Remote control interconnection control management system and method based on artificial intelligence

Abstract

The invention discloses a remote control interconnection control management system and method based on artificial intelligence, and relates to the technical field of control management. The system comprises a resource tagging and parameter set construction module, a task adaptation and resource combination scheme generation module, an instruction issuing and execution state feedback module and a state verification and dynamic adjustment module. According to the method, multi-dimensional tagging is carried out on equipment and edge nodes in a system, a resource available list and a parameter set are constructed, task requirements are analyzed, a resource combination scheme is generated, an instruction is issued after a user selects a feedback, an execution state is collected, state consistency is verified in real time, the scheme is dynamically adjusted when the state is abnormal, and resource parameters are updated. According to the method, the problems of low resource utilization rate, insufficient conflict avoidance and poor user dominance in the prior art are solved, the resource utilization rate, the system stability and the user experience are improved, the multi-scene remote control requirement is met, and the practicability and the expandability are high.

Description

Technical Field

[0001] This invention relates to the field of control and management technology, specifically to a remote-controlled interconnected control and management system and method based on artificial intelligence. Background Technology

[0002] With the popularization of IoT technology, the demand for multi-terminal remote control of the same heterogeneous device in scenarios such as smart homes and industrial control is becoming increasingly urgent. Remote control interconnection systems have become a core support for improving device usability. However, existing remote control interconnection control technologies still have shortcomings, which seriously restrict the system's resource utilization and user experience, making it difficult to meet practical application needs.

[0003] Existing technologies generally employ a task-driven forward logic, which involves first receiving control commands from the user and then traversing system device resources for matching and allocation. This logic leads to significant resource waste due to idleness. In real-world scenarios, devices are rarely completely idle; devices operating under low load have the capability to share tasks, but this is not effectively identified and utilized by existing systems, resulting in low resource utilization. Furthermore, when multiple tasks contend for the same device, existing technologies rely on complex conflict resolution processes, which not only increase system response latency but also lead to issues such as misjudging non-essential conflicts and forced allocation that does not meet user expectations.

[0004] Furthermore, existing systems all employ an automatic execution mode, leaving users with no choice in resource allocation. When system-allocated resources conflict with user habits, adjustments cannot be made flexibly, resulting in a poor user experience. Moreover, when the target device is occupied, tasks can only enter a queue, further reducing execution efficiency.

[0005] In summary, existing remote control and interconnection technologies have significant shortcomings in terms of resource utilization, conflict avoidance, and user autonomy. There is an urgent need for a control management system and method that breaks through the traditional task-driven framework, efficiently utilizes various available resources, and balances flexibility and user experience to solve the above-mentioned technical pain points. Summary of the Invention

[0006] The purpose of this invention is to provide a remote control and interconnection management system and method based on artificial intelligence to solve the problems mentioned in the background art.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A remote control and interconnection management method based on artificial intelligence includes the following steps: S1. Real-time traversal of devices and edge nodes within the system, extraction of functional adaptation features, load redundancy features, and scene matching features for multi-dimensional labeling, construction of a dynamically updated list of available resources, and output of a set of basic resource parameters containing the label features and operating status of each resource. S2. Receive user-controlled tasks and parse the core requirements parameters of the tasks, call the resource tag features of the resource basic parameter set, perform adaptation and correlation analysis between the resource tag features and the core requirements parameters of the tasks, generate different resource combination schemes, and associate the schemes with multi-dimensional tag features, including resource collaboration logic and task execution path. S3. Provide feedback to the user on the resource usage, execution path, and applicable scenario descriptions of each solution. After receiving the user's selected instruction, issue collaborative control instructions to the corresponding resources. Collect task execution progress, resource load changes, and feature matching data in real time to form an execution status feedback set containing parameter deviation information. S4. Based on the execution status feedback set, verify the consistency between the resource running status and the resource basic parameter set in real time. If a deviation or abnormal load is detected, call the updated resource availability list and adaptation correlation analysis logic, regenerate the adaptation scheme for resource switching or task execution path adjustment and feed it back to the user. After the user confirms, the adjustment is executed and the adjustment result is sent back to S1 to update the corresponding resource's tag features and resource basic parameter set.

[0008] Furthermore, S1 includes the following: The system traverses devices and edge nodes according to a preset collection cycle, collecting raw operational data for each resource. This raw operational data includes the set of functions supported by the device, real-time load values, the system's preset rated load limit, the current scene identifier determined by collectable environmental parameters, and historical task execution records. Multi-dimensional feature quantification is performed on the raw operational data. Specifically, the function adaptation feature value is calculated based on the degree of matching between the resource's supported functions and a predefined set of system standard functions, with function item weights used for correction. The load redundancy feature value is calculated based on the ratio between the rated load limit and the real-time load value. The scene matching feature value is calculated based on the success rate of tasks matching the current scene identifier in historical task execution records, with time decay factors and scene similarity factors used for correction. Adaptation levels are defined based on the distribution range of various feature values, forming corresponding multi-dimensional level labels. These labels are associated with feature values ​​and basic resource attributes. Resources that meet the preset availability thresholds for all three level labels are selected and included in the available resource range. Available resources are sorted from high to low according to their functional adaptation level, and within the same level, they are sorted from high to low according to their load redundancy feature values, constructing a dynamically updated list of available resources. The multi-dimensional level labels, real-time load rate, rated load limit, and scenario attribute data of each available resource are extracted, where the real-time load rate is the ratio of the current real-time load value to the rated load limit. These are then summarized to form a set of basic resource parameters containing the feature values ​​and operating status of each resource label and output.

[0009] Furthermore, S2 includes the following: The system receives remote control tasks issued by users and extracts the core requirements parameters of the task corresponding to the multi-dimensional feature dimensions of S1 through semantic parsing. The parameters include the functional requirements of the task, the workload requirements of the task, the task execution scenario identifier, and the task execution time requirements. Call the resource basic parameter set output by S1, extract the multi-dimensional level labels of each available resource and the corresponding function adaptation, load redundancy, and scenario matching feature values, and establish a mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantified values ​​of the resources. Based on the association mapping table, the available resources are initially screened to select those with functional adaptation and scenario matching level tags that are not lower than the lowest level corresponding to the task, and whose current remaining load capacity meets the task load requirements, thus forming a set of candidate resources. The overall adaptability of each resource in the candidate resource set to the current task is calculated. This overall adaptability is obtained by multiplying the feature values ​​of functional adaptability, load redundancy, and scenario matching by their respective task adaptability weights, summing the results, and then multiplying by the task correction factor. Among these, the adaptability weights of the three types of features are positively correlated with the task priority and sum to 1. All weights can be fine-tuned according to the task type and stored in the system configuration library. The task correction factor is obtained by multiplying the task timeliness correction factor and the priority correction factor. The timeliness correction factor is calculated based on the task's preset maximum execution time and estimated execution time, while the priority correction factor is set according to the task priority levels. From the candidate resource set, resources with a comprehensive suitability not lower than a preset suitability threshold are further selected as final candidate resources. Combination suitability analysis is then performed on the final candidate resources to generate different resource combination schemes: the load redundancy capability of a single final candidate resource is calculated, which is the product of the resource load redundancy feature quantification value and the rated load upper limit value. If the load redundancy capability of a single resource meets the task load requirements, the resource is combined as an independent resource to form a single resource scheme, and its overall comprehensive suitability is consistent with the resource's own comprehensive suitability. If the load redundancy capability of a single resource cannot meet the requirements, multiple resources are selected sequentially from high to low comprehensive suitability for collaborative combination until the total load redundancy capability after combination meets the task requirements. Then, the overall comprehensive suitability of the combination is calculated, which is the sum of the products of the comprehensive suitability of each resource in the combination and the corresponding load redundancy feature quantification value, divided by the sum of the load redundancy feature quantification values ​​of each resource in the combination.

[0010] Furthermore, S3 includes the following: For the generated resource combination schemes, resource occupancy characteristics, execution path characteristics, and adaptation scenario suitability characteristics are extracted as core evaluation dimensions. The comprehensive evaluation value is obtained by weighting the above three core evaluation dimensions by assigning preset weights to each of them, with the sum of the weights of each dimension being 1. The weight values ​​are preset according to the task type. Among them, the resource occupancy characteristics are determined based on the matching relationship between the resource allocation load and the resource rated load limit; the execution path characteristics are determined based on the task timeliness requirements and the instruction transmission and resource interaction latency; and the adaptation scenario suitability characteristics are determined based on the resource scenario matching characteristics and the overall resource suitability. After sorting all resource combination schemes from largest to smallest comprehensive evaluation value, a scheme list including scheme identifier, resource combination composition, and core adaptation description is provided to the user. After receiving the user's selected instruction, the system performs real-time status consistency verification on the resources corresponding to the selected scheme, and verifies the deviation rate between the current resource load redundancy capability and the resource load redundancy capability planned by S2. If the deviation rate does not exceed the preset threshold, the instruction is deemed valid. Then, a collaborative control instruction containing resource identifier, load allocation, execution sequence, interaction rules and status reporting requirements is generated and sent to the corresponding resources in order of resource comprehensive adaptability from high to low. Collect task execution progress, resource load changes, and feature matching degree data in real time according to the same collection cycle as S1. Calculate the deviation between the actual and planned values ​​of each dimension parameter based on the collected data. Integrate the collected raw data and the calculated deviation information to construct a structured execution status feedback set that includes task identifier, collection timestamp, real-time running data of each resource, and deviation values ​​of each parameter.

[0011] Furthermore, S4 includes the following: Retrieve the execution status feedback set constructed by S3 and the resource basic parameter set output by S1. Use the planned feature value and planned load value as the benchmark value and the real-time feature value and real-time load value as the actual value to perform multi-dimensional consistency verification on the running status of each resource. The verification dimensions include at least functional adaptation, load redundancy, scenario matching and resource load dimensions. Calculate the deviation rate and combine it with the preset threshold to determine anomalies. If there are no anomalies, continue monitoring. If a deviation or abnormal load is detected, the dynamically updated resource availability list of S1 is retrieved and the adaptation correlation analysis logic of S2 is activated. After the abnormal resource is processed, an adaptation scheme for resource switching or task execution path adjustment is regenerated according to the type of abnormality, and a comprehensive evaluation value is calculated. The evaluation value is sorted and then fed back to the user. After receiving confirmation from the user, the system issues an adjustment execution command and monitors the execution process. If no confirmation is received, the system continues to monitor. After the adjustment is completed, the system collects the actual operating data of the resources and sends it back to S1. The system then updates the resource basic parameter set and the dynamic resource availability list in full, with the updated content corresponding one-to-one with the resource basic parameter set in S1.

[0012] A remote control and interconnection management system based on artificial intelligence includes: a resource tagging and parameter set construction module, a task adaptation and resource combination scheme generation module, an instruction issuance and execution status feedback module, and a status verification and dynamic adjustment module; The resource tagging and parameter set construction module traverses the devices and edge nodes in the system in real time, extracts multi-dimensional features for tagging, constructs a dynamically updated list of available resources, and outputs a resource basic parameter set containing the tag features and operating status of each resource. The task adaptation and resource combination scheme generation module receives user-controlled tasks and parses core requirement parameters. It calls the tag features of the resource basic parameter set to perform adaptation and correlation analysis, and generates multiple resource combination schemes that associate multi-dimensional tag features, resource collaboration logic and task execution paths. The instruction issuance and execution status feedback module provides users with relevant explanations of each resource combination scheme, receives user-selected instructions and issues collaborative control instructions to the corresponding resources; it also collects task execution and resource operation-related data in real time to form an execution status feedback set containing parameter deviation information. The status verification and dynamic adjustment module verifies the consistency between the resource running status and the resource basic parameter set based on the execution status feedback set. When a deviation or abnormal load is detected, it calls the updated resource availability list and adaptation correlation analysis logic to regenerate the adaptation scheme and feeds it back to the user. After the user confirms, the adjustment is executed, and the adjustment result is sent back to the resource tagging and parameter set construction module to complete the update of resource tag features and resource basic parameter set.

[0013] Furthermore, the resource tagging and parameter set construction module includes a data acquisition unit and a tagging and inventory construction unit; The data acquisition unit traverses the devices and edge nodes within the system according to a preset acquisition cycle, collecting the raw operating data of each resource. The raw operating data includes the set of functions supported by the device, real-time load value, rated load upper limit value, current scene identifier and historical task execution records, and performs multi-dimensional feature quantification processing on the raw operating data. The tagging and inventory building unit divides the adaptation level based on the feature quantification results, generates multi-dimensional level tags, filters available resources that meet the preset availability threshold and sorts them, builds a dynamic resource availability inventory, extracts relevant data of available resources, and summarizes and outputs a set of basic resource parameters.

[0014] Furthermore, the task adaptation and resource combination scheme generation module includes a task parsing and association unit and a resource combination scheme construction unit; The task parsing and association unit receives remote control tasks from users, extracts the core requirement parameters of the task through semantic parsing, calls the basic parameter set of resources, establishes an association mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantified values ​​of resources, and initially screens and forms a set of candidate resources based on the mapping table, and calculates the comprehensive suitability of each candidate resource. The resource combination scheme construction unit selects the final candidate resources after the comprehensive adaptability meets the standard, and generates single resource schemes or multi-resource collaborative combination schemes through combination and adaptability analysis. It calculates the overall comprehensive adaptability of each scheme and completes the construction of multiple resource combination schemes.

[0015] Furthermore, the instruction issuance and execution status feedback module includes a scheme feedback and instruction issuance unit and a data acquisition and feedback set construction unit; The scheme feedback and instruction issuance unit comprehensively evaluates and sorts the generated resource combination schemes, provides the scheme list to the user, and after receiving the user's selected instruction, performs real-time status consistency verification on the resources corresponding to the selected scheme. After the verification is passed, it generates and issues collaborative control instructions. The data acquisition and feedback set construction unit collects task execution progress, resource load changes, and feature matching degree data in real time according to a preset acquisition cycle, calculates the deviation between the actual value and the planned value of each dimension parameter, integrates the original data and deviation information, and constructs a structured execution status feedback set.

[0016] Furthermore, the status verification and dynamic adjustment module includes a status verification and anomaly handling unit and a scheme adjustment and parameter update unit; The status verification and exception handling unit retrieves the execution status feedback set and resource basic parameter set, performs multi-dimensional consistency verification on the running status of each resource, calculates the deviation rate and combines it with the preset threshold to determine the exception. When an exception is detected, it processes the resource availability list, activates the adaptation association analysis logic, regenerates the resource switching or path adjustment adaptation scheme, and sorts and feeds it back. The scheme adjustment and parameter update unit receives the user's confirmation instruction for the adaptation scheme, issues the adjustment execution instruction and monitors the execution process. After the execution is completed, it collects the actual operation data of the resources and sends it back, and fully updates the basic parameter set of resources and the dynamic resource availability list to ensure that the resource data is consistent with the actual operation status.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention breaks through the traditional task-driven forward logic, performing multi-dimensional feature tagging processing on devices and edge nodes within the system to construct a dynamically updated resource availability list. It accurately identifies the task-sharing capabilities of low-load devices and generates reasonable resource combination schemes through multi-dimensional adaptation and correlation analysis, fully utilizing low-load devices, avoiding resource idleness and waste, and significantly improving system resource utilization. Secondly, by optimizing resource combination design and real-time status verification and dynamic adjustment mechanisms, it proactively avoids scenarios where multiple tasks compete for the same device, eliminating the need for complex conflict handling processes, effectively reducing system response latency, preventing non-substantive conflict misjudgments and forced allocation that does not meet expectations, and improving system stability and response efficiency. Furthermore, it provides users with feedback on optimized resource combination schemes, giving them full choice of schemes and flexibly adapting to user habits. When a target device is occupied, a new adaptation scheme is generated to replace task queuing, further improving task execution efficiency and user experience. Finally, the resource tag update, solution adaptation, status monitoring and dynamic adjustment are organically combined to form a complete closed loop, which effectively solves the core pain points of existing technologies, adapts to remote control needs in multiple scenarios, and has strong practicality and scalability. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a remote control and interconnection management system based on artificial intelligence according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 The present invention provides the following technical solution: A remote control and interconnection management system based on artificial intelligence includes: a resource tagging and parameter set construction module, a task adaptation and resource combination scheme generation module, an instruction issuance and execution status feedback module, and a status verification and dynamic adjustment module; The resource tagging and parameter set construction module traverses the devices and edge nodes in the system in real time, extracts multi-dimensional features for tagging, constructs a dynamically updated list of available resources, and outputs a resource basic parameter set containing the tag features and operating status of each resource. The task adaptation and resource combination scheme generation module receives user-controlled tasks and parses core requirement parameters. It calls the tag features of the resource basic parameter set to perform adaptation and correlation analysis, and generates multiple resource combination schemes that associate multi-dimensional tag features, resource collaboration logic and task execution paths. The instruction issuance and execution status feedback module provides users with relevant explanations of each resource combination scheme, receives user-selected instructions and issues collaborative control instructions to the corresponding resources; it also collects task execution and resource operation-related data in real time to form an execution status feedback set containing parameter deviation information. The status verification and dynamic adjustment module verifies the consistency between the resource running status and the resource basic parameter set based on the execution status feedback set. When a deviation or abnormal load is detected, it calls the updated resource availability list and adaptation correlation analysis logic to regenerate the adaptation scheme and feeds it back to the user. After the user confirms, the adjustment is executed, and the adjustment result is sent back to the resource tagging and parameter set construction module to complete the update of resource tag features and resource basic parameter set.

[0021] The resource tagging and parameter set construction module includes a data acquisition unit and a tagging and list construction unit; The data acquisition unit traverses the devices and edge nodes within the system according to a preset acquisition cycle, collecting the raw operating data of each resource. The raw operating data includes the set of functions supported by the device, real-time load value, rated load upper limit value, current scene identifier and historical task execution records, and performs multi-dimensional feature quantification processing on the raw operating data. The tagging and inventory building unit divides the adaptation level based on the feature quantification results, generates multi-dimensional level tags, filters available resources that meet the preset availability threshold and sorts them, builds a dynamic resource availability inventory, extracts relevant data of available resources, and summarizes and outputs a set of basic resource parameters.

[0022] The task adaptation and resource combination scheme generation module includes a task parsing and association unit and a resource combination scheme construction unit; The task parsing and association unit receives remote control tasks from users, extracts the core requirement parameters of the task through semantic parsing, calls the basic parameter set of resources, establishes an association mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantified values ​​of resources, and initially screens and forms a set of candidate resources based on the mapping table, and calculates the comprehensive suitability of each candidate resource. The resource combination scheme construction unit selects the final candidate resources after the comprehensive adaptability meets the standard, and generates single resource schemes or multi-resource collaborative combination schemes through combination and adaptability analysis. It calculates the overall comprehensive adaptability of each scheme and completes the construction of multiple resource combination schemes.

[0023] The instruction issuance and execution status feedback module includes a scheme feedback and instruction issuance unit and a data acquisition and feedback set construction unit; The scheme feedback and instruction issuance unit comprehensively evaluates and sorts the generated resource combination schemes, provides the scheme list to the user, and after receiving the user's selected instruction, performs real-time status consistency verification on the resources corresponding to the selected scheme. After the verification is passed, it generates and issues collaborative control instructions. The data acquisition and feedback set construction unit collects task execution progress, resource load changes, and feature matching degree data in real time according to a preset acquisition cycle, calculates the deviation between the actual value and the planned value of each dimension parameter, integrates the original data and deviation information, and constructs a structured execution status feedback set.

[0024] The status verification and dynamic adjustment module includes a status verification and anomaly handling unit and a scheme adjustment and parameter update unit; The status verification and exception handling unit retrieves the execution status feedback set and resource basic parameter set, performs multi-dimensional consistency verification on the running status of each resource, calculates the deviation rate and combines it with the preset threshold to determine the exception. When an exception is detected, it processes the resource availability list, activates the adaptation association analysis logic, regenerates the resource switching or path adjustment adaptation scheme, and sorts and feeds it back. The scheme adjustment and parameter update unit receives the user's confirmation instruction for the adaptation scheme, issues the adjustment execution instruction and monitors the execution process. After the execution is completed, it collects the actual operation data of the resources and sends it back, and fully updates the basic parameter set of resources and the dynamic resource availability list to ensure that the resource data is consistent with the actual operation status.

[0025] A remote control and interconnection management method based on artificial intelligence includes the following steps: S1. Real-time traversal of devices and edge nodes within the system, extraction of functional adaptation features, load redundancy features, and scene matching features for multi-dimensional labeling, construction of a dynamically updated list of available resources, and output of a set of basic resource parameters containing the label features and operating status of each resource. S2. Receive user-controlled tasks and parse the core requirements parameters of the tasks, call the resource tag features of the resource basic parameter set, perform adaptation and correlation analysis between the resource tag features and the core requirements parameters of the tasks, generate different resource combination schemes, and associate the schemes with multi-dimensional tag features, including resource collaboration logic and task execution path. S3. Provide feedback to the user on the resource usage, execution path, and applicable scenario descriptions of each solution. After receiving the user's selected instruction, issue collaborative control instructions to the corresponding resources. Collect task execution progress, resource load changes, and feature matching data in real time to form an execution status feedback set containing parameter deviation information. S4. Based on the execution status feedback set, verify the consistency between the resource running status and the resource basic parameter set in real time. If a deviation or abnormal load is detected, call the updated resource availability list and adaptation correlation analysis logic, regenerate the adaptation scheme for resource switching or task execution path adjustment and feed it back to the user. After the user confirms, the adjustment is executed and the adjustment result is sent back to S1 to update the corresponding resource's tag features and resource basic parameter set.

[0026] Furthermore, S1 includes the following: The system traverses devices and edge nodes according to a preset collection cycle, collecting raw operational data for each resource. This raw operational data includes the set of functions supported by the device, real-time load values, the system's preset rated load limit, the current scene identifier determined by collectable environmental parameters, and historical task execution records. Multi-dimensional feature quantification is performed on the raw operational data. Specifically, the function adaptation feature value is calculated based on the degree of matching between the resource's supported functions and a predefined set of system standard functions, with function item weights used for correction. The load redundancy feature value is calculated based on the ratio between the rated load limit and the real-time load value. The scene matching feature value is calculated based on the success rate of tasks matching the current scene identifier in historical task execution records, with time decay factors and scene similarity factors used for correction. Adaptation levels are defined based on the distribution range of various feature values, forming corresponding multi-dimensional level labels. These labels are associated with feature values ​​and basic resource attributes. Resources that meet the preset availability thresholds for all three level labels are selected and included in the available resource range. Available resources are sorted from high to low according to their functional adaptation level, and within the same level, they are sorted from high to low according to their load redundancy feature values, constructing a dynamically updated list of available resources. The multi-dimensional level labels, real-time load rate, rated load limit, and scenario attribute data of each available resource are extracted, where the real-time load rate is the ratio of the current real-time load value to the rated load limit. These are then summarized to form a set of basic resource parameters containing the feature values ​​and operating status of each resource label and output.

[0027] In this embodiment, the collected raw operational data undergoes feature quantization processing to obtain feature values ​​in three dimensions: functional adaptation, load redundancy, and scene matching; specifically including: Calculate the function adaptation feature value F, and F=(k1×m) / M, where m is the number of intersections between the set of function items supported by the resource and the standard task function library, M is the total number of all function items in the system's predefined standard task function library, which covers the functions required for all typical tasks that the system planning and design can handle; k1 is the function matching correction coefficient, and k1=∑(the preset weights of the m function items supported by the resource) / ∑(the total weights of all M function items in the standard task function library), where the function item weights are preset according to the task's criticality and stored in the system configuration; Calculate the load redundancy characteristic value L, and L=(Pn-Pr) / Pn, where Pn is the rated load limit value preset by the system for this resource, and Pr is the current real-time load value of the resource collected; Pr and Pn have the same unit and jointly represent the absolute amount of resource load; the value range of L is [0,1], and the larger the value, the stronger the remaining available load capacity. Calculate the scene matching feature value S, and S=(k2×t) / T, where t is the number of records in the historical task record set of this resource where the task scene code is the same as the current scene code and the execution result is successful; T is the total number of tasks executed in the history of this resource; k2 is the scene adaptation correction coefficient, and k2=α×β, where α is the time decay factor, α=e^(-λ×Δt), Δt is the time interval between the current time and the most recent successful execution of the same scene code task, λ is a preset decay constant; β is the scene similarity factor, which is obtained by calculating the average cosine similarity between the feature vector of the current scene code and the feature vector of the historical successful task scene code. Based on the aforementioned feature values, multi-dimensional labeling is performed. Each dimension label is divided into adaptation levels according to the distribution range of the corresponding feature value, forming functional adaptation level labels, load redundancy level labels, and scenario matching level labels. The label information is directly associated with the corresponding feature value and basic resource attributes. For example, the F value range [0.8, 1.0] is set as "high adaptation", [0.5, 0.8) as "medium adaptation", and [0, 0.5) as "low adaptation"; the L value range [0.7, 1.0] is set as "high redundancy", [0.3, 0.7) as "medium redundancy", and [0, 0.3) as "low redundancy"; the S value range [0.8, 1.0] is set as "high matching", [0.5, 0.8) as "medium matching", and [0, 0.5) as "low matching". Resources whose functional adaptation level, load redundancy level, and scene matching level all reach or exceed the preset minimum availability level are classified as medium adaptation, medium redundancy, and medium matching, and are included in the scope of available resources. For available resources, they are sorted from high to low by functional adaptation level, and within the same level, they are sorted from high to low by load redundancy feature value L, to construct a dynamically updated list of available resources. The list update cycle is consistent with the data collection cycle. Multi-dimensional level tags, real-time load rate, rated load limit, and scene attribute data of each available resource are extracted. Among them, the real-time load rate is the ratio of the current real-time load value to the rated load limit value. These are summarized to form a set of basic resource parameters containing the tag characteristics and operating status of each resource and output.

[0028] Furthermore, S2 includes the following: The system receives remote control tasks issued by users and extracts the core requirements parameters of the task corresponding to the multi-dimensional feature dimensions of S1 through semantic parsing. The parameters include the functional requirements of the task, the workload requirements of the task, the task execution scenario identifier, and the task execution time requirements. Call the resource basic parameter set output by S1, extract the multi-dimensional level labels of each available resource and the corresponding function adaptation, load redundancy, and scenario matching feature values, and establish a mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantified values ​​of the resources. Based on the association mapping table, the available resources are initially screened to select those with functional adaptation and scenario matching level tags that are not lower than the lowest level corresponding to the task, and whose current remaining load capacity meets the task load requirements, thus forming a set of candidate resources. The overall adaptability of each resource in the candidate resource set to the current task is calculated. This overall adaptability is obtained by multiplying the feature values ​​of functional adaptability, load redundancy, and scenario matching by their respective task adaptability weights, summing the results, and then multiplying by the task correction factor. Among these, the adaptability weights of the three types of features are positively correlated with the task priority and sum to 1. All weights can be fine-tuned according to the task type and stored in the system configuration library. The task correction factor is obtained by multiplying the task timeliness correction factor and the priority correction factor. The timeliness correction factor is calculated based on the task's preset maximum execution time and estimated execution time, while the priority correction factor is set according to the task priority levels. From the candidate resource set, resources with a comprehensive suitability not lower than a preset suitability threshold are further selected as final candidate resources. Combination suitability analysis is then performed on the final candidate resources to generate different resource combination schemes: the load redundancy capability of a single final candidate resource is calculated, which is the product of the resource load redundancy feature quantification value and the rated load upper limit value. If the load redundancy capability of a single resource meets the task load requirements, the resource is combined as an independent resource to form a single resource scheme, and its overall comprehensive suitability is consistent with the resource's own comprehensive suitability. If the load redundancy capability of a single resource cannot meet the requirements, multiple resources are selected sequentially from high to low comprehensive suitability for collaborative combination until the total load redundancy capability after combination meets the task requirements. Then, the overall comprehensive suitability of the combination is calculated, which is the sum of the products of the comprehensive suitability of each resource in the combination and the corresponding load redundancy feature quantification value, divided by the sum of the load redundancy feature quantification values ​​of each resource in the combination.

[0029] In this embodiment, a remote control task issued by the user is received, and the core requirement parameters of the task are extracted through semantic parsing. The core requirement parameters of the task correspond to the multi-dimensional feature dimensions in S1, including the functional requirements of the task, the workload requirements of the task, the task execution scenario identifier, and the task execution time requirements. Call the resource basic parameter set output by S1, extract the multi-dimensional level labels of each available resource and the corresponding function adaptation feature value, load redundancy feature value, and scenario matching feature value, and establish a mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantification values ​​of the resources; Based on the association mapping table, available resources are initially screened. The screening criteria are: the functional adaptation level label of the resource is not lower than the minimum functional adaptation level corresponding to the task requirements, the scenario matching level label is not lower than the minimum scenario matching level corresponding to the task requirements, and the current remaining load capacity of the resource is not lower than the load requirement required for task execution. After the initial screening, a set of candidate resources that meet the basic requirements of the task is obtained. For each resource in the candidate resource set, calculate its comprehensive adaptability A with the current user control task, and A = (wF×F + wL×L + wS×S)×Kt, where wF, wL, and wS are the task adaptability weights for functional adaptation, load redundancy, and scenario matching features, respectively. The weight allocation is positively correlated with the task priority level and satisfies wF + wL + wS = 1. For example, specific calibration rules are as follows: high-priority tasks prioritize execution stability, with wL taking a value of 0.4-0.5, wF taking a value of 0.3-0.4, and wS taking a value of 0.1-0.2; medium-priority tasks have a balanced distribution of the three, with wF, wL, and wS all taking values ​​of 0.3-0.35; low-priority tasks prioritize functional adaptation, with wF taking a value of... 0.4-0.5, wL takes values ​​of 0.2-0.3, and wS takes values ​​of 0.2-0.3; all weights can be fine-tuned according to the task type and stored in the system configuration library; Kt is the task correction factor, which comprehensively considers the task execution time requirements and task priority, and is calculated by the formula Kt=μ×γ, where μ is the task timeliness correction factor, μ=1+0.1×(t0-t1) / t0, t0 is the preset maximum execution time of the task, and t1 is the estimated execution time of the task; γ is the task priority correction factor, for example, high priority tasks γ=1.2, medium priority tasks γ=1.0, and low priority tasks γ=0.9, which is used to distinguish the resource adaptation priority of different priority tasks to ensure that high priority tasks are matched with better resources; The initial set of candidate resources is further screened, and resources with a comprehensive suitability not lower than a preset suitability threshold A0 are selected as final candidate resources. The preset suitability threshold A0 is determined based on the priority of the core task requirements; the higher the task priority, the higher the value. Combination suitability analysis is then performed on the candidate resources to generate different resource combination schemes, specifically including: Calculate the load redundancy capacity R of a single candidate resource, and R = L × U, where U is the rated load upper limit of a single candidate resource, extracted from the resource basic parameter set output by S1; if the load redundancy capacity R of a single candidate resource is not lower than the task load requirement Q in the core requirement parameters of the task, then the candidate resource is used as an independent resource combination to form a single resource scheme, and the overall comprehensive adaptability Ac of the scheme is equal to the comprehensive adaptability A of the candidate resource. If the load redundancy capacity R of a single candidate resource is lower than the task load requirement Q, then multiple candidate resources are selected for collaborative combination. During the combination process, candidate resources are selected in descending order of their overall suitability until the sum of the load redundancy capacities of all candidate resources after combination is not lower than the task load requirement Q. After combination, the overall overall suitability Ac of the resource combination is calculated, and Ac = ∑ i∈[1,n] (Ai×Li) / ∑ i∈[1,n] Li, where Ai is the overall fit of the i-th candidate resource in the combination, Li is the load redundancy characteristic quantification value of the i-th candidate resource in the combination, and n is the total number of collaborative combination candidate resources selected to satisfy the task load requirement Q.

[0030] Furthermore, S3 includes the following: For the generated resource combination schemes, resource occupancy characteristics, execution path characteristics, and adaptation scenario suitability characteristics are extracted as core evaluation dimensions. The comprehensive evaluation value is obtained by weighting the above three core evaluation dimensions by assigning preset weights to each of them, with the sum of the weights of each dimension being 1. The weight values ​​are preset according to the task type. Among them, the resource occupancy characteristics are determined based on the matching relationship between the resource allocation load and the resource rated load limit; the execution path characteristics are determined based on the task timeliness requirements and the instruction transmission and resource interaction latency; and the adaptation scenario suitability characteristics are determined based on the resource scenario matching characteristics and the overall resource suitability. After sorting all resource combination schemes from largest to smallest comprehensive evaluation value, a scheme list including scheme identifier, resource combination composition, and core adaptation description is provided to the user. After receiving the user's selected instruction, the system performs real-time status consistency verification on the resources corresponding to the selected scheme, and verifies the deviation rate between the current resource load redundancy capability and the resource load redundancy capability planned by S2. If the deviation rate does not exceed the preset threshold, the instruction is deemed valid. Then, a collaborative control instruction containing resource identifier, load allocation, execution sequence, interaction rules and status reporting requirements is generated and sent to the corresponding resources in order of resource comprehensive adaptability from high to low. Collect task execution progress, resource load changes, and feature matching degree data in real time according to the same collection cycle as S1. Calculate the deviation between the actual and planned values ​​of each dimension parameter based on the collected data. Integrate the collected raw data and the calculated deviation information to construct a structured execution status feedback set that includes task identifier, collection timestamp, real-time running data of each resource, and deviation values ​​of each parameter.

[0031] In this embodiment, for the generated resource combination scheme, the resource consumption feature value O, the execution path feature value E, and the adaptation scenario fit feature value Sg are extracted as the core dimensions for comprehensive evaluation; wherein, the calculation formula for the resource consumption feature value O is: O=1-(∑ i∈[1,n] Qi / ∑ i∈[1,n]Ui), Qi is the load allocated to the i-th resource, and Ui is the rated load limit of the i-th resource; for a single resource scheme, O = 1 - (Qi / Ui); the formula for calculating the execution path characteristic value E is: E = t0 / (Tr + ∑ i∈[1,n] Ti+t1), t0 is the preset maximum execution time of the task, Tr is the instruction transmission latency, Ti is the latency of the i-th resource collaboration interaction, and t1 is the estimated execution time of the task; the formula for calculating the adaptation scenario fit feature value Sg is: Sg=∑ i∈[1,n] (Si×Ai) / ∑ i∈[1,n] Ai and Si are the matching feature values ​​of the i-th resource scenario, and Ai is the comprehensive adaptability of the i-th resource; for a single resource solution, Sg = S. Calculate the comprehensive evaluation value K for each resource combination scheme, where K = a1 × O + a2 × En + a3 × Sg, where a1, a2, and a3 are the weights of each dimension, satisfying a1 + a2 + a3 = 1, and the specific values ​​are preset according to the task type; En is the normalized result of the execution path feature value E; sort all resource combination schemes in descending order of comprehensive evaluation value K, and provide the user with a list of sorted schemes, which includes the scheme identifier, comprehensive evaluation value K, resource combination composition, and core adaptation description; The system receives the user's resource combination scheme selection instruction, performs real-time status consistency verification on the resources corresponding to the selected scheme, and verifies the deviation between the current resource load redundancy capacity Rc and the load redundancy capacity Rp planned in the scheme in S2, with the deviation rate ΔR=|Rc-Rp| / Rp, where the current load redundancy capacity Rc=Lr×U and the planned load redundancy capacity Rp=Lp×U; if ΔR is less than or equal to the preset deviation threshold ΔR0, the instruction is deemed valid; otherwise, the scheme is reported as unavailable and a suboptimal scheme is recommended; for valid selection instructions, a collaborative control instruction is generated, which includes resource identifier, load allocation, execution sequence, interaction rules, and status reporting requirements; the collaborative control instruction is accurately issued to the corresponding resources in the selected scheme in descending order of resource comprehensive adaptability to ensure the effectiveness of instruction transmission and reception; Collect task execution progress, resource load changes, and feature matching data in real time, using the same collection cycle as S1. Task execution progress data includes the task's execution time (te), the number of completed sub-items (C), and the total number of sub-items (Cn), and calculate the progress rate (P). For decomposable tasks, P = C / Cn; for non-decomposable tasks, P = te / t1. Resource load change data consists of the real-time load value (Pr) for each resource, and calculate the real-time load redundancy feature value (Lr), where Lr = 1 - Pr / U. Feature matching data includes the real-time functional adaptation feature value (Fr) and the real-time scene matching feature value (Sr) for each resource, and the calculation rules are consistent with S1. Based on the collected data, parameter deviation information is calculated. The deviations are all the differences between the actual values ​​and the planned values. For example: schedule deviation ΔP = P - Pt, where Pt is the current preset target schedule rate of the task; load redundancy deviation ΔL = Lr - Lp, where Lp is the resource load redundancy feature value of S2 planning; feature matching degree deviation ΔF = Fr - Fp, ΔS = Sr - Sp, where Fp and Sp are the function and scenario matching feature values ​​of S1 and S2 planning. By integrating the collected raw data with the calculated parameter deviation information, a structured execution status feedback set is constructed. The feedback set should include at least the task identifier, collection timestamp, real-time running data of each resource, and deviation values ​​of each parameter.

[0032] Furthermore, S4 includes the following: Retrieve the execution status feedback set constructed by S3 and the resource basic parameter set output by S1. Use the planned feature value and planned load value as the benchmark value and the real-time feature value and real-time load value as the actual value to perform multi-dimensional consistency verification on the running status of each resource. The verification dimensions include at least functional adaptation, load redundancy, scenario matching and resource load dimensions. Calculate the deviation rate and combine it with the preset threshold to determine anomalies. If there are no anomalies, continue monitoring. If a deviation or abnormal load is detected, the dynamically updated resource availability list of S1 is retrieved and the adaptation correlation analysis logic of S2 is activated. After the abnormal resource is processed, an adaptation scheme for resource switching or task execution path adjustment is regenerated according to the type of abnormality, and a comprehensive evaluation value is calculated. The evaluation value is sorted and then fed back to the user. After receiving confirmation from the user, the system issues an adjustment execution command and monitors the execution process. If no confirmation is received, the system continues to monitor. After the adjustment is completed, the system collects the actual operating data of the resources and sends it back to S1. The system then updates the resource basic parameter set and the dynamic resource availability list in full, with the updated content corresponding one-to-one with the resource basic parameter set in S1.

[0033] In this embodiment, consistency verification is performed on the operational status dimension of each resource. The verification dimension includes at least functional adaptation features, load redundancy features, scenario matching features, and resource load values. Calculate the deviation rate V for each verification dimension, and V = |Xs - Xj| / Xj, where Xs is the actual value of the parameter in the execution status feedback set; Xj is the baseline value of the parameter in the resource basic parameter set; preset deviation threshold V0 for each dimension, and preset load anomaly judgment threshold Ly, where the load anomaly judgment threshold is a preset proportion of the upper limit of the rated load of the resource; if the deviation rate V ≥ V0 for any dimension, it is determined that the resource operation status deviates from the baseline value; if the real-time load value of the resource ≥ Ly, it is determined that the resource load is abnormal; if the deviation rate V < V0 for all dimensions and the real-time load value of the resource < Ly, it is determined that the resource operation status is normal, and the continuous monitoring execution status feedback set is returned. If a deviation in resource operation status or an abnormal load is detected, the dynamically updated resource availability list in S1 is retrieved in real time, and the pre-set resource and task adaptation correlation analysis logic in S2 is activated simultaneously. This logic includes preliminary resource screening rules, comprehensive adaptation calculation rules, and resource combination scheme generation rules. Based on the execution status feedback set, the identifier, abnormal dimension, deviation rate, or actual load value of abnormal resources are extracted. Unrecoverable abnormal resources are removed from the resource availability list. Unrecoverable abnormal resources refer to resources whose load is continuously ≥ Ly and has no remaining redundancy, and whose feature deviation rate is continuously ≥ V0 and cannot be corrected. Resources that are temporarily abnormal but recoverable are marked with an abnormal identifier, retained in the resource availability list, and their comprehensive adaptation weight is reduced. If the anomaly type is a single resource load anomaly or a core feature deviation exceeding the standard, where a core feature deviation exceeding the standard refers to a deviation rate V ≥ V0 in any dimension of functional adaptation feature, load redundancy feature, or scenario matching feature; based on the activated adaptation correlation analysis logic, using the core requirement parameters of the user control task in S3 as a benchmark, re-select candidate resources that meet the basic requirements of the task from the updated resource availability list, with the selection rules being consistent with S2; referring to the calculation formula in S2, calculate the comprehensive adaptation degree A' of the new candidate resources; based on the comprehensive adaptation degree A' of the new candidate resources, regenerate a single resource switching scheme or a multi-resource collaborative switching scheme according to the resource combination rules in S2, specifying the new resource identifier, resource combination composition, load allocation ratio, and switching execution node in the scheme; If the anomaly type is excessive latency in multi-resource collaborative interaction or excessive deviation in execution path characteristics, retain the original resource combination and replan the task execution path based on the updated resource availability list and the real-time running status of each resource. Referring to the calculation formula in S3, calculate the path efficiency E' of the new execution path and generate a task execution path adjustment plan. The plan should clearly define the topology of the new execution path, the execution sequence of each resource, the interaction nodes, and the path efficiency E'. For all regenerated resource switching or path adjustment adaptation plans, calculate the comprehensive evaluation value K' of the plan according to the comprehensive evaluation rules in S3. Sort all regenerated resource switching or path adjustment adaptation schemes in descending order of comprehensive evaluation value K' and provide the user with a sorted list of schemes; receive the user's scheme selection confirmation instruction; if the user selects a certain adaptation scheme, generate an adjustment execution instruction based on that scheme; accurately issue operation instructions to the corresponding resources according to the requirements of the adjustment execution instruction, and monitor the adjustment execution process in real time; if the user does not confirm any scheme, return to continuously monitoring the resource operation status until the user issues a new instruction; After the resource switching or path adjustment is completed, the actual operating data of each resource after the adjustment is collected in real time as the core data of the adjustment result. The core data of the adjustment result is sent back to S1. Based on this data, the resource basic parameter set is fully updated. The updated content corresponds one-to-one with the composition of the resource basic parameter set in S1. The dynamic resource availability list in S1 is updated synchronously.

[0034] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0035] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A remote control and interconnection management method based on artificial intelligence, characterized in that: The method includes the following steps: S1. Real-time traversal of devices and edge nodes within the system, extraction of functional adaptation features, load redundancy features, and scene matching features for multi-dimensional labeling, construction of a dynamically updated list of available resources, and output of a set of basic resource parameters containing the label features and operating status of each resource. S2. Receive user-controlled tasks and parse the core requirements parameters of the tasks, call the resource tag features of the resource basic parameter set, perform adaptation and correlation analysis between the resource tag features and the core requirements parameters of the tasks, generate different resource combination schemes, and associate the schemes with multi-dimensional tag features, including resource collaboration logic and task execution path. S3. Provide feedback to the user on the resource usage, execution path, and applicable scenario descriptions of each solution. After receiving the user's selected instruction, issue collaborative control instructions to the corresponding resources. Collect task execution progress, resource load changes, and feature matching degree data in real time to form an execution status feedback set containing parameter deviation information. S4. Based on the execution status feedback set, verify the consistency between the resource running status and the resource basic parameter set in real time. If a deviation or abnormal load is detected, call the updated resource availability list and adaptation correlation analysis logic, regenerate the adaptation scheme for resource switching or task execution path adjustment and feed it back to the user. After the user confirms, the adjustment is executed and the adjustment result is sent back to S1 to update the corresponding resource's tag features and resource basic parameter set.

2. The remote control and interconnection management method based on artificial intelligence according to claim 1, characterized in that: S1 includes the following: The system traverses devices and edge nodes according to a preset collection cycle, collecting raw operational data for each resource. This raw operational data includes the set of functions supported by the device, real-time load values, the system's preset rated load limit, the current scene identifier determined by collectable environmental parameters, and historical task execution records. Multi-dimensional feature quantification is performed on the raw operational data. Specifically, the function adaptation feature value is calculated based on the degree of matching between the resource's supported functions and a predefined set of system standard functions, with function item weights used for correction. The load redundancy feature value is calculated based on the ratio between the rated load limit and the real-time load value. The scene matching feature value is calculated based on the success rate of tasks matching the current scene identifier in historical task execution records, with time decay factors and scene similarity factors used for correction. Adaptation levels are defined based on the distribution range of various feature values, forming corresponding multi-dimensional level labels. These labels are associated with feature values ​​and basic resource attributes. Resources that meet the preset availability thresholds for all three level labels are selected and included in the available resource range. Available resources are sorted from high to low according to their functional adaptation level, and within the same level, they are sorted from high to low according to their load redundancy feature values, constructing a dynamically updated list of available resources. The multi-dimensional level labels, real-time load rate, rated load limit, and scenario attribute data of each available resource are extracted, where the real-time load rate is the ratio of the current real-time load value to the rated load limit. These are then summarized to form a set of basic resource parameters containing the feature values ​​and operating status of each resource label and output.

3. The remote control and interconnection management method based on artificial intelligence according to claim 2, characterized in that: S2 includes the following: The system receives remote control tasks issued by users and extracts the core requirements parameters of the task corresponding to the multi-dimensional feature dimensions of S1 through semantic parsing. The parameters include the functional requirements of the task, the workload requirements of the task, the task execution scenario identifier, and the task execution time requirements. Call the resource basic parameter set output by S1, extract the multi-dimensional level labels of each available resource and the corresponding function adaptation, load redundancy, and scenario matching feature values, and establish a mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantified values ​​of the resources. Based on the association mapping table, the available resources are initially screened to select those with functional adaptation and scenario matching level tags that are not lower than the lowest level corresponding to the task, and whose current remaining load capacity meets the task load requirements, thus forming a set of candidate resources. The overall adaptability of each resource in the candidate resource set to the current task is calculated. This overall adaptability is obtained by multiplying the feature values ​​of functional adaptability, load redundancy, and scenario matching by their respective task adaptability weights, summing the results, and then multiplying by the task correction factor. Among these, the adaptability weights of the three types of features are positively correlated with the task priority and sum to 1. All weights can be fine-tuned according to the task type and stored in the system configuration library. The task correction factor is obtained by multiplying the task timeliness correction factor and the priority correction factor. The timeliness correction factor is calculated based on the task's preset maximum execution time and estimated execution time, while the priority correction factor is set according to the task priority levels. From the candidate resource set, resources with a comprehensive suitability not lower than a preset suitability threshold are further selected as final candidate resources. Combination suitability analysis is then performed on the final candidate resources to generate different resource combination schemes: the load redundancy capability of a single final candidate resource is calculated, which is the product of the resource load redundancy feature quantification value and the rated load upper limit value. If the load redundancy capability of a single resource meets the task load requirements, the resource is combined as an independent resource to form a single resource scheme, and its overall comprehensive suitability is consistent with the resource's own comprehensive suitability. If the load redundancy capability of a single resource cannot meet the requirements, multiple resources are selected sequentially from high to low comprehensive suitability for collaborative combination until the total load redundancy capability after combination meets the task requirements. Then, the overall comprehensive suitability of the combination is calculated, which is the sum of the products of the comprehensive suitability of each resource in the combination and the corresponding load redundancy feature quantification value, divided by the sum of the load redundancy feature quantification values ​​of each resource in the combination.

4. The remote control and interconnection management method based on artificial intelligence according to claim 3, characterized in that: S3 includes the following: For the generated resource combination schemes, resource occupancy characteristics, execution path characteristics, and adaptation scenario suitability characteristics are extracted as core evaluation dimensions. The comprehensive evaluation value is obtained by weighting the above three core evaluation dimensions by assigning preset weights to each of them, with the sum of the weights of each dimension being 1. The weight values ​​are preset according to the task type. Among them, the resource occupancy characteristics are determined based on the matching relationship between the resource allocation load and the resource rated load limit; the execution path characteristics are determined based on the task timeliness requirements and the instruction transmission and resource interaction latency; and the adaptation scenario suitability characteristics are determined based on the resource scenario matching characteristics and the overall resource suitability. After sorting all resource combination schemes from largest to smallest comprehensive evaluation value, a scheme list including scheme identifier, resource combination composition, and core adaptation description is provided to the user. After receiving the user's selected instruction, the system performs real-time status consistency verification on the resources corresponding to the selected scheme, and verifies the deviation rate between the current resource load redundancy capability and the resource load redundancy capability planned by S2. If the deviation rate does not exceed the preset threshold, the instruction is deemed valid. Then, a collaborative control instruction containing resource identifier, load allocation, execution sequence, interaction rules and status reporting requirements is generated and sent to the corresponding resources in order of resource comprehensive adaptability from high to low. Collect task execution progress, resource load changes, and feature matching degree data in real time according to the same collection cycle as S1. Calculate the deviation between the actual and planned values ​​of each dimension parameter based on the collected data. Integrate the collected raw data and the calculated deviation information to construct a structured execution status feedback set that includes task identifier, collection timestamp, real-time running data of each resource, and deviation values ​​of each parameter.

5. The remote control and interconnection management method based on artificial intelligence according to claim 4, characterized in that: S4 includes the following: Retrieve the execution status feedback set constructed by S3 and the resource basic parameter set output by S1. Use the planned feature value and planned load value as the benchmark value and the real-time feature value and real-time load value as the actual value to perform multi-dimensional consistency verification on the running status of each resource. The verification dimensions include at least functional adaptation, load redundancy, scenario matching and resource load dimensions. Calculate the deviation rate and combine it with the preset threshold to determine anomalies. If there are no anomalies, continue monitoring. If a deviation or abnormal load is detected, the dynamically updated resource availability list of S1 is retrieved and the adaptation correlation analysis logic of S2 is activated. After the abnormal resource is processed, an adaptation scheme for resource switching or task execution path adjustment is regenerated according to the type of abnormality, and a comprehensive evaluation value is calculated. The evaluation value is sorted and then fed back to the user. After receiving confirmation from the user, the system issues an adjustment execution command and monitors the execution process; if no confirmation is received, the system continues to monitor. After the adjustment is completed, the actual operation data of the collected resources is sent back to S1, and the resource basic parameter set and dynamic resource availability list are updated in full. The updated content corresponds one-to-one with the resource basic parameter set of S1.

6. An artificial intelligence-based remote control and interconnection management system, applied to the artificial intelligence-based remote control and interconnection management method according to any one of claims 1-5, characterized in that: The system includes: a resource tagging and parameter set construction module, a task adaptation and resource combination scheme generation module, an instruction issuance and execution status feedback module, and a status verification and dynamic adjustment module. The resource tagging and parameter set construction module traverses the devices and edge nodes in the system in real time, extracts multi-dimensional features for tagging, constructs a dynamically updated list of available resources, and outputs a resource basic parameter set containing the tag features and operating status of each resource. The task adaptation and resource combination scheme generation module receives user-controlled tasks and parses core requirement parameters. It then calls the tag features of the resource basic parameter set to perform adaptation and correlation analysis, generating multiple resource combination schemes that associate multi-dimensional tag features, resource collaboration logic, and task execution paths. The instruction issuance and execution status feedback module provides users with relevant explanations of each resource combination scheme, receives user-selected instructions and issues collaborative control instructions to the corresponding resources; it also collects task execution and resource operation-related data in real time to form an execution status feedback set containing parameter deviation information. The status verification and dynamic adjustment module verifies the consistency between the resource operation status and the resource basic parameter set based on the execution status feedback set. When a deviation or abnormal load is detected, it calls the updated resource availability list and adaptation correlation analysis logic to regenerate the adaptation scheme and feeds it back to the user. After the user confirms, the adjustment is executed, and the adjustment result is sent back to the resource tagging and parameter set construction module to complete the update of resource tag features and resource basic parameter set.

7. The remote control and interconnection management system based on artificial intelligence according to claim 6, characterized in that: The resource tagging and parameter set construction module includes a data acquisition unit and a tagging and list construction unit; The data acquisition unit traverses the devices and edge nodes within the system according to a preset acquisition cycle, and collects the raw operating data of each resource. The raw operating data includes the set of functions supported by the device, real-time load value, rated load upper limit value, current scene identifier and historical task execution record, and performs multi-dimensional feature quantification processing on the raw operating data. The tagging and list construction unit divides the adaptation level based on the feature quantification results, generates multi-dimensional level tags, filters available resources that meet the preset availability threshold and sorts them, constructs a dynamic resource availability list, and extracts relevant data of available resources, summarizes and generates and outputs a resource basic parameter set.

8. The remote control and interconnection management system based on artificial intelligence according to claim 6, characterized in that: The task adaptation and resource combination scheme generation module includes a task parsing and association unit and a resource combination scheme construction unit; The task parsing and association unit receives the user's remote control task, extracts the core requirement parameters of the task through semantic parsing, calls the resource basic parameter set, establishes an association mapping table between the core requirement parameters of the task and the multi-dimensional label features and quantitative values ​​of the resources, and preliminarily screens and forms a set of candidate resources based on the mapping table, and calculates the comprehensive suitability of each candidate resource. The resource combination scheme construction unit screens the final candidate resources after the comprehensive adaptability meets the standard, generates a single resource scheme or a multi-resource collaborative combination scheme through combination and adaptability analysis, calculates the overall comprehensive adaptability of each scheme, and completes the construction of multiple resource combination schemes.

9. The remote control and interconnection management system based on artificial intelligence according to claim 6, characterized in that: The instruction issuance and execution status feedback module includes a scheme feedback and instruction issuance unit and a data acquisition and feedback set construction unit; The scheme feedback and instruction issuance unit comprehensively evaluates and sorts the generated resource combination schemes, provides a scheme list to the user, and after receiving the user's selected instruction, performs real-time status consistency verification on the resources corresponding to the selected scheme. After the verification is passed, it generates and issues collaborative control instructions. The data acquisition and feedback set construction unit collects task execution progress, resource load changes, and feature matching degree data in real time according to a preset acquisition cycle, calculates the deviation between the actual value and the planned value of each dimension parameter, integrates the original data and deviation information, and constructs a structured execution status feedback set.

10. A remote control and interconnection management system based on artificial intelligence according to claim 6, characterized in that: The status verification and dynamic adjustment module includes a status verification and anomaly handling unit and a scheme adjustment and parameter update unit; The status verification and anomaly handling unit retrieves the execution status feedback set and resource basic parameter set, performs multi-dimensional consistency verification on the operating status of each resource, calculates the deviation rate and combines it with a preset threshold to determine anomalies. When an anomaly is detected, it processes the resource availability list, activates the adaptation association analysis logic, regenerates the resource switching or path adjustment adaptation scheme, and sorts and feeds it back. The scheme adjustment and parameter update unit receives the user's confirmation instruction for the adaptation scheme, issues the adjustment execution instruction and monitors the execution process. After the execution is completed, it collects the actual operation data of the resources and sends it back, and fully updates the basic parameter set of resources and the dynamic resource availability list.

Images