Intelligent operation and maintenance method and system for air conditioning system based on Internet of Things technology

Abstract

The invention provides an intelligent operation and maintenance method and system for an air conditioning system based on the Internet of Things technology, and the method comprises the steps: processing operation data through employing an intelligent sensor and an edge calculation node, and generating a comprehensive operation information packet through combining with environment parameters; next, simulation analysis is carried out through a cloud digital twin model, performance prediction and fault point evaluation are generated, the fault probability is evaluated by using a Bayesian network algorithm, and personalized operation and maintenance suggestions are generated; secondly, optimizing operation parameters by adopting a genetic algorithm, and smoothly adjusting the setting through a self-adaptive control theory; and when an abnormality is detected, the fault severity is judged by using a multi-level threshold value and fuzzy logic, an early warning mechanism is triggered in time, and an emergency response scheme is pushed. And finally, recording a maintenance process and user feedback, updating the digital twin model, forming a closed-loop operation and maintenance system, and ensuring efficient and stable operation and service quality improvement.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of Internet of Things technology, and in particular to an intelligent operation and maintenance method of an air-conditioning system based on Internet of Things technology. Background Art

[0002] With the development of Internet of Things (IoT) technology and big data analysis, the operation and maintenance management of air-conditioning systems is gradually shifting towards intelligence and automation. Traditional air-conditioning system operation and maintenance relies on regular manual inspection and maintenance, which is not only inefficient, but also difficult to detect potential faults in time, which easily leads to unplanned downtime and energy waste. In order to improve operation and maintenance efficiency and service quality, modern air-conditioning systems have begun to integrate intelligent sensors and edge computing nodes. These devices can collect operating status data in real time and generate comprehensive operating information packages based on environmental parameters. However, existing technologies still have limitations in data processing and analysis. Especially in the face of complex and changing operating environments, traditional methods are difficult to provide accurate performance prediction and fault diagnosis, which limits the scientificity and timeliness of operation and maintenance decisions.

[0003] At present, although some advanced air-conditioning systems have adopted cloud-based digital twin models for simulation analysis to evaluate system performance trends and potential fault points, most solutions are limited to static data analysis and fail to fully combine historical fault data and current operating information for dynamic joint reasoning, resulting in inaccurate probability distribution assessment of fault occurrence. In addition, existing operation and maintenance recommendations often lack an effective search mechanism for the combination of operating parameters with the lowest energy consumption, and do not fully consider the need for smooth transition when adjusting the operating parameters of the air-conditioning system, which may cause user discomfort or excessive wear of the equipment. When an abnormal situation is detected, the existing early warning mechanism mostly relies on fixed threshold settings, lacks flexibility and adaptability, and cannot dynamically adjust the early warning strategy according to actual working conditions, affecting the timeliness and accuracy of fault response. Summary of the invention

[0004] The embodiment of the present invention provides an intelligent operation and maintenance method for an air-conditioning system based on the Internet of Things technology, which is used to solve the problems in the prior art that the probability distribution assessment of fault occurrence is not accurate enough, there is a lack of an effective search mechanism for the combination of operating parameters with the lowest energy consumption, and the demand for smooth transition is not fully considered when adjusting the operating parameters of the air-conditioning system, which may cause user discomfort or excessive wear of equipment, lack of flexibility and adaptability, and inability to dynamically adjust the early warning strategy according to actual working conditions, affecting the timeliness and accuracy of fault response.

[0005] In a first aspect, an embodiment of the present invention provides an intelligent operation and maintenance method for an air conditioning system based on Internet of Things technology, comprising:

[0006] Receiving real-time data streams from different sources, the real-time data streams containing structured data and unstructured data;

[0007] Using smart sensors deployed in key components of the air-conditioning system and edge computing nodes of the air-conditioning system, the operation status data is preliminarily processed and compressed, and combined with environmental parameters to obtain a comprehensive operation information package;

[0008] Based on the comprehensive operation information package, a digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation for the air-conditioning system, and a Bayesian network algorithm is used to perform joint reasoning on historical fault data and current operation information to evaluate the probability distribution of fault occurrence and generate personalized operation and maintenance recommendations;

[0009] Based on the personalized operation and maintenance suggestions, a genetic algorithm is used to search for an operating parameter combination with the lowest energy consumption, and adaptive control theory is used to smoothly adjust the operating parameters of the air-conditioning system to obtain an optimized air-conditioning system setting;

[0010] When an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel;

[0011] Based on the emergency response plan, the operation and maintenance personnel implement corresponding maintenance actions and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is equipped with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

[0012] Optionally, based on the comprehensive operation information package, a digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation for the air-conditioning system, and a Bayesian network algorithm is used to perform joint reasoning on historical fault data and current operation information to evaluate the probability distribution of fault occurrence and generate personalized operation and maintenance suggestions, including:

[0013] The digital twin model deployed in the cloud is combined with the reinforcement learning algorithm to perform high-precision simulation analysis on the comprehensive operation information package to obtain the performance trend prediction of the air-conditioning system and the possible fault point location;

[0014] Based on the performance trend prediction and the location of potential fault points, the Bayesian network and Markov decision process algorithm are combined to perform joint reasoning processing on historical fault data and current operating status information, evaluate the probability distribution of different faults and the degree of impact of different faults on the performance of the air conditioning system, and generate a preliminary fault impact assessment report;

[0015] Based on the preliminary fault impact assessment report, the graph neural network algorithm is used to model the interdependencies between the internal components of the air-conditioning system, simulate the interactive effects between the components under different fault scenarios, and generate an accurate fault impact assessment report;

[0016] Based on the precise fault impact assessment report, a multi-objective optimization algorithm is used to comprehensively calculate energy-saving standards and user comfort requirements to generate personalized operation and maintenance recommendations;

[0017] Optionally, based on the preliminary fault impact assessment report, a graph neural network algorithm is applied to model the interdependencies between components within the air-conditioning system, simulate the interaction between components under different fault scenarios, and generate an accurate fault impact assessment report, including:

[0018] Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and their importance under different operating conditions, and obtaining an enhanced component dependency model.

[0019] According to the enhanced component dependency model, the interaction effects between components under different fault scenarios are dynamically simulated by combining graph neural network with time series prediction algorithm, and the time propagation characteristics of the interaction effects between components are calculated to obtain the fault propagation path prediction including the time dimension.

[0020] Based on the fault propagation path prediction, a variational autoencoder is used to refine the fault location in the preliminary fault impact assessment report, capture potential fault modes and quantify the uncertainty estimate, and generate an analysis report containing a description of the fault scope and impact degree;

[0021] According to the analysis report, a causal inference method is introduced to analyze the causal relationship between the cause and consequence of the fault, and an integrated learning method is used to comprehensively evaluate the impact of the fault and generate an accurate fault impact assessment report.

[0022] Optionally, based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism to highlight key components and their importance under different operating conditions, thereby obtaining an enhanced component dependency model, including:

[0023] Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and their relative importance under different operating conditions, and obtaining a preliminary component dependency model.

[0024] According to the component dependency model, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the model, thereby obtaining an optimized component dependency model;

[0025] Based on the optimized component dependency model, the adaptive threshold setting is combined with the hierarchical clustering algorithm to perform correlation analysis on the components of the air-conditioning system, identify the key components that play a core role in normal and abnormal conditions, and quantify the importance indicators of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information.

[0026] According to the component dependency description, a Bayesian optimization algorithm is introduced to adjust the model parameters to generate an enhanced component dependency model;

[0027] Optionally, based on the optimized component dependency model, adaptive threshold setting is combined with a hierarchical clustering algorithm to perform association analysis on the components of the air-conditioning system, identify key components that play a core role in normal and abnormal conditions, and quantify the importance indexes of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information, including:

[0028] Based on the optimized component dependency model, the adaptive threshold setting is combined with the hierarchical clustering algorithm, and the community detection algorithm is introduced to perform association analysis on the components of the air-conditioning system, and the component grouping results that distinguish strong and weak dependencies are obtained;

[0029] According to the component grouping results, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the component dependency model, thereby obtaining an adjusted component dependency model;

[0030] Based on the adjusted component dependency model, the importance of components is quantitatively evaluated using the random forest learning algorithm to obtain the importance index of each component;

[0031] According to the importance index of each key component, social network analysis technology is used to analyze the hierarchical structure and functional modules between the key components, and a detailed description of the hierarchical structure and functional modules is generated;

[0032] According to the hierarchical structure and functional module description, a Bayesian optimization algorithm is introduced to adjust the component dependency model parameters to generate a component dependency description containing hierarchical information.

[0033] Optionally, based on the personalized operation and maintenance suggestion, a genetic algorithm is used to search for an operating parameter combination with the lowest energy consumption, and an adaptive control theory is used to smoothly adjust the operating parameters of the air-conditioning system to obtain an optimized air-conditioning system setting, including:

[0034] Based on the personalized operation and maintenance suggestions, a genetic algorithm is used in combination with a multi-objective optimization algorithm to search and process the operating parameter space of the air-conditioning system to obtain a preliminary operating parameter set that minimizes energy consumption and satisfies user comfort and system stability;

[0035] According to the preliminary operating parameter set, the operating parameters of the air conditioning system are smoothly adjusted by using adaptive control theory to obtain an adjusted operating parameter configuration;

[0036] Based on the adjusted operating parameter configuration, a real-time performance monitoring mechanism is introduced, and the actual operating performance of the air-conditioning system is monitored and adjusted using an online learning algorithm to obtain a performance-based operating parameter configuration;

[0037] According to the performance-based operating parameter configuration, a reinforcement learning algorithm is applied to evaluate and optimize the long-term impact of the operating parameters to generate an operating parameter solution with a long-term optimization effect;

[0038] The optimized air conditioning system settings are generated by combining the preliminary operating parameter set, the adjusted operating parameter configuration, the performance-based operating parameter configuration, and the operating parameter scheme with long-term optimization effect.

[0039] Optionally, when an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel, including:

[0040] When an abnormal situation is detected according to the optimized air conditioning system settings, a combination of multi-level threshold settings and fuzzy logic is used to perform severity judgment processing on the abnormal data, distinguish different levels of faults, and evaluate the uncertainty and complexity of the faults to obtain accurate fault severity assessment results;

[0041] Based on the fault severity assessment results, an early warning mechanism is immediately triggered through an event-driven architecture, and a rapid response is given to the operation and maintenance personnel according to different fault levels.

[0042] Based on the content of the emergency response plan, a natural language processing algorithm is used to automatically generate a fault description and recommended maintenance measures, convert technical terms into natural language, and generate an emergency response guide;

[0043] Combined with the real-time positioning system, the faulty equipment is accurately located, and maintenance auxiliary information including the fault location and visual guidance is generated through augmented reality technology;

[0044] The maintenance auxiliary information and emergency response guide are combined with the emergency response plan to generate an emergency response plan conclusion including the maintenance auxiliary information and the emergency response guide, and the conclusion is pushed to the relevant responsible parties through the operation and maintenance platform.

[0045] In a second aspect, an embodiment of the present invention provides an intelligent operation and maintenance system for an air conditioning system based on Internet of Things technology, comprising:

[0046] The compression module is used to use the smart sensors deployed in the key components of the air-conditioning system and the edge computing nodes of the air-conditioning system to preliminarily process and compress the operation status data, and combine it with environmental parameters to obtain a comprehensive operation information package;

[0047] An evaluation module is used to perform simulation analysis using a digital twin model in the cloud based on the comprehensive operation information package to generate a performance trend prediction and potential failure point evaluation for the air-conditioning system, and to perform joint reasoning on historical failure data and current operation information through a Bayesian network algorithm to evaluate the probability distribution of failure occurrence and generate personalized operation and maintenance recommendations;

[0048] An adjustment module, for searching for an operating parameter combination with the lowest energy consumption by using a genetic algorithm based on the personalized operation and maintenance suggestion, and for smoothly adjusting the operating parameters of the air-conditioning system by using an adaptive control theory to obtain an optimized air-conditioning system setting;

[0049] A judgment module, which is used to judge the severity of the fault by combining multi-level threshold setting with fuzzy logic when an abnormal situation is detected according to the optimized air conditioning system settings, and to instantly trigger an early warning mechanism through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel;

[0050] The update module is used for the operation and maintenance personnel to implement corresponding maintenance actions based on the emergency response plan, and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

[0051] In a third aspect, an embodiment of the present invention provides a computing device, comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the methods for intelligent operation and maintenance of an air-conditioning system based on Internet of Things technology as described in the first aspect.

[0052] In a fourth aspect, an embodiment of the present invention provides a computer storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement an intelligent operation and maintenance method for an air-conditioning system based on Internet of Things technology as described in any one of the first aspects.

[0053] In the embodiment of the present invention, intelligent sensors deployed in key components of the air-conditioning system and edge computing nodes of the air-conditioning system are used to preliminarily process and compress the operating status data, and combine the environmental parameters to obtain a comprehensive operating information package; based on the comprehensive operating information package, a digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation of the air-conditioning system, and the historical fault data and current operating information are jointly inferred through the Bayesian network algorithm to evaluate the probability distribution of the occurrence of faults and generate personalized operation and maintenance suggestions; based on the personalized operation and maintenance suggestions, a genetic algorithm is used to search for the operating parameter combination with the lowest energy consumption, and the adaptive control theory is used to smoothly adjust the operating parameters of the air-conditioning system to obtain the optimized air-conditioning system settings; when an abnormal situation is detected according to the optimized air-conditioning system settings, a combination of multi-level threshold settings and fuzzy logic is used to judge the severity of the fault, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response to the operation and maintenance personnel. Plan; Based on the emergency response plan, the operation and maintenance personnel implement corresponding maintenance actions, and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system. The technical solution provided by the present invention realizes real-time monitoring and predictive maintenance of the air-conditioning system by integrating intelligent sensors, edge computing nodes and cloud-based digital twin models. The invention not only improves the fault detection and prevention capabilities and reduces unplanned downtime, but also generates personalized operation and maintenance suggestions through the Bayesian network algorithm, which enhances the scientificity and accuracy of operation and maintenance decisions. In addition, the genetic algorithm is used to optimize energy consumption and the adaptive control theory is adopted to ensure user comfort. Combined with the real-time warning mechanism and the closed-loop operation and maintenance system, the system's operating efficiency and service quality are significantly improved;

[0054] Furthermore, by optimizing the simulation analysis and personalized operation and maintenance suggestion generation process, the reinforcement learning algorithm, Bayesian network and Markov decision process (MDP) algorithm and graph neural network (GNN) algorithm were introduced to achieve high-precision performance trend prediction and fault point location, evaluate the impact of faults through dynamic joint reasoning, and simulate the interaction between components to generate an accurate fault impact assessment report. Finally, a multi-objective optimization algorithm was used to comprehensively consider energy-saving standards and user comfort requirements to generate more scientific and reasonable personalized operation and maintenance suggestions, which significantly improved the intelligence level and practicality of the operation and maintenance methods and ensured the efficient and stable operation of the air-conditioning system.

[0055] These and other aspects of the present invention will become more apparent from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 A flowchart of an intelligent operation and maintenance method for an air conditioning system based on Internet of Things technology provided by an embodiment of the present invention;

[0058] Figure 2 A schematic diagram of the structure of an intelligent operation and maintenance system for an air conditioning system based on Internet of Things technology provided by an embodiment of the present invention;

[0059] Figure 3 A schematic diagram of the structure of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0061] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0063] Figure 1 The present invention provides a flowchart of an intelligent operation and maintenance method for an air conditioning system based on Internet of Things technology, such as Figure 1 As shown, the method includes:

[0064] Step 101: Using smart sensors deployed in key components of the air-conditioning system and edge computing nodes of the air-conditioning system, the operation status data is preliminarily processed and compressed, and combined with environmental parameters to obtain a comprehensive operation information package;

[0065] In this step, smart sensors are installed in key components of the air conditioning system (such as compressors, evaporators, condensers, etc.) to collect real-time operating status data such as temperature, humidity, pressure, vibration, etc.; edge computing nodes are local computing devices responsible for preliminary processing and analysis of data collected from smart sensors, reducing the amount of data transmitted to the cloud and improving response speed; comprehensive operating information packages are integrated with preliminary processed and compressed operating status data and environmental parameters (such as indoor and outdoor temperature, humidity, etc.), providing comprehensive data support for subsequent analysis;

[0066] The operating status data of the air-conditioning system is collected in real time through smart sensors and sent to the edge computing node; the edge computing node performs preliminary processing on the raw data, including data cleaning, formatting and compression, to reduce the data transmission burden and improve efficiency; the processed operating status data is combined with environmental parameters to generate a comprehensive operating information package containing all necessary information for further analysis.

[0067] Step 102: Based on the comprehensive operation information package, a digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation for the air-conditioning system, and a Bayesian network algorithm is used to perform joint reasoning on historical fault data and current operation information to evaluate the probability distribution of fault occurrence and generate personalized operation and maintenance recommendations;

[0068] In this step, the digital twin model refers to a virtual model that accurately simulates the operating conditions of the physical air conditioning system and can perform simulation analysis to predict future performance and identify potential failure points; the Bayesian network algorithm refers to a probabilistic graphical model used to represent the dependencies between variables and evaluate the probability distribution of different failures through joint reasoning;

[0069] The digital twin model in the cloud receives comprehensive operation information packages from edge computing nodes, performs simulation analysis on them, and generates performance trend predictions and potential fault point assessments for the air-conditioning system. It uses the Bayesian network algorithm to combine historical fault data and current operation information to assess the probability distribution of faults and their impact on system performance. Based on the above analysis results, it generates personalized operation and maintenance recommendations to guide operation and maintenance personnel to take appropriate preventive measures or optimization strategies.

[0070] Step 103: Based on the personalized operation and maintenance suggestion, a genetic algorithm is used to search for an operating parameter combination with the lowest energy consumption, and adaptive control theory is used to smoothly adjust the operating parameters of the air-conditioning system to obtain an optimized air-conditioning system setting;

[0071] In this step, genetic algorithm refers to an optimization algorithm that imitates the natural selection process to find the optimal solution and is used to search for the combination of operating parameters with the lowest energy consumption; adaptive control theory refers to a control method that can dynamically adjust the operating parameters during operation to ensure a smooth transition of the system without causing discomfort to users or wear and tear on equipment;

[0072] Based on personalized operation and maintenance suggestions, a genetic algorithm is used to search for operating parameter combinations that meet energy-saving standards and do not affect user comfort. Adaptive control theory is used to gradually adjust the operating parameters of the air-conditioning system to ensure a smooth transition from the current settings to the new settings and avoid the negative impact of sudden changes. Finally, the optimized air-conditioning system settings are generated, making the air-conditioning system efficient and energy-saving while maintaining a good user experience.

[0073] Step 104: when an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel;

[0074] In this step, multi-level threshold setting: setting multiple thresholds of different levels to distinguish between minor abnormalities and serious faults, helping to more accurately determine the severity of the fault.

[0075] Fuzzy logic refers to a logical method for dealing with uncertainty and imprecision, and is suitable for fault assessment under complex working conditions; event-driven architecture refers to a software design pattern that allows the system to automatically perform corresponding actions when specific events occur, such as triggering an early warning mechanism;

[0076] Under the optimized air conditioning system settings, the system operation status is continuously monitored. Once an abnormal situation is detected, the method combining multi-level threshold setting and fuzzy logic is immediately activated to assess the severity of the fault; based on the assessment results, the early warning mechanism is immediately triggered through the event-driven architecture, and a detailed emergency response plan is quickly pushed to the operation and maintenance personnel to ensure that the fault is handled quickly and effectively.

[0077] Step 105: Based on the emergency response plan, the operation and maintenance personnel implement corresponding maintenance actions, and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system;

[0078] In this step, the operation and maintenance platform refers to a management platform that integrates multiple functions, which is used to record the maintenance process, results and user feedback, and support the work of the operation and maintenance personnel; the user feedback interface refers to a functional module that allows users to submit usage experience and satisfaction evaluation, which helps to understand user needs and service quality; the incremental machine learning algorithm refers to a learning method that can continuously update based on the existing model, without retraining the entire model, and can adapt to new data more quickly;

[0079] The operation and maintenance personnel implement maintenance actions in accordance with the emergency response plan, and record the maintenance process and results in detail through the operation and maintenance platform; users can submit their usage experience and satisfaction evaluation through the feedback interface provided by the operation and maintenance platform, and provide valuable suggestions for improving services; the operation and maintenance platform uses incremental machine learning algorithms to learn from user feedback and actual maintenance results, and continuously updates and improves the digital twin model to ensure that it always reflects the latest system status and user needs; forming a closed-loop operation and maintenance system to ensure the long-term stable operation of the air-conditioning system and continuously improve service quality.

[0080] Suppose a large commercial complex has installed an intelligent air-conditioning system, including independent air-conditioning units in multiple rooms. In order to improve operation and maintenance efficiency and service quality, it is decided to adopt an intelligent operation and maintenance method based on the Internet of Things technology. The system deploys intelligent sensors in the key components of each air-conditioning unit to collect various operating status data such as temperature, humidity, pressure, vibration, etc. in real time, and preliminarily processes and compresses these data through edge computing nodes, and generates comprehensive operation information packages in combination with environmental parameters. These information packages are uploaded to the cloud platform, where the digital twin model is used for simulation analysis to predict the performance trend of the air-conditioning system and evaluate potential failure points. At the same time, the Bayesian network algorithm combines historical fault data and current operation information to evaluate the probability distribution of different faults and generate personalized operation and maintenance suggestions.

[0081] Based on these suggestions, a genetic algorithm is used to search for the operating parameter combination with the lowest energy consumption, and adaptive control theory is used to smoothly adjust the air-conditioning system operating parameters to obtain the optimized settings. The optimized air-conditioning system settings not only improve energy efficiency, but also ensure that user comfort is not affected. When an abnormal situation is detected, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and the early warning mechanism is triggered immediately through the event-driven architecture, pushing a detailed emergency response plan to the operation and maintenance personnel to ensure that the fault is handled quickly and effectively.

[0082] The operation and maintenance personnel implement maintenance actions according to the emergency response plan and record the actual maintenance process and results in detail through the operation and maintenance platform. In addition, users can submit their usage experience and satisfaction evaluation through the feedback interface provided by the operation and maintenance platform. The operation and maintenance platform uses incremental machine learning algorithms to learn from user feedback and actual maintenance results, continuously update and improve the digital twin model, and form a closed-loop operation and maintenance system to ensure long-term stable operation of the system and continuous improvement of service quality.

[0083] Through this intelligent operation and maintenance method, the commercial complex has improved its fault detection and prevention capabilities, reduced energy consumption and maintenance costs, and enhanced user experience and overall service levels. Throughout the entire process, from data collection, simulation analysis, generation of personalized operation and maintenance suggestions, optimization settings, anomaly detection and immediate warning, to the final maintenance actions and closed-loop feedback, all steps are closely connected, realizing intelligent management and efficient operation and maintenance, and providing a successful example for other similar projects.

[0084] Based on this, the present invention provides a specific embodiment, wherein step 102, based on the comprehensive operation information package, uses the digital twin model in the cloud to perform simulation analysis, generates a performance trend prediction and potential fault point evaluation for the air-conditioning system, and uses the Bayesian network algorithm to jointly reason with historical fault data and current operation information, evaluates the probability distribution of fault occurrence, and generates personalized operation and maintenance suggestions, specifically including the following steps:

[0085] Step 201: Use the digital twin model deployed in the cloud combined with the reinforcement learning algorithm to perform high-precision simulation analysis on the comprehensive operation information package to obtain the performance trend prediction of the air-conditioning system and the possible fault point location;

[0086] In this step, the digital twin model refers to a virtual model that accurately simulates the operating conditions of the physical air conditioning system and can perform simulation analysis to predict future performance and identify potential failure points; the reinforcement learning algorithm refers to a machine learning method that continuously optimizes the decision-making process by interacting with the environment and is suitable for the control and optimization of complex dynamic systems;

[0087] In the digital twin model deployed in the cloud, the reinforcement learning algorithm is combined to perform high-precision simulation analysis on the comprehensive operation information package uploaded from the edge computing node. By simulating the system behavior under different working conditions, the long-term performance trend of the air-conditioning system is predicted, and the possible fault point location is identified. The reinforcement learning algorithm optimizes the simulation results by continuously adjusting parameters to ensure the accuracy and reliability of the prediction.

[0088] Step 202: Based on the performance trend prediction and the location of the potential fault point, the historical fault data and the current operation status information are jointly reasoned and processed in combination with the Bayesian network and the Markov decision process algorithm to evaluate the probability distribution of different faults and the degree of impact of different faults on the performance of the air conditioning system, and generate a preliminary fault impact assessment report;

[0089] In this step, Bayesian network refers to a probabilistic graphical model used to represent the dependencies between variables and evaluate the probability distribution of different faults through joint reasoning; Markov decision process refers to a mathematical framework used to model decision problems with uncertainty and is suitable for optimal control in dynamic environments;

[0090] Based on the performance trend prediction and potential fault point location generated in step 201, the Bayesian network and MDP algorithm are used to perform joint reasoning on the historical fault data and the current operating status information. This process evaluates the probability distribution of different faults and their impact on the performance of the air conditioning system, generates a detailed preliminary fault impact assessment report, and provides a scientific basis for subsequent fault location and maintenance.

[0091] Step 203: Based on the preliminary fault impact assessment report, a graph neural network algorithm is used to model the interdependencies between components within the air-conditioning system, simulate the interaction between components under different fault scenarios, and generate an accurate fault impact assessment report;

[0092] In this step, graph neural network refers to a deep learning model that is suitable for processing graph structure data, can capture complex dependencies between nodes, and is suitable for simulating the interaction between components within the system;

[0093] Based on the preliminary fault impact assessment report, the graph neural network algorithm was used to model the interdependencies between the components within the air conditioning system. The GNN algorithm simulated the interaction between components under different fault scenarios, revealed the fault propagation path and impact range, and finally generated an accurate fault impact assessment report, which described in detail the performance of each component under different fault conditions, providing operation and maintenance personnel with a more accurate fault diagnosis tool.

[0094] Step 204: Based on the accurate fault impact assessment report, a multi-objective optimization algorithm is used to comprehensively calculate energy-saving standards and user comfort requirements, and generate personalized operation and maintenance suggestions;

[0095] In this step, the multi-objective optimization algorithm refers to an optimization method that can seek the best balance between multiple objectives, such as energy saving standards and user comfort requirements;

[0096] Based on the precise fault impact assessment report, a multi-objective optimization algorithm is used to comprehensively consider energy-saving standards and user comfort requirements to generate personalized operation and maintenance suggestions. The algorithm weighs multiple factors to ensure that the proposed operation and maintenance solution not only improves energy efficiency but also maintains the user's comfort experience. These suggestions include specific parameter adjustments, preventive maintenance measures, etc., aiming to extend the life of the equipment and improve the overall performance of the system.

[0097] Based on this, the present invention provides a specific embodiment, wherein step 303, based on the preliminary fault impact assessment report, applies a graph neural network algorithm to model the interdependencies between components within the air-conditioning system, simulates the interaction between components under different fault scenarios, and generates an accurate fault impact assessment report, specifically comprising the following steps:

[0098] Step 301: Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and the importance of the key components under different operating conditions, and obtaining an enhanced component dependency model;

[0099] In this step, based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are first modeled using a graph neural network algorithm combined with an attention mechanism. The graph neural network captures the direct and indirect connections between components, and the attention mechanism highlights the key components that have a significant impact under different operating conditions. This generates an enhanced component dependency model that not only describes the static connections, but also reflects the dependencies under dynamically changing conditions, providing detailed information support for subsequent analysis.

[0100] Step 302: Based on the enhanced component dependency model, a graph neural network is combined with a time series prediction algorithm to dynamically simulate the interaction between components under different fault scenarios, calculate the time propagation characteristics of the interaction between components, and obtain a fault propagation path prediction including a time dimension;

[0101] In this step, based on the enhanced component dependency model, the graph neural network is combined with the time series prediction algorithm to dynamically simulate the interaction between components under different fault scenarios. This process takes into account the time propagation characteristics of the interaction between components, that is, how the fault gradually spreads to other components over time. In this way, a fault propagation path prediction including the time dimension is obtained, which can more accurately understand the development process and impact range of the fault, and provide a scientific basis for taking timely preventive measures.

[0102] Step 303: Based on the fault propagation path prediction, a variational autoencoder is used to refine the fault location in the preliminary fault impact assessment report, capture potential fault modes and quantify the uncertainty estimation, and generate an analysis report including a description of the fault scope and impact degree;

[0103] In this step, based on the above-mentioned fault propagation path prediction, a variational autoencoder (VAE) is used to refine the fault location in the preliminary fault impact assessment report. The variational autoencoder can capture potential fault modes and represent these modes through implicit variables, thereby improving the accuracy of fault location. In addition, it can quantify uncertainty estimates to help operation and maintenance personnel better understand the possibility of faults and their impact. The final analysis report describes in detail the specific scope and impact of the fault, providing a solid foundation for formulating specific maintenance plans.

[0104] Step 304: Based on the analysis report, a causal inference method is introduced to analyze the causal relationship between the cause and consequence of the fault, and an integrated learning method is used to comprehensively evaluate the impact of the fault to generate an accurate fault impact assessment report;

[0105] In this step, based on the above analysis report, the causal inference method is introduced to analyze the causal relationship between the cause and consequence of the fault. This method can not only identify which factors directly lead to the occurrence of the fault, but also reveal how these factors interact with each other. In order to comprehensively evaluate the impact of the fault, the results of multiple algorithms are integrated through ensemble learning methods to ensure the accuracy and reliability of the evaluation results. The final accurate fault impact assessment report not only includes detailed fault point precise positioning and diffusion path prediction, but also includes the chain reaction that the fault may cause and the impact on the overall performance of the air-conditioning system. It provides detailed information support for subsequent operation and maintenance decisions.

[0106] Based on this, the present invention provides a specific embodiment, wherein step 301, based on the preliminary fault impact assessment report, uses a graph neural network algorithm combined with an attention mechanism to model the interdependencies between components within the air-conditioning system, highlights key components and the importance of key components under different operating conditions, and obtains an enhanced component dependency model, specifically comprising the following steps:

[0107] Step 401: Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, key components and their relative importance under different operating conditions are highlighted, and a preliminary component dependency model is obtained;

[0108] In this step, the attention mechanism refers to a mechanism used to highlight the importance of certain parts of the data and increase the model's attention to key information;

[0109] Based on the preliminary fault impact assessment report, we used a graph neural network combined with an attention mechanism to model the interdependencies between components within the air-conditioning equipment. We captured the direct and indirect connections between components through a graph neural network, and used the attention mechanism to highlight key components that have a significant impact under different operating conditions. This process generated a preliminary component dependency model that reflects the dependencies between components under static connections and dynamic changes, providing detailed information support for subsequent analysis.

[0110] Because traditional methods are difficult to effectively capture the complex interdependencies between components when dealing with complex systems, especially when facing dynamically changing working conditions, it is easy to misjudge or miss key information. Therefore, the graph neural network combined with the attention mechanism is introduced to improve the model's ability to capture the complex dependencies between components. The expression of the model capturing the complex dependencies between components is as follows:

[0111]

[0112] in, Represents the feature vector of node v in the lth layer. This parameter represents the state information of the node in the current layer, and gradually captures more complex dependencies as the number of layers increases; N(v) represents the neighbor set of node v, which defines the connection between each node and its neighboring nodes, ensuring that the component dependency model can capture local and global dependencies; α vu represents the attention weight of node u to node v. By introducing the attention mechanism, the component dependency model can dynamically adjust the importance of different neighbor nodes, highlight key components and their relative importance under different operating conditions. The attention is based on the similarity and interaction between node pairs, so that the component dependency model can focus on the most relevant dependencies.(l) represents the weight matrix of the lth layer, which is used to transform node features so that the component dependency model can learn new feature representations at each layer, thereby better capturing dependencies at different levels; σ represents the activation function, such as ReLU or LeakyReLU. The nonlinear activation function introduces nonlinear transformations, enabling the component dependency model to handle complex dependencies and nonlinear problems; b (l) : Bias increases the flexibility of the component dependency model and allows the component dependency model to be translated during feature conversion, which helps improve the expressiveness of the model; a represents the weight vector in the attention mechanism, which is used to calculate the similarity score between node pairs and determine which node pairs are more important. Through continuous optimization during the training process, the accuracy and generalization ability of the component dependency model can be improved; LeakyReLU represents an improved ReLU activation function that allows small negative values ​​to pass through, avoids the dead zone problem, and enhances the expressiveness of the component dependency model;

[0113] Among them, the attention weight α vu Calculated by:

[0114]

[0115] Among them, α vu represents the attention weight of node u to node v, which is used to measure the importance of node u to node v in the neighbor set of node v. Through the attention mechanism, the component dependency model can dynamically adjust the weights of different neighbor nodes, highlighting the key components and their relative importance under different operating conditions; exp(·) represents an exponential function, which is used to map the input value to the positive real number range, ensure that the attention weight is positive, and can amplify a large similarity score; LeakyReLU(·) represents an improved ReLU activation function that allows small negative values ​​to pass through, which is used to introduce nonlinear transformation, avoid dead zone problems, enhance the expressiveness of the model, and control the slope of the negative part. The specific form is:

[0116] Where α is a small positive number (such as 0.01); a T Represents the transpose of the weight vector a in the attention mechanism, which is used to calculate the similarity scores between node pairs and determine which node pairs are more important. It is continuously optimized through the training process to improve the accuracy and generalization ability of the component dependency model. (l) Represents the weight matrix of the lth layer, which is used to transform node features so that the model can learn new feature representations at each layer, thereby better capturing dependencies at different levels; and Represents the feature vectors of node v and node u in the lth layer, which is used to represent the state information of the node in the current layer, and gradually captures more complex dependencies as the number of layers increases; || represents the connection operator (Concatenation), which is used to concatenate two vectors into a larger vector. In this formula, it is used to connect the feature vectors of node v and node u or k to form a new feature vector as the input for subsequent calculations.

[0117] N(v) represents the set of neighbors of node v, defining the connection between each node and its neighboring nodes, ensuring that the model can capture local and global dependencies. In the denominator, it is used to normalize the attention weights of all neighboring nodes so that the final α vu The value falls in the interval [0,1].

[0118] By introducing graph neural networks combined with attention mechanisms, the component dependency model is able to capture complex dependencies between components, especially strong and weak dependencies between components within air-conditioning equipment. The component dependency model can identify and evaluate them more accurately, thereby improving the accuracy of fault diagnosis and the scientific nature of operation and maintenance decisions. In addition, through a multi-layer structure and appropriate activation function selection, the component dependency model has stronger expressive power and higher robustness, ensuring reliable predictions and support under various working conditions.

[0119] Step 402: According to the component dependency model, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the model, thereby obtaining an optimized component dependency model;

[0120] In this step, the node centrality measure refers to an indicator that measures the importance of a node in the network, such as PageRank or Betweenness Centra l ity;

[0121] Based on the preliminary component dependency model, the weights of each node in the graph neural network are adjusted in combination with the node centrality measurement, which improves the model's attention to key components, enhances learning efficiency and prediction accuracy. After adjustment, the model can more accurately reflect the changes in the dependency relationships between components under different operating conditions, thereby generating an optimized component dependency model.

[0122] Step 403: Based on the optimized component dependency model, adaptive threshold setting is combined with a hierarchical clustering algorithm to perform correlation analysis on the components of the air-conditioning system, identify key components that play a core role in normal and abnormal conditions, and quantify the importance indexes of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information.

[0123] In this step, adaptive threshold setting refers to automatically determining a suitable threshold according to data characteristics to distinguish different levels of dependency strength; hierarchical clustering algorithm refers to a clustering method that gradually groups objects according to similarity to form a tree structure;

[0124] Based on the optimized component dependency model, a method combining adaptive threshold setting and hierarchical clustering algorithm is used to conduct correlation analysis on the components of air-conditioning equipment. It not only identifies the key components that play a core role in normal and abnormal conditions, but also quantifies the importance indicators of these components, and analyzes the hierarchical structure and functional modules between key components. Finally, a component dependency description containing hierarchical information is generated, which records in detail the location, importance and interaction mode of each key component with other components.

[0125] Step 404: According to the component dependency description, a Bayesian optimization algorithm is introduced to adjust the model parameters to generate an enhanced component dependency model;

[0126] In this step, the Bayesian optimization algorithm refers to a method for finding the optimal solution of a function, which is particularly suitable for black box optimization problems.

[0127] Based on the component dependency description containing hierarchical information, the Bayesian optimization algorithm is used to fine-tune the model parameters, ensuring that the component dependency model parameters can most effectively reflect the dynamically changing dependencies between components under different operating conditions. The cross-validation strategy is used to improve the robustness and generalization ability of the component dependency model, and ultimately an enhanced component dependency model is generated, which provides detailed information support for subsequent fault propagation path prediction.

[0128] Based on this, the present invention provides a specific embodiment, wherein step 403, based on the optimized component dependency model, uses adaptive threshold setting combined with a hierarchical clustering algorithm to perform association analysis on the components of the air-conditioning system, identifies key components that play a core role in normal and abnormal conditions, and quantifies the importance indexes of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information, which specifically includes the following steps:

[0129] Step 501: Based on the optimized component dependency model, adaptive threshold setting is combined with a hierarchical clustering algorithm, and a community detection algorithm is introduced to perform association analysis on the components of the air-conditioning system to obtain component grouping results that distinguish strong dependencies and weak dependencies;

[0130] In this step, a method combining adaptive threshold setting and hierarchical clustering algorithm is adopted based on the optimized component dependency model, and a community detection algorithm is introduced to perform association analysis on the components of the air-conditioning equipment. This process automatically determines the critical value of the dependency strength between components through adaptive threshold setting to distinguish strong dependency from weak dependency. The community detection algorithm is used to identify natural groupings between components and improve the accuracy of clustering. Finally, a component grouping result that distinguishes strong dependency from weak dependency is generated, providing a basis for subsequent analysis.

[0131] Step 502: According to the component grouping result, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the component dependency model, and obtain an adjusted component dependency model;

[0132] In this step, the weights of each node in the graph neural network are adjusted based on the component grouping results and combined with the node centrality measurement. The node centrality measurement helps identify key nodes and ensures that the model pays more attention to these important nodes. By adjusting the node weights, the learning efficiency and prediction accuracy of the component dependency model are optimized, and an adjusted component dependency model is generated, which improves the model's ability to reflect changes in component dependencies under different operating conditions.

[0133] Step 503: Based on the adjusted component dependency model, a random forest learning algorithm is used to quantitatively evaluate the importance of the components to obtain an importance index for each component;

[0134] In this step, based on the adjusted component dependency model, the importance of each component is quantitatively evaluated using the random forest learning algorithm. The random forest algorithm can handle nonlinear relationships and provide feature importance scores, thereby more accurately quantifying the importance indicators of each component. It not only considers static connections, but also analyzes component dependencies under dynamically changing conditions to ensure the comprehensiveness and accuracy of importance evaluation.

[0135] Step 504: according to the importance index of each key component, apply social network analysis technology to analyze the hierarchical structure and functional modules between the key components, and generate a detailed hierarchical structure and functional module description;

[0136] In this step, according to the importance index of each key component, social network analysis technology is used to analyze the hierarchical structure and functional modules between key components. SNA technology reveals the interaction patterns between components and their roles in the network, ensuring that the description records in detail the location, importance and interaction mode of each key component with other components. This step generates a detailed description of the hierarchical structure and functional modules, providing detailed information support for the subsequent fault propagation path prediction.

[0137] Step 505: According to the hierarchical structure and functional module description, a Bayesian optimization algorithm is introduced to adjust the component dependency model parameters to generate a component dependency description containing hierarchical information;

[0138] In this step, the Bayesian optimization algorithm is introduced to fine-tune the parameters of the component dependency model based on the hierarchical structure and functional module description. The Bayesian optimization algorithm can efficiently find the optimal solution in the complex parameter space, ensuring that the parameters of the component dependency model most effectively reflect the dynamically changing dependencies between components under different operating conditions. The cross-validation strategy is used to improve the robustness and generalization ability of the component dependency model, and finally a component dependency description containing hierarchical information is generated, providing a scientific basis for operation and maintenance decisions.

[0139] Based on this, the present invention provides a specific embodiment, wherein step 103, based on the personalized operation and maintenance suggestion, uses a genetic algorithm to search for an operating parameter combination with the lowest energy consumption, and uses adaptive control theory to smoothly adjust the operating parameters of the air conditioning system to obtain an optimized air conditioning system setting, specifically comprising the following steps:

[0140] Step 601: Based on the personalized operation and maintenance suggestion, a genetic algorithm is used in combination with a multi-objective optimization algorithm, such as a search process is performed on the operation parameter space of the air conditioning system to obtain a preliminary operation parameter set that minimizes energy consumption and satisfies user comfort and system stability;

[0141] In this step, based on the personalized operation and maintenance suggestions, a genetic algorithm combined with a multi-objective optimization algorithm (such as NSGA-II) is used to comprehensively search the operating parameter space of the air-conditioning equipment. This process aims to find the operating parameter combination with the lowest energy consumption while satisfying user comfort and system stability. The genetic algorithm finds the optimal solution by simulating the natural selection process, while the multi-objective optimization algorithm ensures that the best balance point is found under multiple constraints (such as energy saving and comfort), generating a preliminary set of operating parameters, which provides a basis for subsequent adjustments.

[0142] Step 602: According to the preliminary operating parameter set, the operating parameters of the air conditioning system are smoothly adjusted by using adaptive control theory to obtain an adjusted operating parameter configuration;

[0143] In this step, based on the preliminary operating parameter set, the operating parameters of the air-conditioning equipment are gradually adjusted using the adaptive control theory. The adaptive control theory can dynamically adjust the parameters during operation to ensure a smooth transition from the current settings to the new settings, avoiding the negative impact of sudden changes, and ensuring that the adjustment of the operating parameters is both efficient and will not cause user discomfort or equipment wear. The adjusted operating parameter configuration is generated, providing a stable operating environment for the next step of monitoring.

[0144] Step 603: Based on the adjusted operating parameter configuration, a real-time performance monitoring mechanism is introduced, and the actual operating performance of the air-conditioning system is monitored and adjusted using an online learning algorithm to obtain a performance-based operating parameter configuration;

[0145] In this step, based on the adjusted operating parameter configuration, a real-time performance monitoring mechanism is introduced, and the actual operating performance of the air-conditioning equipment is continuously monitored and fine-tuned using an online learning algorithm. The online learning algorithm can quickly respond to environmental changes and dynamically adjust operating parameters to maintain the efficient operation of the air-conditioning system. This not only improves the operating efficiency, but also ensures that the equipment is always in the best working condition, generates a performance-based operating parameter configuration, and further improves the stability and energy efficiency of the air-conditioning system.

[0146] Step 604: Based on the performance-based operating parameter configuration, a reinforcement learning algorithm is applied to evaluate and optimize the long-term impact of the operating parameters, and an operating parameter solution with a long-term optimization effect is generated;

[0147] In this step, reinforcement learning algorithms, especially deep Q networks, are applied to evaluate and optimize the long-term impact of operating parameters based on performance-based operating parameter configuration. The reinforcement learning algorithm simulates the effects of different operation and maintenance decisions to find the optimal action path to maximize long-term benefits. It not only considers minimizing short-term energy consumption, but also ensures efficient and stable operation of the equipment through long-term performance optimization, and maximizes its service life, generating an operating parameter solution with long-term optimization effects.

[0148] Step 605: generating optimized air conditioning system settings by synthesizing the preliminary operating parameter set, the adjusted operating parameter configuration, the performance-based operating parameter configuration, and the operating parameter scheme with long-term optimization effect;

[0149] In this step, the results of all previous steps are integrated, including: preliminary operating parameter set, adjusted operating parameter configuration, performance-based operating parameter configuration and operating parameter scheme with long-term optimization effect, and finally the optimized air-conditioning equipment settings are generated. This not only takes into account the minimization of short-term energy consumption, but also ensures the efficient and stable operation of the equipment through long-term performance optimization, and maximizes the service life. The whole process forms a complete intelligent management process, which ensures the optimal operation of the air-conditioning equipment under various working conditions, while improving the user experience and service quality.

[0150] Based on this, the present invention provides a specific embodiment, in which step 104, when an abnormal situation is detected according to the optimized air conditioning system setting, the severity of the fault is judged by combining multi-level threshold setting with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel, specifically including the following steps:

[0151] Step 701: when an abnormal situation is detected according to the optimized air conditioning system settings, a combination of multi-level threshold setting and fuzzy logic is used to perform severity judgment processing on the abnormal data, distinguish different levels of faults, and evaluate the uncertainty and complexity of the faults to obtain an accurate fault severity evaluation result;

[0152] In this step, under the optimized air-conditioning equipment settings, the operating status is continuously monitored. Once an abnormal situation is detected, the method combining multi-level threshold setting and fuzzy logic is immediately initiated to judge the severity of the abnormal data. The multi-level threshold setting is used to distinguish between minor abnormalities and serious faults, while fuzzy logic is suitable for handling working conditions with high uncertainty and complexity, helping to more accurately assess the severity of the fault, generate accurate fault severity assessment results, and provide a scientific basis for subsequent responses.

[0153] Step 702: Based on the fault severity assessment result, an early warning mechanism is immediately triggered through an event-driven architecture, and a rapid response is given to the operation and maintenance personnel according to different fault levels;

[0154] In this step, based on the fault severity assessment results, the early warning mechanism is triggered immediately through the event-driven architecture. The architecture can automatically activate the corresponding early warning process according to pre-defined rules and conditions to ensure that potential problems are handled in a timely manner. According to different fault levels, detailed emergency response plans are quickly pushed to operation and maintenance personnel, including fault location, impact assessment, and recommended maintenance measures, to ensure that faults are responded to quickly and effectively.

[0155] Step 703: According to the content of the emergency response plan, a natural language processing algorithm is used to automatically generate a fault description and recommended maintenance measures, convert technical terms into natural language, and generate an emergency response guide;

[0156] In this step, based on the content of the emergency response plan, the natural language processing (NLP) algorithm is used to automatically generate fault descriptions and recommended maintenance measures. The NLP algorithm can convert technical terms into easy-to-understand natural language, ensuring that operation and maintenance personnel and other relevant parties can easily understand the fault situation and handling methods, and generate emergency response guidelines, which not only improves the efficiency of information transmission, but also enhances the clarity of communication, ensuring that all relevant personnel can take appropriate actions quickly.

[0157] Step 704: accurately locate the faulty equipment in combination with the real-time positioning system, and generate maintenance auxiliary information including the fault location and visual guidance through augmented reality technology;

[0158] In this step, the real-time positioning system is used to accurately locate the position of the faulty equipment to ensure that the operation and maintenance personnel can quickly find the fault point. At the same time, augmented reality technology is used to generate maintenance auxiliary information including the fault location and visual guidance. AR technology can overlay virtual information in the real environment and provide intuitive operation guidance, helping the operation and maintenance personnel to complete maintenance tasks more efficiently and reduce the possibility of misoperation.

[0159] Step 705: combining the maintenance auxiliary information and the emergency response guide with the emergency response plan, generating an emergency response plan including the maintenance auxiliary information and the emergency response guide, and pushing it to the relevant responsible party through the operation and maintenance platform;

[0160] In this step, maintenance auxiliary information and emergency response guidelines are integrated and combined with the emergency response plan to generate a comprehensive emergency response plan. This plan records in detail information such as the fault location, maintenance steps, required tools and materials, and is pushed to the relevant responsible parties through the operation and maintenance platform. The operation and maintenance platform ensures the timeliness and accuracy of information transmission, so that all relevant personnel can receive consistent information at the same time, improve collaboration efficiency, and ensure smooth fault handling.

[0161] Figure 2 The present invention provides a schematic diagram of a structure of an intelligent operation and maintenance system for an air conditioning system based on Internet of Things technology, such as Figure 2 As shown, the system includes:

[0162] The compression module 21 is used to use the smart sensors deployed in the key components of the air-conditioning system and the edge computing nodes of the air-conditioning system to preliminarily process and compress the operation status data, and combine it with the environmental parameters to obtain a comprehensive operation information package;

[0163] Evaluation module 22, used to perform simulation analysis using the digital twin model in the cloud based on the comprehensive operation information package, generate performance trend prediction and potential fault point evaluation of the air-conditioning system, and jointly reason historical fault data and current operation information through the Bayesian network algorithm to evaluate the probability distribution of fault occurrence and generate personalized operation and maintenance suggestions;

[0164] An adjustment module 23 is used to search for the operating parameter combination with the lowest energy consumption by using a genetic algorithm based on the personalized operation and maintenance suggestion, and to smoothly adjust the operating parameters of the air-conditioning system by using an adaptive control theory to obtain an optimized air-conditioning system setting;

[0165] The judgment module 24 is used to judge the severity of the fault by combining multi-level threshold setting with fuzzy logic when an abnormal situation is detected according to the optimized air conditioning system setting, and to instantly trigger the early warning mechanism through the event-driven architecture to push the emergency response plan to the operation and maintenance personnel;

[0166] The updating module 25 is used for the operation and maintenance personnel to implement corresponding maintenance actions based on the emergency response plan, and record the actual maintenance process and the actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

[0167] Figure 2 The intelligent operation and maintenance system of air conditioning system based on Internet of Things technology can execute Figure 1 The implementation principle and technical effect of the intelligent operation and maintenance method of an air conditioning system based on the Internet of Things technology described in the embodiment shown are not repeated here. The specific way in which each module and unit performs operations in the intelligent operation and maintenance system of an air conditioning system based on the Internet of Things technology in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0168] In one possible design, Figure 2 The embodiment shown in the figure is an intelligent operation and maintenance system for air conditioning system based on Internet of Things technology, which can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0169] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0170] The processing component 32 is used to use the smart sensors deployed in the key components of the air-conditioning system and the edge computing nodes of the air-conditioning system to perform preliminary processing and compression on the operating status data, and combine it with environmental parameters to obtain a comprehensive operating information package; based on the comprehensive operating information package, the digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation for the air-conditioning system, and the historical fault data and current operating information are jointly inferred through the Bayesian network algorithm to evaluate the probability distribution of the occurrence of faults and generate personalized operation and maintenance suggestions; based on the personalized operation and maintenance suggestions, a genetic algorithm is used to search for the operating parameter combination with the lowest energy consumption, and the operating parameters of the air-conditioning system are smoothly adjusted using adaptive control theory , and obtain the optimized air-conditioning system settings; when an abnormal situation is detected according to the optimized air-conditioning system settings, the severity of the fault is judged by combining multi-level threshold settings with fuzzy logic, and the early warning mechanism is immediately triggered through the event-driven architecture to push the emergency response plan to the operation and maintenance personnel; based on the emergency response plan, the operation and maintenance personnel implement corresponding maintenance actions, and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

[0171] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0172] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0173] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0174] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0175] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0176] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0177] The embodiment of the present invention further provides a computer storage medium storing a computer program, which can achieve the above-mentioned Figure 1 The illustrated embodiment is an intelligent operation and maintenance method for an air conditioning system based on Internet of Things technology.

[0178] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0179] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0180] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0181] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart operation and maintenance method for air conditioning system based on Internet of Things technology, characterized in that: include: Using smart sensors deployed in key components of the air-conditioning system and edge computing nodes of the air-conditioning system, the operation status data is preliminarily processed and compressed, and combined with environmental parameters to obtain a comprehensive operation information package; Based on the comprehensive operation information package, a digital twin model in the cloud is used to perform simulation analysis to generate a performance trend prediction and potential fault point evaluation for the air-conditioning system, and a Bayesian network algorithm is used to perform joint reasoning on historical fault data and current operation information to evaluate the probability distribution of fault occurrence and generate personalized operation and maintenance recommendations; Based on the personalized operation and maintenance suggestions, a genetic algorithm is used to search for an operating parameter combination with the lowest energy consumption, and adaptive control theory is used to smoothly adjust the operating parameters of the air-conditioning system to obtain an optimized air-conditioning system setting; When an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel; Based on the emergency response plan, the operation and maintenance personnel implement corresponding maintenance actions and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is equipped with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

2. The method according to claim 1, characterized in that Based on the comprehensive operation information package, the cloud-based digital twin model is used for simulation analysis to generate performance trend predictions and potential fault point assessments for the air-conditioning system. The Bayesian network algorithm is used to jointly reason historical fault data and current operation information, evaluate the probability distribution of fault occurrence, and generate personalized operation and maintenance recommendations, including: The digital twin model deployed in the cloud is combined with the reinforcement learning algorithm to perform high-precision simulation analysis on the comprehensive operation information package to obtain the performance trend prediction of the air-conditioning system and the possible fault point location; Based on the performance trend prediction and the location of potential fault points, the Bayesian network and Markov decision process algorithm are combined to perform joint reasoning processing on historical fault data and current operating status information, evaluate the probability distribution of different faults and the degree of impact of different faults on the performance of the air conditioning system, and generate a preliminary fault impact assessment report; Based on the preliminary fault impact assessment report, the graph neural network algorithm is used to model the interdependencies between the internal components of the air-conditioning system, simulate the interactive effects between the components under different fault scenarios, and generate an accurate fault impact assessment report; Based on the precise fault impact assessment report, a multi-objective optimization algorithm is used to comprehensively calculate energy-saving standards and user comfort requirements to generate personalized operation and maintenance recommendations; 3. The method according to claim 2, characterized in that Based on the preliminary FIA report, the graph neural network algorithm is used to model the interdependencies between the internal components of the air conditioning system, simulate the interaction between the components under different fault scenarios, and generate an accurate FIA ​​report, including: Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and their importance under different operating conditions, and obtaining an enhanced component dependency model. According to the enhanced component dependency model, the interaction effects between components under different fault scenarios are dynamically simulated by combining graph neural network with time series prediction algorithm, and the time propagation characteristics of the interaction effects between components are calculated to obtain the fault propagation path prediction including the time dimension. Based on the fault propagation path prediction, a variational autoencoder is used to refine the fault location in the preliminary fault impact assessment report, capture potential fault modes and quantify the uncertainty estimate, and generate an analysis report containing a description of the fault scope and impact degree; According to the analysis report, a causal inference method is introduced to analyze the causal relationship between the cause and consequence of the fault, and an integrated learning method is used to comprehensively evaluate the impact of the fault and generate an accurate fault impact assessment report.

4. The method according to claim 3, characterized in that Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and their importance under different operating conditions. An enhanced component dependency model is obtained, including: Based on the preliminary fault impact assessment report, the interdependencies between the internal components of the air-conditioning system are modeled using a graph neural network algorithm combined with an attention mechanism, highlighting the key components and their relative importance under different operating conditions, and obtaining a preliminary component dependency model. According to the component dependency model, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the model, thereby obtaining an optimized component dependency model; Based on the optimized component dependency model, the adaptive threshold setting is combined with the hierarchical clustering algorithm to perform correlation analysis on the components of the air-conditioning system, identify the key components that play a core role in normal and abnormal conditions, and quantify the importance indicators of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information. According to the component dependency description, a Bayesian optimization algorithm is introduced to adjust the model parameters to generate an enhanced component dependency model; 5. The method according to claim 4, characterized in that Based on the optimized component dependency model, the adaptive threshold setting is combined with the hierarchical clustering algorithm to perform correlation analysis on the components of the air-conditioning system, identify the key components that play a core role in normal and abnormal conditions, and quantify the importance indicators of the key components. At the same time, the hierarchical structure and functional modules between the key components are analyzed to generate a component dependency description containing hierarchical information, including: Based on the optimized component dependency model, the adaptive threshold setting is combined with the hierarchical clustering algorithm, and the community detection algorithm is introduced to perform association analysis on the components of the air-conditioning system, and the component grouping results that distinguish strong and weak dependencies are obtained; According to the component grouping results, the node weights in the graph neural network are adjusted in combination with the node centrality measurement to optimize the learning efficiency and prediction accuracy of the component dependency model, thereby obtaining an adjusted component dependency model; Based on the adjusted component dependency model, the importance of components is quantitatively evaluated using the random forest learning algorithm to obtain the importance index of each component; According to the importance index of each key component, social network analysis technology is used to analyze the hierarchical structure and functional modules between the key components, and a detailed description of the hierarchical structure and functional modules is generated; According to the hierarchical structure and functional module description, a Bayesian optimization algorithm is introduced to adjust the component dependency model parameters to generate a component dependency description containing hierarchical information.

6. The method according to claim 1, characterized in that Based on the personalized operation and maintenance suggestions, a genetic algorithm is used to search for the operating parameter combination with the lowest energy consumption, and the adaptive control theory is used to smoothly adjust the air conditioning system operating parameters to obtain the optimized air conditioning system settings, including: Based on the personalized operation and maintenance suggestions, a genetic algorithm is used in combination with a multi-objective optimization algorithm to search and process the operating parameter space of the air-conditioning system to obtain a preliminary operating parameter set that minimizes energy consumption and satisfies user comfort and system stability; According to the preliminary operating parameter set, the operating parameters of the air conditioning system are smoothly adjusted by using adaptive control theory to obtain an adjusted operating parameter configuration; Based on the adjusted operating parameter configuration, a real-time performance monitoring mechanism is introduced, and the actual operating performance of the air-conditioning system is monitored and adjusted using an online learning algorithm to obtain a performance-based operating parameter configuration; According to the performance-based operating parameter configuration, a reinforcement learning algorithm is applied to evaluate and optimize the long-term impact of the operating parameters to generate an operating parameter solution with a long-term optimization effect; The optimized air conditioning system settings are generated by combining the preliminary operating parameter set, the adjusted operating parameter configuration, the performance-based operating parameter configuration, and the operating parameter scheme with long-term optimization effect.

7. The method according to claim 1, characterized in that When an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is determined by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel, including: When an abnormal situation is detected according to the optimized air conditioning system settings, a combination of multi-level threshold settings and fuzzy logic is used to perform severity judgment processing on the abnormal data, distinguish different levels of faults, and evaluate the uncertainty and complexity of the faults to obtain accurate fault severity assessment results; Based on the fault severity assessment results, an early warning mechanism is immediately triggered through an event-driven architecture, and a rapid response is given to the operation and maintenance personnel according to different fault levels. Based on the content of the emergency response plan, a natural language processing algorithm is used to automatically generate a fault description and recommended maintenance measures, convert technical terms into natural language, and generate an emergency response guide; Combined with the real-time positioning system, the faulty equipment is accurately located, and augmented reality technology is used to generate maintenance assistance information including the fault location and visual guidance; The maintenance auxiliary information and emergency response guide are combined with the emergency response plan to generate an emergency response plan conclusion including the maintenance auxiliary information and the emergency response guide, and the conclusion is pushed to the relevant responsible parties through the operation and maintenance platform.

8. An intelligent operation and maintenance system for air conditioning system based on Internet of Things technology, characterized in that: include: The compression module is used to use the smart sensors deployed in the key components of the air-conditioning system and the edge computing nodes of the air-conditioning system to preliminarily process and compress the operation status data, and combine it with environmental parameters to obtain a comprehensive operation information package; An evaluation module is used to perform simulation analysis using a digital twin model in the cloud based on the comprehensive operation information package to generate a performance trend prediction and potential failure point evaluation for the air-conditioning system, and to perform joint reasoning on historical failure data and current operation information through a Bayesian network algorithm to evaluate the probability distribution of failure occurrence and generate personalized operation and maintenance recommendations; An adjustment module, for searching for an operating parameter combination with the lowest energy consumption by using a genetic algorithm based on the personalized operation and maintenance suggestion, and for smoothly adjusting the operating parameters of the air-conditioning system by using an adaptive control theory to obtain an optimized air-conditioning system setting; Judgment module: when an abnormal situation is detected according to the optimized air conditioning system settings, the severity of the fault is judged by combining multi-level threshold settings with fuzzy logic, and an early warning mechanism is immediately triggered through an event-driven architecture to push an emergency response plan to the operation and maintenance personnel; The update module is used for the operation and maintenance personnel to implement corresponding maintenance actions based on the emergency response plan, and record the actual maintenance process and actual maintenance results through the operation and maintenance platform. At the same time, the operation and maintenance platform is provided with a user feedback interface, allowing users to submit usage experience and satisfaction evaluation, obtain user feedback, and use an incremental machine learning algorithm to learn the user feedback and actual maintenance results, update the digital twin model, and form a closed-loop operation and maintenance system.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an intelligent operation and maintenance method for an air-conditioning system based on Internet of Things technology as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, an intelligent operation and maintenance method for an air-conditioning system based on Internet of Things technology as described in any one of claims 1 to 7 is implemented.

Images