Method and device for predicting load and temperature field of information technology equipment
By constructing a three-dimensional model and using CFD solver to predict the temperature field of information technology equipment, the problem of temperature field prediction when the load of information technology equipment in the data center is solved, and fast and low-cost prediction is achieved, reducing energy waste.
Patent Information
- Application Number
- CN202411044745.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-08
AI Technical Summary
The prior art is difficult to quickly and at low cost to predict the temperature field when the load changes in information technology equipment, resulting in waste of energy in data centers.
By constructing a three-dimensional model based on design drawings and on-site information of information technology equipment, using the CFD solver to select control equations and discrete methods, predict the temperature field of information technology equipment under different loads, and combining heat transfer knowledge to predict refrigeration Energy consumption of the system.
Fast, accurate and low-cost load and temperature field prediction of information technology equipment is achieved, reducing dependence on temperature sensors, reducing time and cost, and improving the ease of prediction.
Smart Images

Figure CN118917247A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field, and in particular to a method and device for predicting the load and temperature field of information technology equipment. Background Art
[0002] In a data center, when the load of information technology equipment ITE (Information Technology Equipment) changes, it will affect the thermal environment inside the data center. Data centers mainly rely on air cooling for heat dissipation. After ITE inhales cold air and discharges hot air, the heat in the air is discharged to the outside through the refrigeration system. However, the heat exchange and flow characteristics of air are difficult to predict. If the heat transfer process using air as a medium cannot be mastered, it is impossible to effectively design energy-saving optimization solutions for ITE load changes. As a result, data centers need to maintain the refrigeration system at a higher power, resulting in energy waste.
[0003] In the prior art, there is a method of using artificial intelligence to predict the I TE load and the temperature field around it, but it requires the deployment of a large number of temperature sensors and the collection of tens of thousands of hours of data before a certain degree of prediction can be made. This method is too time-consuming, extremely costly, and difficult to implement. Therefore, how to achieve fast, low-cost and easy-to-implement I TE load and temperature field prediction has become an urgent problem to be solved. Summary of the invention
[0004] Based on the above problems, the present application provides a method and device for predicting the load and temperature field of information technology equipment to achieve fast, low-cost and easy-to-implement I TE load and temperature field prediction.
[0005] The present application discloses a method for predicting load and temperature field of information technology equipment, the method comprising:
[0006] Based on the design drawings and on-site information of the information technology equipment I TE, the I TE and the three-dimensional space where the I TE is placed are modeled to obtain a three-dimensional model;
[0007] Selecting the governing equations and discretization method of the three-dimensional model in a computational fluid dynamics (CFD) solver;
[0008] The control equations are discretized using the discretization method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model;
[0009] According to the prediction model, the temperature field of I TE under different loads is predicted.
[0010] Optionally, the modeling of the information technology equipment I TE and the three-dimensional space where the I TE is placed based on the design drawings and on-site information of the information technology equipment I TE includes:
[0011] The three-dimensional space is modeled according to the position, size and power of the I TE, as well as other objects that affect the thermal environment in the three-dimensional space except the I TE; the objects include air-conditioned compartments and raised floors.
[0012] Optionally, the control equation and discretization method of the three-dimensional model are selected in a computational fluid dynamics (CFD) solver, including:
[0013] The mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations, and the control volume method is selected as the discretization method.
[0014] Optionally, the using the discrete method to discretize the control equations respectively includes:
[0015] In the CFD solver, profiles are selected for the unsteady-state term, convection term, diffusion term, and source term in the control volume method respectively;
[0016] In a target control volume and a target time interval, the unsteady-state term, the convection term, the diffusion term and the source term are respectively integrated according to the profile to obtain an algebraic equation at a target node in the three-dimensional model;
[0017] The algebraic equations on all nodes in the three-dimensional model are obtained by iterative calculation and combined to obtain the prediction model.
[0018] Optionally, after predicting the temperature field of the ITE under different loads according to the prediction model, the method further includes:
[0019] Based on the temperature field, the energy consumption of the refrigeration system in the three-dimensional space under different I TE loads is predicted.
[0020] Based on the above-mentioned method for predicting the load and temperature field of information technology equipment, the present application also discloses a device for predicting the load and temperature field of information technology equipment, including: a modeling unit, an equation selection unit, a discrete unit and a prediction unit;
[0021] The modeling unit is used to model the three-dimensional space where the IT equipment I TE and ITE are placed based on the design drawings and on-site information of the IT equipment I TE to obtain a three-dimensional model;
[0022] The equation selection unit is used to select the control equation and discretization method of the three-dimensional model in a computational fluid dynamics (CFD) solver;
[0023] The discrete unit is used to discretize the control equations respectively using the discrete method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model;
[0024] The prediction unit is used to predict the temperature field of the ITE under different loads according to the prediction model.
[0025] Optionally, the modeling unit is used to:
[0026] The three-dimensional space is modeled according to the position, size and power of the I TE, as well as other objects that affect the thermal environment in the three-dimensional space except the I TE; the objects include air-conditioned compartments and raised floors.
[0027] Optionally, the equation selection unit is used to:
[0028] The mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations, and the control volume method is selected as the discretization method.
[0029] Optionally, the discrete unit is used to:
[0030] In the CFD solver, profiles are selected for the unsteady-state term, convection term, diffusion term, and source term in the control volume method respectively;
[0031] In a target control volume and a target time interval, the unsteady-state term, the convection term, the diffusion term and the source term are respectively integrated according to the profile to obtain an algebraic equation at a target node in the three-dimensional model;
[0032] The algebraic equations on all nodes in the three-dimensional model are obtained by iterative calculation and combined to obtain the prediction model.
[0033] Optionally, the device further comprises:
[0034] The energy-saving unit is used to predict the energy consumption of the refrigeration system of the three-dimensional space under different loads of I TE according to the temperature field.
[0035] The present application discloses a method and device for predicting the load and temperature field of information technology equipment. Based on the design drawings and on-site information of the information technology equipment I TE, a three-dimensional model of the I TE and the three-dimensional space where the I TE is placed is constructed to completely and accurately restore the data center with complex scenes where the I TE is located. The control equation and discretization method of the three-dimensional model are selected in the computational fluid dynamics (CFD) solver, and the control equation is discretized respectively using the discrete method to obtain a prediction model that can reflect the temperature, pressure and velocity values of each point in the three-dimensional model. According to the prediction model, the temperature field of the I TE under different loads is predicted. By combining CFD technology with relevant knowledge of heat transfer, the thermal field changes of the entire data center caused by the change of the I TE load can be accurately described. Thus, the I TE load and its temperature field prediction can be carried out quickly, accurately and at low cost. And there is no need to arrange a large number of temperature sensors and collect data for a long time. It is time-saving, low-cost and easy to implement. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.
[0037] Figure 1a A schematic flow chart of a method for predicting load and temperature field of information technology equipment disclosed in an embodiment of the present application;
[0038] Figure 1b A schematic diagram of a three-dimensional model disclosed in an embodiment of the present application;
[0039] Figure 1c A schematic diagram of the temperature field distribution disclosed in the embodiment of the present application;
[0040] Figure 2a A schematic flow chart of another method for predicting load and temperature field of information technology equipment disclosed in an embodiment of the present application;
[0041] Figure 2b A schematic diagram of the prediction result error analysis disclosed in the embodiment of the present application;
[0042] Figure 3 A schematic diagram of the structure of an apparatus for predicting load and temperature field of information technology equipment disclosed in an embodiment of the present application. DETAILED DESCRIPTION
[0043] This application studies the computational fluid dynamics (CFD) simulation solver and uses CFD technology to analyze the thermodynamic environment of I TE under different loads. It can accurately predict the temperature field of I TE under load changes, as well as the three-dimensional space where I TE is located, that is, the energy consumption of the refrigeration system of the data center.
[0044] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0045] Embodiment 1: This application discloses a method for predicting the load and temperature field of information technology equipment.
[0046] For details, please refer to Figure 1a The present embodiment discloses a method for predicting load and temperature field of information technology equipment, comprising the following steps:
[0047] Step 101: Based on the design drawings of the information technology equipment I TE and the on-site information, the I TE and the three-dimensional space where the I TE is placed are modeled to obtain a three-dimensional model.
[0048] Traditional CFD tool software usually sets the research object in a virtual wind tunnel scene in the form of a "wind tunnel experiment", performs simple boundary definition, and then solves and analyzes. The actual scene of the three-dimensional space (data center) where I TE is placed is more complicated than the wind tunnel test scene. For example, the precision air conditioner in the computer room of the data center will be operated at a variable frequency with a preset temperature and air volume to ensure that the inlet air temperature of I TE remains within the preset range when the load rate changes. At the same time, the infiltration wind at the door of the computer room, the perforation of the bridge water pipe, the heat transfer of the enclosure structure, the start and stop of the fresh air system, and other factors all cause heat input or output in the computer room.
[0049] Therefore, in the method described in this embodiment, Figure 1bAs shown, based on the design drawings and the relevant information collected on site, the data center and I TE are modeled. Specifically, the three-dimensional space can be modeled according to the location of I TE (including the shelf location of I TE in the cabinet), size and power, as well as other objects that affect the thermal environment in the three-dimensional space except I TE. Among them, other objects that affect the thermal environment in the three-dimensional space except I TE include air-conditioning compartments and raised floors. Accurate and complete modeling of the data center in this way will help to apply CFD technology to the prediction results obtained in the subsequent steps of the data center.
[0050] Step 102: Select the control equation and discretization method of the three-dimensional model in a computational fluid dynamics (CFD) solver.
[0051] The basic principle of CFD is to numerically solve the differential equations that control fluid flow, and obtain the discrete distribution of the flow field of the fluid flow in a continuous area, so as to approximately simulate the real fluid flow. In the method described in this embodiment, CFD technology is used to simulate the flow state of air, and combined with relevant knowledge of heat transfer, to describe the temperature field changes affected by I TE load changes.
[0052] The core of the CFD solver consists of a series of control equations, which can exist in a variety of mathematical expressions (such as conservation type, non-conservation type, etc.). Each expression has a common physical principle and can be converted to each other, but when using different numerical methods to solve, choosing the right expression can get a more reasonable solution and avoid problems such as non-physical solutions and instability.
[0053] For example, taking the fluid element as the research object will derive the differential equation, and taking the finite control volume as the research object will derive the integral equation. The difference between the two is that the integral equation allows discontinuities in the control volume, while the differential equation requires the flow field variables to be continuous and differentiable due to the concept of limit. When the divergence theorem is used to transform the integral equation into the differential equation, the divergence theorem requires the variables to be continuous and differentiable, making the integral equation more suitable for calculating the flow field containing discontinuities.
[0054] In the method described in this embodiment, for data centers, the energy transmission medium using air cooling technology or liquid cooling technology is continuous, incompressible viscous fluid (air and insulating oil). Therefore, differential equations with fluid microelement as the research object should be used as mathematical expressions.
[0055] In the method described in this embodiment, the environmental flow of the data center satisfies the Navier-Stokes equations. And because the fluid flow in the HVAC field is basically turbulent flow, it is also necessary to combine the turbulence model and add the turbulent transport equation. Therefore, the mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations. Among them, the gas flow in the data center is an incompressible viscous flow, so the air density is a constant.
[0056] At the same time, due to the complex structure of the data center, it is more suitable to use the control volume method with better conservation properties for discretization, that is, to apply the conservation law to the control volume to establish a discrete equation.
[0057] In the method described in this embodiment, the mass conservation equation, momentum conservation equation, energy conservation equation, k-ε turbulence equation and control volume method general equation are all derived formulas, and thus are not described in detail here.
[0058] Step 103: discretize the control equations respectively using the discretization method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model.
[0059] In the method described in this embodiment, the CFD solver can select profiles for the unsteady-state term, convection term, diffusion term, and source term in the control volume method, respectively. That is, the local distribution curve (profile) of the objective function and its derivative in time-space is selected to specify how to determine the interpolation method of the value being evaluated on the control volume interface from the function values of adjacent nodes.
[0060] In the method described in this embodiment, within the target control volume and the target time interval, the unsteady-state term, convection term, diffusion term and source term are integrated according to the selected number of the model line to obtain the algebraic equation at the target node in the three-dimensional model.
[0061] As an optional method, the above integration steps are performed for all selected control equations. Among them, the non-steady-state term selection line is as follows:
[0062]
[0063] Where t is time, Δt is the time interval, P is the control volume, Φ is a scalar variable, v is the velocity, and e and ω correspond to the surface vectors normal to the element face in the Cartesian grid.
[0064] The convection item selection line is as follows:
[0065]
[0066] Where u is a scalar.
[0067] The diffusion item selection line is as follows:
[0068]
[0069] In the formula, E and W are the same as e and w above, and the writing format in the formula is case-sensitive. δ is the same as Δ.
[0070] Source selection ladder:
[0071]
[0072] Where S is a scalar function.
[0073] In the method described in this embodiment, iterative calculations are performed to obtain algebraic equations on all nodes in the three-dimensional model (in a discrete coordinate system), and the prediction model can be obtained by combining them.
[0074] Step 104: predict the temperature field of the ITE under different loads according to the prediction model.
[0075] In the method described in this embodiment, Figure 1c As shown in the figure, solving the prediction model can comprehensively analyze the cold / heat load formation process of I TE under different loads, and analyze the temperature, velocity, and pressure field distribution in the entire data center room in detail. It can also predict the energy consumption of the refrigeration system in three-dimensional space under different loads of I TE.
[0076] The method described in this embodiment builds a three-dimensional model based on the design drawings and on-site information of I TE, and completely and accurately restores the complex data center where I TE is located. Based on CFD technology, the control equation and discrete method are selected, and combined with relevant knowledge of heat transfer, the temperature field changes and airflow field changes of the entire data center caused by the load changes of I TE are accurately reflected, as well as the energy consumption simulation results of the refrigeration system of the data center. There is no need to arrange a large number of temperature sensors and collect data for a long time. It is fast, time-saving, low-cost and easy to implement.
[0077] Example 2: This application discloses another method for predicting the load and temperature field of information technology equipment. Figure 2a The method described in this embodiment introduces the entire process of predicting the load and temperature field of information technology equipment.
[0078] Step 201: Obtain ITE design drawings and on-site information.
[0079] Step 202: Import the design drawings and site information, model the ITE and data center, and obtain a three-dimensional model.
[0080] Step 203: Select the control equation and discretization method in the CFD solver and assign it to the three-dimensional model.
[0081] Step 204: Discretize the control equation based on a discrete method, and iteratively calculate the discretized control equation to obtain a prediction model.
[0082] Step 205: Check the temperature field simulation results of ITE under different loads and the energy consumption simulation results of the cooling system of the data center according to the output results of the prediction model.
[0083] As an example, Figure 2b The comparison between the simulation results based on the experimental test and the actual measurement results of the sensor is shown to verify the simulation accuracy of the method described in this embodiment. Figure 2b As shown, AC~PH are the cold channel temperature sensors, and AF~PE are the hot channel temperature sensors. It can be seen that the simulated and measured temperature change trends of the 16 sensors in the cold channel are consistent, with an average simulation error of 3.1% and a maximum error of 7.73% for the CC sensor. Correspondingly, the average error between the hot channel simulation and the test data is 3.6%, and the maximum error is 8.19% for the KF sensor.
[0084] Based on the method for predicting the load and temperature field of information technology equipment disclosed in the above embodiment, this embodiment correspondingly discloses an apparatus for predicting the load and temperature field of information technology equipment. Figure 3 , the device for predicting load and temperature field of information technology equipment includes: a modeling unit 301, an equation selection unit 302, a discrete unit 303 and a prediction unit 304;
[0085] The modeling unit 301 is used to model the information technology equipment I TE and the three-dimensional space where the I TE is placed based on the design drawings and on-site information of the information technology equipment I TE to obtain a three-dimensional model;
[0086] The equation selection unit 302 is used to select the control equation and discretization method of the three-dimensional model in the computational fluid dynamics (CFD) solver;
[0087] The discrete unit 303 is used to discretize the control equations respectively using the discrete method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model;
[0088] The prediction unit 304 is used to predict the temperature field of the ITE under different loads according to the prediction model.
[0089] Optionally, the modeling unit 301 is used to:
[0090] The three-dimensional space is modeled according to the position, size and power of the I TE, as well as other objects that affect the thermal environment in the three-dimensional space except the I TE; the objects include air-conditioned compartments and raised floors.
[0091] Optionally, the equation selection unit 302 is used to:
[0092] The mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations, and the control volume method is selected as the discretization method.
[0093] Optionally, the discrete unit 303 is used to:
[0094] In the CFD solver, profiles are selected for the unsteady-state term, convection term, diffusion term, and source term in the control volume method respectively;
[0095] In a target control volume and a target time interval, the unsteady-state term, the convection term, the diffusion term and the source term are respectively integrated according to the profile to obtain an algebraic equation at a target node in the three-dimensional model;
[0096] The algebraic equations on all nodes in the three-dimensional model are obtained by iterative calculation and combined to obtain the prediction model.
[0097] Optionally, the device further comprises:
[0098] The energy-saving unit is used to predict the energy consumption of the refrigeration system of the three-dimensional space under different loads of I TE according to the temperature field.
[0099] The embodiments in this specification are described in a progressive manner. As for the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the methods.
[0100] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0101] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
[0102] The features described in the embodiments of this specification can be replaced or combined with each other, so that professional and technical personnel in this field can implement or use the present application.
[0103] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for predicting load and temperature field of information technology equipment, characterized in that: include: Based on the design drawings and on-site information of the information technology equipment ITE, the ITE and the three-dimensional space where the ITE is placed are modeled to obtain a three-dimensional model; Selecting the governing equations and discretization method of the three-dimensional model in a computational fluid dynamics (CFD) solver; The control equations are discretized using the discretization method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model; According to the prediction model, the temperature field of ITE under different loads is predicted.
2. The method according to claim 1, characterized in that The modeling of the information technology equipment ITE and the three-dimensional space where the ITE is placed based on the design drawings and on-site information of the information technology equipment ITE includes: The three-dimensional space is modeled according to the location, size and power of the ITE, as well as other objects other than the ITE that affect the thermal environment in the three-dimensional space; the objects include air-conditioned compartments and raised floors.
3. The method according to claim 1, characterized in that The control equations and discretization method of the three-dimensional model are selected in a computational fluid dynamics (CFD) solver, including: The mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations, and the control volume method is selected as the discretization method.
4. The method according to claim 1, characterized in that The step of using the discrete method to discretize the control equations respectively includes: In the CFD solver, profiles are selected for the unsteady-state term, convection term, diffusion term, and source term in the control volume method respectively; In a target control volume and a target time interval, the unsteady-state term, the convection term, the diffusion term and the source term are respectively integrated according to the profile to obtain an algebraic equation at a target node in the three-dimensional model; The algebraic equations on all nodes in the three-dimensional model are obtained by iterative calculation and combined to obtain the prediction model.
5. The method according to claim 1, characterized in that After predicting the temperature field of ITE under different loads according to the prediction model, the method further includes: According to the temperature field, the energy consumption of the refrigeration system of the three-dimensional space under different loads of ITE is predicted.
6. A device for predicting load and temperature field of information technology equipment, characterized in that: include: Modeling unit, equation selection unit, discrete unit and prediction unit; The modeling unit is used to model the information technology equipment ITE and the three-dimensional space where the ITE is placed based on the design drawings and on-site information of the ITE to obtain a three-dimensional model; The equation selection unit is used to select the control equation and discretization method of the three-dimensional model in a computational fluid dynamics (CFD) solver; The discrete unit is used to discretize the control equations respectively using the discrete method to obtain a prediction model; the prediction model reflects the temperature, pressure and velocity values of each point in the three-dimensional model; The prediction unit is used to predict the temperature field of ITE under different loads according to the prediction model.
7. The device according to claim 6, characterized in that The modeling unit is used for: The three-dimensional space is modeled according to the location, size and power of the ITE, as well as other objects other than the ITE that affect the thermal environment in the three-dimensional space; the objects include air-conditioned compartments and raised floors.
8. The device according to claim 6, characterized in that The equation selection unit is used to: The mass conservation equation, momentum conservation equation, energy conservation equation and k-ε turbulence equation are selected as the control equations, and the control volume method is selected as the discretization method.
9. The device according to claim 6, characterized in that The discrete unit is used for: In the CFD solver, profiles are selected for the unsteady-state term, convection term, diffusion term, and source term in the control volume method respectively; In a target control volume and a target time interval, the unsteady-state term, the convection term, the diffusion term and the source term are respectively integrated according to the profile to obtain an algebraic equation at a target node in the three-dimensional model; The algebraic equations on all nodes in the three-dimensional model are obtained by iterative calculation and combined to obtain the prediction model.
10. The device according to claim 6, characterized in that The device also includes: The energy-saving unit is used to predict the energy consumption of the refrigeration system of the three-dimensional space under different loads of ITE according to the temperature field.
