Equipment operation and maintenance comprehensive management method and device

Abstract

The invention relates to the technical field of equipment operation and maintenance, in particular to an equipment operation and maintenance comprehensive management method and device.The method comprises the steps that a natural language is input into a target large language model to recognize an operation and maintenance scheduling task corresponding to the natural language, and a task target and task constraints of the operation and maintenance scheduling task are extracted; and based on the task target and the task constraint, generating an operation and maintenance scheduling scheme and / or a spare part ordering plan and sending the plan to the target operation and maintenance personnel, so that the target operation and maintenance personnel execute the operation and maintenance scheduling scheme and / or the spare part ordering plan to complete management of equipment corresponding to the operation and maintenance scheduling task. According to the method and the device, the operation and maintenance requirements expressed by the natural language can be intelligently judged and analyzed through the target large language model, the pre-defined tools in the tool code library are called, and the optimized operation and maintenance scheduling scheme, the spare part ordering scheme and the like are quickly, accurately and automatically generated according to the actual application scene; and the generation efficiency and reliability of the operation and maintenance scheduling scheme and the spare part ordering plan are effectively ensured.

Description

Technical Field

[0001] This application relates to the technical field of equipment operation and maintenance, and particularly relates to a comprehensive management method and device for equipment operation and maintenance. Background Art

[0002] Equipment deployed in fixed positions usually consists of multiple sub-modules, has a high structural complexity, and requires matching the needs of corresponding professional maintenance personnel, spare parts, and instruments. Moreover, since these equipments are often scattered in a vast area, the operation and maintenance costs are relatively high. Formulating precise operation and maintenance strategies is crucial for the reasonable allocation of operation and maintenance resources.

[0003] In related technologies, the algorithms for generating equipment operation and maintenance scheduling plans mainly include methods based on expert experience, methods based on heuristic algorithms, and methods based on deep learning. The method based on expert experience is a method that relies on the knowledge and experience of experts in the field to formulate scheduling plans; the method based on heuristic algorithms conducts mathematical modeling and customized design for specific problems, and can provide near-optimal feasible solutions for complex optimization problems within an acceptable time; common methods of the method based on deep learning are methods that directly solve using machine learning models or combine with heuristic algorithms and methods that obtain solutions using deep reinforcement learning methods.

[0004] However, in related technologies, the method based on expert experience largely depends on the personal experience and knowledge of experts. When facing large-scale, multi-objective scheduling problems with complex constraints, it is difficult to summarize and construct better-performing scheduling knowledge, and there are problems such as low optimization efficiency and poor scheduling performance, making it difficult to ensure the scheduling effect; although the methods based on heuristic algorithms can quickly provide feasible solutions and are simple to implement, they cannot guarantee finding the global optimal solution, may be unstable and sensitive to parameter selection, and different heuristic algorithms are often designed for specific types of problems and may not be applicable or have poor effects for other types of problems, with poor generalization; although the methods based on deep learning can effectively handle complex systems and high-dimensional state spaces, they require a large amount of data and time to train the model, with high training costs and limited generalization ability for new or different problems, and when facing rapid changes in the environment, these algorithms may be difficult to adjust strategies in a timely manner to adapt to the new environmental conditions, affecting the dynamic adaptability of the algorithms. That is, in the actual operation and maintenance management decision-making process, when facing large-scale, multi-objective scheduling problems with complex constraints, these algorithms are difficult to quickly and accurately generate practical operation and maintenance scheduling plans, and in the face of rapid changes in the environment and the characteristics of large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust strategies in a timely manner to adapt to the new environmental conditions, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenarios, and also unable to guarantee the reliability and efficiency of the operation and maintenance scheduling strategy, which urgently needs to be solved. Summary of the Invention

[0005] This application provides a comprehensive management method and device for equipment operation and maintenance, aiming to solve the problems in the related art that the algorithm for generating equipment operation and maintenance scheduling plans is difficult to quickly and accurately generate practical operation and maintenance scheduling plans when facing large-scale, multi-objective, and complex-constrained scheduling problems in the actual operation and maintenance management process. In the face of the rapid changes in the environment and the large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust strategies in a timely manner to adapt to the new environmental conditions, with poor dynamic adaptability and generalization ability, unable to effectively cover the actual application scenarios, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy.

[0006] The first aspect of the embodiments of this application provides a comprehensive management method for equipment operation and maintenance, including the following steps: Input natural language into the target large language model to identify the operation and maintenance scheduling tasks corresponding to the natural language, and extract the task objectives and task constraints of the operation and maintenance scheduling tasks; Based on the task objectives and the task constraints, generate an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks; Send the operation and maintenance scheduling plan and / or spare part ordering plan to the target operation and maintenance personnel, so that the target operation and maintenance personnel execute the operation and maintenance scheduling plan and / or spare part ordering plan to complete the management of the equipment corresponding to the operation and maintenance scheduling tasks.

[0007] Optionally, in an embodiment of this application, before inputting the natural language into the target large language model, it further includes: Collect the tool operation instruction information of the operation and maintenance personnel to generate a target instruction template according to the tool operation instruction information; Based on the target instruction template, use the basic large language model to generate tool call instructions and their corresponding response data, and construct an instruction corpus according to the tool call instructions and their corresponding response data; Correct the instruction corpus, and fine-tune the basic large language model according to the corrected instruction corpus to generate the target large language model.

[0008] Optionally, in an embodiment of this application, the generating an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks based on the task objectives and the task constraints includes: Generate a tool code library according to the target instruction template, the target scheduling mathematical model, and the target solver component; Use the tool code library to generate an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks.

[0009] Optionally, in an embodiment of this application, the generating an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks based on the task objectives and the task constraints further includes: Determine the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling tasks; Determine the operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks according to the target scheduling mathematical model.

[0010] Optionally, in an embodiment of the present application, determining the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task includes: when the actual scheduling scenario is a static scheduling scenario, determining the target scheduling mathematical model as a bilevel programming model, where the bilevel programming model includes an integer linear programming problem model and a vehicle routing problem model with capacity constraints; when the actual scheduling scenario is a dynamic scheduling scenario, determining the target scheduling mathematical model as a dynamic vehicle routing problem model.

[0011] Optionally, in an embodiment of the present application, inputting the natural language into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and extracting the task objective and task constraints of the operation and maintenance scheduling task includes: detecting whether the integrity of the natural language meets the preset integrity requirement; when the integrity meets the preset integrity requirement, extracting the task objective and task constraints of the operation and maintenance scheduling task, otherwise, supplementing the natural language to meet the preset integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task.

[0012] An embodiment of the second aspect of the present application provides an equipment operation and maintenance comprehensive management device, including: an identification module, configured to input natural language into a target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and extract the task objective and task constraints of the operation and maintenance scheduling task; a first generation module, configured to generate an operation and maintenance scheduling plan and / or a spare part order plan for the operation and maintenance scheduling task based on the task objective and the task constraints; a management module, configured to send the operation and maintenance scheduling plan and / or the spare part order plan to the target operation and maintenance personnel to complete the management of the equipment corresponding to the operation and maintenance scheduling task by the target operation and maintenance personnel executing the operation and maintenance scheduling plan and / or the spare part order plan.

[0013] Optionally, in an embodiment of the present application, it further includes: a collection module, configured to collect tool operation instruction information of operation and maintenance personnel before inputting the natural language into the target large language model to generate a target instruction template according to the tool operation instruction information; a construction module, configured to generate a tool call instruction and its corresponding response data based on the target instruction template by using a basic large language model, and construct an instruction corpus according to the tool call instruction and its corresponding response data; a second generation module, configured to correct the instruction corpus to fine-tune the basic large language model according to the corrected instruction corpus to generate the target large language model.

[0014] Optionally, in an embodiment of the present application, the first generation module includes: a first generation unit configured to generate a tool code library according to the target instruction template, the target scheduling mathematical model, and the target solver component; a second generation unit configured to generate an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task by using the tool code library.

[0015] Optionally, in an embodiment of the present application, the first generation module further includes: a first determination unit configured to determine the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task; a second determination unit configured to determine an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task according to the target scheduling mathematical model.

[0016] Optionally, in an embodiment of the present application, the first determination unit includes: a first determination subunit configured to determine that the target scheduling mathematical model is a bilevel programming model when the actual scheduling scenario is a static scheduling scenario, where the bilevel programming model includes an integer linear programming problem model and a vehicle routing problem model with capacity constraints; a second determination subunit configured to determine that the target scheduling mathematical model is a dynamic vehicle routing problem model when the actual scheduling scenario is a dynamic scheduling scenario.

[0017] Optionally, in an embodiment of the present application, the recognition module includes: a detection unit configured to detect whether the integrity of the natural language meets a preset integrity requirement; an extraction unit configured to extract the task objective and task constraints of the operation and maintenance scheduling task when the integrity meets the preset integrity requirement, otherwise, supplement the natural language until it meets the preset integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task.

[0018] An embodiment of the third aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the equipment operation and maintenance comprehensive management method as described in the above embodiments.

[0019] An embodiment of the fourth aspect of the present application provides a computer-readable storage medium storing a computer program, and when the program is executed by a processor, it implements the above equipment operation and maintenance comprehensive management method.

[0020] An embodiment of the fifth aspect of the present application provides a computer program product including a computer program, and when the computer program is executed, it is used to implement the above equipment operation and maintenance comprehensive management method.

[0021] In the embodiments of the present application, natural language can be input into a target large language model to extract the operation and maintenance scheduling tasks corresponding to the natural language, and an operation and maintenance scheduling plan and / or a spare part ordering plan can be generated. Then, the target operation and maintenance personnel execute the operation and maintenance scheduling plan and / or the spare part ordering plan to complete the management of the equipment. Thus, it realizes the intelligent judgment and analysis of the operation and maintenance requirements expressed in natural language through the target large language model, and calls the tools predefined in the tool code library to quickly and accurately generate optimized operation and maintenance scheduling plans and spare part ordering plans according to the actual application scenarios, effectively ensuring the generation efficiency and reliability of the operation and maintenance scheduling plan and the spare part ordering plan. Moreover, the present application can also design and develop an operation and maintenance comprehensive management system based on the target large language model and the tool code library, combine the technical advantages of the large language model and the solver, and more intelligently process and analyze operation and maintenance data and information, thereby effectively improving the digital and intelligent levels of the model equipment health management system, the reliability of the equipment, extending the service life, and reducing the maintenance cost and risk, providing decision-making support for the intelligent operation and maintenance management of the equipment system. Thus, it solves the problems that in the related art, the operation and maintenance scheduling plan generation algorithm for equipment is difficult to quickly and accurately generate practical operation and maintenance scheduling plans and / or spare part ordering plans when facing the scheduling problems with large scale, multiple objectives, and complex constraints in the actual operation and maintenance management process, and in the face of the rapid changes in the environment and the characteristics of large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust the strategy in time to adapt to the new environmental conditions, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenarios, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy.

[0022] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0024] Figure 1 FIG. is a flowchart of an equipment operation and maintenance comprehensive management method according to an embodiment of the present application;

[0025] Figure 2 FIG. is a schematic diagram of the framework of an equipment operation and maintenance comprehensive management system based on a large language model and a solver according to an embodiment of the present application;

[0026] Figure 3 FIG. is a schematic diagram of an equipment operation and maintenance decision-making process according to an embodiment of the present application;

[0027] Figure 4 FIG. is a schematic diagram of a multi-vehicle operation and maintenance task scheduling scenario according to an embodiment of the present application;

[0028] Figure 5 The structural schematic diagram of the equipment operation and maintenance integrated management device provided according to an embodiment of the present application;

[0029] Figure 6 The structural schematic diagram of the electronic device provided according to an embodiment of the present application.

[0030] Reference numerals:

[0031] 10 - Equipment operation and maintenance integrated management device: 100 - Identification module, 200 - First generation module, and 300 - Management module; 601 - Memory, 602 - Processor, and 603 - Communication interface. Detailed implementation manners

[0032] The embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described by referring to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as a limitation to the present application.

[0033] The following describes the equipment operation and maintenance comprehensive management method and device according to the embodiments of the present application. In view of the problem that the algorithm for generating the equipment operation and maintenance scheduling plan in the related technology mentioned in the above background technology is difficult to quickly and accurately generate a practical operation and maintenance scheduling plan when facing the scheduling problems with large scale, multiple objectives, and complex constraints in the actual operation and maintenance management process, and in the face of the rapid changes in the environment and the characteristics of large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust the strategy in a timely manner to adapt to the new environmental conditions, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenarios, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy, the present application provides an equipment operation and maintenance comprehensive management method. In this method, natural language can be input into the target large language model to extract the operation and maintenance scheduling tasks corresponding to the natural language, and generate the operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling tasks. Then, the target operation and maintenance personnel execute the operation and maintenance scheduling plan and / or spare part ordering plan to complete the management of the equipment. Thus, it realizes the intelligent judgment and analysis of the operation and maintenance requirements expressed in natural language through the target large language model, and calls the tools predefined in the tool code library to quickly and accurately automatically generate optimized operation and maintenance scheduling plans and spare part ordering plans according to the actual application scenarios, effectively ensuring the generation efficiency and reliability of the operation and maintenance scheduling plan and spare part ordering plan; and the present application can also design and develop an operation and maintenance comprehensive management system based on the target large language model and the tool code library, combining the technical advantages of the large language model and the solver to more intelligently process and analyze the operation and maintenance data and information, thereby effectively improving the digital and intelligent level of the model equipment health management system, the reliability of the equipment, extending the service life, and reducing the maintenance cost and risk, providing decision support for the intelligent operation and maintenance management of the equipment system. Thus, it solves the problems that the algorithm for generating the equipment operation and maintenance scheduling plan in the related technology is difficult to quickly and accurately generate a practical operation and maintenance scheduling plan when facing the scheduling problems with large scale, multiple objectives, and complex constraints in the actual operation and maintenance management process, and in the face of the rapid changes in the environment and the characteristics of large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust the strategy in a timely manner to adapt to the new environmental conditions, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenarios, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy, etc.

[0034] Specifically, Figure 1 is a flowchart of an equipment operation and maintenance comprehensive management method provided by an embodiment of the present application.

[0035] As Figure 1 shown, the equipment operation and maintenance comprehensive management method includes the following steps:

[0036] In step S101, the natural language is input into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and the task objective and task constraints of the operation and maintenance scheduling task are extracted.

[0037] In some embodiments, when the equipment deployed at a fixed location is undergoing operation and maintenance management, for the convenience of operation and maintenance personnel to manage, the embodiments of the present application can use the target large language model to understand the natural language input by the operation and maintenance personnel, so as to understand the intention of the operation and maintenance personnel.

[0038] For example, when the operation and maintenance personnel are performing operation and maintenance management on the equipment, they can input the requirements into the target large language model through natural language. The target large language model needs to identify the type of requirements corresponding to the natural language and various information existing in the language to identify the operation and maintenance scheduling task corresponding to the natural language.

[0039] Among them, the operation and maintenance scheduling task can be understood here as meeting the requirements of multiple equipment operation and maintenance management by scheduling devices, instruments, and multiple equipment operation and maintenance management personnel. That is, during the process of the equipment operation and maintenance management personnel performing operation and maintenance management on the equipment, they may input the need for certain devices, instruments, or the assistance of other equipment operation and maintenance management personnel, etc. And these devices or instruments may need to be scheduled for use due to quantity issues, and the required equipment operation and maintenance management personnel may be performing operation and maintenance management on other equipment; in order to ensure that the required devices or instruments can be scheduled among multiple pieces of equipment and multiple equipment operation and maintenance management personnel can complete the comprehensive management of multiple pieces of equipment, certain operation and maintenance scheduling tasks will be generated.

[0040] And the target large language model in the embodiments of the present application can obtain the corresponding operation and maintenance scheduling task by identifying the natural language of the equipment operation and maintenance management personnel, and extract the corresponding task objective and task constraints from the operation and maintenance scheduling task. Among them, the target large language model can be understood here as a large language model fine-tuned based on some open-source ChatGLM large language base models.

[0041] Among them, the task objective here refers to the objective that the operation and maintenance scheduling task needs to meet. For example, meeting the operation and maintenance time requirements, minimizing the scheduling time, etc.; the task constraint here refers to the constraints that will be encountered when executing the operation and maintenance scheduling task. For example, resource limitation constraints such as devices and instruments, time window constraints, etc.

[0042] Optionally, in an embodiment of the present application, before inputting natural language into the target large language model, it further includes: collecting tool operation instruction information of operation and maintenance personnel to generate a target instruction template according to the tool operation instruction information; based on the target instruction template, using a basic large language model to generate tool call instructions and their corresponding response data, and constructing an instruction corpus according to the tool call instructions and their corresponding response data; correcting the instruction corpus, and fine-tuning the basic large language model according to the corrected instruction corpus to generate the target large language model.

[0043] Based on the related descriptions of other embodiments, it can be understood that the target large language model in the present application can be understood here as a large language model fine-tuned based on an open-source large language base model. Among them, the specific generation process of the target large language model can but is not limited to being expressed as follows:

[0044] First of all, the embodiment of the present application needs to construct an instruction corpus based on a basic large language model such as the open-source ChatGLM large language base model, and perform manual quality inspection and correction.

[0045] In this process, the embodiment of the present application can collect tool operation instruction information of operation and maintenance personnel to generate a target instruction template according to the tool operation instruction information. Among them, the target instruction template can be understood here as a tool call instruction template defined by the tool operation instructions commonly used by operation and maintenance personnel such as operation and maintenance scheduling management decision-makers or equipment operation and maintenance management personnel in their work.

[0046] For example, the present application can first summarize and sort out various tools commonly used by operation and maintenance scheduling management decision-makers and equipment operation and maintenance management personnel, such as equipment information query tools, spare part information query tools, spare part ordering plan generation tools, and operation and maintenance plan generation tools, etc., and determine the parameter variables involved in the tools. Then, based on this, a set of target instruction templates is defined. This target instruction template needs to contain various descriptions of the tools and parameter variables, and be able to cover various queries and requests that the management decision-maker may put forward. The specific format can be set by those skilled in the art according to actual needs. Here is only an exemplary illustration and no specific limitation is made.

[0047] Taking "equipment information query" as an example, the format of the instruction template defined in the embodiment of the present application can but is not limited to being expressed as follows:

[0048] Dialogue turn 1:

[0049] Role: System;

[0050] Content: Tool settings (Tool name identifier: get_equipment_info; Tool usage description: Query and obtain information about the equipment to be queried and output it in a table in markdown format; Parameters required by the tool and their data types: Equipment name - string, Equipment health status - list);

[0051] Dialogue turn 2:

[0052] Role: User;

[0053] Content: User instruction (Which XX equipment has minor faults);

[0054] Dialogue turn 3:

[0055] Role: AI assistant;

[0056] Content: User instruction parsing (Call tool name identifier: get_equipment_info; Parameter parsing: Equipment name: XX equipment, Equipment health status: ["Minor fault"]);

[0057] Dialogue turn 4:

[0058] Role: Code call tool;

[0059] Content: Return result of the code tool;

[0060] Dialogue turn 5:

[0061] Role: AI assistant;

[0062] Content: Sorting, interpreting, and explaining the return result of the code tool (Based on the query result, all XX equipment with minor faults are sorted into a table in markdwon format as follows: | Equipment type | Equipment code | Equipment status | |---|---|---| | XX equipment | E1 | Minor fault | | XX equipment | E2 | Medium fault | | XX equipment | E3 | Minor fault |).

[0063] Next, the embodiments of the present application can generate tool call instructions and their corresponding response data based on this target instruction template using a basic large language model, so as to construct an instruction corpus according to the tool call instructions and their corresponding response data. For example, based on the above-defined target instruction template and prompt words (relevant prompt words given for the tool call instructions input in real time), using a basic large language model such as the ChatGLM large language model to generate tool call instructions and corresponding response data, including but not limited to user instructions, return results of the tool, and interpretations and answers to these results, forming instruction-response pairs, and constructing an instruction corpus based on all the instruction-response pairs.

[0064] Finally, the embodiments of the present application need to correct the instruction corpus to fine-tune the basic large language model according to the corrected instruction corpus to generate the target large language model. Since the large language model may generate "hallucination" information that is inconsistent with facts or completely fabricated when generating text, the embodiments of the present application also need to correct the corpus, and the correction method can be manual quality inspection. By identifying and correcting these instructions, the data quality of the instruction corpus can be effectively improved to ensure that each instruction is clear, accurate, and the response is appropriate and meaningful.

[0065] Among them, in order to enable the target large language model to better adapt to and complete tasks in the field of operation and maintenance scheduling, when fine-tuning the basic large language model according to the corrected instruction corpus in the embodiments of the present application, the tool call accuracy and parameter extraction accuracy can be used as the ultimate goal, and the open-source ChatGLM large language base model can be fine-tuned based on the instruction corpus using the parameter-efficient fine-tuning method. The available parameter-efficient fine-tuning methods include but are not limited to Adapter-Tuning, Prefix-Tuning, P-tuning, LoRA and its derivative methods, etc. Specifically, it can be selected or adjusted by those skilled in the art according to the actual situation. Here, only an exemplary illustration is given without specific limitation.

[0066] For example, the Low-Rank Adaptation (LoRA) method is used to fine-tune the ChatGLM large language base model in the embodiments of the present application. Specifically, when fine-tuning, the embodiments of the present application can keep the overall structure and parameters of the model unchanged, and insert a learnable low-rank matrix between the hidden layer and the output layer of the model to adjust the weights of the model, so as to achieve local adjustment of the model parameters. For example, keep its original weight matrix W unchanged, only fine-tune the weight matrix of the training update bypass, and this updated weight matrix is decomposed into two low-rank matrices A and B. Among them, the A matrix is initialized with a random Gaussian distribution, and the B matrix is initialized with a 0 matrix.

[0067] Assume that the weight matrix of the pre-trained model is W0 ∈ R d×k , then its update, that is, the mathematical formula of the fine-tuning process of the large language model can be but not limited to expressed as:

[0068] W = W0 + ΔW = W0 + BA

[0069] Among them, B ∈ R d×r , A ∈ R r×d , r << min(d, k), r is the selected rank (a hyperparameter value selected manually).

[0070] Optionally, in an embodiment of the present application, natural language is input into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and the task objective and task constraints of the operation and maintenance scheduling task are extracted, including: detecting whether the integrity of the natural language meets the preset integrity requirement; when the integrity meets the preset integrity requirement, extracting the task objective and task constraints of the operation and maintenance scheduling task, otherwise, supplementing the natural language until it meets the preset integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task.

[0071] It can be understood that when the target large language model in the embodiment of the present application identifies the operation and maintenance scheduling task corresponding to the natural language, the basis for identification is the natural language input by the operation and maintenance personnel.

[0072] However, in some embodiments, there may be various problems with the natural language input by the operation and maintenance personnel to the target large language model. For example, the language is incomplete, only giving the device name "screw" without clear demand words, only giving words such as "two screws" without indicating whether it is necessary to send them to the current place or other places, etc.

[0073] Therefore, in the embodiment of the present application, when identifying the operation and maintenance scheduling task corresponding to the natural language to extract the task objective and task constraints of the operation and maintenance scheduling task, it can be detected whether the integrity of the natural language meets the preset integrity requirement. Here, the preset integrity requirement can be understood as a certain requirement that the natural language needs to meet. For example, it has at least three elements among elements such as time, place, person, and event.

[0074] Then, when the integrity of the natural language meets the preset integrity requirement, the task objective and task constraints of the operation and maintenance scheduling task are extracted. Otherwise, in the embodiment of the present application, the natural language can be supplemented to a certain extent until it meets a certain integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task. For example, if the user only inputs "Need two screws", the embodiment of the present application can display in the UI dialogue interface through the tool call instruction commonly used by the operation and maintenance personnel: "Do you need two screws to repair the equipment at XX location now?" If the user's response is yes, the task objective and task constraints of the operation and maintenance scheduling task can be extracted based on this supplemented complete sentence.

[0075] Step S102, based on the task objective and task constraints, generate an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task.

[0076] In some embodiments, after determining the task objective and task constraints of the operation and maintenance scheduling task, the embodiment of the present application can generate an operation and maintenance scheduling plan or a spare part ordering plan or an operation and maintenance scheduling plan and a spare part ordering plan corresponding to the operation and maintenance scheduling task based on the task objective and task constraints.

[0077] For example, there are many fixed equipment (such as large machinery and production lines) distributed in a large industrial park. In order to ensure the continuous and stable operation of these equipment, the park decided to adopt a multi-vehicle operation and maintenance scheduling solution to complete the operation and maintenance tasks efficiently and safely.

[0078] The mission objectives of this operation and maintenance scheduling task include but are not limited to: ensuring that all fixed equipment receives timely and effective operation and maintenance services to reduce downtime caused by failures; optimizing the use of vehicles and operation and maintenance personnel to improve operation and maintenance efficiency and reduce operation and maintenance costs; complying with safety regulations and operating procedures to ensure the safety of personnel and equipment during the operation and maintenance process.

[0079] The task constraints of the operation and maintenance scheduling task include but are not limited to: the number of operation and maintenance vehicles is limited, and factors such as vehicle load, driving speed and fuel consumption need to be considered; fixed equipment is distributed in different locations in the park, and the operation and maintenance tasks need to consider the distance and traffic conditions; the number of operation and maintenance personnel is limited, and they need to have certain professional skills and experience; operation and maintenance tasks have priorities, and emergency tasks need to be handled first.

[0080] Therefore, the factors that need to be considered in the operation and maintenance scheduling plan include: operation and maintenance team and vehicle configuration, task allocation and priority sorting, route planning and vehicle scheduling, and emergency plan formulation and drills.

[0081] According to the above, assume that there are 5 key equipment in the park that need regular operation and maintenance, 3 operation and maintenance vehicles, and 6 operation and maintenance personnel. The following is the implementation of the operation and maintenance scheduling plan for a certain day:

[0082] 08:00-09:00: The first operation and maintenance vehicle goes to equipment A for routine inspection and maintenance, while the second operation and maintenance vehicle goes to equipment B for preventive maintenance.

[0083] 09:30-10:30: After the first maintenance vehicle completes the maintenance task of equipment A, it goes to equipment C for emergency repair (equipment C has a fault). After the second maintenance vehicle completes the maintenance task of equipment B, it returns to the base and waits.

[0084] 11:00-12:00: The third operation and maintenance vehicle goes to equipment D for routine inspection. At the same time, the first operation and maintenance vehicle returns to the base to stand by after completing the emergency maintenance task of equipment C.

[0085] 13:30-14:30: The second operation and maintenance vehicle goes to equipment E for preventive maintenance. After the third operation and maintenance vehicle completes the operation and maintenance task of equipment D, it returns to the base and stands by.

[0086] 15:00 - 16:00: The first operation and maintenance vehicle goes to Equipment F for daily inspection according to the new operation and maintenance requirements (assuming Equipment F is a newly added operation and maintenance task). After the second operation and maintenance vehicle completes the operation and maintenance task of Equipment E, it also goes to Equipment F to provide assistance (such as carrying tools, materials, etc.).

[0087] By implementing the above operation and maintenance scheduling plan, the fixed equipment in the park can obtain timely and effective operation and maintenance services, and the operation and maintenance efficiency and cost are optimized.

[0088] Optionally, in an embodiment of the present application, based on the task objective and task constraints, an operation and maintenance scheduling plan and / or spare parts ordering plan for the operation and maintenance scheduling task is generated, including: generating a tool code library according to the target instruction template, the target scheduling mathematical model, and the target solver component; using the tool code library to generate an operation and maintenance scheduling plan and / or spare parts ordering plan for the operation and maintenance scheduling task.

[0089] As a possible implementation method, when generating the operation and maintenance scheduling plan for the operation and maintenance scheduling task in the embodiment of the present application, for the convenience of comprehensive management, a certain tool code library can be generated according to the target instruction template, the target scheduling mathematical model, and the target solver component, and provided to the target large language model for calling in the form of a function, but not limited to this.

[0090] Specifically, the embodiment of the present application can develop corresponding tool components based on the target instruction template, that is, the tool call instruction template, to form a code library, and then different tool functions can be implemented by writing different codes according to actual usage needs, forming a tool code library, and provided to the target large language model for calling in the form of a function. Further, the embodiment of the present application can also integrate advanced target solver components in combination with the target scheduling mathematical model to adapt to dynamic constraint conditions and optimization objectives, and finally generate an efficient and highly scalable operation and maintenance scheduling plan or spare parts ordering plan, etc.

[0091] Among them, the target scheduling mathematical model can be understood here as different scheduling mathematical models constructed according to different application scenarios for determining the operation and maintenance scheduling plan and the spare parts ordering plan. The target solver here refers to the solver component that can be used to solve the scheduling mathematical model, including but not limited to open-source solvers such as OR-Tools, SCIP, Gurobi, and self-developed solvers.

[0092] Generally speaking, the tool code library in the embodiments of the present application mainly but not limited to includes two parts: one is the basic tools such as equipment information query, spare part information query, etc. developed in combination with background information and model, spare part ordering plan generation, single / multi-vehicle task scheduling, dynamic task scheduling, etc.; the other is the solver algorithm code library constructed by developing code based on the target scheduling mathematical model, including but not limited to database read and write statements, single / multi-vehicle task mathematical models, dynamic task mathematical models, spare part ordering mathematical models, etc.

[0093] Optionally, in an embodiment of the present application, generating an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task based on the task objective and task constraints further includes: determining the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task; determining the operation and maintenance scheduling plan and / or the spare part ordering plan for the operation and maintenance scheduling task according to the target scheduling mathematical model.

[0094] In the actual execution process, in the actual application scenario, the operation and maintenance resource scheduling of fixed equipment can usually be modeled as a Vehicle Routing Problem (VRP). A common method to solve such problems is to use a solver to develop a customized algorithm to optimize the driving route of the vehicle, meet the operation and maintenance requirements while reducing costs and improving operation and maintenance efficiency.

[0095] However, in the actual operation and maintenance management decision-making process, when generating the operation and maintenance scheduling plan and spare part ordering plan for the operation and maintenance scheduling task, there are many influencing factors, such as: comprehensive consideration of cost and priority, spare part inventory limit, maximum mileage limit of operation and maintenance vehicles, etc. Therefore, it is also necessary to accurately understand the strategic or task requirements and generate a scheduling strategy by comprehensively considering global information such as maintenance costs, inventory information, and maintenance priorities.

[0096] Based on this, the embodiments of the present application need to clarify the constraints and optimization objectives of the operation and maintenance scheduling task according to the actual scheduling scenario of the operation and maintenance scheduling task, and thus establish a scheduling mathematical model corresponding to the actual scenario of the operation and maintenance scheduling task.

[0097] Next, the actual scenario and the corresponding target scheduling mathematical model in the embodiments of the present application will be further described.

[0098] Optionally, in an embodiment of the present application, determining the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task includes: when the actual scheduling scenario is a static scheduling scenario, determining the target scheduling mathematical model as a bilevel programming model, where the bilevel programming model includes an integer linear programming problem model and a vehicle routing problem model with capacity constraints; when the actual scheduling scenario is a dynamic scheduling scenario, determining the target scheduling mathematical model as a dynamic vehicle routing problem model.

[0099] In some embodiments, when determining the target scheduling mathematical model according to the actual scenario of the operation and maintenance scheduling task, the embodiments of the present application mainly but not limited to design two scheduling mathematical models for the static scheduling scenario and the dynamic scheduling scenario respectively.

[0100] (1) For the static scheduling scenario, the embodiments of the present application can design a two-stage algorithm, that is, design a bilevel programming model to improve the efficiency and effect of the overall scheduling. Among them, the upper model of the bilevel programming model mainly determines the optimal solution of resource allocation and scheduling decision by combining the scheduling resource condition constraints and optimization objectives. Therefore, the embodiments of the present application can model it as an integer linear programming problem (Integer Linear Programming, ILP); the lower model mainly ensures that the execution path of each group of tasks not only meets the limitations during the task execution, but also achieves the optimization of efficiency and cost. Therefore, the embodiments of the present application can model it as a capacitated vehicle routing problem model (Capacitated Vehicle Routing Problem, CVRP).

[0101] Among them, the formula of the upper integer linear programming model can be but not limited to expressed as follows:

[0102]

[0103] Among them, Z1 is the objective function of the upper integer linear programming model, which can maximize the total incentive score under the constraint of the spare part inventory quantity; n is the number of operation and maintenance points; s is the number of spare part types related to maintenance; x j and x i respectively represent the decision variables of operation and maintenance point j and operation and maintenance point i. If operation and maintenance point j is selected, then x j is 1, otherwise it is 0; c j is the incentive for completing the task of operation and maintenance point j, and its value is the total priority score of operation and maintenance point j. The priority score is assigned to the operation and maintenance point according to the failure degree of the equipment at operation and maintenance point j and the overall operation and maintenance principle; a ie is the demand of operation and maintenance point i for spare part e; b e is the total inventory of spare part e;

[0104] The formula of the lower capacitated vehicle routing problem model can be but not limited to expressed as follows:

[0105]

[0106] Among them, Z2 is the objective function of the lower capacitated vehicle routing problem model, which is divided into two parts. The first part is the total driving path of the operation and maintenance tasks under the scheduling rule constraints, and the second part is the total penalty cost of all operation and maintenance points that have not been repaired. The goal is to minimize two parts; n is the number of operation and maintenance points; m is the number of operation and maintenance vehicles; c ij is the distance from operation and maintenance point i to operation and maintenance point j for the operation and maintenance vehicle; x i k j indicates whether operation and maintenance vehicle k travels from operation and maintenance point i to operation and maintenance point j. If so, it is 1, otherwise it is 0; p i is the penalty coefficient for operation and maintenance point i not being visited. It is assigned according to the failure situation of the equipment at this operation and maintenance point. Each operation and maintenance point is visited at most once. The warehouse is point 0, which is the starting and ending point of the driving routes of all operation and maintenance vehicles; q i and q j respectively represent the spare part demand quantities for operation and maintenance points i and j. Q is the upper limit of the spare part capacity of the operation and maintenance vehicle; T is the upper limit of the task quantity of a single operation and maintenance vehicle; D is the upper limit of the path length of a single operation and maintenance vehicle. indicates whether operation and maintenance vehicle k travels from operation and maintenance point 0 (i.e., the warehouse) to operation and maintenance point j. If so, it is 1, otherwise it is 0. indicates whether operation and maintenance vehicle k travels from operation and maintenance point j to operation and maintenance point 0 (i.e., the warehouse). If so, it is 1, otherwise it is 0.

[0107] (2) Based on the static scheduling scenario, for the dynamic scheduling scenario, the embodiment of the present application can model the problem as a Dynamic Vehicle Routing Problem With Time Windows (DVRPWTW), which helps to respond in a timely manner to changes in tasks at the decision-making moment, calculate and match the optimal strategy.

[0108] Considering that in the actual application scenario, the dynamic changes of tasks may involve the addition and deletion of quantities, and the urgency (priority) of different tasks is also an important factor to be considered in scheduling. Therefore, the dynamic vehicle routing problem model in the embodiment of the present application includes but is not limited to two parts: task addition and task priority improvement.

[0109] Among them, in the embodiment of the present application, the expression of the task addition part in the dynamic vehicle routing problem model can be but is not limited to expressed as follows:

[0110]

[0111] Among them, Z is the objective function of the dynamic vehicle routing problem model, which is the total driving path of the operation and maintenance tasks under the scheduling rule constraints. The goal is to minimize the overall operation and maintenance cost; n1 is the number of operation and maintenance points that have not been completed before the task addition. Due to the addition of a task, the total number of current uncompleted maintenance points is n1 + 1; is the newly added maintenance point; m is the number of operation and maintenance vehicles; the position coordinate of the k-th operation and maintenance vehicle at the current moment is the point numbered n1 + k + 1, and the starting point of all operation and maintenance vehicle paths is the current position, and the end point is the warehouse; c ij is the distance from operation and maintenance point i to operation and maintenance point j for the operation and maintenance vehicle; indicates whether the operation and maintenance vehicle k travels from operation and maintenance point i to operation and maintenance point j. If so, it is 1, otherwise it is 0; is the order of operation and maintenance point i in the task list of operation and maintenance vehicle k, that is, the warehouse must be visited first before visiting the newly added operation and maintenance point n* (to meet the corresponding requirements such as spare parts and operation and maintenance personnel), indicates the order of the warehouse in the task list of operation and maintenance vehicle k, indicates at (that is, the newly added maintenance point) the order of operation and maintenance vehicle k in the task list; d k is the total distance traveled by operation and maintenance vehicle k. When the operation and maintenance vehicle k returns to the warehouse, the total traveled path will be reset to zero. Each operation and maintenance point is visited at most once. The warehouse is point 0, which is the starting point and end point of all operation and maintenance vehicle travel paths; D is the upper limit of the path length of a single operation and maintenance vehicle, indicates whether the operation and maintenance vehicle k travels from operation and maintenance point j to operation and maintenance point 0 (that is, the warehouse),

[0112] If so, it is 1, otherwise it is 0.

[0113] And, in the embodiment of the present application, the expression of the part for improving the task priority in the vehicle path planning problem model can be but is not limited to being expressed as follows:

[0114]

[0115]

[0116] Among them, Z is the objective function of the dynamic vehicle path planning problem model, which is the sum of the travel paths of operation and maintenance tasks under the restriction of scheduling rules. The goal is to minimize the overall operation and maintenance cost; n2 is the number of uncompleted operation and maintenance points before priority adjustment, is the operation and maintenance point whose priority is improved; m is the number of operation and maintenance vehicles; the position coordinate of the k-th operation and maintenance vehicle at the current moment is the point numbered n2 + k, and the starting point of all operation and maintenance vehicle paths is the current position, and the end point is the warehouse; c ij is the distance from operation and maintenance point i to operation and maintenance point j for the operation and maintenance vehicle; indicates whether the operation and maintenance vehicle k travels from operation and maintenance point i to operation and maintenance point j. If so, it is 1, otherwise it is 0; n k is the operation and maintenance point where the operation and maintenance vehicle k is currently located; is the state variable of the operation and maintenance vehicle k. If the operation and maintenance vehicle k is performing a maintenance task, then the value is 0; if the operation and maintenance vehicle k is in motion, then the value is 1; For the operation and maintenance point The order in the task list of operation and maintenance vehicle k Is the order of operation and maintenance point i in the task list of operation and maintenance vehicle k, that is The task priority of... is higher than that of all other unstarted maintenance tasks; d k Is the total distance traveled by operation and maintenance vehicle k; Each operation and maintenance point is visited at most once, and the warehouse is point 0, which is the starting and ending point of the driving paths of all operation and maintenance vehicles; D is the upper limit of the path length of a single operation and maintenance vehicle Indicates whether operation and maintenance vehicle k travels from operation and maintenance point j to operation and maintenance point 0 (i.e., the warehouse). If it is, it is 1; otherwise, it is 0

[0117] In the embodiments of the present application, a target scheduling mathematical model in the tool code library can be established based on the actual scenario. Thus, when facing scheduling problems with large scale, multiple objectives, and complex constraints in the actual operation and maintenance management process, the strategy can be adjusted quickly, accurately, and in a timely manner, and a practical operation and maintenance scheduling plan can be generated, effectively improving the dynamic adaptability and generalization of the present application, and being able to ensure the reliability and efficiency of the operation and maintenance scheduling strategy while effectively covering the actual application scenario

[0118] Step S103, send the operation and maintenance scheduling plan and / or spare part ordering plan to the target operation and maintenance personnel, so as to complete the management of the equipment corresponding to the operation and maintenance scheduling task through the execution of the operation and maintenance scheduling plan and / or spare part ordering plan by the target operation and maintenance personnel

[0119] In some embodiments, after generating the operation and maintenance scheduling plan and the spare part ordering plan, the present application can send the operation and maintenance scheduling plan and the spare part ordering plan to the target operation and maintenance personnel in the form of intranet information, text messages, etc. After receiving the plan, the target operation and maintenance personnel can execute the plan to complete the operation and maintenance scheduling task, thereby completing the management of the equipment corresponding to the operation and maintenance scheduling task

[0120] The following uses a specific embodiment to explain the equipment operation and maintenance comprehensive management method in the embodiments of the present application

[0121] Figure 2 Is a schematic diagram of the equipment operation and maintenance comprehensive management system framework based on the large language model and the solver in an embodiment of the present application, as Figure 2 shown

[0122] To make full use of operation and maintenance data, improve response speed, and make the management system more intelligent, the embodiment of this application can build an equipment operation and maintenance integrated management system based on a large language model and a solver, which is deployed on equipment at a fixed location. The equipment operation and maintenance integrated management system mainly includes, but is not limited to, five parts: a UI dialogue interaction interface, a large language model, and a tool code library (open-source solver and self-developed solver), independent host private deployment, and specific task allocation.

[0123] Divided by levels, the equipment operation and maintenance integrated management system mainly includes, but is not limited to, seven layers, namely: the interface layer, that is, the UI dialogue interaction interface, which can realize the interaction between users and the equipment operation and maintenance integrated management system; the semantic understanding layer, that is, using the large language model to understand the natural language input by users, including demand type recognition and information extraction (information completion, demand collation), etc.; the business layer is the business services that users can choose, such as data query, single / multi-vehicle task scheduling, dynamic task scheduling, and spare part ordering plan generation, etc.; the code library is the processing code corresponding to the business layer generated based on the solver component, such as database read and write statements, single / multi-vehicle task mathematical models, dynamic task mathematical models, and spare part ordering mathematical models, etc.; the data layer is various scenario parameters of operation and maintenance scheduling tasks, model parameters of various models, corpus, and stored solution files, etc.; the operating environment is the independent host private deployment, and the task execution is that the operation and maintenance personnel execute specific task allocation.

[0124] Figure 3 It is a schematic diagram of the equipment operation and maintenance decision-making process for an embodiment of this application; Figure 4 It is a schematic diagram of a multi-vehicle operation and maintenance task scheduling scenario for an embodiment of this application, such as Figure 3 and Figure 4 shown:

[0125] Users can input natural language. The large language model can accept the operation and maintenance requirements expressed by operation and maintenance personnel in natural language, and accurately call the corresponding tools in the code library through functions such as demand recognition, information completion, and demand collation, quickly and accurately understand the operation and maintenance requirements and intentions, obtain multi-vehicle operation and maintenance scheduling tasks, and then use the target scheduling mathematical model, data query and call, solver component, etc. of the corresponding scenario to generate an operation and maintenance scheduling plan for the multi-vehicle operation and maintenance scheduling task, that is, as Figure 4 shown.

[0126] The embodiments of the present application can build an equipment operation and maintenance integrated management system based on the equipment operation and maintenance integrated management method. Thus, based on the large language model, the services in the tool code library are registered in the callable tool list of the large language model, and an interactive UI usage interface is designed to facilitate users to input natural language, so as to accurately call operation and maintenance integrated management components such as the spare parts inventory management model and the task planning and scheduling model in the way of natural language interaction and multi-round dialogue, generate an optimal operation and maintenance scheduling strategy or spare parts ordering strategy, and send the final operation and maintenance scheduling strategy plan to the operation and maintenance related staff in the form of intranet information, text messages, etc., effectively reducing the maintenance cost and risk in the equipment operation and maintenance process, and providing decision-making support for equipment operation and maintenance managers.

[0127] According to the equipment operation and maintenance integrated management method proposed by the embodiments of the present application, natural language can be input into the target large language model to extract the operation and maintenance scheduling tasks corresponding to the natural language, and an operation and maintenance scheduling plan and / or a spare parts ordering plan can be generated, and then the operation and maintenance scheduling plan and / or the spare parts ordering plan are executed by the target operation and maintenance personnel to complete the management of the equipment. Thus, it realizes the intelligent judgment and parsing of the operation and maintenance requirements expressed in natural language through the target large language model, and calls the tools predefined in the tool code library to quickly and accurately automatically generate optimized operation and maintenance scheduling plans and spare parts ordering plans according to the actual application scenario, effectively ensuring the generation efficiency and reliability of the operation and maintenance scheduling plan and the spare parts ordering plan; moreover, the present application can also design and develop an operation and maintenance integrated management system based on the target large language model and the tool code library, combine the technical advantages of the large language model and the solver, and more intelligently process and analyze operation and maintenance data and information, thereby effectively improving the digital and intelligent level of the model equipment health management system and the reliability of the equipment, extending the service life, and reducing the maintenance cost and risk, providing decision-making support for the intelligent operation and maintenance management of the equipment system. Thus, it solves the problems that the equipment operation and maintenance scheduling plan generation algorithm in the related technology is difficult to quickly and accurately generate practical operation and maintenance scheduling plans when facing large-scale, multi-objective, and complex constraint scheduling problems in the actual operation and maintenance management process, and is difficult to adjust the strategy in time to adapt to the new environmental conditions due to the large data scale and complex data relationship characteristics in the rapid change of the environment and the operation and maintenance decision-making process, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenario, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy.

[0128] Next, the equipment operation and maintenance integrated management device proposed according to the embodiments of the present application is described with reference to the accompanying drawings.

[0129] Figure 5 It is a structural schematic diagram of the equipment operation and maintenance integrated management device of the embodiments of the present application.

[0130] As Figure 5As shown, the integrated management device 10 for equipment operation and maintenance includes: an identification module 100, a first generation module 200, and a management module 300.

[0131] Among them, the identification module 100 is used to input natural language into a target large language model to identify the operation and maintenance scheduling tasks corresponding to the natural language, and extract the task objectives and task constraints of the operation and maintenance scheduling tasks.

[0132] The first generation module 200 is used to generate an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling tasks based on the task objectives and task constraints.

[0133] The management module 300 is used to send the operation and maintenance scheduling plan and / or the spare part ordering plan to the target operation and maintenance personnel, so that the target operation and maintenance personnel execute the operation and maintenance scheduling plan and / or the spare part ordering plan to complete the management of the equipment corresponding to the operation and maintenance scheduling tasks.

[0134] Optionally, in an embodiment of the present application, it further includes: a collection module, a construction module, and a second generation module.

[0135] Among them, the collection module is used to collect the tool operation instruction information of the operation and maintenance personnel before inputting the natural language into the target large language model, so as to generate a target instruction template according to the tool operation instruction information.

[0136] The construction module is used to generate tool call instructions and their corresponding response data based on the target instruction template by using a basic large language model, so as to construct an instruction corpus according to the tool call instructions and their corresponding response data.

[0137] The second generation module is used to correct the instruction corpus, so as to fine-tune the basic large language model according to the corrected instruction corpus to generate a target large language model.

[0138] Optionally, in an embodiment of the present application, the first generation module 200 includes: a first generation unit and a second generation unit.

[0139] Among them, the first generation unit is used to generate a tool code library according to the target instruction template, the target scheduling mathematical model, and the target solver component.

[0140] The second generation unit is used to generate an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling tasks by using the tool code library.

[0141] Optionally, in an embodiment of the present application, the first generation module 200 further includes: a first determination unit and a second determination unit.

[0142] Among them, the first determination unit is used to determine the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling tasks.

[0143] A second determination unit, configured to determine an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task according to the target scheduling mathematical model.

[0144] Optionally, in an embodiment of the present application, the first determination unit includes: a first determination subunit and a second determination subunit.

[0145] Wherein, the first determination subunit is configured to determine that the target scheduling mathematical model is a bilevel programming model when the actual scheduling scenario is a static scheduling scenario, where the bilevel programming model includes an integer linear programming problem model and a vehicle routing problem model with capacity constraints.

[0146] The second determination subunit is configured to determine that the target scheduling mathematical model is a dynamic vehicle routing problem model when the actual scheduling scenario is a dynamic scheduling scenario.

[0147] Optionally, in an embodiment of the present application, the recognition module 100 includes: a detection unit and an extraction unit.

[0148] Wherein, the detection unit is configured to detect whether the integrity of the natural language meets a preset integrity requirement.

[0149] The extraction unit is configured to extract the task objective and task constraints of the operation and maintenance scheduling task when the integrity meets the preset integrity requirement, otherwise, supplement the natural language until it meets the preset integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task.

[0150] It should be noted that the foregoing explanation of the embodiment of the equipment operation and maintenance comprehensive management method also applies to the equipment operation and maintenance comprehensive management device of this embodiment, and will not be repeated here.

[0151] The equipment operation and maintenance integrated management device proposed according to the embodiments of the present application can input natural language into a target large language model to extract the operation and maintenance scheduling tasks corresponding to the natural language, generate an operation and maintenance scheduling plan and / or a spare part ordering plan, and then complete the management of the equipment by the target operation and maintenance personnel executing the operation and maintenance scheduling plan and / or the spare part ordering plan. Thus, it realizes the intelligent judgment and analysis of operation and maintenance requirements expressed in natural language through the target large language model, and calls the tools predefined in the tool code library, and quickly and accurately automatically generates optimized operation and maintenance scheduling plans and spare part ordering plans according to the actual application scenario, effectively ensuring the generation efficiency and reliability of the operation and maintenance scheduling plan and the spare part ordering plan; moreover, the present application can also design and develop an operation and maintenance integrated management system based on the target large language model and the tool code library, combine the technical advantages of the large language model and the solver, and more intelligently process and analyze operation and maintenance data and information, thereby effectively improving the digital and intelligent level of the model equipment health management system and the reliability of the equipment, extending the service life, and reducing the maintenance cost and risk, providing decision-making support for the intelligent operation and maintenance management of the equipment system. Thus, it solves the problems that in the related art, the operation and maintenance scheduling plan generation algorithm is difficult to quickly and accurately generate a practical operation and maintenance scheduling plan when facing the scheduling problems with large scale, multiple objectives, and complex constraints in the actual operation and maintenance management process, and in the face of the rapid changes in the environment and the characteristics of large data scale and complex data relationships in the operation and maintenance decision-making process, it is difficult to adjust the strategy in time to adapt to the new environmental conditions, with poor dynamic adaptability and generalization, unable to effectively cover the actual application scenario, and unable to ensure the reliability and efficiency of the operation and maintenance scheduling strategy, etc.

[0152] Figure 6 The structural schematic diagram of the electronic device provided by the embodiments of the present application. The electronic device may include:

[0153] A memory 601, a processor 602, and a computer program stored on the memory 601 and executable on the processor 602.

[0154] When the processor 602 executes the program, it implements the equipment operation and maintenance integrated management method provided in the above embodiments.

[0155] Further, the electronic device further includes:

[0156] A communication interface 603 for communication between the memory 601 and the processor 602.

[0157] The memory 601 is used to store a computer program executable on the processor 602.

[0158] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0159] If the memory 601, the processor 602, and the communication interface 603 are implemented independently, the communication interface 603, the memory 601, and the processor 602 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity in representation, Figure 6 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0160] Optionally, in a specific implementation, if the memory 601, the processor 602, and the communication interface 603 are integrated on a single chip, the memory 601, the processor 602, and the communication interface 603 can communicate with each other through an internal interface.

[0161] The processor 602 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0162] The embodiments of the present application further provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned comprehensive management method for equipment operation and maintenance is implemented.

[0163] The embodiments of the present application further provide a computer program product, including a computer program, and the computer program can run computer instructions, and when the computer instructions are executed by a processor, the comprehensive management method for equipment operation and maintenance provided by the embodiments of the present application is implemented.

[0164] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0165] In addition, the terms "first" and "second" are used only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of this application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0166] Any process or method description shown in a flowchart or described in other ways herein can be understood as representing a module, segment, or part of code including one or N executable instructions for implementing a customized logical function or process, and the scope of the preferred embodiments of this application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in the reverse order according to the involved functions, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of this application belong.

[0167] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in conjunction with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or N wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0168] It should be understood that various parts of the present application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0169] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method of implementing the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0170] In addition, each functional unit in various embodiments of the present application may be integrated into a processing module, or each unit may exist physically alone, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0171] The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disc, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An integrated management method for equipment operation and maintenance, characterized in that, Including the following steps: Input the natural language into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and extract the task objective and task constraints of the operation and maintenance scheduling task; Based on the task objective and the task constraints, generate an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling task; Send the operation and maintenance scheduling plan and / or spare part ordering plan to the target operation and maintenance personnel, so that the target operation and maintenance personnel execute the operation and maintenance scheduling plan and / or spare part ordering plan to complete the management of the equipment corresponding to the operation and maintenance scheduling task.

2. The method according to claim 1, wherein Before inputting the natural language into the target large language model, it further includes: Collect the tool operation instruction information of the operation and maintenance personnel to generate a target instruction template according to the tool operation instruction information; Based on the target instruction template, use the basic large language model to generate tool call instructions and their corresponding response data, and construct an instruction corpus according to the tool call instructions and their corresponding response data; Correct the instruction corpus, and fine-tune the basic large language model according to the corrected instruction corpus to generate the target large language model.

3. The method according to claim 2, wherein The generating an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling task based on the task objective and the task constraints includes: Generate a tool code library according to the target instruction template, target scheduling mathematical model and target solver component; Use the tool code library to generate an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling task.

4. The method according to claim 3, characterized in that, The generating an operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling task based on the task objective and the task constraints further includes: Determine the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task; Determine the operation and maintenance scheduling plan and / or spare part ordering plan for the operation and maintenance scheduling task according to the target scheduling mathematical model.

5. The method according to claim 4, wherein The determining the target scheduling mathematical model according to the actual scheduling scenario of the operation and maintenance scheduling task includes: When the actual scheduling scenario is a static scheduling scenario, determine that the target scheduling mathematical model is a bilevel programming model, where the bilevel programming model includes an integer linear programming problem model and a vehicle routing problem model with capacity constraints; When the actual scheduling scenario is a dynamic scheduling scenario, determine that the target scheduling mathematical model is a dynamic vehicle routing problem model.

6. The method according to claim 1, characterized in that, The inputting the natural language into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and extracting the task objective and task constraints of the operation and maintenance scheduling task includes: Detect whether the integrity of the natural language meets the preset integrity requirement; When the integrity meets the preset integrity requirement, extract the task objective and task constraints of the operation and maintenance scheduling task, otherwise, supplement the natural language to meet the preset integrity requirement to extract the task objective and task constraints of the operation and maintenance scheduling task.

7. An integrated management device for equipment operation and maintenance, characterized in that, Including: An identification module for inputting the natural language into the target large language model to identify the operation and maintenance scheduling task corresponding to the natural language, and extracting the task objective and task constraints of the operation and maintenance scheduling task; A generation module, configured to generate an operation and maintenance scheduling plan and / or a spare part ordering plan for the operation and maintenance scheduling task based on the task objective and the task constraint; A management module, configured to send the operation and maintenance scheduling plan and / or the spare part ordering plan to the target operation and maintenance personnel, so as to manage the equipment corresponding to the operation and maintenance scheduling task by executing the operation and maintenance scheduling plan and / or the spare part ordering plan by the target operation and maintenance personnel.

8. An electronic device, characterized in that, Comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the equipment operation and maintenance comprehensive management method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to be used for implementing the equipment operation and maintenance comprehensive management method according to any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed, it is used for implementing the equipment operation and maintenance comprehensive management method according to any one of claims 1-6.

Images