Multi-agent processing method and device for abnormal work order

Abstract

The invention relates to an abnormal work order-oriented multi-agent processing method and device, computer equipment, a computer readable storage medium and a computer program product, which can be applied to the technical field of computers. The method comprises the following steps: performing feature extraction processing on an abnormal work order to obtain work order features; key features are extracted from the work order features, and a feature database is constructed; determining a target processing information type, and generating a work order processing scheme of the abnormal work order; obtaining abnormal problem information, obtaining execution process data of the work order processing scheme by implementing the monitoring model, and performing backup processing on the execution process data to obtain process backup data; inputting the abnormal problem information and the process backup data into a problem analysis model to obtain a problem analysis result of the work order processing scheme; determining to-be-adjusted scheme parameters; and adjusting the to-be-adjusted scheme parameters in the work order processing scheme to obtain a target work order processing scheme of the abnormal work order. By adopting the method, the abnormal work order processing efficiency can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and in particular to a multi-agent processing method and device for abnormal work orders, a computer device, a computer readable storage medium, and a computer program product. BACKGROUND

[0002] At present, in the process of enterprise operation management, the timely processing of abnormal work orders is crucial to guarantee the stable operation of a business system. How to efficiently process abnormal work orders is particularly important.

[0003] In the conventional technology, when processing abnormal work orders, a manual processing method is generally used, that is, an operation and maintenance personnel analyzes abnormal work orders according to his own experience and formulates a corresponding processing scheme. However, the abnormal work order processing by this method requires a lot of manual processing time, resulting in low efficiency of abnormal work order processing. SUMMARY

[0004] Therefore, it is necessary to provide a multi-agent processing method and device for abnormal work orders, a computer device, a computer readable storage medium, and a computer program product, which can improve the efficiency of abnormal work order processing.

[0005] In a first aspect, the present application provides a multi-agent processing method for abnormal work orders. The method comprises:

[0006] Obtaining an abnormal work order, performing feature extraction processing on the abnormal work order to obtain work order features;

[0007] Extracting key features from the work order features, and constructing a feature database based on the key features;

[0008] Determining a target processing information type according to the feature database, and generating a work order processing scheme for the abnormal work order based on the target processing information type;

[0009] Obtaining abnormal problem information generated in the execution process of the work order processing scheme, and obtaining execution process data of the work order processing scheme by implementing a monitoring model, performing backup processing on the execution process data to obtain process backup data;

[0010] Inputting the abnormal problem information and the process backup data into a problem analysis model to perform problem analysis processing, and obtaining a problem analysis result of the work order processing scheme;

[0011] Determining a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result;

[0012] Adjusting parameters to be adjusted in the work order processing scheme to obtain a target work order processing scheme of the abnormal work order.

[0013] In one of the embodiments, the determining the target processing information type according to the feature database comprises:

[0014] Determining a plurality of preset processing information types; the plurality of processing information types comprise a centralized coordination type, a distributed coordination type and a hybrid architecture type;

[0015] Determining the target processing information type from the plurality of processing information types according to the key features in the feature database.

[0016] In one of the embodiments, the generating the work order processing scheme of the abnormal work order based on the target processing information type comprises:

[0017] Determining task allocation information, conflict resolution information and collaborative decision information based on the target processing information type;

[0018] Generating the work order processing scheme based on the task allocation information, the conflict resolution information and the collaborative decision information.

[0019] In one of the embodiments, the obtaining execution process data of the work order processing scheme by implementing a monitoring model comprises:

[0020] Monitoring processing timeliness, processing quality and resource efficiency of the work order processing scheme by the monitoring model to obtain processing timeliness information, processing quality information and resource efficiency information of the work order processing scheme;

[0021] Determining the execution process data according to the processing timeliness information, the processing quality information and the resource efficiency information.

[0022] In one of the embodiments, the extracting key features from the work order features comprises:

[0023] Performing multi-stage filtering processing on the work order features to obtain the key features; wherein the multi-stage filtering processing comprises preliminary screening processing based on a variance threshold, feature screening processing based on correlation analysis, feature screening processing based on model evaluation and feature verification processing based on business information.

[0024] In one of the embodiments, the determining parameters to be adjusted in the work order processing scheme according to the problem analysis result comprises:

[0025] determine an adjustment level of the work order processing scheme according to the problem analysis result; the adjustment level comprises a first adjustment level, a second adjustment level or a third adjustment level;

[0026] determine a scheme parameter to be adjusted in the work order processing scheme according to the problem analysis result and the adjustment level.

[0027] In a second aspect, the present application further provides a multi-agent processing device for an abnormal work order. The device comprises:

[0028] a work order acquisition module, configured to acquire an abnormal work order, perform feature extraction processing on the abnormal work order, and obtain work order features;

[0029] a feature construction module, configured to extract key features from the work order features, and construct a feature database based on the key features;

[0030] a scheme generation module, configured to determine a target processing information type according to the feature database, and generate a work order processing scheme for the abnormal work order based on the target processing information type;

[0031] a backup processing module, configured to acquire abnormal problem information generated in an execution process of the work order processing scheme, acquire execution process data of the work order processing scheme by implementing a monitoring model, perform backup processing on the execution process data, and obtain process backup data;

[0032] a data input module, configured to input the abnormal problem information and the process backup data to a problem analysis model for problem analysis processing, and obtain a problem analysis result of the work order processing scheme;

[0033] a parameter determination module, configured to determine a scheme parameter to be adjusted in the work order processing scheme according to the problem analysis result;

[0034] a parameter processing module, configured to perform adjustment processing on the scheme parameter to be adjusted in the work order processing scheme, and obtain a target work order processing scheme for the abnormal work order.

[0035] In a third aspect, the present application further provides a computer device. The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0036] acquire an abnormal work order, perform feature extraction processing on the abnormal work order, and obtain work order features;

[0037] extract key features from the work order features, and construct a feature database based on the key features;

[0038] According to the feature database, a target processing information type is determined, and a work order processing scheme of the abnormal work order is generated based on the target processing information type;

[0039] Abnormal problem information generated in an execution process of the work order processing scheme is acquired, execution process data of the work order processing scheme is acquired through an implementation monitoring model, backup processing is performed on the execution process data, and process backup data is obtained;

[0040] The abnormal problem information and the process backup data are input into a problem analysis model for problem analysis processing, and a problem analysis result of the work order processing scheme is obtained;

[0041] According to the problem analysis result, a to-be-adjusted scheme parameter in the work order processing scheme is determined;

[0042] The to-be-adjusted scheme parameter in the work order processing scheme is adjusted, and a target work order processing scheme of the abnormal work order is obtained.

[0043] In a fourth aspect, the present application further provides a computer readable storage medium. The computer readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to implement the following steps:

[0044] An abnormal work order is acquired, and feature extraction processing is performed on the abnormal work order to obtain work order features;

[0045] Key features are extracted from the work order features, and a feature database is constructed based on the key features;

[0046] According to the feature database, a target processing information type is determined, and a work order processing scheme of the abnormal work order is generated based on the target processing information type;

[0047] Abnormal problem information generated in an execution process of the work order processing scheme is acquired, execution process data of the work order processing scheme is acquired through an implementation monitoring model, backup processing is performed on the execution process data, and process backup data is obtained;

[0048] The abnormal problem information and the process backup data are input into a problem analysis model for problem analysis processing, and a problem analysis result of the work order processing scheme is obtained;

[0049] According to the problem analysis result, a to-be-adjusted scheme parameter in the work order processing scheme is determined;

[0050] The to-be-adjusted scheme parameter in the work order processing scheme is adjusted, and a target work order processing scheme of the abnormal work order is obtained.

[0051] In a fifth aspect, the present application also provides a computer program product. The computer program product comprises a computer program which, when executed by a processor, implements the following steps:

[0052] obtaining an abnormal work order, performing feature extraction processing on the abnormal work order to obtain work order features;

[0053] extracting key features from the work order features, and constructing a feature database based on the key features;

[0054] determining a target processing information type according to the feature database, and generating a work order processing scheme for the abnormal work order based on the target processing information type;

[0055] obtaining abnormal problem information generated in the execution process of the work order processing scheme, and obtaining execution process data of the work order processing scheme by implementing a monitoring model, performing backup processing on the execution process data to obtain process backup data;

[0056] inputting the abnormal problem information and the process backup data into a problem analysis model for problem analysis processing to obtain a problem analysis result of the work order processing scheme;

[0057] determining a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result;

[0058] adjusting the to-be-adjusted scheme parameter in the work order processing scheme to obtain a target work order processing scheme for the abnormal work order.

[0059] The multi-agent processing method, device, computer equipment, computer readable storage medium and computer program product for the abnormal work order described above obtain an abnormal work order, perform feature extraction processing on the abnormal work order to obtain work order features, extract key features from the work order features, construct a feature database based on the key features, determine a target processing information type according to the feature database, generate a work order processing scheme for the abnormal work order based on the target processing information type, obtain abnormal problem information generated in the execution process of the work order processing scheme, and obtain execution process data of the work order processing scheme by implementing a monitoring model, perform backup processing on the execution process data to obtain process backup data, input the abnormal problem information and the process backup data into a problem analysis model to perform problem analysis processing to obtain a problem analysis result of the work order processing scheme, determine a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result, and perform adjustment processing on the to-be-adjusted scheme parameter in the work order processing scheme to obtain a target work order processing scheme for the abnormal work order. The scheme performs feature extraction processing on an abnormal work order to obtain work order features, extracts key features from the work order features to construct a feature database, is conducive to accurately identifying and structurally storing key information of the abnormal work order, and is conducive to quickly matching processing strategies. According to the feature database, a target processing information type is determined, and a work order processing scheme is generated, which is conducive to realizing automatic generation of an abnormal work order processing scheme, thereby reducing the time cost of manually formulating a scheme. The execution process data of the work order processing scheme is obtained by implementing a monitoring model, and backup processing is performed, which is conducive to comprehensively mastering the real-time state and historical record of scheme execution, thereby providing complete data support for subsequent problem analysis. The abnormal problem information and the process backup data are input into a problem analysis model to perform problem analysis processing, which is conducive to accurately locating problems and deficiencies in the work order processing scheme, thereby determining a scheme parameter that needs to be adjusted. The target work order processing scheme is obtained by performing adjustment processing on the to-be-adjusted scheme parameter, which is conducive to optimizing the processing flow and processing strategy of the abnormal work order, thereby improving the efficiency and accuracy of abnormal work order processing. BRIEF DESCRIPTION OF DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the embodiments or related art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating any inventive labor.

[0061] Figure 1 A flowchart of a multi-agent processing method for abnormal work orders in an embodiment;

[0062] Figure 2 This is a flowchart illustrating a multi-agent processing method for abnormal work orders in another embodiment.

[0063] Figure 3 This is a structural block diagram of a multi-agent processing device for abnormal work orders in one embodiment.

[0064] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0066] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0067] In one exemplary embodiment, such as Figure 1 As shown, a multi-agent processing method for abnormal work orders is provided. This embodiment illustrates the method by applying it to a terminal. It is understood that this method can also be applied to a server, or to a system including both a terminal and a server, and is implemented through interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, etc.; the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. In this embodiment, the method includes the following steps:

[0068] Step S101: Obtain abnormal work orders, perform feature extraction processing on abnormal work orders, and obtain work order features;

[0069] Step S102: Extract key features from work order features, and construct a feature database based on the key features;

[0070] Step S103: Determine the target processing information type based on the feature database, and generate a work order processing plan for the abnormal work order based on the target processing information type.

[0071] Step S104, acquire the abnormal problem information generated in the execution process of the work order processing scheme, and acquire the execution process data of the work order processing scheme through the implementation monitoring model, backup the execution process data to obtain process backup data;

[0072] Step S105, input the abnormal problem information and process backup data into the problem analysis model for problem analysis processing to obtain the problem analysis result of the work order processing scheme;

[0073] Step S106, determine the to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result;

[0074] Step S107, adjust the to-be-adjusted scheme parameter in the work order processing scheme to obtain the target work order processing scheme of the abnormal work order.

[0075] Among them, the abnormal work order can be a work order that appears a fault, a problem or needs to be processed in the system running process.

[0076] Among them, the feature extraction processing can be a process of extracting features of the abnormal work order.

[0077] Among them, the work order feature can be various feature information extracted from the abnormal work order, including work order basic attribute, equipment / asset information, time sequence behavior data, text description feature, processing process feature and environment correlation feature.

[0078] Among them, the key feature can be an important feature selected from the work order feature, including dynamic behavior feature, text semantic feature, topology correlation feature and graph neural network feature.

[0079] Among them, the feature database can be a database formed after analyzing and processing the key features.

[0080] Among them, the target processing information type can be a classification type of processing method determined according to the feature database, which is classified into centralized coordination method, distributed collaboration and hybrid architecture.

[0081] Among them, the work order processing scheme can be a processing scheme formulated according to the target processing information type, and the processing scheme is uniformly managed by the scheme management library.

[0082] Among them, the abnormal problem information can be problem information generated in the implementation process of the work order processing scheme, which forms a problem database after being collected.

[0083] Among them, the implementation monitoring model can be a model for monitoring the processed scheme, which monitors the processing timeliness, quality compliance and resource efficiency.

[0084] The execution process data can be data of the work order processing scheme in the execution process obtained by implementing the monitoring model, and the data records the implementation process of the processing scheme.

[0085] The backup processing can be a process of recording and storing the execution process data.

[0086] The process backup data can be data obtained after the backup processing of the execution process data.

[0087] The problem analysis model can be a model (for example, a neural network model) for problem analysis processing of the abnormal problem information and the process backup data, and the model is also called a feedback adjustment model.

[0088] The problem analysis processing can be a process of inputting the abnormal problem information and the process backup data into the problem analysis model for analysis.

[0089] The problem analysis result can be an analysis result obtained after the problem analysis processing of the abnormal problem information and the process backup data, and the result is used to determine a to-be-adjusted scheme parameter in the work order processing scheme.

[0090] The to-be-adjusted scheme parameter can be a parameter in the work order processing scheme that needs to be adjusted according to the problem analysis result.

[0091] The adjustment processing can be a process of updating the to-be-adjusted scheme parameter in the work order processing scheme.

[0092] The target work order processing scheme can be a final processing scheme obtained after the adjustment processing of the to-be-adjusted scheme parameter in the work order processing scheme, and the scheme guarantees the efficiency and accuracy of the implementation scheme in the implementation process.

[0093] Optionally, after the terminal acquires the abnormal work order, the terminal performs feature extraction processing on the abnormal work order, analyzes the source and type of the features in the extraction process, the source and type of the features include work order basic attributes, equipment / asset information, time sequence behavior data, text description features, processing process features and environment related features, so as to obtain work order features; the terminal extracts key features from the work order features, the key features include dynamic behavior features, text semantic features, topology correlation features and graph neural network features, the terminal selects and optimizes the key features, first performs a multi-stage filtering strategy, the multi-stage filtering strategy includes preliminary screening, correlation analysis, model evaluation and business experience, the method of preliminary screening is variance threshold, the correlation analysis method is mutual information, the model evaluation is SHAP (a tool for machine learning model interpretability) value sorting, and the business verification is preset rule calibration, a feature database is constructed based on the key features, the feature database includes feature types and storage schemes, the feature types are real-time features, near-line features and historical features, the real-time feature storage scheme is Redis Time Series (Redis time series database), the near-line feature storage scheme is a distributed real-time analysis database, and the historical feature storage scheme is a data lake storage framework; the terminal determines the target processing information type according to the feature database, the target processing information type is divided into centralized coordination method, distributed collaboration and hybrid architecture, the centralized coordination method is suitable for small systems, the distributed collaboration is suitable for large and scattered systems, and the hybrid architecture is suitable for smart regional management, a work order processing scheme of the abnormal work order is generated based on the target processing information type, and a scheme management library is established to uniformly manage the work order processing scheme; the terminal acquires abnormal problem information generated in the execution process of the work order processing scheme, and collects problems generated in the scheme implementation process to form a problem database, the terminal obtains execution process data of the work order processing scheme through an implementation monitoring model, the implementation monitoring model monitors the work order processing scheme, and the implementation monitoring model monitors processing timeliness monitoring, quality compliance monitoring and resource efficiency, wherein the processing timeliness monitoring includes monitoring of the first response time, and the monitoring method of the first response time is specifically the time difference between work order creation and first operation, and when the time difference is greater than 5 minutes, it is a high-risk work order, when the processing stagnation time is too long and the state is not updated for more than 30 minutes, it is a medium-risk work order, the terminal performs backup processing on the execution process data to obtain process backup data; the terminal inputs the abnormal problem information and the process backup data into a problem analysis model for problem analysis processing, the problem analysis model is also called a feedback adjustment model, the terminal analyzes through artificial and AI (artificial intelligence) models, and the analysis includes the type of the work order, the processing time of the work order and the evaluation of the feedback, when the analysis is completed, the feedback time is identified, so as to obtain the problem analysis result of the work order processing scheme;The terminal determines a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result; the terminal adjusts the to-be-adjusted scheme parameter in the work order processing scheme to formulate a first adjustment scheme, a second adjustment scheme and a third adjustment scheme, wherein the first adjustment scheme has the highest adjustment rate and is adjusted in a short time, and the third adjustment scheme has the lowest adjustment rate and the lowest adjustment time rate, so as to obtain the target work order processing scheme of the abnormal work order.

[0094] In the above multi-agent processing method for abnormal work orders, the abnormal work order is obtained, feature extraction processing is performed on the abnormal work order to obtain work order features; key features are extracted from the work order features, and a feature database is constructed based on the key features; a target processing information type is determined according to the feature database, and a work order processing scheme for the abnormal work order is generated based on the target processing information type; abnormal problem information generated in the execution process of the work order processing scheme is obtained, and execution process data of the work order processing scheme is obtained through an implementation monitoring model, backup processing is performed on the execution process data to obtain process backup data; the abnormal problem information and the process backup data are input into a problem analysis model for problem analysis processing to obtain a problem analysis result of the work order processing scheme; a to-be-adjusted scheme parameter in the work order processing scheme is determined according to the problem analysis result; adjustment processing is performed on the to-be-adjusted scheme parameter in the work order processing scheme to obtain a target work order processing scheme of the abnormal work order. This scheme extracts features from the abnormal work order to obtain work order features, and extracts key features from the work order features to construct a feature database, which is conducive to accurately identifying and structuring key information of abnormal work orders, thereby facilitating rapid matching of processing strategies; by determining a target processing information type according to the feature database and generating a work order processing scheme, it is conducive to realizing automatic generation of the abnormal work order processing scheme, thereby reducing the time cost of manually formulating the scheme; by obtaining execution process data of the work order processing scheme through an implementation monitoring model and performing backup processing, it is conducive to comprehensively mastering the real-time state and historical record of scheme execution, thereby providing complete data support for subsequent problem analysis; by inputting the abnormal problem information and the process backup data into a problem analysis model for problem analysis processing, it is conducive to accurately locating problems and deficiencies in the work order processing scheme, thereby facilitating the determination of scheme parameters that need to be adjusted; by adjusting the to-be-adjusted scheme parameter to obtain the target work order processing scheme, it is conducive to optimizing the processing flow and processing strategy of the abnormal work order, thereby improving the efficiency and accuracy of abnormal work order processing.

[0095] In one exemplary embodiment, determining a target processing information type according to the feature database comprises: determining a plurality of preset processing information types; the plurality of processing information types include a centralized coordination type, a distributed coordination type and a hybrid architecture type; and determining the target processing information type from the plurality of processing information types according to the key features in the feature database.

[0096] The plurality of processing information types can be processing information types classified according to the abnormal work order processing method, and include a centralized coordination type, a distributed coordination type, and a hybrid architecture type.

[0097] The centralized coordination type can be one of the classifications of processing methods, and is suitable for small systems, such as workshop scheduling.

[0098] The distributed coordination type can be one of the classifications of processing methods, and is suitable for large and decentralized systems.

[0099] The hybrid architecture type can be one of the classifications of processing methods, and is suitable for smart regional management, with hierarchical control, balanced efficiency, and fault tolerance.

[0100] Optionally, the terminal determines a plurality of processing information types, including a centralized coordination type, a distributed coordination type, and a hybrid architecture type, wherein the centralized coordination type is suitable for small systems such as workshop scheduling, wherein the distributed coordination type is suitable for large and decentralized systems, and wherein the hybrid architecture type is suitable for smart regional management and has hierarchical control, balanced efficiency, and fault tolerance; the terminal analyzes and judges the system size, the degree of decentralization, and the application scenario of the abnormal work order according to the key features in the feature database, when the system where the abnormal work order is located is determined to be a small system, the terminal determines the target processing information type to be the centralized coordination type among the plurality of processing information types, when the system where the abnormal work order is located is determined to be a large and decentralized system, the terminal determines the target processing information type to be the distributed coordination type among the plurality of processing information types, and when the system where the abnormal work order is located is determined to be a smart regional management system, the terminal determines the target processing information type to be the hybrid architecture type among the plurality of processing information types.

[0101] The technical scheme provided by the embodiment is advantageous in that the target processing information type is determined among the plurality of processing information types according to the key features in the feature database, which is beneficial to automatically selecting the most suitable processing method type according to the actual size, the degree of decentralization, and the application scenario of the system where the abnormal work order is located, thereby improving the pertinence and adaptability of the abnormal work order processing scheme and ensuring a high degree of matching between the processing scheme and the system features.

[0102] In one exemplary embodiment, based on the target processing information type, a work order processing scheme for the abnormal work order is generated, including: determining task allocation information, conflict resolution information, and collaborative decision information based on the target processing information type; and generating the work order processing scheme based on the task allocation information, the conflict resolution information, and the collaborative decision information.

[0103] The task allocation information can be information related to task allocation and optimization determined according to the target processing information type, and the task allocation and optimization adopts a contract net protocol.

[0104] The conflict resolution information can be information related to a conflict resolution mechanism determined according to the target processing information type, the conflict resolution mechanism sets a conflict type, and the conflict type includes resource competition and target conflict, wherein the resource competition solution is a token ring processing method based on priority, and the target conflict solution is a Nash equilibrium game model.

[0105] The collaborative decision information can be information related to a collaborative decision method determined according to the target processing information type, and the collaborative decision method includes Joint Intention, Bayesian Belief Network, and Federated Learning.

[0106] Optionally, the terminal determines the task allocation information, the conflict resolution information, and the collaborative decision information based on the target processing information type, the terminal determines the task allocation information to include that the task allocation and optimization adopt a contract net protocol, the terminal determines the conflict resolution information to include that the conflict resolution mechanism sets a conflict type, and the conflict type includes resource competition and target conflict, wherein the resource competition solution is a token ring processing method based on priority, and the target conflict solution is a Nash equilibrium game model, and the terminal determines the collaborative decision information to include that the collaborative decision method includes Joint Intention, Bayesian Belief Network, and Federated Learning; and the terminal generates a work order processing scheme based on the task allocation information, the conflict resolution information, and the collaborative decision information, and the work order processing scheme unifies management of the formulated work order processing scheme through a scheme management library.

[0107] The technical scheme provided in this embodiment integrates the three key elements of task allocation, conflict resolution, and collaborative decision to form a complete processing scheme, thereby improving the systematicness and executability of the work order processing scheme.

[0108] In one exemplary embodiment, the execution process data of the work order processing scheme is obtained by implementing a monitoring model, including: monitoring the processing timeliness, processing quality, and resource efficiency of the work order processing scheme by the implementation monitoring model to obtain processing timeliness information, processing quality information, and resource efficiency information of the work order processing scheme; and determining the execution process data according to the processing timeliness information, the processing quality information, and the resource efficiency information.

[0109] The processing timeliness can be the time efficiency of the work order processing scheme in the execution process, monitored by implementing a monitoring model, and the monitoring items include the first response time, the processing stagnation duration, and the default risk.

[0110] The processing quality can be the quality level of the work order processing scheme in the execution process, monitored by implementing a monitoring model, and the monitoring items include the operation compliance, the solution effectiveness, and the customer satisfaction prediction.

[0111] The resource efficiency can be the resource use efficiency of the work order processing scheme in the execution process, monitored by implementing a monitoring model, and the monitoring dimensions include the human resource efficiency, the skill matching, and the robot efficiency.

[0112] The processing timeliness information can be information obtained after monitoring the processing timeliness of the work order processing scheme, including that the first response time is the time difference between work order creation and first operation, and the time difference is greater than 5 minutes, which is a high-risk work order, the processing stagnation duration is the duration of state update, which is greater than 30 minutes, which is a medium-risk work order, and the default risk is that the dynamic calculation of the remaining time / standard time ratio is less than 20%, which is a high default risk single.

[0113] The processing quality information can be information obtained after monitoring the processing quality of the work order processing scheme, including that the operation compliance is the NLP (natural language processing) analysis operation log matching SOP (standard operation procedure) keywords, the solution effectiveness is the associated work order reopening rate, that is, the number of new work orders for the same device within 7 days, and the customer satisfaction prediction is the analysis of the emotional tendency in the communication process.

[0114] The resource efficiency information can be information obtained after monitoring the resource efficiency of the work order processing scheme.

[0115] Optionally, the terminal monitors the processing timeliness, processing quality and resource efficiency of the work order processing scheme by implementing the monitoring model, the terminal monitors the processing timeliness, including monitoring the first response time, the monitoring method of the first response time is the time difference between work order creation and first operation, and the time difference greater than 5 minutes is a high-risk work order, monitoring the processing stagnation time, the continuous time greater than 30 minutes is a medium-risk work order, monitoring the default risk, and the proportion of the remaining time to the standard time length less than 20% is a high-default-risk single, so as to obtain the processing timeliness information of the work order processing scheme, the terminal monitors the processing quality, including monitoring the operation compliance, the operation compliance is specifically that the NLP analysis operation log matches the standard SOP keyword, monitoring the solution effectiveness, the solution effectiveness is specifically the associated work order reopening rate, that is, the number of new work orders of the same device within 7 days, monitoring the customer satisfaction prediction, the customer satisfaction prediction is specifically analyzing the sentiment tendency in the communication process, such as angry or urgent word frequency, so as to obtain the processing quality information of the work order processing scheme, the terminal monitors the resource efficiency, including monitoring the human efficiency, the human efficiency monitoring index is the number of work orders processed by each person per unit time, and the optimization time is greater than 8 orders per hour, monitoring the skill matching, the skill matching monitoring index is the difference between the work order difficulty and the skill level of the processor, and the optimization is that the difference is less than or equal to 1 level, monitoring the robot efficiency, the robot efficiency monitoring index is the RPA (robotic process automation) process execution success rate, and the optimization is greater than 98%, so as to obtain the resource efficiency information of the work order processing scheme; the terminal determines the execution process data according to the processing timeliness information (including the first response time information, the processing stagnation time information and the default risk information), the processing quality information (including the operation compliance information, the solution effectiveness information and the customer satisfaction prediction information) and the resource efficiency information (including the human efficiency information, the skill matching information and the robot efficiency information), the execution process data records the complete data of the processing timeliness, the processing quality and the resource efficiency of the work order processing scheme in the execution process.

[0116] The technical scheme provided by the embodiment, by implementing the monitoring model to monitor the processing timeliness, processing quality and resource efficiency of the work order processing scheme, obtaining the processing timeliness information, processing quality information and resource efficiency information, is beneficial to comprehensively master the execution status of the work order processing scheme from the three key dimensions of time efficiency, quality level and resource use, thereby being beneficial to obtaining complete and accurate execution process monitoring data; by determining the execution process data according to the processing timeliness information, processing quality information and resource efficiency information, it is beneficial to integrate multi-dimensional monitoring information to form structured execution process data, thereby being beneficial to providing a comprehensive and reliable data basis for subsequent problem analysis and processing.

[0117] In an exemplary embodiment, the key features are extracted from the ticket features, including: performing a multi-stage filtering process on the ticket features to obtain the key features; wherein the multi-stage filtering process includes a preliminary screening process based on a variance threshold, a feature screening process based on correlation analysis, a feature screening process based on model evaluation, and a feature verification process based on business information.

[0118] The multi-stage filtering process can be a multi-stage filtering and screening process on the ticket features, including a preliminary screening process based on a variance threshold, a feature screening process based on correlation analysis, a feature screening process based on model evaluation, and a feature verification process based on business information.

[0119] The preliminary screening process based on the variance threshold can be a first-stage screening process on the ticket features, and the method is a variance threshold. When the variance threshold is greater than 0.01, remove low-variation features.

[0120] The feature screening process based on correlation analysis can be a second-stage screening process on the ticket features, and the method is mutual information. When the mutual information is greater than 0.3, keep strong correlation features.

[0121] The feature screening process based on model evaluation can be a third-stage screening process on the ticket features, and the method is SHAP value sorting. Select Top30 (top 30) to identify high-contribution features.

[0122] The feature verification process based on business information can be a fourth-stage verification process on the ticket features, and the method is a preset rule calibration to ensure feature interpretability.

[0123] Optionally, the terminal performs multi-stage filtering processing on the ticket features, and the multi-stage filtering processing includes preliminary screening processing based on a variance threshold, feature screening processing based on correlation analysis, feature screening processing based on model evaluation, and feature verification processing based on business information. The terminal first performs the preliminary screening processing based on the variance threshold, removes low-variation features with a variance less than or equal to 0.01 by taking a variance threshold greater than 0.01 as a screening standard, then performs the feature screening processing based on the correlation analysis, retains strongly correlated features with a mutual information greater than 0.3 by taking a mutual information greater than 0.3 as a screening standard, then performs the feature screening processing based on the model evaluation, selects features with SHAP values ranked in the top 30 as high-contribution features by taking a SHAP value ranking method, and finally performs the feature verification processing based on the business information, ensures that the selected features are interpretable by taking an expert rule calibration method, thereby obtaining key features, including dynamic behavior features, text semantic features, topology correlation features, and graph neural network features.

[0124] The technical scheme provided by the embodiment is advantageous in performing multi-stage filtering processing on ticket features, including preliminary screening processing based on a variance threshold, feature screening processing based on correlation analysis, feature screening processing based on model evaluation, and feature verification processing based on business information, thereby facilitating layer-by-layer screening and fine filtering of ticket features from multiple dimensions, and facilitating accurate identification and extraction of key features most valuable for abnormal ticket processing.

[0125] In one exemplary embodiment, according to the problem analysis result, a to-be-adjusted scheme parameter in the ticket processing scheme is determined, including: determining an adjustment level of the ticket processing scheme according to the problem analysis result; the adjustment level includes a first adjustment level, a second adjustment level, or a third adjustment level; and determining the to-be-adjusted scheme parameter in the ticket processing scheme according to the problem analysis result and the adjustment level.

[0126] The adjustment level can be a level classification of adjusting the ticket processing scheme according to the problem analysis result, including a first adjustment level, a second adjustment level, or a third adjustment level.

[0127] The first adjustment level can be a level with the highest adjustment rate in the adjustment level, corresponding to a first adjustment scheme, and the adjustment level is adjusted in a short time.

[0128] The second adjustment level can be a level with a middle adjustment rate in the adjustment level, corresponding to a second adjustment scheme.

[0129] The third adjustment level can be a level with the lowest adjustment rate in the adjustment levels, and the adjustment level corresponds to the third adjustment scheme, and the adjustment level has the lowest adjustment time rate.

[0130] Optionally, the terminal evaluates and judges the problem severity of the problem of the work order processing scheme according to the problem analysis result, determines an adjustment level of the work order processing scheme, the adjustment level includes a first adjustment level, a second adjustment level or a third adjustment level, the first adjustment level corresponds to the highest adjustment rate and adjustment in a short time, the second adjustment level corresponds to the middle adjustment rate, and the third adjustment level corresponds to the lowest adjustment rate and the lowest adjustment time rate; the terminal determines an adjusted scheme parameter in the work order processing scheme according to the problem analysis result and the adjustment level, when the adjustment level is the first adjustment level, the terminal determines a key scheme parameter that needs to be adjusted immediately as the adjusted scheme parameter, when the adjustment level is the second adjustment level, the terminal determines an important scheme parameter that needs to be adjusted in a medium time as the adjusted scheme parameter, and when the adjustment level is the third adjustment level, the terminal determines a secondary scheme parameter that needs to be adjusted in a long time as the adjusted scheme parameter.

[0131] The technical scheme provided in the embodiment is advantageous in determining the adjusted scheme parameter in the work order processing scheme according to the problem analysis result and the adjustment level, so as to facilitate adjustment of scheme parameters with corresponding emergency and importance for different adjustment levels, thereby improving the pertinence and timeliness of scheme adjustment and ensuring that key problems are processed in time.

[0132] The following is an embodiment for explaining the multi-agent processing method for abnormal work orders provided in the application.

[0133] At present, multi-agent processing is a core field of distributed artificial intelligence and complex system research, and aims to solve complex problems that cannot be processed by a single agent through cooperation, competition or game of multiple agents. The existing multi-agent processing method needs to extract features when processing abnormal work orders, and the existing extraction method cannot limit the features in advance, so the accuracy and obviousness of the extracted features cannot be ensured in the later stage, and the scheme cannot be monitored and stored in real time in the processing process, so the problems in the implementation process cannot be solved in the later stage, and the implementation scheme cannot be adjusted.

[0134] Embodiment one:

[0135] Reference Figure 2 The embodiment provides a multi-agent processing method and system for abnormal work orders, including the following steps:

[0136] Step one (S1): First, the characteristics of the abnormal work order are extracted, and the source and type of the characteristics are analyzed during the extraction process. After the feature extraction is completed, the characteristics are analyzed and processed, and a feature database is formed after the analysis and processing.

[0137] Step two (S2): The processing methods are classified, and the processing scheme is developed after the classification is completed. A scheme management library is established to manage the developed scheme uniformly, and the problems generated during the implementation of the scheme are collected to form a problem database.

[0138] Step three (S3): An implementation monitoring model is established to monitor the processing scheme, and backup processing is performed during the process, and a feedback adjustment model is developed to analyze and process the collected problems again, and the processing scheme is adjusted after the analysis and processing.

[0139] The source and type of the characteristics in step one are as follows:

[0140] (1) Work order basic attributes, feature examples include work order type, urgency, and submission channel;

[0141] (2) Equipment / asset information, feature examples include equipment model, service length, and historical failure times;

[0142] (3) Time series behavior data, feature examples include alarm frequency, number of similar work orders in the last hour, and processing time trend;

[0143] (4) Text description features, feature examples include fault phenomenon description and operation log keywords;

[0144] (5) Processing process features, feature examples include the number of transfers and the skill matching degree of the processing person;

[0145] (6) Environment-related features, feature examples include temperature and humidity, power grid fluctuations, and device status in the same area.

[0146] In step one, key features are established, and include dynamic behavior features, text semantic features, topology association features, and graph neural network features. The features are selected and optimized, and a multi-stage filtering strategy is used, including preliminary screening, correlation analysis, model evaluation, and business experience. The preliminary screening method is a method threshold (>0.01) (remove low-variation features), the correlation analysis method is mutual information (MI) >0.3 (retain strongly correlated features), the model evaluation is SHAP value sorting (Top30) (identify high-contribution features), and the business verification is expert rule calibration (ensure feature interpretability).

[0147] and adopt automation feature engineering tools, and include a framework of automation feature engineering: automatically generate deep features and Python libraries for time series data feature extraction: batch extract 450+ time series features.

[0148] The feature database setting in step one includes feature types and storage schemes, specifically real-time features, near-line features, and historical features. The real-time feature storage scheme is a Redis time series database (access delay <10ms), the near-line feature storage scheme is a distributed real-time analytical database (access delay 100ms-1s), and the historical feature storage scheme is a data lake storage framework (access delay >1s).

[0149] The classification of processing methods in step two is divided into centralized coordination method, distributed collaboration and hybrid architecture. The centralized coordination method is suitable for small systems (such as workshop scheduling), the distributed collaboration is suitable for large and scattered systems, and the hybrid architecture is suitable for smart regional management (with hierarchical control, balanced efficiency and fault tolerance).

[0150] In step two, key processing technologies are set, including task allocation and optimization, conflict resolution mechanism and collaborative decision-making method. The task allocation and optimization adopts the contract net protocol, the conflict resolution mechanism sets the conflict types, including resource competition and target conflict, the resource competition solution is the token ring processing method based on priority, the target conflict solution is the Nash equilibrium game model, and the collaborative decision-making method includes joint intent, Bayesian belief network and federated learning.

[0151] The implementation of the monitoring model in step three is specifically constructed as follows:

[0152] (1) The processing timeliness monitoring is the monitoring method of the first response time, and the first response time monitoring method is specifically the time difference between work order creation and the first operation, and the time difference >5 minutes is a high-risk work order. When the processing stagnation time is too long and the state update duration is >30 minutes, it is a medium-risk work order. When the dynamically calculated remaining time / standard time ratio <20%, it is a high default risk single.

[0153] (2) The monitoring points of quality compliance monitoring are operation compliance, solution effectiveness and customer satisfaction prediction. The operation compliance is specifically NLP analysis of operation logs, matching standard SOP keywords. The solution effectiveness is specifically associated with work order reopening rate (new work order within 7 days for the same device). The customer satisfaction prediction is specifically the analysis of emotional tendency (angry / urgent word frequency) in the communication process.

[0154] (3) monitoring the resource performance, and the monitoring dimensions are respectively the human performance, the skill matching and the robot performance, wherein the human performance monitoring index is the number of handled work orders per unit time per capita, and the optimization time is >8 orders / hour, wherein the skill matching monitoring index is the difference between the work order difficulty and the skill level of the handler, and the optimization is that the difference is ≤1 level, and wherein the robot performance monitoring index is the RPA process execution success rate, and the optimization is >98%.

[0155] In step three, a smart early warning strategy classification is established, and is divided into first, second and third levels, wherein the first level condition is that the SLA (Service Level Agreement) remaining <10%+ customer negative emotion, and the response mode is telephone notification of the responsible person+automatic upgrade, wherein the second level condition is that the key step is overdue for one hour, and the response mode is to notify the handler, and wherein the third level condition is that the knowledge base matching degree <60%, and the response mode is to recommend similar solutions.

[0156] In step three, the feedback adjustment model is specifically constructed as follows:

[0157] (1) The data collected by feedback is analyzed by artificial and AI model, and the analysis includes the type of work order, the processing time of work order and the evaluation of feedback, and after the analysis is completed, the feedback time is marked;

[0158] (2) After the analysis is completed, the first, second and third adjustment schemes are formulated, wherein the first adjustment scheme has the highest adjustment rate and adjusts in a short time, and the third adjustment scheme has the lowest adjustment rate and the lowest adjustment time rate.

[0159] Embodiment two:

[0160] A multi-agent processing system for abnormal work orders, comprising a feature extraction unit, a classification processing unit and a monitoring feedback unit;

[0161] The feature extraction unit first extracts the features of the abnormal work orders, analyzes the sources and types of the features during the extraction process, and after the feature extraction is completed, the features are analyzed and processed, and a feature database is formed after the analysis and processing.

[0162] The classification processing unit classifies the processing methods, formulates a processing scheme after the classification is completed, sets up a scheme management library to uniformly manage the formulated scheme, and collects problems generated during the implementation of the scheme to form a problem database;

[0163] The monitoring feedback unit sets up an implementation monitoring model to monitor the processing scheme, and performs backup processing during the recording process, and formulates a feedback adjustment model to analyze and process the collected problems again, and adjusts the processing scheme after the analysis and processing.

[0164] The technical scheme provided by the embodiment can improve the analysis and judgment efficiency, increase the accuracy and obviousness of the extracted features, and can also monitor in real time during processing, so as to extract data during implementation, and adjust the implementation scheme after analysis, thereby ensuring the efficiency and accuracy of the implementation scheme during implementation.

[0165] It should be understood that, although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or steps or stages in other steps.

[0166] Based on the same inventive concept, the embodiment of the present application also provides an abnormal work order oriented multi-agent processing device for implementing the abnormal work order oriented multi-agent processing method described above. The problem solving implementation scheme provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more abnormal work order oriented multi-agent processing device embodiments provided below can refer to the limitations of the abnormal work order oriented multi-agent processing method in the above, which will not be repeated here.

[0167] In one exemplary embodiment, as shown in Figure 3 An abnormal work order oriented multi-agent processing device is provided, which can include:

[0168] The work order acquisition module 301 is configured to acquire an abnormal work order, perform feature extraction processing on the abnormal work order, and obtain work order features.

[0169] The feature construction module 302 is configured to extract key features from the work order features, and construct a feature database based on the key features.

[0170] The scheme generation module 303 is configured to determine a target processing information type according to the feature database, and generate a work order processing scheme for the abnormal work order based on the target processing information type.

[0171] The backup processing module 304 is configured to acquire abnormal problem information generated in an execution process of the work order processing scheme, acquire execution process data of the work order processing scheme by implementing a monitoring model, perform backup processing on the execution process data, and obtain process backup data.

[0172] The data input module 305 is configured to input the abnormal problem information and the process backup data into a problem analysis model to perform problem analysis processing, and obtain a problem analysis result of the work order processing scheme.

[0173] The parameter determination module 306 is configured to determine a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result.

[0174] The parameter processing module 307 is configured to perform adjustment processing on the to-be-adjusted scheme parameter in the work order processing scheme, and obtain a target work order processing scheme of the abnormal work order.

[0175] In an example embodiment, the scheme generation module 303 is further configured to determine a plurality of preset processing information types, wherein the plurality of processing information types include a centralized coordination type, a distributed coordination type, and a hybrid architecture type; and determine a target processing information type from the plurality of processing information types according to the key features in the feature database.

[0176] In an example embodiment, the scheme generation module 303 is further configured to determine task allocation information, conflict resolution information, and collaborative decision-making information based on the target processing information type; and generate the work order processing scheme based on the task allocation information, the conflict resolution information, and the collaborative decision-making information.

[0177] In an example embodiment, the backup processing module 304 is further configured to monitor processing timeliness, processing quality, and resource efficiency of the work order processing scheme by implementing the monitoring model, and obtain processing timeliness information, processing quality information, and resource efficiency information of the work order processing scheme; and determine the execution process data according to the processing timeliness information, the processing quality information, and the resource efficiency information.

[0178] In an example embodiment, the feature construction module 302 is further configured to perform multi-stage filtering processing on the work order features, and obtain the key features; wherein the multi-stage filtering processing includes preliminary screening processing based on a variance threshold, feature screening processing based on correlation analysis, feature screening processing based on model evaluation, and feature verification processing based on business information.

[0179] In an example embodiment, the parameter determination module 306 is further configured to determine an adjustment level of the work order processing scheme according to the problem analysis result; the adjustment level includes a first adjustment level, a second adjustment level, or a third adjustment level; and determine the to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result and the adjustment level.

[0180] Each module in the above multi-agent processing device for exception work order can be implemented by software, hardware and combinations thereof, in whole or in part. Each module can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform operations corresponding to each module.

[0181] In an exemplary embodiment, a computer device, which can be a terminal, is provided, and an internal structure diagram of the computer device can be as shown in Figure 4 The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to perform wired or wireless communication with external terminals, and the wireless communication can be achieved through WIFI, mobile cellular network, NFC (Near Field Communication) or other technologies. The computer program is executed by the processor to implement a multi-agent processing method for exception work order. The display unit of the computer device is configured to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0182] Those skilled in the art can understand that Figure 4 The structure shown in the above

[0183] In an exemplary embodiment, a computer device is also provided, which includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps in each method embodiment.

[0184] In an exemplary embodiment, a computer readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, implements the steps of any of the above method embodiments.

[0185] In an exemplary embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of any of the above method embodiments.

[0186] A person of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium and can include the processes of the above method embodiments when executed. Any reference to a memory, database, or other medium in the embodiments provided in the present application can include at least one of a non-volatile and volatile memory. The non-volatile memory can include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical storage, a high-density embedded non-volatile memory, a resistive memory (ReRAM), a magnetoresistive random access memory (MRAM), a ferroelectric random access memory (FRAM), a phase change memory (PCM), a graphene memory, etc. The volatile memory can include a random access memory (RAM) or an external cache memory, etc. As an illustration but not as a limitation, the RAM can be in various forms such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0187] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combinations of the technical features do not contradict each other, they should be considered as within the scope of the present disclosure.

[0188] The above-described embodiments are merely illustrative of several embodiments of the present application, which are described in more detail and in a specific manner, but should not be construed as limiting the scope of the patent of the present application. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. An abnormal work order oriented multi-agent processing method, characterized in that, The method comprises: obtaining an abnormal work order, performing feature extraction processing on the abnormal work order to obtain work order features; extracting key features from the work order features, and constructing a feature database based on the key features; determining a target processing information type according to the feature database, and generating a work order processing scheme for the abnormal work order based on the target processing information type; obtaining abnormal problem information generated in the execution process of the work order processing scheme, and obtaining execution process data of the work order processing scheme by implementing a monitoring model, performing backup processing on the execution process data to obtain process backup data; inputting the abnormal problem information and the process backup data into a problem analysis model for problem analysis processing to obtain a problem analysis result of the work order processing scheme; determining a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result; adjusting the to-be-adjusted scheme parameter in the work order processing scheme to obtain a target work order processing scheme for the abnormal work order.

2. The method of claim 1, wherein, The determination of the target processing information type according to the feature database comprises: determining a plurality of preset processing information types; the plurality of processing information types include a centralized coordination type, a distributed coordination type, and a hybrid architecture type; determining the target processing information type from the plurality of processing information types according to the key features in the feature database.

3. The method of claim 1, wherein, The generation of the work order processing scheme for the abnormal work order based on the target processing information type comprises: determining task allocation information, conflict resolution information, and collaborative decision information based on the target processing information type; generating the work order processing scheme based on the task allocation information, the conflict resolution information, and the collaborative decision information.

4. The method of claim 1, wherein, The obtaining of the execution process data of the work order processing scheme by implementing the monitoring model comprises: monitoring the processing timeliness, processing quality, and resource efficiency of the work order processing scheme by the implementation monitoring model to obtain processing timeliness information, processing quality information, and resource efficiency information of the work order processing scheme; determining the execution process data according to the processing timeliness information, the processing quality information, and the resource efficiency information.

5. The method of claim 1, wherein, The extraction of the key features from the work order features comprises: performing multi-stage filtering processing on the work order features to obtain the key features; wherein the multi-stage filtering processing comprises preliminary screening processing based on a variance threshold, feature screening processing based on correlation analysis, feature screening processing based on model evaluation, and feature verification processing based on business information.

6. The method according to any one of claims 1 to 5, characterized in that, The determination of the to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result comprises: determining an adjustment level of the work order processing scheme according to the problem analysis result; the adjustment level includes a first adjustment level, a second adjustment level, or a third adjustment level; determining the to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result and the adjustment level.

7. An anomaly ticket oriented multi-agent processing apparatus, characterized by comprising: The device comprises: a work order acquisition module configured to obtain an abnormal work order, perform feature extraction processing on the abnormal work order, and obtain work order features; The feature construction module is configured to extract key features from the work order features, and construct a feature database based on the key features. The scheme generation module is configured to determine a target processing information type according to the feature database, and generate a work order processing scheme of the abnormal work order based on the target processing information type. The backup processing module is configured to acquire abnormal problem information generated in an execution process of the work order processing scheme, acquire execution process data of the work order processing scheme by implementing a monitoring model, perform backup processing on the execution process data, and obtain process backup data. The data input module is configured to input the abnormal problem information and the process backup data into a problem analysis model to perform problem analysis processing, and obtain a problem analysis result of the work order processing scheme. The parameter determination module is configured to determine a to-be-adjusted scheme parameter in the work order processing scheme according to the problem analysis result. The parameter processing module is configured to perform adjustment processing on the to-be-adjusted scheme parameter in the work order processing scheme, and obtain a target work order processing scheme of the abnormal work order. 8.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-7. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6.

Images