Petrochemical engineering pressure pipeline management system based on artificial intelligence

Abstract

The embodiment of the invention provides a petrochemical engineering pressure pipeline management system based on artificial intelligence, and the system is characterized in that a management platform is used for verifying the identity information of a supervisor and issuing a control instruction; the pipe network simulation module is used for constructing a pipe network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the pressure pipeline preset data; the pipeline analysis module is used for distinguishing an abnormal pipeline based on the pipe network operation model and a pre-trained fault analysis neural network model, and positioning and marking a pipeline fault position; the risk alarm module is used for sending risk alarm information; the interruption module is used for interrupting the transportation process of the corresponding pipeline according to the pipeline fault position. According to the embodiment of the invention, the method achieves the automation of the abnormality judgment and fault processing of the pressure pipeline, improves the management efficiency of the pressure pipeline, reduces the management cost, and improves the accuracy and reliability of the abnormality diagnosis of the pressure pipeline through the construction of the pipe network operation model and the fault analysis neural network model.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of computer and automation technology, and in particular to a petrochemical pressure pipeline management system based on artificial intelligence. Background Art

[0002] Petrochemical pressure pipelines are the arteries of the national economy and occupy an important position in the national economy. They are indispensable special equipment in modern industrial production. They are widely used in the fields of petrochemicals, energy, metallurgy, power environmental protection, aviation, medicine and food. They are responsible for transporting oil and natural gas, high temperature and high pressure, flammable and explosive, highly toxic, corrosive or radioactive materials. They have become one of the modes of transportation alongside railways, roads, waterways and air transportation. With the acceleration of my country's modernization and informatization process, the country's various industries are becoming more and more dependent on petrochemical products. As the main transportation equipment for petrochemical enterprises' production equipment, pressure pipelines are developing towards large-scale and intensive equipment construction, as well as the application of new technologies and new processes and the implementation of better environmental protection standards. The laying of new pressure pipelines is more intensive and complex, and the operating conditions of pipelines are more demanding, which brings great difficulties to the operation safety management and safety supervision of pressure pipelines.

[0003] In the current petrochemical pressure pipeline management system, the abnormal judgment and processing of pressure pipelines are usually done manually, which consumes huge time and economic costs. Other pressure pipeline management systems simply perform abnormal diagnosis based on the measurement results of pressure sensors, which has the problem of inaccurate and unreliable diagnosis. Summary of the invention

[0004] In view of the defects of the prior art, an embodiment of the present invention provides a petrochemical pressure pipeline management system based on artificial intelligence.

[0005] The embodiment of the present invention provides a petrochemical pressure pipeline management system based on artificial intelligence, comprising: an intelligent sensor, a management platform, a pipeline network simulation module, a pipeline analysis module, a risk warning module, an interruption module and a terminal device; wherein: the intelligent sensor is used to collect preset data of the pressure pipeline and send it to the management platform and the pipeline network simulation module; the management platform is used to verify the identity information of the supervisor, and issue control instructions according to the operation information of the supervisor, and send the preset data of the pressure pipeline to the terminal device; the pipeline network simulation module is used to obtain the pipeline distribution engineering drawing, the pipeline picture and the pipeline specification information from the management platform, and receive the The preset data of the pressure pipeline sent by the intelligent sensor is used to build a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the preset data of the pressure pipeline; the pipeline analysis module is used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, locate and mark the pipeline fault position, and send the pipeline fault position to the interruption module and the risk warning module; the risk warning module is used to send risk warning information to the management platform and the terminal device according to the pipeline fault position; the interruption module is used to interrupt the transportation process of the corresponding pipeline according to the pipeline fault position.

[0006] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, when the pipeline network simulation module is used to construct a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the pressure pipeline preset data, it is specifically used to: block the pipeline picture according to the display ratio of the pipeline picture, extract high-frequency components and smooth the pipeline picture after blocking; calculate the grayscale average value of the pipeline picture after blocking, compare the grayscale value of the pixels of the pipeline picture after blocking with the grayscale average value, divide the pixels with grayscale values ​​greater than the grayscale average value into segmentation targets, divide the pixels with grayscale values ​​less than the grayscale average value into segmentation backgrounds, perform target separation according to the classification results of the segmentation targets and the segmentation background, and obtain the pipeline picture separated from the background; construct a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture separated from the background, the pipeline specification information and the pressure pipeline preset data; wherein the pressure pipeline preset data is annotated at the corresponding pipeline position as annotation information.

[0007] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, when the pipeline analysis module is used to identify abnormal pipelines based on the pipeline network operation model and a pre-trained fault analysis neural network model, it is specifically used to: input the preset data of the pressure pipeline into the pipeline network operation model to obtain a pressure pipeline operation prediction curve, and identify abnormal pipelines according to the pressure pipeline operation prediction curve.

[0008] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, before the pipeline analysis module is used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, it is also used to: extract pressure pipeline accident data from the pipeline accident database as a sample data set, and eliminate abnormal data in the sample data set by calculating the standard deviation; standardize and normalize the remaining sample data set, and divide the processed sample data set into test set samples and training set samples; input the training set samples into a convolutional neural network, and use the least squares recursive method to define the energy function of multiple rounds of learning of the convolutional neural network. When the value of the energy function is less than the target error, the training process ends and the fault analysis neural network model is output; the test set samples are input into the fault analysis neural network model for testing, and the parameters of the fault analysis neural network model are updated.

[0009] According to an embodiment of the present invention, a petrochemical pressure pipeline management system based on artificial intelligence is provided, the system also includes a performance optimization module, the performance optimization module is used to: generate a corresponding startup list for each group of functional interfaces of the management platform, determine the order of each functional interface in the startup list according to the least recent use, and further link each group of startup lists in order of the number of accesses from small to large; update the order of each functional interface in each group of startup lists and the link order of each group of startup lists in real time according to the real-time interaction information of each group of functional interfaces; select the functional page with the least number of accesses from the head of each startup list in turn as the page to be recycled according to the number of accesses from small to large according to the preset time interval; merge the selected pages to be recycled into a block and mark them, wake up the compression driver to parse the marked block, and obtain the physical page belonging to the block, copy the physical page to a buffer, call the compression algorithm to compress the physical page in the buffer into a compressed block, and store the compressed block in the compression area of ​​the performance optimization module.

[0010] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the system also includes a log analysis module, which is used to: obtain log data recorded in the management platform through a log collection plug-in or a syslog server deployed on the management platform, filter out log data to be analyzed based on the log data, and process the log data to be analyzed into a preset unified format; obtain user operation behavior recorded in the processed log data to be analyzed, extract user behavior features according to the user operation behavior, match the user behavior features with abnormal behavior features, and obtain a matching result; in response to the matching result being greater than a preset matching threshold, generate corresponding alarm information according to the matching result; send the alarm information to the management platform, and send it to the terminal device through the management platform.

[0011] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module is also used to generate an alarm risk score for the alarm information and add the alarm risk score to the alarm information; wherein the alarm risk score is expressed as:

[0012]

[0013] Where S represents the alarm risk score, R1 represents the importance of abnormal control instructions, R2 represents the number of historical hits of abnormal control instructions, R3 represents the number of hits of abnormal control instructions, and A S Indicates the importance of assets, V i It indicates the severity of the vulnerability caused by abnormal control instructions, and n indicates the total number of abnormal control instructions.

[0014] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module is also used to send the alarm information to the interrupt module so that the interrupt module can terminate the response process of the relevant abnormal control instructions.

[0015] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module is also used to record the IP address and user information of the device associated with the alarm information and send them to the management platform.

[0016] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the terminal device includes a smart phone, a tablet computer, a laptop computer or a desktop computer.

[0017] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention collects preset data of the pressure pipeline, builds a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the preset data of the pressure pipeline, identifies abnormal pipelines based on the pipeline network operation model and a pre-trained fault analysis neural network model, locates and marks the pipeline fault position and automatically interrupts the transportation process of the corresponding pipeline, realizes the automation of pressure pipeline abnormality judgment and fault processing, improves the management efficiency of the pressure pipeline, reduces the management cost, and at the same time, improves the accuracy and reliability of pressure pipeline abnormality diagnosis by building a pipeline network operation model and a fault analysis neural network model. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 This is one of the structural schematic diagrams of the petrochemical pressure pipeline management system based on artificial intelligence provided by an embodiment of the present invention;

[0020] Figure 2 This is the second structural schematic diagram of the petrochemical pressure pipeline management system based on artificial intelligence provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0022] Figure 1 This is one of the structural diagrams of the petrochemical pressure pipeline management system based on artificial intelligence provided by the embodiment of the present invention. Figure 1As shown, the system includes an intelligent sensor 1, a management platform 2, a pipe network simulation module 3, a pipeline analysis module 4, a risk warning module 5, an interruption module 6 and a terminal device 7; the management platform 2 is connected to the intelligent sensor 1, the pipe network simulation module 3, the pipeline analysis module 4, the risk warning module 5, the interruption module 6 and the terminal device 7, the intelligent sensor 1 is also connected to the pipe network simulation module 3, the pipe network simulation module 3 is also connected to the pipeline analysis module 4, the pipeline analysis module 4 is also connected to the risk warning module 5, and the risk warning module 5 is also connected to the terminal device 7.

[0023] The intelligent sensor 1 is used to collect preset data of the pressure pipeline and send it to the management platform 2 and the pipe network simulation module 3; there can be one or more intelligent sensors 1. The preset data of the pressure pipeline includes the operation data of the pipeline, the pressure inside and outside the pipeline, and the basic parameter data of the pipeline. Different intelligent sensors 1 can be set according to different test or measurement requirements.

[0024] The management platform 2 is used to verify the identity information of the supervisor, issue control instructions based on the supervisor's operation information, and send the pressure pipeline preset data to the terminal device 7.

[0025] The pipe network simulation module 3 is used to obtain the pipeline distribution engineering drawing, pipeline pictures and pipeline specification information from the management platform 2, and receive the preset data of the pressure pipeline sent by the intelligent sensor 1, and build a pipe network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the preset data of the pressure pipeline. The pipeline distribution engineering drawing is uploaded to the management platform 2 by the staff in advance, and the pipe network simulation module 3 obtains the pipeline distribution engineering drawing by communicating with the management platform 2. The pipe network simulation module 3 can obtain the pipeline picture and pipeline specification information by communicating with the management platform 2, or it can collect the pipeline picture and pipeline specification information by itself or through other channels. The pipe network simulation module 3 builds a pipe network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the preset data of the pressure pipeline.

[0026] The pipeline analysis module 4 is used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, locate and mark the pipeline fault position, and send the pipeline fault position to the interruption module 6 and the risk alarm module 5. The pipeline analysis module obtains the pressure pipeline preset data of the pressure pipeline at each position based on the pipeline network operation model constructed by the pipeline simulation module 3, performs abnormal analysis on the pressure pipeline based on the pressure pipeline preset data and the pre-trained fault analysis neural network model, identifies abnormal pipelines, and locates and marks the pipeline fault position.

[0027] The risk warning module 5 is used to send risk warning information to the management platform 2 and the terminal device 7 according to the pipeline fault location. The risk warning module 5 can send risk warning information to the terminal device 7 using different communication methods according to different types of the terminal device 7.

[0028] The interruption module 6 is used to interrupt the transportation process of the corresponding pipeline according to the pipeline fault position. The interruption module 6 is used to interrupt the transportation process of the abnormal pipeline and feed back the interruption information to the staff.

[0029] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention collects preset data of the pressure pipeline, builds a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the preset data of the pressure pipeline, identifies abnormal pipelines based on the pipeline network operation model and a pre-trained fault analysis neural network model, locates and marks the pipeline fault position and automatically interrupts the transportation process of the corresponding pipeline, realizes the automation of pressure pipeline abnormality judgment and fault processing, improves the management efficiency of the pressure pipeline, reduces the management cost, and at the same time, improves the accuracy and reliability of pressure pipeline abnormality diagnosis by building a pipeline network operation model and a fault analysis neural network model.

[0030] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the pipeline network simulation module 3 is used to construct a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the pressure pipeline preset data, and is specifically used to: block the pipeline picture according to the display ratio of the pipeline picture, extract high-frequency components and smooth the pipeline picture after blocking; calculate the grayscale average value of the pipeline picture after blocking, compare the grayscale value of the pixels of the pipeline picture after blocking with the grayscale average value, divide the pixels with grayscale values ​​greater than the grayscale average value into segmentation targets, divide the pixels with grayscale values ​​less than the grayscale average value into segmentation backgrounds, separate the targets according to the classification results of the segmentation targets and the segmentation backgrounds, and obtain the pipeline picture separated from the background; construct a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture separated from the background, the pipeline specification information and the pressure pipeline preset data; wherein the pressure pipeline preset data is annotated at the corresponding pipeline position as annotation information.

[0031] An implementation method of constructing a pipeline network operation model by the pipeline network simulation module 3 is as follows:

[0032] Step (1): The staff uploads the pressure pipeline distribution engineering drawing to the management platform 2 through the terminal device 7, and then the pipeline network simulation module 3 extracts the corresponding pipeline distribution engineering drawing from the management platform 2, and then collects the specification information of each pressure pipeline and collects the image information of each pipeline;

[0033] Step (2): Divide the image information into blocks according to their display ratios, and then analyze and extract the high-frequency components of the data through Fourier transform based on the divided image information, and smooth it through Gaussian filtering;

[0034] Step (3): Calculate the average grayscale value of each image information respectively, then compare the grayscale value of each group of pixels in each image information after segmentation with the calculated average value, and classify all pixels with grayscale values ​​greater than the average value as segmentation targets, and classify all pixels with grayscale values ​​less than the average value as segmentation backgrounds, and perform target separation based on the classification results;

[0035] Step (4): Construct a corresponding pipeline network operation model based on pipeline specification information, pipeline distribution engineering drawings and processed pressure pipeline actual image information, and mark the preset data of each type of pressure pipeline at the corresponding pressure pipeline position in the pipeline network operation model.

[0036] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention obtains a pipeline image separated from the background by preprocessing the pipeline image, builds a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline image separated from the background, pipeline specification information and pressure pipeline preset data, and annotates the pressure pipeline preset data as annotation information at the corresponding pipeline position, thereby improving the standardization and accuracy of the pipeline network operation model.

[0037] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the pipeline analysis module 4, when used to identify abnormal pipelines based on the pipeline network operation model and a pre-trained fault analysis neural network model, is specifically used to: input the preset data of the pressure pipeline into the pipeline network operation model, obtain a pressure pipeline operation prediction curve, and identify abnormal pipelines according to the pressure pipeline operation prediction curve.

[0038] When training the fault analysis neural network model, the output data of the model can be pre-set, such as one or several indicators for pipeline analysis. Then, the preset data of the pressure pipeline is input into the pipeline network operation model to obtain the corresponding indicator data. The preset data of the pressure pipeline within a period of time is input into the pipeline network operation model to obtain the pressure pipeline operation prediction curve describing the change of the corresponding indicator data, and the abnormal pipeline is identified according to the pressure pipeline operation prediction curve.

[0039] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention obtains a pressure pipeline operation prediction curve by inputting preset data of the pressure pipeline into the pipeline network operation model, and identifies abnormal pipelines according to the pressure pipeline operation prediction curve, thereby further improving the accuracy and reliability of pressure pipeline abnormality diagnosis.

[0040] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the pipeline analysis module 4, before being used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, is also used to: extract pressure pipeline accident data from the pipeline accident database as a sample data set, and eliminate abnormal data in the sample data set by calculating the standard deviation; standardize and normalize the remaining sample data set, and divide the processed sample data set into test set samples and training set samples; input the training set samples into a convolutional neural network, and use the least squares recursive method to define an energy function of multiple rounds of learning of the convolutional neural network. When the value of the energy function is less than the target error, the training process ends and the fault analysis neural network model is output; the test set samples are input into the fault analysis neural network model for testing, and the parameters of the fault analysis neural network model are updated.

[0041] An implementation method of the pipeline analysis module 4 for performing abnormal analysis on the pipeline is as follows:

[0042] Step (1): The pipeline analysis module 4 extracts pressure pipeline accident data from each pipeline accident database as a sample data set, then calculates the average value of the sample data set, and then calculates its standard deviation based on the sample data set and the average value of the sample data set, and then removes abnormal data in the sample data set based on the calculated standard deviation;

[0043] Step (2): After the remaining data is standardized, it is normalized and then divided into test set samples and training set samples. Then a set of convolutional neural networks is constructed and its parameter setting vector is assigned. Then, the number of neurons in each neural network layer is determined according to the preset information of the staff.

[0044] Step (3): Input the training set samples into the neural network input layer, and determine the center vector value (referring to the center vector of the hidden layer function clustering of each neural unit) to obtain the output layer as a linear combination of the hidden node outputs (referring to the optimal parameter combination when the output value tends to 0), and then use the least squares recursive method to define the energy function of the convolutional neural network for multiple rounds of learning. When the energy function is less than the target error, the training process ends and the fault analysis neural network model is output. Otherwise, the training continues;

[0045] Step (4): Import the test set samples into the fault analysis neural network model for testing, and calculate the loss value of the fault analysis neural network model, then update the parameters of the fault analysis neural network model, then input the current pipeline information (pressure pipeline preset data) into the fault analysis neural network model, and output the operation prediction curve of each pressure pipeline, then mark the pressure pipeline with abnormalities, and interrupt the operation of the pressure pipeline through the interruption module 6.

[0046] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention extracts pressure pipeline accident data from each pipeline accident database as a sample data set through a pipeline analysis module, then calculates the standard deviation of the sample data set and removes abnormal data in the sample data set, processes the remaining data, and then divides it into a test set and a training set, then constructs a group of convolutional neural networks, determines the number of neurons in each neural network layer, inputs the training set into the neural network input layer, defines the energy function of the convolutional neural network for multiple rounds of learning, and when the energy function is less than the target error, the training process ends and outputs the fault analysis neural network model, otherwise, the training continues, then the test set is imported into the fault analysis neural network model for testing, and the loss value of the fault analysis neural network model is calculated, and then the parameters of the fault analysis neural network model are updated, thereby improving the accuracy and reliability of the fault analysis neural network model, inputting the current pipeline information into the fault analysis neural network model, and outputting the operation prediction curve of each pressure pipeline, then marking the pressure pipeline with abnormalities, and interrupting the operation of the pressure pipeline, which can automatically interrupt the operation of the abnormal pipeline without manual interruption by the staff, thereby ensuring the timeliness of fault handling, reducing the difficulty of operation, and effectively improving the efficiency of pipeline management.

[0047] Figure 2 This is the second structural diagram of the petrochemical pressure pipeline management system based on artificial intelligence provided by the embodiment of the present invention. Figure 2As shown, the system also includes a performance optimization module 8, which is connected to the management platform 2. The performance optimization module 8 is used to: generate a corresponding startup list for each group of function interfaces of the management platform 2, determine the order of each function interface in the startup list according to the least recent use, and further link each group of startup lists in order of the number of visits from small to large; update the order of each function interface in each group of startup lists and the link order of each group of startup lists in real time according to the real-time interaction information of each group of function interfaces; select the function page with the least number of visits from the head of each startup list in turn as the page to be recycled according to the number of visits from small to large according to the preset time interval; merge the selected pages to be recycled into a block and mark them, wake up the compression driver to parse the marked block, and obtain the physical page belonging to the block, copy the physical page to the buffer, call the compression algorithm to compress the physical page in the buffer into the compressed block, and store the compressed block in the compression area of ​​the performance optimization module.

[0048] The performance optimization module 8 can optimize and adjust the performance of data interaction. The steps of one implementation method include:

[0049] Step (1): the performance optimization module 8 generates a startup linked list for each group of functional interfaces of the management platform 2, and determines the order of each functional interface in the startup linked list in the LRU (least recently used) linked list order, and further links each group of startup linked lists according to the number of accesses from small to large;

[0050] Step (2): updating the order of each function interface in each group of startup linked lists and the link order of each group of startup linked lists in real time according to the interaction information of each group of function interfaces; and selecting the function interface with the least number of accesses from the head of the LRU linked list as the page to be recycled, and stopping after recycling enough pages to be recycled;

[0051] Step (3): The selected pages to be recycled are merged into a block and marked, and then a compression driver is awakened to parse the marked block and obtain the physical page belonging to the block, and then the physical page is copied to the buffer, and then the compression algorithm is called to compress the physical page in the buffer into a compressed block, and the compressed block is stored in the compression area of ​​the performance optimization module.

[0052] Step (4): Repeat steps (2) and (3) at preset time intervals.

[0053] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention generates a corresponding startup chain list for each group of functional interfaces of the management platform through a performance optimization module, determines the order of each functional interface in the startup chain list according to the least recent use, further links each group of startup chain lists according to the number of accesses from small to large, updates the order of each functional interface in each group of startup chain lists and the link order of each group of startup chain lists in real time according to the real-time interaction information of each group of functional interfaces, and selects the function interface with the least number of accesses from the head of each startup chain list in turn according to the number of accesses from small to large according to the preset time interval. The system selects pages to be recycled until enough pages to be recycled are recycled, merges the selected pages to be recycled into a block and marks them, then wakes up a compression driver to parse the marked block, obtains the physical page belonging to the block, copies the physical page to the buffer, and then calls the compression algorithm to compress the physical page in the buffer into a compressed block, and stores the compressed block in the compression area of ​​the performance optimization module. It can perform large-grained memory compression on the management platform, effectively improve the running performance of the management platform, and at the same time ensure the stability of the management platform when synchronizing data, and reduce the probability of jamming during data interaction.

[0054] like Figure 2 As shown, according to an embodiment of the present invention, a petrochemical pressure pipeline management system based on artificial intelligence is provided, and the system also includes a log analysis module 9, which is used to: obtain the log data recorded in the management platform 2 through a log collection plug-in or a syslog server deployed on the management platform 2, filter out the log data to be analyzed based on the log data, and process the log data to be analyzed into a preset unified format; obtain the user operation behavior recorded in the processed log data to be analyzed, extract user behavior features according to the user operation behavior, match the user behavior features with abnormal behavior features, and obtain a matching result; in response to the matching result being greater than a preset matching threshold, generate corresponding alarm information according to the matching result; send the alarm information to the management platform 2, and send it to the terminal device 7 through the management platform 2.

[0055] The log analysis module 9 is used to perform risk analysis on user operation behaviors. The steps of one implementation are as follows:

[0056] Step (1): deploy relevant log collection plug-ins on the management platforms 2 of different systems or obtain the log data recorded in the management platforms 2 of different systems through the syslog server, and filter out the log data that meets the staff's preset conditions (determine that the relevant operations are accurate), and then process the remaining log data into log data in a unified format;

[0057] Step (2): extract user behavior features based on the user operation behaviors recorded in the processed log data, match the user behavior features with the abnormal behavior features, and obtain matching results. If the matching results are greater than a preset matching threshold, generate corresponding alarm information based on the matching results. The alarm information is then sent to the management platform 2, and then sent to the terminal device 7 through the management platform 2, so as to feed back the alarm information to relevant staff.

[0058] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention performs risk analysis on user operation behaviors through a log analysis module, generates alarm information, sends the alarm information to a management platform, and sends it to a terminal device through the management platform, thereby realizing early warning of user operation behaviors.

[0059] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module 9 is also used to generate an alarm risk score for the alarm information and add the alarm risk score to the alarm information; wherein the alarm risk score is expressed as:

[0060]

[0061] Where S represents the alarm risk score, R1 represents the importance of abnormal control instructions, R2 represents the number of historical hits of abnormal control instructions, R3 represents the number of hits of abnormal control instructions, and A S Indicates the importance of assets, V i It indicates the severity of the vulnerability caused by abnormal control instructions, and n indicates the total number of abnormal control instructions.

[0062] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention improves the accuracy of the alarm risk score by determining the alarm risk score according to the importance of abnormal control instructions, the historical number of hits of abnormal control instructions, the number of hits of abnormal control instructions, the importance of assets, the severity of vulnerabilities caused by abnormal control instructions, and the total number of abnormal control instructions.

[0063] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module 9 is also used to send the alarm information to the interruption module 6 so that the interruption module 6 terminates the response process of the relevant abnormal control instructions.

[0064] The log analysis module 9 is further used to send the alarm information to the interruption module 6 so that the interruption module 6 can terminate the response process of the relevant abnormal control instruction, that is, interrupt the response to the corresponding user operation.

[0065] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention performs risk analysis on user operation behaviors through a log analysis module, interrupts the response process of abnormal operation behaviors, and is conducive to timely elimination of potential faults.

[0066] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the log analysis module 9 is also used to record the IP address and user information of the device associated with the alarm information and send them to the management platform 2.

[0067] After obtaining the alarm information, the log analysis module 9 can also record the IP address of the device associated with the alarm information and the user information and send it to the management platform 2 for troubleshooting.

[0068] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention records the IP address of the device associated with the alarm information and the user information and sends them to the management platform, which is conducive to the rapid troubleshooting.

[0069] According to an artificial intelligence-based petrochemical pressure pipeline management system provided by an embodiment of the present invention, the terminal device 7 includes a smart phone, a tablet computer, a laptop computer or a desktop computer.

[0070] The terminal device 7 may be of various types, such as a smart phone, a tablet computer, a laptop computer, or a desktop computer.

[0071] The artificial intelligence-based petrochemical pressure pipeline management system provided by the embodiment of the present invention improves the flexibility of using terminal devices by setting terminal devices including smart phones, tablet computers, laptops or desktop computers.

[0072] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0073] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A petrochemical pressure pipeline management system based on artificial intelligence, characterized in that: It includes intelligent sensors, management platform, pipe network simulation module, pipeline analysis module, risk warning module, interruption module and terminal equipment; among which: The intelligent sensor is used to collect preset data of the pressure pipeline and send it to the management platform and the pipe network simulation module; The management platform is used to verify the identity information of the supervisor, issue control instructions based on the supervisor's operation information, and send the pressure pipeline preset data to the terminal device; The pipe network simulation module is used to obtain the pipeline distribution engineering drawing, the pipeline picture and the pipeline specification information from the management platform, and receive the pressure pipeline preset data sent by the intelligent sensor, and build a pipe network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the pressure pipeline preset data; The pipeline analysis module is used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, locate and mark the pipeline fault position, and send the pipeline fault position to the interruption module and the risk warning module; The risk warning module is used to send risk warning information to the management platform and the terminal device according to the pipeline fault location; The interruption module is used to interrupt the transportation process of the corresponding pipeline according to the pipeline fault position.

2. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1 is characterized in that: When the pipeline network simulation module is used to construct a pipeline network operation model according to the pipeline distribution engineering drawing, the pipeline picture, the pipeline specification information and the pressure pipeline preset data, it is specifically used to: The pipeline image is divided into blocks according to the display ratio of the pipeline image, and high-frequency components are extracted and smoothed from the divided pipeline image; Calculate the grayscale average value of the pipeline image after segmentation, compare the grayscale values ​​of the pixels of the pipeline image after segmentation with the grayscale average value, classify the pixels whose grayscale values ​​are greater than the grayscale average value as segmentation targets, and classify the pixels whose grayscale values ​​are less than the grayscale average value as segmentation backgrounds, perform target separation according to the classification results of the segmentation targets and the segmentation backgrounds, and obtain the pipeline image separated from the background; A pipeline network operation model is constructed according to the pipeline distribution engineering drawing, the pipeline image separated from the background, the pipeline specification information and the pressure pipeline preset data; wherein the pressure pipeline preset data is marked as annotation information at the corresponding pipeline position.

3. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1 is characterized in that: When the pipeline analysis module is used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, it is specifically used to: The preset data of the pressure pipeline is input into the pipeline network operation model to obtain a pressure pipeline operation prediction curve, and abnormal pipelines are identified according to the pressure pipeline operation prediction curve.

4. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1 is characterized in that: Before being used to identify abnormal pipelines based on the pipeline network operation model and the pre-trained fault analysis neural network model, the pipeline analysis module is also used to: Extracting pressure pipeline accident data from a pipeline accident database as a sample data set, and eliminating abnormal data in the sample data set by calculating a standard deviation; Performing standardization and normalization processing on the remaining sample data set, and dividing the processed sample data set into test set samples and training set samples; Input the training set samples into the convolutional neural network, define the energy function of the convolutional neural network for multiple rounds of learning by the least squares recursive method, and when the value of the energy function is less than the target error, the training process ends and the fault analysis neural network model is output; The test set samples are input into the fault analysis neural network model for testing, and the parameters of the fault analysis neural network model are updated.

5. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1 is characterized in that: The system further comprises a performance optimization module, wherein the performance optimization module is used to: Generate a corresponding startup chain list for each group of functional interfaces of the management platform, determine the order of each functional interface in the startup chain list according to the least recent use, and further link each group of startup chain lists according to the number of visits from the least to the most; According to the real-time interaction information of each group of functional interfaces, the order of each functional interface in each group of startup linked lists and the link order of each group of startup linked lists are updated in real time; At a preset time interval, the function page with the least number of accesses is selected from the head of each startup linked list in the order of the number of accesses from the least to the most as the page to be recycled; The selected pages to be reclaimed are merged into a block and marked, the compression driver is awakened to parse the marked block, and the physical pages belonging to the block are obtained, the physical pages are copied to a buffer, the compression algorithm is called to compress the physical pages in the buffer into a compressed block, and the compressed block is stored in the compression area of ​​the performance optimization module.

6. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1 is characterized in that: The system further comprises a log analysis module, wherein the log analysis module is used to: Obtaining log data recorded in the management platform through a log collection plug-in or a syslog server deployed on the management platform, filtering out log data to be analyzed based on the log data, and processing the log data to be analyzed into a preset unified format; Obtaining the user operation behavior recorded in the processed log data to be analyzed, extracting user behavior features according to the user operation behavior, matching the user behavior features with abnormal behavior features, and obtaining a matching result; In response to the matching result being greater than a preset matching threshold, generating corresponding alarm information according to the matching result; The alarm information is sent to the management platform, and then sent to the terminal device through the management platform.

7. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 6 is characterized in that: The log analysis module is further used to generate an alarm risk score for the alarm information and add the alarm risk score to the alarm information; wherein the alarm risk score is expressed as: Where S represents the alarm risk score, R1 represents the importance of abnormal control instructions, R2 represents the number of historical hits of abnormal control instructions, R3 represents the number of hits of abnormal control instructions, and A S Indicates the importance of assets, V i It indicates the severity of the vulnerability caused by abnormal control instructions, and n indicates the total number of abnormal control instructions.

8. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 6 is characterized in that: The log analysis module is further used to send the alarm information to the interruption module so that the interruption module can terminate the response process of the relevant abnormal control instruction.

9. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 6 is characterized in that: The log analysis module is also used to record the IP address and user information of the device associated with the alarm information and send them to the management platform.

10. The artificial intelligence-based petrochemical pressure pipeline management system according to claim 1, characterized in that: The terminal device includes a smart phone, a tablet computer, a laptop computer or a desktop computer.

Images