Sensor scene identifying and transferring method in edge server

Abstract

The invention discloses a method for identifying and transferring a sensor scene in an edge server. The method comprises the following steps: acquiring a current scene; analyzing according to the current scene, and determining an algorithm matched with the current scene through scene recognition; when the scene changes, algorithm updating and scene evolution are carried out; acquiring real-time sensor data of the current scene; analyzing the real-time sensor data through an intelligent diagnosis and decision support module, determining an equipment state predicted value and a decision suggestion, and generating a diagnosis report; generating a corresponding diagnosis model according to the current scene; and analyzing the real-time sensor data of the current scene and the historical data set of the sensor through the diagnosis model, determining a future state prediction value of the equipment, and generating an intelligent adjustment suggestion. Through the comprehensive application of scene-driven algorithm selection, intelligent scene evolution, local calculation and data fusion, the edge server can provide customized and accurate intelligent diagnosis and decision support in different scenes.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular, to a method for identifying and mobilizing sensor scenes in an edge server. Background Art

[0002] With the popularization of the Internet of Things and smart devices, a single algorithm cannot meet the different needs in multiple scenarios. Future edge servers will need to have scenario-driven intelligent diagnosis functions, that is, to select and call different algorithms for local computing and intelligent diagnosis according to the specific needs of the scenario. This means that in different scenarios, the system will automatically switch to different algorithms to most effectively solve the problems in the scenario, improve the accuracy of diagnosis and the real-time nature of decision-making.

[0003] For example, in the monitoring scenario of power distribution cabinets, edge servers need to have diagnostic capabilities such as health status assessment, load monitoring, and temperature anomalies of power equipment; while in the fire monitoring scenario, diagnostic algorithms based on smoke sensor data and fire alarm thresholds are required. Each scenario requires different algorithms, which need to be optimized and selected based on real-time sensor data and historical data learning.

[0004] However, existing sensor scene recognition and mobilization methods cannot call corresponding algorithms for intelligent diagnosis according to scene requirements, nor can they adjust algorithms when the scene changes.

[0005] Therefore, the prior art has defects and is in urgent need of improvement. Summary of the invention

[0006] In view of the above problems, the purpose of the present invention is to provide a method for identifying and mobilizing sensor scenarios in edge servers, which enables edge servers to provide customized, accurate intelligent diagnosis and decision support in different scenarios through the comprehensive application of scenario-driven algorithm selection, intelligent scenario evolution, local computing and data fusion. This not only greatly improves the adaptability of edge servers in industrial environments and intelligent application scenarios, but also enables them to have long-term development potential in scenario evolution and algorithm upgrades.

[0007] A first aspect of the present invention provides a method for identifying and mobilizing a sensor scene in an edge server, comprising:

[0008] Get the current scene;

[0009] Analyze the current scene and determine an algorithm that matches the current scene through scene recognition;

[0010] Monitor the current scene, and update the algorithm and evolve the scene when the scene changes;

[0011] Get real-time sensor data of the current scene;

[0012] Analyze the real-time sensor data through the intelligent diagnosis and decision support module, determine the equipment status prediction value and decision suggestions, and generate a diagnosis report;

[0013] Generate a corresponding diagnosis model according to the current scenario;

[0014] The diagnostic model is used to analyze the real-time sensor data of the current scene and the historical data set of the sensor to determine the predicted value of the future state of the equipment and generate intelligent adjustment suggestions.

[0015] In this solution, the analysis based on the current scene and determining the algorithm matching the current scene through scene recognition include:

[0016] Define a scene set S = {S1, S2, ..., S n} and algorithm set A={A1,A2,...,A m};

[0017] Establish the current scene S according to the scene set S and the algorithm set A j and matching algorithm A i The scene mapping formula;

[0018] A i =f(S j ,D,T);

[0019] Among them, S j is the current scene, S j ∈S, D is the sensor data set, D={d1,d2,...,d k}, T is the sensor type set, T={t1,t2,...,t k}, f is the mapping function.

[0020] This plan also includes:

[0021] When the number of scenes increases or the number of sensors in a scene increases, the scene-algorithm mapping matrix M[a,b] is automatically generated through multi-dimensional vectors to expand the algorithm;

[0022] M[a,b]=f(S a ,T b );

[0023] Among them, M[a,b] represents the scene S a and sensor T b The corresponding scene-algorithm mapping matrix, sensor T b For scene S a The bth type of sensor configured inside.

[0024] In this solution, the current scene is monitored, and the algorithm is updated and the scene is evolved when the scene changes, including:

[0025] Through data learning, the current algorithm A t Perform algorithm optimization and determine the optimized algorithm A t+1 ;

[0026]

[0027] Among them, η is the learning rate, is the gradient based on the loss function L, H is the historical data set, and R is the real-time data set;

[0028] According to the historical scene data S history and the current scene S j Predict new scene S new ;

[0029] S new =g(S history ,S j ,D);

[0030] Among them, g is the scene prediction function, and D is the sensor data set;

[0031] The new scene S new With the optimized algorithm A t+1 to make a match.

[0032] In this solution, the real-time sensor data is analyzed by the intelligent diagnosis and decision support module to determine the equipment status prediction value and decision suggestions, and generate a diagnosis report, including:

[0033] The real-time sensor data D at time t is analyzed by preset intelligent diagnosis formula (t) Perform diagnostic analysis to determine the predicted value E^(t) of the equipment status at time t;

[0034] E^(t)=h(T scene(t) ,D (t) ,H);

[0035] Among them, T scene(t) is the scene adaptive threshold, D (t) is the real-time sensor data at time t, H is the historical data set, and h is the diagnostic function;

[0036] Analyze the equipment state prediction value E^(t) based on the rule base R to generate a decision suggestion P;

[0037] P = q(E^(t), R);

[0038] Among them, q is the decision function;

[0039] Generate a diagnostic report based on the equipment state prediction value E^(t) and the decision suggestion P.

[0040] In this solution, generating a corresponding diagnostic model according to the current scenario includes:

[0041] Determine the scene S based on the scene model generation formula j and diagnostic model M j The generative relationship of

[0042] M j =p(S j ,D,C);

[0043] Among them, C is the scene constraint, D is the sensor data set, and p is the constraint function.

[0044] This plan also includes:

[0045] When the new scene S new When it appears, the new scene S is automatically generated according to the scene model generation formula new The corresponding diagnostic model M new .

[0046] In this solution, the real-time sensor data of the current scene and the historical data set of the sensor are analyzed by the diagnostic model to determine the predicted value of the future state of the equipment and generate intelligent adjustment suggestions, including:

[0047] Perform data fusion on the real-time sensor data and the sensor's historical data set to obtain the fused data set F (t) ;

[0048] F (t) =α×D (t) +β×H;

[0049] Among them, α is the fusion weight of real-time sensor data, β is the fusion weight of the sensor's historical data set, and D (t) is the real-time sensor data at time t, H is the historical data set;

[0050] Through the fused data set F (t) Generate the predicted value of the future state of the equipment E′^(t);

[0051] E′^(t)=h(F(t),M);

[0052] Among them, h is the diagnostic function and M is the diagnostic model;

[0053] Analyze the predicted value E′^(t) of the future state of the equipment based on the rule base R to generate an intelligent adjustment suggestion Q;

[0054] Q = q(E′^(t), R);

[0055] Among them, q is the decision function.

[0056] The present invention discloses a method for identifying and mobilizing sensor scenes in an edge server, the method comprising: obtaining the current scene; analyzing the current scene, and determining an algorithm that matches the current scene through scene identification; updating the algorithm and evolving the scene when the scene changes; obtaining real-time sensor data of the current scene; analyzing the real-time sensor data through an intelligent diagnosis and decision support module, determining a device state prediction value and a decision suggestion, and generating a diagnosis report; generating a corresponding diagnosis model according to the current scene; analyzing the real-time sensor data of the current scene and a set of historical data of the sensor through the diagnosis model, determining a device future state prediction value, and generating an intelligent adjustment suggestion. The present invention enables the edge server to provide customized, accurate intelligent diagnosis and decision support in different scenarios through the comprehensive application of scenario-driven algorithm selection, intelligent scenario evolution, local computing, and data fusion. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A flow chart of a sensor scene recognition and mobilization method in an edge server provided by the present invention is shown;

[0058] Figure 2 A flow chart showing the algorithm update and scene evolution method provided by the present invention is shown;

[0059] Figure 3 A flow chart of the diagnostic report generating method provided by the present invention is shown. DETAILED DESCRIPTION

[0060] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0061] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0062] Figure 1 A flow chart of a sensor scene recognition and mobilization method in an edge server provided by the present invention is shown.

[0063] like Figure 1 As shown, the present invention discloses a method for identifying and mobilizing a sensor scene in an edge server, comprising:

[0064] S102, obtaining the current scene;

[0065] S104, analyzing the current scene and determining an algorithm matching the current scene through scene recognition;

[0066] S106, monitoring the current scene, and updating the algorithm and evolving the scene when the scene changes;

[0067] S108, obtaining real-time sensor data of the current scene;

[0068] S110, analyzing the real-time sensor data through the intelligent diagnosis and decision support module, determining the equipment status prediction value and decision suggestions, and generating a diagnosis report;

[0069] S112, generating a corresponding diagnosis model according to the current scenario;

[0070] S114, analyzing the real-time sensor data of the current scene and the historical data set of the sensor through the diagnostic model, determining the predicted value of the future state of the equipment, and generating intelligent adjustment suggestions.

[0071] According to an embodiment of the present invention, the edge server automatically selects an algorithm that matches the current scene through scene recognition to ensure the efficiency and accuracy of the system in different scenes. According to the real-time sensor data, device type (sensor type) and application scenario of the current scene, the corresponding algorithm is automatically called through the scene mapping formula to achieve intelligent diagnosis. Among them, the scene is set according to actual needs, such as power monitoring scene, fire monitoring scene, environmental monitoring scene, intelligent traffic scene, etc.

[0072] As devices continue to evolve and scenarios change, edge servers can automatically adapt and update algorithms through intelligent scenario evolution and algorithm adaptation mechanisms, determine algorithms that match new scenarios, and automatically update algorithms when scenarios change, so that they can better solve problems that arise in new scenarios. Through data learning and algorithm optimization mechanisms, the system gradually accumulates experience and continuously improves the accuracy and execution efficiency of the algorithm. This algorithm adaptive evolution mechanism not only improves the efficiency of edge computing, but also ensures that devices can always meet new industrial needs.

[0073] The edge server can perform intelligent diagnosis locally, using local computing power to generate diagnostic reports and decision recommendations in real time, without relying on cloud processing, reducing cloud dependence and ensuring the real-time and reliability of the scene. The real-time sensor data is analyzed through intelligent diagnostic formulas, and the device status prediction value is output. The device status prediction value is analyzed based on the rule base to generate decision recommendations. This local calculation not only saves cloud bandwidth, but also ensures the real-time and reliability of the scene without a network connection. Combined with sensor data, the algorithm will conduct an in-depth analysis of the device or environment and automatically generate a multi-dimensional intelligent diagnostic report to provide users with intelligent decision support.

[0074] In addition, the system can generate independent diagnostic models for different scenarios to ensure high adaptability and flexibility. It also performs intelligent scenario diagnosis through an intelligent diagnostic mechanism that integrates real-time data with historical data, integrates real-time sensor data with historical data, and determines the predicted value of the future state of the equipment through data learning and prediction. It analyzes the predicted value of the future state of the equipment based on the rule base, predicts the possible future state of the equipment, and provides preventive maintenance or intelligent adjustment suggestions based on this. Through the comprehensive application of scenario-driven algorithm selection, intelligent scenario evolution, local computing and data fusion, efficient diagnosis and decision support of edge servers in complex industrial scenarios can be achieved.

[0075] According to an embodiment of the present invention, analyzing the current scene and determining an algorithm matching the current scene through scene recognition includes:

[0076] Define a scene set S = {S1, S2, ..., S n} and algorithm set A={A1,A2,...,A m};

[0077] Establish the current scene S according to the scene set S and algorithm set A j and matching algorithm A i The scene mapping formula;

[0078] A i =f(S j ,D,T);

[0079] Among them, S j is the current scene, S j ∈S, D is the sensor data set, D={d1,d2,...,d k}, T is the sensor type set, T={t1,t2,...,t k}, f is the mapping function.

[0080] It should be noted that the edge server automatically selects the algorithm that matches the current scenario through scene recognition to ensure the efficiency and accuracy of the system in different scenarios. This dynamic selection and calling mechanism not only improves the adaptability of the system, but also lays the foundation for multi-scenario intelligent diagnosis.

[0081] The specific process of scene recognition and algorithm calling is to identify the current scene S through the scene set j (such as electricity, fire protection, etc.), according to the current scene S j Analyze the sensor type set T in the current scene and call the scene mapping formula to match the current scene S j Matching Algorithm A i For example, if the current scene S j For power monitoring scenarios, call algorithm A i is the load assessment and temperature and humidity monitoring algorithm for the distribution cabinet. If the current scenario S j For fire monitoring scenarios, call algorithm A i For smoke detection and fire alarm algorithms.

[0082] According to an embodiment of the present invention, it also includes:

[0083] When the number of scenes increases or the number of sensors in a scene increases, the scene-algorithm mapping matrix M[a,b] is automatically generated through multi-dimensional vectors to expand the algorithm;

[0084] M[a,b]=f(S a ,T b );

[0085] Among them, M[a,b] represents the scene S a and sensor T b The corresponding scene-algorithm mapping matrix, sensor T b For scene S a The bth type of sensor configured inside.

[0086] It should be noted that when the number of monitoring scenarios of the edge server increases or the sensor data in any monitoring scenario increases, the system automatically generates the scene-algorithm mapping matrix M[a, b] through a multi-dimensional vector and automatically calls the algorithm that matches the updated scene to ensure the adaptability and scalability of scene recognition.

[0087] Figure 2 A flow chart of the algorithm updating and scene evolution method provided by the present invention is shown.

[0088] like Figure 2 As shown, according to an embodiment of the present invention, the current scene is monitored, and when the scene changes, the algorithm is updated and the scene evolves, including:

[0089] S202, the current algorithm A is improved through data learning t Perform algorithm optimization and determine the optimized algorithm A t+1 ;

[0090]

[0091] Among them, η is the learning rate, is the gradient based on the loss function L, H is the historical data set, and R is the real-time data set;

[0092] S204, based on historical scene data S history and the current scene S j Predict new scene S new ;

[0093] S new =g(S history ,S j ,D);

[0094] Among them, g is the scene prediction function, and D is the sensor data set;

[0095] S206, the new scene S new With the optimized algorithm A t+1 to make a match.

[0096] It should be noted that the system can automatically update the algorithm when the scene changes through data learning and algorithm optimization. This adaptive mechanism supports the continuous evolution of the system as industrial needs change. t Input into the algorithm optimization formula, control the optimization step size by setting the learning rate η, based on the gradient of the loss function L For the current algorithm A t , analyze the historical data set H and real-time data set R of the sensor to determine the optimized algorithm A t+1 Then the system uses the scenario evolution formula according to the historical scenario data S history and the current scene S j Predict new scene S new , when determining the new scene S new After that, call the optimized algorithm A t+1 This intelligent adaptive algorithm optimization based on historical data and real-time data enables the edge server to continuously optimize its diagnostic capabilities in the face of changing scenarios.

[0097] In addition, when the number and complexity of scenarios increase, the system can expand the scenario evolution capability through scene vectorization and automatic clustering technology to generate new scenario algorithm matching rules.

[0098] Figure 3A flow chart of the diagnostic report generating method provided by the present invention is shown.

[0099] like Figure 3 As shown, according to an embodiment of the present invention, the real-time sensor data is analyzed by an intelligent diagnosis and decision support module to determine the device state prediction value and decision suggestions, and generate a diagnosis report, including:

[0100] S302, using a preset intelligent diagnosis formula to diagnose the real-time sensor data D at time t (t) Perform diagnostic analysis to determine the predicted value E^(t) of the equipment status at time t;

[0101] E^(t)=h(T scene(t) ,D (t) ,H);

[0102] Among them, T scene(t) is the scene adaptive threshold, D (t) is the real-time sensor data at time t, H is the historical data set, and h is the diagnostic function;

[0103] S304, analyzing the equipment state prediction value E^(t) based on the rule base R, and generating a decision suggestion P;

[0104] P = q(E^(t), R);

[0105] Among them, q is the decision function;

[0106] S306, generating a diagnosis report according to the device state prediction value E^(t) and the decision suggestion P.

[0107] It should be noted that the edge server can perform intelligent diagnosis locally without relying on cloud processing, and provide instant diagnosis reports and decision support for the scenario. (t) Input into the intelligent diagnosis formula, call the diagnosis function, and combine the scene adaptive threshold T at time t scene(t) The sensor's historical data set H is analyzed to output the predicted value of the device state E^(t) of the sensor device. Based on the rule base R preset by the system, the predicted value of the device state E^(t) is analyzed by calling the decision function, and a decision suggestion P is generated according to the matching decision rule. A multi-dimensional intelligent diagnosis report is automatically generated through the predicted value of the device state E^(t) of the sensor device and the corresponding decision suggestion P, providing intelligent decision support for users.

[0108] This combination of local computing and intelligent diagnosis enables edge servers to respond quickly and process efficiently, greatly improving the intelligence level of the system.

[0109] According to an embodiment of the present invention, generating a corresponding diagnosis model according to the current scenario includes:

[0110] Determine the scene S based on the scene model generation formula j and diagnostic model M j The generative relationship of

[0111] M j =p(S j ,D,C);

[0112] Among them, C is the scene constraint, D is the sensor data set, and p is the constraint function.

[0113] It should be noted that the edge server can not only adapt to different scenarios, but also generate customized diagnostic models according to each specific scenario to meet the specific needs of each industry and each scenario. These scenarios include but are not limited to: power monitoring, fire monitoring, environmental monitoring, intelligent transportation, etc. The sensor data in each scenario will call a customized algorithm model for intelligent analysis and judgment. j Input into the scene model generation formula, and call the constraint function to constrain the scene S j Analyze the corresponding sensor data set D and determine the corresponding diagnostic model M j .

[0114] This customized algorithm model and dynamic scene adaptability for different application scenarios ensures the high adaptability of the diagnostic model and provides great flexibility and intelligent advantages for the system.

[0115] According to an embodiment of the present invention, it also includes:

[0116] When the new scene S new When it appears, a new scene S is automatically generated according to the scene model generation formula new The corresponding diagnostic model M new .

[0117] It should be noted that when determining the new scene S new After that, the new scene S new Input into the scene model generation formula, and call the constraint function to generate the new scene S new Analyze the corresponding sensor data set D and determine the corresponding diagnostic model M new , in order to improve the new scene S new data analysis efficiency.

[0118] According to an embodiment of the present invention, the real-time sensor data of the current scene and the historical data set of the sensor are analyzed by a diagnostic model to determine the predicted value of the future state of the device and generate intelligent adjustment suggestions, including:

[0119] Perform data fusion on the real-time sensor data and the sensor's historical data set to obtain the fused data set F (t) ;

[0120] F (t) =α×D (t) +β×H;

[0121] Among them, α is the fusion weight of real-time sensor data, β is the fusion weight of the sensor's historical data set, and D (t) is the real-time sensor data at time t, H is the historical data set;

[0122] Through the fused dataset F (t) Generate the predicted value of the future state of the equipment E′^(t);

[0123] E′^(t)=h(F(t),M);

[0124] Among them, h is the diagnostic function and M is the diagnostic model;

[0125] Analyze the predicted value of the future state of the equipment E′^(t) based on the rule base R and generate intelligent adjustment suggestions Q;

[0126] Q = q(E′^(t), R);

[0127] Among them, q is the decision function.

[0128] It should be noted that the edge server can perform intelligent scene diagnosis through an intelligent diagnosis mechanism that integrates real-time data and historical data. When performing intelligent scene diagnosis, the edge server multiplies the real-time sensor data and the sensor's historical data set by the corresponding fusion weights, integrates the obtained data, and determines the fused data set F (t) According to the fused dataset F (t) Select the corresponding diagnostic model M, analyze it by calling the diagnostic function h, determine the predicted value of the future state of the equipment E′^(t), predict the possible future state of the equipment, and provide preventive maintenance or intelligent adjustment suggestions based on this.

[0129] Through this fusion mechanism of historical data and real-time data, the edge server's comprehensive perception of devices and environments is enhanced, especially in high-risk environments, which can effectively improve the reliability and security of device operation.

[0130] In addition, the system dynamically adjusts the values ​​of fusion weights α and β according to real-time sensor data to adapt to changes in data characteristics and ensure the flexibility and accuracy of diagnosis. Among them, the values ​​of fusion weights α and β satisfy α+β=1.

[0131] Various functions used in this application (including but not limited to mapping function f, loss function L, scene prediction function g, diagnosis function h, decision function q and constraint function p) are set by those skilled in the art according to requirements.

[0132] The information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals (including but not limited to signals transmitted between user terminals and other devices, etc.) involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions. For example, the "current scene" and "real-time sensor data" involved in this disclosure are all obtained with full authorization.

[0133] The present invention discloses a method for identifying and mobilizing sensor scenes in an edge server, the method comprising: obtaining the current scene; analyzing the current scene, and determining an algorithm that matches the current scene through scene identification; updating the algorithm and evolving the scene when the scene changes; obtaining real-time sensor data of the current scene; analyzing the real-time sensor data through an intelligent diagnosis and decision support module, determining a device state prediction value and a decision suggestion, and generating a diagnosis report; generating a corresponding diagnosis model according to the current scene; analyzing the real-time sensor data of the current scene and a set of historical data of the sensor through the diagnosis model, determining a device future state prediction value, and generating an intelligent adjustment suggestion. The present invention enables the edge server to provide customized, accurate intelligent diagnosis and decision support in different scenarios through the comprehensive application of scenario-driven algorithm selection, intelligent scenario evolution, local computing, and data fusion.

[0134] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0135] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0136] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0137] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, and other media that can store program codes.

[0138] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

Claims

1. A method for mobilizing sensor scene recognition in an edge server, characterized in that: include: Get the current scene; Analyze the current scene and determine an algorithm that matches the current scene through scene recognition; Monitor the current scene, and update the algorithm and evolve the scene when the scene changes; Get real-time sensor data of the current scene; Analyze the real-time sensor data through the intelligent diagnosis and decision support module, determine the equipment status prediction value and decision suggestions, and generate a diagnosis report; Generate a corresponding diagnosis model according to the current scenario; The diagnostic model is used to analyze the real-time sensor data of the current scene and the historical data set of the sensor to determine the predicted value of the future state of the equipment and generate intelligent adjustment suggestions.

2. The method for mobilizing sensor scene recognition in an edge server according to claim 1, characterized in that: The analyzing the current scene and determining an algorithm matching the current scene through scene recognition includes: Define a scene set S = {S1, S2, ..., S n } and algorithm set A={A1,A2,...,A m }; Establish the current scene S according to the scene set S and the algorithm set A j and matching algorithm A i The scene mapping formula; A i =f(S j ,D,T); Among them, S j is the current scene, S j ∈S, D is the sensor data set, D={d1,d2,...,d k }, T is the sensor type set, T={t1,t2,...,t k }, f is the mapping function.

3. The method for mobilizing sensor scene recognition in an edge server according to claim 2, characterized in that: Also includes: When the number of scenes increases or the number of sensors in a scene increases, the scene-algorithm mapping matrix M[a,b] is automatically generated through multi-dimensional vectors to expand the algorithm; M[a,b]=f(S a ,T b ); Among them, M[a,b] represents the scene S a and sensor T b The corresponding scene-algorithm mapping matrix, sensor T b For scene S a The bth type of sensor configured inside.

4. The method for mobilizing sensor scene recognition in an edge server according to claim 1, characterized in that: The monitoring of the current scene and updating the algorithm and evolving the scene when the scene changes include: Through data learning, the current algorithm A t Perform algorithm optimization and determine the optimized algorithm A t+1 ; Among them, η is the learning rate, is the gradient based on the loss function L, H is the historical data set, and R is the real-time data set; According to the historical scene data S history and the current scene S j Predict new scene S new ; S new =g(S history ,S j ,D); Among them, g is the scene prediction function, and D is the sensor data set; The new scene S new With the optimized algorithm A t+1 to make a match.

5. The method for mobilizing sensor scene recognition in an edge server according to claim 1, characterized in that: The intelligent diagnosis and decision support module analyzes the real-time sensor data, determines the equipment status prediction value and decision suggestions, and generates a diagnosis report, including: The real-time sensor data D at time t is analyzed by preset intelligent diagnosis formula (t) Perform diagnostic analysis to determine the predicted value E^(t) of the equipment status at time t; E^(t)=h(T scene(t) ,D (t) ,H); Among them, T scene(t) is the scene adaptive threshold, D (t) is the real-time sensor data at time t, H is the historical data set, and h is the diagnostic function; Analyze the equipment state prediction value E^(t) based on the rule base R to generate a decision suggestion P; P = q(E^(t), R); Among them, q is the decision function; Generate a diagnostic report based on the equipment state prediction value E^(t) and the decision suggestion P.

6. The method for mobilizing sensor scene recognition in an edge server according to claim 1, characterized in that: The generating a corresponding diagnosis model according to the current scenario includes: Determine the scene S based on the scene model generation formula j and diagnostic model M j The generative relationship of M j =p(S j ,D,C); Among them, C is the scene constraint, D is the sensor data set, and p is the constraint function.

7. The method for mobilizing sensor scene recognition in an edge server according to claim 4, characterized in that: Also includes: When the new scene S new When it appears, the new scene S is automatically generated according to the scene model generation formula new The corresponding diagnostic model M new .

8. The method for mobilizing sensor scene recognition in an edge server according to claim 1, characterized in that: The method of analyzing the real-time sensor data of the current scene and the historical data set of the sensor by the diagnostic model, determining the predicted value of the future state of the device, and generating intelligent adjustment suggestions includes: Perform data fusion on the real-time sensor data and the sensor's historical data set to obtain the fused data set F (t) ; F (t) =α×D (t) +β×H; Among them, α is the fusion weight of real-time sensor data, β is the fusion weight of the sensor's historical data set, and D (t) is the real-time sensor data at time t, H is the historical data set; Through the fused data set F (t) Generate the predicted value of the future state of the equipment E′^(t); E′^(t)=h(F(t),M); Among them, h is the diagnostic function and M is the diagnostic model; Analyze the predicted value E′^(t) of the future state of the equipment based on the rule base R to generate an intelligent adjustment suggestion Q; Q = q(E′^(t), R); Among them, q is the decision function.

Images