Pre-cooling mode cold machine energy consumption prediction model training method and device, equipment and medium

Abstract

The application discloses a chiller energy consumption prediction model training method, device and equipment in a precooling mode, and belongs to the technical field of system energy consumption monitoring. The method comprises the following steps: obtaining sample data of a water cooling system of a data center in a precooling mode; based on the water inlet temperature data of a plate exchanger, the water outlet temperature data of a chiller and the flow data in the sample data, obtaining a comprehensive load rate corresponding to the sample data; inputting the sample data and the comprehensive load rate into a chiller energy consumption prediction model in the precooling mode, and training the energy consumption prediction model in the precooling mode; inputting next sample data and a comprehensive load rate of the water cooling system corresponding to the next sample data into the chiller energy consumption prediction model, and performing next-round iteration training on the chiller energy consumption prediction model until the chiller energy consumption prediction model meets a preset training completion condition. The chiller energy consumption prediction model obtained through the above method is simple and has high precision in predicting the energy consumption of the chiller.

Description

Technical Field

[0001] The present application relates to the technical field of system energy consumption monitoring, and in particular to a method, device, equipment and medium for training a refrigeration machine energy consumption prediction model in a pre-cooling mode. Background Art

[0002] As data centers continue to expand, energy consumption becomes increasingly prominent, with water cooling systems accounting for a significant portion of their energy consumption. To meet data center cooling needs while reducing the energy efficiency of water cooling systems, chillers (chillers) are used in high-temperature environments, such as summer. Plate heat exchangers (plate heat exchangers) are used in lower temperatures, such as winter. During transitional seasons, such as spring and autumn, a pre-cooling mode is employed. In this mode, the plate heat exchanger and chiller are typically connected in series. Plate heat exchangers utilize natural cooling to exchange heat, and the cooling process consumes no electricity, whereas chillers do consume electricity.

[0003] In pre-cooling mode, chiller energy consumption is affected by the following factors: chiller chilled water outlet temperature; chiller load factor; and chiller cooling water inlet temperature. In practice, the chiller chilled water inlet temperature is affected by the plate-cooling unit's chilled water outlet temperature, necessitating the development of two prediction models for chiller chilled water inlet temperature and plate-cooling unit's chilled water outlet temperature. Similarly, the chiller cooling water inlet temperature is affected by the plate-cooling unit's chilled water outlet temperature, necessitating the development of two prediction models for chiller cooling water inlet temperature and plate-cooling unit's chilled water outlet temperature.

[0004] In summary, prior art requires multiple prediction models to calculate chiller energy consumption. This not only increases system complexity and computational burden, but also leads to cumulative errors in each model, ultimately impacting overall system prediction accuracy and energy consumption analysis. This accumulated error can lead to biased energy consumption analysis and affect system optimization decisions. Summary of the Invention

[0005] The present application provides a method, device, equipment and medium for training a refrigeration machine energy consumption prediction model in a pre-cooling mode, which can at least solve the problem of complex refrigeration machine energy consumption prediction process and large errors.

[0006] In order to solve the above technical problems, this application is implemented as follows:

[0007] In a first aspect, an embodiment of the present application provides a method for training a chiller energy consumption prediction model in a pre-cooling mode, the method comprising: obtaining sample data of a water cooling system of a data center in the pre-cooling mode, wherein the water cooling system comprises a chiller and a plate exchanger, and the sample data comprises: water inlet temperature data of the plate exchanger, water outlet temperature data of the chiller, and flow data; based on the water inlet temperature data of the plate exchanger, water outlet temperature data of the chiller, and flow data in the sample data, obtaining a comprehensive load rate corresponding to the sample data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate exchanger; inputting the sample data and the comprehensive load rate into the chiller energy consumption prediction model in the pre-cooling mode, and training the energy consumption prediction model in the pre-cooling mode, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the power of the chiller in the pre-cooling mode; inputting the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data into the chiller energy consumption prediction model, and performing the next round of iterative training on the chiller energy consumption prediction model until the chiller energy consumption prediction model meets the preset training completion conditions.

[0008] In a second aspect, an embodiment of the present application provides an energy-saving method based on a water cooling system, the method comprising: obtaining target data in a water cooling system of a data center in a pre-cooling mode, wherein the water cooling system comprises a chiller and a plate exchanger, and the target data comprises: inlet water temperature data of the plate exchanger, outlet water temperature data of the chiller, and flow data; based on the inlet water temperature data of the plate exchanger, outlet water temperature data of the chiller, and flow data in the target data, obtaining a comprehensive load rate of the water cooling system, wherein the comprehensive load rate is the overall load rate of the chiller and the plate exchanger; obtaining the predicted power of the chiller in the pre-cooling mode by inputting the target data and the comprehensive load rate into a pre-trained chiller energy consumption prediction model, wherein the chiller energy consumption prediction model is trained by the method described in the first aspect above; performing energy-saving analysis on the chiller and the plate exchanger based on the predicted power of the chiller, wherein the plate exchanger does not consume energy.

[0009] In a third aspect, an embodiment of the present application provides a training device for a refrigeration machine energy consumption prediction model in a pre-cooling mode, the device comprising: The first acquisition module is used to obtain sample data of the water cooling system of the data center in the pre-cooling mode, wherein the water cooling system includes a chiller and a plate heat exchanger, and the sample data includes: the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow data; the second acquisition module is used to obtain the comprehensive load rate corresponding to the sample data based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow data in the sample data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate heat exchanger; the first training module is used to input the sample data and the comprehensive load rate into the chiller energy consumption prediction model in the pre-cooling mode, and train the energy consumption prediction model in the pre-cooling mode, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the power of the chiller in the pre-cooling mode; the second training module is used to input the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data into the chiller energy consumption prediction model, and perform the next round of iterative training on the chiller energy consumption prediction model until the chiller energy consumption prediction model meets the preset training completion conditions.

[0010] In a fourth aspect, an embodiment of the present application provides an energy-saving device based on a water cooling system, the device comprising: a target data acquisition module for acquiring target data in a water cooling system of a data center in a pre-cooling mode, wherein the water cooling system comprises a chiller and a plate exchanger, and the target data comprises: water inlet temperature data of the plate exchanger, water outlet temperature data of the chiller, and flow data; a load rate acquisition module for acquiring a comprehensive load rate of the water cooling system based on the water inlet temperature data of the plate exchanger, water outlet temperature data of the chiller, and flow data in the target data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate exchanger; a power acquisition module for obtaining the predicted power of the chiller in the pre-cooling mode by inputting the target data and the comprehensive load rate into a pre-trained chiller energy consumption prediction model, wherein the chiller energy consumption prediction model is trained by the method described in the first aspect above; an analysis module for performing energy-saving analysis on the chiller and the plate exchanger based on the predicted power of the chiller, wherein the plate exchanger does not consume energy.

[0011] In a fifth aspect, an embodiment of the present application provides an electronic device comprising a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the method described in the first or second aspect above are implemented.

[0012] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first or second aspect above are implemented.

[0013] In the seventh aspect, an embodiment of the present application provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer performs the steps of the method described in the first or second aspect above.

[0014] The technical solution provided by this application may have the following beneficial effects:

[0015] In an embodiment of the present application, sample data of a water cooling system of a data center in a pre-cooling mode is obtained, wherein the water cooling system includes a chiller and a plate heat exchanger, and based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow data in the sample data, the comprehensive load rate corresponding to the sample data is obtained, and then the sample data and the comprehensive load rate are input into the chiller energy consumption prediction model in the pre-cooling mode, and the energy consumption prediction model in the pre-cooling mode is trained, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the power of the chiller in the pre-cooling mode; the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data are input into the chiller energy consumption prediction model, and the chiller energy consumption prediction model is iteratively trained for the next round until the chiller energy consumption prediction model meets the preset training completion conditions. In this way, the chiller and the plate heat exchanger are regarded as a whole, and the chiller energy consumption prediction model in the pre-cooling mode is trained. There is no need to predict various temperature parameters of the chiller. This is not only simple and convenient, but also can reduce the error in predicting the chiller energy consumption and improve the prediction accuracy.

[0016] In the embodiments of the present application, it should be understood that the above general description and the following detailed description are merely exemplary and explanatory and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0018] Figure 1 A flow chart of a method for training a chiller energy consumption prediction model in a pre-cooling mode provided in an embodiment of the present application is shown;

[0019] Figure 2 A schematic diagram of a process for an energy-saving method based on a water cooling system provided in an embodiment of the present application is shown;

[0020] Figure 3 A schematic diagram of the structure of a refrigeration machine energy consumption prediction model training device in a pre-cooling mode provided by an embodiment of the present application is shown;

[0021] Figure 4 A schematic structural diagram of an energy-saving device based on a water cooling system provided in an embodiment of the present application is shown;

[0022] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown;

[0023] Figure 6 A schematic structural diagram of another electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0024] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0025] As data centers continue to expand, energy consumption becomes increasingly prominent, with cooling systems accounting for a significant portion of their energy consumption. Traditional data center cooling systems primarily rely on chillers (chillers). While these systems can meet data center cooling needs, they operate inefficiently under partial load or low ambient temperatures, resulting in significant energy waste. To improve the energy efficiency of cooling systems, pre-cooling has been introduced. This system utilizes natural cooling sources (such as low-temperature outdoor air or chilled water) to pre-cool the data center, reducing chiller operating time and energy consumption.

[0026] However, these pre-cooling models still face challenges in practical application. In particular, the coordinated control strategy for the chiller and plate heat exchanger (PHE) has not yet reached optimality, limiting improvements in overall system energy efficiency. Therefore, in-depth analysis of the combined energy consumption of the chiller and PHE under different control strategies is crucial. This analysis not only reveals the energy consumption distribution and key influencing factors during system operation but also provides a scientific basis for optimizing control strategies, thereby maximizing the overall energy efficiency of the refrigeration system and achieving more efficient energy utilization.

[0027] In the pre-cooling mode of a data center's water-cooled cooling system, a plate heat exchanger (PHE) and chiller are typically connected in series. The PHE utilizes a natural cooling source for heat exchange, consuming no electricity during the cooling process, whereas the chiller consumes electricity during the cooling process. The PHE's heat transfer efficiency significantly impacts the chiller's energy consumption. Key indicators considered during the pre-cooling process include:

[0028] Plate replacement related indicators:

[0029] (1) Water outlet temperature on the chiller side

[0030] (2) Inlet water temperature of plate-cooled chiller

[0031] (3) Water outlet temperature on the plate exchange cooling side

[0032] (4) Inlet water temperature of plate exchanger cooling side

[0033] Refrigeration related indicators:

[0034] (1) Chiller chilled water outlet temperature

[0035] (2) Chiller chilled water inlet temperature

[0036] (3) Cooling water inlet temperature of chiller

[0037] (4) Cooling water outlet temperature of chiller

[0038] In addition, the instantaneous flow rate of chilled water outlet and the instantaneous flow rate of cooling water outlet of the chiller will also affect the performance of the plate exchanger and the chiller.

[0039] When analyzing the combined energy consumption of chillers and plate heat exchangers, the focus is on the chiller's energy consumption, as plate heat exchangers do not consume energy. Chiller energy consumption is affected by the following factors:

[0040] (1) Chiller chilled water outlet temperature

[0041] (2) Refrigeration machine load rate

[0042] (3) Cooling water inlet temperature of chiller

[0043] The calculation formula of the cooling machine load rate is:

[0044] Chiller load factor = (Chiller chilled water inlet temperature - Chiller chilled water outlet temperature) * Chiller outlet instantaneous flow rate * Specific heat capacity / Chiller rated cooling capacity

[0045] In pre-cooling mode, the chiller's chilled water inlet temperature is affected by the plate-cooling side's chilled water outlet temperature, so two prediction models are needed for the chiller's chilled water inlet temperature and the plate-cooling side's chilled water outlet temperature. Similarly, the chiller's cooling water inlet temperature is affected by the plate-cooling side's cooling water outlet temperature, so two prediction models are needed for the chiller's cooling water inlet temperature and the plate-cooling side's cooling water outlet temperature.

[0046] In summary, the related chiller energy consumption prediction has the following problems:

[0047] (1) Numerous indicators: The relevant scheme involves multiple key indicators, including multiple temperature parameters and flow parameters of the plate exchanger and chiller. The increase of these indicators significantly increases the complexity of the system.

[0048] (2) Multiple prediction models: To indirectly obtain the chiller's chilled water inlet temperature and chiller's cooling water inlet temperature, it is necessary to establish multiple prediction models for the plate exchanger's chilled water outlet temperature, plate exchanger's cooling water outlet temperature, chiller's chilled water inlet temperature, and chiller's cooling water inlet temperature. The use of multiple prediction models increases the system's complexity and computational burden.

[0049] (3) Model error accumulation: Due to the existence of multiple prediction models, the errors of each model may accumulate, ultimately affecting the prediction accuracy and energy consumption analysis effect of the entire system. The accumulation of errors may lead to deviations in energy consumption analysis and affect the optimization decision of the system.

[0050] (4) Complex logic: The relevant solutions are relatively complex in logic, involving the connection of multiple steps and models. This complexity not only increases the difficulty of system design and maintenance, but also may make it difficult to respond and adjust quickly in actual applications.

[0051] In response to the problems in the related art, the embodiment of the present application proposes a refrigeration energy consumption prediction solution to solve at least one of the above-mentioned technical problems. The technical solution provided by the embodiment of the present application is described below in conjunction with the accompanying drawings.

[0052] Figure 1 The flowchart of a method for training a chiller energy consumption prediction model in a pre-cooling mode provided by an exemplary embodiment of the present application is shown. The method can be executed by an electronic device. The electronic device can be a terminal such as a mobile phone or a computer. Figure 1 As shown, the method mainly includes the following steps:

[0053] S101: Obtain sample data of a water cooling system in a pre-cooling mode of a data center.

[0054] The water cooling system includes a chiller and a plate heat exchanger, and the sample data includes: water inlet temperature data of the plate heat exchanger, water outlet temperature data of the chiller, and flow rate data.

[0055] In this embodiment, the plate heat exchanger consumes no energy in pre-cooling mode, so the chiller and plate heat exchanger can be considered as a single entity (a water-cooling system) to obtain corresponding sample data. This sample data, including the plate heat exchanger's inlet water temperature, the chiller's outlet water temperature, and flow rate data, can be obtained through measurement or historical record querying, eliminating the need for model prediction. This reduces the number of models and computing resource consumption while also streamlining system operation.

[0056] In actual applications, when collecting data, data such as operating mode, instantaneous chilled water outlet flow rate, instantaneous cooling water outlet flow rate, chiller chilled water outlet temperature, plate exchanger chilling side inlet water temperature, plate exchanger cooling side inlet water temperature, and real power can be collected.

[0057] The collected data is then processed. The operating mode values ​​are 0, 1, and 2, representing the chiller, pre-cooling, and plate exchange modes of the data center water cooling system, respectively. The pre-cooling mode data is filtered out by setting the operating mode to 1. This allows us to obtain sample data for the water cooling system in pre-cooling mode.

[0058] S102: Based on the water inlet temperature data of the plate heat exchanger, the water outlet temperature data of the chiller, and the flow rate data in the sample data, obtaining a comprehensive load rate corresponding to the sample data.

[0059] The comprehensive load rate is the overall load rate of the chiller and the plate exchanger.

[0060] In the embodiment of the present application, in the pre-cooling mode, both the plate heat exchanger and the chiller are working and can be regarded as a whole. Therefore, the comprehensive load rate corresponding to the sample data can be obtained through the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller and the flow data. In this way, the operating status of the water cooling system in the pre-cooling mode can be accurately described.

[0061] S103: Inputting the sample data and the comprehensive load rate into a refrigeration machine energy consumption prediction model in a pre-cooling mode, and training the energy consumption prediction model in the pre-cooling mode.

[0062] The energy consumption prediction model in the pre-cooling mode is used to predict the power of the refrigerator in the pre-cooling mode.

[0063] In an embodiment of the present application, the sample data and the corresponding integrated load rate can be input into a chiller energy consumption prediction model to obtain the chiller power output from the chiller energy consumption prediction model. This allows the chiller power to be predicted using a single chiller energy consumption prediction model, eliminating the need for multiple prediction models. This not only improves model training efficiency but also reduces the error accumulation caused by multiple prediction models.

[0064] To some extent, the predicted power of a chiller can be used to represent the predicted energy consumption of the chiller. In practical applications, when predicting energy consumption, it is sufficient to predict the power without calculating the energy consumption.

[0065] The programming language and tools used to train the chiller energy consumption prediction model are not specifically limited in the embodiments of the present application.

[0066] In an embodiment of the present application, the refrigeration machine energy consumption prediction model can obtain the predicted power of the refrigeration machine based on the comprehensive load rate and sample data. Compared with the related technology that requires multiple prediction models and multiple parameters of the refrigeration machine and plate exchanger, the above-mentioned refrigeration machine energy consumption prediction model can not only reduce the consumption of computing resources and improve operating efficiency, but also reduce the accumulation of model errors, improve prediction accuracy, and ensure the reliability of energy consumption analysis results.

[0067] S104: Input the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data into the refrigeration machine energy consumption prediction model, and perform the next round of iterative training on the refrigeration machine energy consumption prediction model until the refrigeration machine energy consumption prediction model meets the preset training completion conditions.

[0068] In practical applications, the preset training completion condition may be that the corresponding value of the loss function falls within a certain range or reaches a certain number of iterations, which can be determined based on actual conditions and is not specifically limited in the embodiments of this application.

[0069] In an embodiment of the present application, sample data of a water cooling system of a data center in a pre-cooling mode is obtained, wherein the water cooling system includes a chiller and a plate heat exchanger, and based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow data in the sample data, the comprehensive load rate corresponding to the sample data is obtained, and then the sample data and the comprehensive load rate are input into the chiller energy consumption prediction model in the pre-cooling mode, and the energy consumption prediction model in the pre-cooling mode is trained, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the power of the chiller in the pre-cooling mode; the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data are input into the chiller energy consumption prediction model, and the chiller energy consumption prediction model is iteratively trained for the next round until the chiller energy consumption prediction model meets the preset training completion conditions. In this way, the chiller and the plate heat exchanger are regarded as a whole, and the chiller energy consumption prediction model in the pre-cooling mode is trained. There is no need to predict various temperature parameters of the chiller. This is not only simple and convenient, but also can reduce the error in predicting the chiller energy consumption and improve the prediction accuracy.

[0070] In practical applications, the load factor is a parameter that describes the operating status of a device or system, generally referring to the ratio of the actual load carried by the device or system to the rated load. In this embodiment of the present application, the actual load can be obtained based on the sample data, and then the rated load of the sample device can be obtained by querying the chiller information, such as the nameplate. The combined load factor of the sample device can then be obtained based on the actual load and the rated load.

[0071] In an optional implementation, the plate exchanger water inlet temperature data includes: the plate exchanger chiller side water inlet temperature, the chiller water outlet temperature data includes: the chilled water outlet temperature of the chiller, and the flow data includes: the chilled water outlet instantaneous flow rate;

[0072] The obtaining of the comprehensive load rate corresponding to the sample data based on the water inlet temperature data of the plate heat exchanger, the water outlet temperature data of the chiller, and the flow rate data in the sample data includes:

[0073] Step 1021: determining the actual load corresponding to the sample data based on the chilled water inlet temperature of the plate-type heat exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate;

[0074] Step 1022: Obtain a comprehensive load rate corresponding to the sample data based on the actual load and the rated load of the water cooling system, wherein the comprehensive load rate is a ratio of the actual load to the rated load, and the rated load is the sum of the rated cooling capacity of the chiller and the rated cooling capacity of the plate exchanger.

[0075] In the embodiment of the present application, the actual load of the sample device (referring to the actual load borne by the device, system, or object during operation) can be determined based on the chilled water inlet temperature of the plate heat exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate. Then, based on the actual load and the rated load of the water cooling system, the comprehensive load rate corresponding to the sample data can be obtained. The rated cooling capacity of a chiller is the ability of the chiller to remove heat from a confined space per unit time, and can generally be marked on the body of the chiller. The rated cooling capacity of a plate heat exchanger is similar to this. This is not only simple and fast, but also more accurate.

[0076] In an optional implementation, determining the actual load corresponding to the sample data according to the chilled-side water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate includes:

[0077] The actual load corresponding to the sample data is the product of the temperature difference, the instantaneous flow rate of the chilled water outlet and the specific heat capacity, wherein the temperature difference is the difference between the chilled side inlet water temperature of the plate exchanger and the chilled water outlet temperature of the refrigerator.

[0078] For example, in practical applications, the combined load rate of the chiller and plate heat exchanger can be calculated using the following formula:

[0079] The combined load factor of the chiller and plate exchanger = (plate exchanger chilled water inlet temperature - chiller chilled water outlet temperature) * instantaneous chilled water outlet flow rate * specific heat capacity / (rated cooling capacity of the chiller + rated cooling capacity of the plate exchanger).

[0080] The specific heat capacity refers to the specific heat capacity of chilled water, and the specific value can be found by querying. The rated cooling capacity of a chiller is its ability to remove heat from a confined space per unit time, and is typically marked on the chiller body.

[0081] In practical applications, the energy consumption of a chiller is affected by the inlet water temperature of the plate heat exchanger on the cooling side, the chiller's chilled water outlet temperature, the instantaneous flow rate of the cooling water outlet, and the combined load rate of the plate heat exchanger and the chiller. Therefore, the chiller energy consumption prediction model can determine the predicted power of the chiller based on the above four factors.

[0082] In an optional implementation, the plate heat exchanger inlet water temperature data includes: the cooling side inlet water temperature of the plate heat exchanger, the chiller outlet water temperature data includes: the chilled water outlet temperature of the chiller, and the flow data includes: the instantaneous flow rate of the cooling water outlet;

[0083] Inputting the sample data and the comprehensive load rate into a refrigeration machine energy consumption prediction model in the pre-cooling mode to train the refrigeration machine energy consumption prediction model may include:

[0084] Input the cooling side water inlet temperature of the plate heat exchanger, the chilled water outlet temperature of the chiller, the instantaneous flow rate of the cooling water outlet and the comprehensive load rate into the chiller energy consumption prediction model to obtain the predicted power output by the chiller energy consumption prediction model;

[0085] Based on the predicted power and the label power corresponding to the sample data, the fitting parameters of the refrigerator energy consumption prediction model are adjusted.

[0086] In an embodiment of the present application, the energy consumption prediction model of the chiller can obtain the predicted power of the chiller based on the cooling side inlet water temperature of the plate exchanger, the chilled water outlet temperature of the chiller, the instantaneous flow rate of the cooling water outlet and the comprehensive load rate. The specific relationship therein can be determined by continuously adjusting the connection relationship and coefficients between the above-mentioned various factors. Then, the fitting parameters of the chiller energy consumption prediction model can be adjusted based on the predicted power and the label power corresponding to the sample data (which can be the real power or actual power corresponding to the chiller).

[0087] In an optional implementation, the refrigerator power Y in the pre-cooling mode satisfies the following formula:

[0088] Y = a0 + a1 * (X[0] - X[1]) + a2 * ((X[0] - X[1]) ^ 2) + a3 * X[2] +a4 * (X[2]^ 2) + a5 * X[2] * (X[0]- X[1]) * X[3];

[0089] Wherein, X[0] represents the cooling side inlet water temperature of the plate exchanger, X[1] represents the chilled water outlet temperature of the chiller, X[2] represents the comprehensive load rate, X[3] represents the instantaneous flow rate of the cooling water outlet, and a0, a1, a2, a3, a4, a5, a6, a7, b and a8 represent the fitting parameters of the chiller energy consumption prediction model.

[0090] In the embodiments of the present application, the above formula is relatively excellent and can accurately represent a possible implementation method for obtaining the predicted power of the chiller based on the cooling-side water inlet temperature of the plate heat exchanger, the chiller's chilled water outlet temperature, the instantaneous cooling water outlet flow rate, and the overall load rate. Other formulas can also obtain the predicted power of the chiller, but the results may be relatively poor. In practical applications, if Python is used for model training, Scipy (an open source library based on Python, mainly used for scientific computing. It can help users solve various problems by providing a series of algorithms and mathematical tools) in Python can be used to fit the parameters (params) a0, a1, a2, a3, a4, a5, a6, a7, b, and a8. If other programming languages, such as C++ or Java, are used, other tools can also be used to fit the parameters, and the embodiments of the present application do not specifically limit this.

[0091] Using the training set (train) and the test set (test), calculate the determination coefficient R 2 Three metrics are used to evaluate the fit and accuracy of regression models: the mean absolute percentage error (MAPE), and the mean squared error (MSE). In practical applications, the coefficient of determination measures the model's ability to explain the variation in the dependent variable. Its value ranges from 0 to 1, with values ​​closer to 1 indicating a better fit. The mean absolute percentage error (MAPE) is a metric used to assess the accuracy of a predictive model. It is primarily used in regression analysis to measure the relative error between the predicted value and the true value. The mean squared error (MSE) can be used to measure the deviation between the model's predicted value and the actual value.

[0092] In practical applications, the specific performance of the training set (train) and the test set (test) is shown in Table 1. 2 All exceeded 0.95, reaching above 0.97; the mape values ​​were all less than 0.04, with the minimum reaching 0.028; and the mse values ​​were all less than 20, with the minimum reaching 18.35. It can be seen that the model performs very well on both the training and test sets.

[0093] Table 1

[0094]

[0095] Figure 2 The flowchart of an energy-saving method based on a water cooling system provided by an exemplary embodiment of the present application is shown. The method can be executed by an electronic device. The electronic device can be a terminal such as a mobile phone or a computer. Figure 2 As shown, the method mainly includes the following steps:

[0096] S201: Acquire target data of a water cooling system in a data center in a pre-cooling mode.

[0097] The water cooling system includes a chiller and a plate heat exchanger, and the target data includes: water inlet temperature data of the plate heat exchanger, water outlet temperature data of the chiller, and flow rate data.

[0098] In this embodiment, the chiller and plate heat exchanger in pre-cooling mode can be considered as a single entity (a water-cooling system). The target data for this water-cooling system can include the plate heat exchanger's inlet water temperature, the chiller's outlet water temperature, and flow rate data. Compared to related techniques for predicting chiller energy consumption, the required target data is relatively minimal, reducing system complexity and improving efficiency.

[0099] S202: Obtaining a comprehensive load rate of the water cooling system based on the water inlet temperature data of the plate heat exchanger, the water outlet temperature data of the chiller, and the flow rate data in the target data.

[0100] The comprehensive load rate is the overall load rate of the chiller and the plate exchanger.

[0101] In practical applications, the overall load factor of the water cooling system can be determined based on the inlet water temperature data of the plate heat exchanger and the outlet water temperature and flow data of the chiller. This helps determine the operating status of the water cooling system and quickly determine the predicted power of the chiller.

[0102] S203: The predicted power of the refrigerator in the pre-cooling mode is obtained by inputting the target data and the comprehensive load rate into a pre-trained refrigerator energy consumption prediction model.

[0103] The energy consumption prediction model of the refrigeration machine is Figure 1 The refrigeration machine energy consumption prediction model training method in the pre-cooling mode is used for training.

[0104] In the embodiment of the present application, the target data can be input through Figure 1 The chiller energy consumption prediction model obtained by the training method for the chiller energy consumption prediction model in the pre-cooling mode shown in the figure obtains the predicted chiller power in the pre-cooling mode outputted by the chiller energy consumption prediction model. Compared to solutions in related arts, the predicted power obtained by the chiller energy consumption prediction model obtained by this embodiment of the application is less complex and more operational and maintainable. Furthermore, it consumes less computing resources, has higher operational efficiency, and can ensure prediction accuracy.

[0105] S204: performing energy-saving analysis on the chiller and the plate heat exchanger according to the predicted power of the chiller.

[0106] The plate converter does not consume energy.

[0107] In this embodiment of the present application, the predicted power of the chiller indicates its predicted energy consumption, and the plate heat exchanger consumes no energy. Therefore, energy-saving analysis can be performed on the chiller and plate heat exchanger based on the chiller's predicted power. This energy-saving analysis not only reveals the energy consumption distribution and key influencing factors in the operation of the refrigeration system, but also provides a scientific basis for optimizing control strategies, thereby maximizing the overall energy efficiency of the refrigeration system and achieving more efficient energy utilization.

[0108] In an embodiment of the present application, target data for the water cooling system in pre-cooling mode of a data center can be obtained. Based on the plate heat exchanger inlet water temperature data, chiller outlet water temperature data, and flow rate data in the target data, the comprehensive load rate of the water cooling system can be obtained. Then, by inputting the target data and the comprehensive load rate into a pre-trained chiller energy consumption prediction model, the predicted power of the chiller in pre-cooling mode can be obtained. Based on the predicted power of the chiller, an energy-saving analysis of the chiller and plate heat exchanger can be performed. This helps optimize the operating efficiency of the cooling system and significantly reduces the overall energy consumption of the data center, thereby promoting the sustainable development of the data center.

[0109] The chiller energy consumption prediction model training method in the pre-cooling mode provided in the embodiments of the present application can be executed by a chiller energy consumption prediction model training device in the pre-cooling mode. In the embodiments of the present application, the chiller energy consumption prediction model training method in the pre-cooling mode is executed by the chiller energy consumption prediction model training device in the pre-cooling mode as an example to illustrate the chiller energy consumption prediction model training method in the pre-cooling mode provided in the embodiments of the present application.

[0110] Figure 3A structure schematic diagram of a pre-cooling mode under cold machine energy consumption prediction model training device provided by an example embodiment of the present application is shown. The pre-cooling mode under cold machine energy consumption prediction model training device can realize all or part of the contents in the example embodiment shown in the present application. The pre-cooling mode under cold machine energy consumption prediction model training device comprises a first acquisition module 301, a second acquisition module 302, a first training module 303 and a second training module 304. Figure 1

[0111] In the embodiment of the present application, the first acquisition module 301 is configured to acquire sample data of a water cooling system in a pre-cooling mode, wherein the water cooling system comprises a cold machine and a plate exchanger, and the sample data comprises water inlet temperature data of the plate exchanger, water outlet temperature data of the cold machine and flow data. The second acquisition module 302 is configured to acquire a comprehensive load rate corresponding to the sample data based on the water inlet temperature data of the plate exchanger, the water outlet temperature data of the cold machine and the flow data in the sample data, wherein the comprehensive load rate is an overall load rate of the cold machine and the plate exchanger. The first training module 303 is configured to input the sample data and the comprehensive load rate into a pre-cooling mode under cold machine energy consumption prediction model, and train the pre-cooling mode under energy consumption prediction model, wherein the pre-cooling mode under energy consumption prediction model is used to predict power of the cold machine in the pre-cooling mode. The second training module 304 is configured to input next sample data and a comprehensive load rate of the water cooling system corresponding to the next sample data into the cold machine energy consumption prediction model, and perform next round of iteration training on the cold machine energy consumption prediction model until the cold machine energy consumption prediction model meets a preset training completion condition.

[0112] In an optional implementation, the water inlet temperature data of the plate exchanger comprises water inlet temperature of a freezing side of the plate exchanger, the water outlet temperature data of the cold machine comprises water outlet temperature of freezing water of the cold machine, and the flow data comprises freezing water outlet instantaneous flow.

[0113] The second acquisition module 302 is configured to acquire the comprehensive load rate corresponding to the sample data based on the water inlet temperature data of the plate exchanger, the water outlet temperature data of the cold machine and the flow data in the sample data, comprising:

[0114] According to the water inlet temperature of the freezing side of the plate exchanger, the water outlet temperature of freezing water of the cold machine and the freezing water outlet instantaneous flow, the actual load corresponding to the sample data is determined.

[0115] According to the actual load and a rated load of the water cooling system, the comprehensive load rate corresponding to the sample data is obtained, wherein the comprehensive load rate is a ratio of the actual load to the rated load, and the rated load is a sum of a rated refrigerating capacity of the cold machine and a rated refrigerating capacity of the plate exchanger.​

[0116] In an optional implementation, the second acquisition module 302 is configured to determine the actual load corresponding to the sample data based on the chilled side water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate, including:

[0117] The actual load corresponding to the sample data is the product of the temperature difference, the instantaneous flow rate of the chilled water outlet and the specific heat capacity, wherein the temperature difference is the difference between the chilled side inlet water temperature of the plate exchanger and the chilled water outlet temperature of the refrigerator.

[0118] In an optional implementation, the plate heat exchanger inlet water temperature data includes: the cooling side inlet water temperature of the plate heat exchanger, the chiller outlet water temperature data includes: the chilled water outlet temperature of the chiller, and the flow data includes: the instantaneous flow rate of the cooling water outlet;

[0119] The first training module 303 is used to input the sample data and the comprehensive load rate into the refrigeration machine energy consumption prediction model in the pre-cooling mode to train the refrigeration machine energy consumption prediction model, including:

[0120] Input the cooling side water inlet temperature of the plate heat exchanger, the chilled water outlet temperature of the chiller, the instantaneous flow rate of the cooling water outlet and the comprehensive load rate into the chiller energy consumption prediction model to obtain the predicted power output by the chiller energy consumption prediction model;

[0121] Based on the predicted power and the label power corresponding to the sample data, the fitting parameters of the refrigerator energy consumption prediction model are adjusted.

[0122] In an optional implementation, the refrigerator power Y in the pre-cooling mode satisfies the following formula:

[0123] Y = a0 + a1 * (X[0] - X[1]) + a2 * ((X[0] - X[1]) ^ 2) + a3 * X[2] +a4 * (X[2]^ 2) + a5 * X[2] * (X[0]- X[1]) * X[3];

[0124] Wherein, X[0] represents the cooling side inlet water temperature of the plate exchanger, X[1] represents the chilled water outlet temperature of the chiller, X[2] represents the comprehensive load rate, X[3] represents the instantaneous flow rate of the cooling water outlet, and a0, a1, a2, a3, a4, a5, a6, a7, b and a8 represent the fitting parameters of the chiller energy consumption prediction model.

[0125] The training device for predicting the energy consumption of a chiller in the pre-cooling mode in the embodiment of the present application can be an electronic device or a component of an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other device other than a terminal. For example, the electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, a mobile internet device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc. It can also be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc., and the embodiment of the present application does not specifically limit it.

[0126] The training device for predicting the energy consumption of a chiller in the pre-cooling mode in the embodiment of the present application can be a device having an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0127] The refrigeration machine energy consumption prediction model training device provided in the pre-cooling mode of the embodiment of the present application can achieve Figure 1 To avoid repetition, the various processes implemented in the method embodiment are not described here.

[0128] The energy-saving method based on a water cooling system provided in the embodiment of the present application can be executed by an energy-saving device based on a water cooling system. In the embodiment of the present application, the energy-saving device based on a water cooling system is used as an example to illustrate the energy-saving method based on a water cooling system.

[0129] Figure 4 The schematic diagram of the structure of the energy-saving device based on the water cooling system provided by an exemplary embodiment of the present application is shown. The energy-saving device based on the water cooling system can achieve the following Figure 2 In the embodiment shown, all or part of the contents, the energy-saving device based on the water cooling system includes: a target data acquisition module 401 , a load rate acquisition module 402 , a power acquisition module 403 and an analysis module 404 .

[0130] In an embodiment of the present application, a target data acquisition module 401 is used to obtain target data in a water cooling system of a data center in a pre-cooling mode, wherein the water cooling system includes a chiller and a plate exchanger, and the target data includes: inlet water temperature data of the plate exchanger, outlet water temperature data of the chiller, and flow data; a load rate acquisition module 402 is used to obtain a comprehensive load rate of the water cooling system based on the inlet water temperature data of the plate exchanger, outlet water temperature data of the chiller, and flow data in the target data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate exchanger; a power acquisition module 403 is used to obtain the predicted power of the chiller in the pre-cooling mode by inputting the target data and the comprehensive load rate into a pre-trained chiller energy consumption prediction model, wherein the chiller energy consumption prediction model is obtained through the above Figure 1 The training method for predicting the energy consumption of the refrigerator in the pre-cooling mode is shown; the analysis module 404 is used to perform energy-saving analysis on the refrigerator and the plate converter according to the predicted power of the refrigerator, wherein the plate converter does not consume energy.

[0131] The energy-saving device based on the water cooling system in the embodiment of the present application can be an electronic device or a component of the electronic device, such as an integrated circuit or chip. The electronic device can be a terminal or other devices other than a terminal. For example, the electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, a mobile Internet device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a personal digital assistant (PDA), etc. It can also be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), a teller machine or a self-service machine, etc., and the embodiment of the present application does not specifically limit it.

[0132] The energy-saving device based on the water cooling system in the embodiment of the present application can be a device having an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0133] The energy-saving device based on the water cooling system provided in the embodiment of the present application can achieve Figure 2To avoid repetition, the various processes implemented in the method embodiment are not described here.

[0134] Alternatively, as Figure 5 As shown, the embodiment of the present application further provides an electronic device 500, including a processor 501 and a memory 502, wherein the memory 502 stores a program or instruction that can be run on the processor 501, and when the program or instruction is executed by the processor 501, the above Figure 1 The training method of the cooling machine energy consumption prediction model in the pre-cooling mode shown or the above Figure 2 The various steps of the energy-saving method based on the water cooling system shown can achieve the same technical effect. To avoid repetition, they will not be repeated here.

[0135] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0136] Figure 6 The following is a block diagram of another electronic device 600 according to an exemplary embodiment of the present application. The electronic device 600 can be implemented as a smartphone, tablet computer, laptop computer, desktop computer, smartwatch, television, etc. The electronic device 600 may also be referred to as user equipment, portable terminal, laptop terminal, desktop terminal, or other names.

[0137] Typically, the electronic device 600 includes a processor 601 and a memory 602 .

[0138] Processor 601 may include one or more processing cores, such as an octa-core processor or a decitalic processor. Processor 601 may be implemented in hardware using at least one of the following: a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), or a PLA (Programmable Logic Array). Processor 601 may also include a main processor and a coprocessor. The main processor is used to process data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 601 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing content required for display. In some embodiments, processor 601 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0139] The memory 602 may include one or more computer-readable storage media, which may be non-transitory. The memory 602 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 602 is used to store at least one instruction, which is used to be executed by the processor 601 to implement all or part of the steps in the chiller energy consumption prediction model training method in the pre-cooling mode or the energy-saving method based on the water cooling system shown in the method embodiment of the present application.

[0140] In some embodiments, electronic device 600 may optionally include a peripheral device interface 603 and at least one peripheral device. Processor 601, memory 602, and peripheral device interface 603 may be connected via a bus or signal lines. Each peripheral device may be connected to peripheral device interface 603 via a bus, signal lines, or circuit boards. Specifically, the peripheral device may include at least one of a radio frequency circuit 604, a display screen 605, a camera assembly 606, an audio circuit 607, and a power supply 608.

[0141] In some embodiments, the electronic device 600 further includes one or more sensors 609 , including but not limited to: an acceleration sensor 610 , a gyroscope sensor 611 , a pressure sensor 612 , an optical sensor 613 , and a proximity sensor 614 .

[0142] Those skilled in the art will understand that Figure 6 The structure shown in the figure does not constitute a limitation on the electronic device 600, and the electronic device 600 may include more or fewer components than shown in the figure, or combine certain components, or adopt a different component arrangement.

[0143] An embodiment of the present application also provides a computer-readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned refrigeration machine energy consumption prediction model training method in the pre-cooling mode or the energy-saving method based on the water cooling system are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0144] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0145] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned refrigeration machine energy consumption prediction model training method in the pre-cooling mode or the energy-saving method based on the water cooling system, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0146] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0147] An embodiment of the present application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the steps of the above-mentioned refrigeration machine energy consumption prediction model training method in the pre-cooling mode or the energy-saving method based on the water cooling system are implemented, and the same technical effect can be achieved. To avoid repetition, they will not be repeated here.

[0148] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered merely as exemplary, and the true scope and spirit of the present application are indicated by the claims.

[0149] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for training a chiller energy consumption prediction model in a pre-cooling mode, characterized in that: include: Obtain sample data of a water cooling system in a pre-cooling mode of a data center, wherein the water cooling system includes a chiller and a plate heat exchanger, and the sample data includes: water inlet temperature data of the plate heat exchanger, water outlet temperature data of the chiller, and flow rate data; Based on the water inlet temperature data of the plate heat exchanger, the water outlet temperature data of the chiller, and the flow rate data in the sample data, obtaining a comprehensive load rate corresponding to the sample data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate heat exchanger; Inputting the sample data and the comprehensive load rate into a refrigerator energy consumption prediction model in a pre-cooling mode, and training the energy consumption prediction model in the pre-cooling mode, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the predicted power of the refrigerator in the pre-cooling mode; Inputting the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data into the refrigeration machine energy consumption prediction model, and performing the next round of iterative training on the refrigeration machine energy consumption prediction model until the refrigeration machine energy consumption prediction model meets the preset training completion condition; The plate exchanger water inlet temperature data includes: the plate exchanger chiller side water inlet temperature; the chiller water outlet temperature data includes: the chilled water outlet temperature of the chiller; the flow rate data includes: the chilled water outlet instantaneous flow rate; The obtaining of the comprehensive load rate corresponding to the sample data based on the water inlet temperature data of the plate heat exchanger, the water outlet temperature data of the chiller, and the flow rate data in the sample data includes: Determining the actual load corresponding to the sample data according to the chilled water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate; Obtaining a comprehensive load rate corresponding to the sample data based on the actual load and the rated load of the water cooling system, wherein the comprehensive load rate is a ratio of the actual load to the rated load, and the rated load is the sum of the rated cooling capacity of the chiller and the rated cooling capacity of the plate heat exchanger; The determining of the actual load corresponding to the sample data according to the chilled water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate includes: The actual load corresponding to the sample data is the product of the temperature difference, the instantaneous flow rate of the chilled water outlet and the specific heat capacity, wherein the temperature difference is the difference between the chilled side inlet water temperature of the plate exchanger and the chilled water outlet temperature of the refrigerator.

2. The method according to claim 1, characterized in that The plate exchanger inlet water temperature data includes: the plate exchanger cooling side inlet water temperature, the chiller outlet water temperature data includes: the chilled water outlet temperature of the chiller, and the flow data includes: the cooling water outlet instantaneous flow rate; Inputting the sample data and the comprehensive load rate into a refrigeration machine energy consumption prediction model in a pre-cooling mode and training the refrigeration machine energy consumption prediction model includes: Input the cooling side water inlet temperature of the plate heat exchanger, the chilled water outlet temperature of the chiller, the instantaneous flow rate of the cooling water outlet and the comprehensive load rate into the chiller energy consumption prediction model to obtain the predicted power output by the chiller energy consumption prediction model; Based on the predicted power and the label power corresponding to the sample data, the fitting parameters of the refrigerator energy consumption prediction model are adjusted.

3. The method according to claim 2, characterized in that The predicted power Y in the pre-cooling mode satisfies the following formula: Y = a0 + a1 * (X[0] - X[1]) + a2 * ((X[0] - X[1]) ^ 2) + a3 * X[2] + a4 *(X[2] ^2) + a5 * - X[1]) * X[3]; Wherein, X[0] represents the cooling side inlet water temperature of the plate exchanger, X[1] represents the chilled water outlet temperature of the chiller, X[2] represents the comprehensive load rate, X[3] represents the instantaneous flow rate of the cooling water outlet, and a0, a1, a2, a3, a4, a5, a6, a7, b and a8 represent the fitting parameters of the chiller energy consumption prediction model.

4. An energy-saving method based on a water cooling system, characterized in that: include: Obtain target data in a pre-cooling mode of a water cooling system in a data center, wherein the water cooling system includes a chiller and a plate heat exchanger, and the target data includes: inlet water temperature data of the plate heat exchanger, outlet water temperature data of the chiller, and flow rate data; Based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow rate data in the target data, obtaining a comprehensive load rate of the water cooling system, wherein the comprehensive load rate is the overall load rate of the chiller and the plate heat exchanger; The predicted power of the refrigerator in the pre-cooling mode is obtained by inputting the target data and the comprehensive load rate into a pre-trained refrigerator energy consumption prediction model, wherein the refrigerator energy consumption prediction model is trained by the method according to any one of claims 1 to 3; An energy-saving analysis is performed on the chiller and the plate heat exchanger according to the predicted power of the chiller, wherein the plate heat exchanger does not consume energy.

5. A training device for a refrigeration machine energy consumption prediction model in a pre-cooling mode, characterized in that: include: A first acquisition module is configured to acquire sample data of a water cooling system of a data center in a pre-cooling mode, wherein the water cooling system includes a chiller and a plate heat exchanger, and the sample data includes: water inlet temperature data of the plate heat exchanger, water outlet temperature data of the chiller, and flow rate data; A second acquisition module is configured to acquire a comprehensive load rate corresponding to the sample data based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow rate data in the sample data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate heat exchanger; a first training module, configured to input the sample data and the comprehensive load rate into a refrigerator energy consumption prediction model in a pre-cooling mode, and train the energy consumption prediction model in the pre-cooling mode, wherein the energy consumption prediction model in the pre-cooling mode is used to predict the predicted power of the refrigerator in the pre-cooling mode; The second training module is used to input the next sample data and the comprehensive load rate of the water cooling system corresponding to the next sample data into the chiller energy consumption prediction model, and perform the next round of iterative training on the chiller energy consumption prediction model until the chiller energy consumption prediction model meets the preset training completion condition; wherein the plate heat exchanger inlet water temperature data includes: the chiller side inlet water temperature of the plate heat exchanger, the chiller outlet water temperature data includes: the chilled water outlet temperature of the chiller, and the flow data includes: the chilled water outlet instantaneous flow rate; The second acquisition module is used to acquire the comprehensive load rate corresponding to the sample data based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow data in the sample data, including: Determining the actual load corresponding to the sample data according to the chilled water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the instantaneous chilled water outlet flow rate; Obtaining a comprehensive load rate corresponding to the sample data based on the actual load and the rated load of the water cooling system, wherein the comprehensive load rate is a ratio of the actual load to the rated load, and the rated load is the sum of the rated cooling capacity of the chiller and the rated cooling capacity of the plate heat exchanger; The second acquisition module is configured to determine the actual load corresponding to the sample data based on the chilled water inlet temperature of the plate exchanger, the chilled water outlet temperature of the chiller, and the chilled water outlet instantaneous flow rate, including: The actual load corresponding to the sample data is the product of the temperature difference, the instantaneous flow rate of the chilled water outlet and the specific heat capacity, wherein the temperature difference is the difference between the chilled side inlet water temperature of the plate exchanger and the chilled water outlet temperature of the refrigerator.

6. An energy-saving device based on a water cooling system, characterized in that: include: A target data acquisition module is configured to acquire target data in a water cooling system of a data center in a pre-cooling mode, wherein the water cooling system includes a chiller and a plate heat exchanger, and the target data includes: water inlet temperature data of the plate heat exchanger, water outlet temperature data of the chiller, and flow rate data; a load rate acquisition module, configured to acquire a comprehensive load rate of the water cooling system based on the inlet water temperature data of the plate heat exchanger, the outlet water temperature data of the chiller, and the flow rate data in the target data, wherein the comprehensive load rate is the overall load rate of the chiller and the plate heat exchanger; a power acquisition module, configured to obtain the predicted power of the refrigerator in a pre-cooling mode by inputting the target data and the comprehensive load rate into a pre-trained refrigerator energy consumption prediction model, wherein the refrigerator energy consumption prediction model is trained by the method according to any one of claims 1 to 3; The analysis module is used to perform energy-saving analysis on the refrigerator and the plate converter according to the predicted power of the refrigerator, wherein the plate converter does not consume energy.

7. An electronic device, characterized in that: The electronic device includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the method according to any one of claims 1 to 4 are implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images