Method and system for diagnosing heating scene anomaly based on large model, and storage medium

Abstract

The application relates to a large model-based heating scene anomaly diagnosis method and system and a storage medium, relates to the field of large model application in the heating industry, and comprises the following steps: establishing a diagnosis rule library comprising a plurality of preset diagnosis rules; creating a plurality of diagnosis tasks; configuring corresponding abnormal parameter rules as diagnosis starting conditions for each diagnosis task; custom-configuring corresponding diagnosis processes for each diagnosis task, including customizing a diagnosis sequence and selecting a plurality of preset diagnosis rules in sequence; monitoring heat exchange system operation parameters in real time according to the created plurality of diagnosis tasks, and performing diagnosis through a large model according to a preset diagnosis process after the diagnosis starting conditions are met, and giving summary suggestions and final conclusions according to diagnosis results. The application has the effects of forming a diagnosis process integrating professional knowledge and expert experience, greatly shortening an abnormal response time, reducing experience requirements of operation and maintenance personnel, and improving fault diagnosis accuracy and professionalism.

Description

Technical Field

[0001] This application relates to the field of large-scale model applications in the heating industry, and in particular to a method, system, and storage medium for diagnosing anomalies in heating scenarios based on large-scale models. Background Technology

[0002] In the operation and maintenance of heat exchange systems, traditional methods have long been used for data analysis, which present data results in the form of data reports and trend curves. This method can only show "what happened", but cannot penetrate the data surface to explain "why it happened". Operation and maintenance personnel need to rely on their own professional skills to analyze data reports, trend curves and so on.

[0003] Because the operating status of heat exchange systems is affected by multiple factors, including external environmental meteorological data (such as sudden drops in temperature, changes in wind speed, and humidity fluctuations), internal system pipeline data (such as abnormal pipeline pressure, imbalance of medium flow, and degree of pipe scaling), and operational data of core equipment (such as pump speed deviations, abnormal temperature differences in heat exchangers, and valve on / off status malfunctions), when an anomaly occurs in a certain data point of the heat exchange system, maintenance personnel need to manually collect, organize, and correlate these multidimensional data points scattered across different systems and formats during the analysis process. This often requires repeatedly switching between meteorological monitoring platforms, pipeline monitoring systems, and equipment management software, manually comparing data times and verifying data accuracy, and then relying on their personal professional knowledge and experience to make a fault diagnosis based on these data results.

[0004] The traditional data anomaly analysis process described above is time-consuming, with troubleshooting cycles often measured in hours. This results in severely delayed anomaly responses and missed opportunities for fault repair. Furthermore, the troubleshooting process heavily relies on the individual skills and experience of maintenance personnel. Inexperienced personnel are prone to attribution biases and misjudgments, which not only fail to resolve the problem but may also lead to additional equipment wear and increased maintenance costs. While experienced experts have high decision-making accuracy, their knowledge is difficult to standardize. In addition, traditional methods completely lack early warning capabilities for system anomalies and faults, only allowing for reactive "post-incident remediation." This makes it difficult to avoid the negative impact of faults on the stability of heat exchange station operations, heating quality, and even the efficiency of the entire heating system, thus hindering the intelligent upgrading and high-quality development of heat exchange station operation and maintenance management in the long term. Summary of the Invention

[0005] To address the aforementioned technical challenges, this application integrates professional knowledge and expert experience, and transforms them into structured and algorithmic approaches to form a standardized diagnostic process and precise diagnostic model for real-time monitoring and autonomous diagnosis. This significantly shortens anomaly response time, reduces the experience requirements for maintenance personnel, improves fault diagnosis accuracy, and enables a shift in maintenance mode from "post-event remediation" to "early warning and proactive prevention." This application provides a method, system, and storage medium for anomaly diagnosis in heating scenarios based on a large model.

[0006] Firstly, this application provides a method for diagnosing anomalies in heating scenarios based on a large model, employing the following technical solution:

[0007] A method for anomaly diagnosis in heating scenarios based on a large model includes the following steps:

[0008] A diagnostic rule base is established, which includes several pre-set diagnostic rules based on the extraction of theoretical knowledge and expert experience in the heating industry. The diagnostic rules include several pre-set related parameters, judgment conditions for the related parameters, and judgment conclusions and processing suggestions for satisfying the judgment conditions.

[0009] Create several diagnostic tasks and set the basic information for each diagnostic task;

[0010] Configure corresponding abnormal parameter rules as diagnostic initiation conditions for each diagnostic task;

[0011] For each diagnostic task, a corresponding diagnostic process can be customized, including customizing the diagnostic order and selecting several preset diagnostic rules in sequence.

[0012] Obtain real-time and historical data of the operating parameters of the heating system;

[0013] Based on the abnormal parameter rules of each of the created diagnostic tasks, it is determined in real time whether the real-time data and / or historical data meet the diagnostic initiation conditions;

[0014] When the diagnostic initiation conditions are met, the large model performs diagnosis sequentially based on each of the pre-set diagnostic rules in the corresponding diagnostic process. Then, the root cause analysis is performed in combination with the diagnostic results of each diagnostic rule, and summary suggestions and final conclusions are given.

[0015] By adopting the above technical solution, theoretical knowledge in the heat exchanger industry's operation and maintenance field is integrated with the practical experience of senior experts. Based on the pre-set diagnostic rules, the related parameters of fault occurrence, the judgment conditions for abnormal related parameters, and the judgment conclusions and handling suggestions after the related parameters meet the judgment conditions, professional knowledge and expert experience are structured and algorithmically transformed. This achieves the digital accumulation and continuous evolution of professional knowledge and expert experience, solving the problems of the gap in the inheritance of expert experience and the low utilization rate of knowledge. This facilitates the formation of standardized diagnostic analysis based on professional knowledge and expert experience during the diagnosis process. It also helps operation and maintenance personnel to carry out actual operations based on the subsequent summary suggestions and final conclusions, reducing the experience requirements of operation and maintenance personnel, improving the accuracy of operation and maintenance personnel's decision-making, and effectively improving the accuracy and professionalism of anomaly analysis and diagnostic results.

[0016] During the operation of the heat exchange system, real-time and historical data of the heat exchange system are acquired. Based on the abnormal parameter rules set in the diagnostic task, the system determines whether the acquired real-time and historical data meet the diagnostic initiation conditions. If they do, a large model performs diagnosis according to a preset diagnostic process, providing summary suggestions and final conclusions. Through a closed loop of "real-time monitoring - anomaly identification - automatic diagnosis - fault location - suggestion provision," the system achieves fully intelligent and autonomous anomaly diagnosis. It continuously monitors the operating parameters of the heat exchange system and, after anomaly detection, conducts professional analysis through a traceable diagnostic path to accurately locate the root cause of the fault and assess the scope of impact. It automatically generates natural language guidance containing summary suggestions and final conclusions, upgrading traditional manual troubleshooting to a standardized intelligent process. Anomaly response time is reduced to minutes, and diagnostic accuracy is significantly improved, realizing a transformation from "post-event remediation" to "early warning and proactive prevention" in operation and maintenance.

[0017] By establishing a diagnostic rule base, several diagnostic rules related to the current diagnostic task can be selected from the diagnostic rule base during the configuration of the diagnostic process, and the order of these diagnostic rules can be adjusted. This enables personalized and autonomous configuration in the diagnostic process and facilitates the reuse of diagnostic rules in different diagnostic tasks, simplifying the operation steps for operation and maintenance personnel to configure the diagnostic process.

[0018] In a specific implementation scheme, when the diagnostic initiation conditions are met, the large model performs diagnosis sequentially based on each of the pre-set diagnostic rules in the corresponding diagnostic process, according to the order of the diagnostic rules. Based on the diagnostic results of each diagnostic rule, a summary suggestion and a final conclusion are given, including:

[0019] Obtain real-time and / or historical data of the preset associated parameters in the current diagnostic rule;

[0020] Identify abnormal parameters in the real-time and / or historical data of the associated parameters that meet the preset judgment conditions.

[0021] The large model provides a diagnostic conclusion based on the abnormal parameters and preset judgment conclusions, according to the current diagnostic rules.

[0022] By integrating the diagnostic conclusions of all the diagnostic rules in the diagnostic process and the preset treatment suggestions through a large model, root cause analysis is performed, and a final conclusion and summary suggestions are given.

[0023] In one specific implementation, the preset steps of the diagnostic rule include:

[0024] Set the name of the diagnostic rule;

[0025] Select the abnormal diagnosis type and the diagnostic items under the abnormal diagnosis type;

[0026] Explain the causal relationship between the diagnostic item malfunction and parameter abnormality and use it as the basis for inspection;

[0027] Set several of the associated parameters that need to be checked;

[0028] Set several judgment conditions for each of the aforementioned associated parameters;

[0029] Explain the judgment conclusion when all judgment conditions are met simultaneously or partially;

[0030] Based on the judgment conclusion, the corresponding processing suggestions are set.

[0031] By adopting the above technical solution, when setting up diagnostic rules, the name of the diagnostic rule and the diagnostic type and diagnostic items are set, which facilitates the selection of diagnostic rules during the configuration of the diagnostic process; the inspection basis, related parameters, judgment conditions, judgment conclusions and handling suggestions are set, which realizes the utilization and digital accumulation of professional knowledge and expert experience, making it easy to identify anomalies based on the set inspection basis, related parameters and judgment conditions, and to perform root cause analysis and fault diagnosis on abnormal data based on the inspection basis, judgment conclusions and handling suggestions, and to give professional summary suggestions and final conclusions.

[0032] In one specific implementation, configuring corresponding abnormal parameter rules for each diagnostic task as diagnostic initiation conditions includes:

[0033] Set several abnormal parameter conditions and configure the association between several abnormal parameter conditions, and configure the abnormal parameter conditions and the association as the abnormal parameter rules;

[0034] The abnormal parameter conditions include specific parameters, data types, data ranges, value acquisition methods, set values, and the relationship between data values ​​and set values;

[0035] The relationships include "AND" and "OR".

[0036] By adopting the above technical solutions, and based on theoretical knowledge in the heat exchange industry's operation and maintenance field and the practical experience of senior experts, abnormal parameter conditions and correlations are set. This allows for timely identification of abnormal parameters and autonomous initiation of diagnostic processes during real-time monitoring of heating system operation data. This not only enables the utilization of professional knowledge and expert experience but also significantly shortens the response time to heating system anomalies, achieving a shift from "passive response" to "proactive prediction." It also reduces energy consumption costs by 15%-20% and lowers user complaint rates by more than 30%, ushering in a new stage of efficient, stable, and intelligent modern heating.

[0037] In a specific feasible implementation, it also includes:

[0038] The diagnostic tasks include public tasks and personal tasks. The basic information of the public tasks includes the task name, and the basic information of the personal tasks includes the task name and the heat exchanger units covered by the task.

[0039] Based on the theoretical knowledge and expert experience of the heating industry, corresponding abnormal parameter rules are configured for each of the public tasks as diagnostic start conditions; based on the heat exchange units covered by the task, specific abnormal parameter rules are configured for each of the individual tasks as diagnostic start conditions.

[0040] When creating the personal task, it is supported to choose to reuse the already created public task and directly call the basic information, abnormal parameter rules and diagnostic process of the public task, and then set the heat exchanger unit covered by the task;

[0041] When a personal task reuses a previously created public task, it supports personalized adjustments to the basic information, abnormal parameter rules, and diagnostic process of the public task within the current personal task.

[0042] By adopting the above technical solution, since the basic information of the common task does not include the heat exchanger units covered by the task, the common task cannot actually run. However, when creating multiple individual tasks, the same common task can be reused, thereby effectively shortening the configuration time of the diagnostic task and simplifying the operation steps of the operation and maintenance personnel. At the same time, when individual tasks reuse common tasks, the basic information, abnormal parameter rules and diagnostic process of the common task can be personalized to suit different heat exchanger units.

[0043] In a specific feasible implementation, it also includes:

[0044] When customizing the diagnostic process, it is possible to personalize the associated parameters, judgment conditions, judgment conclusions and processing suggestions preset in each selected diagnostic rule within the diagnostic process of the current diagnostic task.

[0045] By adopting the above technical solution, when configuring the diagnostic process, the specific content of the selected diagnostic rules, such as the associated parameters, judgment conditions, judgment conclusions and processing suggestions, can be customized. This not only makes the diagnostic rules applicable to the current diagnostic task, but also avoids resetting new diagnostic rules and simplifies the operation steps.

[0046] In a specific feasible implementation, it also includes:

[0047] The diagnostic rules, when preset, support the option to reuse the preset diagnostic rules and directly call the associated parameters, judgment conditions, judgment conclusions and processing suggestions of the preset diagnostic rules;

[0048] When the diagnostic rule reuses a preset diagnostic rule, it supports personalized adjustments to the associated parameters, judgment conditions, judgment conclusions, and processing suggestions of the preset diagnostic rule in the current diagnostic rule.

[0049] By adopting the above technical solution, existing diagnostic rules can be reused when setting up diagnostic rules, and adjustments can be made based on existing diagnostic rules, which effectively simplifies the operation steps of setting up diagnostic rules.

[0050] In one specific implementation, the preset steps of the diagnostic rule further include setting the fault urgency level of the diagnostic rule.

[0051] By adopting the above technical solution and setting the urgency level of the fault when setting the diagnostic rules, the operation and maintenance personnel can intuitively understand the severity and urgency of the anomaly, so that they can take corresponding measures in a timely and rapid manner to reduce the scope of the fault's impact.

[0052] Secondly, this application provides a heating scenario anomaly diagnosis system based on a large model, employing the following technical solution:

[0053] A heating scenario anomaly diagnosis system based on a large model includes:

[0054] The diagnostic rule configuration module is used to establish a diagnostic rule library. The diagnostic rule library includes several preset diagnostic rules based on the extraction of theoretical knowledge and expert experience in the heating industry. The diagnostic rules include several preset related parameters, judgment conditions of the related parameters, and judgment conclusions and processing suggestions that meet the judgment conditions.

[0055] The diagnostic task configuration module is used to create several diagnostic tasks and set the basic information of each diagnostic task; configure corresponding abnormal parameter rules for each diagnostic task as diagnostic start conditions; and customize the corresponding diagnostic process for each diagnostic task, including custom selection of several preset diagnostic rules and setting the diagnostic order of the selected diagnostic rules.

[0056] The unit operation monitoring module is used to acquire real-time and historical data of the heating system; determine in real time whether the real-time data and / or historical data meet the diagnostic start conditions according to the abnormal parameter rules of each created diagnostic task; and when the diagnostic start conditions are met, perform diagnosis based on each diagnostic rule in sequence according to the preset diagnostic rules in the corresponding diagnostic process through the large model, and then perform root cause analysis based on the diagnostic results of each diagnostic rule and give summary suggestions and final conclusions.

[0057] Thirdly, this application provides a computer-readable storage medium, which adopts the following technical solution:

[0058] A computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above for the anomaly diagnosis method for heating scenarios based on a large model.

[0059] In summary, this application includes at least one of the following beneficial technical effects:

[0060] Based on theoretical knowledge in the heat exchanger industry's operation and maintenance field and the practical experience of senior experts, abnormal parameter rules and diagnostic rules are configured to facilitate abnormal diagnosis based on professional knowledge and expert experience during the diagnostic process and form standardized diagnostic analysis. This reduces the experience requirements of operation and maintenance personnel, improves the accuracy of their decision-making, and effectively enhances the accuracy and professionalism of abnormal analysis and diagnostic results.

[0061] By implementing a closed loop of "real-time monitoring - anomaly identification - automatic diagnosis - fault location - suggestion provision", the entire process of intelligent and autonomous anomaly diagnosis is achieved, upgrading the traditional manual investigation to a standardized intelligent process, shortening the anomaly response time to the minute level, and realizing the transformation of the operation and maintenance model from "post-event remediation" to "early warning and proactive prevention".

[0062] It supports flexible configuration and adjustment of abnormal parameter rules, diagnostic processes and diagnostic rules, enabling personalized and autonomous configuration and adjustment for different heat exchanger units and different diagnostic tasks. It also facilitates the reuse of common tasks and diagnostic rules, effectively simplifying the operation steps for operation and maintenance personnel. Attached Figure Description

[0063] Figure 1This is a flowchart illustrating the anomaly diagnosis method for heating scenarios based on a large model, as described in this application embodiment.

[0064] Figure 2 This is a flowchart illustrating the pre-setting steps of the diagnostic rules in an embodiment of this application. Detailed Implementation

[0065] The following is in conjunction with the appendix Figure 1-2 This application will be described in further detail.

[0066] This application discloses a method for diagnosing anomalies in heating scenarios based on a large model.

[0067] Reference Figure 1 The anomaly diagnosis method for heating scenarios based on large models includes the following steps:

[0068] S100: Establish a diagnostic rule base, which includes several pre-set diagnostic rules based on theoretical knowledge and expert experience extracted from the heating industry.

[0069] Specifically, we collect and organize theoretical knowledge and expert experience documents in the heating industry, select theoretical knowledge and practical experience of senior experts in the field of heat exchange operation and maintenance, refine and integrate them, and pre-set several diagnostic rules for different diagnostic contents when different equipment, energy consumption and the overall system malfunction during the operation of the heating system. We then store these pre-set diagnostic rules in a diagnostic rule library for later use.

[0070] The diagnostic rules include several preset related parameters, judgment conditions for the related parameters, judgment conclusions and processing suggestions when the judgment conditions are met.

[0071] In addition, when setting up diagnostic rules, it is possible to select and reuse pre-set diagnostic rules and directly call the associated parameters, judgment conditions, judgment conclusions and processing suggestions in the pre-set diagnostic rules; and when reusing pre-set diagnostic rules, it is possible to make personalized adjustments to the associated parameters, judgment conditions, judgment conclusions and processing suggestions in the pre-set diagnostic rules in the current diagnostic rules.

[0072] Specifically, the preset steps of the diagnostic rules include:

[0073] S101: Select whether to reuse the preset diagnostic rules. If you select yes, you will select to reuse the specific preset diagnostic rule and directly call all the contents of the selected preset diagnostic rule. You can also modify all the contents of the called preset diagnostic rule under the current diagnostic rule. If you select no, you will customize all the contents of the current diagnostic rule.

[0074] S102: Set the name of the diagnostic rule so that the appropriate diagnostic rule can be selected and called according to the name of the diagnostic rule.

[0075] S103: Select the abnormal diagnosis type and the diagnostic items under the abnormal diagnosis type.

[0076] Based on the different abnormal indicators of the heating system during operation, several diagnostic types are preset, and based on the different specific abnormal locations under each diagnostic type, several diagnostic items are preset to classify the abnormal rules, so as to facilitate the subsequent invocation of diagnostic rules. When presetting diagnostic rules, the appropriate abnormal diagnostic type and diagnostic items are selected according to the content of the current diagnostic rule.

[0077] Specifically, in this embodiment, the types of abnormal diagnosis include equipment fault diagnosis, system operation abnormality diagnosis, and energy consumption index abnormality diagnosis. Among them, the diagnostic items for equipment fault diagnosis include various equipment in the heating system, including circulating pumps, heat exchangers, electric regulating valves, etc.; the diagnostic items for system operation abnormality diagnosis include the primary network system and the secondary network system; and the diagnostic items for energy consumption index abnormality diagnosis include water consumption, electricity consumption, heat consumption, etc.

[0078] S104: Explain the causal relationship between diagnostic item failures and parameter abnormalities and use it as the basis for inspection.

[0079] Specifically, explain which related parameters will change and the reasons for the changes when the selected diagnostic item in the chosen abnormal diagnostic type becomes abnormal or malfunctions in the current diagnostic rule. For example, when the abnormal diagnostic type is system operation abnormality diagnosis, the diagnostic item is the primary network system, the diagnostic rule name is insufficient heat source / hydraulic imbalance, and the inspection basis is set to cause insufficient flow and heat distribution of the heat exchange station units at the end of the heating network when there is insufficient heat source or hydraulic imbalance in the primary network, the primary electric regulating valve of the unit will gradually open, and in extreme cases, even if the valve is fully open, the target temperature cannot be achieved.

[0080] S105: Set several related parameters that need to be checked.

[0081] Specifically, based on the inspection criteria, set the associated parameters and select the data type of the associated parameters. The data type of the associated parameters includes real-time data and historical data. When the data type is historical data, select the data range to be queried. The unit of the data range includes minutes, hours, and days. For example, the associated parameters can be set to real-time data of the primary control valve's feedback opening and historical data of the primary control valve over the past 24 hours.

[0082] S106: Set several judgment conditions for each associated parameter.

[0083] Specifically, by extracting theoretical knowledge from the heat exchanger industry's operation and maintenance field and combining it with the practical experience of senior experts, we set judgment conditions that related parameters must meet when specific anomalies or faults occur in the current fault diagnosis project. For example, the judgment condition is set as calculating the average value of the primary control valve's feedback opening over the past 24 hours. If the average value of the primary control valve's feedback opening over the past 24 hours is less than 90%, and the real-time data of the primary control valve's feedback opening is greater than 95%, then the judgment condition is met.

[0084] S107: Explain the judgment conclusion when all judgment conditions are met simultaneously or partially.

[0085] Specifically, when multiple judgment conditions are set, the judgment conclusions for all conditions being met simultaneously or partially are explained; when a single judgment condition is set, the judgment conclusion when the condition is met is explained. For example, if the single judgment condition in the example above is met, it indicates that the heat source is insufficient or the primary network hydraulics are out of balance.

[0086] S108: Set corresponding processing suggestions based on the judgment conclusion.

[0087] Specifically, by extracting theoretical knowledge from the heat exchanger industry's operation and maintenance field and combining it with the practical experience of senior experts, this document provides suggested measures to be taken when current anomalies or malfunctions occur, facilitating practical operation by maintenance personnel. For example, when there is insufficient heat source or hydraulic imbalance in the primary network, the suggested measures could include considering increasing the heat source load, or reducing the heat load of units surrounding the unit or at the front end of the pipeline network, or reducing the secondary heat supply, thereby increasing the heat load that can be allocated to the unit.

[0088] S109: Set the fault urgency level of the diagnostic rules so that operation and maintenance personnel can intuitively understand the severity and urgency of the anomaly, and then take corresponding measures in a timely and rapid manner to reduce the scope of the fault's impact.

[0089] S200: Create several diagnostic tasks.

[0090] Diagnostic tasks include public tasks and personal tasks. When creating a diagnostic task, you can choose to create a public task or a personal task. When creating a personal task, you can choose to reuse a previously created public task and directly call all the contents of the public task; you can also personalize all the contents of the called public task in the current personal task.

[0091] Specifically, the steps for creating a diagnostic task include:

[0092] S201: Set the basic information for each diagnostic task.

[0093] The basic information for a public task includes the task name and task description, with the task description explaining the task content and purpose.

[0094] The basic information for individual tasks includes whether to reuse public tasks and the public tasks to be reused, as well as the task name, task description, and the heat exchanger units covered by the task.

[0095] If you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the basic information of the common task. It also supports adjusting the basic information of the reused common task in the current personal task. At this time, you only need to select the heat exchanger unit covered by the current personal task.

[0096] If you choose not to reuse public tasks and instead set your current personal tasks through customization, you can directly set the task name, task description, and the heat exchanger units covered by the task.

[0097] S202: Configure corresponding abnormal parameter rules for each diagnostic task as diagnostic initiation conditions.

[0098] Specifically, based on the extraction of theoretical knowledge and expert experience in the heating industry, corresponding abnormal parameter rules are configured for each public task as diagnostic start conditions; based on the heat exchange units covered by the task, specific abnormal parameter rules are configured for each individual task as diagnostic start conditions.

[0099] Configuring abnormal parameter rules for diagnostic tasks includes setting several abnormal parameter conditions and configuring the relationships between these abnormal parameter conditions.

[0100] The abnormal parameter conditions include specific parameters, data types, data ranges, value retrieval methods, set values, and the relationship between data values ​​and set values.

[0101] Furthermore, firstly, select the specific parameters and data types in the abnormal parameter conditions. Data types include real-time data and historical data. When the data type is real-time data, it is necessary to set the setpoint and select the relationship between the data value and the setpoint. The relationship between the data value and the setpoint includes greater than, less than, equal to, greater than or equal to, less than or equal to, and not equal to. When the data type is historical data, it is necessary to set the data range, value retrieval method, setpoint, and relationship between the data value and the setpoint. The unit of the data range includes minutes, hours, and days. The value retrieval method includes threshold, interruption, jump, and no change. The relationship between the data value and the setpoint includes greater than, less than, equal to, greater than or equal to, less than or equal to, not equal to, number of times, and duration. The association relationships between several abnormal parameter conditions include "AND" and "OR".

[0102] When creating a personal task, if you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the exception parameter rules of the common task, and you can adjust the exception parameter rules of the reused common task in the current personal task.

[0103] S203: Customize the corresponding diagnostic process for each diagnostic task, including customizing the diagnostic sequence and selecting several preset diagnostic rules in sequence.

[0104] Specifically, you can select several diagnostic rules in the diagnostic process in the required diagnostic order. When selecting diagnostic rules, you can select the abnormal diagnostic type, diagnostic item and name of the diagnostic rule in order to quickly locate the diagnostic rule to be selected. When customizing the diagnostic process, you can personalize the preset associated parameters, judgment conditions, judgment conclusions and processing suggestions in the diagnostic process of the current diagnostic task.

[0105] When creating a personal task, if you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the diagnostic process of the common task. It also supports adjusting the diagnostic process of the reused common task in the current personal task.

[0106] S300: Based on several created diagnostic tasks, it monitors the operating parameters of the heat exchange system in real time, and after meeting the diagnostic start-up conditions, performs diagnostics through a large model according to a preset diagnostic process. Based on the diagnostic results, it provides summary suggestions and final conclusions, specifically including:

[0107] S301: Obtain real-time and historical data of the operating parameters of the heating system.

[0108] S302: Based on the abnormal parameter rules of each created diagnostic task, determine in real time whether the real-time data and / or historical data meet the diagnostic start conditions.

[0109] Specifically, it retrieves real-time and / or historical data of specific parameters in the abnormal parameter rules of each created diagnostic task; determines whether the diagnostic start conditions are met; and automatically starts the diagnostic process after the diagnostic start conditions are met.

[0110] S303: When the diagnostic initiation conditions are met, the large model performs diagnosis according to the preset diagnostic rules in the corresponding diagnostic process, based on each diagnostic rule in turn. Then, the root cause analysis is performed in combination with the diagnostic results of each diagnostic rule, and summary suggestions and final conclusions are given.

[0111] Specifically, it acquires real-time and / or historical data of the preset related parameters in the current diagnostic rules; identifies abnormal parameters in the real-time and / or historical data of the related parameters that meet the judgment conditions according to the preset judgment conditions; provides a diagnostic conclusion based on the current diagnostic rules through the large model based on the abnormal parameters and the preset judgment conclusions; and integrates the diagnostic conclusions of all diagnostic rules in the diagnostic process and the preset treatment suggestions through the large model to perform root cause analysis and provide a final conclusion and summary suggestions.

[0112] The implementation principle of the anomaly diagnosis method for heating scenarios based on a large model in this application is as follows:

[0113] During the operation of the heat exchange system, real-time and historical data are acquired. Based on the abnormal parameter rules set in the diagnostic task, the system determines whether the acquired real-time and historical data meet the diagnostic initiation conditions. If they do, a large model performs diagnosis according to a preset diagnostic process, providing summary suggestions and final conclusions. This closed-loop process of "real-time monitoring - anomaly identification - automatic diagnosis - fault location - suggestion provision" achieves fully intelligent and autonomous anomaly diagnosis, upgrading traditional manual troubleshooting to a standardized intelligent process, reducing anomaly response time to minutes. By extracting and integrating theoretical knowledge from the heat exchange industry's operation and maintenance field with the practical experience of senior experts, and based on this, preset diagnostic rules facilitate professional analysis through a traceable diagnostic path after anomaly detection. This accurately locates the root cause of the fault and assesses the scope of impact, automatically generating natural language guidance containing summary suggestions and final conclusions. This allows maintenance personnel to perform practical operations based on the subsequently provided summary suggestions and final conclusions, reducing the experience requirements for maintenance personnel, improving the accuracy of their decisions, and realizing a transformation from "post-event remediation" to "early warning and proactive prevention" in the operation and maintenance model.

[0114] It should be understood that, although Figure 1 and Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1 and Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0115] This application also discloses an anomaly diagnosis system for heating scenarios based on a large model.

[0116] A heating scenario anomaly diagnosis system based on a large model includes:

[0117] The diagnostic rule configuration module is used to establish a diagnostic rule library, which includes several pre-set diagnostic rules based on theoretical knowledge and expert experience extracted from the heating industry.

[0118] Specifically, we collect and organize theoretical knowledge and expert experience documents in the heating industry, select theoretical knowledge and practical experience of senior experts in the field of heat exchange operation and maintenance, refine and integrate them, and pre-set several diagnostic rules for different diagnostic contents when different equipment, energy consumption and the overall system malfunction during the operation of the heating system. We then store these pre-set diagnostic rules in a diagnostic rule library for later use.

[0119] The diagnostic rules include several preset related parameters, judgment conditions for the related parameters, judgment conclusions and processing suggestions when the judgment conditions are met.

[0120] In addition, when setting up diagnostic rules, it is possible to select and reuse pre-set diagnostic rules and directly call the associated parameters, judgment conditions, judgment conclusions and processing suggestions in the pre-set diagnostic rules; and when reusing pre-set diagnostic rules, it is possible to make personalized adjustments to the associated parameters, judgment conditions, judgment conclusions and processing suggestions in the pre-set diagnostic rules in the current diagnostic rules.

[0121] Specifically, you can choose whether to reuse the preset diagnostic rules. If you choose yes, you can reuse the specific preset diagnostic rules and directly call all the contents of the selected preset diagnostic rules. You can also modify all the contents of the called preset diagnostic rules under the current diagnostic rules. If you choose no, you can customize all the contents of the current diagnostic rules.

[0122] Set a name for the diagnostic rule so that the appropriate diagnostic rule can be selected and invoked based on its name.

[0123] Select the abnormal diagnosis type and the diagnostic items under the abnormal diagnosis type.

[0124] Based on the different abnormal indicators of the heating system during operation, several diagnostic types are preset, and based on the different specific abnormal locations under each diagnostic type, several diagnostic items are preset to classify the abnormal rules, so as to facilitate the subsequent invocation of diagnostic rules. When presetting diagnostic rules, the appropriate abnormal diagnostic type and diagnostic items are selected according to the content of the current diagnostic rule.

[0125] Specifically, in this embodiment, the types of abnormal diagnosis include equipment fault diagnosis, system operation abnormality diagnosis, and energy consumption index abnormality diagnosis. Among them, the diagnostic items for equipment fault diagnosis include various equipment in the heating system, including circulating pumps, heat exchangers, electric regulating valves, etc.; the diagnostic items for system operation abnormality diagnosis include the primary network system and the secondary network system; and the diagnostic items for energy consumption index abnormality diagnosis include water consumption, electricity consumption, heat consumption, etc.

[0126] Explain the causal relationship between diagnostic item failures and parameter abnormalities and use it as the basis for inspection.

[0127] Specifically, explain which related parameters will change and the reasons for the changes when the selected diagnostic item in the chosen abnormal diagnostic type becomes abnormal or malfunctions in the current diagnostic rule. For example, when the abnormal diagnostic type is system operation abnormality diagnosis, the diagnostic item is the primary network system, the diagnostic rule name is insufficient heat source / hydraulic imbalance, and the inspection basis is set to cause insufficient flow and heat distribution of the heat exchange station units at the end of the heating network when there is insufficient heat source or hydraulic imbalance in the primary network, the primary electric regulating valve of the unit will gradually open, and in extreme cases, even if the valve is fully open, the target temperature cannot be achieved.

[0128] Set several related parameters that need to be checked.

[0129] Specifically, based on the inspection criteria, set the associated parameters and select the data type of the associated parameters. The data type of the associated parameters includes real-time data and historical data. When the data type is historical data, select the data range to be queried. The unit of the data range includes minutes, hours, and days. For example, the associated parameters can be set to real-time data of the primary control valve's feedback opening and historical data of the primary control valve over the past 24 hours.

[0130] Set several judgment conditions for each associated parameter.

[0131] Specifically, by extracting theoretical knowledge from the heat exchanger industry's operation and maintenance field and combining it with the practical experience of senior experts, we set judgment conditions that related parameters must meet when specific anomalies or faults occur in the current fault diagnosis project. For example, the judgment condition is set as calculating the average value of the primary control valve's feedback opening over the past 24 hours. If the average value of the primary control valve's feedback opening over the past 24 hours is less than 90%, and the real-time data of the primary control valve's feedback opening is greater than 95%, then the judgment condition is met.

[0132] Explain the judgment conclusion when all judgment conditions are met simultaneously or partially.

[0133] Specifically, when multiple judgment conditions are set, the judgment conclusions for all conditions being met simultaneously or partially are explained; when a single judgment condition is set, the judgment conclusion when the condition is met is explained. For example, if the single judgment condition in the example above is met, it indicates that the heat source is insufficient or the primary network hydraulics are out of balance.

[0134] Based on the judgment conclusion, corresponding processing suggestions are set.

[0135] Specifically, by extracting theoretical knowledge from the heat exchanger industry's operation and maintenance field and combining it with the practical experience of senior experts, this document provides suggested measures to be taken when current anomalies or malfunctions occur, facilitating practical operation by maintenance personnel. For example, when there is insufficient heat source or hydraulic imbalance in the primary network, the suggested measures could include considering increasing the heat source load, or reducing the heat load of units surrounding the unit or at the front end of the pipeline network, or reducing the secondary heat supply, thereby increasing the heat load that can be allocated to the unit.

[0136] Set the urgency level of the fault in the diagnostic rules so that maintenance personnel can intuitively understand the severity and urgency of the anomaly, and then take corresponding measures in a timely and rapid manner to reduce the scope of the fault's impact.

[0137] The diagnostic task configuration module is used to create several diagnostic tasks and set the basic information of each diagnostic task; configure corresponding abnormal parameter rules for each diagnostic task as diagnostic start conditions; and customize the corresponding diagnostic process for each diagnostic task, including customizing the diagnostic order and selecting several preset diagnostic rules in sequence.

[0138] Diagnostic tasks include public tasks and personal tasks. When creating a diagnostic task, you can choose to create a public task or a personal task. When creating a personal task, you can choose to reuse a previously created public task and directly call all the contents of the public task; you can also personalize all the contents of the called public task in the current personal task.

[0139] Specifically, set the basic information for each diagnostic task.

[0140] The basic information for a public task includes the task name and task description, with the task description explaining the task content and purpose.

[0141] The basic information for individual tasks includes whether to reuse public tasks and the public tasks to be reused, as well as the task name, task description, and the heat exchanger units covered by the task.

[0142] If you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the basic information of the common task. It also supports adjusting the basic information of the reused common task in the current personal task. At this time, you only need to select the heat exchanger unit covered by the current personal task.

[0143] If you choose not to reuse public tasks and instead set your current personal tasks through customization, you can directly set the task name, task description, and the heat exchanger units covered by the task.

[0144] Configure corresponding abnormal parameter rules for each diagnostic task as diagnostic initiation conditions.

[0145] Specifically, based on the extraction of theoretical knowledge and expert experience in the heating industry, corresponding abnormal parameter rules are configured for each public task as diagnostic start conditions; based on the heat exchange units covered by the task, specific abnormal parameter rules are configured for each individual task as diagnostic start conditions.

[0146] Configuring abnormal parameter rules for diagnostic tasks includes setting several abnormal parameter conditions and configuring the relationships between these abnormal parameter conditions.

[0147] The abnormal parameter conditions include specific parameters, data types, data ranges, value retrieval methods, set values, and the relationship between data values ​​and set values.

[0148] Furthermore, firstly, select the specific parameters and data types in the abnormal parameter conditions. Data types include real-time data and historical data. When the data type is real-time data, it is necessary to set the setpoint and select the relationship between the data value and the setpoint. The relationship between the data value and the setpoint includes greater than, less than, equal to, greater than or equal to, less than or equal to, and not equal to. When the data type is historical data, it is necessary to set the data range, value retrieval method, setpoint, and relationship between the data value and the setpoint. The unit of the data range includes minutes, hours, and days. The value retrieval method includes threshold, interruption, jump, and no change. The relationship between the data value and the setpoint includes greater than, less than, equal to, greater than or equal to, less than or equal to, not equal to, number of times, and duration. The association relationships between several abnormal parameter conditions include "AND" and "OR".

[0149] When creating a personal task, if you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the exception parameter rules of the common task, and you can adjust the exception parameter rules of the reused common task in the current personal task.

[0150] Specifically, you can select several diagnostic rules in the diagnostic process in the required diagnostic order. When selecting diagnostic rules, you can select the abnormal diagnostic type, diagnostic item and name of the diagnostic rule in order to quickly locate the diagnostic rule to be selected. When customizing the diagnostic process, you can personalize the preset associated parameters, judgment conditions, judgment conclusions and processing suggestions in the diagnostic process of the current diagnostic task.

[0151] When creating a personal task, if you choose to reuse a common task, all the contents of the reused common task will be automatically called after the selection, including the diagnostic process of the common task. It also supports adjusting the diagnostic process of the reused common task in the current personal task.

[0152] The unit operation monitoring module is used to monitor the operating parameters of the heat exchange system in real time according to several created diagnostic tasks. After the diagnostic start-up conditions are met, it performs diagnosis through a large model according to the preset diagnostic process, and provides summary suggestions and final conclusions based on the diagnostic results.

[0153] Specifically, real-time and historical data of the heating system's operating parameters are obtained, and then the abnormal parameter rules for each created diagnostic task are used to determine in real time whether the real-time and / or historical data meet the diagnostic activation conditions.

[0154] Furthermore, real-time and / or historical data of specific parameters in the abnormal parameter rules of each created diagnostic task are obtained; it is determined whether the diagnostic start conditions are met; and the diagnostic process is automatically started after the diagnostic start conditions are met.

[0155] Specifically, when the diagnostic initiation conditions are met, the large model performs diagnosis sequentially based on each pre-set diagnostic rule in the corresponding diagnostic process. Then, the root cause analysis is performed by combining the diagnostic results of each diagnostic rule, and summary suggestions and final conclusions are given.

[0156] Furthermore, the system acquires real-time and / or historical data of the pre-set related parameters in the current diagnostic rules; identifies abnormal parameters in the real-time and / or historical data of the related parameters that meet the judgment conditions according to the pre-set judgment conditions; provides a diagnostic conclusion based on the current diagnostic rules through the large model based on the abnormal parameters and the pre-set judgment conclusions; and integrates the diagnostic conclusions of all diagnostic rules in the diagnostic process and the pre-set processing suggestions through the large model to perform root cause analysis and provide a final conclusion and summary suggestions.

[0157] This application also discloses a computer-readable storage medium.

[0158] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the above-described method for diagnosing abnormal heating scenarios based on a large model. The computer-readable storage medium includes, for example, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0159] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A method for diagnosing a heating scene anomaly based on a large model, the method comprising: The method comprises the following steps: establishing a diagnostic rule base comprising a plurality of diagnostic rules preset according to refined theoretical knowledge and expert experience in the heating industry, the diagnostic rules comprising a plurality of associated parameters, judgment conditions of the associated parameters, and judgment conclusions and processing suggestions meeting the judgment conditions; creating a plurality of diagnostic tasks and setting basic information of each diagnostic task; configuring corresponding abnormal parameter rules as diagnostic starting conditions for each diagnostic task; configuring corresponding diagnostic processes for each diagnostic task, including customizing diagnostic order and selecting a plurality of preset diagnostic rules in sequence; obtaining real-time data and historical data of heating system operation parameters; real-time judging whether the real-time data and / or historical data meet the diagnostic starting conditions according to the abnormal parameter rules of each created diagnostic task; when the diagnostic starting conditions are met, sequentially diagnosing based on each diagnostic rule in the preset diagnostic rule order in the corresponding diagnostic process through a large model, and then performing root cause analysis and giving summary suggestions and final conclusions in combination with the diagnostic results of each diagnostic rule; the diagnostic tasks include public tasks and personal tasks, the basic information of the public tasks includes a task name, and the basic information of the personal tasks includes a task name and a heat exchange unit covered by the task; configuring corresponding abnormal parameter rules as diagnostic starting conditions for each public task according to refined theoretical knowledge and expert experience in the heating industry, and configuring specific abnormal parameter rules as diagnostic starting conditions for each personal task according to the heat exchange unit covered by the task; supporting selection of reuse of the created public tasks when creating the personal tasks, and directly invoking the basic information, abnormal parameter rules and diagnostic processes of the public tasks, and then setting the heat exchange unit covered by the task; when the personal tasks reuse the created public tasks, supporting individualized adjustment of the basic information, abnormal parameter rules and diagnostic processes of the public tasks in the current personal task; wherein the presetting step of the diagnostic rules comprises: setting a plurality of associated parameters to be checked; setting a plurality of judgment conditions for each associated parameter; explaining the judgment conclusion in the case that all judgment conditions are met or partially met; setting corresponding processing suggestions according to the judgment conclusion. 2.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, when the diagnostic starting conditions are met, sequentially diagnosing based on each diagnostic rule in the preset diagnostic rule order in the corresponding diagnostic process through a large model, and then giving summary suggestions and final conclusions according to the diagnostic results of each diagnostic rule, which comprises: obtaining real-time data and / or historical data of the associated parameters preset in the current diagnostic rule; identifying abnormal parameters meeting the judgment conditions in the real-time data and / or historical data of the associated parameters according to the preset judgment conditions; giving diagnostic conclusions based on the current diagnostic rule through a large model according to the abnormal parameters and the preset judgment conclusion; The diagnostic conclusions of all the diagnostic rules in the diagnostic process and the preset processing suggestions are integrated by a large model to perform root cause analysis and give a final conclusion and summary suggestions. 3.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, The presetting step of the diagnostic rule further includes: setting the name of the diagnostic rule; selecting an abnormality diagnosis type and a diagnosis item under the abnormality diagnosis type; explaining the cause-and-effect relationship between the failure of the diagnosis item and the parameter abnormality and serving as a check basis. 4.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, The corresponding abnormal parameter rule is configured for each diagnostic task as a diagnostic start condition, including: setting a plurality of abnormal parameter conditions and configuring the association relationship between a plurality of the abnormal parameter conditions, and configuring the abnormal parameter conditions and the association relationship as the abnormal parameter rule; the abnormal parameter condition includes a specific parameter, a data type, a data range, a value mode, a set value, and a relationship between a data value and a set value; the association relationship includes "and" and "or". 5.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, Further including: when the diagnostic process is self-defined, the preset associated parameters, judgment conditions, judgment conclusions, and processing suggestions in each selected diagnostic rule are individually adjusted in the diagnostic process of the current diagnostic task. 6.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, Further including: the diagnostic rule supports selecting a pre-set diagnostic rule and directly calling the associated parameters, judgment conditions, judgment conclusions, and processing suggestions of the pre-set diagnostic rule when pre-setting the diagnostic rule; when the diagnostic rule reuses the pre-set diagnostic rule, the associated parameters, judgment conditions, judgment conclusions, and processing suggestions of the pre-set diagnostic rule are individually adjusted in the current diagnostic rule. 7.The large model-based heating scene anomaly diagnosis method according to claim 1, characterized in that, The presetting step of the diagnostic rule further includes setting the failure urgency of the diagnostic rule. 8.A large model-based heating scene anomaly diagnosis system, characterized in that, A large model-based heating scene abnormality diagnosis method according to any one of claims 1 to 7, including: a diagnostic rule configuration module for establishing a diagnostic rule library, the diagnostic rule library including a plurality of diagnostic rules pre-set according to the extraction of theoretical knowledge and expert experience in the heating industry, the diagnostic rules including a plurality of pre-set associated parameters, judgment conditions of the associated parameters, and judgment conclusions and processing suggestions meeting the judgment conditions; a diagnostic task configuration module for creating a plurality of diagnostic tasks and setting the basic information of each diagnostic task, configuring the corresponding abnormal parameter rule for each diagnostic task as the diagnostic start condition, and customizing the corresponding diagnostic process for each diagnostic task, including customizing the selection of a plurality of pre-set diagnostic rules and setting the diagnostic order of the selected diagnostic rules; The unit operation monitoring module is configured to acquire real-time data and historical data of the heating system, determine whether the real-time data and / or the historical data satisfy the diagnostic start condition according to the abnormal parameter rule of each diagnostic task that has been created, and when the diagnostic start condition is satisfied, sequentially perform diagnosis based on each diagnostic rule in the order of the diagnostic rule preset in the corresponding diagnostic process through the large model, and then perform root cause analysis and give a summary suggestion and a final conclusion in combination with the diagnostic result of each diagnostic rule.

9. A computer-readable storage medium, characterized in that, The computer program is stored and can be loaded and executed by the processor to perform the large model-based heating scene abnormal diagnosis method according to any one of claims 1 to 7.

Images