AI big data intelligent processing method and system based on edge computing

Abstract

The invention provides an AI big data intelligent processing method and system based on edge computing, and the method comprises the steps: determining a data source, carrying out the monitoring and collection of source data according to the data source, and transmitting the monitored and collected data to an edge node; performing edge calculation on the monitoring acquisition data through an edge node to obtain monitoring acquisition processing data; uploading the monitored, collected and processed data to a cloud computing center based on a concurrent data transfer control model; aI big data is adopted in the cloud computing center to automatically analyze, monitor, collect and process data; and result feedback is carried out. According to the method, edge calculation and AI big data intelligent processing are combined, the delay and bandwidth requirements of data transmission can be reduced, the real-time performance and the response capability of AI big data intelligent processing are improved, the concurrent data transfer control model is adopted for uploading regulation and control, the parallel performance and stability of uploaded data are improved, and the data transmission efficiency is improved. And a guarantee is provided for a cloud computing center to carry out AI big data intelligent processing.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and in particular to an AI big data intelligent processing method and system based on edge computing. Background Art

[0002] With the rapid development of mobile Internet and Internet of Things, various intelligent terminals and sensor devices have generated massive amounts of data. The demand for storage and processing of massive data has also increased, which has put pressure on centralized data processing centers. The big data industry takes big data as its core resource and realizes the value of data through collection, storage, processing, analysis, application and display of the generated data.

[0003] To meet the needs of real-time data analysis, AI is usually used in cloud computing centers to realize big data intelligent processing, thereby improving the efficiency of data processing. However, when uploading monitoring and collected data, problems such as data transmission delay, bandwidth bottleneck, and poor stability often occur, which affect the data analysis of the cloud computing center, making it impossible for the cloud computing center to perform big data analysis on the monitoring and collected data in a timely manner, resulting in poor real-time and responsiveness of the AI ​​big data intelligent processing of the cloud computing center. Therefore, the present invention proposes an AI big data intelligent processing method and system based on edge computing, which realizes the combination of edge computing and AI big data intelligent processing, can reduce data transmission delay and bandwidth requirements, improve the real-time and responsiveness of AI big data intelligent processing, and adopts a concurrent data control model to regulate the uploaded data, improve the parallel performance and stability of the uploaded data, and provide guarantee for the cloud computing center to perform AI big data intelligent processing. Summary of the invention

[0004] The purpose of the present invention is to provide an AI big data intelligent processing method and system based on edge computing to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: an AI big data intelligent processing method based on edge computing, comprising:

[0006] Determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node;

[0007] Perform edge computing on the monitoring and collection data through edge nodes to obtain monitoring and collection processing data;

[0008] Upload the monitored, collected and processed data to the cloud computing center based on the concurrent data transfer control model;

[0009] In the cloud computing center, the monitoring, collection and processing data are received, and AI big data automated analysis is performed on the monitoring, collection and processing data to obtain the monitoring data analysis results;

[0010] Provide feedback based on the monitoring data analysis results.

[0011] Furthermore, when monitoring and collecting source data according to the data source, the monitoring and collecting device is deployed in advance in the data source for monitoring, and the edge node is configured according to the monitoring and collecting device to establish a relationship between the edge node and the monitoring and collecting device.

[0012] Furthermore, edge computing is performed on the monitoring and collection data through edge nodes, including:

[0013] Perform reception detection on the monitoring and collection data, analyze the compatibility of the monitoring and collection data with the edge nodes, and obtain preliminary detection results;

[0014] According to the preliminary detection result, when the monitoring and collection data are adapted to the edge node, a first analysis is performed on the received monitoring and collection data to determine whether there is missing data in the monitoring and collection data, and a first analysis result is obtained;

[0015] Supplement the monitoring and collection data according to the first analysis result to obtain the first processed data of the monitoring and collection;

[0016] Performing a second analysis on the first processed data collected by monitoring, determining whether there is information that needs to be corrected in the first processed data collected by monitoring, and obtaining a second analysis result;

[0017] Correcting the first processed data collected by monitoring according to the second analysis result to obtain second processed data collected by monitoring;

[0018] The edge computing screening model is used to perform model processing on the second processed data collected by monitoring to obtain the third processed data collected by monitoring.

[0019] Furthermore, the edge computing screening model is a data processing model created for edge nodes, including:

[0020] Analyze the target data for the edge nodes, obtain the target data features, and obtain the edge node processing data features;

[0021] Building a recognition and screening model according to the data features processed by the edge node to obtain a first building model;

[0022] Perform security analysis based on the characteristics of data processed by edge nodes to determine the security performance of data processed by edge nodes, and determine the security performance enhancement strategy based on the security performance of data processed by edge nodes to obtain a security assurance strategy;

[0023] Based on edge computing and security assurance strategies, a secondary model is constructed for the first construction model to obtain an edge computing screening model.

[0024] Furthermore, the concurrent data transfer control model includes: a node access unit, a forwarding analysis unit, a traffic diversion control unit and a data sending unit;

[0025] The node access unit is used to access the edge node and obtain monitoring, collection and processing data for the edge node;

[0026] The forwarding analysis unit is used to perform data screening and upload analysis on the monitoring, collection and processing data to obtain data screening results and upload analysis data;

[0027] The diversion control unit is used to divert and buffer the monitoring, collection and processing data according to the data screening results in combination with the uploaded analysis data, and to control the buffer sequence for the buffer;

[0028] The data sending unit is used to send and upload the data packets to the cloud computing center according to the buffer sequence of the cache area.

[0029] Furthermore, the forwarding analysis unit performs data screening and upload analysis on the monitoring, collection and processing data, including:

[0030] Determining a data packet of the monitoring, collecting and processing data based on the monitoring, collecting and processing data;

[0031] Performing data packet feature recognition on the data packet, obtaining the data packet recognition feature, and judging whether the data packet conforms to the target data flow according to the data packet recognition feature analysis, and obtaining the recognition judgment result;

[0032] The data packets are screened according to the identification and judgment results to obtain data screening results;

[0033] Different analysis methods are used to analyze the uploaded data according to the data screening results. When the data screening result is that the data packet conforms to the target data stream, the first analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet; when the data screening result is that the data packet does not conform to the target data stream, the second analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet.

[0034] Furthermore, the diversion control unit performs diversion buffering on the monitoring, collection and processing data in combination with the uploaded analysis data according to the data screening result, and performs buffer sequence control on the buffer area, including:

[0035] Divert the data packets of the monitored, collected and processed data according to the data screening results. When the data screening results show that the data packets meet the target data flow, the corresponding data packets are diverted to the first buffer. When the data screening results show that the data packets do not meet the target data flow, the corresponding data packets are diverted to the second buffer.

[0036] In the first buffer and / or the second buffer, the sending port information is acquired for the data generating end to determine the sending characteristics of the sending port and simultaneously obtain the current buffering status of the buffer sequence. The sending characteristics of the sending port are combined with the current buffering status of the buffer sequence according to the correspondence between the sending port and the buffer sequence to perform upload data flow analysis to obtain upload traffic analysis data. Then, based on the data screening result, an optimization analysis is performed according to the upload traffic analysis data and the upload analysis data of the data packet to determine the buffer target sequence of the data packet and buffer the data packet to the buffer target sequence.

[0037] Furthermore, when AI big data automated analysis is performed on the monitoring, collection and processing data, an AI big data intelligent analysis model is used to perform automated analysis on the monitoring, collection and processing data, including:

[0038] Identify monitoring, acquisition and processing data and determine the characteristics of the monitoring, acquisition and processing data;

[0039] Configure a big data analysis solution based on the characteristics of the monitored, collected and processed data to obtain an AI big data analysis solution;

[0040] According to the AI ​​big data analysis plan, retrieve the big data AI model and obtain the AI ​​big data target analysis model;

[0041] The AI ​​big data target analysis model is used to automatically analyze the monitoring collection and processing data to obtain the monitoring data analysis results.

[0042] Furthermore, when feedback is provided for the monitoring data analysis results, data storage and AI big data automated analysis records are performed for the monitoring data analysis results and the monitoring collection and processing data based on the monitoring data analysis results, and at the same time, the result feedback type is determined, and the format of the monitoring data analysis results is converted according to the result feedback type to obtain the monitoring data analysis results in the target format, and then the monitoring data analysis results in the target format are fed back according to the result feedback type.

[0043] An AI big data intelligent processing system based on edge computing, comprising: a monitoring and acquisition module, an edge computing module, a data upload module, an AI analysis module and a result feedback module;

[0044] The monitoring and collection module is used to determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node;

[0045] The edge computing module is used to perform edge computing on the monitoring and collection data through the edge node to obtain the monitoring and collection processing data;

[0046] The data uploading module is used to upload the monitoring, collection and processing data to the cloud computing center based on the concurrent data transfer control model;

[0047] The AI ​​analysis module is used to perform AI big data automated analysis on the monitoring, collection and processing data to obtain monitoring data analysis results;

[0048] The result feedback module is used to provide result feedback for the monitoring data analysis results.

[0049] The present invention realizes the combination of edge computing and AI big data intelligent processing, which can reduce the delay and bandwidth requirement of data transmission, improve the real-time and responsiveness of AI big data intelligent processing, and adopts a concurrent data transfer control model to regulate the uploaded data, thereby improving the parallel performance and stability of the uploaded data, and providing guarantee for the cloud computing center to perform AI big data intelligent processing.

[0050] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0051] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 Schematic diagram of the AI ​​big data intelligent processing method of the present invention;

[0054] Figure 2 This is a schematic diagram of the edge computing steps in the AI ​​big data intelligent processing method of the present invention;

[0055] Figure 3 It is a schematic diagram of the composition of the concurrent data transfer control model in the AI ​​big data intelligent processing method described in the present invention;

[0056] Figure 4This is a schematic diagram of the AI ​​big data intelligent processing system described in the present invention. DETAILED DESCRIPTION

[0057] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0058] like Figure 1 As shown, an embodiment of the present invention provides an AI big data intelligent processing method based on edge computing, including:

[0059] Determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node;

[0060] Perform edge computing on the monitoring and collection data through edge nodes to obtain monitoring and collection processing data;

[0061] Upload the monitored, collected and processed data to the cloud computing center based on the concurrent data transfer control model;

[0062] In the cloud computing center, the monitoring, collection and processing data are received, and AI big data automated analysis is performed on the monitoring, collection and processing data to obtain the monitoring data analysis results;

[0063] Provide feedback based on the monitoring data analysis results.

[0064] In the above technical solution, when collecting source data information for data sources, data collection devices are deployed according to the data sources, and source data information is collected for the data sources through the data collection devices, thereby obtaining monitoring and collection data.

[0065] In the above technical solution, there are multiple edge nodes.

[0066] In the above technical solution, the concurrent data transfer control model realizes the concurrent reception and forwarding processing of the monitoring, collection and processing data obtained by the edge node.

[0067] In the above technical solution, when the monitoring and collected data is transmitted to the edge node, the monitoring and collected data is transmitted in real time.

[0068] In the above technical solution, the cloud computing center is the data service center, which uses AI for big data analysis.

[0069] The above technical solution combines edge computing with AI big data intelligent processing to realize data monitoring, collection and remote analysis, so that after the monitoring and collection data is obtained, it can be uploaded to the cloud computing center in time for AI big data automated analysis. Edge computing is performed on the collection source data through edge nodes to optimize the monitoring and collection data, so that when uploading the monitoring and collection processing data, the data transmission delay and bandwidth requirements can be reduced, thereby improving the real-time and responsiveness of AI big data intelligent processing. When uploading the monitoring and collection processing data to the cloud computing center, a concurrent data transfer model is used to improve the parallel performance and stability of the uploaded data, providing guarantee for AI big data intelligent processing, so that the cloud computing center can better realize AI big data automated analysis.

[0070] In an embodiment provided by the present invention, when monitoring and collecting source data according to the data source, the monitoring and collecting device is deployed in advance in the data source for monitoring, and the edge node is configured according to the monitoring and collecting device to establish a relationship between the edge node and the monitoring and collecting device.

[0071] In the above technical solution, source data is monitored by the monitoring and collection device for the data source to obtain the monitoring and collection data.

[0072] In the above technical solution, when the monitoring and collection device obtains the monitoring and collection data, it transmits the monitoring and collection data to the corresponding edge node in real time according to the relationship between the edge node and the monitoring and collection device.

[0073] In the above technical solution, when edge nodes are configured according to the monitoring and collection device, data volume analysis is performed on the monitoring and collection device to determine the edge node demand, and a corresponding number of edge nodes are retrieved according to the edge node demand, and then a relationship between the retrieved edge nodes and the monitoring and collection device is established.

[0074] The above technical solution adopts a monitoring and collection device to realize the monitoring of source data in the data source, so that the monitoring and collection data of the source data can be automatically obtained, and then the current status can be analyzed according to the monitoring and collection data to realize the monitoring function, and by establishing a relationship between the edge node and the monitoring and collection device, the monitoring and collection device can process the monitoring and collection data immediately when it is obtained, reducing the dependence on the central server, optimizing the data transmission process, reducing the network bandwidth occupied by data transmission, and using multiple edge nodes for edge computing processing of monitoring and collection data, realizing decentralized processing of tasks, and can also configure the edge nodes according to needs, improving the scalability and fault tolerance of edge computing, and can adapt to different scenario requirements.

[0075] like Figure 2As shown, in one embodiment provided by the present invention, edge computing is performed on monitoring and collecting data through edge nodes, including:

[0076] A1. Perform reception detection on the monitoring and collection data, analyze the compatibility of the monitoring and collection data with the edge nodes, and obtain preliminary detection results;

[0077] A2. According to the preliminary detection result, when the monitoring and collection data are adapted to the edge node, a first analysis is performed on the received monitoring and collection data to determine whether there is missing data in the monitoring and collection data, and a first analysis result is obtained;

[0078] A3. Supplement the monitoring and collection data according to the first analysis result to obtain the first processed data of monitoring and collection;

[0079] A4. Perform a second analysis on the first processed data collected by monitoring, determine whether there is information that needs to be corrected in the first processed data collected by monitoring, and obtain a second analysis result;

[0080] A5. Correct the first processed data for monitoring and acquisition according to the second analysis result to obtain second processed data for monitoring and acquisition;

[0081] A6. Use the edge computing screening model to perform model processing on the second processed data collected by monitoring to obtain the third processed data collected by monitoring.

[0082] In the above technical scheme, when analyzing the compatibility of the monitoring and collection data with the edge nodes, data feature analysis is performed on the monitoring and collection data to obtain the monitoring and collection data features, and at the same time, node feature analysis is performed on the edge nodes to obtain the edge node features. The matching degree between the monitoring and collection data and the edge nodes is judged based on the monitoring and collection data features and the edge node features, and whether the monitoring and collection data and the edge nodes are compatible is determined based on the matching degree, thereby obtaining preliminary detection results.

[0083] In the above technical solution, when the first analysis result shows that there are missing data in the monitoring and collection data, the monitoring and collection data are supplemented. When the first analysis result shows that there are no missing data in the monitoring and collection data, the second analysis is directly performed on the monitoring and collection data.

[0084] In the above technical solution, situations that need to be corrected include: erroneous data information, improperly expressed information, logical errors, etc.

[0085] In the above technical solution, when the second analysis result is that there is information that needs to be corrected in the first processed data collected for monitoring, the first processed data collected for monitoring is corrected; when the second analysis result is that there is no information that needs to be corrected in the first processed data collected for monitoring, the edge computing screening model is directly used to perform data processing on the first processed data collected for monitoring.

[0086] In the above technical solution, model processing includes: data feature recognition and screening, data information encryption and other processing.

[0087] In the above technical solution, the third processed data collected by monitoring is the final processed data collected by monitoring.

[0088] The above technical scheme realizes the sorting of monitoring and collection data through edge nodes, improves the accuracy and completeness of monitoring and collection processing data, provides AI big data automatic analysis guarantee for cloud computing center, and through receiving detection, subsequent analysis and processing are performed only when the monitoring and collection data are adapted to the edge node, avoiding edge computing of monitoring and collection data that does not conform to the edge node and occupying the performance of the edge node, thereby improving the edge computing effectiveness of the edge node, and timely discovers the existence of data missing in the monitoring and collection data through the first analysis, thereby ensuring the integrity of the monitoring and collection processing data, and realizes the correction of the monitoring and collection data through the second analysis, avoiding large errors in AI big data automatic analysis caused by errors or inaccuracies in the monitoring and collection data, thereby ensuring the accuracy of the monitoring and collection processing data, and realizing edge computing by adopting the edge computing screening model to perform model processing on the second processed data of monitoring and collection, so that the model processing of the second processed data of monitoring and collection can be efficiently realized, the quality of the third processed data of monitoring and collection can be improved, and the security of the third processed data of monitoring and collection can be ensured.

[0089] In one embodiment provided by the present invention, the edge computing screening model is a data processing model created for edge nodes, including:

[0090] Analyze the target data for the edge nodes, obtain the target data features, and obtain the edge node processing data features;

[0091] Building a recognition and screening model according to the data features processed by the edge node to obtain a first building model;

[0092] Perform security analysis based on the characteristics of data processed by edge nodes to determine the security performance of data processed by edge nodes, and determine the security performance enhancement strategy based on the security performance of data processed by edge nodes to obtain a security assurance strategy;

[0093] Based on edge computing and security assurance strategies, a secondary model is constructed for the first construction model to obtain an edge computing screening model.

[0094] In the above technical solution, the first constructed model is an identification and screening model for edge node processing data features.

[0095] In the above technical solution, the edge computing screening model realizes edge computing of the monitoring and collection data in the edge node.

[0096] In the above technical solution, when determining the security performance enhancement strategy based on the security performance of edge node processing data, the security performance of edge node processing data is analyzed to determine whether the edge node processing data is safe and obtain the security analysis result. When the security analysis result shows that the edge node processing data is not safe, the data type analysis is performed on the edge node processing data to determine the edge node processing data type, and whether the edge node processing data can be disassembled according to the edge node processing data type. If the edge node processing data can be disassembled, the distribution and data importance analysis are performed on the edge node processing data, and the edge node processing data is split into multiple edge node processing data sets according to the data importance analysis result, and then the security enhancement scheme is matched for the edge node processing data set in combination with the data importance analysis result to obtain the target security enhancement processing scheme, and then the security performance enhancement strategy is determined to obtain the security assurance strategy. If the edge node processing data cannot be disassembled, the data importance analysis is performed on the edge node processing data to determine the situation with the highest data importance in the edge node processing data, and the security enhancement scheme is matched according to the situation with the highest data importance in the edge node processing data to obtain the target security enhancement processing scheme, and then the security performance enhancement strategy is determined to obtain the security assurance strategy.

[0097] The above technical solution combines the data processing characteristics of edge nodes to build a model, so that there is a corresponding relationship between the edge computing screening model and the edge node. Therefore, in the edge node, the edge computing screening model can be used to better perform edge computing on the corresponding monitoring and collection data, thereby improving the edge computing efficiency of the edge node, improving the reading and utilization performance of the monitoring and collection processing data, and ensuring the security of the monitoring and collection processing data.

[0098] like Figure 3 As shown, in one embodiment provided by the present invention, the concurrent data transfer control model includes: a node access unit, a forwarding analysis unit, a traffic diversion control unit and a data sending unit;

[0099] The node access unit is used to access the edge node and obtain monitoring, collection and processing data for the edge node;

[0100] The forwarding analysis unit is used to perform data screening and upload analysis on the monitoring, collection and processing data to obtain data screening results and upload analysis data;

[0101] The diversion control unit is used to divert and buffer the monitoring, collection and processing data according to the data screening results in combination with the uploaded analysis data, and to control the buffer sequence for the buffer;

[0102] The data sending unit is used to send and upload the data packets to the cloud computing center according to the buffer sequence of the cache area.

[0103] In the above technical solution, the node access unit, the forwarding analysis unit, the diversion control unit and the data sending unit are connected in sequence.

[0104] In the above technical solution, the node access unit obtains monitoring, collection and processing data from the edge node in real time.

[0105] In the above technical solution, when the monitoring, acquisition and processing data are diverted and buffered according to the data screening result in combination with the uploaded analysis data, the monitoring, acquisition and processing data are formed into a buffer sequence in the buffer in the form of data packets.

[0106] The above technical solution realizes block processing of the concurrent data transfer control model through the node access unit, forwarding analysis unit, diversion control unit and data sending unit, so that the node access unit, forwarding analysis unit, diversion control unit and data sending unit can be operated based on different CPUs, thereby improving the parallelism of the concurrent data transfer control model, avoiding excessive CPU load affecting the concurrent data transfer control model to upload the monitoring, collection and processing data to the cloud computing center, and can also reduce the invalid performance overhead caused by switching between CPUs. At the same time, it can also ensure the stability of data upload, so that the cloud computing center can better realize AI big data automatic analysis and provide protection for the cloud computing center.

[0107] In one embodiment of the present invention, the forwarding analysis unit performs data screening and upload analysis on the monitoring, collection and processing data, including:

[0108] Determining a data packet of the monitoring, collecting and processing data based on the monitoring, collecting and processing data;

[0109] Performing data packet feature recognition on the data packet, obtaining the data packet recognition feature, and judging whether the data packet conforms to the target data flow according to the data packet recognition feature analysis, and obtaining the recognition judgment result;

[0110] The data packets are screened according to the identification and judgment results to obtain data screening results;

[0111] Different analysis methods are used to analyze the uploaded data according to the data screening results. When the data screening result is that the data packet conforms to the target data stream, the first analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet; when the data screening result is that the data packet does not conform to the target data stream, the second analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet.

[0112] In the above technical solution, the data packet characteristics include: protocol, port and encoding identifier and other characteristics.

[0113] In the above technical solution, the target data stream is a data packet with preset characteristic feature attributes.

[0114] In the above technical solution, the first analysis method is different from the second analysis method. The first analysis method is an upload analysis method formed by combining the characteristics of the target data stream, and the second analysis method is an upload analysis method formed by combining the characteristics of the non-target data stream.

[0115] The above technical solution is not only convenient for analysis in the form of data packets, but also can be uploaded directly based on data packets, which provides convenience for the data sending unit to upload data. Moreover, by obtaining the data packet identification features and analyzing the data packet identification features to determine whether the data packet conforms to the characteristics of the target data flow, it is possible to screen based on the identification and judgment results, thereby dividing the data packets into two categories. Then, different analysis methods can be used according to different data screening results to better analyze the uploaded data, thereby ensuring the efficiency and effectiveness of data uploading.

[0116] In one embodiment provided by the present invention, the shunt control unit performs shunt buffering on the monitoring, collection and processing data in combination with the uploaded analysis data according to the data screening result, and performs buffer sequence control on the buffer, including:

[0117] Divert the data packets of the monitored, collected and processed data according to the data screening results. When the data screening results show that the data packets meet the target data flow, the corresponding data packets are diverted to the first buffer. When the data screening results show that the data packets do not meet the target data flow, the corresponding data packets are diverted to the second buffer.

[0118] In the first buffer and / or the second buffer, the sending port information is acquired for the data generating end to determine the sending characteristics of the sending port and simultaneously obtain the current buffering status of the buffer sequence. The sending characteristics of the sending port are combined with the current buffering status of the buffer sequence according to the correspondence between the sending port and the buffer sequence to perform upload data flow analysis to obtain upload traffic analysis data. Then, based on the data screening result, an optimization analysis is performed according to the upload traffic analysis data and the upload analysis data of the data packet to determine the buffer target sequence of the data packet and buffer the data packet to the buffer target sequence.

[0119] In the above technical solution, the first buffer and the second buffer correspond to a data sending end respectively, the data sending end includes multiple sending ports, and the first buffer and the second buffer include multiple buffer sequences respectively, each buffer sequence corresponds to a sending port.

[0120] The above technical solution realizes the diversion of different buffers of data packets through the diversion control unit, so that data can be uploaded synchronously based on the first buffer and the second buffer, thereby improving the parallelism and upload efficiency of the data sending unit, and optimizing and analyzing the uploaded traffic analysis data and the uploaded analysis data of the data packet based on the data screening result, so that when the buffer sequence of the data packet in the corresponding buffer is determined, the buffer can be regulated in combination with the current situation of the buffer sequence and the performance of the sending port, thereby improving the balance of the buffer sequence in the same buffer, so that the data sending unit can upload data more stably based on the buffer sequence when uploading data, thereby ensuring the stability of data upload.

[0121] In one embodiment provided by the present invention, when performing AI big data automated analysis on monitoring, collecting and processing data, an AI big data intelligent analysis model is used to perform automated analysis on the monitoring, collecting and processing data, including:

[0122] Identify monitoring, acquisition and processing data and determine the characteristics of the monitoring, acquisition and processing data;

[0123] Configure a big data analysis solution based on the characteristics of the monitored, collected and processed data to obtain an AI big data analysis solution;

[0124] According to the AI ​​big data analysis plan, retrieve the big data AI model and obtain the AI ​​big data target analysis model;

[0125] The AI ​​big data target analysis model is used to automatically analyze the monitoring collection and processing data to obtain the monitoring data analysis results.

[0126] In the above technical solution, the characteristics of the monitored, collected and processed data include: images, texts, sounds, etc.

[0127] In the above technical solution, a big data AI model is established in advance for different features, and the established big data AI model is saved.

[0128] In the above technical scheme, the big data AI model is retrieved according to the AI ​​big data analysis scheme, including: performing feature analysis according to the AI ​​big data analysis scheme to obtain the features of the AI ​​big data analysis scheme; determining the number of the features of the AI ​​big data analysis scheme, and when there is only one AI big data analysis scheme feature, performing a matching degree analysis in the big data AI model library according to the features of the AI ​​big data analysis scheme to obtain the best big data AI model, and then performing a suitability test on the best big data AI model. If the suitability test passes, the best big data AI model is the AI ​​big data target analysis model. If the suitability test fails, the AI ​​big data model is established according to the features of the AI ​​big data analysis scheme to obtain the AI ​​big data initial model, and then model training and optimization are performed on the AI ​​big data initial model, and the trained and optimized AI big data initial model is used as the AI ​​big data target analysis model, and at the same time, the trained and optimized AI big data initial model is stored in the big data AI model library in combination with the features of the AI ​​big data analysis scheme; when there are two or more AI big data analysis scheme features, a hierarchical analysis is performed on the features of the AI ​​big data analysis scheme to determine the AI ​​big data. The first feature of the data analysis solution and the second feature of the AI ​​big data analysis solution are matched in the big data AI model library according to the first feature of the AI ​​big data analysis solution to obtain the first feature matching result, and the second feature matching is performed in the first feature matching result according to the second feature of the AI ​​big data analysis solution, and the target number of AI big data models are screened to best correspond to the matching result of the second feature of the AI ​​big data analysis solution to obtain the second feature matching result, and then the same model analysis is performed on the second feature matching result. If the same AI big data model exists in the second feature matching result, the same AI big data model is the AI ​​big data target analysis model. If the same AI big data model does not exist in the second feature matching result, the model commonality analysis is performed on the first feature matching result to determine the AI ​​big data basic model, and then the AI ​​big data basic model is improved and optimized according to the second feature of the AI ​​big data analysis solution to obtain the optimized AI big data improved model, and the optimized AI big data improved model is used as the AI ​​big data target analysis model, and at the same time, the optimized AI big data improved model is stored in the big data AI model library in combination with the characteristics of the AI ​​big data analysis solution.

[0129] The above technical solution combines AI big data analysis to realize automated analysis of monitoring, collection and processing data. It can not only improve the efficiency of obtaining monitoring data analysis results, making it possible to analyze a large amount of monitoring, collection and processing data in a short period of time, but also does not require human participation, saving manpower consumption, reducing the probability of errors and the influence of human subjective factors, and ensuring the accuracy of monitoring data analysis results.

[0130] In one embodiment provided by the present invention, when feedback is provided for the monitoring data analysis results, data storage and AI big data automated analysis and recording are performed for the monitoring data analysis results and the monitoring collection and processing data according to the monitoring data analysis results, and at the same time, the result feedback type is determined, and the format of the monitoring data analysis results is converted according to the result feedback type to obtain the monitoring data analysis results in the target format, and then the monitoring data analysis results in the target format are fed back according to the result feedback type.

[0131] In the above technical solution, the result feedback types include: result response feedback and information presentation feedback.

[0132] In the above technical scheme, when the feedback type is result response feedback, the feedback target is determined, identification information is obtained for the feedback target, the identification information is combined with the monitoring data analysis result to generate a result response feedback signal, and then the target position feedback is performed for the result response feedback signal. When the feedback type is information presentation feedback, the presentation information is determined based on the monitoring data analysis result, and the presentation is performed based on the presentation information.

[0133] When the above technical solution provides feedback on the monitoring data analysis results, it can not only record and store the AI ​​big data automated analysis of the cloud computing center, but also provide feedback according to different result feedback types. It can be applied to different situations and provide convenience for result feedback.

[0134] like Figure 4 As shown, an embodiment of the present invention provides an AI big data intelligent processing system based on edge computing, including: a monitoring and collection module, an edge computing module, a data upload module, an AI analysis module and a result feedback module;

[0135] The monitoring and collection module is used to determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node;

[0136] The edge computing module is used to perform edge computing on the monitoring and collection data through the edge node to obtain the monitoring and collection processing data;

[0137] The data uploading module is used to upload the monitoring, collection and processing data to the cloud computing center based on the concurrent data transfer control model;

[0138] The AI ​​analysis module is used to perform AI big data automated analysis on the monitoring, collection and processing data to obtain monitoring data analysis results;

[0139] The result feedback module is used to provide result feedback for the monitoring data analysis results.

[0140] In the above technical scheme, an AI big data intelligent processing system based on edge computing is used to implement an AI big data intelligent processing method based on edge computing. The AI ​​big data intelligent processing system and the AI ​​big data intelligent processing method are one-to-one corresponding. The technical effect of the AI ​​big data intelligent processing system based on edge computing has been explained in the corresponding method embodiment and will not be repeated here.

[0141] Those skilled in the art should understand that the first and second in the present invention merely refer to different application stages.

[0142] Other embodiments of the present disclosure will be readily apparent to those skilled in the art after considering the specification and practicing the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0143] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An AI big data intelligent processing method based on edge computing, characterized in that: include: Determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node; Perform edge computing on the monitoring and collection data through edge nodes to obtain monitoring and collection processing data; Upload the monitored, collected and processed data to the cloud computing center based on the concurrent data transfer control model; In the cloud computing center, the monitoring, collection and processing data are received, and AI big data automated analysis is performed on the monitoring, collection and processing data to obtain the monitoring data analysis results; Provide feedback based on the monitoring data analysis results.

2. The AI ​​big data intelligent processing method according to claim 1 is characterized in that: When monitoring and collecting source data according to the data source, the monitoring and collecting device is deployed in advance in the data source for monitoring, and the edge node is configured according to the monitoring and collecting device to establish the relationship between the edge node and the monitoring and collecting device.

3. The AI ​​big data intelligent processing method according to claim 1 is characterized in that: Edge computing is performed on monitoring and collection data through edge nodes, including: Perform reception detection on the monitoring and collection data, analyze the compatibility of the monitoring and collection data with the edge nodes, and obtain preliminary detection results; According to the preliminary detection result, when the monitoring and collection data are adapted to the edge node, a first analysis is performed on the received monitoring and collection data to determine whether there is missing data in the monitoring and collection data, and a first analysis result is obtained; Supplement the monitoring and collection data according to the first analysis result to obtain the first processed data of the monitoring and collection; Performing a second analysis on the first processed data collected by monitoring, determining whether there is information that needs to be corrected in the first processed data collected by monitoring, and obtaining a second analysis result; Correcting the first processed data collected by monitoring according to the second analysis result to obtain second processed data collected by monitoring; The edge computing screening model is used to perform model processing on the second processed data collected by monitoring to obtain the third processed data collected by monitoring.

4. The AI ​​big data intelligent processing method according to claim 3 is characterized in that: The edge computing screening model is a data processing model created for edge nodes, including: Analyze the target data for the edge nodes, obtain the target data features, and obtain the edge node processing data features; Building a recognition and screening model according to the data features processed by the edge node to obtain a first building model; Perform security analysis based on the characteristics of data processed by edge nodes to determine the security performance of data processed by edge nodes, and determine the security performance enhancement strategy based on the security performance of data processed by edge nodes to obtain a security assurance strategy; Based on edge computing and security assurance strategies, a secondary model is constructed for the first construction model to obtain an edge computing screening model.

5. The AI ​​big data intelligent processing method according to claim 1 is characterized in that: The concurrent data transfer control model includes: node access unit, forwarding analysis unit, diversion control unit and data sending unit; The node access unit is used to access the edge node and obtain monitoring, collection and processing data for the edge node; The forwarding analysis unit is used to perform data screening and upload analysis on the monitoring, collection and processing data to obtain data screening results and upload analysis data; The diversion control unit is used to divert and buffer the monitoring, collection and processing data according to the data screening results in combination with the uploaded analysis data, and to control the buffer sequence for the buffer; The data sending unit is used to send and upload the data packets to the cloud computing center according to the buffer sequence of the cache area.

6. The AI ​​big data intelligent processing method according to claim 5 is characterized in that: The forwarding analysis unit performs data screening and upload analysis on the monitoring, collection and processing data, including: Determining a data packet of the monitoring, collecting and processing data based on the monitoring, collecting and processing data; Performing data packet feature recognition on the data packet, obtaining the data packet recognition feature, and judging whether the data packet conforms to the target data flow according to the data packet recognition feature analysis, and obtaining the recognition judgment result; The data packets are screened according to the identification and judgment results to obtain data screening results; Different analysis methods are used to analyze the uploaded data according to the data screening results. When the data screening result is that the data packet conforms to the target data stream, the first analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet; when the data screening result is that the data packet does not conform to the target data stream, the second analysis method is used to analyze the uploaded data for the data packet, determine the upload parameters of the data packet, and obtain the upload analysis result of the data packet.

7. The AI ​​big data intelligent processing method according to claim 6 is characterized in that: The shunt control unit performs shunt buffering on the monitored, collected and processed data in combination with the uploaded analysis data according to the data screening result, and performs buffer sequence control on the buffer, including: Divert the data packets of the monitored, collected and processed data according to the data screening results. When the data screening results show that the data packets meet the target data flow, the corresponding data packets are diverted to the first buffer. When the data screening results show that the data packets do not meet the target data flow, the corresponding data packets are diverted to the second buffer. In the first buffer and / or the second buffer, the sending port information is acquired for the data generating end to determine the sending characteristics of the sending port and simultaneously obtain the current buffering status of the buffer sequence. The sending characteristics of the sending port are combined with the current buffering status of the buffer sequence according to the correspondence between the sending port and the buffer sequence to perform upload data flow analysis to obtain upload traffic analysis data. Then, based on the data screening result, an optimization analysis is performed according to the upload traffic analysis data and the upload analysis data of the data packet to determine the buffer target sequence of the data packet and buffer the data packet to the buffer target sequence.

8. The AI ​​big data intelligent processing method according to claim 1 is characterized in that: When performing AI big data automated analysis on monitoring, collection and processing data, an AI big data intelligent analysis model is used to perform automated analysis on monitoring, collection and processing data, including: Identify monitoring, acquisition and processing data and determine the characteristics of the monitoring, acquisition and processing data; Configure a big data analysis solution based on the characteristics of the monitored, collected and processed data to obtain an AI big data analysis solution; According to the AI ​​big data analysis plan, retrieve the big data AI model and obtain the AI ​​big data target analysis model; The AI ​​big data target analysis model is used to automatically analyze the monitoring collection and processing data to obtain the monitoring data analysis results.

9. The AI ​​big data intelligent processing method according to claim 1, characterized in that: When feedback is provided for the monitoring data analysis results, data storage and AI big data automated analysis records are performed for the monitoring data analysis results and the monitoring collection and processing data according to the monitoring data analysis results. At the same time, the result feedback type is determined, and the format of the monitoring data analysis results is converted according to the result feedback type to obtain the monitoring data analysis results in the target format. The monitoring data analysis results in the target format are then fed back according to the result feedback type.

10. An AI big data intelligent processing system based on edge computing, characterized in that: The AI ​​big data intelligent processing system includes: a monitoring and acquisition module, an edge computing module, a data upload module, an AI analysis module and a result feedback module; The monitoring and collection module is used to determine the data source, monitor and collect the source data according to the data source, obtain the monitoring and collection data, and transmit the monitoring and collection data to the edge node; The edge computing module is used to perform edge computing on the monitoring and collection data through the edge node to obtain the monitoring and collection processing data; The data uploading module is used to upload the monitoring, collection and processing data to the cloud computing center based on the concurrent data transfer control model; The AI ​​analysis module is used to perform AI big data automated analysis on the monitoring, collection and processing data to obtain monitoring data analysis results; The result feedback module is used to provide result feedback for the monitoring data analysis results.

Images