A multi-level agent-based industrial system evaluation method and device

Abstract

The application discloses an industrial system evaluation method and device based on multi-level agents, and relates to the technical field of artificial intelligence. The method comprises the following steps: acquiring real-time data streams and / or historical data streams of each measuring point in each industrial equipment in an industrial system; inputting a measuring point evaluation prompt word into a measuring point agent corresponding to the measuring point, so as to obtain a measuring point evaluation result corresponding to the measuring point; inputting an equipment evaluation prompt word into an equipment agent corresponding to each industrial equipment, so as to obtain an equipment evaluation result corresponding to the industrial equipment; and inputting a system evaluation prompt word into a system agent corresponding to the industrial system, so as to obtain a system evaluation result corresponding to the industrial system. The above technical scheme realizes intelligent evaluation of the industrial system.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to an industrial system evaluation method and device based on multi-level intelligent agents. Background Technology

[0002] In the operation and maintenance of large industrial systems, staff typically check instruments on each device, summarize the health scores and alarm events of each device, analyze whether there are systemic problems such as regional temperature rises and power fluctuations, and submit the inspection results in the form of manual reports. The inspection results can be used to assess the overall operating status of the industrial system.

[0003] This assessment method is primarily human-based, typically relying on manual data collection and status assessment, which limits its efficiency and accuracy. Summary of the Invention

[0004] This invention provides an industrial system evaluation method and device based on multi-level intelligent agents to achieve automated evaluation of industrial systems.

[0005] According to one aspect of the present invention, an industrial system evaluation method based on multi-level intelligent agents is provided, comprising: Acquire real-time and / or historical data streams from various measuring points within industrial equipment in an industrial system; For each of the aforementioned measurement points, a measurement point evaluation prompt word is constructed based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule, and the measurement point evaluation prompt word template. The measurement point evaluation prompt word is then input into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point. For each of the industrial devices, an equipment evaluation prompt is constructed based on the evaluation results of the measurement points corresponding to each measurement point in the industrial device, the equipment information corresponding to the industrial device, the equipment evaluation rules, and the equipment evaluation prompt template. The equipment evaluation prompt is then input into the equipment agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device. Based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed. The system evaluation prompt words are then input into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0006] Furthermore, for each of the aforementioned industrial devices, the measuring points within the industrial device include at least the host output frequency, host output power, host speed, device status, and device operating conditions.

[0007] Further, based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rules, and the measurement point evaluation prompt word template, measurement point evaluation prompt words are constructed, including: The real-time data stream and / or the historical data stream corresponding to the measurement point, along with the measurement point evaluation rules, are filled into the measurement point evaluation prompt word template to obtain the measurement point evaluation prompt word. The measurement point evaluation prompt word template includes format limitations for the measurement point evaluation result.

[0008] Furthermore, the evaluation results of the measurement points include: The diagnostic conclusions include an overview of the data and a description of the trends, anomaly identification and location, anomaly cause inference, maintenance recommendations and handling steps, cross-measurement point linkage inspection recommendations, and verifiable evidence.

[0009] Furthermore, based on the measurement point evaluation results corresponding to each measurement point within the industrial equipment, the corresponding equipment information, equipment evaluation rules, and equipment evaluation prompt word template, equipment evaluation prompt words are constructed, including: The evaluation results of the measurement points corresponding to each measurement point in the industrial equipment, the corresponding equipment information of the industrial equipment, and the equipment evaluation rules are filled into the equipment evaluation prompt word template to obtain the equipment evaluation prompt word. The equipment evaluation prompt word template includes format restrictions for the equipment evaluation results.

[0010] Furthermore, the equipment evaluation results include: Basic equipment information, abnormal measurement points, fault type location, fault risk prediction, and equipment maintenance suggestions.

[0011] Furthermore, based on the equipment evaluation results corresponding to each piece of industrial equipment within the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed, including: The system evaluation results of each industrial device in the industrial system and the system evaluation rules of the industrial system are filled into the system evaluation prompt word template to obtain the system evaluation prompt word. The system evaluation prompt word template includes format restrictions for the system evaluation results.

[0012] Furthermore, the system evaluation results include: Overall operational conclusions, equipment operation overview, system-level correlations, comprehensive cause inferences, and risk assessment and recommendations.

[0013] Furthermore, the measurement point evaluation rules, the equipment evaluation rules, and the system evaluation rules are all determined based on the information input by the staff.

[0014] According to another aspect of the present invention, an industrial system evaluation apparatus based on multi-level intelligent agents is provided, comprising: The acquisition module is used to acquire real-time data streams and / or historical data streams from various measuring points within each piece of industrial equipment in the industrial system. The measurement point evaluation module is used to construct a measurement point evaluation prompt word for each measurement point based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule and the measurement point evaluation prompt word template, input the measurement point evaluation prompt word into the measurement point agent corresponding to the measurement point, and obtain the measurement point evaluation result corresponding to the measurement point. The equipment evaluation module is used to construct equipment evaluation prompts for each of the industrial equipment based on the evaluation results of the measurement points corresponding to each measurement point in the industrial equipment, the equipment information corresponding to the industrial equipment, the equipment evaluation rules, and the equipment evaluation prompt template, and input the equipment evaluation prompts into the equipment intelligent agent corresponding to the industrial equipment to obtain the equipment evaluation result corresponding to the industrial equipment. The system evaluation module is used to construct system evaluation prompts based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template, and input the system evaluation prompts into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0015] According to another aspect of the present invention, a computer device is provided, the computer device comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the industrial system evaluation method based on multi-level intelligent agents as described in any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the industrial system evaluation method based on multi-level intelligent agents as described in any embodiment of the present invention.

[0017] The technical solution of this invention firstly acquires the data streams corresponding to each measuring point within each industrial device in an industrial system, specifically real-time data streams and / or historical data streams. Secondly, it constructs measuring point evaluation prompts for each measuring point based on the real-time and / or historical data streams, measuring point evaluation rules, and measuring point evaluation prompt templates. These prompts are then input into the corresponding measuring point agent. The measuring point agent evaluates each measuring point based on these prompts, thus obtaining the measuring point evaluation results and achieving intelligent evaluation of each measuring point. Finally, the evaluation results and equipment information of each industrial device can be used to further refine the evaluation. Based on equipment evaluation rules and equipment evaluation prompt word templates, equipment evaluation prompt words are constructed for each industrial device. These prompt words are then input into the corresponding intelligent agent for each industrial device. The intelligent agent performs equipment evaluation based on these prompt words, thus obtaining the equipment evaluation results for each industrial device. This achieves intelligent evaluation of each industrial device. Furthermore, based on the system evaluation results and system evaluation prompt word templates for each industrial device within the industrial system, system evaluation prompt words are constructed. These prompt words are then input into the system intelligent agent, which performs system evaluation based on them, thus obtaining the system evaluation results. This achieves intelligent evaluation of the industrial system.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an industrial system evaluation method based on multi-level intelligent agents provided in Embodiment 1; Figure 2 This is a flowchart of an industrial system evaluation method based on multi-level intelligent agents provided in Embodiment 2; Figures 3a-3c A system interface diagram created for the intelligent agent provided in this second embodiment; Figures 4a-4e This is a system interface diagram for the system evaluation provided in Embodiment 2; Figures 5a-5cThis is a system interface diagram for querying evaluation results provided in Embodiment 2; Figure 6 This is a schematic diagram of the structure of an industrial system evaluation device based on a multi-level intelligent agent provided in Embodiment 3; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] Example 1 Figure 1 This is a flowchart of an industrial system evaluation method based on multi-level intelligent agents provided in Embodiment 1. This embodiment is applicable to situations requiring automated evaluation of industrial systems. The method can be executed by an industrial system evaluation device based on multi-level intelligent agents. This device can be implemented in hardware and / or software and can be configured in a computer device. Figure 1 As shown, the method includes: Step 110: Obtain the real-time data stream and / or historical data stream of each measuring point in each industrial device in the industrial system.

[0024] Industrial equipment can be understood as equipment within an industrial system. For example, an industrial system can be a home appliance manufacturing system, and industrial equipment can include production equipment and monitoring equipment. Measuring points within industrial equipment include at least the host output frequency, host output power, host speed, equipment status, and equipment operating conditions. These measuring points typically generate real-time data streams at a frequency of seconds or minutes, and a database can store the corresponding real-time data streams for these measuring points.

[0025] Specifically, during the operation of an industrial system, real-time data streams from various measuring points within each piece of industrial equipment can be acquired in real time, and historical data streams from each measuring point can also be retrieved from a database. Evaluation of measuring points can typically be performed using real-time data streams and / or historical data streams. Therefore, when evaluating measuring points, one can acquire real-time data streams, or historical data streams, or both.

[0026] For each measuring point agent, the tool can acquire real-time and / or historical data streams of the measuring points based on the Model Context Protocol (MCP) corresponding to the measuring point agent.

[0027] In this embodiment of the invention, the data stream corresponding to each measuring point within each industrial device in an industrial system is acquired.

[0028] Step 120: For each measurement point, construct a measurement point evaluation prompt word based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule, and the measurement point evaluation prompt word template. Input the measurement point evaluation prompt word into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point.

[0029] With the continuous iteration of artificial intelligence technology, large models have rapidly emerged due to their powerful general reasoning, knowledge memory and cross-scenario adaptability, providing a core capability foundation for intelligent systems; intelligent agent technology has become a key carrier for achieving goal-oriented intelligent tasks through autonomous planning, tool invocation and closed-loop interaction mechanisms.

[0030] The measurement point intelligent agent uses measurement point assessment prompts as trigger conditions to complete measurement point assessment through three core steps: data acquisition and knowledge base retrieval, specifically by acquiring real-time and / or historical data streams of the measurement points using the MCP tool, and also by retrieving the corresponding measurement point assessment rules from the measurement point knowledge base; knowledge fusion and reasoning analysis, specifically by integrating real-time and / or historical data streams and measurement point assessment rules to conduct multi-dimensional analysis, covering anomaly judgment and quantitative description, anomaly authenticity verification, anomaly root cause identification and causal chain construction, anomaly propagation risk assessment, and health status quantitative assessment; and measurement point assessment result output, specifically outputting measurement point assessment results, including a summary of diagnostic conclusions, data overview and trend description, anomaly identification and location, anomaly cause inference, maintenance suggestions and handling steps, cross-measurement point linkage inspection suggestions, and verifiable evidence.

[0031] In practical applications, the MCP tool supports flexible time parameter configuration for acquiring real-time and / or historical data streams from measurement points. The data time period can be set to time-series data segments of seconds, minutes, or hours, and the time granularity can be configured as needed to be at the second, minute, or hour level, which can adapt to the differentiated data acquisition needs of different industrial system evaluation scenarios.

[0032] Specifically, for each measurement point, the measurement point evaluation rules corresponding to the measurement point can first be obtained from the knowledge base. Then, measurement point evaluation prompts can be constructed based on the real-time data stream and / or historical data stream corresponding to the measurement point, the measurement point evaluation rules, and the measurement point evaluation prompt template. Specifically, the real-time data stream and / or historical data stream and the measurement point evaluation rules can be filled into the measurement point evaluation prompt template according to the prompts. This will yield the measurement point diagnostic prompts. The measurement point diagnostic prompts can then be input into the measurement point agent. The measurement point agent can perform multi-dimensional analysis of the measurement point based on the measurement point diagnostic prompts. Specifically, this includes anomaly judgment and quantitative description, anomaly authenticity verification, anomaly root cause identification and causal chain construction, anomaly propagation risk assessment, and health status quantitative assessment. This will yield the measurement point evaluation results, which may include a summary of diagnostic conclusions, data overview and trend description, anomaly identification and location, anomaly cause inference, maintenance suggestions and handling steps, cross-measurement point linkage inspection suggestions, and verifiable evidence.

[0033] In this embodiment of the invention, after constructing the measurement point evaluation prompt words based on the real-time data stream and / or historical data stream corresponding to the measurement point, the measurement point evaluation rules, and the measurement point evaluation prompt word template, the measurement point evaluation prompt words are input into the measurement point intelligent agent. The measurement point intelligent agent performs measurement point evaluation based on the measurement point evaluation prompt words, and the measurement point evaluation result can be obtained, thereby realizing intelligent evaluation of the measurement point.

[0034] Step 130: For each of the industrial devices, construct equipment evaluation prompts based on the evaluation results of the measurement points corresponding to each measurement point within the industrial device, the equipment information corresponding to the industrial device, the equipment evaluation rules, and the equipment evaluation prompt template. Input the equipment evaluation prompts into the equipment agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device.

[0035] As the core of equipment diagnosis, the intelligent agent of the equipment outputs the evaluation results of each measurement point in the industrial equipment. Based on the attached equipment knowledge base, it completes the structured integration and reliable quantification of multi-source information. It can also carry out cross-measurement point anomaly correlation analysis and reasoning and equipment-level anomaly root cause localization. On this basis, it can formulate scientific and feasible maintenance strategies, generate standardized maintenance operation procedures, and finally output complete equipment evaluation results.

[0036] Specifically, for each piece of industrial equipment, the equipment evaluation rules corresponding to the industrial equipment can first be obtained from the knowledge base. Then, equipment evaluation prompts can be constructed based on the evaluation results of each measurement point within the industrial equipment, the corresponding equipment information, the equipment evaluation rules, and the equipment evaluation prompt template. Specifically, the evaluation results of each measurement point within the industrial equipment, the corresponding equipment information, and the equipment evaluation rules can be filled into the equipment evaluation prompt template according to the prompts. This will yield the equipment evaluation prompts. These prompts can then be input into the equipment agent, which can perform multi-dimensional analysis of the industrial equipment based on the prompts. Specifically, this includes cross-measurement point correlation analysis, abnormal data capture, fault type localization, fault root cause analysis, and equipment maintenance suggestions. The resulting equipment evaluation results can include basic equipment information, fault type localization results, fault root cause analysis results, fault risk prediction, and equipment maintenance suggestions.

[0037] In this embodiment of the invention, after constructing equipment evaluation prompts based on the evaluation results of each measurement point within the industrial equipment, the corresponding equipment information, equipment evaluation rules, and equipment evaluation prompt templates, the equipment evaluation prompts are input into the equipment intelligent agent. The equipment intelligent agent then performs equipment evaluation based on the equipment evaluation prompts to obtain the equipment evaluation results, thereby achieving intelligent evaluation of each industrial piece of equipment.

[0038] Step 140: Construct system evaluation prompts based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template. Input the system evaluation prompts into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0039] The system's intelligent agent is equipped with a corresponding knowledge base and possesses experience in industrial system health assessment and diagnosis. It can handle the operational analysis and troubleshooting of complex industrial systems. Based on the equipment assessment results of multiple industrial devices, it can identify the overall status of the industrial system, anomaly distribution patterns, potential risks, and system-level problems, achieving a comprehensive diagnosis of the overall operational status of the industrial system and outputting complete system assessment results.

[0040] Specifically, firstly, system evaluation rules corresponding to the industrial system can be obtained from the knowledge base. Secondly, system evaluation prompts can be constructed based on the system evaluation results of each industrial device within the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template. Specifically, the system evaluation results of each industrial device within the industrial system and the system evaluation rules of the industrial system can be filled into the system evaluation prompt template according to the prompts. This will yield the system evaluation prompts. Then, the system evaluation prompts can be input into the system agent. The system agent can perform multi-dimensional analysis of the industrial system based on the system evaluation prompts. Specifically, it can perform anomaly distribution analysis, equipment anomaly correlation inference, equipment anomaly root cause analysis, and system improvement measures, etc., thereby obtaining the system evaluation results. The system evaluation results can include overall operation conclusions, equipment operation overview, system-level correlation analysis, comprehensive cause inference, and risk assessment and recommendations, etc.

[0041] In this embodiment of the invention, after constructing system evaluation prompts based on the system evaluation results and system evaluation prompt templates corresponding to each industrial device in the industrial system, the system evaluation prompts are input into the system intelligent agent. The system intelligent agent performs system evaluation based on the system evaluation prompts to obtain the system evaluation results, thereby realizing intelligent evaluation of the industrial system.

[0042] The industrial system evaluation method based on multi-level intelligent agents provided in this embodiment includes: acquiring real-time data streams and / or historical data streams of each measuring point within each industrial device in the industrial system; for each measuring point, constructing a measuring point evaluation prompt word based on the real-time data stream and / or historical data stream corresponding to the measuring point, measuring point evaluation rules, and a measuring point evaluation prompt word template, and inputting the measuring point evaluation prompt word into the measuring point intelligent agent corresponding to the measuring point to obtain the measuring point evaluation result corresponding to the measuring point; for each industrial device, constructing an equipment evaluation prompt word based on the measuring point evaluation results corresponding to each measuring point within the industrial device, the corresponding equipment information of the industrial device, equipment evaluation rules, and an equipment evaluation prompt word template, and inputting the equipment evaluation prompt word into the equipment intelligent agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device; and constructing a system evaluation prompt word based on the equipment evaluation results corresponding to each industrial device within the industrial system, the system evaluation rules of the industrial system, and a system evaluation prompt word template, and inputting the system evaluation prompt word into the system intelligent agent corresponding to the industrial system to obtain the system evaluation result corresponding to the industrial system. The above technical solution first acquires the data streams corresponding to each measuring point within each industrial device in the industrial system, specifically real-time and / or historical data streams. Second, it constructs measuring point evaluation prompts for each measuring point based on the real-time and / or historical data streams, measuring point evaluation rules, and measuring point evaluation prompt templates. These prompts are then input into the corresponding measuring point agent. The measuring point agent evaluates each measuring point based on these prompts, thus obtaining the measuring point evaluation results and achieving intelligent evaluation of each measuring point. Finally, the solution can be further developed based on the measuring point evaluation results, equipment information, and other relevant data within each industrial device. The system constructs equipment evaluation prompts for each industrial device using equipment evaluation rules and prompt templates. These prompts are then input into the corresponding intelligent agents for each device. The intelligent agents perform equipment evaluations based on these prompts, yielding evaluation results for each industrial device. This enables intelligent evaluation of each industrial device. Furthermore, the system can construct system evaluation prompts based on the system evaluation results and prompt templates for each industrial device within the industrial system. These prompts are input into the system intelligent agents, which then perform system evaluations based on these prompts, yielding system evaluation results. This also enables intelligent evaluation of the industrial system.

[0043] Example 2 Figure 2 This is a flowchart illustrating an industrial system evaluation method based on multi-level intelligent agents, provided in Embodiment 2. This embodiment is a specific implementation based on the above embodiments. Figure 2As shown, the method includes: Step 210: Obtain the real-time data stream and / or historical data stream of each measuring point in each industrial device in the industrial system.

[0044] Step 220: For each measurement point, construct a measurement point evaluation prompt word based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rules, and the measurement point evaluation prompt word template.

[0045] In one implementation, step 220 may specifically include: The real-time data stream and / or the historical data stream corresponding to the measurement point, along with the measurement point evaluation rules, are filled into the measurement point evaluation prompt word template to obtain the measurement point evaluation prompt word.

[0046] The measurement point assessment prompt template includes format restrictions for the measurement point assessment results. Specifically, it can be: based on the data stream returned by the MCP tool, analyze the operational status of the measurement points and locate problems, and provide scientific, reliable and engineering-based measurement point assessment results. The measurement point assessment results need to meet specific format restrictions, which can include output format requirements, field descriptions and output specifications. Field descriptions can include a summary of the corresponding diagnostic conclusions in the field, a data overview and trend description, anomaly identification and location, anomaly cause inference, maintenance suggestions and handling steps, cross-measurement point linkage inspection suggestions, and specific explanations of verifiable evidence.

[0047] For example, `status` (string): indicates the device's operating status; "0" represents normal operation, and "1" represents an abnormality. `last_value` (string): extracts the value at the last moment in the data stream returned by the tool. If the tool returned no data (returning "0" or "0.0" is not considered as no data returned), it will be null. `score` (string): calculates a health score based on the device's operating status as reflected in the tool's data. The score ranges from 0 to 100, with higher scores indicating better device health. `result` (string): provides a scientifically reliable diagnostic conclusion based on the tool's data, including main conclusions and key information, strictly limited to 40-60 Chinese characters. `detail_result` (string): provides detailed report content, as comprehensive as possible to facilitate subsequent comprehensive analysis by the model.

[0048] Determining the measurement point assessment results requires strictly adhering to the following assessment process: For anomaly judgment and quantitative description: anomalies are judged based on the measurement point data stream, preset static thresholds, and dynamic operating range; quantification is mandatory, including: key indicators and actual values, deviation from normal range, anomaly timestamps, duration, and trend. For anomaly authenticity verification: clearly determine whether the anomaly is a genuine fault signal or a pseudo-anomaly such as noise or short-term operating condition fluctuations, and provide matching criteria. For anomaly root cause identification and causal chain construction: provide the essential cause of the anomaly, establish a cause-phenomenon causal link; specify the related components, material conditions (wear, corrosion, etc.), and operating conditions (load, voltage, environment, etc.). For anomaly propagation risk assessment: determine whether the anomaly is a local problem or can propagate to the whole; describe the affected components / systems / processes, and provide a conclusion on whether to initiate cross-measurement point collaborative diagnosis. For health status quantitative assessment: output health index (e.g., 92→75), health level, and trend (slow decline / accelerated deterioration / tending to stability), and provide a judgment on equipment reliability degradation.

[0049] The measurement point evaluation rules corresponding to the measurement points are obtained from the measurement point knowledge base.

[0050] Specifically, for each measurement point, the real-time data stream and / or historical data stream corresponding to the measurement point obtained based on the MCP tool, as well as the measurement point evaluation rules corresponding to the measurement point obtained from the measurement point knowledge base, can be filled into the measurement point evaluation prompt word template to obtain the measurement point evaluation prompt words corresponding to the measurement point.

[0051] In this embodiment of the invention, the construction of measurement point evaluation prompts corresponding to the measurement points is realized.

[0052] Step 230: Input the measurement point evaluation prompt word into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point.

[0053] Specifically, after inputting the measurement point assessment prompts into the corresponding measurement point agent, the measurement point agent can perform multi-dimensional analysis of the measurement point based on the diagnostic prompts. This includes anomaly detection and quantitative description, anomaly authenticity verification, anomaly root cause identification and causal chain construction, anomaly propagation risk assessment, and health status quantitative assessment. When detecting and quantitatively describing anomalies, based on the collected data stream and user-preset diagnostic requirements such as static thresholds and dynamic operating ranges for the measurement point, potential anomalies are identified. The resulting anomalies are then objectively quantitatively described, including key indicators and values ​​triggering the anomaly, and deviations from the normal range of these indicators. The magnitude (e.g., exceeding the threshold by 20%, falling below the lower limit by 15%), the timestamp and duration of the anomaly, and the trend of indicator changes (e.g., sudden rise, gradual decline, periodic fluctuations) can be used to verify the authenticity of anomalies. This allows us to determine whether the current anomaly matches the fault characteristics in the knowledge base, clarifying whether it is a real fault signal or a pseudo-anomaly caused by noise or short-term operating condition fluctuations, thus improving diagnostic reliability from the source. When identifying the root cause of anomalies and constructing causal chains, we can explain the essential causes of anomalies (e.g., temperature rise caused by increased load, pressure drop due to aging and leakage of seals), establishing a clear causal link from cause to phenomenon. Further analysis of key components, material conditions (wear, corrosion, etc.), and operating conditions (load, voltage, environment, etc.) with high probability of correlation provides clear direction for subsequent diagnosis. When conducting anomaly propagation risk assessment, it can determine whether the anomaly is a localized problem limited to a single measuring point or may spread and affect the overall operation of the equipment, identifying potentially affected components, systems, and processes (e.g., sensor anomalies causing control module adjustment deviations, leading to actuator overload), and providing a basis for decision-making regarding whether to initiate cross-measuring point collaborative diagnosis. When conducting quantitative health status assessment, the degree of anomaly deviation and failure development rate can be combined to evaluate the health index and its changing trend at that measuring point (e.g., a drop from 92 points to 75 points), clarifying the range of health level decline and future reliability degradation (e.g., slow decline or accelerated deterioration), providing intuitive quantitative results for equipment health management.

[0054] After the above analysis, the measurement point intelligent agent can output measurement point evaluation results, which may include a summary of diagnostic conclusions, data overview and trend description, anomaly identification and location, anomaly cause inference, maintenance suggestions and handling steps, cross-measurement point linkage inspection suggestions, and verifiable evidence. Specifically, the diagnostic conclusion summary can include the main conclusions and key information, within 20-40 words (normal, slightly abnormal, faulty, deteriorating, and expected to worsen further within the next x minutes); the data overview and trend description can include a description of whether the measurement point data is complete, a list of key statistical indicators of the sequence data (including maximum, minimum, average, and standard deviation, etc.), and a description of the data's changing trend and rate of change; anomaly identification and localization can describe the abnormal intervals and types of anomalies in the measurement point time series data, extract key features, assess the duration, and predict possible risk development trends; anomaly cause inference can list possible causes of anomalies, clarify the judgment criteria, and provide the confidence level of each cause; maintenance suggestions and handling steps can provide targeted maintenance solutions and establish standardized maintenance procedures; cross-measurement point linkage inspection suggestions can clarify other measurement points that need to be inspected collaboratively, key indicators, inspection time windows, and judgment criteria; verifiable evidence can provide traceable evidence to support the diagnostic conclusions, including the original data stream of measurement points with marked abnormal intervals, knowledge base matching entries, etc., ensuring that the entire diagnostic process is traceable and verifiable.

[0055] The measurement point intelligent agent automatically retrieves, semantically matches, and fuses reasoning data with measurement point-level structured knowledge (historical cases, rules, structural associations, etc.), and simultaneously combines static thresholds, dynamic ranges, and knowledge matching results when determining anomalies. Measurement point evaluation results, based on both data characteristics and domain knowledge, can distinguish between real faults and short-term noise or operating condition fluctuations, significantly reducing false alarms and false negatives. Verifiable results can be quickly verified manually by outputting key indicators, deviation levels, durations, and evidence chains.

[0056] In this embodiment of the invention, by inputting measurement point evaluation prompts into the measurement point agent, the measurement point agent can evaluate the measurement point based on the measurement point evaluation prompts and obtain the measurement point evaluation results, thereby realizing intelligent evaluation of the measurement point.

[0057] Step 240: For each of the industrial devices, construct equipment evaluation prompts based on the evaluation results of the measurement points corresponding to each measurement point within the industrial device, the corresponding equipment information, equipment evaluation rules, and equipment evaluation prompt template.

[0058] The equipment evaluation prompt template includes format constraints for the equipment evaluation results, which can include background, role positioning, objectives, constraints, and output format.

[0059] The background could be: generating equipment assessment results for industrial equipment. These results, based on provided equipment information and diagnostic items, would produce structured and professional diagnostic conclusions. Your role could be that of an industrial equipment diagnostic expert, proficient in transforming sensor data and diagnostic items into clear, actionable conclusions and recommendations, ensuring the report is highly professional and practical. The goal could be to strictly adhere to the specified template to generate a diagnostic report, with the report title including the equipment name; and to fully utilize the provided "Diagnostic Report Requirements" and "Diagnostic Item Analysis" to ensure the generated conclusions are professional, credible, and logically rigorous. Constraints may include: 1) Strictly follow the specified OutputFormat to maintain structural consistency; 2) The language should be concise, clear, and precise, with standardized terminology to avoid ambiguity; 3) Note that "Diagnostic agent failure" should not be considered abnormal data; 4) If no "Diagnostic report requirements" are provided, this section should be ignored, but it will not affect the generation of other content; 5) The diagnostic time should use the "report generation time" provided in the "Maintenance Personnel Information" section above; 6) The "Data Source" should be extracted from the diagnostic item name, and multiple diagnostic item names should be connected with commas ","; 7) Second-level headings and column punctuation should be kept consistent with the template; 8) Note: When the number of abnormal core parameters is 0, the "Abnormal Data Capture" section of the report should not display any "Name" or "Result."

[0060] The output format can include basic equipment information, abnormal data capture, evaluation results, and equipment maintenance suggestions + re-inspection plan. The basic equipment information is obtained by the backend program and organized based on the equipment intelligent agent. Specifically, it can include evaluation time, evaluation object, measurement point number, measurement point intelligent agent, and data source. Abnormal data capture can include the following N core parameters that are abnormal when continuously monitoring the data stream for 1 hour: First abnormal parameter: abnormal result; Second abnormal parameter: abnormal result... Nth abnormal parameter: abnormal result. The evaluation results include fault type location, fault root cause analysis and fault risk prediction. Fault type location requirements: (1) Focus on the “cross-measurement point linkage inspection suggestions” in the output of the measurement point intelligent agent and carry out cross-measurement point abnormal correlation analysis; (2) Infer the fault propagation path and identify the abnormal source measurement point, key propagation measurement point, and affected terminal measurement point; (3) The output content must include: fault type, inference confidence level, and supporting basis (list key evidence, such as: abnormal measurement point information, cross-measurement point correlation results, and propagation path); (4) When there is no abnormal data, it is uniformly stated as: After AI diagnosis, all core parameters of the equipment are within the standard range, there is no fault risk, the operating status is good, and the diagnostic confidence level is 100%. Fault root cause analysis requirements: (1) At least two root causes must be provided, sorted from high to low probability of impact; (2) Each root cause must be supported by evidence (from abnormal data of measurement points or cross-measurement point linkage analysis), and the root cause measurement points and derivation logic must be clearly defined; (3) When there is no abnormal data, it should be uniformly stated as: the equipment is operating stably and no faults have occurred, root cause analysis is not required, and it is recommended to maintain the existing operation and maintenance rhythm. Fault risk prediction requirements: (1) Distinguish between short-term risks (such as downtime or decreased production efficiency within 12-24 hours) and long-term risks (such as life loss, quality fluctuations, and increased maintenance costs); (2) Potential risk measurement points (measurement points that are currently normal but affected by abnormalities) need to be predicted; (3) When there is no abnormal data, it should be uniformly stated as: there are no operational risks in the short term, and the equipment can be put into production normally; in the long term, if routine maintenance is maintained, various potential faults can be effectively avoided, production continuity can be guaranteed, and there will be no additional cost losses. Equipment maintenance recommendations should include: providing specific and actionable maintenance suggestions based on abnormal data, diagnostic conclusions, and root cause analysis, clearly specifying maintenance steps, required tools / consumables, and precautions, and prioritizing them. If no abnormal data is available, provide recommendations for routine equipment maintenance, specifying the maintenance cycle and content. Re-inspection plan requirements: Based on the maintenance recommendations (or routine maintenance recommendations), clearly specify the re-inspection time, re-inspection items, re-inspection standards, and re-inspection personnel (which can be labeled as {re-inspection personnel}), ensuring that the maintenance effect is verifiable and the equipment status is traceable. Re-inspection items must correspond to diagnostic items and abnormal parameters, and the re-inspection standards must clearly define the acceptable range.

[0061] The equipment evaluation rules for industrial equipment are obtained from the equipment knowledge base.

[0062] Specifically, for each piece of industrial equipment, the evaluation results of each measurement point within the industrial equipment, the corresponding equipment information, and the equipment evaluation rules obtained from the equipment knowledge base can be filled into the equipment evaluation prompt word template to obtain the corresponding equipment evaluation prompt words for the industrial equipment.

[0063] In this embodiment of the invention, the construction of equipment evaluation prompts for industrial equipment is realized.

[0064] Step 250: Input the equipment evaluation prompt into the equipment agent corresponding to the industrial equipment to obtain the equipment evaluation result corresponding to the industrial equipment.

[0065] Specifically, after inputting the equipment assessment prompts into the corresponding intelligent agent of the industrial equipment, the intelligent agent can perform multi-dimensional analysis of the industrial equipment based on the equipment assessment prompts. Specifically, it can perform cross-measurement point correlation analysis, abnormal data capture, fault type location, fault root cause analysis, and equipment maintenance suggestions, thereby obtaining equipment assessment results. The equipment assessment results can include basic equipment information, fault type location results, fault root cause analysis results, fault risk prediction, and equipment maintenance suggestions.

[0066] Basic equipment information can include diagnosis time, equipment name, equipment model, equipment code, diagnosis execution entity (equipment agent / human diagnostic personnel), and a list of diagnostic coverage test points (including the names of all test points involved in the diagnosis). Anomaly data capture, by fusing the output results of various test point agents, describes the situation of all abnormal test points, including test point name, anomaly overview, anomaly characteristics, anomaly range, anomaly type, and data trend; and based on the above information, provides a scientifically reliable equipment diagnostic conclusion. When locating fault types, the equipment agent, leveraging the reasoning capabilities of a large model, receives the output results of various test point agents, paying particular attention to "cross-test point linkage inspection suggestions," to perform cross-test point anomaly correlation analysis, identify fault propagation paths and patterns, and determine which test point anomalies originate from chain reactions of core test points and which are independent faults. The equipment diagnostic agent traces the fault propagation path of abnormal measuring points based on the correlation between measuring points (e.g., "abnormal pressure at measuring point 1 → flow fluctuation at measuring point 2 → temperature increase at measuring point 3 → decreased efficiency of the core system"), quantifies the scope, speed, and impact of the anomaly propagation, and identifies the "origin measuring point," "critical propagation measuring point," and "affected terminal measuring point." The deduction process outputs the most likely root cause, associated anomalies, and the links between affected measuring points, achieving systematic judgment and precise location of anomalies at multiple measuring points. Finally, based on the reasoning and analysis results, the equipment diagnostic agent outputs the "Fault Type Location" section of the report, specifically including: Fault Type: Based on fault mode reasoning, determining whether it is a single-point independent fault, a multi-point related fault, or a system-level root cause fault. Inference Confidence: Based on data support and the reasoning process, the agent provides the inference confidence for each fault type, indicating the reliability of the diagnostic results. Supporting Evidence: The agent provides detailed key evidence supporting the conclusions, including anomaly information from abnormal measuring points, cross-measuring point correlation analysis results, and anomaly propagation paths, helping to verify and interpret the diagnostic results. When conducting root cause analysis, the equipment agent first provides at least two possible root causes and ranks them according to their probability of impact to ensure that the most likely cause is addressed first. Each root cause analysis is accompanied by supporting evidence, which typically comes from abnormal data from measurement points or cross-measure point linkage analysis. Secondly, the agent determines the root cause measurement point and outputs a detailed chain of evidence explaining how the root cause of the fault was derived through association rules and data analysis. When predicting fault risks, the equipment agent combines the operational trends of abnormal and normal measurement points, historical global fault cases, and current anomaly propagation patterns to predict measurement points or systems that are not yet abnormal but have potential risks. This primarily involves predicting short-term and long-term risks of faults. Short-term risks mainly involve potential equipment downtime or decreased production efficiency, affecting recent production operations; long-term risks include equipment lifespan depletion, fluctuations in production quality, and increased maintenance costs, which may adversely affect the long-term stability and maintenance costs of the equipment.For example, "Measurement point E is currently normal, but due to the abnormality of measurement point F, the health index of measurement point E is expected to drop to a critical value within the next 12 hours." Through this prediction, the equipment diagnostic agent can help engineers identify potential fault risks in advance, reducing the possibility of unexpected downtime and production interruptions. When determining equipment maintenance recommendations, the equipment agent ensures the accuracy and effectiveness of maintenance through a multi-dimensional solution matching mechanism: automatically linking expert maintenance experience databases, standard fault handling specifications, and equipment maintenance manual operation guidelines to form a scientifically feasible maintenance strategy and generate a standardized maintenance process, specifically including the following core elements: Maintenance priority ranking: following the principle of "root cause first, then correlation," prioritizing the handling of core measurement points causing the fault, and then gradually resolving related abnormal measurement points. Precise maintenance solutions: outputting equipment handling measures, standardized operating procedures, and key risk warnings. For example: developing enhanced monitoring and early warning plans for potential risk measurement points; generating collaborative maintenance strategies for cross-measurement point related faults (example: "When repairing measurement point B, simultaneously check measurement points C and D to reduce equipment downtime and improve maintenance efficiency"). Measurement point retest plan: The equipment diagnostic intelligent agent simultaneously formulates a full-cycle retest plan, clarifies the measurement points that must be retested after repair and the judgment criteria, and recommends a reasonable observation time window based on the equipment operating characteristics to ensure that the maintenance effect meets the standards and the equipment operates stably for a long time.

[0067] The intelligent device performs structured integration, confidence quantification, and cross-measurement point correlation reasoning on the measurement point evaluation results corresponding to each measurement point, automatically generating a causal chain of "source anomaly → propagation chain → affected end". It can organize scattered measurement point anomalies into complete fault propagation paths, distinguish between independent anomalies and chain effects, and accurately locate the root cause of the equipment.

[0068] In this embodiment of the invention, by inputting equipment evaluation prompts into the equipment intelligent agent, the equipment intelligent agent evaluates the industrial equipment based on the equipment evaluation prompts and obtains the equipment evaluation results, thereby realizing intelligent evaluation of each piece of industrial equipment.

[0069] Step 260: Construct system evaluation prompts based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template.

[0070] The system evaluation prompt template includes format constraints for the system evaluation results, which may include role settings, task objectives, core analysis requirements, output format, and core user requirements.

[0071] The role / equipment expert can be a senior industrial station diagnostic specialist with extensive experience in equipment group health assessment. They are proficient in the operational logic, equipment linkages, and typical failure modes of various industrial systems. Based on the diagnostic results of multiple industrial devices, they can comprehensively identify the overall operational status, anomaly distribution patterns, potential risks, and system-level problems of the entire station. They strictly adhere to professional technical specifications and write accurate and rigorous station inspection and diagnostic reports in a standardized format.

[0072] The task objective is to conduct a comprehensive diagnostic analysis of the entire station based on the equipment evaluation results of multiple industrial devices, focusing on identifying cross-device related anomalies, common-cause anomalies, and system-level problems. The final output is a structured and professional system evaluation result (Markdown format), ensuring that all parts of the report fully meet the preset requirements and provide accurate basis for management decision-making and operation and maintenance personnel operations.

[0073] Core analysis requirements may include: Core principle: The output content must fully comply with the specific provisions of each chapter of the report in the subsequent core user requirements, without omitting any key points or deviating from any requirements. Equipment operation overview specific requirements: In the second section of the report, "Equipment Operation Overview," the description of the "Current Status" field must be completely consistent with the value of the corresponding "Equipment Status" field in the input "Diagnostic Results of All Equipment." Secondary processing, modification, abbreviation, supplementary explanation, or any form of adjustment is strictly prohibited to ensure the accuracy and consistency of the data. Global comprehensive analysis: Systematically summarize and aggregate the diagnostic results of all equipment, and combine the equipment health status, anomaly distribution, and anomaly severity to determine the overall health status of the entire station, providing support for the "Overall Operation Conclusion." Correlation analysis: Deeply explore the abnormal linkage relationships between equipment, accurately identify the related scenarios of "one equipment anomaly causing other equipment anomalies" (such as a high temperature of a certain equipment causing system pressure fluctuations, or a failure of a certain equipment causing abnormal load of other equipment), and clarify the linkage logic. Pattern Recognition: Based on multi-device anomaly data, determine whether there are systemic problems (such as insufficient cooling system, power fluctuations, uneven gas supply, load imbalance, etc.), and clarify the manifestation and scope of impact of system-level problems. Key Risk Extraction: Extract core anomalies and potential safety hazards across the entire station, clearly label the severity level (minor / moderate / severe) and handling priority of each type of anomaly, highlighting key areas of focus. Cause Inference: Combining the common characteristics, temporal relationships, and operating principles of multiple device anomalies, infer possible system-level root causes, and label the confidence level (high / medium / low) of each inference result to provide a reliable basis for subsequent risk assessment and recommendations. Report Quality Requirements: The content should be professional and standardized, logically rigorous and consistent, and accurately expressed without redundant statements; the tone should conform to the rigorous and objective style of industrial technical reports, avoiding colloquial or vague expressions.

[0074] The output format includes: overall operational conclusions, equipment operation overview, system-level correlation analysis, comprehensive cause inference, and risk assessment and recommendations. The overall operational conclusions specifically include: Operational status: normal / overall stable operation with minor anomalies / system-level anomalies detected requiring attention / serious anomalies; Anomaly distribution: clearly stating the number of devices involved, specific device names, and concentrated areas (e.g., gas source processing area, cooling system area, etc.); Main manifestations: a brief and accurate description of the specific phenomena of the main anomalies (e.g., pressure fluctuation range, temperature exceeding upper limit, equipment failure shutdown type, etc.). The equipment operation overview specifically includes: Equipment name (fill in the actual equipment name), current status (completely consistent with the "Equipment Status" in the equipment assessment results), anomaly level (minor / medium / high), and the main problem being a core anomaly of the refining equipment, such as exhaust temperature approaching the upper limit, cooling fan failure leading to high-temperature shutdown, etc. If no anomalies are found, fill in "-". System-level correlation analysis specifically includes: clearly describing the abnormal linkage relationships between equipment (e.g., "The exhaust temperature of air compressor A rises, causing the load on dryer C to increase and the cooling pressure to fluctuate abnormally"), explaining the linkage logic; determining whether there are common cause anomalies, and identifying the common cause problem (e.g., "The efficiency of the cooling water system decreases, causing multiple pieces of equipment such as air compressors and dryers to simultaneously experience abnormal temperature rises"); detailing the scope of equipment affected by the anomaly, the duration of the anomaly, and the degree of impact on the overall system's operational stability and production efficiency. Comprehensive cause inference specifically includes: possible causes, potential system-level root causes, such as insufficient cooling system efficiency, power fluctuations, etc.; basis and manifestations, combining the commonalities, temporal relationships, and phenomena of multiple equipment anomalies, explaining the basis for inference, such as "The temperature of multiple pieces of equipment rises simultaneously, and the temperature difference between the inlet and outlet of the cooling system is abnormal"; confidence level, high / medium / low. Risk assessment and recommendations may include: Current risk level: Low / Medium / High (determined comprehensively based on the severity of the anomaly, the scope of impact, and potential consequences); Key areas / equipment to focus on: Clearly mark the areas (e.g., "cooling system area") and equipment (e.g., "air compressor A, dryer C") that require key monitoring; Recommended measures: Based on the diagnostic results, propose specific, feasible, and targeted improvement suggestions, such as: 1. Check the operating efficiency of the cooling circulation system, clean debris from the heat dissipation pipes, and monitor the cooling water temperature difference; 2. Monitor the load balance among the air compressor group in real time, adjust operating parameters, and avoid overloading of individual equipment; 3. Regularly collect system-level energy consumption and efficiency data, track trends, and predict anomalies in advance; 4. For equipment with an anomaly level of "high," immediately shut down for maintenance and investigate the root cause of the fault.

[0075] The core requirements for users include: (1) Overall operation conclusion: It is necessary to comprehensively evaluate the overall operation status of the industrial system by analyzing the diagnostic data of all industrial equipment; combine the health status of industrial equipment, the distribution of abnormalities and the severity of various abnormalities to clearly determine whether the industrial system is in normal operation or has systemic abnormalities; if there are abnormalities, it is necessary to clearly identify the industrial equipment and areas involved, and briefly describe the specific manifestations of the main problems (such as excessive temperature, pressure fluctuations, and shutdown failures). The core purpose is to provide the management with a comprehensive and clear overview of the station's operation, helping them to quickly grasp the current health status and potential risks; (2) Equipment operation overview: describe the diagnostic results of each industrial equipment, clearly assess the current status, abnormality level (minor / moderate / serious) and main problems of each industrial equipment; the "current status" must be completely consistent with the input equipment assessment results and must not be modified; the abnormality level must correspond to the severity of the industrial equipment abnormality, and the main problems must be accurately extracted to help maintenance personnel quickly identify the health status of a single industrial equipment and key points of concern; (3) System-level correlation analysis: focus on analyzing the abnormal linkage relationship and common cause abnormalities between industrial equipment to determine whether there are systemic problems; based on the data of multiple industrial equipment, identify the correlation logic between industrial equipment (such as the failure of a certain industrial equipment causing load changes and parameter abnormalities in other industrial equipment); through cross-equipment abnormality pattern analysis, infer possible common cause problems (such as cooling system efficiency decline, power fluctuations). (3) Determine whether such problems will affect the operational stability of the entire station system; (4) Comprehensive cause inference: Combine the abnormal data of multiple industrial equipment to identify common abnormal characteristics and infer possible systemic root causes; Based on the abnormal patterns, timing relationships and working principles of the equipment, ensure that the inference logic is rigorous; At the same time, mark the confidence level (high / medium / low) for each inference result to clarify the possibility and severity of the problem and provide a solid basis for subsequent risk assessment and recommendations; (5) Risk assessment and recommendations: Based on the equipment assessment results, system-level problems and cause inference, comprehensively assess the risk level (low / medium / high) of the current industrial system operation; Clearly identify key areas or equipment to avoid missing key hidden dangers; Combine specific abnormal problems to propose targeted and feasible improvement suggestions. The suggestions should be in line with the actual operation and maintenance scenario and have operability (e.g., for cooling system failure, it is recommended to check the system operating efficiency; for unbalanced load, it is recommended to monitor the load balance of the air compressor group).

[0076] The system evaluation rules for industrial systems are obtained from the system knowledge base.

[0077] Specifically, the system evaluation results of each piece of industrial equipment within the industrial system and the system evaluation rules of the industrial system can be filled into the system evaluation prompt word template to obtain the system evaluation prompt words for the industrial system.

[0078] In this embodiment of the invention, the construction of evaluation prompts for industrial corresponding systems is realized.

[0079] Step 270: Input the system evaluation prompt into the system agent corresponding to the industrial system to obtain the system evaluation result corresponding to the industrial system.

[0080] Specifically, after inputting the system evaluation prompts into the system agent corresponding to the industrial system, the system agent can perform multi-dimensional analysis of the industrial system based on the system evaluation prompts. Specifically, it can perform anomaly distribution analysis, equipment anomaly correlation inference, equipment anomaly root cause analysis, and system improvement measures, thereby obtaining system evaluation results. The system evaluation results can include overall operation conclusions, equipment operation overview, system-level correlation analysis, comprehensive cause inference, and risk assessment and recommendations.

[0081] For the overall operational conclusions, the system agent assesses the overall operational status of the industrial system by comprehensively analyzing the equipment evaluation results of all industrial equipment. Based on the health status of the industrial equipment, the distribution of anomalies, and the severity of various anomalies, it determines whether the industrial system is operating normally or if systemic anomalies exist. If anomalies are found, the system agent identifies the affected industrial equipment and areas and briefly describes the manifestations of the main problems, such as excessively high temperatures or pressure fluctuations. The core purpose of this part is to provide management with a comprehensive and clear overview of the industrial system's operation, helping them quickly understand the current health status and potential risks. For the equipment operation summary, the diagnostic results for each piece of industrial equipment can be described, assessing the current status, anomaly level, and main problems of each piece of equipment. Anomalies are categorized into minor, moderate, and severe levels. The main problems of each piece of industrial equipment are also extracted, such as excessively high temperatures or downtime due to malfunction, thereby helping maintenance personnel quickly identify the health status of industrial equipment and key issues requiring attention. For system-level correlation analysis, the existence of systemic problems is typically determined by analyzing the linkage and common causes of anomalies between industrial equipment. Based on data from multiple industrial equipment, the system agent identifies the correlations between them; for example, a failure in one piece of industrial equipment may lead to load changes in other industrial equipment. By analyzing anomaly patterns across devices, the intelligent agent can infer potential common causes, such as decreased cooling system efficiency or power fluctuations, thereby determining whether there are potential faults affecting the stability of the entire system. For comprehensive cause inference: by analyzing anomaly data from multiple industrial devices, commonalities are identified and potential systemic problems are inferred. The system agent infers potential root causes based on anomaly patterns, temporal relationships, and operating principles among industrial devices. Simultaneously, the system agent assesses the confidence level of each inference, determining the likelihood and severity of the problem, thus providing a basis for subsequent risk assessment and recommendations. For risk assessment and recommendations, based on device assessment results, the current risk level (low, medium, high) can be comprehensively evaluated, and key areas or industrial devices of concern can be identified. The system agent assesses various risks by analyzing the severity, scope of impact, and potential consequences of anomalies, and proposes specific improvement suggestions based on the diagnostic results. For example, for cooling system failures, it is recommended to check system operating efficiency; for load imbalance, it is recommended to monitor the load balance of the air compressor group.

[0082] The system's intelligent agent takes the equipment assessment results corresponding to each factory's equipment as input and utilizes the system knowledge base (system topology, process dependencies, environmental information, etc.) to conduct cross-equipment correlation analysis and common cause inference. It can identify common potential hazards affecting multiple devices and predict potential risks, providing system-level early warnings for operation and maintenance planning. It outputs system health, risk classification, and impact scope for adjusting operation and maintenance strategies and resource allocation.

[0083] The evaluation process for each level of intelligent agents includes a structured chain of evidence, comprising data fragments, trend information, knowledge base matching items, and inference links. The evaluation process is transparent and verifiable, no longer a black box, allowing on-site personnel to verify the reasonableness of conclusions based on evidence. Evaluation results are traceable to specific data and knowledge entries, meeting audit and compliance requirements. Furthermore, by constructing a three-tiered knowledge base—measurement point, equipment, and system—and automatically accumulating manual confirmations, evaluation reports, maintenance results, and handling procedures into new knowledge for knowledge base updates and model training, the longer the evaluation period, the more accurate the evaluation, the greater the knowledge coverage, and the more reliable the inference, reducing long-term maintenance costs and adapting to new failure modes. The false positive rate for similar failures with verifiable effects decreases, processing time shortens, and the number of knowledge base items and rules continuously grows. It provides standardized evaluation reports, prioritized maintenance recommendations, and re-inspection plans, combined with cross-measurement point collaborative inspection strategies. Maintenance personnel can quickly locate key components and priorities, reducing blind troubleshooting and repetitive repairs, and shortening the failure handling cycle.

[0084] In this embodiment of the invention, by inputting system evaluation prompts into the system agent, the system agent evaluates the industrial system based on the system evaluation prompts and obtains the system evaluation results, thereby realizing intelligent evaluation of the industrial system.

[0085] In practical applications, the measurement point knowledge base, equipment knowledge base, system knowledge base, and corresponding large model training dataset can be synchronously updated based on the return results of each intelligent agent (measurement point diagnostic conclusions, equipment diagnostic reports, system inspection reports) and human feedback on the return results (such as confirmation of diagnostic accuracy, evaluation of fault handling effectiveness, and supplementation of uncovered scenarios). This continuously improves the model's inference accuracy, expands the coverage of the knowledge base, and optimizes the diagnostic logic, forming a closed-loop iterative intelligent diagnostic capability. Through a knowledge adaptive enhancement mechanism, the intelligent agent can achieve continuous learning based on online feedback, sharing abnormal patterns, thresholds, and experience rules among similar devices and systems, accelerating the adaptation process of new devices and systems, and improving the overall level of intelligent diagnosis.

[0086] Figures 3a-3c This is a diagram of the system interface created for the intelligent agent provided in Embodiment 2. Figure 3a As shown, the tool's capabilities can first be structurally defined based on the OpenAI Schema interface specification, supporting dual-mode adaptation of both MCP native calling methods and general interface calling methods. It also incorporates a flexible and configurable authentication mechanism to ensure the security and compliance of interface access. The system's operation steps are: MCP Service -> Service Management -> Enterprise Service -> Create Enterprise Service. Figure 3bAs shown, users can import various diagnostic knowledge files such as equipment operation manuals and user manuals into the knowledge base to provide data support for intelligent agent diagnosis. In addition, the system automatically configures a dedicated knowledge base for each factory, including measurement points, equipment, and system diagnostic results. The system operation steps are: Equipment Knowledge Base -> Equipment Knowledge Base -> Create New Knowledge Base -> Create Custom Knowledge Base -> Fill in the knowledge base name and description -> Add File -> Parse File. Figure 3c As shown, users can create device agents by manually filling in device information and attaching appropriate knowledge bases to them. Alternatively, they can download standard templates from the system, fill in the device information according to the template requirements, and upload them for batch creation. The system operation steps are: Device Intelligence -> Device AI Profile -> Add Device / Import Device. Users can also create measurement point agents by manually filling in measurement point information and attaching appropriate knowledge bases and MCP tools to them. Alternatively, they can download standard templates from the system, fill in the measurement point information according to the template requirements, and upload them for batch creation. The system operation steps are: Device Intelligence -> Device AI Profile -> Device X - Device Diagnostic Configuration - Create Diagnostic Rule - Add Diagnostic Item.

[0087] Figures 4a-4e This is a system interface diagram for the system evaluation provided in Embodiment 2. (See diagram below.) Figure 4a As shown, you can activate the device to check the measurement points, such as... Figure 4b As shown, when performing real-time evaluation of measurement points, selecting a measurement point and clicking "Measurement Point Diagnosis" will enable simultaneous evaluation of the measurement point and generation of an evaluation report. The diagnostic progress will also be synchronized on the page in real time. Figure 4c As shown, when performing timed evaluations of measurement points, a timed task can be set for the measurement point agent to achieve timed evaluation of the measurement points. For example... Figure 4d As shown, evaluation tasks can be configured, and scheduled evaluation times can be set. The devices to be evaluated by the system can be selected, and scheduled tasks can be set for the system's intelligent agent to achieve scheduled system evaluation. The system's operation steps are: Device Intelligence -> Device AI Inspection. Figure 4e As shown, clicking "Inspect Now" on the page will start the system inspection in real time, and the inspection progress will be synchronized in real time in the form of a progress bar within the system.

[0088] Figures 5a-5c This is a system interface diagram for querying evaluation results provided in Embodiment 2. For example... Figure 5a As shown, you can view the system's evaluation results, such as... Figure 5b As shown, you can view the evaluation results of each device within the system, such as... Figure 5c As shown, you can view the evaluation results for each measurement point within the device. The system operation steps are: AI Report Record -> AI Diagnostic Record -> View.

[0089] The industrial system evaluation method based on multi-level intelligent agents provided in this embodiment includes: acquiring real-time data streams and / or historical data streams of each measurement point within each industrial device in the industrial system; for each measurement point, constructing measurement point evaluation prompts based on the real-time data stream and / or historical data stream corresponding to the measurement point, measurement point evaluation rules, and measurement point evaluation prompt templates; inputting the measurement point evaluation prompts into the measurement point intelligent agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point; for each industrial device, constructing device evaluation prompts based on the measurement point evaluation results corresponding to each measurement point within the industrial device, the corresponding device information, device evaluation rules, and device evaluation prompt templates; inputting the device evaluation prompts into the device intelligent agent corresponding to the industrial device to obtain the device evaluation result corresponding to the industrial device; constructing system evaluation prompts based on the device evaluation results corresponding to each industrial device within the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt templates; inputting the system evaluation prompts into the system intelligent agent corresponding to the industrial system to obtain the system evaluation result corresponding to the industrial system. The above technical solution first acquires the data streams corresponding to each measuring point within each industrial device in the industrial system, specifically real-time and / or historical data streams. Second, it constructs measuring point evaluation prompts for each measuring point based on the real-time and / or historical data streams, measuring point evaluation rules, and measuring point evaluation prompt templates. These prompts are then input into the corresponding measuring point agent. The measuring point agent evaluates each measuring point based on these prompts, thus obtaining the measuring point evaluation results and achieving intelligent evaluation of each measuring point. Finally, the solution can be further developed based on the measuring point evaluation results, equipment information, and design parameters within each industrial device. The system constructs equipment evaluation prompts for each industrial device based on evaluation rules and equipment evaluation prompt templates. These prompts are then input into the corresponding intelligent agents for each industrial device. The intelligent agents perform equipment evaluations based on these prompts, yielding evaluation results for each device. This achieves intelligent evaluation of each industrial device. Furthermore, the system can construct system evaluation prompts based on the system evaluation results for each industrial device and the system evaluation prompt templates. These prompts are then input into the system intelligent agents, which perform system evaluations based on these prompts, yielding system evaluation results. This also achieves intelligent evaluation of the industrial system.

[0090] Furthermore, through a hierarchical design of measurement point intelligent agents, equipment diagnostic intelligent agents, and system intelligent agents, continuous reasoning from microscopic data to system-level operational status is achieved. A data flow and structured knowledge fusion reasoning mechanism is used to perform high-precision calculations for anomaly identification, root cause analysis of faults, and health status assessment. Multi-level, multi-object intelligent diagnosis is achieved through cross-measurement point causal chain construction, equipment-level propagation path analysis, and system-level common cause identification. Simultaneously, a three-level knowledge base closed-loop learning mechanism automatically precipitates assessment results into knowledge entries and continuously optimizes reasoning logic. Automatic report generation templates and the MCP toolchain standardize, structure, and automate the diagnostic process, thereby ensuring high precision, interpretability, and sustainable optimization capabilities for industrial intelligent diagnosis.

[0091] Example 3 Figure 6 This is a schematic diagram of an industrial system evaluation device based on a multi-level intelligent agent provided in Embodiment 3. As shown in Figure 3, the device includes: The acquisition module 610 is used to acquire real-time data streams and / or historical data streams of each measuring point in each industrial device in the industrial system; The measurement point evaluation module 620 is used to construct a measurement point evaluation prompt word for each measurement point based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule and the measurement point evaluation prompt word template, input the measurement point evaluation prompt word into the measurement point agent corresponding to the measurement point, and obtain the measurement point evaluation result corresponding to the measurement point. The equipment evaluation module 630 is used to construct equipment evaluation prompts for each of the industrial equipment based on the evaluation results of the measurement points corresponding to each measurement point in the industrial equipment, the equipment information corresponding to the industrial equipment, the equipment evaluation rules, and the equipment evaluation prompt template, and input the equipment evaluation prompts into the equipment intelligent agent corresponding to the industrial equipment to obtain the equipment evaluation result corresponding to the industrial equipment. The system evaluation module 640 is used to construct system evaluation prompts based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template, and input the system evaluation prompts into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0092] The industrial system evaluation device based on multi-level intelligent agents provided in this embodiment acquires real-time data streams and / or historical data streams of each measurement point within each industrial device in an industrial system. For each measurement point, a measurement point evaluation prompt is constructed based on the real-time data stream and / or historical data stream corresponding to the measurement point, measurement point evaluation rules, and measurement point evaluation prompt template. The measurement point evaluation prompt is then input into the measurement point intelligent agent corresponding to the measurement point to obtain the measurement point evaluation result. For each industrial device, a device evaluation prompt is constructed based on the measurement point evaluation results corresponding to each measurement point within the industrial device, the corresponding device information, device evaluation rules, and device evaluation prompt template. The device evaluation prompt is then input into the device intelligent agent corresponding to the industrial device to obtain the device evaluation result. Finally, a system evaluation prompt is constructed based on the device evaluation results corresponding to each industrial device within the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt template. The system evaluation prompt is then input into the system intelligent agent corresponding to the industrial system to obtain the system evaluation result. The above technical solution first acquires the data streams corresponding to each measuring point within each industrial device in the industrial system, specifically real-time and / or historical data streams. Second, it constructs measuring point evaluation prompts for each measuring point based on the real-time and / or historical data streams, measuring point evaluation rules, and measuring point evaluation prompt templates. These prompts are then input into the corresponding measuring point agent. The measuring point agent evaluates each measuring point based on these prompts, thus obtaining the measuring point evaluation results and achieving intelligent evaluation of each measuring point. Finally, the solution can be further developed based on the measuring point evaluation results, equipment information, and other relevant data within each industrial device. The system constructs equipment evaluation prompts for each industrial device using equipment evaluation rules and prompt templates. These prompts are then input into the corresponding intelligent agents for each device. The intelligent agents perform equipment evaluations based on these prompts, yielding evaluation results for each industrial device. This enables intelligent evaluation of each industrial device. Furthermore, the system can construct system evaluation prompts based on the system evaluation results and prompt templates for each industrial device within the industrial system. These prompts are input into the system intelligent agents, which then perform system evaluations based on these prompts, yielding system evaluation results. This also enables intelligent evaluation of the industrial system.

[0093] In one embodiment, for each of the industrial devices, the measuring points within the industrial device include at least the host output frequency, host output power, host speed, device status, and device operating conditions.

[0094] Based on the above embodiments, the measurement point evaluation module 620 is specifically used for: The real-time data stream and / or the historical data stream corresponding to the measurement point, along with the measurement point evaluation rules, are filled into the measurement point evaluation prompt word template to obtain the measurement point evaluation prompt word. The measurement point evaluation prompt word template includes format limitations for the measurement point evaluation result. The measurement point evaluation prompt word is then input into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point.

[0095] In one embodiment, the measurement point evaluation results include: The diagnostic conclusions include an overview of the data and a description of the trends, anomaly identification and location, anomaly cause inference, maintenance recommendations and handling steps, cross-measurement point linkage inspection recommendations, and verifiable evidence.

[0096] Based on the above embodiments, the equipment evaluation module 630 is specifically used for: The evaluation results of each measurement point within the industrial equipment, the corresponding equipment information, and the equipment evaluation rules are filled into the equipment evaluation prompt word template to obtain the equipment evaluation prompt word. The equipment evaluation prompt word template includes format limitations for the equipment evaluation results. The equipment evaluation prompt word is then input into the equipment agent corresponding to the industrial equipment to obtain the equipment evaluation result corresponding to the industrial equipment.

[0097] In one embodiment, the equipment evaluation results include: Basic equipment information, abnormal measurement points, fault type location, fault risk prediction, and equipment maintenance suggestions.

[0098] Based on the above embodiments, the system evaluation module 640 is specifically used for: The system evaluation results of each industrial device in the industrial system and the system evaluation rules of the industrial system are filled into the system evaluation prompt word template to obtain the system evaluation prompt word. The system evaluation prompt word template includes format restrictions for the system evaluation results. The system evaluation prompt word is then input into the system agent corresponding to the industrial system to obtain the system evaluation result corresponding to the industrial system.

[0099] In one implementation, the system evaluation results include: Overall operational conclusions, equipment operation overview, system-level correlations, comprehensive cause inferences, and risk assessment and recommendations.

[0100] In one embodiment, the measurement point evaluation rule, the equipment evaluation rule, and the system evaluation rule are all determined based on information input by the staff.

[0101] The industrial system evaluation device based on multi-level intelligent agents provided in this embodiment can execute the industrial system evaluation method based on multi-level intelligent agents provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the industrial system evaluation method based on multi-level intelligent agents.

[0102] It is worth noting that in the above embodiments of the industrial system evaluation device based on multi-level intelligent agents, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the scope of protection of the present invention.

[0103] Example 4 Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Figure 7 A block diagram of an exemplary electronic device 7 suitable for implementing embodiments of the present invention is shown. Figure 7 The electronic device 7 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0104] like Figure 7 As shown, the electronic device 7 is represented in the form of a general-purpose computing electronic device. The components of the electronic device 7 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0105] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0106] Electronic device 7 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by electronic device 7, including volatile and non-volatile media, removable and non-removable media.

[0107] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Electronic device 7 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (… Figure 7 Not shown; usually referred to as a "hard drive"). Although Figure 7 As not shown, disk drives for reading and writing to removable non-volatile disks (e.g., "floppy disks") and optical disc drives for reading and writing to removable non-volatile optical discs (e.g., CD-ROMs, DVD-ROMs, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0108] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of the present invention.

[0109] Electronic device 7 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), and with one or more devices that enable a user to interact with electronic device 7, and / or with any device that enables electronic device 7 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed through input / output (I / O) interface 22. Furthermore, electronic device 7 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) through network adapter 20. Figure 7 As shown, network adapter 20 communicates with other modules of electronic device 7 via bus 18. It should be understood that, although... Figure 7 Not shown, it can be combined with electronic device 7 to use other hardware and / or software modules, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0110] Processing unit 16 executes various functional applications and page displays by running programs stored in system memory 28, such as implementing the industrial system evaluation method based on multi-level intelligent agents provided in this embodiment of the invention, which includes: Acquire real-time and / or historical data streams from various measuring points within industrial equipment in an industrial system; For each of the aforementioned measurement points, a measurement point evaluation prompt word is constructed based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule, and the measurement point evaluation prompt word template. The measurement point evaluation prompt word is then input into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point. For each of the industrial devices, an equipment evaluation prompt is constructed based on the evaluation results of the measurement points corresponding to each measurement point in the industrial device, the equipment information corresponding to the industrial device, the equipment evaluation rules, and the equipment evaluation prompt template. The equipment evaluation prompt is then input into the equipment agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device. Based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed. The system evaluation prompt words are then input into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0111] Of course, those skilled in the art will understand that the processor can also implement the technical solution of the industrial system evaluation method based on multi-level intelligent agents provided in any embodiment of the present invention.

[0112] Example 5 This invention provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the program implements, for example, the industrial system evaluation method based on multi-level intelligent agents provided in this invention. The method includes: Acquire real-time and / or historical data streams from various measuring points within industrial equipment in an industrial system; For each of the aforementioned measurement points, a measurement point evaluation prompt word is constructed based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule, and the measurement point evaluation prompt word template. The measurement point evaluation prompt word is then input into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point. For each of the industrial devices, an equipment evaluation prompt is constructed based on the evaluation results of the measurement points corresponding to each measurement point in the industrial device, the equipment information corresponding to the industrial device, the equipment evaluation rules, and the equipment evaluation prompt template. The equipment evaluation prompt is then input into the equipment agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device. Based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed. The system evaluation prompt words are then input into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

[0113] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0114] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0115] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0116] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0117] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computing device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0118] Furthermore, the acquisition, storage, use, and processing of data in the technical solution of this invention all comply with relevant laws and regulations.

[0119] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. An industrial system evaluation method based on multi-level intelligent agents, characterized in that, include: Acquire real-time and / or historical data streams from various measuring points within industrial equipment in an industrial system; For each of the aforementioned measurement points, a measurement point evaluation prompt word is constructed based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rule, and the measurement point evaluation prompt word template. The measurement point evaluation prompt word is then input into the measurement point agent corresponding to the measurement point to obtain the measurement point evaluation result corresponding to the measurement point. For each of the industrial devices, an equipment evaluation prompt is constructed based on the evaluation results of the measurement points corresponding to each measurement point in the industrial device, the equipment information corresponding to the industrial device, the equipment evaluation rules, and the equipment evaluation prompt template. The equipment evaluation prompt is then input into the equipment agent corresponding to the industrial device to obtain the equipment evaluation result corresponding to the industrial device. Based on the equipment evaluation results corresponding to each industrial device in the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed. The system evaluation prompt words are then input into the system agent corresponding to the industrial system to obtain the system evaluation results corresponding to the industrial system.

2. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, For each of the aforementioned industrial devices, the measuring points within the industrial device shall include at least the host output frequency, host output power, host speed, device status, and device operating conditions.

3. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, Based on the real-time data stream and / or the historical data stream corresponding to the measurement point, the measurement point evaluation rules, and the measurement point evaluation prompt word template, a measurement point evaluation prompt word is constructed, including: The real-time data stream and / or the historical data stream corresponding to the measurement point, along with the measurement point evaluation rules, are filled into the measurement point evaluation prompt word template to obtain the measurement point evaluation prompt word. The measurement point evaluation prompt word template includes format limitations for the measurement point evaluation result.

4. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, The evaluation results of the measurement points include: The diagnostic conclusions include an overview of the data and a description of the trends, anomaly identification and location, anomaly cause inference, maintenance recommendations and handling steps, cross-measurement point linkage inspection recommendations, and verifiable evidence.

5. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, Based on the evaluation results of the measurement points corresponding to each measurement point within the industrial equipment, the corresponding equipment information, equipment evaluation rules, and equipment evaluation prompt word templates, equipment evaluation prompt words are constructed, including: The evaluation results of the measurement points corresponding to each measurement point in the industrial equipment, the corresponding equipment information of the industrial equipment, and the equipment evaluation rules are filled into the equipment evaluation prompt word template to obtain the equipment evaluation prompt word. The equipment evaluation prompt word template includes format restrictions for the equipment evaluation results.

6. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, The equipment evaluation results include: Basic equipment information, abnormal measurement points, fault type location, fault risk prediction, and equipment maintenance suggestions.

7. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, Based on the equipment evaluation results corresponding to each piece of industrial equipment within the industrial system, the system evaluation rules of the industrial system, and the system evaluation prompt word template, system evaluation prompt words are constructed, including: The system evaluation results of each industrial device in the industrial system and the system evaluation rules of the industrial system are filled into the system evaluation prompt word template to obtain the system evaluation prompt word. The system evaluation prompt word template includes format restrictions for the system evaluation results.

8. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, The system evaluation results include: Overall operational conclusions, equipment operation overview, system-level correlations, comprehensive cause inferences, and risk assessment and recommendations.

9. The industrial system evaluation method based on multi-level intelligent agents according to claim 1, characterized in that, The measurement point evaluation rules, the equipment evaluation rules, and the system evaluation rules are all determined based on the information input by the staff.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the industrial system evaluation method based on multi-level intelligent agents as described in any one of claims 1-9.

Images