Abnormality detection and diagnosis method, system and equipment for water chilling unit of central air conditioner and medium

Abstract

The invention provides a central air conditioner water chilling unit anomaly detection and diagnosis method, system and device and a medium, and belongs to the technical field of industrial equipment predictive maintenance. The method comprises the following steps that time sequence data and log data of operation of an air conditioner unit are collected and preprocessed; performing fine-tuning training on the time sequence detection model by using the time sequence data of the operation of the air conditioning unit, and performing fine-tuning training on the large language model by using the log data of the operation of the air conditioning unit; and the real-time operation data of the air conditioning unit are preprocessed and then input into the time sequence prediction model for anomaly detection, and when anomaly is detected, log data of operation of the air conditioning unit in an abnormal time period are obtained and input into the large language model for anomaly diagnosis. The operation state of the water chilling unit can be monitored in real time, the diagnosis process is automatically triggered when abnormity is detected, automation and intelligentization of operation and maintenance management are achieved, the abnormity response speed is increased, and the maintenance cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of predictive maintenance of industrial equipment, and specifically relates to a method, system, equipment and medium for detecting and diagnosing abnormalities of a central air-conditioning chiller. Background Art

[0002] Traditional central air-conditioning chiller operation and maintenance practices have always relied on manual monitoring and diagnosis. This method is not only inefficient, but also due to the limitations of professional knowledge and experience, it is often difficult to make quick and accurate judgments on complex abnormal patterns and deep-seated system problems. This not only increases maintenance costs, but also prolongs equipment downtime, affecting overall operating efficiency and production stability.

[0003] For anomaly detection, currently some anomaly detection methods based on deep learning are more commonly used, such as autoencoders, convolutional neural networks, and long short-term memory networks, which can learn and identify complex patterns and abnormal behaviors in data. These methods have certain effects in the fields of image and speech recognition, but these methods usually require a large amount of labeled data for training, which is often not feasible in industrial applications because obtaining a large amount of labeled abnormal data is time-consuming and expensive. In addition, deep learning models have poor interpretability. When the model predicts an abnormality, it is difficult to explain why such a prediction occurs, which is a major disadvantage for industrial systems that require high reliability and interpretability.

[0004] Another way to detect anomalies is to analyze the operation logs of important air-conditioning chillers. However, the current mainstream method of log analysis mainly relies on regular expression matching, keyword search and simple statistical analysis. However, the implementation of these methods has obvious limitations, such as the inability to capture complex patterns and abnormal behaviors in log data. On the other hand, a large amount of manual participation is required for log screening, classification and interpretation, which is inefficient and easily affected by human factors. Summary of the invention

[0005] In a first aspect, an embodiment of the present application provides a method for detecting and diagnosing abnormalities of a central air-conditioning chiller, comprising the following steps: S1. Collect the time series data and log data of the air conditioning unit operation and pre-process them; S2. Use the time series data of the air conditioning unit operation to fine-tune the time series detection model, and use the log data of the air conditioning unit operation to fine-tune the large language model; S3. The real-time operation data of the air-conditioning unit is pre-processed and then input into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the operation of the air-conditioning unit during the abnormal period is obtained and input into the large language model for anomaly diagnosis.

[0006] Furthermore, the specific steps of step S1 are as follows: S11. Collecting the time series data of the operation of the air conditioning unit, the time series data of the operation of the air conditioning unit includes the inlet and outlet temperature of the chilled water, the inlet and outlet temperature of the cooling water, the inlet and outlet temperature of the cooling water loop heat exchanger, the inlet and outlet temperature of the chilled water loop heat exchanger and the compressor power; S12. Clean the time series data of the operation of the air-conditioning unit by removing random noise through filtering algorithm, fill in missing data through linear interpolation, unify the data format and measurement unit, and complete standardization; S13. Collecting log data of the air conditioning unit, the log data of the air conditioning unit includes equipment operation status, equipment alarm information, maintenance and diagnostic information and historical abnormal information; S14. Perform text normalization on the log data of the collection group, truncate the log text according to the maximum input length of the large language model, and convert the log text into a format supported by the large language model input.

[0007] Furthermore, the specific steps of step S2 are as follows: S21. fine-tune the structure of the Moirai time series detection model according to the time series data of the air conditioning unit; S22. Use the time series data of the air conditioning unit to train and optimize the fine-tuned Moirai time series detection model to obtain an optimized Moirai time series detection model; S23. Fine-tune the backend layers of the Qwen large language model, add an adaptive layer to the feature extraction part, add an external knowledge base to the retrieval function of the RAG architecture, and limit the output exception report format to match the abnormal diagnosis of the air-conditioning unit.

[0008] Furthermore, the specific steps of step S21 are as follows: S211. Divide the time series data of the air conditioning unit operation into a training set and a validation set according to a set ratio; S212. Select N patch sizes, use multi-patch size input projection to segment the time series data of the training set and the validation set into time windows according to each patch size, and obtain N segmentation results; S213. Map each window of each segmentation result to a high-dimensional feature space using a patch embedding method to obtain a processed test set and a validation set; S214. Input the test set or validation set after the patch embedding into the Moirai time series detection model, use the attention mechanism to process the data layer by layer, and in the process of data processing, adjust the layers of the back end of the Moirai time series detection model within a set range or add an adaptive layer to match the air conditioning unit anomaly detection task; S215. Introduce a multi-layer arbitrary variable attention mechanism in the Moirai time series detection model, and process sensor data of different quantities and types.

[0009] Furthermore, the specific steps of step S22 are as follows: S221. Taking the negative log-likelihood function of the predicted distribution of the normal operation data of the air-conditioning unit as the objective function, the prediction error of the training sample is characterized; S222. The Moirai time series detection model is iteratively optimized on the training set and the validation set by using a cross-validation method, and the training is stopped when the objective function is minimized or the number of iterations is completed, thereby obtaining the optimized Moirai time series detection model.

[0010] Furthermore, the specific steps of step S23 are as follows: S231. Lock the front-end layer of the Qwen large language model, and adjust the back-end layer within a set range or add an adaptive layer; S232. Adding an adaptive layer to the feature extraction part of the Qwen large language model to adapt to the abnormal data distribution characteristics of the air-conditioning unit; S233. Use the retrieval function in the RAG architecture to build an external knowledge base containing air conditioning unit abnormality diagnosis knowledge, and introduce air conditioning chiller failure modes, processing suggestions and historical cases; S234. Simulate the input of the air conditioning unit abnormal diagnosis related problems into the Qwen language model, so that the Qwen language model calls the relevant historical cases and diagnostic information in the retrieval function of the RAG architecture, and adjusts the Qwen language model according to the retrieval results; S235. Modify the output layer of the Qwen large language model, add formatted content to the output abnormal diagnosis report, and match the annotation format of the abnormal diagnosis of the air-conditioning unit; the formatted content includes the abnormal report header, abnormal description, root cause analysis and modification suggestions.

[0011] Furthermore, the specific steps of step S3 are as follows: S31. Collect real-time operation data of air-conditioning units, and input them into the optimized Moirai time series detection model after preprocessing; S32. The Moirai time series detection model analyzes various sensor data in the time dimension and feature dimension through the attention mechanism, compares the learned normal operation characteristics of the air-conditioning unit with the input data, and obtains the prediction results of the input data; S33. Analyze the prediction results, regard the probability that the data falls outside the normal range as an abnormal score, and determine whether the abnormal score exceeds the warning threshold; If not, return to step S31; If yes, go to step S34; S34 triggers real-time warning, determines abnormal period, and extracts log data and time series data of air-conditioning unit operation during abnormal period; S35. After preprocessing the extracted log data of the air conditioning unit during the abnormal period, it is input into the fine-tuned Qwen large language model together with the abnormal score; S36. The fine-tuned Qwen large language model uses the model's natural language processing capabilities to analyze log and sensor data during the abnormal period, and performs feature analysis on multimodal data. By comparing context and abnormal features, and combining historical data, root cause analysis is performed. At the same time, the retrieval function in RAG is activated to retrieve abnormal records and diagnostic suggestions similar to the input data from the knowledge base; S37. The fine-tuned Qwen large language model generates a structured anomaly diagnosis report through multimodal fusion in accordance with the agreed format and combined with anomaly scores, historical similar cases, and contextual information; S38. Store the generated abnormal diagnosis report in the knowledge base to build the air-conditioning unit failure mode library, and further fine-tune the input and output of the Qwen large language model based on the abnormal diagnosis report.

[0012] In a second aspect, the embodiment of the present application further provides a central air-conditioning chiller abnormality detection and diagnosis system, comprising: A data acquisition unit, used to collect the time series data and log data of the operation of the air-conditioning unit and perform pre-processing; A model fine-tuning unit, used to fine-tune the time series detection model using the time series data of the air-conditioning unit operation, and to fine-tune the large language model using the log data of the air-conditioning unit operation; The anomaly detection and diagnosis unit is used to pre-process the real-time operation data of the air-conditioning unit and input it into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the operation of the air-conditioning unit during the abnormal period is obtained and input into the large language model for anomaly diagnosis.

[0013] In a third aspect, an embodiment of the present application further provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for abnormal detection and diagnosis of a central air-conditioning chiller unit as described in the first aspect are implemented.

[0014] In a fourth aspect, an embodiment of the present application further provides a storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the central air-conditioning chiller abnormality detection and diagnosis method as described in the first aspect are implemented.

[0015] It can be seen from the above technical solutions that the present invention has the following advantages: In the central air-conditioning chiller abnormality detection and diagnosis method, system, device and medium provided by the present application, by collecting and preprocessing the time series data and log data of the air-conditioning unit, and using the fine-tuned time series detection model for anomaly detection, the abnormal patterns in the data can be accurately identified, and the accuracy of detection is significantly improved. With the help of efficient attention mechanism and multi-layer arbitrary variable attention mechanism, comprehensive analysis of multimodal data is achieved, and the detection efficiency is improved; when an abnormality is detected, the log data of the abnormal period is deeply analyzed by using the fine-tuned large language model, and the root cause of the abnormality can be intelligently diagnosed and modification suggestions are given by combining historical data and similar cases in the knowledge base, which not only improves the intelligent level of diagnosis, but also enhances the interpretability of the diagnosis results, so that operation and maintenance personnel can more clearly understand the causes and solutions of the abnormality; through the automated anomaly detection and diagnosis process, the dependence on manual monitoring is reduced, and the misjudgment and missed judgment caused by insufficient professional knowledge and experience are reduced. At the same time, timely abnormal warnings and accurate diagnosis results help operation and maintenance personnel to quickly locate and solve problems, thereby effectively shortening the downtime of equipment and reducing maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1 The present invention is a flowchart of an embodiment of a method for detecting and diagnosing abnormalities of a central air-conditioning chiller.

[0018] Figure 2 The present invention is a flowchart of another embodiment of the method for detecting and diagnosing abnormalities of a central air-conditioning chiller.

[0019] Figure 3 It is a schematic diagram of the abnormality detection and diagnosis system of the central air-conditioning chiller unit of the present invention. DETAILED DESCRIPTION

[0020] In the following detailed description of the central air conditioning chiller abnormality detection and diagnosis method, various embodiments of the present disclosure will be described more fully. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0021] For example, in the traditional operation and maintenance mode of central air-conditioning chillers, manual monitoring and detection have always dominated. However, this mode is not only inefficient, but also limited by professional knowledge and experience. It is often difficult to make quick and accurate judgments in the face of complex abnormal patterns and deep-seated system problems. This not only pushes up maintenance costs, but also prolongs the downtime of equipment, which in turn affects the overall operating efficiency and production stability.

[0022] In the field of anomaly detection, the current popular approach is to use deep learning-based technologies, such as autoencoders, convolutional neural networks, and long short-term memory networks. These technologies can learn and identify complex patterns and abnormal behaviors in data, and have shown certain results in the fields of image and speech recognition. However, in actual industrial applications, these methods face many challenges. On the one hand, they usually require a large amount of labeled data for training, and in industrial environments, obtaining these labeled abnormal data is time-consuming and costly. On the other hand, the interpretability of deep learning models is relatively poor. When the model predicts an anomaly, it is difficult to clearly explain the reason for the prediction. This is undoubtedly a major flaw for industrial systems that require high reliability and interpretability.

[0023] Another common anomaly detection method is to analyze the operation logs of important air-conditioning chillers. However, the current mainstream methods of log analysis still rely mainly on regular expression matching, keyword search, and simple statistical analysis. The limitations of these methods are obvious. They cannot capture complex patterns and abnormal behaviors in log data. At the same time, these methods also require a lot of manual participation, including the screening, classification, and interpretation of logs, which is not only inefficient but also easily interfered by human factors.

[0024] In response to the above problems, the present embodiment provides a central air-conditioning chiller abnormality detection and diagnosis method, which can fully mine time series data information and log data information at the same time. The former realizes early identification and early warning of abnormal conditions, thereby reducing potential equipment abnormalities and downtime; the latter helps operation and maintenance personnel quickly understand the causes of abnormalities and provide targeted repair suggestions. Combining the two can greatly improve the operation and maintenance efficiency of the chiller, and can also significantly improve the diagnostic accuracy and reduce the energy consumption of the equipment, thereby improving the operation and maintenance efficiency and diagnostic accuracy of the chiller.

[0025] 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] See also Figure 1 The figure is a flow chart of a method for detecting and diagnosing abnormality of a central air-conditioning chiller in a specific embodiment, the method comprising the following steps: S1. Collect the time series data and log data of the air conditioning unit operation and pre-process them; It should be noted that the time series data of the air conditioning unit operation is preprocessed to ensure that it matches the time series detection model, and the log data of the air conditioning unit is preprocessed to ensure that it matches the large language model; S2. Use the time series data of the air conditioning unit operation to fine-tune the time series detection model, and use the log data of the air conditioning unit operation to fine-tune the large language model; It should be noted that by fine-tuning the training of the time series detection model, the multimodal data characteristics of the normal operation of the air-conditioning unit equipment will be learned, and by fine-tuning the training of the large language model, the abnormal pattern recognition ability of the air-conditioning unit will be improved; S3. The real-time operation data of the air-conditioning unit is pre-processed and input into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the air-conditioning unit operation during the abnormal period is obtained and input into the large language model for anomaly diagnosis; It should be noted that by integrating the time series prediction model with the large language model to detect and diagnose anomalies in the air-conditioning unit, the operation and maintenance efficiency of the air-conditioning unit can be greatly improved, and the diagnosis accuracy can be significantly improved, and the energy consumption of the equipment can be reduced.

[0027] The abnormality detection and diagnosis method for the central air-conditioning chiller of this embodiment can monitor the operating status of the chiller in real time and automatically trigger the diagnosis process when an abnormality is detected, thereby realizing the automation and intelligence of operation and maintenance management, which not only improves the abnormality response speed but also reduces the maintenance cost, and is of great significance to the stable operation of the industrial chiller.

[0028] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another central air-conditioning chiller abnormality detection and diagnosis method is provided, such as Figure 2 As shown, the method comprises the following steps: S1. Collect the time series data and log data of the air conditioning unit operation and pre-process them; the specific steps of step S1 are as follows: S11. Collecting the time series data of the operation of the air conditioning unit, the time series data of the operation of the air conditioning unit includes the inlet and outlet temperature of the chilled water, the inlet and outlet temperature of the cooling water, the inlet and outlet temperature of the cooling water loop heat exchanger, the inlet and outlet temperature of the chilled water loop heat exchanger and the compressor power; It should be noted that the inlet and outlet temperatures of chilled water, the inlet and outlet temperatures of cooling water, the inlet and outlet temperatures of the cooling water loop heat exchanger, and the near outlet temperature of the chilled water loop heat exchanger are measured using thermistors respectively; S12. Clean the time series data of the operation of the air-conditioning unit by removing random noise through filtering algorithm, fill in missing data through linear interpolation, unify the data format and measurement unit, and complete standardization; It should be noted that filtering provides the signal-to-noise ratio of data, filling in missing data ensures data integrity, and unifying data formats and measurement units ensures that data from different sensor sources can be compared and processed under the same standards; S13. Collecting log data of the air conditioning unit, the log data of the air conditioning unit includes equipment operation status, equipment alarm information, maintenance and diagnostic information and historical abnormal information; It should be noted that the equipment operation status data can be the on / off status change of the chiller, the status information of different working modes such as cooling, standby, and defrosting; the equipment alarm information can be high / low pressure alarm; overheating or overcooling alarm; current too high or too low alarm; sensor abnormality alarm; maintenance and diagnosis information can be equipment maintenance records, including cleaning, lubrication, inspection and other information; historical abnormality information can be the time of occurrence, abnormality type and processing result; S14. Normalize the log data of the collection unit, truncate the log text according to the maximum input length of the large language model, and convert the log text into a format supported by the large language model input; Exemplarily, the log text is truncated according to the maximum input length of the Qwen large language model, and the log text is converted into an input format supported by the Qwen large language model; S2. Use the time series data of the air conditioning unit operation to fine-tune the time series detection model, and use the log data of the air conditioning unit operation to fine-tune the large language model; the specific steps of step S2 are as follows: S21. fine-tune the structure of the Moirai time series detection model according to the time series data of the air conditioning unit; S22. Use the time series data of the air conditioning unit to train and optimize the fine-tuned Moirai time series detection model to obtain an optimized Moirai time series detection model; It should be noted that the normal time series data set is used to fine-tune and train the structure of the Moirai time series detection model, so that the Moirai time series detection model can learn the time series regularity information of the central air-conditioning unit under normal operation mode; S23. Fine-tune the back-end layers of the Qwen large language model, add an adaptive layer to the feature extraction part, add an external knowledge base to the retrieval function of the RAG architecture, and limit the output abnormality report format to match the abnormality diagnosis of the air conditioning unit; It should be noted that the fine-tuning of the Qwen large language model is to prepare and process abnormal data. By adding the RAG mechanism, it can more accurately analyze the air conditioning unit log data and generate answers that better meet the requirements of abnormal diagnosis; S3. The real-time operation data of the air-conditioning unit is preprocessed and input into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the air-conditioning unit operation during the abnormal period is obtained and input into the large language model for anomaly diagnosis; the specific steps of step S3 are as follows: S31. Collect real-time operation data of air-conditioning units, and input them into the optimized Moirai time series detection model after preprocessing; S32. The Moirai time series detection model analyzes various sensor data in the time dimension and feature dimension through the attention mechanism, compares the learned normal operation characteristics of the air-conditioning unit with the input data, and obtains the prediction results of the input data; It should be noted that the pre-processed data is input into the time series Transformer architecture of the Moirai time series detection model, and the attention mechanism is used to analyze the multimodal data in the time dimension and feature dimension; the Moirai time series detection model will compare the input data based on the normal operation characteristics of the equipment learned during the pre-training and fine-tuning process; through the arbitrary variable attention mechanism, the model can adapt to the different dimensions of the chiller operation data and realize the comprehensive analysis of various sensor data; this process allows the Moirai time series detection model to maintain accurate monitoring of each dimension when processing multiple sensor inputs; S33. Analyze the prediction results, regard the probability that the data falls outside the normal range as an abnormal score, and determine whether the abnormal score exceeds the warning threshold; If not, return to step S31; If yes, go to step S34; It should be noted that the Moirai time series detection model generates the prediction results of the input data, using the mixed distribution probability of the model output, and regards the probability that the data falls outside the normal distribution as an abnormal score to determine whether the current operating state deviates from the normal mode. is the input data collected in real time. is the length of the time series, Indicates at time Sensor data characteristics represents the distribution parameters learned by the model during the fine-tuning phase. The calculation formula for the anomaly score is as follows:

[0029] S34 triggers real-time warning, determines abnormal period, and extracts log data and time series data of air-conditioning unit operation during abnormal period; It should be noted that if the anomaly score exceeds the warning threshold generated by the model, the system will trigger a real-time warning and automatically extract relevant log data for possible abnormal periods; S35. After preprocessing the extracted log data of the air conditioning unit during the abnormal period, it is input into the fine-tuned Qwen large language model together with the abnormal score; It should be noted that all test results, including records of normal and abnormal conditions, are stored in the database. These historical data can be used for further analysis and model optimization to provide support for long-term chiller operation status monitoring and maintenance strategy improvement; S36. The fine-tuned Qwen large language model uses the model's natural language processing capabilities to analyze log and sensor data during the abnormal period, and performs feature analysis on multimodal data. By comparing context and abnormal features, and combining historical data, root cause analysis is performed. At the same time, the retrieval function in RAG is activated to retrieve abnormal records and diagnostic suggestions similar to the input data from the knowledge base; It should be noted that the multimodal data may be temperature, pressure, and alarm data; S37. The fine-tuned Qwen large language model generates a structured anomaly diagnosis report through multimodal fusion in accordance with the agreed format and combined with anomaly scores, historical similar cases, and contextual information; It should be noted that during the model reasoning process, the Qwen large language model identifies abnormal patterns through the comparison of context and abnormal features, performs root cause analysis in combination with historical data, and simultaneously activates the retrieval function in RAG to retrieve abnormal records and diagnostic suggestions similar to the input data from the knowledge base; For example, let the feature vector of the current abnormal data be , the feature vector of the 𝑖th historical anomaly case in the knowledge base is , the similarity calculation formula between the two is:

[0030] By calculating the similarity between current anomaly data and historical anomaly cases, the system can identify the most similar anomaly patterns and add their root causes and solutions to the report:

[0031] This ensures that the model can use historical data to improve the accuracy of abnormal root cause analysis during the diagnosis process; S38. The generated abnormal diagnosis report is stored in the knowledge base to build the air conditioning unit failure mode library, and the input and output of the Qwen large language model are further fine-tuned according to the abnormal diagnosis report; Add structured report template prompts to the input of the Qwen large language model, such as the report structure: exception description, root cause analysis, and repair suggestions, to ensure that the output of the Qwen model conforms to the specified report format; provide sample reports and paragraph sequence guidance for the Qwen large language model so that the output report has a clear logical hierarchy and standardized structure, including exception description, cause analysis, and repair measures; based on the reasoning results, the Qwen large language model generates a detailed exception report, which covers the description of the abnormal event, the location of the root cause device, and the scope of impact. When generating an exception report, the Qwen large language model combines the anomaly score, historical similar cases, and contextual information to generate a structured report through multimodal fusion. For example, suppose the time series feature and log features The fusion vector is F, and It is a hyperparameter for adjusting weights, which is used to balance the contribution of time series and log features. The calculation formula of F is:

[0032] Based on the fusion vector F, the Qwen large language model generates structured exception report content, including exception description, root cause analysis, and repair suggestions; In addition, a feedback loop mechanism is established to allow operation and maintenance personnel to provide feedback on the generated reports to optimize the report generation quality of the Qwen large language model; the generated diagnostic reports are stored in the knowledge base, and the fault mode library of the chiller is gradually built and enriched. In future diagnoses, when encountering similar anomalies, the Qwen large language model can directly retrieve relevant records from the knowledge base to improve diagnostic efficiency; based on feedback, the input format and prompts are gradually adjusted to optimize the performance of the Qwen large language model in generating anomaly reports that meet actual needs. At the same time, the feedback will also be used to update the knowledge base to ensure that the RAG module can call the latest and more adaptable anomaly handling cases in the future reasoning process.

[0033] In an embodiment of the present invention, based on step S21, step S22 and step S23, a possible embodiment is given below to illustrate its specific implementation scheme in a non-limiting manner.

[0034] The specific steps of step S21 are as follows: S211. Divide the time series data of the air conditioning unit operation into a training set and a validation set according to a set ratio; It should be noted that after the data preprocessing is completed, the processed normal operation time series data of the air-conditioning unit is divided into a training set and a validation set, and the division ratio is usually 80%:20% to ensure that the model can learn sufficient samples of normal equipment behavior during the training process; S212. Select N patch sizes, use multi-patch size input projection to segment the time series data of the training set and the validation set into time windows according to each patch size, and obtain N segmentation results; It should be noted that the input time series data is projected with multiple patch sizes, which divides the time series data into time windows of specific lengths according to the set multiple patch sizes, so that the model can capture features at different time scales; Assume the input time series data of the chiller is a matrix ,in Indicates at time The collected feature vector contains the readings of multiple sensors. The projection formula for multiple patch size inputs is as follows:

[0035] Where 𝑘 represents the size of the patch. By using different 𝑘 values, the data can be segmented at different time scales, for example, small patches are used to capture short-term fluctuations, and large patches are used to capture long-term trends; Two patch sizes are chosen exemplarily: k 1=3 and k 2=10 for k 1=3, the generated time window is: [ x 1, x 4, x 7,…, x 997 ] (Total 333 windows) for k 2=10, the resulting time window is: [ x 1, x 11 , x 21 ,…, x 991 ] (99 windows in total) S213. Map each window of each segmentation result to a high-dimensional feature space using a patch embedding method to obtain a processed test set and a validation set; It should be noted that patch embedding is used to map these windows into a high-dimensional feature space so that the model can effectively capture the temporal patterns and correlations of device operation data in subsequent layers; S214. Input the test set or validation set after the patch embedding into the Moirai time series detection model, use the attention mechanism to process the data layer by layer, and in the process of data processing, adjust the layers of the back end of the Moirai time series detection model within a set range or add an adaptive layer to match the air conditioning unit anomaly detection task; S215. Introducing a multi-layer arbitrary variable attention mechanism in the Moirai time series detection model, and processing different amounts and types of sensor data; It should be noted that the embedded time series data is input into the core time series Transformer architecture of the Moirai time series detection model, and its attention mechanism is used to process the data layer by layer. In the fine-tuning stage, some pre-trained parameters are locked, and only the last few layers or the added adaptive layers are fine-tuned to adapt to the specific data distribution of the chiller. A multi-layer arbitrary variable attention mechanism is used, which allows the model to maintain the independence between dimensions when processing different numbers and types of sensor data, thereby improving the adaptability of the Moirai time series detection model to the multimodal data of the chiller. The specific steps of step S22 are as follows: S221. Taking the negative log-likelihood function of the predicted distribution of the normal operation data of the air-conditioning unit as the objective function, the prediction error of the training sample is characterized; It should be noted that during the fine-tuning process, a specific objective function is used to optimize the parameters of the Moirai time series detection model; this objective function is usually the negative log-likelihood, which ensures that the Moirai time series detection model is sensitive to abnormal patterns while predicting normal data distribution; Let the output of the model be the predictive distribution of normal data ,in Indicates the normal operating data of the chiller. is the number of training samples, and the objective function is:

[0036] S222. Using cross-validation, the Moirai time series detection model is iteratively optimized on the training set and the validation set, and the training is stopped when the objective function is minimized or the number of iterations is completed to obtain the optimized Moirai time series detection model; It should be noted that the cross-validation method is used to iteratively optimize the training set and the validation set to reduce overfitting and make the Moirai time series detection model more generalizable; The specific steps of step S23 are as follows: S231. Lock the front-end layer of the Qwen large language model, and adjust the back-end layer within a set range or add an adaptive layer; It should be noted that in order to retain the basic language understanding ability obtained by the Qwen large language model during the pre-training process, the parameters of the first few layers will be locked during fine-tuning, and only the latter few layers will be fine-tuned or adaptive layers will be added, so that the Qwen large language model can focus on the abnormal characteristics of the chiller based on the original general knowledge; S232. Adding an adaptive layer to the feature extraction part of the Qwen large language model to adapt to the abnormal data distribution characteristics of the air-conditioning unit; It should be noted that task-specific adaptive layers such as small-scale fully connected layers or convolutional layers will be added to the feature extraction module of the Qwen large language model. These layers are specifically used to adapt to the distribution characteristics of abnormal data of the chiller so that the Qwen large language model can better capture the subtle differences in the abnormal state of the chiller; the adaptive layer makes the Qwen large language model more sensitive in distinguishing normal and abnormal states by learning the specific abnormal patterns of the chiller; S233. Use the retrieval function in the RAG architecture to build an external knowledge base containing air conditioning unit abnormality diagnosis knowledge, and introduce air conditioning chiller failure modes, processing suggestions and historical cases; S234. Simulate the input of the air conditioning unit abnormal diagnosis related problems into the Qwen language model, so that the Qwen language model calls the relevant historical cases and diagnostic information in the retrieval function of the RAG architecture, and adjusts the Qwen language model according to the retrieval results; It should be noted that the use of an external knowledge base can improve the Qwen large language model's ability to recognize abnormal patterns of chillers and enrich its diagnostic output; S235. Modify the output layer of the Qwen large language model, add formatted content to the output abnormal diagnosis report, and match the annotation format of the abnormal diagnosis of the air-conditioning unit; the formatted content includes the abnormal report header, abnormal description, root cause analysis and modification suggestions; It should be noted that the output layer of the model is modified to make it more suitable for generating structured abnormality report content, ensuring that the Qwen large language model can cover the key abnormality information of the chiller when generating the report and can comply with the standardized structure of the diagnostic report.

[0037] The present invention can fully mine time series data information and log data information at the same time, realize early identification and early warning of abnormal conditions, thereby reducing potential equipment abnormalities and downtime, helping operation and maintenance personnel to quickly understand the causes of abnormalities, and provide targeted repair suggestions. Combining the two can greatly improve the operation and maintenance efficiency of the chiller, and can also significantly improve the diagnostic accuracy and reduce the energy consumption of the equipment.

[0038] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0039] like Figure 3As shown, the following is an embodiment of the central air-conditioning chiller abnormality detection and diagnosis system provided by the embodiment of the present disclosure. The system and the central air-conditioning chiller abnormality detection and diagnosis method of the above-mentioned embodiments belong to the same inventive concept. For details not fully described in the embodiment of the central air-conditioning chiller abnormality detection and diagnosis system, please refer to the embodiment of the above-mentioned central air-conditioning chiller abnormality detection and diagnosis method.

[0040] The system includes: A data acquisition unit, used to collect the time series data and log data of the operation of the air-conditioning unit and perform pre-processing; A model fine-tuning unit, used to fine-tune the time series detection model using the time series data of the air-conditioning unit operation, and to fine-tune the large language model using the log data of the air-conditioning unit operation; The anomaly detection and diagnosis unit is used to pre-process the real-time operation data of the air-conditioning unit and input it into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the operation of the air-conditioning unit during the abnormal period is obtained and input into the large language model for anomaly diagnosis.

[0041] The central air-conditioning chiller abnormality detection and diagnosis system of this embodiment can monitor the operating status of the chiller in real time and automatically trigger the diagnosis process when an abnormality is detected, thereby realizing the automation and intelligence of operation and maintenance management, which not only improves the abnormality response speed but also reduces the maintenance cost, and is of great significance to the stable operation of the industrial chiller.

[0042] The central air-conditioning chiller abnormality detection and diagnosis method provided in the embodiment of the present application can be applied to electronic devices. It can be understood by those skilled in the art that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or less components than shown, or combine certain components, or arrange different components. In an embodiment of the present invention, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0043] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, buttons, a camera, a display, and a SIM card interface, etc.

[0044] It is to be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0045] The processor may include one or more processing units, for example, the processor may include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units may be independent devices or integrated into one or more processors.

[0046] The processor can be the nerve center and command center of the electronic device. The controller can generate an operation control signal according to the instruction operation code and timing signal to complete the control of fetching and executing instructions.

[0047] A memory may also be provided in the processor for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory may store instructions or data that the processor has just used or is cyclically used. If the processor needs to use the instruction or data again, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor, and thus improves system efficiency.

[0048] The above-mentioned electronic device implements the method for abnormal detection and diagnosis of central air-conditioning chillers of the present application, which collects the time series data and log data of the operation of the air-conditioning unit and performs preprocessing; uses the time series data of the operation of the air-conditioning unit to fine-tune the time series detection model, and uses the log data of the operation of the air-conditioning unit to fine-tune the large language model; pre-processes the real-time operation data of the air-conditioning unit and inputs it into the time series prediction model for abnormality detection. When an abnormality is detected, the log data of the operation of the air-conditioning unit within the abnormal period is obtained and input into the large language model for abnormality diagnosis. The technical solution achieves real-time monitoring of the operating status of the chiller and automatically triggers the diagnosis process when an abnormality is detected, thereby realizing the automation and intelligence of operation and maintenance management, which not only improves the abnormal response speed, but also reduces the beneficial effect of maintenance costs.

[0049] The storage medium provided in the present application stores a program product that can implement a method for detecting and diagnosing abnormalities of a central air-conditioning chiller.

[0050] The abnormality detection and diagnosis method of the central air-conditioning chiller unit includes: collecting the time series data and log data of the air-conditioning unit operation and preprocessing them; using the time series data of the air-conditioning unit operation to fine-tune the time series detection model, and using the log data of the air-conditioning unit operation to fine-tune the large language model; preprocessing the real-time operation data of the air-conditioning unit and inputting it into the time series prediction model for abnormality detection, and when an abnormality is detected, obtaining the log data of the air-conditioning unit operation during the abnormal period and inputting it into the large language model for abnormality diagnosis.

[0051] In some possible embodiments, the central air-conditioning chiller abnormality detection and diagnosis method disclosed herein can be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above "Exemplary Method" section of this specification according to various exemplary embodiments of the present disclosure.

[0052] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0053] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting and diagnosing abnormality of a central air-conditioning chiller, characterized in that: The steps include: S1. Collect the time series data and log data of the air conditioning unit operation and pre-process them; S2. Use the time series data of the air conditioning unit operation to fine-tune the time series detection model, and use the log data of the air conditioning unit operation to fine-tune the large language model; S3. The real-time operation data of the air-conditioning unit is pre-processed and then input into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the operation of the air-conditioning unit during the abnormal period is obtained and input into the large language model for anomaly diagnosis.

2. The method for detecting and diagnosing abnormality of a central air-conditioning chiller according to claim 1, characterized in that: The specific steps of step S1 are as follows: S11. Collecting the time series data of the operation of the air conditioning unit, the time series data of the operation of the air conditioning unit includes the inlet and outlet temperature of the chilled water, the inlet and outlet temperature of the cooling water, the inlet and outlet temperature of the cooling water loop heat exchanger, the inlet and outlet temperature of the chilled water loop heat exchanger and the compressor power; S12. Clean the time series data of the operation of the air-conditioning unit by removing random noise through filtering algorithm, fill in missing data through linear interpolation, unify the data format and measurement unit, and complete standardization; S13. Collecting log data of the air conditioning unit, the log data of the air conditioning unit includes equipment operation status, equipment alarm information, maintenance and diagnostic information and historical abnormal information; S14. Perform text normalization on the log data of the collection group, truncate the log text according to the maximum input length of the large language model, and convert the log text into a format supported by the large language model input.

3. The central air-conditioning chiller abnormality detection and diagnosis method according to claim 2, characterized in that: The specific steps of step S2 are as follows: S21. fine-tune the structure of the Moirai time series detection model according to the time series data of the air conditioning unit; S22. Use the time series data of the air conditioning unit to train and optimize the fine-tuned Moirai time series detection model to obtain an optimized Moirai time series detection model; S23. Fine-tune the backend layers of the Qwen large language model, add an adaptive layer to the feature extraction part, add an external knowledge base to the retrieval function of the RAG architecture, and limit the output exception report format to match the abnormal diagnosis of the air-conditioning unit.

4. The central air-conditioning chiller abnormality detection and diagnosis method according to claim 3, characterized in that: The specific steps of step S21 are as follows: S211. Divide the time series data of the air conditioning unit operation into a training set and a validation set according to a set ratio; S212. Select N patch sizes, use multi-patch size input projection to segment the time series data of the training set and the validation set into time windows according to each patch size, and obtain N segmentation results; S213. Map each window of each segmentation result to a high-dimensional feature space using a patch embedding method to obtain a processed test set and a validation set; S214. Input the test set or validation set after the patch embedding into the Moirai time series detection model, use the attention mechanism to process the data layer by layer, and in the process of data processing, adjust the layers of the back end of the Moirai time series detection model within a set range or add an adaptive layer to match the air conditioning unit anomaly detection task; S215. Introduce a multi-layer arbitrary variable attention mechanism in the Moirai time series detection model, and process sensor data of different quantities and types.

5. The central air-conditioning chiller abnormality detection and diagnosis method according to claim 3, characterized in that: The specific steps of step S22 are as follows: S221. Taking the negative log-likelihood function of the predicted distribution of the normal operation data of the air-conditioning unit as the objective function, the prediction error of the training sample is characterized; S222. The Moirai time series detection model is iteratively optimized on the training set and the validation set by using a cross-validation method, and the training is stopped when the objective function is minimized or the number of iterations is completed, thereby obtaining the optimized Moirai time series detection model.

6. The central air-conditioning chiller abnormality detection and diagnosis method according to claim 3, characterized in that: The specific steps of step S23 are as follows: S231. Lock the front-end layer of the Qwen large language model, and adjust the back-end layer within a set range or add an adaptive layer; S232. Adding an adaptive layer to the feature extraction part of the Qwen large language model to adapt to the abnormal data distribution characteristics of the air-conditioning unit; S233. Use the retrieval function in the RAG architecture to build an external knowledge base containing air conditioning unit abnormality diagnosis knowledge, and introduce air conditioning chiller failure modes, processing suggestions and historical cases; S234. Simulate the input of the air conditioning unit abnormal diagnosis related problems into the Qwen language model, so that the Qwen language model calls the relevant historical cases and diagnostic information in the retrieval function of the RAG architecture, and adjusts the Qwen language model according to the retrieval results; S235. Modify the output layer of the Qwen large language model, add formatted content to the output abnormal diagnosis report, and match the annotation format of the abnormal diagnosis of the air-conditioning unit; the formatted content includes the abnormal report header, abnormal description, root cause analysis and modification suggestions.

7. The central air-conditioning chiller abnormality detection and diagnosis method according to claim 3, characterized in that: The specific steps of step S3 are as follows: S31. Collect real-time operation data of air-conditioning units, and input them into the optimized Moirai time series detection model after preprocessing; S32. The Moirai time series detection model analyzes various sensor data in the time dimension and feature dimension through the attention mechanism, compares the learned normal operation characteristics of the air-conditioning unit with the input data, and obtains the prediction results of the input data; S33. Analyze the prediction results, regard the probability that the data falls outside the normal range as an abnormal score, and determine whether the abnormal score exceeds the warning threshold; If not, return to step S31; If yes, go to step S34; S34 triggers real-time warning, determines abnormal period, and extracts log data and time series data of air-conditioning unit operation during abnormal period; S35. After preprocessing the extracted log data of the air conditioning unit during the abnormal period, it is input into the fine-tuned Qwen large language model together with the abnormal score; S36. The fine-tuned Qwen large language model uses the model's natural language processing capabilities to analyze log and sensor data during the abnormal period, and performs feature analysis on multimodal data. By comparing context and abnormal features, and combining historical data, root cause analysis is performed. At the same time, the retrieval function in RAG is activated to retrieve abnormal records and diagnostic suggestions similar to the input data from the knowledge base; S37. The fine-tuned Qwen large language model generates a structured anomaly diagnosis report through multimodal fusion in accordance with the agreed format and combined with anomaly scores, historical similar cases, and contextual information; S38. Store the generated abnormal diagnosis report in the knowledge base to build the air-conditioning unit failure mode library, and further fine-tune the input and output of the Qwen large language model based on the abnormal diagnosis report.

8. A central air-conditioning chiller abnormality detection and diagnosis system, characterized in that: include: A data acquisition unit, used to collect the time series data and log data of the operation of the air-conditioning unit and perform pre-processing; A model fine-tuning unit, used to fine-tune the time series detection model using the time series data of the air-conditioning unit operation, and to fine-tune the large language model using the log data of the air-conditioning unit operation; The anomaly detection and diagnosis unit is used to pre-process the real-time operation data of the air-conditioning unit and input it into the time series prediction model for anomaly detection. When an anomaly is detected, the log data of the operation of the air-conditioning unit during the abnormal period is obtained and input into the large language model for anomaly diagnosis.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the central air-conditioning chiller abnormality detection and diagnosis method as described in any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of the method for detecting and diagnosing abnormalities of a central air-conditioning chiller unit as claimed in any one of claims 1 to 7 are implemented.

Images