CN114359781B - Intelligent recognition system for cloud-side collaborative autonomous learning - Google Patents
Intelligent recognition system for cloud-side collaborative autonomous learning Download PDFInfo
- Publication number
- CN114359781B CN114359781B CN202111465719.9A CN202111465719A CN114359781B CN 114359781 B CN114359781 B CN 114359781B CN 202111465719 A CN202111465719 A CN 202111465719A CN 114359781 B CN114359781 B CN 114359781B
- Authority
- CN
- China
- Prior art keywords
- identification
- information
- algorithm
- model
- alarm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses an intelligent recognition system for cloud-edge collaborative autonomous learning, which comprises: acquiring a risk identification element set of an oil and gas pipeline; constructing a pipeline risk identification database; performing algorithm identification matching writing to generate an algorithm identification matching model; inputting first monitoring video information of an oil-gas pipeline line into an algorithm identification matching model to obtain first alarm information; carrying out false alarm label identification on the first alarm information to obtain a first label classification result; inputting the classification result of the first label into an intelligent recognition algorithm training platform to perform false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model; and optimizing the algorithm identification matching model in the edge computing device. The method solves the technical problems that in the prior art, the computing power of a chip adopted by edge computing is limited, the realized intelligent identification items are limited, and meanwhile, the identification accuracy of an algorithm is low, so that the false alarm rate is increased and the management efficiency is reduced.
Description
Technical Field
The invention relates to the field of intelligent recognition, in particular to an intelligent recognition system for cloud-side collaborative autonomous learning.
Background
With the rapid development of national economy, the national demand for energy is increasing, and pipelines are the most economic and important way for transporting oil and gas. However, with the rapid development of urbanization, the requirements for pipeline safety protection are higher and higher, and especially, the safety supervision pressure of the fruit area is higher and higher after the pipeline is high. At present, the main safety protection measures for the high back fruit areas of pipeline lines adopt a mode of manual regular patrol, and are assisted by video monitoring and unmanned aerial vehicle patrol. Because the personnel regularly patrol and protect the pipeline, the pipeline is easy to be mastered by lawless persons, and is not beneficial to the safety supervision of the pipeline. Although video monitoring equipment is installed on pipeline in succession in recent years, visualization of line security risk supervision is basically achieved, problems are brought about, such as a series of problems of who sees, what sees and the like of video pictures. At present, some cameras are provided with intelligent recognition algorithms and functions, and due to the fact that the pertinence of the algorithms of the cameras is not strong, a large amount of false alarms and invalid alarms are generated when the cameras are applied to the pipeline industry. With the continuous development of artificial intelligence technology and hardware products, edge calculation is gradually developed, so that intelligent identification of security risks by writing a complex identification algorithm in the front end becomes possible.
However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the inventor of the present application finds that the above technology has at least the following technical problems:
the method has the problems that the computing power of a chip adopted by edge computing is limited, the realized intelligent identification items are limited, and meanwhile, the identification accuracy of the algorithm is low, so that the false alarm rate is increased, and the management efficiency is reduced.
Disclosure of Invention
The embodiment of the application provides an intelligent recognition system for cloud-edge collaborative autonomous learning, and solves the technical problems that in the prior art, the computing power of a chip adopted by edge computing is limited, the achievable intelligent recognition items are limited, meanwhile, the recognition accuracy of an algorithm is low, the false alarm rate is increased, and the management efficiency is reduced.
In view of the above problems, the embodiment of the present application provides an intelligent recognition system for cloud-edge collaborative autonomous learning.
In a first aspect, an embodiment of the present application provides an intelligent recognition system for cloud-side collaborative autonomous learning, where the system is in communication connection with a cloud server, and the system includes: a first obtaining unit for obtaining a risk identification element set of an oil and gas pipeline line; the first construction unit is used for constructing a pipeline risk identification database according to the risk identification element set; the first generation unit is used for inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model; the second obtaining unit is used for inputting the first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and obtaining first alarm information according to the algorithm identification matching model; the first execution unit is used for transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform; a third obtaining unit, configured to perform false alarm tag identification on the first alarm information according to the intelligent identification management platform, so as to obtain a first tag classification result; a fourth obtaining unit, configured to input the first label classification result into the intelligent recognition algorithm training platform to perform a false alarm optimization model training, and obtain first optimization information according to the false alarm optimization model; a second execution unit to optimize the algorithmically identified matching model in the edge computing device according to the first optimization information.
On the other hand, the application also provides an intelligent identification method for cloud-edge collaborative autonomous learning, which comprises the following steps: acquiring a risk identification element set of an oil and gas pipeline; constructing a pipeline risk identification database according to the risk identification element set; inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model; inputting first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and identifying the matching model according to the algorithm to obtain first alarm information; transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform; carrying out false alarm label identification on the first alarm information according to the intelligent identification management platform to obtain a first label classification result; inputting the first label classification result into the intelligent recognition algorithm training platform for carrying out false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model; optimizing the algorithm identification matching model in the edge computing device according to the first optimization information.
In a third aspect, the present invention provides an intelligent recognition system for cloud-edge collaborative autonomous learning, including a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein the processor implements the steps of the method according to the first aspect when executing the program.
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
the risk identification element set of the oil and gas pipeline is obtained; constructing a pipeline risk identification database according to the risk identification element set; inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model; inputting first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and identifying the matching model according to the algorithm to obtain first alarm information; transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform; carrying out false alarm label identification on the first alarm information according to the intelligent identification management platform to obtain a first label classification result; inputting the first label classification result into the intelligent recognition algorithm training platform for carrying out false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model; optimizing the algorithm identification matching model in the edge computing device according to the first optimization information. The intelligent risk identification is realized, false alarm and invalid alarm are reduced, the intelligent management level and the management capacity are improved, the algorithm of the edge computing equipment can be updated according to the intelligent identification items required by the field application environment, and the functions of the edge computing equipment are enriched, so that the technical effects of cloud-edge cooperation and autonomous training and learning are realized.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
Fig. 1 is a schematic flow diagram of an intelligent identification system for cloud-edge collaborative autonomous learning according to an embodiment of the present application;
fig. 2 is a schematic diagram of a false alarm marking process of the intelligent recognition system for cloud-side collaborative autonomous learning according to the embodiment of the present application;
fig. 3 is a schematic flow chart illustrating a first accurate coefficient generation process of an intelligent recognition system for cloud-edge collaborative autonomous learning according to an embodiment of the present application;
fig. 4 is a schematic flow chart of a process of acquiring first optimization information of an intelligent recognition system for cloud-edge collaborative autonomous learning according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an intelligent recognition system for cloud-edge collaborative autonomous learning according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of an exemplary electronic device according to an embodiment of the present application.
Description of reference numerals: a first obtaining unit 11, a first constructing unit 12, a first generating unit 13, a second obtaining unit 14, a first executing unit 15, a third obtaining unit 16, a fourth obtaining unit 17, a second executing unit 18, a computing device 90, a memory 91, a processor 92, and an input-output interface 93.
Detailed Description
The embodiment of the application provides an intelligent recognition system for cloud-side collaborative autonomous learning, and solves the technical problems that in the prior art, the computing power of a chip adopted by edge computing is limited, the realized intelligent recognition items are limited, and meanwhile, the recognition accuracy of an algorithm is low, so that the false alarm rate is increased, and the management efficiency is reduced. The intelligent risk identification is realized, false alarm and invalid alarm are reduced, the intelligent management level and the management capacity are improved, the algorithm of the edge computing equipment can be updated according to the intelligent identification items required by the field application environment, and the functions of the edge computing equipment are enriched, so that the technical effects of cloud-edge cooperation and autonomous training and learning are realized.
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are merely some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.
Summary of the application
With the rapid development of national economy, the national demand for energy is increasing, and pipelines are the most economic and important way for transporting oil and gas. However, with the rapid development of urbanization, the requirements for pipeline safety protection are higher and higher, and especially the safety supervision pressure of the fruit area is higher and higher after the pipeline is high. At present, the main safety protection measures for the high back fruit area of the pipeline line adopt a mode of manual regular patrol, and are assisted by video monitoring and unmanned aerial vehicle patrol. The regular patrol of personnel has a certain time rule, so that the patrol is easy to be mastered by lawless persons, and is not beneficial to the safety supervision of pipelines. Although video monitoring equipment is installed on pipeline in succession in recent years, visualization of line security risk supervision is basically achieved, problems are brought about, such as a series of problems of who sees, what sees and the like of video pictures. At present, some cameras are provided with intelligent recognition algorithms and functions, and due to the fact that the pertinence of the algorithms of the cameras is not strong, a large amount of false alarms and invalid alarms are generated when the cameras are applied to the pipeline industry. With the continuous development of artificial intelligence technology and hardware products, edge calculation is gradually developed, so that intelligent identification of security risks by writing a complex identification algorithm in the front end becomes possible. In the prior art, the chip computing power adopted by the edge computing is limited, the realized intelligent identification items are limited, and meanwhile, the identification accuracy of the algorithm is low, so that the false alarm rate is increased, and the management efficiency is reduced.
In view of the above technical problems, the technical solution provided by the present application has the following general idea:
the embodiment of the application provides an intelligent recognition system of autonomous learning in coordination on cloud limit, wherein, the system and a cloud server communication connection, the system includes: a first obtaining unit for obtaining a risk identification element set of an oil and gas pipeline line; the first construction unit is used for constructing a pipeline risk identification database according to the risk identification element set; the first generation unit is used for inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model; the second obtaining unit is used for inputting the first monitoring video information of the oil and gas pipeline into the algorithm recognition matching model, and obtaining first alarm information according to the algorithm recognition matching model; the first execution unit is used for transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform; a third obtaining unit, configured to perform false alarm tag identification on the first alarm information according to the intelligent identification management platform, and obtain a first tag classification result; a fourth obtaining unit, configured to input the first label classification result into the intelligent recognition algorithm training platform to perform false alarm optimization model training, and obtain first optimization information according to the false alarm optimization model; a second execution unit to optimize the algorithmically identified matching model in the edge computing device according to the first optimization information.
Having thus described the general principles of the present application, embodiments thereof will now be described with reference to the accompanying drawings. As can be known to those skilled in the art, with the development of technology and the emergence of new scenes, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.
Example one
As shown in fig. 1, an embodiment of the present application provides an intelligent identification method for cloud-side collaborative autonomous learning, where the method is applied to an intelligent identification system for cloud-side collaborative autonomous learning, where the method is communicatively connected to a cloud server, and the method includes:
step S100: acquiring a risk identification element set of an oil and gas pipeline;
step S200: constructing a pipeline risk identification database according to the risk identification element set;
particularly, the mode of manual regular patrol and protection assistance, video monitoring and unmanned aerial vehicle patrol and protection combination is adopted for main safety protection measures of a high-back fruit area of a pipeline line at present. Some cameras have some intelligent recognition algorithms and functions, but due to the fact that the pertinence of the algorithms of the cameras is not strong, a large number of false alarms and invalid alarms are generated when the cameras are applied to the pipeline industry. In order to effectively solve the pain point of a manager in the aspect of safety risk video intelligent identification, a risk identification element set of an oil and gas pipeline is collected, wherein the risk identification element set comprises risk elements such as identification personnel staying for a long time, vehicles staying for a long time, an excavator and fireworks. And constructing a pipeline risk identification database according to the risk identification element set, wherein the pipeline risk identification database covers main risk elements of the fruit area after the pipeline is high, and can lay a foundation for intelligent risk identification, reduction of false alarm and ineffective alarm.
Step S300: inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model;
specifically, with the continuous development of artificial intelligence technology and hardware products, edge calculation is gradually developed, so that intelligent identification of security risks by writing a complex identification algorithm in the front end is possible. An edge computing device, which refers to a device installed in the field and having intelligent recognition functionality, is capable of performing resource-intensive processes or analysis at or near the data source. The apparatus has two forms: one is to integrate the chip and the mainboard with written algorithm into the camera; one is to separately install a device with intelligent identification function on the site. The edge computing equipment has an intelligent identification function, and can write in an intelligent identification algorithm suitable for the pipeline industry, such as long-time stay of personnel, long-time stay of vehicles, excavator, firework identification and the like. The pipeline risk identification database is input into edge computing equipment, algorithm intelligent identification matching writing is carried out, namely, the edge computing equipment is used for carrying out matching and writing of identification algorithms, and an algorithm identification matching model is generated, so that an accurate matching identification algorithm is realized.
Step S400: inputting first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and identifying the matching model according to the algorithm to obtain first alarm information;
step S500: transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform;
further, step S500 in the embodiment of the present application further includes:
step S510: the transmission process of transmitting the first alarm information to the cloud server may be implemented by a network communication device.
Specifically, the first monitoring video information comprises a real-time monitoring video, the first monitoring video information of the oil and gas pipeline is input into the algorithm identification matching model, and intelligent identification analysis is carried out on objects in a video picture to obtain first alarm information, wherein the first alarm information comprises a picture form and a short video form. And further transmitting the alarm information to the rear end through network communication equipment, wherein the network communication equipment has a network transmission function, and can transmit the alarm information identified by the edge computing equipment back to the cloud server through a mobile network signal and transmit an instruction sent by the cloud server to the edge computing equipment. The network communication equipment supports mobile/telecommunication/Unicom network transmission and simultaneously supports communication protocols such as VPDN, APN and the like customized by enterprises.
The cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform. The intelligent recognition training platform can be in seamless butt joint with the intelligent recognition management platform, and the intelligent recognition management platform has the capacity of receiving and processing alarm information sent back by the edge computing equipment. The intelligent recognition algorithm training platform is a training platform with functions of data marking, model training, model checking and model publishing. The alarm information is transmitted to the cloud server through the network communication equipment, and analysis processing is carried out, so that the intelligent management level, the management capacity and the management efficiency can be improved.
Step S600: carrying out false alarm label identification on the first alarm information according to the intelligent identification management platform to obtain a first label classification result;
step S700: inputting the first label classification result into the intelligent recognition algorithm training platform to perform false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model;
step S800: optimizing the algorithm identification matching model in the edge computing device according to the first optimization information.
Specifically, the intelligent recognition management platform confirms the accuracy of the first alarm information recognition, if the first alarm information is a false alarm, a false alarm tag is selected during processing, the alarm information is classified into a false alarm database and is classified according to different recognition items, and a first tag classification result is obtained and comprises various false alarm data. The intelligent recognition algorithm training platform automatically leads alarm information sent back by edge computing equipment into the training platform through seamless butt joint with the intelligent recognition management platform, automatically classifies according to preset model types and names to form sample data required by training, trains a misinformation optimization model, transmits the trained model to front-end edge computing equipment through a network to obtain first optimization information, and optimizes the algorithm recognition matching model in the edge computing equipment according to the first optimization information. The algorithm of the edge computing equipment can be updated according to the intelligent identification items required by the field application environment, and the functions of the edge computing equipment are enriched, so that the technical effects of cloud edge cooperation and autonomous training learning are achieved.
Further, as shown in fig. 2, the embodiment of the present application further includes:
step S910: carrying out accuracy identification on the first alarm information to obtain a first accuracy coefficient;
step S920: inputting the first accurate coefficient into a logistic regression model for accuracy judgment to obtain a logistic regression result, wherein the logistic regression result comprises a first result and a second result, the first result is inaccurate, and the second result is accurate;
step S930: if the logistic regression result is the first result, generating first false alarm identification information;
step S940: and carrying out false alarm marking on the first alarm information according to the first false alarm identification information.
Specifically, due to the problem of pertinence of the identification algorithm, accurate identification needs to be performed on the acquired first alarm information. And obtaining the first accuracy coefficient, wherein the larger the accuracy coefficient is, the more accurate the first alarm information is. Logistic regression is a pattern recognition algorithm based on probability, is a classification method, and is a binary classification problem. And inputting the first accurate coefficient into a logistic regression model to carry out accuracy judgment to obtain a logistic regression result, wherein the logistic regression result comprises a first result and a second result, the first result is inaccurate, and the second result is accurate. And if the logistic regression result is the first result, namely, the identification is wrong, first false alarm identification information is generated, and false alarm marking is carried out on the first alarm information. Through the judgment of the logistic regression model, the judgment capability of the classification accuracy of the first alarm information can be improved.
Further, wherein, the false alarm tag identification is performed on the first alarm information according to the intelligent identification management platform to obtain a first tag classification result, step S600 in the embodiment of the present application further includes:
step S610: obtaining first screening alarm information by identifying and screening the first alarm information, wherein the first screening alarm information is alarm information with a false alarm tag;
step S620: generating a first characteristic set by carrying out characteristic identification on the first screening alarm information;
step S630: and carrying out feature recognition item analysis on the first feature set, and carrying out intelligent feature direction classification on the first screening alarm information according to the feature recognition item to obtain a first label classification result.
Specifically, first alarm information comprises wrong alarm information and correct alarm information, and through discernment screening, the alarm information of screening to have the wrong report label obtains first screening alarm information. And performing feature recognition on the first screening alarm information through a feature recognition technology to generate a false alarm information feature data set, namely the first feature set. And further carrying out feature recognition item analysis on the first feature set, analyzing false alarm caused by which feature items, carrying out intelligent feature direction classification on the screening alarm information according to the feature recognition items to obtain a first label classification result, and carrying out error feature recognition and classification to facilitate summarizing of a rule of false alarm information, thereby optimizing intelligent recognition capability and reducing false alarm rate.
Further, as shown in fig. 3, the embodiment of the present application further includes:
step S911: acquiring a first manager of the intelligent management platform;
step S912: acquiring first person identification information according to the first manager;
step S913: inputting the first alarm information into an accuracy check model for self-checking to obtain first self-checking identification information;
step S914: and generating the first accuracy coefficient according to the first person identification information and the first self-checking identification information.
Further, the embodiment of the present application further includes:
step S915: taking the first person identification information as an abscissa axis;
step S916: constructing a two-dimensional rectangular coordinate system by taking the first self-checking identification information as a vertical coordinate axis;
step S917: and generating the logistic regression model according to the rectangular coordinate system, wherein the logistic regression model comprises a first logistic regression line, one side of the first logistic regression line is the inaccurate result of the first alarm information, and the other side of the first logistic regression line is the accurate result of the first alarm information.
Specifically, a first manager of the intelligent management platform is obtained through registration information of the system, the first manager is a person on duty on the day, first person identification information is obtained according to the first manager, and the first person identification information is information for identifying and confirming accuracy of alarm information identification through a login platform by the manager. Inputting the first alarm information into an accuracy check model for self-checking to obtain first self-checking identification information, wherein the first self-checking identification information is obtained by performing model checking on accuracy of alarm information identification, and the first accuracy coefficient is generated according to the first person identification information and the first self-checking identification information. And constructing a two-dimensional rectangular coordinate system by taking the first person identification information as an abscissa axis and the first self-checking identification information as an ordinate axis. And generating the logistic regression model according to the rectangular coordinate system, wherein the logistic regression model comprises a first logistic regression line, the rectangular coordinate system is divided into two areas by the first logistic regression line, one side of the first logistic regression line is a result that the first alarm information is inaccurate, and the other side of the first logistic regression line is a result that the first alarm information is accurate. By generating the first accuracy coefficient and constructing the logistic regression model, the accuracy of alarm information processing can be improved.
Further, as shown in fig. 4, the step S700 in the embodiment of the present application further includes, as a result of inputting the first label classification result into the intelligent recognition algorithm training platform to perform a false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model:
step S710: using the first label classification result as a negative sample training data set;
step S720: all alarm information which is judged as the second result in the logistic regression result is used as a positive sample training data set;
step S730: inputting the negative sample training data set and the positive sample training data set as input information into the misinformation optimization model for model training;
step S740: performing model verification on the misinformation optimization model to obtain a first verification result;
step S750: and if the first verification result is that the verification is passed, obtaining the first optimization information according to the false alarm optimization model.
Specifically, all data in the negative sample training data set are false alarm data, and all data in the positive sample training data set are correct alarm data, which is accurate data to be identified. Inputting the negative sample training data set and the positive sample training data set into the misinformation optimization model as input information to perform model training, performing model verification on the misinformation optimization model after training is completed to obtain a first verification result, and if the first verification result is verification passing, indicating that the model verification meets the verification requirement, further obtaining the first optimization information through the misinformation optimization model, and obtaining scientific and reliable optimization information through training of positive and negative samples and verification of the model.
Through the above description of the embodiments, those skilled in the art will clearly understand that the present application can be implemented by software plus necessary general-purpose hardware, and certainly can also be implemented by special-purpose hardware including special-purpose integrated circuits, special-purpose CPUs, special-purpose memories, special-purpose components and the like. Generally, functions performed by computer programs can be easily implemented by corresponding hardware, and specific hardware structures for implementing the same functions may be various, such as analog circuits, digital circuits, or dedicated circuits. However, for the present application, the implementation of a software program is more preferable. Based on such understanding, the technical solutions of the present application or portions contributing to the prior art may be embodied in the form of a software product, where the computer software product is stored in a readable storage medium, such as a floppy disk, a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk of a computer, and includes several instructions for causing a computer device to execute the method according to the embodiments of the present application.
To sum up, the intelligent recognition system for cloud-edge collaborative autonomous learning provided by the embodiment of the application has the following technical effects:
1. the risk identification element set of the oil and gas pipeline is obtained; constructing a pipeline risk identification database according to the risk identification element set; inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model; inputting first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and identifying the matching model according to the algorithm to obtain first alarm information; transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform; carrying out false alarm tag identification on the first alarm information according to the intelligent identification management platform to obtain a first tag classification result; inputting the first label classification result into the intelligent recognition algorithm training platform for carrying out false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model; optimizing the algorithm identification matching model in the edge computing device according to the first optimization information. The intelligent risk identification is realized, false alarm and invalid alarm are reduced, the intelligent management level and the management capacity are improved, the algorithm of the edge computing equipment can be updated according to the intelligent identification items required by the field application environment, and the functions of the edge computing equipment are enriched, so that the technical effects of cloud-edge cooperation and autonomous training and learning are realized.
2. Due to the adoption of the method for carrying out accuracy identification and accuracy judgment on the alarm information, the judgment capability of improving the classification accuracy of the alarm information can be improved, the intelligent identification capability is optimized, and the false alarm rate is reduced.
Example two
Based on the same inventive concept as the intelligent recognition system for cloud-edge collaborative autonomous learning in the foregoing embodiment, the present invention further provides an intelligent recognition system for cloud-edge collaborative autonomous learning, as shown in fig. 5, the system includes:
a first obtaining unit 11, wherein the first obtaining unit 11 is used for obtaining a risk identification element set of an oil and gas pipeline;
a first constructing unit 12, where the first constructing unit 12 is configured to construct a pipeline risk identification database according to the risk identification element set;
the first generating unit 13 is configured to input the pipeline risk identification database into edge computing equipment to perform algorithm identification matching writing, and generate an algorithm identification matching model;
the second obtaining unit 14 is configured to input the first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and obtain first alarm information according to the algorithm identification matching model;
the first execution unit 15 is configured to transmit the first alarm information to a cloud server, where the cloud server includes an intelligent recognition management platform and an intelligent recognition algorithm training platform;
a third obtaining unit 16, where the third obtaining unit 16 is configured to perform false alarm tag identification on the first alarm information according to the intelligent identification management platform, and obtain a first tag classification result;
a fourth obtaining unit 17, where the fourth obtaining unit 17 is configured to input the first label classification result into the intelligent recognition algorithm training platform to perform a false alarm optimization model training, and obtain first optimization information according to the false alarm optimization model;
a second execution unit 18, the second execution unit 18 configured to optimize the algorithmically identified matching model in the edge computing device according to the first optimization information.
Further, the system further comprises:
a fifth obtaining unit, configured to perform accuracy identification on the first alarm information to obtain a first accuracy coefficient;
a sixth obtaining unit, configured to input the first accurate coefficient into a logistic regression model to perform accuracy judgment, and obtain a logistic regression result, where the logistic regression result includes a first result and a second result, the first result is inaccurate, and the second result is accurate;
a second generating unit, configured to generate first false positive identification information if the logistic regression result is the first result;
and the third execution unit is used for carrying out false alarm marking on the first alarm information according to the first false alarm identification information.
Further, the system further comprises:
a seventh obtaining unit, configured to obtain first screening alarm information by identifying and screening the first alarm information, where the first screening alarm information is alarm information with a false alarm tag;
a third generating unit, configured to generate a first feature set by performing feature recognition on the first screening alarm information;
and the eighth obtaining unit is used for carrying out feature identification item analysis on the first feature set, carrying out intelligent feature direction classification on the first screening alarm information according to the feature identification item, and obtaining the first label classification result.
Further, the system further comprises:
a ninth obtaining unit, configured to obtain a first administrator of the intelligent management platform;
a tenth obtaining unit, configured to obtain first person identification information according to the first administrator;
an eleventh obtaining unit, configured to input the first alarm information into an accuracy check model for self-checking, and obtain first self-checking identification information;
a fourth generating unit configured to generate the first accuracy coefficient based on the first person identification information and the first self-test identification information.
Further, the system further comprises:
a fourth execution unit configured to take the first person identification information as an abscissa axis;
a second constructing unit, configured to construct a two-dimensional rectangular coordinate system by using the first self-inspection identification information as a vertical coordinate axis;
and the fifth generation unit is used for generating the logistic regression model according to the rectangular coordinate system, wherein the logistic regression model comprises a first logistic regression line, one side of the first logistic regression line is a result that the first alarm information is inaccurate, and the other side of the first logistic regression line is a result that the first alarm information is accurate.
Further, the system further comprises:
a fifth execution unit, where a transmission process for transmitting the first alarm information to the cloud server may be implemented by a network communication device.
Further, the system further comprises:
a sixth execution unit, configured to use the first label classification result as a negative sample training data set;
a seventh execution unit, configured to use all alarm information in the logistic regression result that is determined as the second result as a positive sample training data set;
the eighth execution unit is used for inputting the negative sample training data set and the positive sample training data set into the misinformation optimization model as input information for model training;
a twelfth obtaining unit, configured to obtain a first verification result by performing model verification on the false alarm optimization model;
a thirteenth obtaining unit, configured to obtain the first optimization information according to the false alarm optimization model if the first verification result is that the verification is passed.
In the embodiment of the present application, the network device and the terminal device may be divided into functional modules according to the above method example, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one receiving module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. It should be noted that, in the embodiment of the present application, the division of the module is schematic, and is only one logic function division, and another division manner may be available in actual implementation. Through the foregoing detailed description of the intelligent recognition system for cloud-side collaborative autonomous learning, those skilled in the art can clearly know the implementation method of the intelligent recognition system for cloud-side collaborative autonomous learning in this embodiment, so for the brevity of the description, detailed description is not provided here.
Exemplary electronic device
FIG. 6 is a schematic diagram of a computing device of the present application. The computing device 90 shown in fig. 6 may include: memory 91, processor 92, input/output interface 93. The memory 91, the processor 92 and the input/output interface 93 are connected via an internal connection path, the memory 33 is used for storing instructions, and the processor 92 is used for executing the instructions stored in the memory 91 to control the input/output interface 93 to receive input data and information and output data such as operation results.
FIG. 6 is a schematic diagram of a computing device of another embodiment of the present application. The computing device 90 shown in fig. 6 may include a memory 91, a processor 92, and an input/output interface 93. The memory 91, the processor 92 and the input/output interface 93 are connected through an internal connection path, the memory 91 is used for storing instructions, and the processor 92 is used for executing the instructions stored in the memory 92 so as to control the input/output interface 93 to receive input data and information and output data such as operation results.
In implementation, the steps of the above method may be performed by instructions in the form of hardware, integrated logic circuits, or software in the processor 92. The method for recognizing the abnormal message and/or the method for training the abnormal message recognition model disclosed in the embodiments of the present application may be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor. The software module may be located in ram, flash memory, rom, prom or eprom, registers, etc. storage media that are well known in the art. The storage medium is located in the memory 91, and the processor 92 reads the information in the memory 91 and completes the steps of the method in combination with the hardware. To avoid repetition, it is not described in detail here.
It should be understood that in the embodiment of the present application, the processor may be a Central Processing Unit (CPU), and the processor may also be other general-purpose processors, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
It will be appreciated that in embodiments of the application, the memory may comprise both read-only memory and random access memory, and may provide instructions and data to the processor. A portion of the processor may also include non-volatile random access memory. For example, the processor may also store information of the device type.
It should be understood that, in the various embodiments of the present application, the sequence numbers of the above-mentioned processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.
In the several embodiments provided in the present application, it should be understood that the disclosed system, system and method may be implemented in other ways. For example, the above-described system embodiments are merely illustrative, and for example, the division of the units is only one type of logical functional division, and other divisions may be realized in practice, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed coupling or direct coupling or communication connection between each other may be through some interfaces, indirect coupling or communication connection of systems or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the application to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable system. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer readable storage medium may be any available medium that can be read by a computer or a data storage device including one or more available media integrated servers, data centers, and the like. The usable medium may be a magnetic medium (e.g., a floppy Disk, a hard Disk, a magnetic tape), an optical medium (e.g., a Digital Versatile Disk (DVD)), or a semiconductor medium (e.g., a Solid State Disk (SSD)), among others.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (9)
1. An intelligent recognition system for cloud-edge collaborative autonomous learning, wherein the system is in communication connection with a cloud server, the system comprising:
the first obtaining unit is used for obtaining a risk identification element set of an oil and gas pipeline;
the first construction unit is used for constructing a pipeline risk identification database according to the risk identification element set;
the first generation unit is used for inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model;
the second obtaining unit is used for inputting the first monitoring video information of the oil and gas pipeline into the algorithm recognition matching model, and obtaining first alarm information according to the algorithm recognition matching model;
the first execution unit is used for transmitting the first alarm information to the cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform;
a third obtaining unit, configured to perform false alarm tag identification on the first alarm information according to the intelligent identification management platform, and obtain a first tag classification result;
a fourth obtaining unit, configured to input the first label classification result into the intelligent recognition algorithm training platform to perform false alarm optimization model training, and obtain first optimization information according to the false alarm optimization model;
a second execution unit to optimize the algorithmically identified matching model in the edge computing device according to the first optimization information.
2. The system of claim 1, wherein the system comprises:
a fifth obtaining unit, configured to perform accuracy identification on the first alarm information to obtain a first accuracy coefficient;
a sixth obtaining unit, configured to input the first accurate coefficient into a logistic regression model for accuracy judgment to obtain a logistic regression result, where the logistic regression result includes a first result and a second result, the first result is inaccurate, and the second result is accurate;
a second generating unit, configured to generate first false positive identification information if the logistic regression result is the first result;
and the third execution unit is used for carrying out false alarm marking on the first alarm information according to the first false alarm identification information.
3. The system of claim 2, wherein the third obtaining unit further comprises:
a seventh obtaining unit, configured to obtain first screening alarm information by identifying and screening the first alarm information, where the first screening alarm information is alarm information with a false alarm tag;
a third generating unit, configured to generate a first feature set by performing feature recognition on the first screening alarm information;
and the eighth obtaining unit is used for carrying out feature identification item analysis on the first feature set, carrying out intelligent feature direction classification on the first screening alarm information according to the feature identification item, and obtaining the first label classification result.
4. The system of claim 2, wherein the system comprises:
a ninth obtaining unit, configured to obtain a first administrator of the intelligent management platform;
a tenth obtaining unit, configured to obtain first person identification information according to the first administrator;
an eleventh obtaining unit, configured to input the first alarm information into an accuracy check model for self-checking, and obtain first self-checking identification information;
a fourth generating unit configured to generate the first accuracy coefficient based on the first person identification information and the first self-test identification information.
5. The system of claim 4, wherein the system comprises:
a fourth execution unit configured to take the first person identification information as an abscissa axis;
the second construction unit is used for constructing a two-dimensional rectangular coordinate system by taking the first self-checking identification information as an ordinate axis;
and the fifth generation unit is used for generating the logistic regression model according to the rectangular coordinate system, wherein the logistic regression model comprises a first logistic regression line, one side of the first logistic regression line is a result that the first alarm information is inaccurate, and the other side of the first logistic regression line is a result that the first alarm information is accurate.
6. The system of claim 1, wherein the first execution unit further comprises:
a fifth execution unit, where a transmission process for transmitting the first alarm information to the cloud server may be implemented by a network communication device.
7. The system of claim 2, wherein the fourth obtaining unit further comprises:
a sixth execution unit, configured to use the first label classification result as a negative sample training data set;
a seventh execution unit, configured to use all alarm information in the logistic regression result that is determined as the second result as a positive sample training data set;
the eighth execution unit is used for inputting the negative sample training data set and the positive sample training data set into the false alarm optimization model as input information for model training;
a twelfth obtaining unit, configured to obtain a first verification result by performing model verification on the misinformation optimization model;
a thirteenth obtaining unit, configured to, if the first verification result is a verification pass, obtain the first optimization information according to the false alarm optimization model.
8. An intelligent recognition matching method for cloud edge collaborative autonomous learning, wherein the method comprises the following steps:
acquiring a risk identification element set of an oil and gas pipeline;
constructing a pipeline risk identification database according to the risk identification element set;
inputting the pipeline risk identification database into edge computing equipment for algorithm identification matching writing, and generating an algorithm identification matching model;
inputting first monitoring video information of the oil and gas pipeline into the algorithm identification matching model, and identifying the matching model according to the algorithm to obtain first alarm information;
transmitting the first alarm information to a cloud server, wherein the cloud server comprises an intelligent recognition management platform and an intelligent recognition algorithm training platform;
carrying out false alarm label identification on the first alarm information according to the intelligent identification management platform to obtain a first label classification result;
inputting the first label classification result into the intelligent recognition algorithm training platform for carrying out false alarm optimization model training, and obtaining first optimization information according to the false alarm optimization model;
optimizing the algorithm identification matching model in the edge computing device according to the first optimization information.
9. A cloud-edge collaborative autonomous learning smart recognition system comprising at least one processor and a memory, the at least one processor coupled with the memory for reading and executing instructions in the memory to perform the system of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111465719.9A CN114359781B (en) | 2021-12-02 | 2021-12-02 | Intelligent recognition system for cloud-side collaborative autonomous learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111465719.9A CN114359781B (en) | 2021-12-02 | 2021-12-02 | Intelligent recognition system for cloud-side collaborative autonomous learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114359781A CN114359781A (en) | 2022-04-15 |
| CN114359781B true CN114359781B (en) | 2022-12-27 |
Family
ID=81097089
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111465719.9A Active CN114359781B (en) | 2021-12-02 | 2021-12-02 | Intelligent recognition system for cloud-side collaborative autonomous learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114359781B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117149361B (en) * | 2023-10-30 | 2023-12-29 | 北京万界数据科技有限责任公司 | Multi-terminal collaborative optimization system for training model |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10529318B2 (en) * | 2015-07-31 | 2020-01-07 | International Business Machines Corporation | Implementing a classification model for recognition processing |
| CN109117871A (en) * | 2018-07-20 | 2019-01-01 | 上海市肺科医院 | A kind of system and method that the pernicious probability of pulmonary lesions is obtained by deep learning |
| CN111582016A (en) * | 2020-03-18 | 2020-08-25 | 宁波送变电建设有限公司永耀科技分公司 | Intelligent maintenance-free power grid monitoring method and system based on cloud edge collaborative deep learning |
| CN112257500A (en) * | 2020-09-16 | 2021-01-22 | 江苏方天电力技术有限公司 | Intelligent image recognition system and method for power equipment based on cloud edge cooperation technology |
| CN113569682A (en) * | 2021-07-19 | 2021-10-29 | 江苏宏晨矿业设备科技有限公司 | Video monitoring method and device for intelligently capturing mine identification elements |
-
2021
- 2021-12-02 CN CN202111465719.9A patent/CN114359781B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114359781A (en) | 2022-04-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108197658B (en) | Image annotation information processing method, device, server and system | |
| CN112183577A (en) | Training method of semi-supervised learning model, image processing method and equipment | |
| EP3637317A1 (en) | Method and apparatus for generating vehicle damage information | |
| CN111639648B (en) | Certificate identification method, device, computing equipment and storage medium | |
| CN113837151B (en) | Table image processing method and device, computer equipment and readable storage medium | |
| CN114359781B (en) | Intelligent recognition system for cloud-side collaborative autonomous learning | |
| CN113158869A (en) | Image recognition method and device, terminal equipment and computer readable storage medium | |
| CN113393487A (en) | Moving object detection method, moving object detection device, electronic equipment and medium | |
| CN111325181A (en) | State monitoring method and device, electronic equipment and storage medium | |
| CN109035666B (en) | Fire and smoke detection method and device and terminal equipment | |
| CN114692778A (en) | Multi-modal sample set generation method, training method and device for intelligent inspection | |
| CN112396060B (en) | Identification card recognition method based on identification card segmentation model and related equipment thereof | |
| CN117036947A (en) | Image recognition-based agricultural risk early warning method, device, equipment and medium | |
| CN111091099A (en) | Scene recognition model construction method, scene recognition method and device | |
| CN116842384A (en) | Multi-mode model training method and device, electronic equipment and readable storage medium | |
| CN114241411B (en) | Counting model processing method and device based on target detection and computer equipment | |
| CN115860024A (en) | Two-dimensional code identification method and identification device | |
| CN115762704A (en) | Prescription auditing method, device, equipment and storage medium | |
| CN114998665A (en) | Image category identification method and device, electronic equipment and storage medium | |
| CN113822313A (en) | Method and device for detecting abnormity of graph nodes | |
| CN112270257A (en) | Motion trajectory determination method and device and computer readable storage medium | |
| CN112989869A (en) | Optimization method, device and equipment of face quality detection model and storage medium | |
| WO2024088031A1 (en) | Data acquisition method and apparatus, and related device | |
| CN116883764B (en) | Battery system fault identification method and device | |
| WO2022222143A1 (en) | Security test method and apparatus for artificial intelligence system, and terminal device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |