Refrigerating machine room equipment performance model health degree evaluation method and system, medium and terminal

Abstract

The invention provides a refrigeration machine room equipment performance model health degree evaluation method and system, a medium and a terminal, and the method comprises the steps: monitoring the real-time operation data, historical operation data and equipment performance prediction data generated in the operation process of a refrigeration machine room, carrying out the prediction accuracy of an equipment performance model, and employing a multi-mode fusion technology, and obtaining the health degree of the refrigeration machine room. The weights of different input dimensions in the health degree calculation model are dynamically adjusted in combination with the current operation condition, the environmental parameters and the equipment aging condition, and self-adaptive health degree evaluation of the equipment performance model is realized, so that the health degree of the model can be evaluated more comprehensively and accurately, errors of a single data source are effectively reduced, and the user experience is improved. And the comprehensive health degree score of the most representative equipment performance model is calculated. And when the comprehensive health degree score of the equipment performance model is lower than the preset health degree threshold value, model retraining is automatically triggered, so that adaptive adjustment of the model is realized to ensure that the equipment performance model is always in an optimal state.

Description

Technical Field

[0001] The present application relates to the field of industrial refrigeration technology, and in particular to a method, system, medium and terminal for evaluating the health of a refrigeration room equipment performance model. Background Art

[0002] In the operation and management of efficient industrial refrigeration rooms, chillers, water pumps and cooling towers are the core equipment of the refrigeration system. Their operating status and performance are directly related to the energy efficiency and stability of the entire system. In order to maximize the operating efficiency of these equipment and ensure their long-term stable operation, modeling their operating data to fit the equipment performance curve has become a common practice in the industry. This modeling not only helps predict the operating trend of the equipment, but also provides a scientific basis for equipment maintenance and optimization.

[0003] However, there are still many challenges and shortcomings in the current field of equipment performance model health monitoring. Traditional monitoring methods mostly focus on monitoring a single performance parameter, such as energy consumption, temperature difference or operating efficiency, and use the abnormality of these indicators to indirectly reflect the health of the model. However, this single-dimensional evaluation method ignores many factors that may affect the accuracy of the model, such as changes in ambient temperature, aging of the equipment itself, and dynamic fluctuations in operating conditions. Therefore, it is often difficult to obtain comprehensive and accurate evaluation results by monitoring only a single parameter.

[0004] Periodic calibration is another commonly used method for evaluating model health. By regularly recalibrating the equipment performance model, the deviation of the model caused by environmental factors or equipment aging can be corrected to a certain extent. However, the limitation of this method is that it cannot reflect the changes in the health status of the equipment during dynamic operation in real time, resulting in a certain lag in the evaluation results, making it difficult to guide equipment maintenance and optimization work in a timely manner.

[0005] Manual maintenance is an auxiliary means, usually when the model shows obvious deviation, the operation and maintenance personnel intervene to adjust or retrain. However, this method is not only inefficient, but also cannot achieve adaptive adjustment of the model, and it is difficult to cope with complex and changing operating environments. In addition, excessive reliance on manual maintenance will increase operating costs and reduce the overall reliability and stability of the system.

[0006] More importantly, existing methods often ignore the importance of environmental and equipment aging factors in model health assessment. The environmental conditions, service life, and degree of equipment aging of the equipment will have a significant impact on the accuracy of the model. However, in the current assessment system, these factors have not been fully considered and quantified, resulting in incomplete and inaccurate assessment results.

[0007] In summary, the existing equipment performance model health assessment methods have deficiencies in comprehensiveness, real-time and accuracy, and cannot accurately reflect the health status of the equipment performance model. Summary of the invention

[0008] In view of the shortcomings of the prior art mentioned above, the purpose of the present application is to provide a refrigeration room equipment performance model health assessment method, system, medium and terminal, which are used to solve the technical problems of the existing equipment performance model health assessment method, such as relying on a single performance parameter monitoring, failing to reflect the changes in the health status of the equipment during dynamic operation in real time, and failing to achieve adaptive adjustment of the model.

[0009] To achieve the above-mentioned purpose and other related purposes, the first aspect of the present application provides a method for evaluating the health of a refrigeration room equipment performance model, including: collecting real-time operation data and historical operation data generated during the operation of the refrigeration room; preprocessing the collected real-time operation data and historical operation data respectively; evaluating the prediction accuracy of a pre-built equipment performance model based on the preprocessed real-time operation data and the pre-processed historical operation data; adjusting the weights in a pre-built health calculation model based on the prediction accuracy of the equipment performance model and based on multimodal fusion technology; and calculating the comprehensive health score of the equipment performance model based on the health calculation model after the weight adjustment.

[0010] In some embodiments of the first aspect of the present application, the calculation formula of the health degree calculation model includes:

[0011] ;

[0012] Where H represents the comprehensive health score of the equipment performance model; i represents the operating parameters of different equipment in the refrigeration room; w i Indicates the weight of the operating coefficient corresponding to the operating parameters of different equipment in the refrigeration room; β i Indicates the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room; T i represents the real-time running data after preprocessing; P i represents the equipment performance prediction data; γ(E) represents the equipment aging weight function; n represents the total number of different equipment operating parameters in the refrigeration room; exp represents the exponential function.

[0013] In some embodiments of the first aspect of the present application, according to the prediction accuracy of the equipment performance model and based on the multimodal fusion technology, the method of adjusting the weights in the pre-built health calculation model includes: based on the multimodal fusion technology, adjusting the operating coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, and the equipment aging weight function.

[0014] In some embodiments of the first aspect of the present application, the historical operating data includes one or more combinations of operating parameter data, equipment aging data, and operating environment data; the method of adjusting the equipment aging weight function includes: evaluating the aging degree of different equipment in the refrigeration room based on the operating parameter data, equipment aging data, and operating environment data, and based on a pre-built equipment aging model; and adjusting the equipment aging weight function based on the aging degree of different equipment in the refrigeration room using multivariate regression analysis technology.

[0015] In some embodiments of the first aspect of the present application, the method further includes: comparing the comprehensive health score of the device performance model with a preset health threshold, and determining whether the device performance model needs to be retrained based on the comparison result.

[0016] In some embodiments of the first aspect of the present application, the method for evaluating the prediction accuracy of a pre-built equipment performance model includes: based on a data offset detection algorithm, comparing the pre-processed real-time operation data with the training data for building the equipment performance model, and calculating the offset value between the pre-processed real-time operation data and the training data; and evaluating the prediction accuracy of the pre-built equipment performance model based on the offset value between the pre-processed real-time operation data and the training data.

[0017] In some embodiments of the first aspect of the present application, the method for evaluating the prediction accuracy of a pre-built equipment performance model includes: obtaining equipment performance prediction data according to the pre-processed real-time operation data and based on the pre-built equipment performance model; calculating the deviation value between the pre-processed real-time operation data and the equipment performance prediction data; calculating the fitting error of the equipment performance model according to the deviation value between the pre-processed real-time operation data and the equipment performance prediction data and through curve fitting technology; and evaluating the prediction accuracy of the pre-built equipment performance model according to the fitting error of the equipment performance model.

[0018] To achieve the above-mentioned purpose and other related purposes, the second aspect of the present application provides a refrigeration room equipment performance model health assessment system, including: a data acquisition module, used to collect real-time operation data and historical operation data generated during the operation of the refrigeration room; a data preprocessing module, used to preprocess the collected real-time operation data and historical operation data respectively; a model accuracy assessment module, used to evaluate the prediction accuracy of a pre-built equipment performance model based on the preprocessed real-time operation data and the preprocessed historical operation data; a weight adjustment module, used to adjust the weights in a pre-built health calculation model based on the prediction accuracy of the equipment performance model and based on multimodal fusion technology; a health assessment module, used to calculate the comprehensive health score of the equipment performance model based on the health calculation model after adjusting the weight.

[0019] 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 refrigeration room equipment performance model health assessment method as described above.

[0020] To achieve the above-mentioned purpose and other related purposes, the fourth aspect of the present application provides an electronic terminal, including a memory, a processor and a computer program stored in the memory, and the processor executes the computer program to implement the refrigeration room equipment performance model health assessment method as described above.

[0021] As described above, the refrigeration room equipment performance model health evaluation method, system, medium and terminal of the present application have the following beneficial effects: by monitoring the real-time operation data, historical operation data and equipment performance prediction data generated by the refrigeration room during operation, according to the prediction accuracy of the equipment performance model, and using multimodal fusion technology, combined with the current operating status, environmental parameters and equipment aging, the weights of different input dimensions in the health calculation model are dynamically adjusted, and the adaptive health evaluation of the equipment performance model is realized, so that the model health can be evaluated more comprehensively and accurately, and the error of a single data source can be effectively reduced, and then the most representative equipment performance model comprehensive health score is calculated. When the comprehensive health score of the equipment performance model is lower than the preset health threshold, the model retraining is automatically triggered, and the model is adaptively adjusted to ensure that the equipment performance model is always in the optimal state, so that the equipment performance model can comprehensively evaluate the current operating status and future operating status of the equipment, so that the health of the refrigeration room in the dynamic operation process can be evaluated in real time and accurately, providing strong support for the operation optimization of the refrigeration room. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1Shown is a flow chart of a method for evaluating the health status of a refrigeration room equipment performance model in one embodiment of the present application.

[0023] Figure 2 Shown is a schematic diagram of a flow chart for evaluating the prediction accuracy of a device performance model in one embodiment of the present application.

[0024] Figure 3 Shown is another flow chart of evaluating the prediction accuracy of the equipment performance model in one embodiment of the present application.

[0025] Figure 4 Shown is a schematic block diagram of a refrigeration room equipment performance model health assessment system in one embodiment of the present application.

[0026] Figure 5 Shown is a schematic diagram of the structure of an electronic terminal in one embodiment of the present application. DETAILED DESCRIPTION

[0027] 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.

[0028] The existing equipment performance model health assessment methods are insufficient in terms of comprehensiveness, real-timeness and accuracy, and cannot accurately reflect the health status of the equipment performance model. In order to solve the problems in the above-mentioned background technology, the present application provides a refrigeration room equipment performance model health assessment method, system, medium and terminal, which are used to solve the technical problems of the existing equipment performance model health assessment methods, such as relying on a single performance parameter monitoring, being unable to reflect the changes in the health status of the equipment during dynamic operation in real time, and being unable to achieve adaptive adjustment of the model.

[0029] To facilitate understanding of the embodiments of the present application, first Figure 1 Detailed description. Figure 1 The flowchart of the refrigeration room equipment performance model health evaluation method in the embodiment of the present invention is shown. The refrigeration room equipment performance model health evaluation method in this embodiment mainly includes the following steps:

[0030] S101: Collecting real-time operation data and historical operation data generated during the operation of the refrigeration room.

[0031] In this embodiment, the real-time operation data and historical operation data of the refrigeration machine, water pump, cooling tower and other equipment in the refrigeration room are collected. The real-time operation data can be used to reflect the current operation status of the refrigeration room. The historical operation data is stored in the database and can be used for long-term trend analysis and comparison. Various sensors are installed in the refrigeration room, including temperature sensors, humidity sensors, energy consumption metering equipment, and operation status monitoring equipment. The sensors can collect the temperature, humidity, energy consumption, pressure, flow, equipment operation status and other operation data of the refrigeration room in real time. At the same time, the data transmission function of the equipment in the refrigeration room can also output operation data, such as equipment efficiency, water volume and other data. Various operation data are stored in the data center to facilitate the call for analysis and judgment of the current status of the refrigeration room.

[0032] S102: Pre-processing the collected real-time operation data and historical operation data respectively.

[0033] In this embodiment, the collected real-time operation data and historical operation data are preprocessed respectively, and the preprocessing methods include but are not limited to: data cleaning preprocessing, data outlier preprocessing, data normalization preprocessing, data type conversion preprocessing, etc.

[0034] In this embodiment, the data cleaning preprocessing is to delete duplicate data, fill in missing values, etc. to ensure the integrity and accuracy of the data.

[0035] In this embodiment, the data outlier preprocessing is to identify and process outliers (such as maximum or minimum values) in the data, which can be detected and processed by setting thresholds, using statistical methods (such as the 3σ principle) or machine learning algorithms. The threshold setting is to set reasonable upper and lower thresholds according to the characteristics of the data and business needs, and the data exceeding the threshold is regarded as an outlier. Use statistical methods (such as the 3σ principle) to detect and process outliers. The 3σ principle means that the probability that the data value is within the range of the mean plus or minus 3 times the standard deviation is 99.73%, and data outside this range can be regarded as an outlier. Use machine learning algorithms (such as isolation forests) to automatically detect and process outliers. This method can adapt to the dynamic changes of data and improve the accuracy and efficiency of outlier detection.

[0036] In this embodiment, the data normalization preprocessing is to scale the data to a specific range so that data of different magnitudes can be compared and comprehensively analyzed. For the key operating parameters of equipment such as chillers, water pumps, cooling towers, etc. collected in real time, normalization processing is required because their numerical ranges and units may be different.

[0037] In this embodiment, the data type conversion preprocessing refers to converting different data types in the original data into a format or type suitable for analysis and modeling before performing data analysis and modeling.

[0038] S103: Evaluate the prediction accuracy of the pre-built equipment performance model based on the pre-processed real-time operation data and the pre-processed historical operation data.

[0039] In this embodiment, if Figure 2 As shown, a flow chart of evaluating the prediction accuracy of a device performance model in an embodiment of the present invention is shown. The method of evaluating the prediction accuracy of a pre-built device performance model includes:

[0040] S1031: Based on a data offset detection algorithm, the preprocessed real-time operation data is compared with the training data for constructing the device performance model, and an offset value between the preprocessed real-time operation data and the training data is calculated.

[0041] In this embodiment, the equipment performance model is trained using historical operation data stored in the data center and is used to predict the operation status of the equipment. The preprocessed real-time operation data is compared with the training data for constructing the equipment performance model to obtain an offset value between the preprocessed real-time operation data and the training data, thereby evaluating the offset degree between the preprocessed real-time operation data and the training data.

[0042] In this embodiment, the data offset detection algorithm includes: KL divergence algorithm, Mahalanobis distance algorithm, etc. The KL divergence algorithm (Kullback-Leibler Divergence) is an algorithm for measuring the difference between two probability distributions. First, the pre-processed real-time operation data and model training data are converted into the form of probability distribution. Then, the probability distribution of the pre-processed real-time operation data (set as P) is compared with the probability distribution of the model training data (set as Q), and the KL divergence value is calculated. The KL divergence value represents the degree of difference between the two data distributions, thereby obtaining the offset value between the pre-processed real-time operation data and the training data. If the KL divergence value is small, it means that the distribution of the real-time operation data of the device is similar to the distribution of the model training data, and the model has good generalization ability. If the KL divergence value is large, it indicates that there is a large difference between the two data distributions, which may be caused by changes in the device state, interference from environmental factors, or insufficient model training.

[0043] In this embodiment, the Mahalanobis distance algorithm is an effective multi-dimensional space data point similarity measurement algorithm. First, the mean and covariance matrix of the pre-processed real-time operation data and model training data are calculated respectively, which is the basis of Mahalanobis distance calculation. Then, for each data point in the pre-processed real-time operation data, the Mahalanobis distance between it and the mean of the model training data is calculated. The obtained Mahalanobis distance value reflects the statistical distance between the pre-processed real-time operation data point and the mean of the model training data, thereby obtaining the offset value between the pre-processed real-time operation data and the training data. If the distance value is small, it means that the distribution of the data point is similar to that of the model training data, and the model has a good prediction ability for the data point. If the distance value is large, it indicates that there is a large difference in the distribution of the data point and the model training data, which may be caused by changes in the device state, interference from environmental factors, and other reasons.

[0044] S1032: Evaluate the prediction accuracy of the pre-built equipment performance model according to the offset value between the pre-processed real-time operation data and the training data.

[0045] In this embodiment, if the offset value between the preprocessed real-time operation data and the training data exceeds the preset offset threshold, it means that the prediction accuracy of the device performance model has decreased, and the device performance cannot be accurately predicted based on the real-time operation data. The model health score will decrease, and it is necessary to collect training data again and train the model to adapt to the new environment. If the offset value between the preprocessed real-time operation data and the training data does not exceed the preset offset threshold, it means that the prediction accuracy of the device performance model remains within an acceptable range, the model can more accurately predict the device performance based on the real-time operation data, and the model health is good.

[0046] In this embodiment, if Figure 3 As shown, another flow chart of evaluating the prediction accuracy of the device performance model in an embodiment of the present invention is shown. The method of evaluating the prediction accuracy of the pre-built device performance model includes:

[0047] S1031A: Acquire equipment performance prediction data according to the preprocessed real-time operation data and based on a pre-built equipment performance model.

[0048] In this embodiment, the pre-processed real-time operation data is applied to a pre-built equipment performance model, and equipment performance prediction data is obtained by inputting the pre-processed real-time operation data into the equipment performance model.

[0049] S1032B: Calculate the deviation value between the preprocessed real-time operation data and the equipment performance prediction data.

[0050] In this embodiment, the deviation value between the preprocessed real-time operation data and the device performance prediction data is obtained by calculating the difference between the two or a certain distance measurement. The deviation value can be a scalar (indicating the size of the overall deviation) or a vector (indicating the deviation of each data point).

[0051] S1033C: Calculate the fitting error of the equipment performance model based on the deviation value between the preprocessed real-time operation data and the equipment performance prediction data by using curve fitting technology.

[0052] In this embodiment, the curve fitting technique is a mathematical method for finding a curve that best describes a set of data points. The pre-processed real-time operation data is used as the independent variable (input), the equipment performance prediction data is used as the dependent variable (output), and the best fitting curve is found through the curve fitting technique. The fitting error is a measure of the difference between the fitting curve and the actual data points, and is usually expressed using indicators such as mean square error or weighted square error.

[0053] S1034D: Evaluate the prediction accuracy of the pre-built equipment performance model based on the fitting error of the equipment performance model.

[0054] In this embodiment, the prediction accuracy of the pre-built device performance model can be evaluated based on the fitting error of the device performance model. If the fitting error does not exceed the preset error threshold, it means that the model can fit the actual data well and the prediction accuracy is high. If the fitting error exceeds the preset error threshold, it means that the difference between the model and the actual data is large, the prediction accuracy is low, the model health score will be reduced, and it is necessary to collect training data again and train the model to adapt to the new environment.

[0055] S104: According to the prediction accuracy of the equipment performance model and based on multimodal fusion technology, adjust the weights in the pre-built health calculation model.

[0056] In this embodiment, the calculation formula of the health calculation model includes:

[0057] ; Formula (I)

[0058] Where H represents the comprehensive health score of the equipment performance model; i represents the operating parameters of different equipment in the refrigeration room; w i Indicates the weight of the operating coefficient corresponding to the operating parameters of different equipment in the refrigeration room; β iIndicates the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room; T i represents the real-time running data after preprocessing; P i represents the equipment performance prediction data; γ(E) represents the equipment aging weight function; n represents the total number of different equipment operating parameters in the refrigeration room; exp represents the exponential function.

[0059] In this embodiment, the health status of the equipment in the refrigeration room will be affected by many factors, including the passage of time, changes in the external environment, and the aging of the equipment itself. These factors will cause changes in equipment performance, which in turn affects the prediction accuracy of the pre-built equipment performance model. This application uses multimodal fusion technology to dynamically adjust the weights of each item in the pre-built health calculation model in a timely manner according to actual conditions to ensure the accuracy and reliability of the model health calculation.

[0060] In this embodiment, according to the prediction accuracy of the equipment performance model and based on the multimodal fusion technology, the method of adjusting the weights in the pre-built health calculation model includes: based on the multimodal fusion technology, adjusting the operating coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, and the equipment aging weight function.

[0061] In this embodiment, the refrigeration room contains many devices, such as chillers, water pumps, cooling towers, etc. Their operating parameters (such as temperature, pressure, flow, etc.) have a significant impact on system performance. The chillers, water pumps, cooling towers and other equipment in the refrigeration room each have sensor monitoring data in multiple dimensions. These data can be divided into:

[0062] (1) Physical dimension: real-time operating data of the equipment, such as temperature, pressure, flow, etc.

[0063] (2) Model dimension: predictive data output by the equipment performance model, including energy efficiency prediction, flow prediction, etc.

[0064] (3) Historical dimension: historical operating data of equipment, including equipment aging, failure rate, maintenance records, etc.

[0065] In this embodiment, the multimodal fusion technology can fuse data from different sensors and monitoring systems to comprehensively evaluate the current operating status and future health status of the equipment, enriching the input dimensions of health. Subsequently, based on the prediction accuracy of the equipment performance model and using multimodal fusion technology, the impact of different input dimensions on health is dynamically adjusted. If a parameter has a greater impact on the model prediction accuracy, its weight is increased accordingly; otherwise, the weight is reduced to achieve adaptive model health assessment.

[0066] In this embodiment, in the multimodal fusion technology, commonly used fusion strategies include but are not limited to: weighted average fusion algorithm, deep learning fusion algorithm, inter-modal error adjustment fusion algorithm, etc.

[0067] In this embodiment, the weighted average fusion algorithm assigns different weights to each dimension data according to their reliability and importance, and then calculates the weighted average as the final fusion result. The weight allocation is usually based on the accuracy and stability of the data and their contribution to the equipment health assessment. By calculating the weighted average, a health assessment result that integrates multiple dimensional information can be obtained.

[0068] In this embodiment, the deep learning fusion algorithm uses a deep learning model (such as a neural network) to automatically learn and extract features from multi-dimensional data and fuse them. The deep learning model can extract useful features from the original data. The extracted features can be fused at different levels of the model to generate a comprehensive health assessment. The deep learning model needs to be trained to optimize its parameters to improve the accuracy of the health assessment.

[0069] In this embodiment, the inter-modal error adjustment fusion algorithm aims to improve the accuracy of the fusion result by adjusting the errors between different dimensional data. First, it is necessary to analyze the source and size of the errors between different dimensional data. According to the error analysis results, the different dimensional data are appropriately adjusted to reduce the errors between them. The dimensional data after error adjustment can be fused to generate more accurate health assessment results.

[0070] In this embodiment, the historical operation data includes one or more of operation parameter data, equipment aging data, and operation environment data; the manner of adjusting the equipment aging weight function includes:

[0071] (1) Evaluate the aging degree of different equipment in the refrigeration room based on the operating parameter data, equipment aging data, operating environment data, and a pre-built equipment aging model.

[0072] In this embodiment, historical operating parameter data, historical equipment aging data, and historical operating environment data are integrated into one data set, and each data point contains the operating parameters, aging data, and environmental data of the equipment. The integrated data set is divided into a training set and a validation set (or test set), usually in a ratio of 80% training set and 20% validation set. Ensure that the data distribution in the training set and the validation set is similar to avoid overfitting or underfitting of the model. Use the training set data to train the deep learning model, and during the training process, continuously adjust the parameters of the model to minimize the loss function (such as mean square error, cross entropy, etc.). Use the validation set data to monitor the performance of the model to avoid overfitting, so as to construct the equipment aging model, which can comprehensively consider the operating parameters, equipment aging data, and operating environment data to evaluate the degree of aging of the equipment.

[0073] (2) According to the aging degree of different equipment in the refrigeration room, the equipment aging weight function is adjusted using multiple regression analysis technology.

[0074] In this embodiment, according to the aging degree of different equipment in the refrigeration room, the multivariate regression analysis technology is used to calculate the impact of equipment aging on the health of the model, and dynamically adjust the equipment aging weight function in the health calculation model. The multivariate regression analysis technology includes but is not limited to: linear regression algorithm, polynomial regression algorithm, ridge regression algorithm, etc.

[0075] In this embodiment, the linear regression algorithm assumes that there is a linear relationship between the dependent variable and the independent variable, and the value of the dependent variable can be predicted by fitting a straight line or plane (for multiple independent variables). In the linear regression algorithm, each independent variable has a corresponding coefficient, which represents the degree of influence of the independent variable on the dependent variable. The polynomial regression algorithm is an extension of linear regression, which allows the relationship between the independent variable and the dependent variable to be nonlinear. By introducing polynomial terms (such as square, cube, etc.) of the independent variable, polynomial regression can fit more complex curves. This makes polynomial regression more flexible and accurate than linear regression when processing data with nonlinear relationships. The ridge regression algorithm is a linear regression method for dealing with multicollinearity problems, which is to limit the size of the regression coefficient by adding an L2 regularization term (i.e., the sum of the squares of all regression coefficients) to the regression equation, thereby reducing the impact of multicollinearity. By adjusting the parameters of the regularization term (i.e., the ridge parameter), the fit and generalization ability of the model can be balanced.

[0076] In this embodiment, the regression coefficients of each variable are extracted by using multiple regression analysis technology, and these coefficients reflect the influence of each variable (including variables related to equipment aging) on ​​the model health index. Specifically, the model health index of the equipment is used as the dependent variable, and the operating parameter data, equipment aging data (such as service life, number of repairs, etc.) and operating environment data are used as independent variables. Through multiple regression analysis, the regression coefficient of each independent variable can be obtained, and these coefficients represent the specific influence of each variable on the model health. In order to quantify the influence of equipment aging on the model health, a quantitative value reflecting the influence of equipment aging can be obtained by calculating the product of the coefficient of the aging variable and the corresponding data value, and summing or weighted summing these products. Based on this quantitative value, the equipment aging weight function of the equipment aging in the health calculation model is dynamically adjusted. The equipment aging weight function can comprehensively consider the influence of different aging factors on the model health, and give corresponding weights according to the actual situation. For example, if the service life of a certain equipment has a particularly large influence on the model health, then this factor should be given a higher weight during the evaluation.

[0077] It is worth noting that this application uses multimodal fusion technology to effectively fuse multi-dimensional input data (real-time operation data, historical operation data, equipment performance prediction data), so that the weight of each input dimension can be dynamically adjusted according to environmental changes, equipment aging and current operating conditions, to achieve adaptive health assessment of the equipment performance model, and then comprehensively evaluate the current operating status and future health status of the equipment, so that the equipment performance model can provide a more comprehensive and accurate assessment when facing complex industrial systems, which is mainly reflected in the following aspects:

[0078] (1) Multimodal fusion technology can combine real-time operation data collected by different sensors, historical operation data, and equipment performance prediction data output by the model to achieve adaptive health assessment of the equipment performance model, so that the equipment performance model can comprehensively evaluate the current operating status and future health status of the equipment. This multimodal fusion technology does not rely on a single data source or model result, but improves the comprehensiveness and accuracy of the assessment through multimodal feature fusion. For example, when the cooling tower cannot dissipate heat efficiently due to climatic conditions, multimodal fusion technology can readjust the weights in the health calculation model according to different input signals to avoid equipment loss caused by excessive system load.

[0079] (2) Multimodal fusion technology takes health as one of the core dimensions in the state space. Through multimodal data fusion, it enriches the input dimension of health and expands the state space of health. Usually, a single data source can only reflect the local state of the device, while the multimodal fusion algorithm can provide a more comprehensive description of the current health status of the device by integrating multi-dimensional data.

[0080] (3) The multimodal fusion technology can be used to dynamically adjust the impact of different input dimensions on the health calculation model. Specifically, the weights of each input dimension can be dynamically adjusted according to environmental changes, equipment aging and current operating conditions to achieve adaptive health assessment. For example, weight adjustment is performed due to environmental changes. When the external temperature rises, the temperature load of the cooling tower increases. This application will increase the weight of the operating parameters related to the cooling tower to ensure the accuracy of the assessment. Dynamic adjustment of health due to equipment aging. When the equipment enters the later stage of use, the weight of the impact of historical data will be increased to better reflect the impact of aging on the health of the model.

[0081] (4) Using multimodal fusion technology to fuse multi-dimensional input data can effectively reduce the error caused by a single mode. In particular, when the equipment performance model fitting error is large or a sensor fails, multimodal fusion technology can automatically assign more weight to other modes to ensure the stability of the model health assessment. For example, if the data of a certain sensor of the refrigeration machine fluctuates greatly or the model prediction deviates greatly from the actual, multimodal fusion technology can automatically reduce the weight of the mode, thereby reducing the impact of the error signal on the model health assessment, thereby identifying abnormal data and automatically correcting it, realizing automatic adjustment of modal failure, and ensuring the accuracy of the model health assessment results.

[0082] (5) The use of multimodal fusion technology can take into account both long-term and short-term goals in model health assessment. Among them, short-term optimization is to give a positive evaluation when the real-time operation data of the equipment shows that it is in an efficient state, thereby improving short-term health. Long-term balance is to balance the relationship between short-term efficiency and long-term health by introducing equipment aging data and long-term historical data, thereby avoiding excessive consumption of equipment.

[0083] S105: Calculate the comprehensive health score of the equipment performance model according to the health calculation model after adjusting the weight.

[0084] In this embodiment, the comprehensive health score of the equipment performance model is calculated using formula (1) according to the health calculation model after weight adjustment.

[0085] In this embodiment, the method further includes: comparing the comprehensive health score of the equipment performance model with a preset health threshold, and judging whether the equipment performance model needs to be retrained according to the comparison result. If the comprehensive health score of the equipment performance model is higher than or equal to the preset health threshold, it indicates that the equipment performance model is currently in a good health state and does not need to be retrained. If the comprehensive health score of the equipment performance model is lower than the preset health threshold, it indicates that there is a deviation or degradation in the equipment performance model, and the model retraining is automatically triggered to ensure that the equipment performance model is always in the optimal state, providing strong support for the operation optimization of the refrigeration room. Based on the updated equipment performance model, combined with real-time or recent operation data, equipment aging data and operation environment data, the health of the refrigeration room during dynamic operation can be evaluated in real time and accurately to adjust the equipment operation strategy, such as optimizing the start and stop frequency of the chiller or adjusting the wind speed of the cooling tower, optimizing the load distribution and operation time of the equipment, reducing unnecessary energy consumption, and improving energy utilization efficiency. In this way, through continuous monitoring and optimization, the health status problems of the equipment can be identified in advance, maintenance or load adjustment can be carried out in time, the service life of the equipment can be extended, and the refrigeration room can always be kept in an efficient, stable and environmentally friendly operation state.

[0086] In this embodiment, the method further includes: displaying the comprehensive health score and change trend of the equipment performance model through a visual interface, so that the user can intuitively understand the equipment status.

[0087] It is worth noting that the refrigeration room equipment performance model health assessment method of the present application, by monitoring the real-time operation data, historical operation data and equipment performance prediction data generated by the refrigeration room during operation, according to the prediction accuracy of the equipment performance model, and using multimodal fusion technology, combined with the current operating conditions, environmental parameters and equipment aging, dynamically adjusts the weights of different input dimensions in the health calculation model, and realizes the adaptive health assessment of the equipment performance model, so that the model health can be evaluated more comprehensively and accurately, effectively reducing the error of a single data source, and then calculating the most representative equipment performance model. The comprehensive health score. When the comprehensive health score of the equipment performance model is lower than the preset health threshold, the model retraining is automatically triggered, and the model is adaptively adjusted to ensure that the equipment performance model is always in the optimal state, so that the equipment performance model can comprehensively evaluate the current and future operating states of the equipment, so that the health of the refrigeration room in the dynamic operation process can be evaluated in real time and accurately, providing strong support for the operation optimization of the refrigeration room.

[0088] 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.

[0089] 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.

[0090] Figure 4 Schematic diagram of a refrigeration room equipment performance model health evaluation system provided in an embodiment of the present application. Figure 4 As shown, the refrigeration room equipment performance model health evaluation system includes:

[0091] The data collection module 401 is used to collect real-time operation data and historical operation data generated during the operation of the refrigeration room.

[0092] The data preprocessing module 402 is used to preprocess the collected real-time operation data and historical operation data respectively.

[0093] The model accuracy evaluation module 403 is used to evaluate the prediction accuracy of the pre-built equipment performance model based on the pre-processed real-time operation data and the pre-processed historical operation data.

[0094] The weight adjustment module 404 is used to adjust the weights in the pre-built health calculation model according to the prediction accuracy of the equipment performance model and based on multimodal fusion technology.

[0095] The health evaluation module 405 is used to calculate the comprehensive health score of the equipment performance model according to the health calculation model after adjusting the weight.

[0096] In this embodiment, the health assessment module 405 is also used to perform the following steps: compare the comprehensive health score of the equipment performance model with the preset health threshold, and determine whether the equipment performance model needs to be retrained based on the comparison result. If the comprehensive health score of the equipment performance model is higher than or equal to the preset health threshold, it indicates that the equipment performance model is currently in a good health state and does not need to be retrained. If the comprehensive health score of the equipment performance model is lower than the preset health threshold, it indicates that there is a deviation or degradation in the equipment performance model, and the model retraining is automatically triggered to ensure that the equipment performance model is always in the optimal state, providing strong support for the operation optimization of the refrigeration room.

[0097] 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.

[0098] It should also be understood that the division of modules in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present application may be integrated into a processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0099] Figure 5 is a schematic block diagram of an electronic terminal provided in an embodiment of the present application. Figure 5 As shown, the electronic terminal 500 includes: at least one processor 501, a memory 502, at least one network interface 503 and a user interface 505. The various components in the device are coupled together through a bus system 504. It can be understood that the bus system 504 is used to realize the connection and communication between these components. In addition to the data bus, the bus system 504 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, Figure 5 In the specification, various buses are labeled as bus systems.

[0100] The user interface 505 may include a display, a keyboard, a mouse, a trackball, a click gun, keys, buttons, a touch pad or a touch screen.

[0101] It is understood that the memory 502 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.

[0102] The memory 502 in the embodiment of the present invention is used to store various categories of data to support the operation of the electronic terminal 500. Examples of these data include: any executable program for operating on the electronic terminal 500, such as an operating system 5021 and an application 5022; the operating system 5021 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 5022 may include various applications, such as a media player (Media Player), a browser (Browser), etc., for implementing various application services. The implementation of the refrigeration room equipment performance model health assessment method provided in the embodiment of the present invention may be included in the application 5022.

[0103] The method disclosed in the above embodiment of the present invention can be applied to the processor 501, or implemented by the processor 501. The processor 501 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 501 or the instruction in the form of software. The above processor 501 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 501 can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiment of the present invention. The general processor 501 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.

[0104] In an exemplary embodiment, the electronic terminal 500 may be implemented by one or more application specific integrated circuits (ASIC), DSP, programmable logic device (PLD), complex programmable logic device (CPLD) to execute the aforementioned method.

[0105] 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 3 A method according to any one of the embodiments shown.

[0106] 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, and when the program code is run on a computer, the computer executes Figures 1 to 3 A method according to any one of the embodiments shown.

[0107] 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).

[0108] 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.

[0109] 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.

[0110] 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.

[0111] 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.

[0112] 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.

[0113] 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)).

[0114] 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.

[0115] 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.

[0116] In summary, the present application provides a method, system, medium and terminal for evaluating the health of a refrigeration room equipment performance model. By monitoring the real-time operation data, historical operation data and equipment performance prediction data generated by the refrigeration room during operation, according to the prediction accuracy of the equipment performance model, and using multimodal fusion technology, combined with the current operating status, environmental parameters and equipment aging, the weights of different input dimensions in the health calculation model are dynamically adjusted, and the adaptive health evaluation of the equipment performance model is realized, so that the model health can be evaluated more comprehensively and accurately, and the error of a single data source can be effectively reduced, thereby calculating the most representative comprehensive health score of the equipment performance model. When the comprehensive health score of the equipment performance model is lower than the preset health threshold, the model retraining is automatically triggered, and the adaptive adjustment of the model is realized to ensure that the equipment performance model is always in the optimal state, so that the equipment performance model can comprehensively evaluate the current operating state and future operating state of the equipment, so as to be able to evaluate the health of the refrigeration room in the dynamic operation process in real time and accurately, and provide strong support for the operation optimization of the refrigeration room. Therefore, the present application effectively overcomes the various shortcomings in the prior art and has a high industrial utilization value.

[0117] 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 method for evaluating the health of a refrigeration room equipment performance model, characterized in that: include: Collect real-time and historical operation data generated by the refrigeration room during operation; Preprocessing the collected real-time operation data and historical operation data respectively; Evaluating the prediction accuracy of the pre-built equipment performance model based on the pre-processed real-time operation data and the pre-processed historical operation data; According to the prediction accuracy of the equipment performance model and based on multimodal fusion technology, adjusting the weights in the pre-built health calculation model; The calculation formula of the health degree calculation model is include: ； Where H represents the comprehensive health score of the equipment performance model; i represents the operating parameters of different equipment in the refrigeration room; w i Indicates the weight of the operating coefficient corresponding to the operating parameters of different equipment in the refrigeration room; β i Indicates the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room; T i represents the real-time running data after preprocessing; P i represents the equipment performance prediction data; γ(E) represents the equipment aging weight function; n represents the total number of different equipment operating parameters in the refrigeration room; exp represents the exponential function; The comprehensive health score of the equipment performance model is calculated according to the health calculation model after the weight is adjusted.

2. The refrigeration room equipment performance model health evaluation method according to claim 1 is characterized in that: According to the prediction accuracy of the equipment performance model and based on the multimodal fusion technology, the weights in the pre-built health calculation model are adjusted in the following ways: Based on multimodal fusion technology, the operating coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room, and the equipment aging weight function are adjusted respectively.

3. The refrigeration room equipment performance model health evaluation method according to claim 2 is characterized in that: The historical operation data includes one or a combination of operating parameter data, equipment aging data, and operating environment data; The method of adjusting the device aging weight function includes: According to the operating parameter data, equipment aging data, operating environment data, and based on a pre-built equipment aging model, evaluating the aging degree of different equipment in the refrigeration room; According to the aging degree of different equipment in the refrigeration room, the equipment aging weight function is adjusted using multiple regression analysis technology.

4. The refrigeration room equipment performance model health evaluation method according to claim 1 is characterized in that: The method further comprises: The comprehensive health score of the device performance model is compared with a preset health threshold, and it is determined whether the device performance model needs to be retrained based on the comparison result.

5. The refrigeration room equipment performance model health evaluation method according to claim 1 is characterized in that: Ways to evaluate the predictive accuracy of pre-built equipment performance models include: Based on a data offset detection algorithm, the preprocessed real-time operation data is compared with the training data for building the equipment performance model, and an offset value between the preprocessed real-time operation data and the training data is calculated; The prediction accuracy of the pre-built equipment performance model is evaluated according to the offset value between the pre-processed real-time operation data and the training data.

6. The method for evaluating the health of a refrigeration room equipment performance model according to claim 1, characterized in that: Ways to evaluate the predictive accuracy of pre-built equipment performance models include: According to the pre-processed real-time operation data and based on a pre-built equipment performance model, obtaining equipment performance prediction data; Calculating a deviation value between the preprocessed real-time operation data and the equipment performance prediction data; Calculating the fitting error of the equipment performance model based on the deviation value between the preprocessed real-time operation data and the equipment performance prediction data by using a curve fitting technique; The prediction accuracy of the pre-built equipment performance model is evaluated based on the fitting error of the equipment performance model.

7. A refrigeration room equipment performance model health assessment system, characterized in that: include: Data collection module, used to collect real-time operation data and historical operation data generated during the operation of the refrigeration room; A data preprocessing module, used to preprocess the collected real-time operation data and historical operation data respectively; A model accuracy evaluation module, used to evaluate the prediction accuracy of the pre-built equipment performance model based on the pre-processed real-time operation data and the pre-processed historical operation data; A weight adjustment module, used to adjust the weights in a pre-built health calculation model according to the prediction accuracy of the equipment performance model and based on multimodal fusion technology; The calculation formula of the health degree calculation model is include: ； Where H represents the comprehensive health score of the equipment performance model; i represents the operating parameters of different equipment in the refrigeration room; w i Indicates the weight of the operating coefficient corresponding to the operating parameters of different equipment in the refrigeration room; β i Indicates the sensitivity coefficient weights corresponding to the operating parameters of different equipment in the refrigeration room; T i represents the real-time running data after preprocessing; P i represents the equipment performance prediction data; γ(E) represents the equipment aging weight function; n represents the total number of different equipment operating parameters in the refrigeration room; exp represents the exponential function; The health evaluation module is used to calculate the comprehensive health score of the equipment performance model according to the health calculation model after adjusting the weight.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for evaluating the health status of a refrigeration room equipment performance model according to any one of claims 1 to 6 is implemented.

9. An electronic terminal 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 refrigeration room equipment performance model health assessment method according to any one of claims 1 to 6.

Images