Control Method and System for Wide-Area Optimal Efficiency of Data Center Cold Source System

Abstract

The present application discloses a control method and system for wide-area optimal efficiency of a cold source system in a data center. The method includes: obtaining the energy consumption models of each cooling component in the cooling system, determining the associated variables between the energy consumption models and transforming the expression of the energy consumption models based on the above-mentioned associated variables, then obtaining the fixed components, variable coefficient values and corresponding energy consumption data in the energy consumption models through historical data, then obtaining multiple system energy consumption values by adjusting the values of the associated variables, and finally when the variable function values corresponding to each energy consumption model are a certain value, the system energy consumption value is the lowest. At this time, the input parameters corresponding to the variable function, that is, the associated variables, are inversely analyzed to obtain the control parameters for minimizing the energy consumption of the entire cooling system, thereby minimizing the operating energy consumption of the entire cooling system and effectively reducing the operating cost of the data center.

Description

Technical Field

[0001] The present application belongs to the technical field of system energy consumption optimization control, and relates to a control method and system for wide-area optimal efficiency of a cooling source system in a data center. Background Art

[0002] With the development and popularization of artificial intelligence, big data and cloud computing technologies, data centers have become an important infrastructure in the big data era. Data centers are a set of complex facilities, and their core modules include IT equipment such as servers, storage and switches. The above IT equipment will generate a lot of heat during operation. In order to ensure the stable operation of the data center, current data centers are equipped with corresponding cooling systems.

[0003] The cooling system of a data center usually includes a refrigeration host, a cooling tower, and a water pump for delivering coolant. In the process of calculating the energy consumption of the entire data center, we found that the data center cooling system itself consumes a lot of electricity in the process of cooling the IT equipment inside the data center. According to statistics, the current data center power usage efficiency PUE (Power Usage Effectiveness) is about 1.52, that is, in addition to the energy consumption of the above-mentioned IT equipment itself, the energy consumption of the cooling system heat dissipation will account for 30~40% of the total energy consumption of the data center (the rest of the energy consumption also includes the lighting of the computer room, etc.). It can be seen that how to effectively reduce the energy consumption of the cooling system itself is the key to reducing the operating costs of the entire data center.

[0004] During the operation of the existing cooling system, the chilled water output from the refrigeration host will be pumped into the coolant tank by the water pump, thereby providing a cold source for the precision air conditioner installed in the data room, and the heat generated by the refrigeration host during refrigeration will be taken to the cooling tower by the cooling water for dissipation. The cooling water cooled in the cooling tower will finally be transported back to the refrigeration host, thus realizing the heat dissipation cycle. Usually, based on the operating parameters of each heat dissipation component, such as the chiller, cooling tower, water pump, etc., the energy consumption model of the component can be theoretically obtained, and then the energy consumption of the entire cooling system can be minimized by controlling the operating parameters of each component. However, in actual applications, there are many factors that affect the energy consumption of each cooling component. For example, changes in the wet-bulb temperature of the cooling tower will not only affect its own energy consumption, but also have a linkage effect on the energy consumption of the refrigeration host. If there are multiple refrigeration hosts and cooling towers that can be connected arbitrarily in the entire cooling system, the situation will become more complicated.

[0005] Obviously, it is difficult to achieve the goal of minimizing the energy consumption of the entire cooling system by simply obtaining the optimal system control parameters according to the theoretical model. How to effectively reduce the energy consumption of the data center cooling system as a whole when there are a large number of variable parameters is a problem that needs to be solved urgently. Summary of the invention

[0006] In view of the problem that it is difficult to minimize the overall energy consumption during the control process of the data center cooling system, resulting in high operating costs of the data center, the first object of this application is to provide a control method for the wide-area optimal efficiency of the data center cold source system. By setting the overall goal and then optimizing layer by layer, it can quickly and effectively determine the relevant control parameters of each cooling component in the cooling system, thereby minimizing the operating energy consumption of the entire cooling system and effectively reducing the operating costs of the data center. In addition, to implement the above control method, this application also proposes a control system for the wide-area optimal efficiency of the data center cold source system, and the specific solution is as follows:

[0007] A control method for the wide-area optimal efficiency of a data center cold source system includes:

[0008] Obtain and store the energy consumption models of each cooling component in the cooling system, and determine the model variables and the constraint conditions of each model variable;

[0009] Determine the associated variables and non-associated variables according to the correlation of the model variables among the energy consumption models;

[0010] Fit or transform the expressions of the energy consumption models into the following form:

[0011] P = X * f(T[i]) + K;

[0012] Preset the system energy consumption model according to the energy consumption models of the above cooling components:

[0013] P total = P1 + P2 + … + P n = X1 * f1(T[i]) + X2 * f2(T[i]) + K1 + K2 + … + X n * f n (T[i])+ K n ;

[0014] Among them, P total is the system energy consumption value, X is the variable coefficient composed of non-associated variables in the energy consumption model, f(T[i]) is the variable function composed of associated variables in each energy consumption model, T[i] is a combination of one associated variable or multiple associated variables, and K is the fixed component in each energy consumption model;

[0015] Based on the historical energy consumption data of each cooling component, obtain the variable coefficient and fixed component in the energy consumption model of each cooling component, and combine the preset system energy consumption model to obtain the target variable function value that minimizes the value of P total ;

[0016] Based on the above target variable function value and combined with the variable function formula, inversely analyze to obtain at least one set of target associated variable values corresponding to each variable function value;

[0017] Take the above target associated variable value and target non-associated variable value as the control parameters of the corresponding cooling components in the cooling system and store them.

[0018] Through the above technical solution, first obtain the energy consumption models of each cooling component in the cooling system, determine the associated variables between the energy consumption models and transform the energy consumption models, and then obtain the values of the fixed components and variable coefficients in the energy consumption models and the corresponding energy consumption data through historical data. Thus, when the value of the associated variable is changed, the output values of each energy consumption model included in the system energy consumption model will change dynamically. Finally, when the variable function value corresponding to each energy consumption model is a certain value, the system energy consumption is the lowest. At this time, reverse-analyze to obtain the parameters corresponding to the variable function, that is, the associated variables, and then the control parameters for minimizing the energy consumption of the entire cooling system can be obtained, thereby minimizing the operating energy consumption of the entire cooling system and effectively reducing the operating cost of the data center.

[0019] Further, determine the associated variables and non-associated variables according to the correlation between the model variables of each energy consumption model, including:

[0020] Obtain the historical energy consumption data and model variable data of each cooling component in the cooling system;

[0021] Calculate the correlation coefficients between every two different model variables respectively;

[0022] Set a correlation threshold. If the correlation coefficient between every two model variables is higher than the above correlation threshold, then determine the above model variables as associated variables;

[0023] If there are multiple groups of model variables whose correlation coefficients are higher than the correlation threshold, then use the principal component analysis method to analyze and obtain multiple associated variables.

[0024] Through the above technical solution, the associated variables between each energy consumption model can be accurately determined, making the minimum energy consumption value obtained in the subsequent steps more reliable.

[0025] Further, obtain the target variable function value that minimizes the system energy consumption value P of the preset system energy consumption model based on the historical energy consumption data of each cooling component total including:

[0026] Obtain and generate a reference energy consumption data group reflecting the correlation between the numerical changes of each model variable and the energy consumption change based on the historical energy consumption data of each cooling component during operation, and store the variable coefficient values and fixed component values corresponding to the energy consumption models under each energy consumption data in an associated manner;

[0027] Select a reference energy consumption data, substitute the corresponding variable coefficient value and fixed component value into the preset system energy consumption model, adjust the variable function value to obtain the current system minimum energy consumption value and store it;

[0028] Select the remaining reference energy consumption data and configure variable function values to obtain the corresponding system minimum energy consumption values respectively;

[0029] Compare the obtained multiple system minimum energy consumption values to obtain the system minimum energy consumption and its corresponding target variable function value, target variable coefficient value, and each target non-associated variable value.

[0030] Furthermore, based on each target variable function value, resolve to obtain each target associated variable value corresponding to the output of the variable function of the above variable function values, including:

[0031] Perform energy consumption ranking on multiple system minimum energy consumption values based on the magnitude of the energy consumption values;

[0032] Reverse resolve the corresponding target associated variable value according to the current target variable function value;

[0033] Compare the target associated variable value with the constraint conditions:

[0034] When the target associated variable value obtained by reverse resolution is a group, if the target associated variable value does not meet the constraint conditions, then select the target variable function value corresponding to the next system minimum energy consumption according to the energy consumption ranking, and reverse resolve to obtain the corresponding target associated variable value;

[0035] Repeat the above steps until the target associated variable value meets the constraint conditions;

[0036] When the target associated variable values obtained by reverse resolution are multiple groups, determine whether each group of target associated variable values meets the constraint conditions in turn, and select one group of target associated variable values that meet the constraint conditions;

[0037] If each group of target associated variables does not meet the constraint conditions, then select the target variable function value corresponding to the next system minimum energy consumption according to the energy consumption ranking, and reverse resolve to obtain the corresponding target associated variable value;

[0038] Repeat the above steps until the target associated variable value meets the constraint conditions.

[0039] Through the above technical solutions, it can be ensured that the optimized target associated variable values meet the constraint conditions and can be directly used as system control parameters in system control.

[0040] Furthermore, obtain the energy consumption models of each cooling component in the cooling system, including:

[0041] Establish a theoretical model based on the input and output parameters of each cooling model, and / or generate by fitting through a regression algorithm based on the historical energy consumption data of each cooling model combined with the input and output parameters.

[0042] Through the above technical solution, the energy consumption models of each cooling component can be obtained flexibly and accurately according to the different properties of each cooling component.

[0043] Further, the cooling component includes at least one refrigeration host, a water pump, and a cooling tower;

[0044] The preset system energy consumption model is configured as:

[0045] P total =(A1·P ch1 + A2·P ch2 +…+A n ·P chn )+ (B1·P pump1 + B2·P pump2 +…+B n ·P pumpn )+(C1·P ct1 + C2·P ct2 +…+C n ·P ctn );

[0046] Among them, P total is the system energy consumption value, and P ch is the energy consumption of the refrigeration host, which is configured as:

[0047] ;

[0048] The above Q cooling is the cooling load (kW) of the refrigeration host;

[0049] P pump is the energy consumption of the water pump, which is configured as:

[0050] ;

[0051] P ct is the energy consumption of the cooling tower, which is configured as:

[0052] ;

[0053] In the preset system energy consumption model, A, B, and C are respectively used to represent the states of each refrigeration host, water pump, and cooling tower connected to the system, and their values are defined as 0 or 1 according to the connection relationship and operating state of each cooling component;

[0054] In the refrigeration host energy consumption model, COP(T cw,in ,T cw,out ,m cw ) is the performance coefficient of the refrigeration host, which can be determined by the cooling water inlet temperature T cw,in , the cooling water outlet temperature T cw,out , and the cooling water mass flow rate mcw Obtained by fitting:

[0055] ;

[0056] In the water pump energy consumption model, m cw is the mass flow rate of cooling water (kg / s), g is the acceleration due to gravity (m / s 2 ), H is the water pump head (m), η pump is the water pump efficiency;

[0057] In the cooling tower energy consumption model, ;

[0058] K fan is the fan energy consumption coefficient;

[0059] ΔT approach is the approach temperature of the cooling tower (°C), m and n are empirical coefficients, usually taken as 2 - 3, P pump,ct is the energy consumption of the internal water pump of the cooling tower, and its energy consumption model is the same as that of the water pump for transporting cooling water in the system. Q ct is the heat dissipation requirement of the cooling tower, Q ct,ref is the reference heat load (kW), T wb is the wet bulb temperature of the chiller, T cw,out is the outlet temperature of the cooling water, T cw,in is the inlet temperature of the cooling water, m cw is the mass flow rate of the cooling water.

[0060] Furthermore, the associated variables are configured as the cooling tower load Q ct , the wet bulb temperature T wb of the chiller, the outlet temperature T cw,out of the cooling water, the inlet temperature T cw,in of the cooling water, and the mass flow rate m cw of the cooling water.

[0061] Based on the above control method for the wide - area optimal efficiency of the data center cold source system, the present application also discloses a control system for the wide - area optimal efficiency of the data center cold source system, including:

[0062] A controller, configured to be connected to each cooling component for control, receive control parameters and convert them into corresponding control signals for controlling the actions of each cooling component;

[0063] A data memory, configured to be connected to each cooling component for data, and used to receive and store the operation data of each cooling component, the energy consumption models of each cooling component, and the constraint conditions of each model variable;

[0064] A data processor, configured to be loaded with a computer-readable program for implementing a control method for achieving wide-area optimal efficiency of the cold source system of the data center, is data-connected to the data memory and each cooling component, and is used to obtain the real-time operation data of each cooling component, and parse and obtain the control parameters of each cooling component based on the computer-readable program, and output them to the controller.

[0065] The beneficial effects of this application are as follows:

[0066] First, accurately obtain the energy consumption models of each cooling component in the cooling system according to the requirements, determine the associated variables between the energy consumption models and transform the energy consumption models. Then, obtain the values of the fixed components and variable coefficients in the energy consumption models and the corresponding energy consumption data through historical data. Thus, when the value of the associated variable is changed, the output values of each energy consumption model included in the system energy consumption model will change dynamically. Finally, when the variable function values corresponding to each energy consumption model are a certain value, the system energy consumption is the lowest, and the associated variables corresponding to the variable function are inversely parsed to obtain the control parameters for achieving the lowest energy consumption of the entire cooling system. Overall, by using the method of setting the overall goal first and then optimizing layer by layer, the operating energy consumption of the entire cooling system is reduced to the lowest, effectively reducing the operating cost of the data center. Description of the Drawings

[0067] Figure 1 It is a schematic connection structure diagram of the cooling tower and the refrigeration host;

[0068] Figure 2 It is a schematic diagram of the control method of this application;

[0069] Figure 3 It is a schematic diagram of the method for obtaining the values of each associated variable. Detailed Embodiments

[0070] The following details the embodiments of this application, and the examples of the embodiments are shown in the appendix Figures 1-3 shown.

[0071] In the description of this specification, the description referring to the terms "certain embodiments", "one embodiment", "some embodiments", "schematic embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of this application. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0072] As Figure 1 shown, the data center cooling system mainly includes at least one refrigeration host, a water pump, and a cooling tower. For simplicity of illustration,Figure 1 Only the case where a refrigeration main unit is connected to a cooling tower is schematically shown. In practical applications, in order to meet the cooling requirements of the data center and improve the reliability of the system at the same time, the number of the above cooling components is usually configured as multiple. When in use, different numbers of cooling components can be connected to the cooling system as needed.

[0073] Since there are multiple associated energy consumption influencing factors among the cooling components, for example, the cooling water temperature output by the cooling tower is not only related to the energy consumption of the cooling tower itself, but also associated with the energy consumption of the refrigeration main unit. Therefore, adjusting the control parameter of any cooling component may have a linkage effect on the energy consumption of the entire cooling system. For this reason, the embodiment of the present application provides a control method for the wide-area optimal efficiency of a data center cold source system to find the control parameters that minimize the overall energy consumption of the entire cooling system, such as Figure 2 shown, mainly including the following steps:

[0074] S100, obtain and store the energy consumption models of the cooling components in the cooling system, and determine the model variables and the constraint conditions of each model variable;

[0075] S200, determine the associated variables and non-associated variables according to the correlation of the model variables among the energy consumption models;

[0076] S300, fit or transform the expressions of the energy consumption models into the following form:

[0077] P = X * f(T[i]) + K;

[0078] And preset a system energy consumption model according to the energy consumption models of the above cooling components:

[0079] P total = P1 + P2 + … + P n = X1 * f1(T[i]) + X2 * f2(T[i]) + K1 + K2 + … + X n * f n (T[i]) + K n ;

[0080] Among them, P total is the system energy consumption value, X is the variable coefficient composed of non-associated variables in the energy consumption model, is a fixed value or a function, f(T[i]) is the variable function composed of associated variables in each energy consumption model, T[i] is a combination of one associated variable or multiple associated variables, and K is the fixed component in each energy consumption model;

[0081] S400, based on the historical energy consumption data of each cooling component, obtain the variable coefficient and the fixed component in the energy consumption model of each cooling component, and combine the preset system energy consumption model to obtain the target variable function value that minimizes the system energy consumption value P total ;

[0082] S500, based on the above-mentioned objective variable function values and in combination with the variable function formula, inversely resolve to obtain at least one set of target associated variable values corresponding to each variable function value;

[0083] S600, use the above-mentioned target associated variable values and target non-associated variable values as the control parameters of the corresponding cooling components in the cooling system and store them.

[0084] In the above step S100, obtaining the energy consumption models of the cooling components in the cooling system includes:

[0085] S110, establish a theoretical model based on the input and output parameters of each cooling model,

[0086] and / or S120, generate by fitting via a regression algorithm based on the historical energy consumption data of each cooling model in combination with the input and output parameters.

[0087] In the specific implementation manner, the configuration of the energy consumption model is mostly completed by combining the above two sub-steps. For example, for the energy consumption of the water pump in the cooling system, since its theoretical model and the model variables affecting the water pump energy consumption are relatively clear, the method in step S110 can be directly used for modeling, such as:

[0088] .

[0089] In practical applications, the energy consumption of the refrigeration host is usually affected by multiple different types of model variables. Therefore, the energy consumption model of the refrigeration host mostly uses the method in step S120 for modeling, that is, obtained by fitting via a regression algorithm according to the historical energy consumption data. Usually, the above-mentioned fitted energy consumption model also needs to go through parameter optimization and verification steps. Obviously, by combining the two ways of theoretical model modeling and data fitting modeling, the energy consumption models of each cooling component can be obtained flexibly and accurately according to the different properties of each cooling component, providing a basis for the subsequent calculation of minimizing the energy consumption of the entire cooling system.

[0090] After establishing the energy consumption models of each cooling component, it is also necessary to limit the value ranges of each model variable so that the obtained model variables can be directly applied to the cooling system in the later stage. For this reason, step S100 also includes the confirmation of model variables and the determination steps of the constraint conditions of each model variable. For the constraint conditions of each model variable, they can be obtained from the control parameter descriptions provided by the manufacturers of each cooling component. For example, the maximum lift and rated power of the water pump, the operating range of the compressor in the refrigeration host, etc. They can also be obtained by analyzing the historical energy consumption data of each cooling component. For example, find the maximum and minimum values of the cooling water temperature during the normal operation of the cooling tower in the historical energy consumption data, and thus obtain the limited interval of the cooling water temperature. Finally, the above-mentioned constraint conditions can also be determined by theoretical derivation. For example, the cooling water mass flow rate:

[0091] .

[0092] In step S200, the associated variables and the non-associated variables are determined according to the correlation of the model variables between the energy consumption models, further comprising:

[0093] S210, obtaining historical energy consumption data and model variable data of each cooling component in the cooling system;

[0094] S220, respectively calculating correlation coefficients between different model variables. In practical applications, the Pearson correlation coefficient method can be used for calculation, and the specific process will not be repeated here.

[0095] S230, setting a correlation threshold, if the correlation coefficient between two model variables is higher than the correlation threshold, then determining the model variables as associated variables;

[0096] S240, if there are multiple groups of model variables whose correlation coefficients are higher than the correlation threshold, principal component analysis is used to obtain multiple interrelated associated variables, for example, the three model variables of the cooling water inlet temperature, cooling water outlet temperature and cooling water mass flow rate of the refrigeration host are used as associated variables.

[0097] Through the above technical solution, the associated variables between the various energy consumption models can be accurately determined, so that the minimum energy consumption value obtained in the subsequent steps is more reliable.

[0098] In the implementation of the method of the present application, in order to facilitate the calculation and optimization of the subsequent cooling system energy consumption data, the above-mentioned associated variables are configured as cooling tower load, cooler wet bulb temperature T wb , Cooling water outlet temperature T cw,out , Cooling water inlet temperature T cw,in , cooling water mass flow m cw .

[0099] In step S300, in order to facilitate the numerical optimization of model variables and highlight the linkage effect of changes in associated variables on the energy consumption of the entire cooling system, the expressions of each energy consumption model are first fitted or transformed into the following form: P = X*f(T[i])+K, where X is the variable coefficient composed of non-associated variables in the energy consumption model, f(T[i]) is the variable function composed of associated variables in each energy consumption model, T[i] is a combination of one or more associated variables, and K is a fixed component in each energy consumption model. For example, the energy consumption model of the water pump is converted into the following form:

[0100]

[0101] in It can be regarded as a variable function f(m cw) coefficient. Similarly, the energy consumption model of the refrigeration host and cooling tower is converted or fitted.

[0102] Then, the system energy consumption model is preset according to the energy consumption models of the above cooling components:

[0103] P total =P1+P2+…+P n = X1*f1(T[i])+X2*f2(T[i])+K1+K2+…+X n *f n (T[i])+K n .

[0104] From the configuration method of the aforementioned associated variables and non-associated variables, it can be known that the variable coefficient value X in the preset system energy consumption model is basically stable and will not fluctuate greatly with the change of the value of the associated variable T[i]. However, the function values ​​of f1 (T[i]) and f2 (T[i]) will change in conjunction with the change of the value of T[i]. Therefore, to obtain the minimum value of the above-mentioned preset system energy consumption model, it is only necessary to change the value of T[i] to obtain a set of variable function values. In the implementation manner of the present application, a set of variable function values ​​is first set to obtain, and then the functional formula of the variable function is combined with reverse analysis to obtain the values ​​of each associated variable. To this end, combined with Figure 3 As shown, the purpose of step S400 is to obtain the values ​​of each associated variable.

[0105] In detail, step S400 further includes:

[0106] S410, acquiring and generating a reference energy consumption data group based on the historical energy consumption data of each cooling component operation to reflect the correlation between the value change of each model variable and the energy consumption change, and storing the variable coefficient value and the fixed component value corresponding to the energy consumption model under each energy consumption data;

[0107] S420, selecting a reference energy consumption data, substituting the corresponding variable coefficient value and fixed component value into the preset system energy consumption model, adjusting the variable function value to obtain the current system minimum energy consumption value and storing it;

[0108] S430, selecting the remaining reference energy consumption data and configuring the variable function value to obtain the corresponding system minimum energy consumption value;

[0109] S440, comparing multiple system minimum energy consumption values ​​to obtain the system minimum energy consumption and its corresponding target variable function value, target variable coefficient value and each target non-associated variable value.

[0110] In detail, in step S410, the reference energy consumption data set may select different data within a specific range as reference to simplify the data.

[0111] In step S500, based on the values of each objective variable function, resolving to obtain the values of each objective associated variable corresponding to the output of each variable function by the variable function further includes:

[0112] S510, performing energy consumption ranking on the minimum energy consumption values of multiple systems based on the magnitudes of the energy consumption values;

[0113] S520, resolving in reverse according to the current objective variable function value to obtain the corresponding objective associated variable value;

[0114] S530, comparing the objective associated variable value with the constraint conditions:

[0115] S5301, when the objective associated variable value obtained by reverse resolution is a set, if the objective associated variable value does not satisfy the constraint conditions, then select the objective variable function value corresponding to the next lowest energy consumption of the system according to the energy consumption ranking, and resolve in reverse to obtain the corresponding objective associated variable value;

[0116] Repeat the above steps until the objective associated variable value satisfies the constraint conditions;

[0117] S5302, when the objective associated variable values obtained by reverse resolution are multiple sets, sequentially determine whether each set of objective associated variable values satisfies the constraint conditions, and select one set of objective associated variable values that satisfies the constraint conditions;

[0118] If each set of objective associated variables does not satisfy the constraint conditions, then select the objective variable function value corresponding to the next lowest energy consumption of the system according to the energy consumption ranking, and resolve in reverse to obtain the corresponding objective associated variable value;

[0119] Repeat the above steps until the objective associated variable value satisfies the constraint conditions.

[0120] In the embodiment of the present application, the cooling component includes at least one refrigeration host, a water pump, and a cooling tower; for facilitating the optimization calculation of the energy consumption of a cooling system including multiple cooling components, the preset system energy consumption model in the embodiment of the present application is configured as:

[0121] Further, the cooling component includes at least one refrigeration host, a water pump, and a cooling tower;

[0122] The preset system energy consumption model is configured as:

[0123] P total =(A1·P ch1 + A2·P ch2 +…+A n ·P chn )+ (B1·P pump1 + B2·P pump2 +…+B n ·P pumpn)+(C1·P ct1 + C2·P ct2 +…+C n ·P ctn );

[0124] Among them, P total is the system energy consumption value, and P ch is the energy consumption of the chiller, and the configuration is:

[0125] ;

[0126] Q cooling is the cooling load (kW) of the chiller;

[0127] P pump is the energy consumption of the water pump, and the configuration is:

[0128] ;

[0129] P ct is the energy consumption of the cooling tower, and the configuration is:

[0130] ;

[0131] In the preset system energy consumption model, A, B, and C are used to represent the states of each chiller, water pump, and cooling tower connected to the system, and their values are defined as 0 or 1 according to the connection relationship and operating state of each cooling component;

[0132] In the chiller energy consumption model, COP(T cw,in ,T cw,out ,m cw ) is the coefficient of performance of the chiller, which can be obtained by fitting from the inlet cooling water temperature T cw,in , the outlet cooling water temperature T cw,out , and the cooling water mass flow rate m cw :

[0133]

[0134] where a1, a2, a3, a4, a5, and a6 are fitting coefficients.

[0135] In the water pump energy consumption model, m cw is the cooling water mass flow rate (kg / s), g is the acceleration due to gravity (m / s 2 ), H is the water pump head (m), and η pump is the water pump efficiency;

[0136] In the cooling tower energy consumption model:

[0137] , c p is the specific heat capacity of the cooling water;

[0138] K fan is the fan energy consumption coefficient;

[0139] ΔT approach is the approach temperature (°C) of the cooling tower, which is a value set during system design. m and n are empirical coefficients, usually taken as 2 - 3, and P pump,ct is the energy consumption of the water pump inside the cooling tower, and its energy consumption model is the same as that of the water pump for transporting cooling water in the system. Q ct is the heat dissipation requirement of the cooling tower. Q ct,ref is the reference heat load (kW), and T wb is the wet bulb temperature of the chiller, and T cw,out is the outlet water temperature of the cooling water, and T cw,in is the inlet water temperature of the cooling water, and m cw is the mass flow rate of the cooling water.

[0140] During the optimization calculation, the associated variables are preferably configured as the cooling tower load, the wet bulb temperature T of the chiller wb , the outlet water temperature T of the cooling water cw,out , the inlet water temperature T of the cooling water cw,in , the mass flow rate m of the cooling water cw .

[0141] Based on the above control method for the wide - area optimal efficiency of the data center cold source system, the embodiments of the present application also disclose a control system for the wide - area optimal efficiency of the data center cold source system, which mainly includes a controller, a data memory, and a data processor.

[0142] The controller is configured to be connected to each cooling component for control, receive control parameters and convert them into corresponding control signals for controlling the actions of each cooling component. In practical applications, the above - mentioned controller can be implemented by a single - chip microcomputer or an FPGA control module.

[0143] The data memory is configured to be connected to each cooling component for data, and is used to receive and store the operation data of each cooling component, the energy consumption models of each cooling component, and the constraint conditions of each model variable, etc. The above - mentioned data memory includes two - level data storage units. Among them, the historical energy consumption data is refined into the form of a reference energy consumption data group and stored in the first - level data storage unit. Similarly, the energy consumption models of each cooling component are also stored in the first - level data storage unit for easy calling by the data processor, while the historical energy consumption data itself and the real - time operation data of each cooling component are preferably stored in the second - level data storage unit.

[0144] The data processor is loaded with a computer-readable program for implementing the control method for achieving the wide-area optimal efficiency of the cold source system of the data center. Multiple data processing units are configured in the above program, such as a function inverse parsing unit for inversely parsing function variables according to a functional formula and a function output value; and a numerical comparison unit for comparing whether the model variables obtained by inverse parsing satisfy the constraint conditions. In a specific implementation process, the data processor is data-connected to the data memory and each cooling component, and is used to obtain the real-time operation data of each cooling component, and based on the computer-readable program, parse and obtain the control parameters of each cooling component, that is, the model variables, and then output them to the controller to generate corresponding control instruction signals.

[0145] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for controlling the wide-area optimal efficiency of a data center cooling system, characterized in that: include: Obtain and store energy consumption models of each cooling component in the cooling system, and determine model variables and constraints of each model variable; Determine the associated variables and the non-associated variables according to the correlation of the model variables among the energy consumption models; Fit or transform the expressions of each energy consumption model into the following forms: P = X*f(T[i])+K; The system energy consumption model is preset according to the energy consumption models of the above cooling components: P total =P1+P2+…+P n = X1*f1(T[i])+X2*f2(T[i])+K1+K2+…+X n *f n (T[i])+K n ; Among them, P total is the system energy consumption value, X is the variable coefficient composed of non-correlated variables in the energy consumption model, f(T[i]) is the variable function composed of correlated variables in each energy consumption model, T[i] is a correlated variable or a combination of multiple correlated variables, and K is the fixed component in each energy consumption model; Based on the historical energy consumption data of each cooling component, the variable coefficient and fixed component in the energy consumption model of each cooling component are obtained, and the system energy consumption value P is obtained by combining the preset system energy consumption model. total Minimum objective variable function value; Based on the above target variable function value and in combination with the variable function formula, reverse analysis is performed to obtain at least one set of target associated variable values ​​corresponding to each variable function value; The above target associated variable values ​​and target non-associated variable values ​​are used as control parameters of corresponding cooling components in the cooling system and stored.

2. The method according to claim 1, characterized in that According to the correlation of model variables among energy consumption models, associated variables and non-associated variables are determined, including: Obtain historical energy consumption data and model variable data of each cooling component in the cooling system; Calculate the correlation coefficients between different model variables respectively; A correlation threshold is set, and if the correlation coefficient between two model variables is higher than the correlation threshold, the model variables are determined to be associated variables; If there are multiple groups of model variables whose correlation coefficients are higher than the correlation threshold, a principal component analysis method is used to obtain multiple associated variables.

3. The method according to claim 1, characterized in that Based on the historical energy consumption data of each cooling component, the system energy consumption value P of the preset system energy consumption model is obtained. total The minimum objective variable function value includes: Obtain and generate a reference energy consumption data group based on the historical energy consumption data of each cooling component operation to reflect the correlation between the value change of each model variable and the energy consumption change, and store the variable coefficient value and the fixed component value corresponding to the energy consumption model under each energy consumption data; Select a reference energy consumption data, substitute the corresponding variable coefficient value and fixed component value into the preset system energy consumption model, adjust the variable function value to obtain the current system minimum energy consumption value and store it; Select the remaining reference energy consumption data and configure the variable function value to obtain the corresponding minimum energy consumption value of the system; By comparing the obtained multiple system minimum energy consumption values, the system minimum energy consumption and its corresponding target variable function value, target variable coefficient value and each target non-associated variable value are obtained.

4. The method according to claim 3, characterized in that Based on the values ​​of each target variable function, the values ​​of each target associated variable corresponding to each variable function outputting the above variable function value are obtained by analysis, including: Sort the energy consumption of multiple systems by their lowest energy consumption values ​​based on their energy consumption values; Reverse analysis is performed based on the current target variable function value to obtain the corresponding target associated variable value; Compare the target-linked variable value to the constraint: When the target associated variable value obtained by reverse analysis is a group, if the target associated variable value does not meet the constraint condition, the target variable function value corresponding to the next system minimum energy consumption is selected according to the energy consumption sorting, and the corresponding target associated variable value is obtained by reverse analysis; Repeat the above steps until the target associated variable value meets the constraint conditions; When there are multiple groups of target associated variable values ​​obtained by reverse parsing, determine in turn whether each group of target associated variable values ​​meets the constraint conditions, and select one group of target associated variable values ​​that meets the constraint conditions; If all groups of target-related variables do not meet the constraints, the target variable function value corresponding to the next system's minimum energy consumption is selected according to the energy consumption ranking, and the corresponding target-related variable value is obtained by reverse analysis; Repeat the above steps until the target associated variable value satisfies the constraint conditions.

5. The method according to claim 1, characterized in that Obtain the energy consumption model of each cooling component in the cooling system, including: Establishing a theoretical model based on the input and output parameters of each cooling model, and / or Based on the historical energy consumption data of each cooling model combined with the input and output parameters, it is generated through regression algorithm fitting.

6. The method according to claim 3, characterized in that The cooling assembly includes at least one refrigeration host, a water pump and a cooling tower; The preset system energy consumption model is configured as follows: P total =(A1·P ch1 + A2·P ch2 +…+A n ·P chn )+ (B1·P pump1 + B2·P pump2 +…+B n ·P pumpn )+(C1·P ct1 + C2·P ct2 +…+C n ·P ctn ); Among them, P total is the energy consumption value of the system, P ch The energy consumption of the cooling unit is: ； The above Q cooling is the cooling load of the refrigeration host; P pump is the energy consumption of the water pump, and the configuration is: ； P ct The energy consumption of the cooling tower is: ； In the preset system energy consumption model, A, B, and C are used to represent the status of each refrigeration host, water pump, and cooling tower connected to the system, and their values ​​are defined as 0 or 1 according to the connection relationship and operating status of each cooling component; In the refrigeration host energy consumption model, COP (T cw,in ,T cw,out ,m cw ) is the performance coefficient of the refrigeration unit, which can be obtained from the cooling water inlet temperature T cw,in , Cooling water outlet temperature T cw,out And cooling water mass flow m cw The fitting result is: ； In the water pump energy consumption model, m cw is the cooling water mass flow rate, g is the acceleration of gravity, H is the pump head, η pump is the pump efficiency; In the cooling tower energy consumption model: , c p is the specific heat capacity of cooling water; K fan is the fan energy consumption coefficient; ΔT approach is the approach temperature of the cooling tower, m and n are empirical coefficients, usually 2~3, P pump,ct is the energy consumption of the water pump inside the cooling tower. Its energy consumption model is the same as the energy consumption model of the water pump that delivers cooling water in the system. ct The heat dissipation demand of the cooling tower, Q ct,ref is the reference heat load, T wb is the cooler wet bulb temperature, T cw,out is the cooling water outlet temperature, T cw,in is the cooling water inlet temperature, m cw is the cooling water mass flow rate.

7. The method according to claim 6, characterized in that The associated variables are configured as cooling tower load, cooler wet bulb temperature T wb , Cooling water outlet temperature T cw,out , Cooling water inlet temperature T cw,in , cooling water mass flow m cw .

8. A control system for wide-area optimal efficiency of a data center cooling system, characterized in that: include: A controller configured to be in control connection with each cooling component, receive control parameters and convert them into corresponding control signals for controlling the actions of each cooling component; A data storage device configured to be data-connected to each cooling component, and used to receive and store operation data of each cooling component, energy consumption model of each cooling component, and constraint conditions of each model variable; A data processor is configured to be loaded with a computer-readable program for implementing a control method for wide-area optimal efficiency of a cooling source system of a data center as described in any one of claims 1 to 7, and is connected to the data storage device and data of each cooling component to obtain real-time operating data of each cooling component, and to parse and obtain control parameters of each cooling component based on the computer-readable program, and output them to a controller.

Images