Intelligent management system, method, device and medium for power plant operation and maintenance personnel

Abstract

The application discloses a kind of power plant operation and maintenance personnel's intelligent management system, method, equipment and medium, system includes: data acquisition module, knowledge base module, digital person engine module and central processing module;Central processing module includes: stage determination unit, induction training unit, on-the-job guidance unit and knowledge extraction unit.New employee stage, growth period stage and expert stage respectively to the operation and maintenance personnel in corresponding knowledge support and service are provided, not only realize the digitization inheritance of power plant relevant knowledge and improve the professional ability and operation and maintenance standardization of operation and maintenance personnel.

Description

Technical Field

[0001] This invention relates to the field of power plant management technology, and in particular to an intelligent management system, method, equipment and medium for power plant operation and maintenance personnel. Background Technology

[0002] As large-scale industrial facilities, the safe and stable operation of power plants depends heavily on the professional skills and practical experience of operation and maintenance personnel. The complexity and high risk of operation and maintenance work place stringent requirements on personnel's knowledge reserves, operational proficiency, and emergency response capabilities.

[0003] However, the current knowledge management and personnel training system in the power plant operation and maintenance field faces many prominent problems. On the one hand, a large number of experienced expert personnel are about to retire, and the tacit knowledge they have accumulated is difficult to fully pass on through traditional mentorship or operation manuals. This leads to the risk of a knowledge cliff, where companies face the loss of core skills. This tacit knowledge covers key aspects such as equipment failure prediction, emergency decision-making, and handling of non-standard operating conditions, and is the core asset for ensuring the efficient operation and maintenance work. On the other hand, traditional training models have obvious limitations. Classroom teaching focuses on theory and is disconnected from practical operation. The quality of mentorship depends on individual factors and is difficult to standardize. Furthermore, high-risk, low-frequency emergency response procedures cannot be practiced in real-world scenarios, resulting in slow skill improvement for new employees and making them easily helpless in the face of emergencies. At the same time, existing knowledge management systems are mostly based on document storage and are essentially passive query databases. They cannot provide accurate knowledge delivery based on the real-time work scenarios of operation and maintenance personnel, and the correlation between knowledge and actual application scenarios is insufficient, making it difficult to play a supporting role at key operational nodes. Furthermore, the growth path of operations and maintenance personnel from onboarding to becoming experts lacks an objective and quantitative assessment and tracking system. Companies cannot accurately grasp the skill gaps of their employees, making it difficult to develop personalized training programs, resulting in low efficiency in personnel growth.

[0004] In existing technologies, for example, Chinese patent CN117808945A discloses a digital human generation system based on a large-scale pre-trained language model. However, its training concentration is relatively shallow. In the specific training process, it cannot make the implicit knowledge of experts explicit, the training process contextualized, the knowledge service contextualized, or the personnel growth path clear. It is not suitable for training power plant operation and maintenance personnel and cannot systematically solve the technical problems of difficult knowledge transfer, low training efficiency, passive knowledge service, and unclear growth path.

[0005] In view of the above, this application is hereby submitted. Summary of the Invention

[0006] This application provides an intelligent management system, method, equipment, and medium for power plant operation and maintenance personnel, which solves the technical problem of difficult knowledge transfer among power plant operation and maintenance personnel in the prior art, and achieves the technical effect of efficiently improving the knowledge transfer of power plant operation and maintenance personnel.

[0007] Firstly, this application provides an intelligent management system for power plant operation and maintenance personnel, comprising: The data acquisition module is used to acquire operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments. The knowledge base module is used to store structured knowledge, unstructured knowledge, and tacit knowledge of the power plant. The digital human engine module is used to generate and drive digital human avatars, which are used to interact with maintenance personnel based on their voice interaction data. The central processing module communicates with the data acquisition module, knowledge base module, and digital human engine module, and includes: The stage determination unit is used to determine the training stage of the operation and maintenance personnel based on their preset information and historical operation data. The training stage includes the new employee stage, the growth stage, and the expert stage. The onboarding training unit is designed to provide training and assessment of standard operating procedures for power plants to maintenance personnel who are identified as new employees, using digital avatars in physical and / or virtual environments. The on-the-job coaching unit is used to respond to operations and maintenance personnel who are determined to be in the growth stage. It analyzes the matching degree between the operational behavior data of operations and maintenance personnel and standard operation data, and provides auxiliary guidance by controlling the digital human image during non-interference times. The knowledge extraction unit is used to respond to the fact that the operation and maintenance personnel are judged to be in the expert stage, identify the optimized operation data in the operation behavior data that has a positive deviation from the standard operation data, and control the digital human image to interact with the operation and maintenance personnel through language in order to obtain the implicit experience knowledge corresponding to the optimized operation data.

[0008] In some embodiments of this application, based on the foregoing scheme, the on-the-job coaching unit includes a deviation quantification analysis module, used for: Based on the operational behavior data, the actual operation sequence of the operation and maintenance personnel is obtained. The normalized path distance between the actual operation sequence and the standard operation sequence in the standard operation data is determined by the dynamic time warping algorithm. The normalized path distance is used to quantify the matching degree between the operation and maintenance personnel's operational behavior data and the standard operation data. When the distance of the normalized path exceeds the preset slight deviation threshold but is lower than the preset serious violation threshold, an auxiliary guidance prompt signal is generated. The auxiliary guidance prompt signal is used to control the digital human image to provide auxiliary guidance during non-interference times.

[0009] In some embodiments of this application, based on the foregoing scheme, the knowledge extraction unit includes a positive deviation identification module and a tacit knowledge conversion module; The positive deviation identification module is used to obtain the operation result, actual time consumption, and deviation degree between the actual operation and the standard operation when the operation behavior data does not match the standard operation data. Based on the operation result, actual time consumption, and deviation degree, it determines whether there is a positive deviation between the operation behavior data and the standard operation data, and identifies and optimizes the operation data based on the judgment result. The tacit knowledge conversion module is used to control the digital human image to interact with maintenance personnel in the event of a positive deviation between the operational behavior data and the standard operational data. It converts the voice interaction data of the maintenance personnel into a structured knowledge graph as tacit experience knowledge, sends the tacit experience knowledge to the knowledge base module for storage, and generates an update signal. The update signal is used to update the standard operating procedures called by the onboarding training unit based on the tacit experience knowledge.

[0010] In some embodiments of this application, based on the foregoing scheme, the onboarding training unit includes a performance evaluation module for: Based on the accuracy of operations, compliance of processes, and time spent on operations and maintenance personnel during training in physical and / or virtual environments, an assessment report corresponding to the operations and maintenance personnel is generated.

[0011] In some embodiments of this application, based on the foregoing scheme, the data acquisition module includes: The visual acquisition device includes at least one of the following: a fixed camera installed in the physical environment, an augmented reality (AR) device worn by maintenance personnel, smart inspection glasses, and a virtual camera installed in a simulated operating room corresponding to the virtual environment; the visual acquisition device is used to acquire the body posture data, hand operation data, and interaction data with the corresponding device of the maintenance personnel, as operational behavior data. Microphone arrays are used to acquire voice interaction data between maintenance personnel and the digital avatar; The system log interface is used to access and parse the operation records of maintenance personnel in the production system to obtain the operation log data of the production system.

[0012] In some embodiments of this application, based on the foregoing scheme, the knowledge base module includes: An explicit knowledge base is used to store structured knowledge, which includes at least one of the following: equipment ledgers, standard operating procedure documents, historical fault reports, and equipment drawings. Implicit knowledge base, used to store implicit experiential knowledge; An unstructured knowledge base is used to store unstructured knowledge, which includes at least one of the following: basic information of operations and maintenance personnel, job title, length of service, historical training and assessment records, evaluation reports, and on-the-job performance.

[0013] In some embodiments of this application, based on the aforementioned scheme, the stage determination unit includes an operation and maintenance personnel skill proficiency assessment model. The operation and maintenance personnel skill proficiency assessment model is used to represent the preset information and historical operation data of the operation and maintenance personnel as multi-dimensional feature vectors and calculate the corresponding comprehensive proficiency score, and determine the trainee stage of the operation and maintenance personnel based on the comprehensive proficiency score.

[0014] Secondly, this application provides an intelligent management method for power plant operation and maintenance personnel, including: Acquire operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments; Generate and drive digital human figures, which are used to interact with maintenance personnel based on their voice interaction data; Based on the preset information and historical operation data of the operation and maintenance personnel, the student stage of the operation and maintenance personnel is determined. The student stage includes the new employee stage, the growth stage, and the expert stage. When maintenance personnel are identified as new employees, digital avatars can be used to provide them with training and assessment on standard operating procedures for power plants in physical and / or virtual environments. When maintenance personnel are identified as being in the growth stage, analyze the matching degree between the maintenance personnel's operational behavior data and standard operation data, and control the digital human image to provide auxiliary guidance during non-interference times; When maintenance personnel are deemed to be at the expert level, the system identifies optimized operational data that deviates positively from standard operational data. It then controls a digital human to interact with the maintenance personnel verbally to obtain implicit experiential knowledge corresponding to the optimized operational data.

[0015] Thirdly, this application provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to execute an intelligent management method for power plant maintenance personnel, as provided in the second aspect.

[0016] Fourthly, this application provides a non-transitory computer-readable storage medium that, when the instructions in the storage medium are executed by the processor of an electronic device, enables the electronic device to perform an intelligent management method for power plant operation and maintenance personnel as provided in the second aspect.

[0017] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By identifying positive and negative deviations in expert operational behavior data, we proactively mine and extract experts' tacit experience, transforming it into a structured knowledge graph for storage. This effectively solves the problem of tacit knowledge being lost with the retirement of personnel, and realizes the digital inheritance and accumulation of valuable corporate knowledge.

[0018] 2. For new employees, we provide immersive standard operating procedure training in virtual reality or physical environments. We use performance evaluation modules to quantitatively assess the accuracy of operations, compliance of procedures, and operation time. This breaks through the training bottleneck of high-risk, low-frequency operations that cannot be practiced in real-world scenarios. It helps new employees quickly master the essential skills for their positions and significantly improves training effectiveness and safety.

[0019] 3. For employees in the growth stage, an on-the-job coaching model is adopted. The dynamic time warping algorithm is used to quantitatively analyze the deviation between actual operation and standard operation. At non-interference times, a digital human image is used to provide auxiliary guidance, realizing real-time, non-intrusive skill assistance for operation and maintenance personnel. This helps employees standardize their operation behavior, make up for skill gaps, and steadily improve their professional capabilities and operation and maintenance standardization.

[0020] 4. Based on the preset information and historical operation data of maintenance personnel, a skill proficiency assessment model is established. Through multi-dimensional feature vectors and comprehensive proficiency scores, the training stage is dynamically determined. This can objectively and clearly divide the personnel growth path, break through the passive query limitations of traditional knowledge management systems, realize the orderly management and precise retrieval of knowledge, and provide timely and effective knowledge support for maintenance personnel at all stages.

[0021] 5. Optimize the interaction experience between maintenance personnel and the system. By using a digital human-like avatar adapted to the power plant scenario to achieve natural language interaction, combined with functions such as voice recognition and intent understanding, the training, tutoring, and knowledge query services will be more immersive and professional, thereby improving employees' willingness to use the system and work collaboration efficiency. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A schematic diagram of the structure of an intelligent management system for power plant operation and maintenance personnel provided in an embodiment of this application; Figure 2 A flowchart illustrating an intelligent management method for power plant operation and maintenance personnel provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] This application provides an intelligent management system, method, equipment, and medium for power plant operation and maintenance personnel, solving the technical problem of knowledge transfer difficulties among power plant operation and maintenance personnel in the prior art.

[0025] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0026] Example 1 This application provides embodiments such as Figure 1 The intelligent management system shown includes: a data acquisition module, a knowledge base module, a digital human engine module, and a central processing module.

[0027] The data acquisition module is used to acquire operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments.

[0028] The data acquisition module includes: a visual acquisition device, a microphone array, and a system log interface.

[0029] The visual acquisition device includes at least one of the following: a fixed camera installed in the physical environment, an augmented reality (AR) device worn by maintenance personnel, smart inspection glasses, and a virtual camera installed in a simulated operating room corresponding to the virtual environment; the visual acquisition device is used to acquire the body posture data, hand operation data, and interaction data with the corresponding device of the maintenance personnel as operational behavior data.

[0030] Microphone arrays are used to acquire voice interaction data between operations and maintenance personnel and the digital avatar. For example, they can collect natural language dialogues between operations and maintenance personnel and the digital avatar, or record the thought process of experts performing operations.

[0031] The system log interface is used to access and parse the operation records of maintenance personnel in the production system (such as distributed control system, plant-level monitoring information system, operation ticket system) to obtain the operation log data of the production system.

[0032] Furthermore, the data acquisition module also includes a VR / AR device interface, used to collect high-precision interactive data such as trainees' head orientation and controller operation during virtual training.

[0033] The knowledge base module is used to store structured knowledge, unstructured knowledge, and tacit experiential knowledge of the power plant. The knowledge base module includes: an explicit knowledge base, a tacit knowledge base, and an unstructured knowledge base.

[0034] The explicit knowledge base is used to store structured knowledge, which includes at least one of the following: equipment ledger, standard operating procedure documents, historical fault reports, and equipment drawings.

[0035] Implicit knowledge bases are used to store implicit experiential knowledge. For example, decision logic stored in the form of IF-THEN rules (if the inlet pressure of pump A is lower than X and valve B is open, then check whether pipeline C is blocked first), or the relationships between equipment, faults, causes, and phenomena stored in the form of knowledge graphs.

[0036] An unstructured knowledge base is used to store unstructured knowledge, which includes at least one of the following: basic information of operations and maintenance personnel, job title, length of service, historical training and assessment records, evaluation reports, and on-the-job performance.

[0037] The digital human engine module generates and drives a digital human avatar, which interacts with maintenance personnel based on their voice interaction data. It includes: The image rendering engine is responsible for rendering high-fidelity 3D models of digital humans in real time. The text-to-speech (TTS) and audio-visual recognition (ASR) engine is responsible for smoothly synthesizing the system's speech into speech and accurately recognizing the user's speech into text. The Natural Language Understanding (NLU) and Generation (NLG) engine is responsible for understanding the user's question intent and generating logical and naturally linguistic answers.

[0038] The central processing module communicates with the data acquisition module, knowledge base module, and digital human engine module. The central processing module includes: a stage determination unit, an onboarding training unit, an on-the-job coaching unit, and a knowledge extraction unit.

[0039] The stage determination unit is used to determine the trainee stage of the operation and maintenance personnel based on their preset information and historical operation data. The trainee stage includes the new employee stage, the growth stage, and the expert stage.

[0040] The stage determination unit includes the operation and maintenance personnel skill proficiency assessment model. The operation and maintenance personnel skill proficiency assessment model is used to represent the preset information and historical operation data of operation and maintenance personnel as multi-dimensional feature vectors and calculate the corresponding comprehensive proficiency score, and determine the trainee stage of operation and maintenance personnel based on the comprehensive proficiency score.

[0041] For example, when an operations and maintenance personnel logs into the system, the unit first retrieves their information from the personnel database and automatically determines their current career lifecycle stage based on a set of preset rules or models. For instance, the rules could be set as follows: less than 1 year of service is the new employee stage; 1 year to less than 5 years of service is the growth stage; and 5 years of service or more and having passed advanced skills certification is the expert stage.

[0042] The assessment is based on a dynamically updated skills proficiency evaluation model for operations and maintenance personnel. This model represents each operations and maintenance personnel as a multi-dimensional feature vector (Vop) and calculates their comprehensive proficiency score (Sprof). The calculation method is defined by the following formula:

[0043] in: It is an N-dimensional feature vector describing the capabilities of operations and maintenance personnel, and is a quantitative representation of their capabilities.

[0044] Fi represents the i-th normalized capability characteristic factor. For example: F1 is the seniority factor, F2 is the quantitative score of the advanced certification, F3 is the average assessment score of historical virtual training, and F4 is the reciprocal of the frequency of risky operations recorded by the system during on-the-job training, etc.

[0045] ωi is the weight coefficient corresponding to the i-th feature, and ∑ωi=1. These weights can be set by combining the analytic hierarchy process (AHP) with expert opinions, or they can be dynamically optimized by machine learning models to adapt to the core competency requirements of different positions.

[0046] Through the above model, the system can more accurately and dynamically divide the growth stages of individuals, thereby providing a scientific basis for activating the most suitable mentor mode.

[0047] The onboarding training unit is designed to respond to maintenance personnel who are identified as new employees by providing them with training and assessment on standard operating procedures for power plants in physical and / or virtual environments using digital avatars.

[0048] The onboarding training unit includes a performance evaluation module, which is used to generate an evaluation report corresponding to the operations and maintenance personnel based on their operational accuracy, process compliance, and operation time during the training process in physical and / or virtual environments.

[0049] For example, when an operations and maintenance personnel is identified as a new employee, the onboarding training unit is activated. It calls upon the VR / AR device interface and the digital human engine module to create an immersive virtual training environment for the new employee. For instance, it simulates a complete startup process for a hydroelectric generator unit. The digital human avatar first demonstrates and explains the standard operation in the virtual space. Then, the new employee is required to repeat the operation independently. During the operation, the performance evaluation module assesses the accuracy and time taken for each step in real time, immediately interrupting and correcting any errors. After the training, a detailed evaluation report containing scores for each operation and analysis of weaknesses is automatically generated and stored in the employee's personal file.

[0050] The on-the-job coaching unit is used to respond to operations and maintenance personnel who are determined to be in the growth stage. It analyzes the matching degree between the operational behavior data of operations and maintenance personnel and standard operation data, and provides auxiliary guidance by controlling the digital human image during non-interference times.

[0051] The on-the-job coaching unit includes a deviation quantification analysis module, used for: Based on the operational behavior data, the actual operation sequence of the operation and maintenance personnel is obtained. The normalized path distance between the actual operation sequence and the standard operation sequence in the standard operation data is determined by the dynamic time warping algorithm. The normalized path distance is used to quantify the matching degree between the operation and maintenance personnel's operational behavior data and the standard operation data. When the distance of the normalized path exceeds the preset slight deviation threshold but is lower than the preset serious violation threshold, an auxiliary guidance prompt signal is generated. The auxiliary guidance prompt signal is used to control the digital human image to provide auxiliary guidance during non-interference times.

[0052] For example, when an operations and maintenance personnel is determined to be in the growth stage, the on-the-job coaching unit runs silently in the background. It continuously observes the employee's operational behavior in actual work through a data acquisition module. When it detects that the employee's operational sequence deviates from the standard operating procedure (SOP) while performing an operation ticket, but does not immediately pose a danger, the digital avatar will not immediately issue an alarm. Instead, at an appropriate time (such as during an operation break), it will gently remind the employee through headphones or a screen next to them: "Just a reminder, if you changed the order of steps 3 and 4, it might be more in line with the standard procedure and would also reduce an unnecessary confirmation step. This is for your reference." This approach serves a coaching purpose while protecting the employee's self-esteem.

[0053] Deviation identification is based on a quantitative calculation of the compliance of the operation sequence. Specifically, the system defines the Standard Operating Procedure (SOP) as a standard sequence of behaviors in the behavioral paradigm space. , where a nThis is a standard atomic operation. Simultaneously, the actual operation sequence of the operations personnel is obtained in real time through computer vision and system logs. .

[0054] The on-the-job coaching unit uses the Dynamic Time Warping (DTW) algorithm to calculate the warped path distance D(ASOP,Breal) between two sequences: D(ASOP,Breal) = DTW(ASOP,Breal). This algorithm can effectively calculate the minimum distance between two sequences of unequal time durations, and this distance D objectively quantifies the degree of deviation of the actual operation from the standard operation. When D exceeds a preset slight deviation threshold but is below a serious violation threshold, the system triggers the aforementioned advisory reminder.

[0055] The knowledge extraction unit is used in response to operations and maintenance personnel being assessed as being in the expert stage. It identifies optimized operational data that deviates positively from standard operational data and controls a digital avatar to interact with the personnel verbally, thereby obtaining tacit experiential knowledge corresponding to the optimized operational data. The knowledge extraction unit includes a positive deviation identification module and a tacit knowledge conversion module.

[0056] The positive deviation identification module is used to obtain the operation result, actual time consumption, and deviation degree between the actual operation and the standard operation when the operation behavior data does not match the standard operation data. Based on the operation result, actual time consumption, and deviation degree, it determines whether there is a positive deviation between the operation behavior data and the standard operation data, and identifies and optimizes the operation data based on the judgment result.

[0057] The tacit knowledge conversion module is used to control the digital human image to interact with maintenance personnel in the event of a positive deviation between the operational behavior data and the standard operational data. It converts the voice interaction data of the maintenance personnel into a structured knowledge graph as tacit experience knowledge, sends the tacit experience knowledge to the knowledge base module for storage, and generates an update signal. The update signal is used to update the standard operating procedures called by the onboarding training unit based on the tacit experience knowledge.

[0058] For example, when an operations and maintenance (O&M) personnel are identified as being at the expert level, the knowledge extraction unit continuously compares the expert's real-time operations with the standard procedures in the explicit knowledge base. When it discovers that the expert has adopted an operational method that differs significantly from the standard procedure but yields the correct result and is more efficient (i.e., a positive deviation), it determines that this is a potential and valuable tacit knowledge point. At this point, it will drive the digital human to initiate an interview via natural language afterward: "Hello, Mr. / Ms. X, I am your digital apprentice. I noticed that when you were handling the XX fault today, you did not follow the steps in the manual, but instead checked XX points first, which resulted in locating the problem more quickly. I would very much like to learn your decision-making process at that time. Could you explain it to me?" The expert's verbal response is recorded and transcribed into text, then processed by the tacit knowledge conversion module, ultimately forming a new knowledge entry and generating an update signal to proactively update the standard operating procedures corresponding to the tacit experience knowledge invoked by the onboarding training unit, thereby enriching the entire knowledge base.

[0059] The positive deviation behavior that triggers knowledge extraction is precisely determined by the Positive Deviation Behavior Recognition Model in the Positive Deviation Recognition Module. The conditions for an operation event Eop to be determined as a positive deviation event Epid can be formally described as follows:

[0060] Wherein: T op This refers to the actual time it took for the expert to complete the operation; T SOP Refers to the standard time specified in the standard operating procedure (SOP); η refers to a preset operational efficiency coefficient, such as 0.8, used to define significantly higher efficiency; R op The final result of the operation must be successful; D(ASOP,Bop) refers to the deviation between the actual operation and the standard operation calculated by the DTW algorithm. This value must be greater than a significant deviation threshold θ to ensure that the behavior is non-routine.

[0061] The system will only initiate an interview with the expert when all of the above conditions are met. This model ensures that each extraction targets truly valuable, innovative, and efficient operational experience, avoiding unnecessary disruption to the expert.

[0062] Furthermore, the system also includes a knowledge transfer unit. When an expert retires, the system uses all the data accumulated in their personnel database (including their extracted knowledge, their voice recordings of conversations with the digital avatar, their written reports, etc.) to train a personalized model of the expert. This model (for example, it can be fine-tuned based on a large language model like Qwen) aims to learn and imitate the expert's language style, logical habits, and knowledge domain. When subsequent employees encounter problems, especially those related to the retired expert's field, they can choose to "consult X's digital avatar." The digital avatar's answering style, tone, and even the cases it cites will highly mimic the expert's, making knowledge transfer more human and relatable.

[0063] Example 2 Based on the same inventive concept, embodiments of this application also provide, as follows: Figure 2 The method for intelligent management of power plant operation and maintenance personnel is shown, including steps S1-S6.

[0064] Step S1: Obtain operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments; Step S2: Generate and drive the digital human avatar, which is used to interact with the maintenance personnel based on the voice interaction data of the maintenance personnel; Step S3: Based on the preset information and historical operation data of the operation and maintenance personnel, determine the training stage of the operation and maintenance personnel. The training stage includes the new employee stage, the growth stage, and the expert stage. Step S4: When the maintenance personnel are determined to be in the new employee stage, use digital human figures to provide the maintenance personnel with training and assessment on the standard operating procedures of the power plant in the physical environment and / or virtual environment. Step S5: When the operation and maintenance personnel are determined to be in the growth stage, analyze the matching degree between the operation and maintenance personnel's operational behavior data and standard operation data, and control the digital human image to provide auxiliary guidance during non-interference times. Step S6: When the operation and maintenance personnel are determined to be in the expert stage, identify the optimized operation data in the operation behavior data that has a positive deviation from the standard operation data, and control the digital human image to interact with the operation and maintenance personnel through language in order to obtain the implicit experience knowledge corresponding to the optimized operation data.

[0065] For example, by deploying visual cameras and microphone arrays in the actual control room, key equipment area, and simulated operation room in the physical environment, the body posture, hand movements, equipment interaction process, and voice dialogue of maintenance personnel can be continuously captured. At the same time, the operation records in the power plant's existing production system (such as DCS, SIS, or operation ticket system) can be accessed and parsed through interfaces, thereby comprehensively obtaining the operation behavior data, voice interaction data, and operation log data of the production system of maintenance personnel in the physical and / or virtual environments.

[0066] By using real-time rendering technology to generate a high-fidelity 3D digital human model, the system's feedback text is smoothly synthesized into natural speech, and the user's voice commands are accurately recognized as text. At the same time, the semantic intent of the user's questions is deeply analyzed and logical natural language answers are generated, thereby generating and driving the digital human image, enabling it to conduct smooth multimodal interactions with maintenance personnel through anthropomorphic voice, actions, and expressions based on the collected voice interaction data.

[0067] The system retrieves pre-defined information (such as seniority, job title, and certificate level) and historical operation data (such as past training results and on-the-job operation records) of maintenance personnel. It then uses a pre-defined maintenance personnel skill proficiency assessment model to represent these multi-dimensional data as feature vectors and calculates a comprehensive proficiency score. Based on the numerical range of this score, the system accurately determines whether the maintenance personnel are in the new employee stage, the growth stage, or the expert stage, and dynamically updates their stage based on subsequent changes in skill assessment scores.

[0068] When determining that maintenance personnel are in the new employee stage, a virtual simulation or augmented reality training environment containing the standard operating procedures of power plants is constructed. First, the standard operation is demonstrated and explained through a digital human figure. Then, the maintenance personnel's independent operation process in the environment is monitored. Real-time quantitative evaluation is carried out based on the accuracy of their operation steps, the compliance of the process, and the operation time. If an operation error occurs, it is interrupted and corrected immediately. An evaluation report containing scores and weaknesses is generated after the training.

[0069] When it is determined that the operation and maintenance personnel are in the growth stage, the matching degree between their actual operation behavior and standard operation data is silently analyzed. Specifically, this includes constructing an actual operation sequence based on the operation behavior data, using a dynamic time warping algorithm to calculate the warped path distance between the sequence and the standard operation sequence. When the distance exceeds the preset slight deviation threshold but is lower than the serious violation threshold, it is determined that there is a non-risk deviation, and the digital human image is controlled to issue auxiliary guidance during non-interference times such as operation breaks.

[0070] When it is determined that the operation and maintenance personnel are in the expert stage, the system will identify in real time whether there are optimized operations in their operation behavior data that deviate positively from the standard operation data. That is, the operation results are successful, the actual time is significantly better than the standard time, and the operation process is significantly different. Once such optimized operations are identified, the system will control the digital human image to initiate a natural language interview to ask about the decision-making logic. The expert's answer will be parsed and transformed into a structured knowledge graph or rule entries, thereby extracting implicit experience knowledge for subsequent use.

[0071] Example 3 Based on the same inventive concept, embodiments of this application also provide an electronic device, including: Processor 31; Memory 32 is used to store executable instructions of processor 31; The processor 31 is configured to execute an intelligent management method for power plant operation and maintenance personnel as described above.

[0072] Example 4 Based on the same inventive concept, this application also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor 31 of an electronic device, enables the electronic device to implement the intelligent management method for power plant operation and maintenance personnel as described above.

[0073] Since the electronic device described in this embodiment is an electronic device used to implement the information processing method in the embodiments of this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the information processing method described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the information processing method in the embodiments of this application falls within the scope of protection of this application.

[0074] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0075] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0076] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0077] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0078] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0079] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent management system for power plant operation and maintenance personnel, characterized in that, include: The data acquisition module is used to acquire operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments. The knowledge base module is used to store structured knowledge, unstructured knowledge, and tacit knowledge of the power plant. The digital human engine module is used to generate and drive a digital human avatar, which is used to interact with maintenance personnel based on their voice interaction data. The central processing module is communicatively connected to the data acquisition module, the knowledge base module, and the digital human engine module, and includes: The stage determination unit is used to determine the training stage of the operation and maintenance personnel based on their preset information and historical operation data. The training stage includes the new employee stage, the growth stage, and the expert stage. The onboarding training unit is designed to provide training and assessment of standard operating procedures for power plants to maintenance personnel who are identified as new employees, using digital avatars in physical and / or virtual environments. The on-the-job coaching unit is used to respond to the fact that the operation and maintenance personnel are determined to be in the growth stage, analyze the matching degree between the operation and maintenance personnel's operational behavior data and standard operation data, and control the digital human image to provide auxiliary guidance when there is no interference. The knowledge extraction unit is used to respond to the operation and maintenance personnel being determined to be in the expert stage, identify optimized operation data in the operation behavior data that have a positive deviation from the standard operation data, and control the digital human image to interact with the operation and maintenance personnel through language in order to obtain implicit experience knowledge corresponding to the optimized operation data.

2. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The on-the-job coaching unit includes a deviation quantification analysis module, used for: The actual operation sequence of the operation and maintenance personnel is obtained based on the operation behavior data. The normalized path distance between the actual operation sequence and the standard operation sequence in the standard operation data is determined by the dynamic time warping algorithm. The normalized path distance is used to quantify the matching degree between the operation and maintenance personnel's operation behavior data and the standard operation data. When the distance of the normalized path exceeds a preset slight deviation threshold but is lower than a preset serious violation threshold, an auxiliary guidance prompt signal is generated. The auxiliary guidance prompt signal is used to control the digital human image to provide auxiliary guidance during non-interference times.

3. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The knowledge extraction unit includes a positive deviation identification module and a tacit knowledge conversion module; The positive deviation identification module is used to obtain the operation result, actual time consumption, and deviation degree between the actual operation and the standard operation when the operation behavior data does not match the standard operation data. Based on the operation result, actual time consumption, and deviation degree, it determines whether there is a positive deviation between the operation behavior data and the standard operation data, and identifies and optimizes the operation data based on the judgment result. The implicit knowledge conversion module is used to control the digital human image to interact with the operation and maintenance personnel in the event of a positive deviation between the operation behavior data and the standard operation data. The module converts the voice interaction data of the operation and maintenance personnel into a structured knowledge graph as implicit experience knowledge, sends the implicit experience knowledge to the knowledge base module for storage, and generates an update signal. The update signal is used to update the standard operation process called by the onboarding training unit based on the implicit experience knowledge.

4. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The onboarding training unit includes a performance evaluation module, used for: Based on the accuracy of operations, compliance of processes, and time spent on operations and maintenance personnel during training in physical and / or virtual environments, an assessment report corresponding to the operations and maintenance personnel is generated.

5. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The data acquisition module includes: The visual acquisition device includes at least one of the following: a fixed camera installed in the physical environment, an augmented reality (AR) device worn by maintenance personnel, smart inspection glasses, and a virtual camera installed in a simulated operating room corresponding to the virtual environment; the visual acquisition device is used to acquire the body posture data, hand operation data, and interaction data with the corresponding device of the maintenance personnel, as operational behavior data. Microphone arrays are used to acquire voice interaction data between maintenance personnel and the digital avatar; The system log interface is used to access and parse the operation records of maintenance personnel in the production system to obtain the operation log data of the production system.

6. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The knowledge base module includes: An explicit knowledge base is used to store structured knowledge, which includes at least one of equipment ledgers, standard operating procedure documents, historical fault reports, and equipment drawings. Implicit knowledge base, used to store implicit experiential knowledge; An unstructured knowledge base is used to store unstructured knowledge, which includes at least one of the following: basic information of operation and maintenance personnel, job position, length of service, historical training and assessment records, evaluation reports, and on-the-job performance.

7. The intelligent management system for power plant operation and maintenance personnel as described in claim 1, characterized in that, The stage determination unit includes an operation and maintenance personnel skill proficiency assessment model. The operation and maintenance personnel skill proficiency assessment model is used to represent the preset information and historical operation data of operation and maintenance personnel as multi-dimensional feature vectors and calculate the corresponding comprehensive proficiency score, and determine the trainee stage of operation and maintenance personnel based on the comprehensive proficiency score.

8. A method for intelligent management of power plant operation and maintenance personnel, characterized in that, include: Acquire operational behavior data, voice interaction data, and production system operation log data of maintenance personnel in physical and / or virtual environments; Generate and drive a digital human avatar, which is used to interact with maintenance personnel based on their voice interaction data; Based on the preset information and historical operation data of the operation and maintenance personnel, the student stage of the operation and maintenance personnel is determined. The student stage includes the new employee stage, the growth stage, and the expert stage. When maintenance personnel are identified as new employees, digital avatars can be used to provide them with training and assessment on standard operating procedures for power plants in physical and / or virtual environments. When maintenance personnel are determined to be in the growth stage, the matching degree between the maintenance personnel's operational behavior data and standard operational data is analyzed, and the digital human image is controlled to provide auxiliary guidance during non-interference times. When the maintenance personnel are determined to be in the expert stage, the system identifies optimized operation data that deviates positively from the standard operation data in the operation behavior data, and controls the digital human image to interact with the maintenance personnel through language in order to obtain implicit experience knowledge corresponding to the optimized operation data.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute a method for intelligent management of power plant operation and maintenance personnel as described in claim 8.

10. A non-transitory computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to implement the intelligent management method for power plant operation and maintenance personnel as described in claim 8.

Images