gradient weight method to sort the dependent variable data of the second data in sequence according to the importance of the feature, and calculates accordingly, and adds the second field to each data using the second weighting rule; the third weighting rule selects the third data in the real-time operation data as the key value assignment, and uses the gradient weight method to sort the dependent variable data of the third data in sequence according to the importance of the feature, and calculates accordingly, and adds the third field to each data using the third weighting rule; wherein, the nth weighting rule selects the nth data in the real-time operation data as the key value assignment, and uses the gradient weight method to sort the dependent variable data of the nth data in sequence according to the importance of the feature, and calculates accordingly, and adds the nth field to each data using the nth weighting rule; Among them, according to the calculation rules of the neural network, the following formula is used to calculate the gradient weight: a) Forward propagation: During the forward propagation process, the activation value A_i of each layer is calculated: A_i=f(\sum_{j=1}^{n}W_{ij}\cdot A_{i-1}+b_i); Among them, Wij represents the connection weight, A_i-1 represents the activation value of the previous layer, bi represents the bias term, and f represents the activation function; b) Backward Propagation: During back-propagation, the gradient of each layer is calculated: \frac{\partial L}{\partial A_i}=\frac{\partial L}{\partial A_{i+1}}\cdot\frac{\partial A_{i+1}}{\partial A_i}; Where L represents the loss function, and A_{i+1} represents the activation value of the next layer; Assume that the nth data, i.e. the independent variable, is X, the dependent variable is Y, and there is a proportional relationship between them; According to the proportional relationship, \frac{\partial L}{\partial A_i} can be expressed as: \frac{\partial L}{\partial A_i}=k\cdot\frac{\partial L}{\partial Y}\cdot\frac{\partial Y}{\partial A_i}; Where k is the proportionality coefficient, \frac{\partial L}{\partial Y} represents the partial derivative of the loss function with respect to the dependent variable Y, and \frac{\partial Y}{\partial A_i} represents the partial derivative of the dependent variable Y with respect to the activation value A_i; Assume that the nth data, i.e., the independent variable, is X, the dependent variable is Y, and there is an inverse relationship between them; According to the inverse relationship, \frac{\partial L}{\partial A_i} can be expressed as: \frac{\partial L}{\partial A_i}=-k\cdot\frac{\partial L}{\partial Y}\cdot\frac{\partial Y}{\partial A_i}; Among them, k is the proportional coefficient, \frac{\partial L}{\partial Y} represents the partial derivative of the loss function with respect to the dependent variable Y, and \frac{\partial Y}{\partial A_i} represents the partial derivative of the dependent variable Y with respect to the activation value A_i.

9. The oil and gas pipeline network SCADA pipeline network abnormality determination method according to claim 8, characterized in that: In the step S6, the weight value of the problem data is updated, wherein only the problem weight rule for the problem data as the key value is updated, and the problem weight rules for all the updated data are calculated by sorting them in sequence according to the importance of the features using the gradient weight method, and the alarm threshold of the data in the abnormal determination is optimized according to the key value assignment object in the problem weight rule, and one or more dependent variable data related to the problem data are optimized one by one according to the problem weight rule; Among them, the algorithm for updating the weight value is: Use the gradient descent algorithm to update the connection weights Wij: W_{ij}^{new}=W_{ij}^{old}-\eta\cdot\frac{\partial L}{\partial W_{ij}}; Among them, W_{ij}^{new} represents the updated connection weight, W_{ij}^{old} represents the connection weight before update, \eta represents the learning rate, and \frac{\partial L}{\partial W_{ij}} represents the partial derivative of the loss function with respect to the connection weight.

10. An oil and gas pipeline network SCADA pipeline network anomaly detection system, applicable to an oil and gas pipeline network SCADA pipeline network anomaly detection method as claimed in any one of claims 1 to 9, characterized in that: include: Monitoring module, which obtains real-time data of each line node of the oil and gas network management and performs abnormality detection; Security module, which conducts security assessment and audit on SCADA network and obtains vulnerability information; A security maintenance module generates a vulnerability maintenance plan according to the vulnerability information, generates an abnormal measurement cycle each time the vulnerability plan is completed, and records the absence of vulnerability information as a normal operation cycle; The operation and maintenance module extracts the abnormal measurement results of the SCADA pipe network, processes the abnormal measurement results, diagnoses the abnormal cause, corrects it, and records the corrected cause as a fault cycle; The data storage module stores the data of each period in a distributed manner on the big data platform, sets a weight value for each type of information in the big data platform, and establishes a neural network data analysis model; The data processing module updates the weight value of the problem data according to the abnormal cause every time a new fault cycle is recorded, and optimizes the abnormality determination rules.

Images