Method, device and system for evaluating health degree of comprehensive digital intelligent management platform of refrigerating machine room, medium and program product

Abstract

The invention provides a health degree assessment method, device and system for a comprehensive digital intelligent management platform of a refrigerating machine room, a medium and a program product, real-time monitoring on the health degree can be realized through a sparse Bayesian learning algorithm and a Bayesian optimal classification algorithm and a machine room energy efficiency supervision platform, and corresponding adjustment is quickly made to maintain an optimal operation state. The sparse Bayesian learning algorithm not only can select features, but also can dynamically adjust weights, so as to ensure that the features can flexibly adapt to constantly changing actual demands. By comprehensively evaluating multiple dimensions such as system performance, data processing capability, model prediction accuracy and user experience, the platform provides comprehensive health condition analysis, and comprehensive mastering of the operation state of the platform is ensured. And according to the predicted health degree, the operation state of the platform is further evaluated, and early warning and maintenance are made in time.

Description

Technical Field

[0001] The present application relates to the field of energy efficiency supervision technology, and in particular to a health assessment method, device, system, medium and program product for a comprehensive digital management platform for refrigeration rooms. Background Art

[0002] With the popularization of intelligent management of buildings and industrial computer rooms, energy efficiency monitoring platforms are widely used to monitor and optimize the energy efficiency of computer room equipment. However, as the platform operation time increases and the amount of data increases, the operation health of the platform itself may gradually decline, affecting its performance and stability. Although existing technologies can monitor the energy efficiency of equipment in real time, the means of evaluating the health of the platform are still insufficient.

[0003] Current health assessment methods focus on the following aspects: system resource monitoring, monitoring system performance indicators such as CPU, memory, and network bandwidth, and evaluating the load of the platform; regular maintenance and manual adjustments, relying on regular maintenance and manual adjustments by operation and maintenance personnel, lack of real-time adaptability, and prone to lag problems.

[0004] However, the above-mentioned existing technologies still have obvious problems and shortcomings: the evaluation dimension is single, and the existing evaluation methods mainly focus on the system hardware performance, ignoring key dimensions such as data processing capabilities and model prediction accuracy; lack of real-time adjustment capabilities, health assessment cannot dynamically reflect the real-time status of the platform, which may lead to delayed processing of problems. Summary of the invention

[0005] In view of the shortcomings of the prior art mentioned above, the purpose of this application is to provide a health assessment method, device, system, medium and program product for a comprehensive digital management platform for refrigeration rooms, so as to solve the technical problems of the existing health assessment having a single assessment dimension and lacking real-time adjustment capability.

[0006] To achieve the above-mentioned purpose and other related purposes, the first aspect of the present application provides a health assessment method for a comprehensive digital management platform of a refrigeration room, including: collecting a number of key performance data of the computer room energy efficiency supervision platform to be evaluated based on multiple preset dimensions, and preprocessing the key performance data; the multiple preset dimensions include: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension; the preprocessed key performance data is input into a sparse Bayesian learning algorithm for feature selection and dynamically adjusts the weights of the features of each dimension, and calculates the health of the computer room energy efficiency supervision platform in real time; the calculated health data is input into a Bayesian optimal classification algorithm for analysis to obtain a classification result of the operating status of the computer room energy efficiency supervision platform, and an early warning is triggered based on the classification result.

[0007] In some embodiments of the first aspect of the present application, the preprocessed key performance data is input into a sparse Bayesian learning algorithm for feature selection and dynamic adjustment of the weights of the features of each dimension, and the health of the computer room energy efficiency monitoring platform is calculated in real time, including: inputting the preprocessed and normalized key performance data into a sparse Bayesian learning algorithm for sparse feature selection to select several features that have the greatest impact on health and dynamically adjust the weights; according to the output results of the sparse Bayesian learning algorithm, the health of the computer room energy efficiency monitoring platform is calculated in real time through a pre-built health prediction model; the cross-validation method is used to evaluate the generalization ability of the model, and the hyperparameters of the model are adjusted to optimize the model performance.

[0008] In some embodiments of the first aspect of the present application, the preprocessed and normalized key performance data is input into a sparse Bayesian learning algorithm for sparse feature selection to select several features that have the greatest impact on health and dynamically adjust the weights, which includes: initializing the parameters of the sparse Bayesian model; using L1 regularization, adjusting the parameters of the sparse Bayesian model through an optimization algorithm so that the weights of unimportant features tend to zero, so as to select several features that have the greatest impact on health; evaluating the performance of the sparse Bayesian model, and dynamically adjusting the feature weights according to the performance of the sparse Bayesian model in actual applications; the dynamic adjustment method includes reallocating the weights of the selected features and / or reselecting the features.

[0009] In some embodiments of the first aspect of the present application, the health prediction model includes: ;in, Indicates The weight of the indicator, The energy efficiency monitoring platform of the computer room is The actual performance on key performance indicators, The energy efficiency monitoring platform of the computer room is Ideal performance on key performance indicators; Indicates the degree of closeness between actual performance and ideal performance; is an adjustment factor used to fine-tune the health score.

[0010] In some embodiments of the first aspect of the present application, the calculated health data is input into a Bayesian optimal classification algorithm for analysis to obtain a classification result of the operating status of the computer room energy efficiency monitoring platform, including the following steps: collecting historical health data and corresponding operating status data of the computer room energy efficiency monitoring platform, and dividing them into a training set and a validation set according to a certain ratio; inputting the training set into a Bayesian optimal classification model for model training; the Bayesian optimal classification model learns how to calculate the posterior probability based on the input health history data and predict the most likely platform operating status; inputting the validation set into the trained Bayesian optimal classification model to evaluate the performance of the model.

[0011] In some embodiments of the first aspect of the present application, the Bayesian optimal classification model selects the category label that maximizes the posterior probability as the prediction result according to the classification goal of minimizing the classification error rate.

[0012] To achieve the above-mentioned purpose and other related purposes, the second aspect of the present application provides a health assessment system for a comprehensive digital management platform of a refrigeration room, including: a data acquisition module, used to collect a number of key performance data of the computer room energy efficiency supervision platform to be evaluated based on multiple preset dimensions, and pre-process the key performance data; the multiple preset dimensions include: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension; a health calculation module, used to input the preprocessed key performance data into a sparse Bayesian learning algorithm for feature selection and dynamically adjust the weights of the features of each dimension, and calculate the health of the computer room energy efficiency supervision platform in real time; a status classification module, used to input the calculated health data into a Bayesian optimal classification algorithm for analysis, so as to obtain a classification result of the operating status of the computer room energy efficiency supervision platform, and trigger an early warning based on the classification result.

[0013] To achieve the above-mentioned purpose and other related purposes, the third aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the health assessment method of the comprehensive digital management platform of the refrigeration room.

[0014] To achieve the above-mentioned purpose and other related purposes, the fourth aspect of the present application provides a computer program product, which includes a computer program code. When the computer program code runs on a computer, the computer implements the health assessment method of the comprehensive digital management platform of the refrigeration room.

[0015] To achieve the above-mentioned purpose and other related purposes, the fifth aspect of the present application provides a computer device, including a memory, a processor and a computer program stored in the memory; the processor executes the computer program to implement the health assessment method of the comprehensive digital management platform of the refrigeration room.

[0016] As described above, the health assessment method, device, system, medium and program product of the comprehensive digital management platform for refrigeration rooms of the present application have the following beneficial effects.

[0017] (1) Through the sparse Bayesian learning algorithm and the Bayesian optimal classification algorithm, the computer room energy efficiency monitoring platform can achieve real-time monitoring of health and make corresponding adjustments quickly to maintain the optimal operating state.

[0018] (2) Sparse Bayesian learning algorithms can not only select features but also dynamically adjust weights, ensuring that they can flexibly adapt to changing actual needs.

[0019] (3) By comprehensively evaluating multiple dimensions such as system performance, data processing capabilities, model prediction accuracy, and user experience, the platform provides a comprehensive health status analysis to ensure a comprehensive understanding of the platform's operating status. Based on the predicted health, the platform's operating status is further evaluated to provide timely warnings and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Shown is a flow chart of a method for evaluating the health of a comprehensive digital management platform for a refrigeration room based on sparse Bayesian and Bayesian optimal classification in one embodiment of the present application.

[0021] Figure 2 Shown is a schematic diagram of key performance data in different dimensions in one embodiment of the present application.

[0022] Figure 3 Shown is a schematic diagram of the execution process of step S2 in an embodiment of the present application.

[0023] Figure 4 Shown is a schematic diagram of the process of performing sparse feature selection by the SBL algorithm in one embodiment of the present application.

[0024] Figure 5 Shown is a structural schematic diagram of a health assessment system for a comprehensive digital management platform for a refrigeration room based on sparse Bayesian and Bayesian optimal classification in one embodiment of the present application.

[0025] Figure 6 Shown is a schematic diagram of the structure of a computer device in one embodiment of the present application. DETAILED DESCRIPTION

[0026] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0027] Figure 1 A flowchart of a health assessment method for a refrigeration room integrated digital management platform based on sparse Bayesian and Bayesian optimal classification in an embodiment of the present invention is shown. This embodiment aims to propose a comprehensive and efficient health assessment method for a refrigeration room integrated digital management platform, which dynamically assesses the operating status of the supervision platform through multi-dimensional data fusion and real-time monitoring, and combines advanced algorithms to achieve accurate health prediction and adjustment.

[0028] Specifically, the computer room energy efficiency monitoring method in this embodiment mainly includes the following steps.

[0029] Step S1: collecting a number of key performance data of the computer room energy efficiency monitoring platform to be evaluated based on a plurality of preset dimensions, and preprocessing the key performance data.

[0030] like Figure 2 As shown, the multiple preset dimensions in the embodiments of the present application include but are not limited to: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension. The system performance dimension is used to monitor the usage of hardware resources such as CPU, memory, network bandwidth, etc. The data processing dimension is used to evaluate the speed, delay and error rate of data processing by the platform. The model prediction dimension evaluates the energy efficiency prediction ability of the platform through the accuracy and stability of model prediction. The user interaction dimension is used to evaluate the user experience by monitoring the response time and operation smoothness. By collecting key performance data under these dimensions in real time, the operation status of the computer room can be comprehensively monitored and managed, potential problems can be discovered and solved in time, and the stability of the system and user satisfaction can be improved.

[0031] Key performance data collected from the dimension of system performance include but are not limited to CPU usage data, memory usage data, disk I / O data, network traffic data, system load data, service response time data, etc. CPU usage data can be used to monitor CPU load, including core usage, idle time, etc.; memory usage data can be used to monitor memory usage, free memory, memory usage peak, etc.; disk I / O data can be used to monitor disk read and write speed, I / O operation times, disk queue length, etc.; network traffic data can be used to monitor network throughput, bandwidth usage, packet loss rate, etc.; system load data can be used to monitor the system load average and understand the system load at different times; service response time data can be used to monitor the response time of the service to ensure the timeliness of the service.

[0032] Key performance data collected from the dimension of data processing include but are not limited to data throughput data, data storage capacity data, data processing time data, data integrity data, etc. Data throughput data is used to monitor the input and output rate of data, including the speed of data reception and transmission; data storage capacity data is used to monitor the usage of storage devices, including used space and remaining space; data processing time data is used to monitor the time consumed for data processing, including the time for data cleaning, conversion and loading; data integrity data is used to monitor the integrity of data to ensure that data is not lost or damaged during processing.

[0033] Key performance data collected from the model prediction dimension include but are not limited to prediction accuracy data, model performance index data, model training time data, model resource usage data, etc. Prediction accuracy data is used to monitor the accuracy of model prediction results, including prediction error and confidence interval; model performance index data is used to monitor the performance indicators of the model, such as accuracy, recall rate, F1 score, etc.; model training time data is used to monitor the time required for model training, including data preprocessing, model training and verification time; model resource usage data is used to monitor resource usage during model training and prediction, such as CPU, memory and storage.

[0034] Key performance data collected from the user interaction dimension include but are not limited to user request volume data, user response time data, user error rate data, user satisfaction data, user behavior analysis data, etc. User request volume data is used to monitor the number of user requests, including the total number of requests and the number of requests for each service; user response time data is used to monitor the response time of user requests, including the average response time and the maximum response time; user error rate data is used to monitor the error rate in user requests, including the number of failed requests and exception handling; user satisfaction data is used to collect user satisfaction with the service through surveys or feedback; user behavior analysis data is used to analyze user behavior patterns, including access frequency, dwell time, and page jump paths.

[0035] In the embodiments of the present application, the methods of preprocessing the key performance data include but are not limited to missing value processing, outlier detection and processing, data deduplication, data type conversion, time series data processing, data resampling, data denoising, etc. Among them, missing value processing refers to filling missing values ​​with mean, median, mode or prediction models based on other data; data deduplication refers to identifying and deleting duplicate records to avoid bias in analysis; data type conversion refers to converting data into an appropriate format, such as converting dates from strings to date types or converting fractional data to numeric types; time series data processing refers to performing differential, moving average or seasonal adjustment on time series data; data resampling refers to resampling time series data to change the time granularity of the data, such as converting daily data to monthly data; data denoising refers to removing noise data that deviates significantly from the mean or normal value range.

[0036] Normalizing the collected data allows for comprehensive comparison of data of different dimensions to ensure the accuracy of the evaluation. Normalization is to scale the data so that it falls into a small undetermined interval, usually [0,1] or [-1,1]. The data normalization methods used in the embodiments of the present application include but are not limited to: Min-Max normalization, Z-Score normalization, nonlinear normalization, and other methods.

[0037] Step S2: Input the preprocessed key performance data into the sparse Bayesian learning algorithm for feature selection and dynamically adjust the weights of the features in each dimension, and calculate the health of the energy efficiency monitoring platform of the computer room in real time.

[0038] like Figure 3 As shown, the above step S2 specifically includes the following steps.

[0039] Step S21: input the preprocessed and normalized key performance data into the sparse Bayesian learning algorithm for sparse feature selection, so as to select several features that have the greatest impact on health and dynamically adjust the weights.

[0040] Specifically, during the data processing, the Sparse Bayesian Learning (SBL) algorithm automatically identifies features that have a greater impact on health assessment and assigns significant weights, and assigns zero weights to features that have a smaller impact on health assessment to eliminate them. In this way, redundant or noisy data can be eliminated, improving computational efficiency.

[0041] like Figure 4 As shown, the process of sparse feature selection using the SBL algorithm includes the following.

[0042] Step S21a: Initialize the parameters of the sparse Bayesian model.

[0043] When building a sparse Bayesian model, parameter initialization is a key step, which directly affects the convergence and final performance of the model. The methods for initializing parameters of a sparse Bayesian model include the following.

[0044] A) Initialization of model parameters and ARD parameters: In the sparse Bayesian model, the weight parameter w and the correlation parameter α need to be initialized. Usually, these parameters are initialized to small positive values ​​to avoid numerical instability during the optimization process. This is because in the Bayesian framework, these parameters usually have prior distributions, and small positive values ​​can serve as the initial points of these distributions.

[0045] B) Initialization of weight parameters: The weight parameter w can be initialized to zero or a small random value. This is because in sparse models, it is expected that most weights are close to zero after optimization to achieve feature selection. Initialization to zero or a small value can help the model learn from a simple state and gradually adjust the weights to capture important features in the data.

[0046] C) Initialization of the correlation parameter (ARD parameter): The correlation parameter α controls the variance of the weight parameter, and its initial value has a great influence on the sparsity of the model. These parameters can be initialized to smaller values ​​so that the weight parameters corresponding to unimportant features tend to zero more easily during the optimization process. Such initialization helps the model tend to sparse solutions from the beginning.

[0047] Step S21b: Using L1 regularization, the parameters of the sparse Bayesian model are adjusted through an optimization algorithm so that the weights of unimportant features tend to zero, so as to select several features that have the greatest impact on health.

[0048] L1 regularization is shown in formula (1), also known as Lasso (Least Absolute Shrinkage and Selection Operator). By adding a penalty term of the sum of the absolute values ​​of the parameters to the loss function, some parameters are forced to zero, thereby achieving feature selection and model sparsification. Various optimization algorithms can be used to adjust model parameters, such as gradient descent, stochastic gradient descent, Adam, etc. These algorithms iteratively update parameters to minimize the loss function containing the L1 regularization term. During training, L1 regularization will cause the weights of unimportant features to tend to zero, so the features that have the greatest impact on health can be selected by observing the model parameters. Only features corresponding to non-zero weights are retained to improve prediction efficiency and model interpretability.

[0049] ;Formula (1).

[0050] in, is the original loss function (such as mean squared error), is a hyperparameter of the regularization strength, controlling the weight of the penalty term, are model parameters.

[0051] Step S21c: Evaluate the performance of the sparse Bayesian model and dynamically adjust feature weights according to the performance of the sparse Bayesian model in practical applications; the dynamic adjustment method includes reallocating the weights of the selected features and / or reselecting features.

[0052] Exemplarily, the performance evaluation indicators of the sparse Bayesian model include but are not limited to: accuracy, precision, recall, F1 score, AUC-ROC, mean square error and determination coefficient, etc. Accuracy indicates the proportion of samples predicted correctly, precision indicates the proportion of samples predicted as positive that are actually positive, recall indicates the proportion of samples actually positive that are correctly predicted as positive, F1 score indicates the harmonic mean of precision and recall, which comprehensively measures the performance of the model, AUC-ROC indicates the classification ability of the evaluation model at different thresholds, and mean square error and determination coefficient are used for performance evaluation of regression tasks.

[0053] It should be understood that Sparse Bayesian Learning (SBL) is a machine learning algorithm based on Bayesian probability theory, which is mainly used in regression and classification problems. Its core feature is to automatically select features or dimensions that have a greater impact on prediction by introducing sparsity priors, thereby effectively reducing the complexity of the model and preventing overfitting. SBL shows good robustness and accuracy when processing high-dimensional and large-scale data sets. The SBL algorithm can gradually update the parameters of the model based on given observation data and prior assumptions. Compared with traditional regression or classification methods, SBL can automatically eliminate irrelevant or redundant information, making the model more efficient, concise, and accurate. Specifically, in the health assessment of the energy efficiency supervision platform of the high-efficiency computer room, due to the large number of data dimensions, at least there are system performance dimensions, data processing dimensions, model prediction dimensions, and user interaction dimensions, the impact of different dimensions on the health of the supervision platform has inconsistent weights. Therefore, by introducing the SBL algorithm, the platform can automatically identify the dimensions that have the greatest impact on health, thereby reducing the dependence on secondary dimensions and improving the accuracy and robustness of health assessment.

[0054] The SBL algorithm automatically adjusts the weights of the key performances of the system performance dimension, data processing dimension, model prediction dimension, and user interaction dimension based on the key performance data of the computer room energy efficiency supervision platform to be evaluated, and gives priority to the dimensions that have a greater impact on the health of the computer room energy efficiency supervision platform. For example, the impact of the system performance dimension and the data processing dimension on the health of the computer room energy efficiency supervision platform is greater than that of the model prediction dimension and the user interaction dimension, so the weights of the first two dimensions are increased and given priority.

[0055] Step S22: According to the output result of the sparse Bayesian learning algorithm, the health of the computer room energy efficiency monitoring platform is calculated in real time through a pre-built health prediction model.

[0056] In some examples, the actual performance of the platform on each indicator is compared with the ideal performance, and a comprehensive health score is calculated based on the weight of each indicator. The score is further fine-tuned by an adjustment factor. The specific health prediction model is as follows.

[0057] ;Formula (2).

[0058] in, Indicates The weight of the indicator, The energy efficiency monitoring platform of the computer room is The actual performance on key performance indicators, The energy efficiency monitoring platform of the computer room is Ideal performance on key performance indicators; Indicates the degree of closeness between actual performance and ideal performance; is an adjustment factor, which is used to fine-tune the health score to more accurately reflect the actual health status. The adjustment factor can be dynamically adjusted according to environmental conditions (such as temperature, humidity, system load, etc.) so that the health score can respond to these changes in a timely manner. Among them, C represents dynamic factors related to system environment, load, external conditions, etc. The parameters used in the above formula (2) are a subset of the data collected in the above dimensions. Specifically, the collected data is screened by the sparse Bayesian learning algorithm to extract relatively more important parameters. Subsequently, these key parameters are input into the health prediction model to further predict the health.

[0059] To facilitate understanding, this paper explains the practical application case of sparse Bayesian learning algorithm in health assessment.

[0060] The application scenario is: In a large industrial computer room, the energy efficiency monitoring platform needs to process a large amount of data from different equipment in real time, including cooling towers, refrigeration pumps, and air handling equipment. Due to the diversity of equipment and complex data, the operation status of the platform is easily affected by load fluctuations.

[0061] The goal is: Through the SBL algorithm, the platform can automatically adjust the weights of each dimension and perform sparsity screening on real-time data to ensure the accuracy and rapid response of health assessment.

[0062] The implementation steps are: S22a: Data collection and feature extraction. Collect real-time data from various monitoring modules in the platform (such as system performance, data processing, and model prediction) and extract key features.

[0063] S22b: Apply the sparse Bayesian learning algorithm. Input the collected feature data into the SBL algorithm, automatically select several features that have the greatest impact on health, and ignore the noise data.

[0064] S22c: Health calculation and dynamic adjustment: According to the output of the SBL algorithm, the weight of each feature is dynamically adjusted, and the platform health is calculated in real time.

[0065] S22d: Feedback and optimization. Use the health results to provide feedback to operation and maintenance personnel to help them maintain or optimize the platform configuration when necessary.

[0066] Therefore, through the SBL algorithm, the platform can more accurately identify the main factors affecting health. Sparse Bayesian learning can effectively reduce the data dimension and reduce the computing burden of the platform. Due to real-time adjustment and feature screening, the platform can respond quickly when the data load changes and maintain efficient operation.

[0067] Step S23: Use cross-validation to evaluate the generalization ability of the model and adjust the hyperparameters of the model to optimize the model performance.

[0068] In some examples, cross-validation methods include K-fold cross-validation, Leave-One-Out Cross-Validation (LOOCV), and Stratified K-fold cross-validation. Divide the dataset into a training set and a validation set, or divide the dataset into multiple subsets according to the cross-validation method. Train the model on the training set, using different hyperparameter combinations. For each hyperparameter combination, use the validation set or the subset in the cross-validation to evaluate the performance of the model. Available performance metrics include accuracy, recall, F1 score, mean squared error (MSE), etc., depending on the nature of the problem (classification or regression). Based on the evaluation results, select the hyperparameter combination that performs best on the validation set. For example, select parameters that maximize the performance metric or minimize the loss function. The hyperparameter space can be systematically explored through methods such as Grid Search or Random Search to find the best hyperparameter combination. Retrain the model based on the best hyperparameter combination and perform a final evaluation on an independent test set to ensure the generalization ability of the model. Finally, the test set is used to verify the final performance of the model to ensure that the model can perform well on unseen data. Through the above steps, the SBL algorithm can correct the errors of the prediction model and improve the prediction accuracy.

[0069] Step S3: Input the calculated health data into the Bayesian optimal classification algorithm for analysis to obtain a classification result of the operating status of the computer room energy efficiency monitoring platform, and issue an early warning based on the classification result.

[0070] It should be understood that Bayesian Optimal Classification is a classification algorithm based on Bayesian theory, which aims to find the classification result that maximizes the posterior probability by deriving the prior distribution and likelihood function of each type of data. Bayesian optimal classification is widely used to process multi-category data sets. It can automatically classify according to different input data and modify the classification results according to prior and current data. The core of this algorithm is that it can give the best classification result by considering uncertainty factors. In system health assessment, Bayesian optimal classification can help the platform judge the current status (such as "normal", "warning" or "fault") in real time and provide targeted feedback.

[0071] In the health evaluation of this application, the health status of the computer room energy efficiency monitoring platform can usually be divided into different levels such as "normal", "slightly abnormal", "severely abnormal", "fault", etc. Through the Bayesian optimal classification algorithm, the computer room energy efficiency monitoring platform can automatically classify the health score results and predict possible future status changes based on the combination of historical data and real-time data.

[0072] In some examples, the calculated health is classified using the Bayesian optimal classification algorithm, and the operating status of the computer room energy efficiency monitoring platform can be divided into different health levels in real time, helping operation and maintenance personnel to quickly determine the health status of the computer room energy efficiency monitoring platform. The Bayesian optimal classification algorithm is based on posterior probability, that is, in the ideal situation where all relevant probabilities are known, Bayesian decision theory considers how to select the optimal category label based on these probabilities and misjudgment losses. In this way, operation and maintenance personnel can quickly determine the health status of the platform and take corresponding maintenance measures.

[0073] In some examples, the Bayesian optimal classification algorithm can adjust the health classification results in real time according to the newly input data to ensure the accuracy of the classification. An important feature of the Bayesian optimal classifier is that it can adjust the classification results in real time according to the newly input data to ensure the accuracy of the classification. This is because the Bayesian classifier considers the prior probability and the posterior probability in the decision-making process, and can update these probabilities as new data is added, thereby achieving adaptive correction.

[0074] In some examples, through the Bayesian optimal classifier, the platform can issue early warnings when the health level drops to a critical threshold based on the classification results. This is based on the principle of conditional risk minimization in Bayesian decision theory, that is, selecting the category label that minimizes the conditional risk for each sample. In this way, potential failures can be prevented and the stability and reliability of the system can be improved.

[0075] Therefore, the application of Bayesian optimal classifier in health assessment can effectively help operation and maintenance personnel monitor and predict the operating status of the platform in real time, thereby improving the stability of the system and reducing the risk of failure.

[0076] It is worth noting that the probabilistic decision-making and real-time dynamic update capabilities of the Bayesian optimal classifier are particularly important in health assessment, because the classification of health status often involves uncertainty and risk assessment. The Bayesian optimal classifier can provide probabilistic estimates of the health status of the system and can update these estimates in real time to reflect the latest system status. This capability makes the Bayesian optimal classifier particularly effective in health assessment, because it can not only provide an accurate classification of the current status, but also predict and warn of possible future problems, thereby achieving preventive maintenance and risk management.

[0077] To facilitate understanding, this paper explains the practical application case of Bayesian optimal classification algorithm in health assessment.

[0078] The application scenario is: in a high-load building computer room, the energy efficiency monitoring platform needs to monitor the operating status of multiple subsystems in real time and predict possible failures in advance based on changes in health.

[0079] The goal is: Through the Bayesian optimal classification algorithm, the energy efficiency monitoring platform can automatically classify different health levels, and provide early warning of potential system failures, helping operation and maintenance personnel take preventive measures.

[0080] The implementation steps are: First, the health history data and corresponding operation status data of the computer room energy efficiency monitoring platform are collected and divided into training set and verification set according to a certain ratio. For example, the health score data of each subsystem in the energy efficiency monitoring platform is collected, and the three states of "normal", "warning" and "fault" are defined according to the historical data, so as to preliminarily establish the association between the health score data and the operation status of the platform.

[0081] It should be understood that the collected health score data of each subsystem in the energy efficiency supervision platform is the basis for training and prediction of the Bayesian optimal classification model. Based on historical data analysis, the characteristics and thresholds of each category are determined. For example, if the health score is within a certain range, the system is defined as "normal", if the score is close to the fault threshold, the system is defined as "warning", and if the score is below a certain threshold, it is defined as "fault".

[0082] Secondly, the training set is input into the Bayesian optimal classification model for model training; the Bayesian optimal classification model learns how to calculate the posterior probability based on the input health history data and predict the most likely platform operation status.

[0083] When the Bayesian optimal classification model is used for classification, the Bayesian optimal classification model selects the category label with the maximum posterior probability as the prediction result according to the classification goal of minimizing the classification error rate.

[0084] Before Bayesian classification, the Bayesian optimal classification model needs to be trained using historical data so that the model can learn the relationship between different health scores and system status. Once the model training is completed, the real-time health score can be input into the model, and the model will automatically classify the current system status into "normal", "warning", and "fault" according to the learned rules.

[0085] Finally, the validation set is input into the trained Bayesian optimal classification model to evaluate the performance of the model. Based on the health classification results of the Bayesian optimal classification model, the energy efficiency supervision platform issues real-time warning information to the operation and maintenance personnel and provides corresponding maintenance suggestions.

[0086] Therefore, by using the Bayesian optimal classification algorithm to classify the calculated health, the energy efficiency monitoring platform can identify possible performance bottlenecks or failures in advance. Based on the prediction results, operation and maintenance personnel can arrange preventive maintenance in advance to reduce the risk of unplanned downtime. Through health classification, the energy efficiency monitoring platform can more accurately judge the system status and ensure stable operation.

[0087] In the above, the implementation process and principle of a health evaluation method for a room energy efficiency monitoring platform provided by an embodiment of the present invention are explained in detail. In the following, the health evaluation system, medium, and program product of a room energy efficiency monitoring platform provided by an embodiment of the present invention will be described in conjunction with the accompanying drawings.

[0088] like Figure 5 As shown, a structural schematic diagram of a health assessment system for a comprehensive digital management platform of a refrigeration room in an embodiment of the present invention is shown.

[0089] The health evaluation system 500 of the comprehensive digital management platform of the refrigeration room includes: a data acquisition module 501, a health calculation module 502, and a state classification module 503. The data acquisition module 501 is used to collect several key performance data of the energy efficiency supervision platform of the computer room to be evaluated based on multiple preset dimensions, and pre-process the key performance data; the multiple preset dimensions include: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension. The health calculation module 502 is used to input the pre-processed key performance data into the sparse Bayesian learning algorithm for feature selection and dynamically adjust the weights of the features of each dimension, and calculate the health of the energy efficiency supervision platform of the computer room in real time. The state classification module 503 is used to input the calculated health data into the Bayesian optimal classification algorithm for analysis, so as to obtain the classification result of the operating status of the energy efficiency supervision platform of the computer room, and trigger an early warning based on the classification result.

[0090] It should be understood that the specific process of each module executing the above corresponding steps has been described in detail in the above method embodiment, and for the sake of brevity, it will not be repeated here. It should also be understood that the division of modules in the embodiment of the present application is schematic, which is only a logical function division, and there may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present application can be integrated into a processor, or it can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0091] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" represent examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0092] In the embodiments of the present application, "at least one" refers to one or more, and "plurality" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can represent: a, b, c, ab, ac, bc or abc, where a, b, c can be single or multiple.

[0093] Figure 6 is a schematic block diagram of a computer device provided in an embodiment of the present application. Figure 6 As shown in FIG. 6 , the computer device includes: at least one processor 601, a memory 602, at least one network interface 603, and a user interface 605. The various components in the device are coupled together via a bus system 604. It can be understood that the bus system 604 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 604 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, the bus system 604 is not described in detail. Figure 6 In the specification, various buses are labeled as bus systems.

[0094] The user interface 605 may include a display, a keyboard, a mouse, a trackball, a click gun, keys, buttons, a touch pad or a touch screen.

[0095] It is understood that the memory 602 can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM). The memory described in the embodiments of the present invention is intended to include but is not limited to these and any other suitable categories of memory.

[0096] The memory 602 in the embodiment of the present invention is used to store various categories of data to support the operation of the computer device 600. Examples of these data include: any executable program for operating on the computer device 600, such as an operating system 6021 and an application 6022; the operating system 6021 includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application 6022 may include various applications, such as a media player (Media Player), a browser (Browser), etc., for implementing various application services. The health assessment method of the comprehensive digital management platform for the refrigeration room provided in the embodiment of the present invention may be included in the application 6022.

[0097] The method disclosed in the above embodiment of the present invention can be applied to the processor 601, or implemented by the processor 601. The processor 601 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 601 or the instruction in the form of software. The above processor 601 may be a general processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor 601 can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiment of the present invention. The general processor 601 may be a microprocessor or any conventional processor, etc. In combination with the steps of the accessory optimization method provided in the embodiment of the present invention, it can be directly embodied as a hardware decoding processor to execute, or it can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium, which is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0098] In an exemplary embodiment, the computer device 600 may be one or more application specific integrated circuits (ASIC), DSP, programmable logic device (PLD), complex programmable logic device (CPLD) to execute the aforementioned method.

[0099] According to the method provided in the embodiment of the present application, the present application also provides a computer program product, which includes: a computer program code, when the computer program code is run on a computer, the computer executes Figures 1 to 4 A method for assessing the health of a comprehensive digital management platform for a refrigeration room in any of the embodiments shown.

[0100] According to the method provided in the embodiment of the present application, the present application also provides a computer-readable storage medium, which stores a program code. When the program code runs on a computer, the computer executes the above method.

[0101] The terms "component", "module", "system", etc. used in this specification are used to represent computer-related entities, hardware, firmware, a combination of hardware and software, software, or software in execution. For example, a component can be, but is not limited to, a process, a processor, an object, an executable file, an execution thread, a program and / or a computer running on a processor. By way of illustration, both applications and computing devices running on a computing device can be components. One or more components may reside in a process and / or an execution thread, and a component may be located on a computer and / or distributed between two or more computers. In addition, these components may be executed from various computer-readable media having various data structures stored thereon. Components may, for example, communicate through local and / or remote processes according to signals having one or more data packets (e.g., data from two components interacting with another component between a local system, a distributed system and / or a network, such as the Internet interacting with other systems through signals).

[0102] Those of ordinary skill in the art will appreciate that the various illustrative logical blocks and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0103] 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.

[0104] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0105] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0106] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0107] In the above embodiments, the functions of each functional unit can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions (programs). When the computer program instructions (programs) are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. Available media may be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., high-density digital video discs (DVDs), or semiconductor media (e.g., solid state disks (SSDs)).

[0108] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0109] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0110] In summary, the present application provides a health assessment method, device, system, medium and program product for a comprehensive digital management platform for refrigeration rooms. Through the sparse Bayesian learning algorithm and the Bayesian optimal classification algorithm, the energy efficiency supervision platform of the room can achieve real-time monitoring of health and quickly make corresponding adjustments to maintain the optimal operating state. The sparse Bayesian learning algorithm can not only select features but also dynamically adjust weights to ensure that they can flexibly adapt to changing actual needs. By comprehensively evaluating multiple dimensions such as system performance, data processing capabilities, model prediction accuracy, and user experience, the platform provides a comprehensive health status analysis to ensure a comprehensive grasp of the platform's operating status. The platform's operating status is further evaluated based on the predicted health, and early warnings and maintenance are made in a timely manner. Therefore, the present application effectively overcomes the various shortcomings in the prior art and has a high industrial utilization value.

[0111] The above embodiments are merely illustrative of the principles and effects of the present application and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A health assessment method for a comprehensive digital management platform for a refrigeration room, characterized in that: include: Collecting a number of key performance data of the computer room energy efficiency monitoring platform to be evaluated based on multiple preset dimensions, and preprocessing the key performance data; The multiple preset dimensions include: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension; The pre-processed key performance data is input into the sparse Bayesian learning algorithm for feature selection and dynamic adjustment of the weights of the features in each dimension, and the health of the energy efficiency monitoring platform of the computer room is calculated in real time; The calculated health data is input into the Bayesian optimal classification algorithm for analysis to obtain the classification result of the operating status of the computer room energy efficiency monitoring platform, and an early warning is issued based on the classification result.

2. The health evaluation method of the comprehensive digital management platform of the refrigeration room according to claim 1 is characterized in that: The pre-processed key performance data is input into the sparse Bayesian learning algorithm to select features and dynamically adjust the weights of features in each dimension, and calculate the health of the energy efficiency monitoring platform of the computer room in real time, including: The preprocessed and normalized key performance data is input into the sparse Bayesian learning algorithm for sparse feature selection to select several features that have the greatest impact on health and dynamically adjust the weights; According to the output results of the sparse Bayesian learning algorithm, the health status of the energy efficiency monitoring platform of the computer room is calculated in real time through the pre-built health prediction model; Cross-validation was used to evaluate the generalization ability of the model and the hyperparameters of the model were adjusted to optimize the model performance.

3. The health evaluation method of the comprehensive digital management platform of the refrigeration room according to claim 2 is characterized in that: The preprocessed and normalized key performance data is input into the sparse Bayesian learning algorithm for sparse feature selection to select several features that have the greatest impact on health and dynamically adjust the weights, which includes: Initialize the parameters of the sparse Bayesian model; Using L1 regularization, the parameters of the sparse Bayesian model are adjusted through an optimization algorithm so that the weights of unimportant features tend to zero, so as to select several features that have the greatest impact on health. Evaluate the performance of the sparse Bayesian model and dynamically adjust feature weights based on the performance of the sparse Bayesian model in practical applications; the dynamic adjustment method includes reallocating the weights of selected features and / or reselecting features.

4. The method according to claim 2, characterized in that: The health prediction model includes: ;in, Indicates The weight of the indicator, The energy efficiency monitoring platform of the computer room is The actual performance on key performance indicators, The energy efficiency monitoring platform of the computer room is Ideal performance on key performance indicators; Indicates the degree of closeness between actual performance and ideal performance; is an adjustment factor used to fine-tune the health score.

5. The health evaluation method of the comprehensive digital management platform of the refrigeration room according to claim 1 is characterized in that: The step of inputting the calculated health data into the Bayesian optimal classification algorithm for analysis to obtain a classification result of the operating status of the computer room energy efficiency monitoring platform includes the following steps: Collect the health history data and corresponding operation status data of the computer room energy efficiency monitoring platform, and divide them into training set and validation set according to a certain ratio; Inputting the training set into the Bayesian optimal classification model for model training; the Bayesian optimal classification model learns how to calculate the posterior probability based on the input health history data and predict the most likely platform operation status; The validation set was input into the trained Bayesian optimal classification model to evaluate the performance of the model.

6. The health evaluation method of the integrated digital management platform of the refrigeration room according to claim 5 is characterized in that: The Bayesian optimal classification model selects the category label with the maximum posterior probability as the prediction result according to the classification goal of minimizing the classification error rate.

7. A health assessment system for a comprehensive digital management platform of a refrigeration room, characterized in that: include: A data collection module, used to collect a number of key performance data of the computer room energy efficiency monitoring platform to be evaluated based on multiple preset dimensions, and pre-process the key performance data; The multiple preset dimensions include: system performance dimension, data processing dimension, model prediction dimension and user interaction dimension; The health calculation module is used to input the pre-processed key performance data into the sparse Bayesian learning algorithm for feature selection and dynamically adjust the weights of the features of each dimension, and calculate the health of the energy efficiency monitoring platform of the computer room in real time; The status classification module is used to input the calculated health data into the Bayesian optimal classification algorithm for analysis to obtain the classification results of the operating status of the computer room energy efficiency supervision platform and to issue an early warning based on the classification results.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the health assessment method of the comprehensive digital management platform of the refrigeration room as described in any one of claims 1 to 6.

9. A computer program product, characterized in that The computer program product includes computer program code. When the computer program code runs on a computer, the computer implements the health assessment method of the comprehensive digital management platform of the refrigeration room as described in any one of claims 1 to 6.

10. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the health assessment method of the comprehensive digital management platform of the refrigeration room as described in any one of claims 1 to 6.

Images