Multi-device collaborative management method and system based on Internet of Things

Abstract

The invention relates to the technical field of Internet of Things and industrial automation, and discloses a multi-device collaborative management method and system based on the Internet of Things, and the method comprises the steps: obtaining industrial device data, and recognizing the industrial device data as an abnormal event marked with a priority; obtaining and classifying the abnormal events, extracting an approval process template and querying a resource schedule, and generating a conflict-free allocation scheme through conflict detection; a collaborative management instruction is obtained based on the conflict-free distribution scheme updating template and is distributed and executed; and obtaining an execution feedback record, and if an energy consumption deviation detection result in the record exceeds a threshold value, judging that the abnormal event is a new abnormal event and marking a priority. According to the method, dynamic analysis and scheduling are carried out on the edge side, so that the problem of insufficient real-time performance of multi-device collaborative management in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things and industrial automation, and in particular to a multi-device collaborative management method and system based on Internet of Things. BACKGROUND

[0002] At present, with the rapid development of industrial Internet of Things, device management is changing from single device monitoring to multi-device collaborative management, which is the key to improving production efficiency and resource utilization. A large number of devices generate heterogeneous data such as running state, fault code and energy consumption in real time. How to efficiently process these data and perform collaborative scheduling is a challenge currently faced.

[0003] In one prior art, multi-device management usually relies on a centralized cloud processing architecture. This architecture uploads all device data to the cloud server for unified analysis and decision-making. However, this approach has obvious bottlenecks. The round-trip transmission of data from the device end to the cloud end will cause significant delay, limiting the real-time response capability of the system. Especially in scenarios that require rapid handling of device conflicts or personnel schedule coordination, the lack of dynamic edge-side verification mechanism often leads to chaotic resource allocation or delayed response.

[0004] In summary, the prior art has the problem of insufficient real-time performance of multi-device collaborative management. SUMMARY

[0005] The present application provides a multi-device collaborative management method and system based on Internet of Things to solve the problem of insufficient real-time performance of multi-device collaborative management in the prior art.

[0006] In a first aspect, to solve the above technical problems, the present application provides a multi-device collaborative management method based on Internet of Things, comprising: Obtaining running state data and fault code data of multi-source heterogeneous industrial devices and processing to obtain abnormal events marked with priority labels; Obtaining the abnormal events marked with priority labels and performing classification processing to extract corresponding approval process templates and generate an initial flow sequence; According to the initial flow sequence, querying associated resource records to obtain personnel schedule information and device occupation information, and determining a set of available time periods; According to the set of available time periods, performing conflict detection to generate a conflict-free allocation scheme; Updating the approval process templates based on the conflict-free allocation scheme to obtain collaborative management instructions; Distributing the collaborative management instructions and obtaining execution feedback data to generate a response confirmation record; According to the response confirmation record, energy consumption deviation detection is performed, and if it is detected that an energy consumption deviation index exceeds a preset energy consumption deviation threshold, a new abnormal event is determined and a priority label is marked.

[0007] Preferably, the running state data and fault code data of the multi-source heterogeneous industrial equipment are acquired and processed to obtain an abnormal event marked with a priority label, including: The running state data and the fault code data are acquired through an edge gateway and standardized protocol conversion is performed to obtain a uniform format data set; Real-time data aggregation is performed according to the uniform format data set, and energy consumption deviation detection is performed to obtain an energy consumption deviation index; If the energy consumption deviation index exceeds a preset energy consumption deviation threshold, the current state is determined to be an abnormal event, and a priority label is marked for the abnormal event, to obtain the abnormal event marked with the priority label.

[0008] Preferably, the abnormal event marked with the priority label is acquired and classified, a corresponding approval process template is extracted, and an initial flow sequence is generated, including: The abnormal event marked with the priority label is classified to obtain a hierarchical abnormal type; A corresponding approval process template is extracted from a preset template database according to the hierarchical abnormal type; The initial flow sequence is generated based on the approval process template.

[0009] Preferably, the initial flow sequence is queried to obtain personnel schedule information and equipment occupation information, and an available time period set is determined, including: Personnel identifiers and equipment identifiers are extracted from the initial flow sequence, and a pre-established schedule database and a pre-established equipment management database are queried according to the personnel identifiers and the equipment identifiers, to obtain personnel schedule information and equipment occupation information; The personnel schedule information and the equipment occupation information are time axis combined and processed, and a common available time period meeting a preset minimum time length requirement is screened out, to obtain the available time period set.

[0010] Preferably, the available time period set is subjected to conflict detection to generate a conflict-free allocation scheme, including: The available time period set is subjected to time overlap analysis to obtain a conflict time period; The conflict time period is subjected to conflict adjudication according to a preset priority rule, and the conflict-free time period is combined to generate the conflict-free allocation scheme.

[0011] Preferably, the updating the approval process template based on the conflict-free allocation scheme comprises: updating a field in the approval process template according to the conflict-free allocation scheme to obtain an updated approval process template; serializing the updated approval process template to generate the collaborative management instruction.

[0012] Preferably, the distributing the collaborative management instruction and obtaining execution feedback data to generate a response confirmation record comprises: distributing the collaborative management instruction to relevant device terminals and personnel terminals through a collaborative instruction distribution interface; obtaining execution feedback data returned by the device terminals and the personnel terminals; generating the response confirmation record according to the execution feedback data; if there is an unconfirmed state in the response confirmation record, retrying distribution to unconfirmed terminals after a preset time interval.

[0013] Preferably, the performing energy consumption deviation detection according to the response confirmation record comprises: extracting energy consumption related fields from the response confirmation record; calculating a new energy consumption deviation index according to the energy consumption related fields; if the new energy consumption deviation index exceeds a preset energy consumption deviation threshold, judging the current state as a new abnormal event and marking a priority label for the new abnormal event.

[0014] In a second aspect, the present application provides a multi-device collaborative management system based on Internet of Things, comprising: an abnormality identification module configured to obtain running state data and fault code data of multi-source heterogeneous industrial devices, and process the data to obtain abnormal events marked with priority labels; a process generation module configured to obtain the abnormal events marked with priority labels, and perform classification processing to extract corresponding approval process templates and generate initial process sequences; an availability analysis module configured to query associated resource records according to the initial process sequences, obtain personnel schedule information and device occupation information, and determine available time period sets; a conflict detection module configured to perform conflict detection according to the available time period sets to generate a conflict-free allocation scheme; an instruction generation module configured to update the approval process templates based on the conflict-free allocation scheme to obtain collaborative management instructions; A cooperative execution module is configured to distribute the cooperative management instruction and obtain execution feedback data, and generate a response confirmation record; A closed-loop monitoring module is configured to perform energy consumption deviation detection according to the response confirmation record, and if it is detected that an energy consumption deviation index exceeds a preset energy consumption deviation threshold, a new abnormal event is determined and a priority label is marked.

[0015] In a third aspect, the present application further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the multi-device cooperative management method based on Internet of Things according to any one of the above.

[0016] In a fourth aspect, the present application further provides a computer readable storage medium, comprising a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to execute the multi-device cooperative management method based on Internet of Things according to any one of the above when the computer program runs.

[0017] Compared with the prior art, the present application has the following beneficial effects: (1) The present application converts and aggregates the multi-source heterogeneous data in real time through the edge gateway, completes the data analysis task on the edge side close to the data source, avoids the network delay caused by transmitting massive raw data to the cloud, and can quickly and accurately identify abnormal events such as energy consumption deviation, thereby solving the problem of untimely abnormal response caused by data processing delay in the prior art.

[0018] (2) The present application queries the personnel schedule and device occupancy information, detects conflicts according to preset priority rules, dynamically associates the abnormal event processing with real-time resource availability, and establishes a clear conflict arbitration mechanism, thereby realizing the transition from static allocation to dynamic intelligent scheduling, and generating a conflict-free resource allocation scheme, thereby solving the problems of resource allocation confusion and low efficiency caused by the lack of dynamic checking in the prior art.

[0019] (3) The present application obtains execution feedback after distributing the cooperative management instruction, and re-performs energy consumption deviation detection according to the feedback record, thereby constructing a closed-loop feedback system from instruction execution to effect verification, which can autonomously evaluate the effectiveness of management operations and automatically trigger a new round of management process when the problem is not solved, thereby ensuring that the abnormal event is effectively disposed, solving the problem of lack of verification link and inability to guarantee management reliability in the prior art, and improving the running stability of the whole system. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1It is a flow chart of a multi-device collaborative management method based on Internet of Things provided by the first embodiment of the present application. Figure 2 It is a structure diagram of a multi-device collaborative management system based on Internet of Things provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0022] Referring to Figure 1 The first embodiment of the present application provides a multi-device collaborative management method based on Internet of Things, comprising the following steps: S11, obtaining running state data and fault code data of multi-source heterogeneous industrial devices, and processing to obtain an abnormal event marked with a priority label; S12, obtaining the abnormal event marked with the priority label, and classifying and processing to extract a corresponding approval process template, and generating an initial flow sequence; S13, querying associated resource records according to the initial flow sequence, obtaining personnel schedule information and device occupation information, and determining a set of available time periods; S14, performing conflict detection according to the set of available time periods, and generating a conflict-free allocation scheme; S15, updating the approval process template based on the conflict-free allocation scheme, and obtaining a collaborative management instruction; S16, distributing the collaborative management instruction, and obtaining execution feedback data, and generating a response confirmation record; S17, performing energy consumption deviation detection according to the response confirmation record, and if the detected energy consumption deviation index exceeds a preset energy consumption deviation threshold, judging as a new abnormal event and marking a priority label.

[0023] In step S11, the running state data and the fault code data of the multi-source heterogeneous industrial devices are obtained and processed to obtain an abnormal event marked with a priority label, comprising: The running state data and the fault code data are obtained through an edge gateway, and standardized protocol conversion is performed to obtain a unified format data set; Real-time data aggregation is performed according to the unified format data set, and energy consumption deviation detection is performed to obtain an energy consumption deviation index; If the energy consumption deviation index exceeds a preset energy consumption deviation threshold, the current state is determined as an abnormal event, and a priority label is marked for the abnormal event, to obtain the abnormal event with the marked priority label.

[0024] In an implementation, through an edge gateway deployed in an industrial site, running state data (such as rotation speed, current, and temperature) and fault code data are acquired from a plurality of different types and different manufacturers of industrial equipment (such as motors, sensors, and PLC controllers) in real time. Since these devices can use different communication protocols, the edge gateway has the parsing capability of multiple protocols, such as MQTT protocol and OPC UA protocol.

[0025] It should be noted that, to solve the compatibility problem of multi-source heterogeneous data, standardized protocol conversion is required. The conversion converts the original data using different protocols and data structures into a standard data structure, such as JSON format, through a predefined field mapping rule. The construction of the predefined field mapping rule is based on the analysis of the official technical documents and communication protocol specifications of each type of device to be accessed, the extraction of the data frame structure and key field definition, and the establishment of a mapping relationship table between the original field name and the standardized universal field name (such as “rotation_speed” and “temperature”) in the form of key-value pair. Through this processing, specific fields of different devices (such as the rotation speed field “speed” of a motor and the temperature field “temp” of a sensor) are accurately mapped, thereby generating a unified format data set with regular content and unified format.

[0026] It should be noted that, after obtaining the unified format data set, the system performs real-time data aggregation. The aggregation calculates the arithmetic mean of consecutive data points within a preset time window to obtain an aggregated data point that reflects the recent trend, thereby reducing the interference caused by instantaneous data fluctuations. The determination of the preset time window is the result of offline experiments, in which the system response real-time performance and the calculation resource load are balanced and optimized; for example, a shorter window (such as 10 seconds) responds quickly but has large calculation overhead, and a longer window (such as 5 minutes) has good smoothing effect but lags in response, and through testing the system comprehensive performance under different window lengths, an optimal value is finally selected, which can be 5 minutes in this embodiment. In this process, the system calculates the energy consumption deviation index according to the aggregated device power data and the reference energy consumption data pre-stored in the device configuration file. The calculation process of the energy consumption deviation index is as follows: calculating the difference between the actual average power consumption and the reference power consumption; dividing the difference by the reference power consumption to obtain a ratio; and expressing the ratio in percentage form, and the result is the energy consumption deviation index.

[0027] It should be noted that the determination of the preset energy consumption deviation threshold is based on statistical analysis of a large amount of energy consumption data of the device in a historical normal operation period. Specifically, the determination process includes calculating the probability density distribution of the historical energy consumption deviation index, and selecting the upper limit value of the confidence interval (for example, selecting the deviation value corresponding to the 95% quantile) that can cover most of the normal fluctuation range as the preset energy consumption deviation threshold. This method ensures that the threshold has scientific basis and can effectively distinguish between normal energy consumption fluctuations of the device and potential abnormal events.

[0028] In another implementation, the process of labeling the priority label can also be specifically limited. When the energy consumption deviation index is detected to exceed the preset energy consumption deviation threshold, the system determines this as an abnormal event, and uses a pre-trained decision tree classification model to assign a priority label to the abnormal event. The input features of the decision tree model include the energy consumption deviation degree of the current abnormal event, the importance level of the device in the production process, and the specific fault code type; the output of the model is a specific priority label such as "high", "medium", and "low".

[0029] It should be noted that the decision tree model is obtained by supervised learning training on a labeled data set containing thousands of historical abnormal events. Each record in the data set contains the above-mentioned input features, and the real priority label determined by analyzing the historical system running log and the maintenance work order corresponding thereto with clear handling level record. Using a standard decision tree construction algorithm (such as CART algorithm), the Gini impurity is used as the splitting criterion, and the decision tree is recursively constructed until the preset stopping condition is met, for example, the number of node samples is less than the minimum number of divisions. The minimum number of divisions is a preset hyperparameter for preventing model overfitting, and its value is determined by the method of grid search combined with k-fold cross-validation. On an independent validation set, the value that makes the model classification accuracy and generalization ability achieve the optimal balance is selected, which can be 20 in this embodiment.

[0030] Exemplarily, assume that an edge gateway obtains the running data of a critical motor through the MQTT protocol, the motor's rotation speed is 1800 rpm, the current is 5 A, and a fault code E101 is reported. The system first converts these data into a unified format dataset in JSON format: {“device_id”:“M001”,“timestamp”:“2025-09-17 10:30:00”,“rotation_speed”:1800,“fault_code”:“E101”}. According to the real-time current and voltage, the system calculates that the current actual average power consumption is 1050 W, and the reference power consumption of the motor is 1000 W, so the energy consumption deviation index is 5%. If the preset energy consumption deviation threshold determined based on historical data statistics is 4%, the system determines that this is an abnormal event. Subsequently, the system inputs the deviation degree (5%), the device importance (high), and the fault code type (E101 represents overload) into the pre-trained decision tree model as features, the model outputs a“high priority”label according to the internal rules, and finally generates an abnormal event record with the label; the preset energy consumption deviation threshold is not a fixed value, and can be dynamically configured according to the device type, historical running conditions and process requirements. For example, for high-precision devices, the threshold can be set to 2% to 3%; for ordinary load motors, the threshold can be set to 5% to 10%.

[0031] In step S12, the abnormal event with the priority label is obtained, and classification processing is performed to extract a corresponding approval process template and generate an initial flow sequence, including: The abnormal event with the priority label is classified to obtain a classified abnormal type; According to the classified abnormal type, a corresponding approval process template is extracted from a preset template database; The approval process template is serialized to generate the initial flow sequence.

[0032] In an implementation manner, the system obtains the abnormal event with the priority label generated in S11, and performs classification processing to determine the specific nature, so as to match the most suitable processing flow. The classification processing is realized by a pre-trained support vector machine (SVM) classification model. The support vector machine (SVM) classification model receives a standardized input feature vector, which includes key attributes of the abnormal event, such as energy consumption deviation value, fault code, and priority label assigned in S11. The output of the model is a specific, standardized classified abnormal type, such as“severe-device overload abnormality”or“mild-sensor communication abnormality”.

[0033] It should be noted that the support vector machine model is obtained by supervised learning training on a labeled data set containing thousands of historical abnormal event samples. Each record in the data set contains the above input features, and the true abnormal type label determined according to the historical maintenance record and the device fault analysis report. A support vector machine classifier with a radial basis function (RBF) kernel is used for training, and a grid search combined with k-fold cross-validation method is used to optimize the two key hyperparameters of the penalty coefficient C and the kernel function parameter gamma. Finally, a set of hyperparameters that achieve a preset accuracy threshold (for example, 98%) on the independent validation set is selected for the final model training. The preset accuracy threshold is determined according to the business reliability requirements (SLA, Service-Level Agreement) of the collaborative management system.

[0034] It is worth noting that after obtaining the hierarchical abnormal type, the system indexes it to extract the exactly corresponding approval process template from a preset template database. The construction process of the preset template database includes: inductive classification of the types of all historical abnormal events; according to the internal device maintenance manual, safety production procedures (SOP) and historical best practice cases, design a standardized processing workflow for each abnormal type; finally, store these workflows in a structured data format (such as XML or JSON) to form the approval process template, which clearly defines the approval level, responsible person role, recommended time limit and other fields required for processing this type of abnormality.

[0035] In another implementation, the system serializes the extracted approval process template on the edge computing node to generate an initial flow sequence. The serialization here refers to converting the structured and static field information in the template into an instruction sequence with a timestamp, following a specific order, and directly executable by the scheduling system. This processing is completed on the edge node close to the device, reducing the communication round trip with the cloud server, thereby improving the response speed. The serialization processing arranges the approval steps defined in the template according to a preset logical order. The preset logical order is determined by a set of sorting rules, which takes the priority of the event as the first sorting key (high priority first), and the timestamp of the event as the second sorting key (occurred first), ensuring that critical and urgent tasks can be processed first.

[0036] Exemplarily, the system obtains an abnormal event labeled as "high priority", whose feature vector is {energy consumption deviation: 6%, fault code: "overload", priority: "high"}. The feature vector is input into the pre-trained support vector machine model, and the model outputs that the classified abnormal type is "serious - device overload abnormality". Subsequently, the system extracts the approval template with ID T001 from the pre-set template database according to the type, which stipulates that "first-level approval" is required and the approver is "manager_A". Finally, the edge computing node serializes the template according to the sorting rule of "high priority, time stamp in front", and generates the initial flow sequence: {"sequence_id": "S001", "device_id": "M002", "timestamp": "2025-09-17 10:32:00", "approval_level": 1}.

[0037] In step S13, the available time period set is determined according to the initial flow sequence and by querying the associated resource records to obtain personnel schedule information and device occupation information, including: The personnel identifier and the device identifier are extracted from the initial flow sequence, and the pre-established schedule database and the pre-established device management database are queried according to the personnel identifier and the device identifier, respectively, to obtain the personnel schedule information and the device occupation information. The personnel schedule information and the device occupation information are subjected to time axis merging processing, and a common available time period meeting a pre-set minimum time length requirement is screened out to obtain the available time period set.

[0038] In an implementation manner, the system queries and extracts the resource identifiers necessary for performing the current collaborative management task according to the initial flow sequence generated in S12. The personnel identifier is a unique employee number assigned to each technical personnel in the enterprise human resource system, and the device identifier is a unique asset number assigned to each industrial device in the asset management system. The system initiates data query requests to the pre-established schedule database and the pre-established device management database according to the extracted identifiers.

[0039] It should be noted that the pre-established schedule database is usually an electronic calendar system integrated with the existing office automation (OA) or human resource (HR) system of the enterprise, in which the scheduled information of each staff member, such as meetings, tasks or vacations, is recorded. The pre-established equipment management database is part of the enterprise asset management (EAM) or manufacturing execution system (MES), in which the scheduled production plan, maintenance window or current occupancy status of each equipment is recorded. By querying these two databases, the system can obtain the idle time period of the staff member and the idle time period of the target equipment, respectively.

[0040] It should be noted that after obtaining the schedule information of the staff member and the equipment, respectively, the system needs to perform a time axis merging process to find the common idle time period of the two. The specific implementation of this process is as follows: the idle time period set of the staff member and the idle time period set of the equipment are regarded as two interval sets distributed on the time axis; by calculating the intersection of the two interval sets, a new set containing all the common available time periods in which the staff member and the equipment are both in an idle state is obtained.

[0041] In another implementation, the screening process can also be specifically limited. The system will perform an effectiveness verification on each common available time period obtained after the above merging process, that is, judge whether its length meets the preset minimum time length requirement. The determination of the preset minimum time length requirement is based on a pre-established knowledge base that maps different "classification exception types" determined in S12 to their standard disposal time lengths. This knowledge base is constructed by analyzing the maintenance work orders of historical similar exception events and combining the standard operating procedures (SOP) provided by the equipment manufacturer. Only the common available time period whose length is greater than or equal to the requirement will be confirmed as the final effective one and be included in the available time period set.

[0042] Exemplarily, assuming that the resources specified in the initial flow sequence are personnel identification "T001" and equipment identification "M003", and the classification of the current abnormality is "serious-equipment overload abnormality". The system first queries the schedule database and finds that the free time of "T001" is "08:00-12:00"; at the same time, it queries the equipment management database and finds that the free time of "M003" is "10:00-12:00" and "15:00-17:00". The system performs time axis merging processing on the two time sets, calculates the intersection, and obtains the common available time period as "10:00-12:00". Subsequently, the system queries the knowledge base and finds that the standard disposal time length (i.e., the preset minimum time length requirement) of "serious-equipment overload abnormality" is 2 hours. Since the length of the time period "10:00-12:00" is exactly 2 hours, it meets the requirement, so the system confirms this time period as valid and adds it to the final available time period set.

[0043] In step S14, conflict detection is performed according to the available time period set, and a conflict-free allocation scheme is generated, including: time overlap analysis is performed on the available time period set to obtain a conflict time period; conflict adjudication is performed on the conflict time period according to a preset priority rule, and the conflict-free time period is combined to generate the conflict-free allocation scheme.

[0044] In an implementation manner, after obtaining the available time period set determined in S13, the system needs to perform time overlap analysis on each time period in the set. The specific process of the analysis is as follows: for a certain available time period, the system performs time overlap comparison between the time period and other abnormality processing tasks generated by the system and to be processed or already allocated, to check whether there is a resource occupation conflict. If there is, the time period is identified and marked as a conflict time period.

[0045] It is worth noting that for the identified conflict time period, the system will perform conflict adjudication according to a set of preset priority rules to determine the final attribution of the resource (i.e., the time period). The establishment of the preset priority rules is realized through a quantitative priority scoring model, which maps each possible task to a specific basic priority score. The quantification process includes the following steps: The first step is abnormality severity quantification. The system first maps the "classified abnormality type" determined in S12 to a severity score according to a pre-established "abnormality severity score table". The establishment of this score table is obtained by stratified quantification of the objective impact data of historical similar abnormal events. This process includes: according to the historical work order records, extracting the "production interruption time" and "safety risk level" caused by each event; according to the pre-set stratification standard, the event is divided into different severity levels. The determination of the pre-set stratification standard is based on the clustering analysis of historical impact data: specifically, the system uses, for example, K-Means Clustering algorithm to process the "production interruption time" data points of all historical events, and divides them into three clusters (i.e. 'high','medium', 'low' impact clusters); the system then calculates the boundary values between these clusters (for example, 1 hour and 15 minutes), and uses these boundary values as time thresholds to define severity levels. For non-continuous data such as "safety risk level", its stratification standard is directly mapped according to the enterprise safety management regulations. For example, events with interruption time exceeding 1 hour or triggering safety alarms are defined as "severe" level; finally, a fixed quantitative score is assigned to each level ('severe','moderate','minor'). For example, "severe abnormality" leading to line shutdown can be quantified as 10 points, "moderate abnormality" leading to efficiency reduction can be quantified as 5 points, and "minor abnormality" requiring only observation can be quantified as 1 point.

[0046] The second step is equipment importance quantification. The system maps the importance level of the abnormal equipment to an importance score according to a pre-established "equipment importance level score table". The establishment of this score table is based on the position and role of the equipment in the production process flow diagram (P&ID), for example, the bottleneck equipment located in the main process path is defined as the highest level. For example, "core equipment" can be quantified as 10 points, "important equipment" as 5 points, and "auxiliary equipment" as 1 point.

[0047] Third, calculate the base priority score. Before performing the weighted summation, the system first normalizes the anomaly severity score and the device importance score to eliminate the dimensional and scale differences between the two scores. For example, the maximum-minimum value normalization method is used to map both scores to the unified interval [0, 1]. Then, the system obtains a comprehensive evaluation value by performing a weighted summation on the normalized two scores. The weight value is a configurable system parameter, which is obtained based on the analysis of historical operation and maintenance data. By analyzing a large number of processing records and final impacts of historical events, it is found that the weight coefficient of the impact of the severity of the anomaly on the overall production is about 0.7, and the weight coefficient of the importance of the device itself is about 0.3. This weight is the system default configuration, which can be adjusted according to the actual operation and maintenance strategy. In order to facilitate subsequent comparison and grade division, the system further linearly amplifies the comprehensive evaluation value obtained, for example, by multiplying it by 100 to obtain the final base priority score.

[0048] Fourth, generate adjudication rules. The system solidifies the logic of "comparing the priority scores of tasks" into a series of "IF-THEN-ELSE" form adjudication rules and stores them in the rule library.

[0049] It should be noted that the specific execution logic of the conflict adjudication is as follows: when a conflict is detected, the system obtains the priority score of the new task to be allocated and the priority score of the existing task that has occupied the time period. The system compares the two scores: if the priority score of the new task is higher than that of the existing task, the system will perform preemptive scheduling, that is, automatically suspend or reschedule the existing task, and allocate the conflict time period to the new task; if the priority score of the new task is lower than or equal to that of the existing task, the system determines that the time period is unavailable, and continues to analyze the next time period in the set of available time periods. After the system traverses all available time periods and completes all necessary conflict adjudications, it will finally determine the time period available for executing the new task, and encapsulate it into a conflict-free allocation scheme containing specific execution time, personnel and device information.

[0050] Exemplarily, assume that the set of available time slots determined by S13 contains one time slot "10:00-12:00", and the assigned resource is technician "T001". After the time overlap analysis, the system finds that in the schedule of "T001", "10:00-11:00" is occupied by a "routine device inspection" task, and thus "10:00-11:00" is identified as a conflict time slot. The system then initiates a conflict adjudication: the current to-be-assigned "severe-device overload exception" task has a severity score of 10 (normalized to 1.0), and is associated with a "core device" with an importance score of 10 (normalized to 1.0). With the weights of 0.7 and 0.3 and the amplification of 100, the final priority score of the task is 100. The occupied "routine device inspection" task has an exception type of "mild" (a severity score of 1, normalized to 0.1), and is associated with an "important device" (an importance score of 5, normalized to 0.5). The final priority score of the task is 22. Since 100 > 22, the system adjudicates that the new task wins, and automatically reschedules the "routine device inspection" task to other idle time. In this way, the complete "10:00-12:00" time slot is successfully assigned, and the system finally generates a conflict-free assignment scheme: { "device_id": "M003", "technician_id": "T001", "time_slot": "10:00-12:00"}.

[0051] In step S15, the conflict-free assignment scheme is used to update the approval process template, to obtain a collaborative management instruction, including: updating the fields in the approval process template according to the conflict-free assignment scheme, to obtain an updated approval process template; serializing the updated approval process template, to generate the collaborative management instruction.

[0052] In an implementation manner, after the system generates the conflict-free assignment scheme in S14, the specific resource scheduling information in the scheme is used to update the general approval process template extracted in S12. The updating process is a field mapping and filling operation, in which the personnel identifier, the device identifier, and the specific execution time slot in the conflict-free assignment scheme are filled into the placeholder fields in the approval process template that are preset for defining the execution resources, to obtain an updated approval process template with complete content and clear information.

[0053] It should be noted that the updated approval process template is a task blueprint to be executed, which contains complete context information. In order to convert it into a command that can be directly parsed and executed by the device and personnel terminal, the system needs to perform serialization processing on it. The serialization processing here refers to extracting and integrating the multi-level, structured information in the updated template into a flattened, single data object (such as a JSON object), which is the final collaborative management instruction. This process also includes assigning a unique instruction ID to the instruction for subsequent tracking, and placing it in the correct position in the execution queue according to the priority and timestamp of the event, to ensure that the instruction can be timely and orderly distributed.

[0054] Exemplarily, assuming that the approval process template extracted by S12 is {“template_id”:“T001”,“approver”:“manager_B”,“assigned_technician”:null,“scheduled_time”:null}. The conflict-free allocation scheme generated by S14 is {“device_id”:“M003”,“technician_id”:“T001”,“time_slot”:“10:00-12:00”}. The system fills“technician_id”:“T001” and“time_slot”:“10:00-12:00” in the allocation scheme into the assigned_technician and scheduled_time fields of the template, respectively, to obtain the updated approval process template: {...,“assigned_technician”:“T001”,“scheduled_time”:“10:00-12:00”}. The system performs serialization processing on the updated template, extracts the key information, and generates the final collaborative management instruction: {“instruction_id”:“I001”,“device_id”:“M003”,“technician_id”:“T001”,“time_slot”:“10:00-12:00”,“action”:“repair”}.

[0055] In step S16, the collaborative management instruction is distributed, and execution feedback data is obtained, and a response confirmation record is generated, including: Distribute the collaborative management instruction to the related device terminal and personnel terminal through the collaborative instruction distribution interface; Obtain the execution feedback data returned by the device terminal and the personnel terminal; Generate the response confirmation record according to the execution feedback data; If the response confirmation record shows an unconfirmed status, the system will automatically trigger a retry distribution operation for the unconfirmed terminal after a preset time interval.

[0056] In one implementation, the system distributes the collaborative management instructions generated in S15 to the designated device terminals (e.g. the control panel of the device) and personnel terminals (e.g. the mobile device of the technician) in the form of digital messages through a standardized collaborative instruction distribution interface (e.g. based on RESTful API or message queue (MQ)).

[0057] It is worth noting that after receiving and executing the instructions, the device terminals and personnel terminals will return execution feedback data. This feedback data is a structured data set that contains the execution status of the instructions (e.g. "completed" "failed"), the completion timestamp, and key performance parameters (e.g. the energy consumption value of the device after maintenance). The system aggregates and analyzes all returned feedback data and associates it with the corresponding instruction ID, thereby generating a response confirmation record containing the timestamp and execution status.

[0058] It is worth noting that in order to ensure the reliability of instruction distribution, the present application contains an automatic retry mechanism. The system monitors the status of the response confirmation record while distributing the instructions. If the status of a terminal is still "unconfirmed" within a preset time interval, the system will automatically trigger a retry distribution operation for that terminal. The determination of the preset time interval is based on statistical analysis of historical network communication delay data and the real-time response requirements of the business. This process includes calculating the 95th percentile of historical instruction confirmation time and setting it as a reference value (e.g. 30 seconds) to ensure that the interval can tolerate normal network fluctuations and start a retry in time when communication anomalies occur. The preset time interval can be dynamically adjusted according to network conditions and device types, for example, for large devices with slow response, the interval can be extended to 60-120 seconds.

[0059] Exemplarily, the system dispatches a co-management instruction to the mobile device of technician "T005" and the control panel of device "M005" through the message queue. The system then generates a response confirmation record: {"instruction_id": "I003", "device_M005_status": "pending", "technician_T005_status": "pending"}. Device "M005" soon returns a confirmation, and the record is updated to {..., "device_M005_status": "confirmed",...}. However, after the preset 30-second time interval, the status of "T005" is still "pending". The system then triggers the retry mechanism and re-dispatches the instruction to the mobile device of "T005". After the second dispatch, "T005" returns a confirmation, and the final response confirmation record is updated to {..., "device_M005_status": "confirmed", "technician_T005_status": "confirmed"}.

[0060] In step S17, energy consumption deviation detection is performed according to the response confirmation record, and if it is detected that the energy consumption deviation index exceeds the preset energy consumption deviation threshold, a new abnormal event is determined and a priority label is marked, including: extracting an energy consumption related field from the response confirmation record; calculating a new energy consumption deviation index according to the energy consumption related field; if the new energy consumption deviation index exceeds the preset energy consumption deviation threshold, the current state is determined as a new abnormal event, and a priority label is marked for the new abnormal event.

[0061] In an implementation manner, this step serves as a verification and feedback closed loop link of the entire co-management process. After obtaining the response confirmation record generated in S16, the system parses the record to extract the energy consumption related field related to the device performance evaluation. It should be noted that the extraction process is based on a predefined field matching rule: since the response confirmation record is in a structured data format (for example, JSON), the data parsing module of the system will search in the structured data according to the pre-configured keyword (for example, "post_maintenance_power") as the energy consumption field identifier, and the value corresponding to the keyword will be extracted as the energy consumption related field.

[0062] It should be noted that the system calculates the energy consumption deviation index of the new round by using the same calculation method as in S11 according to the extracted energy consumption related field. Then, the index is compared with the same preset energy consumption deviation threshold defined in S11.

[0063] It should be noted that if the newly calculated energy consumption deviation index still exceeds the preset energy consumption deviation threshold, it indicates that the previous collaborative management operation has not completely solved the problem or a new problem has occurred in the operation process. In this case, the system judges the current state as a new abnormal event and uses the same decision tree classification model as in S11 to assign a priority label to the new abnormal event according to the current latest event characteristics. The newly generated abnormal event with the priority label will automatically re-enter the management process of the application as the input of S12, thereby starting a new round of analysis, scheduling and execution to realize real closed-loop adaptive control.

[0064] For example, assume that the response confirmation record generated in S16 is {“instruction_id”:“I003”,...,“post_maintenance_power”:1045W}. The system extracts the value of the energy consumption related field as 1045W from it. Given that the reference power consumption of the device is 1000W, the system calculates the new energy consumption deviation index as 4.5%. This value still exceeds the 4% preset energy consumption deviation threshold determined in S11. Therefore, the system determines that it is a new abnormal event and assigns a priority to it again (for example, a higher “urgent” priority may be assigned because the problem persists), thereby triggering a new round of collaborative management process.

[0065] It should be noted that the application takes energy consumption deviation detection as an example of closed-loop monitoring, but the triggering conditions of the closed-loop monitoring module are not limited to this. In actual applications, corresponding thresholds and detection logic can be set according to other performance indicators (such as device vibration value, temperature, yield, etc.) extracted from the response confirmation record, thereby realizing closed-loop verification of the solution effect of multiple types of abnormalities.

[0066] In order to facilitate the understanding of the application, some preferred embodiments of the application will be described further below.

[0067] In an implementation manner, the method of the application can be applied to a complete and closed-loop intelligent factory device maintenance scenario. Taking an energy consumption abnormality of a key motor on a production line as an example, the complete working process is as follows: Step one, the edge gateway deployed in the workshop real-time collects the current data of the motor (ID: "M003") and finds that the current data is continuously high. Through the method described in S11, the energy consumption deviation index of the motor is calculated as 6%, which exceeds the preset energy consumption deviation threshold of 4%. The system determines that this is an abnormal event and labels the event as "high priority" through the decision tree model combined with the "high" equipment importance level.

[0068] Step two, the system obtains the high-priority abnormal event and classifies it as "serious-equipment overload abnormality" through the support vector machine model described in S12. Then, the system extracts the corresponding approval process template from the template database and generates an initial process sequence containing the first-level approval requirements through the serialization processing of the edge node.

[0069] Step three, the process requires a senior maintenance technician (e.g., ID: "T001"). The system queries the schedule database of "T001" and the equipment management database of "M003" through the method described in S13, and determines that there is a common available period of 2 hours that meets the standard disposal duration (2 hours) on "2025-09-17 14:00-16:00" after time axis merging and filtering.

[0070] Step four, the system performs conflict detection on the "14:00-16:00" period as described in S14 and finds that technician "T001" has scheduled a "routine equipment inspection" task during this period. According to the quantitative scoring model described in S14, the priority score of the current "serious-equipment overload abnormality" task is 100 points, while the priority score of the occupied "routine equipment inspection" task is 22 points. Since 100>22, the system determines that the new task wins according to the priority rules, automatically suspends the "routine equipment inspection" task and notifies the relevant personnel to reschedule.

[0071] Step five, after conflict resolution, the system updates the conflict-free allocation scheme containing the specific information of technician "T001" and the period "14:00-16:00" to the approval process template through the method described in S15, and serializes to generate a unique and executable collaborative management instruction.

[0072] Step six, the instruction is sent to the mobile terminal of "T001" and the local console of "M003" through the distribution interface described in S16. After the task is completed, the system receives the execution feedback containing the "completed" state and the energy consumption data after maintenance, and generates a response confirmation record.

[0073] Step seven, the system analyzes the post-maintenance energy consumption data in the response confirmation record according to the method described in S17, and calculates a new energy consumption deviation index of 1.5%. Since the value is lower than the threshold of 4%, the system determines that the abnormality has been successfully resolved, and the management process loop is closed.

[0074] In another implementation, the system deployment architecture of the present application can also be configured differently. In the foregoing embodiment, all data analysis, model inference and decision generation are completed on the local edge gateway. This edge computing architecture can minimize latency and is suitable for single factories or production lines with extremely high real-time control requirements. In addition, the present application can also be deployed under a side-cloud collaborative architecture: in this architecture, the edge gateway is responsible for the preliminary collection, cleaning and aggregation of data, and uploads the aggregated data to the cloud server; the cloud server, with its powerful computing power, is responsible for performing complex model training, global resource optimization and cross-factory collaborative scheduling strategy calculation; the final collaborative management instructions calculated are issued to the edge gateway for local execution. This architecture is suitable for large enterprises or cross-regional industrial groups that need to perform large-scale data analysis and centralized management.

[0075] In summary, the present application discloses a multi-device collaborative management method based on the Internet of Things, which includes obtaining and processing the running state data and fault code data of multiple source heterogeneous industrial devices to obtain abnormal events marked with priority labels; obtaining and analyzing the abnormal events to generate an initial flow sequence containing resource allocation; querying and analyzing resource availability based on the flow sequence to generate a conflict-free allocation scheme through conflict detection; generating and distributing collaborative management instructions based on the scheme; obtaining feedback of instruction execution and performing verification, and if the problem persists, generating a new abnormal event to start the next round of management. The present application solves the problem of insufficient real-time performance and low resource allocation efficiency in the prior art caused by rigid collaborative strategies and lack of real-time dynamic verification by integrating multiple source heterogeneous data in real time on the edge side, dynamically detecting resource conflicts, and establishing a closed-loop feedback mechanism for execution verification.

[0076] Reference Figure 2 The second embodiment of the present application provides a multi-device collaborative management system based on the Internet of Things, comprising: An abnormality identification module for obtaining and processing the running state data and fault code data of multiple source heterogeneous industrial devices to obtain abnormal events marked with priority labels; A flow generation module for obtaining and classifying the abnormal events marked with priority labels, extracting corresponding approval process templates, and generating an initial flow sequence; An availability analysis module is configured to query associated resource records according to the initial flow procedure sequence, acquire personnel schedule information and equipment occupation information, and determine a set of available time periods; A conflict detection module is configured to perform conflict detection according to the set of available time periods, and generate a conflict-free allocation scheme; An instruction generation module is configured to update the approval flow template based on the conflict-free allocation scheme, and obtain a collaborative management instruction; A collaborative execution module is configured to distribute the collaborative management instruction, acquire execution feedback data, and generate a response confirmation record; A closed-loop monitoring module is configured to perform energy consumption deviation detection according to the response confirmation record, and if it is detected that an energy consumption deviation index exceeds a preset energy consumption deviation threshold, a new abnormal event is determined and a priority label is marked.

[0077] It should be noted that the multi-device collaborative management system based on the Internet of Things provided by the embodiments of the present application is used to execute all process steps of the multi-device collaborative management method based on the Internet of Things of the above-mentioned embodiments, and the working principles and advantages of the two are one-to-one correspondence, so they will not be repeated.

[0078] The embodiments of the present application also provide an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a multi-device collaborative management program based on the Internet of Things. The processor executes the computer program to implement the steps in each of the multi-device collaborative management methods based on the Internet of Things described above, such as Figure 1 The steps S11 shown. Alternatively, the processor executes the computer program to implement the functions of each module / unit in each system embodiment, such as the abnormality identification module.

[0079] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments that can complete a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0080] The electronic device can be a desktop computer, a notebook, a palm computer, and a smart tablet, etc. The electronic device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and can include more or fewer components than the above, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0081] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like, and is a control center of the electronic device, which connects various parts of the electronic device through various interfaces and lines.

[0082] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function (such as a sound playing function, an image playing function, etc.), etc.; and the data storage area can store data created based on use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash storage device, or other volatile solid-state storage device.

[0083] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. that can carry the computer program code. It should be noted that the contents included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0084] It should be noted that the above-described device embodiments are only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. In addition, the connection relationship between the modules in the device embodiment provided by the present application indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0085] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-described specific embodiments are only for the specific embodiments of the present application and are not used to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for managing a plurality of devices based on Internet of Things, characterized in that, The application comprises the following steps: acquiring running state data and fault code data of multi-source heterogeneous industrial equipment, processing the data, and obtaining an abnormal event marked with a priority label; acquiring the abnormal event marked with the priority label and performing classification processing to extract a corresponding approval process template and generate an initial flow sequence; querying associated resource records according to the initial flow sequence, acquiring personnel schedule information and equipment occupation information, and determining a set of available time periods; performing conflict detection according to the set of available time periods to generate a conflict-free allocation scheme; updating the approval process template based on the conflict-free allocation scheme to obtain a collaborative management instruction; distributing the collaborative management instruction and acquiring execution feedback data to generate a response confirmation record; performing energy consumption deviation detection according to the response confirmation record, and if the detected energy consumption deviation index exceeds a preset energy consumption deviation threshold, determining a new abnormal event and marking a priority label. 2.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, The application comprises the following steps: acquiring running state data and fault code data of multi-source heterogeneous industrial equipment, processing the data, and obtaining an abnormal event marked with a priority label; acquiring the abnormal event marked with the priority label and performing classification processing to extract a corresponding approval process template and generate an initial flow sequence; querying associated resource records according to the initial flow sequence, acquiring personnel schedule information and equipment occupation information, and determining a set of available time periods; 3.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, performing conflict detection according to the set of available time periods to generate a conflict-free allocation scheme; updating the approval process template based on the conflict-free allocation scheme to obtain a collaborative management instruction; distributing the collaborative management instruction and acquiring execution feedback data to generate a response confirmation record; performing energy consumption deviation detection according to the response confirmation record, and if the detected energy consumption deviation index exceeds a preset energy consumption deviation threshold, determining a new abnormal event and marking a priority label. 4.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, The application comprises the following steps: acquiring running state data and fault code data of multi-source heterogeneous industrial equipment, processing the data, and obtaining an abnormal event marked with a priority label; acquiring the abnormal event marked with the priority label and performing classification processing to extract a corresponding approval process template and generate an initial flow sequence; 5.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, querying associated resource records according to the initial flow sequence, acquiring personnel schedule information and equipment occupation information, and determining a set of available time periods; performing conflict detection according to the set of available time periods to generate a conflict-free allocation scheme; updating the approval process template based on the conflict-free allocation scheme to obtain a collaborative management instruction; 6.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, distributing the collaborative management instruction and acquiring execution feedback data to generate a response confirmation record; performing energy consumption deviation detection according to the response confirmation record, and if the detected energy consumption deviation index exceeds a preset energy consumption deviation threshold, determining a new abnormal event and marking a priority label. The application comprises the following steps: acquiring running state data and fault code data of multi-source heterogeneous industrial equipment, processing the data, and obtaining an abnormal event marked with a priority label; acquiring the abnormal event marked with the priority label and performing classification processing to extract a corresponding approval process template and generate an initial flow sequence; querying associated resource records according to the initial flow sequence, acquiring personnel schedule information and equipment occupation information, and determining a set of available time periods; performing conflict detection according to the set of available time periods to generate a conflict-free allocation scheme; updating the approval process template based on the conflict-free allocation scheme to obtain a collaborative management instruction; distributing the collaborative management instruction and acquiring execution feedback data to generate a response confirmation record; performing energy consumption deviation detection according to the response confirmation record, and if the detected energy consumption deviation index exceeds a preset energy consumption deviation threshold, determining a new abnormal event and marking a priority label. According to the conflict-free distribution scheme, a field in the approval process template is updated, and an updated approval process template is obtained; The updated approval process template is serialized to generate the collaborative management instruction. 7.The Internet-of-Things based multi-device cooperative management method according to claim 1, characterized in that, The collaborative management instruction is distributed, and execution feedback data is obtained to generate a response confirmation record, including: The collaborative management instruction is distributed to related device terminals and personnel terminals through a collaborative instruction distribution interface; Execution feedback data returned by the device terminals and the personnel terminals is obtained; The response confirmation record is generated according to the execution feedback data; If there is an unconfirmed state in the response confirmation record, the unconfirmed terminal is retried and distributed after a preset time interval. 8.The Internet of Things based multi-device cooperative management method according to claim 1, characterized in that, The energy consumption deviation is detected according to the response confirmation record, and if the energy consumption deviation index exceeds the preset energy consumption deviation threshold, it is judged as a new abnormal event and a priority label is marked, including: Energy consumption related fields are extracted from the response confirmation record; A new energy consumption deviation index is calculated according to the energy consumption related fields; If the new energy consumption deviation index exceeds the preset energy consumption deviation threshold, the current state is judged as a new abnormal event, and a priority label is marked for the new abnormal event.

9. A multi-device cooperative management system based on Internet of Things, characterized in that, Including: An abnormality identification module is configured to obtain running state data and fault code data of a plurality of source heterogeneous industrial devices, and process the data to obtain an abnormal event marked with a priority label; A process generation module is configured to obtain the abnormal event marked with the priority label, and perform classification processing to extract a corresponding approval process template, and generate an initial process sequence; An availability analysis module is configured to query associated resource records according to the initial process sequence, obtain personnel schedule information and device occupation information, and determine a set of available time periods; A conflict detection module is configured to perform conflict detection according to the set of available time periods, and generate a conflict-free distribution scheme; An instruction generation module is configured to update the approval process template based on the conflict-free distribution scheme to obtain a collaborative management instruction; A collaborative execution module is configured to distribute the collaborative management instruction, and obtain execution feedback data to generate a response confirmation record; A closed-loop monitoring module is configured to detect energy consumption deviation according to the response confirmation record, and if the energy consumption deviation index exceeds the preset energy consumption deviation threshold, it is judged as a new abnormal event and a priority label is marked.

Images