CN114625229A - Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center - Google Patents
Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center Download PDFInfo
- Publication number
- CN114625229A CN114625229A CN202210268074.8A CN202210268074A CN114625229A CN 114625229 A CN114625229 A CN 114625229A CN 202210268074 A CN202210268074 A CN 202210268074A CN 114625229 A CN114625229 A CN 114625229A
- Authority
- CN
- China
- Prior art keywords
- heat dissipation
- cooled
- temperature
- rotating speed
- liquid cooling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000017525 heat dissipation Effects 0.000 title claims abstract description 122
- 239000007788 liquid Substances 0.000 title claims abstract description 91
- 238000001816 cooling Methods 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000007654 immersion Methods 0.000 title claims abstract description 55
- 230000033228 biological regulation Effects 0.000 title claims abstract description 48
- 238000013528 artificial neural network Methods 0.000 claims abstract description 50
- 238000005457 optimization Methods 0.000 claims abstract description 33
- 239000012530 fluid Substances 0.000 claims abstract description 22
- 238000005265 energy consumption Methods 0.000 claims abstract description 15
- 230000008569 process Effects 0.000 claims description 25
- 230000008859 change Effects 0.000 claims description 15
- 239000000758 substrate Substances 0.000 claims description 14
- 230000001276 controlling effect Effects 0.000 claims description 11
- 239000003507 refrigerant Substances 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 6
- 230000001105 regulatory effect Effects 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 4
- 230000006870 function Effects 0.000 abstract description 11
- 238000013473 artificial intelligence Methods 0.000 abstract description 9
- 210000002569 neuron Anatomy 0.000 description 24
- 238000012545 processing Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 8
- 230000010365 information processing Effects 0.000 description 6
- 238000013507 mapping Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000004907 flux Effects 0.000 description 5
- 239000002699 waste material Substances 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000011478 gradient descent method Methods 0.000 description 3
- 230000000638 stimulation Effects 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 238000005057 refrigeration Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/16—Constructional details or arrangements
- G06F1/20—Cooling means
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D23/00—Control of temperature
- G05D23/19—Control of temperature characterised by the use of electric means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2200/00—Indexing scheme relating to G06F1/04 - G06F1/32
- G06F2200/20—Indexing scheme relating to G06F1/20
- G06F2200/201—Cooling arrangements using cooling fluid
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Automation & Control Theory (AREA)
- Human Computer Interaction (AREA)
- Cooling Or The Like Of Electrical Apparatus (AREA)
Abstract
The invention provides an optimization method, a device and a data center for regulation and control of immersion liquid cooling heat dissipation, wherein the method comprises the steps of obtaining the current temperature and the real-time temperature of each part to be dissipated in a liquid cooling cabinet, and calculating the real-time temperature variation of each part to be dissipated; and inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction by using the neural network operation as a target function that each part to be dissipated meets the temperature of a regulation point and the energy consumption of the system is the lowest. Based on the method, an immersion liquid cooling heat dissipation regulation and control optimization device and a data center are further provided. According to the invention, a driving pump is additionally arranged in front of a high-power-consumption part in a node of a traditional immersion system, so that the local flow velocity is increased, and the over-reaction of the flow of the whole system is reduced. The control is carried out in an artificial intelligence mode, each machine automatically learns, and is continuously optimized in continuous summary learning, and the system flow and the rotating speed control logic are calculated to achieve the purpose of controlling the optimal power consumption.
Description
Technical Field
The invention belongs to the technical field of server heat dissipation management, and particularly relates to an immersion liquid cooling heat dissipation regulation and control optimization method and device and a data center.
Background
With the development of new infrastructure such as cloud computing and big data, the requirement on data computing speed is higher and higher, the computing speed and the computation amount of a processor are also higher and higher, so that the power consumption of components such as a CPU (central processing unit) is increased and soared, the temperature spec is also reduced, especially, the power consumption of the CPU is increased by 80% each year, the heat dissipation of electronic devices becomes a problem which is quite scorching at present, the requirement on the power consumption of a system heat dissipation fan in the current society is higher and higher, the requirement on PUE (power output equipment) in a machine room is reduced, and energy conservation is a mainstream trend at present. How to effectively solve the problem of overhigh temperature of each electronic component, the immersion liquid cooling is generated at the same time, and the PUE of the machine room is greatly reduced. The PUE is a short hand for Power Usage efficiency, is an index for evaluating the energy efficiency of the data center, and is a ratio of all energy consumed by the data center to energy consumed by the IT load.
However, immersion liquid cooling in the prior art is still in a development stage, and the linear flow of the whole system is increased only by an external pump in the current adopted mode, so that the regulation and control of heat dissipation are relatively single. The flow that gets into Tank is adjusted through the outside CDU system of Tank to current submergence system, generally can 12 ~ 24 node servers in a Tank, it can lead to the flow increase of whole Tank system because of the part of a certain node is overheated also to deposit, cause the waste of flow, too big flow, the burden of pump has been increased, the pressure of whole pipeline has been increased, the risk of revealing has been increased, and cause more serious washing to the node in the system, the erosion and corrosion of part has been accelerated, the device life has been reduced. Even more worrying is that this increase in pipe system flow is limited in the flow allocated to individual nodes and may not be sufficient to solve the heat dissipation problem of large power consuming devices such as the corresponding CPU. Wherein, Tank is a water Tank; the CDU is a liquid cooling distribution device.
In the traditional concept, immersion liquid cooling is that a fan and other power-assisted heat dissipation measures are not needed in a server. With the increasing heat flux density of high heat flux devices, the limitation of immersion liquid cooling is also appearing, and the heat dissipation effect is not infinite. The immersion liquid cooling has the advantages that the density of the heat dissipation medium is 1800 times of that of air, the specific heat capacity is obviously higher than that of the air, so the heat dissipation effect is obvious, but the flow rate is also obviously reduced, so the final heat dissipation capacity is also limited.
Disclosure of Invention
In order to solve the technical problem, the invention provides an immersion liquid cooling heat dissipation regulation and control optimization method and device and a data center. From the angle of a server framework, the secondary joint debugging of the temperature control of the whole immersion system is realized by adopting an artificial intelligence learning mode, and the problem of component heat dissipation is solved to the greatest extent.
In order to achieve the purpose, the invention adopts the following technical scheme:
an optimization method for regulation and control of immersion liquid cooling heat dissipation comprises the following steps:
acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet, and calculating the real-time temperature variation of each part to be cooled;
and inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the operation of the neural network by taking each part to be dissipated meeting the temperature of a regulation point and the lowest energy consumption of the system as an objective function.
Further, the method also comprises the step of calculating the flow change in the liquid cooling cabinet according to the rotating speed of the heat dissipation pump corresponding to each part to be dissipated.
Further, the method for calculating the flow change in the liquid cooling cabinet comprises the following steps: q ═ cm Δ t;
wherein Q is the system flow in the liquid-cooled cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature variation.
Further, the neural network comprises an input layer, a hidden layer and an output layer;
the input layer is used for receiving real-time temperature variation of each part to be cooled;
the output layer is used for outputting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated;
the hidden layer is located between the input layer and the output layer, the target function is that each part to be cooled meets the temperature of a control point and the energy consumption of the system is the lowest, and the hidden layer is used for converting the real-time temperature variation into the rotating speed of the heat dissipation pump corresponding to each part to be cooled.
Further, the method further comprises:
when the rotating speed of the heat dissipation pump corresponding to each part to be dissipated, which is output by the output layer of the neural network, does not accord with the expected rotating speed, the error between the output rotating speed and the expected rotating speed is corrected through the output layer in a mode of error gradient reduction, and the weight of each layer is reversely transmitted to the hidden layer and the input layer by layer;
and continuously adjusting the weight of each layer until the error meets the heat dissipation requirement.
Further, the detailed process of predicting the rotating speed of the heat dissipation pump corresponding to each component to be heat dissipated in the fluid entering direction through the neural network operation includes:
if the temperature of any part to be cooled exceeds 10%, the temperature of other parts is lower than 10%, and the flow rate of the system is unchanged, increasing the rotating speed of a cooling pump corresponding to the heated part to be cooled and increasing the local flow rate;
when the temperature rise of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is increased by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is increased by 30%;
if the temperature of any part to be cooled is more than 10%, the temperature of other parts is less than 10%, and the flow of the system is unchanged, the rotating speed of a heat-radiating pump corresponding to the part to be cooled is increased, and the local flow is reduced;
when the cooling of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is reduced by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is reduced by 30%.
The invention also provides an immersion liquid cooling heat dissipation regulation and control optimization device, which comprises a substrate management controller, parts to be dissipated and heat dissipation pumps corresponding to the parts to be dissipated in the fluid entering direction;
the substrate management controller is in communication connection with each part to be cooled; the system comprises a liquid cooling cabinet, a temperature sensor and a temperature controller, wherein the liquid cooling cabinet is used for acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet and calculating the real-time temperature variation of each part to be cooled; inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the neural network operation by taking each part to be dissipated meeting the temperature of a regulation point and the lowest system energy consumption as a target function;
the heat dissipation pump is used for adjusting the rotating speed calculated by the substrate management controller.
The apparatus of claim 7, further comprising a cold liquid distribution device; the substrate management controller is in communication connection with the cold liquid distribution device; and the cold liquid distribution device is used for controlling the output flow of the system according to the rotating speed predicted by the neural network.
The immersion liquid cooling heat dissipation regulation and optimization device of claim 9, wherein the method for the cold liquid distribution device to calculate the flow change in the liquid cooling cabinet comprises: q ═ cm Δ t; wherein Q is the system flow in the liquid-cooled cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature variation.
The invention also provides a data center which comprises the immersion liquid cooling heat dissipation regulation and control optimization device.
The effect provided in the summary of the invention is only the effect of the embodiment, not all the effects of the invention, and one of the above technical solutions has the following advantages or beneficial effects:
the invention provides an optimization method, a device and a data center for regulation and control of immersion liquid cooling heat dissipation, wherein the method comprises the steps of obtaining the current temperature and the real-time temperature of each part to be dissipated in a liquid cooling cabinet, and calculating the real-time temperature variation of each part to be dissipated; and inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction by using the neural network operation as a target function that each part to be dissipated meets the temperature of a regulation point and the energy consumption of the system is the lowest. Based on the method, the immersion liquid cooling heat dissipation regulation and control optimization device and the data center are further provided. The invention optimizes the structure system based on the original structure, performs assistant optimization on the heat dissipation bottleneck component in the immersion system, and adds a heat dissipation pump in front of the component radiator with high power consumption and high heat flow density such as a CPU (central processing unit) for heat dissipation. The heat dissipation pump and the CDU carry out combined two-stage intelligent regulation and control on the whole system, the problem of component heat dissipation and the problem of the heat dissipation power consumption ratio of the whole system are solved to the maximum extent, the PUE of a machine room is reduced to the maximum extent, and the best efficiency is achieved. The immersion liquid cooling scheme is eliminated, the primary concept that a fan or a motor is not needed in the system is achieved, the system is in accordance with the Internet and the great era, and the requirements of future machine rooms are met.
Drawings
Fig. 1 is a flowchart of an optimization method for regulating and controlling immersion liquid cooling heat dissipation in embodiment 1 of the present invention;
fig. 2 is a schematic diagram of a neural network in an immersion liquid cooling heat dissipation regulation and optimization method according to embodiment 1 of the present invention;
fig. 3 is a schematic connection diagram of an immersion liquid cooling heat dissipation regulation and control optimization device in embodiment 2 of the present invention;
fig. 4 is a schematic diagram of a position of a heat dissipation pump added in an immersion liquid cooling heat dissipation regulation and control optimization device in embodiment 2 of the present invention.
Detailed Description
In order to clearly explain the technical features of the present invention, the following detailed description of the present invention is provided with reference to the accompanying drawings. The following disclosure provides many different embodiments, or examples, for implementing different features of the invention. To simplify the disclosure of the present invention, specific example components and arrangements are described below. Furthermore, the present invention may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and procedures are omitted so as to not unnecessarily limit the invention.
Example 1
The embodiment 1 of the invention provides an immersion liquid cooling heat dissipation regulation and control optimization method, which adopts an artificial intelligence learning mode to realize two-stage joint regulation of temperature control of the whole immersion system, so as to reduce the PUE value of a machine room as much as possible and achieve the purpose of lowest overall power consumption. When the method is implemented, each device in the whole system is used as a regulation parameter to guide the flow control of a CDU (cold liquid distribution device), and when the temperature of a certain device is too high, the CDU controls the flow of the whole liquid cooling cabinet system to increase until the temperature meets the requirement, so that the flow of the whole system is greatly increased, the waste of the flow is caused, and the excessive flushing of the internal devices is caused. Reducing the service life.
Fig. 1 shows a flow chart of an optimization method for regulating and controlling immersion liquid cooling heat dissipation.
In step S100, obtaining the current temperature and the real-time temperature of each component to be cooled in the liquid-cooled cabinet, and calculating the real-time temperature variation of each component to be cooled;
each component to be cooled within the liquid-cooled cabinet includes, but is not limited to, a high heat flux CPU (central processing unit), a GPU (graphics processing unit), and a TPU (timing unit). Firstly, a heat dissipation pump is added at the position of a fluid inlet of a part to be dissipated for flow control, and the flow of an immersion system flows from bottom to top.
In step S110, the temperature variation is input into the neural network, and the rotation speed of the heat dissipation pump corresponding to each component to be heat dissipated in the fluid entering direction is predicted through the neural network operation with the objective function that each component to be heat dissipated satisfies the temperature of the control point and the energy consumption of the system is the lowest.
Fig. 2 is a schematic diagram of a neural network in an immersion liquid cooling heat dissipation regulation and optimization method in embodiment 1 of the present invention; the neural network comprises an input layer, a hidden layer and an output layer; the input layer is used for receiving the real-time temperature variation of each part to be cooled; the output layer is used for outputting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated; the hidden layer is positioned between the input layer and the output layer, and the target function is that each part to be cooled meets the temperature of a control point and the energy consumption of the system is the lowest, so that the real-time temperature variation is converted into the rotating speed of the heat dissipation pump corresponding to each part to be cooled.
Neural networks are capable of learning and storing a large number of input-output pattern mappings without prior disclosure of mathematical equations describing such mappings. The learning rule is that a gradient descent method is used, and the weight and the threshold value of the network are continuously adjusted through back propagation, so that the error square sum of the network is minimum.
The weight is the probability of implementation of this path. The connection path between neurons in each layer and adjacent layers.
The threshold value is a critical value, and when the external stimulation reaches a certain threshold value, the neuron is stimulated to influence the next neuron. On each neuron. (input layer neurons do not)
The initial values of the weights and thresholds of the neural network are generally randomly generated.
The neural network uses the BP algorithm. The BP algorithm consists of two processes, forward computation of the data stream (forward propagation) and back propagation of the error signal.
The idea of the BP algorithm is as follows: each neuron of the input layer is responsible for receiving input temperature information from the outside and transmitting the input temperature information to each neuron of the middle layer; the middle layer is an internal information processing layer and is responsible for information transformation, and can be designed into a single hidden layer or a multi-hidden layer structure according to the requirement of information change capability; the information of each neuron transmitted to the output layer by the last hidden layer is further processed to complete the forward propagation processing process of one learning, and the output layer outputs the information processing result, namely the rotating speed value, to the outside. When the actual output does not match the desired output, the error back-propagation phase is entered. And the error passes through the output layer, the weight of each layer is corrected in a mode of error gradient reduction, and the weight is reversely transmitted to the hidden layer and the input layer by layer.
The repeated information forward propagation and error backward propagation process is a process of continuously adjusting weights of all layers and a process of learning and training the neural network, and the process is carried out until the error output by the network is reduced to an acceptable degree or preset learning times.
The detailed process for predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the neural network operation comprises the following steps:
if the temperature of any part to be cooled exceeds 10%, the temperature of other parts is lower than 10%, and the flow of the system is unchanged, increasing the rotating speed of a cooling pump corresponding to the heated part to be cooled to increase the local flow;
when the temperature rise of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is increased by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is increased by 30%;
if the temperature of any part to be cooled is more than 10%, the temperature of other parts is less than 10%, and the flow of the system is unchanged, the rotating speed of a heat-radiating pump corresponding to the part to be cooled is increased, and the local flow is reduced;
when the cooling of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is reduced by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is reduced by 30%.
In step S120, the flow change in the liquid-cooled cabinet is calculated according to the rotation speed of the heat dissipation pump corresponding to each component to be cooled.
The method for calculating the flow change in the liquid cooling cabinet comprises the following steps: q ═ cm Δ t;
q is the system flow in the liquid cooling cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature change.
The system automatically performs statistics, collection, statistics and learning, and calculates out reasonable CDU flow of the system and flow rotating speed of a pump in a node, so that the temperature of each component in the system is controlled to meet requirements at the lowest flow rotating speed, the PUE value of the whole machine room is reduced as far as possible, finally, each device meets the corresponding temperature of a regulation and control point, and the heat dissipation energy consumption of the system is lowest through synchronous statistics.
In the method for optimizing regulation and control of immersion liquid cooling heat dissipation provided in embodiment 1 of the present invention, from the perspective of server architecture, a two-stage joint regulation of temperature control of the entire immersion system is implemented in an artificial intelligence learning manner, so as to reduce the PUE value of a machine room as much as possible, and achieve the purpose of lowest overall power consumption. Particularly, a heat dissipation pump is additionally arranged in front of a high-power-consumption part in a node of a traditional immersion system, so that the local flow speed is increased, and the over-reaction of the flow of the whole system is reduced. And secondly, the original temperature components of the replacement points are linearly controlled, the control is carried out in an artificial intelligence mode, all machines automatically learn, the optimization is continuously carried out in continuous summary learning, and a set of flow and rotating speed control logic of the system is calculated to achieve the purpose of controlling the optimal power consumption.
Example 2
Based on the method for regulating and controlling the immersion liquid cooling heat dissipation provided by the embodiment 1 of the present invention, the embodiment 2 of the present invention further provides an apparatus for regulating and controlling the immersion liquid cooling heat dissipation, as shown in fig. 3, which is a schematic connection diagram of the apparatus for regulating and controlling the immersion liquid cooling heat dissipation provided by the embodiment 2 of the present invention. The device comprises a substrate management controller, parts to be cooled and heat dissipation pumps corresponding to the parts to be cooled in the fluid inlet direction;
the substrate management controller is in communication connection with each part to be cooled; the system comprises a liquid cooling cabinet, a temperature sensor and a temperature controller, wherein the liquid cooling cabinet is used for acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet and calculating the real-time temperature variation of each part to be cooled; inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction by using the neural network operation with the temperature of each part to be dissipated meeting the regulation and control point and the lowest system energy consumption as a target function;
the base plate management controller transmits the rotation speed signals after analysis and calculation to the original fan signal slot position, and the heat dissipation pump receives the original fan signals and the power supply and rotation speed signals of the power supply slot. The heat dissipation pump is used for adjusting the rotating speed calculated by the substrate management controller.
Fig. 4 is a schematic diagram of positions of a heat dissipation pump added to the immersion liquid cooling heat dissipation regulation and control optimization device in embodiment 2 of the present invention. The parts to be radiated include, but are not limited to, a high heat flux density CPU (central processing unit), a GPU (graphic processing unit), and a TPU (timing processor). Firstly, a heat dissipation pump is added at the position of a fluid inlet of a part to be dissipated for flow control, and the flow of an immersion system flows from bottom to top.
Each device in the whole system is used as a regulation parameter to guide the flow control of the CDU, when the temperature of a certain device is too high, the CDU controls the flow of the whole tank system to increase until the temperature meets the requirement, so that the flow of the whole system is greatly increased, the waste of the flow is caused, and the internal devices are excessively flushed. Reducing the service life.
In embodiment 2 of the present invention, a neural network is operated in a baseboard management controller to predict traffic, and the neural network is schematically illustrated; the neural network comprises an input layer, a hidden layer and an output layer; the input layer is used for receiving the real-time temperature variation of each part to be cooled; the output layer is used for outputting the rotating speed of the heat dissipation pump corresponding to each part to be cooled; the hidden layer is positioned between the input layer and the output layer, and the target function is that each part to be cooled meets the temperature of a control point and the energy consumption of the system is the lowest, so that the real-time temperature variation is converted into the rotating speed of the heat dissipation pump corresponding to each part to be cooled.
Neural networks are capable of learning and storing a large number of input-output pattern mappings without prior disclosure of mathematical equations describing such mappings. The learning rule is that a gradient descent method is used, and the weight and the threshold value of the network are continuously adjusted through back propagation, so that the error square sum of the network is minimum.
The weight is the probability of implementation of this path. The connection path between neurons in each layer and adjacent layers.
The threshold value is a critical value, and when the external stimulation reaches a certain threshold value, the neuron is stimulated to influence the next neuron. On each neuron. (input layer neurons do not)
The initial values of the weights and thresholds of the neural network are generally randomly generated.
The neural network uses the BP algorithm. The BP algorithm consists of two processes, forward computation of the data stream (forward propagation) and back propagation of the error signal.
The idea of the BP algorithm is as follows: each neuron of the input layer is responsible for receiving input temperature information from the outside and transmitting the input temperature information to each neuron of the middle layer; the middle layer is an internal information processing layer and is responsible for information transformation, and can be designed into a single hidden layer or a multi-hidden layer structure according to the requirement of information change capability; the information of each neuron transmitted to the output layer by the last hidden layer is further processed to complete the forward propagation processing process of one learning, and the output layer outputs the information processing result, namely the rotating speed value, to the outside. When the actual output does not match the desired output, the error back-propagation phase is entered. And the error passes through the output layer, the weight of each layer is corrected in a mode of error gradient reduction, and the weight is reversely transmitted to the hidden layer and the input layer by layer.
The repeated information forward propagation and error backward propagation process is a process of continuously adjusting weights of all layers and a process of learning and training the neural network, and the process is carried out until the error output by the network is reduced to an acceptable degree or preset learning times.
The detailed process for predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the neural network operation comprises the following steps:
if the temperature of any part to be cooled exceeds 10%, the temperature of other parts is lower than 10%, and the flow of the system is unchanged, increasing the rotating speed of a cooling pump corresponding to the heated part to be cooled to increase the local flow;
when the temperature rise of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is increased by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is increased by 30%;
if the temperature of any part to be cooled is more than 10%, the temperature of other parts is less than 10%, and the flow of the system is unchanged, the rotating speed of a heat-radiating pump corresponding to the part to be cooled is increased, and the local flow is reduced;
when the cooling of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is reduced by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is reduced by 30%.
The invention provides an immersion liquid cooling heat dissipation regulation and control optimization device, which also comprises a CDU (central control unit), namely a cold liquid distribution device, wherein the method for calculating the flow change in a liquid cooling cabinet by the cold liquid distribution device comprises the following steps: q ═ cm Δ t; wherein Q is the system flow in the liquid-cooled cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature change.
In the immersion liquid cooling heat dissipation regulation and control optimization device provided in embodiment 2 of the present invention, from the perspective of server architecture, two-stage joint regulation of temperature control of the entire immersion system is implemented in an artificial intelligence learning manner, so as to reduce the PUE value of a machine room as much as possible, and achieve the purpose of lowest overall power consumption. Particularly, a heat dissipation pump is additionally arranged in front of a high-power-consumption part in a node of a traditional immersion system, so that the local flow speed is increased, and the over-reaction of the flow of the whole system is reduced. And secondly, the original temperature components of the replacement points are linearly controlled, the control is carried out in an artificial intelligence mode, all machines automatically learn, the optimization is continuously carried out in continuous summary learning, and a set of flow and rotating speed control logic of the system is calculated to achieve the purpose of controlling the optimal power consumption.
Example 3
Based on the immersion liquid cooling heat dissipation regulation and control optimization device provided by the embodiment 2 of the invention, the embodiment 3 of the invention also provides a data center, and the data center comprises the immersion liquid cooling heat dissipation regulation and control optimization device.
Fig. 3 is a schematic connection diagram of an immersion liquid cooling heat dissipation regulation and control optimization device in embodiment 2 of the present invention. The device comprises a substrate management controller, parts to be cooled and heat dissipation pumps corresponding to the parts to be cooled in the fluid inlet direction;
the substrate management controller is in communication connection with each part to be cooled; the system comprises a liquid cooling cabinet, a temperature sensor and a temperature controller, wherein the liquid cooling cabinet is used for acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet and calculating the real-time temperature variation of each part to be cooled; inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction by using the neural network operation with the temperature of each part to be dissipated meeting the regulation and control point and the lowest system energy consumption as a target function;
the base plate management controller transmits the rotation speed signals after analysis and calculation to the original fan signal slot position, and the heat dissipation pump receives the original fan signals and the power supply and rotation speed signals of the power supply slot. The heat dissipation pump is used for adjusting the rotating speed calculated by the substrate management controller.
Fig. 4 is a schematic diagram of a position of a heat dissipation pump added in an immersion liquid cooling heat dissipation regulation and control optimization device in embodiment 2 of the present invention. The parts to be radiated include, but are not limited to, a high heat flux density CPU (central processing unit), a GPU (graphic processing unit), and a TPU (timing processor). Firstly, a heat dissipation pump is added at the position of a fluid inlet of a part to be dissipated for flow control, and the flow of an immersion system flows from bottom to top.
Each device in the whole system is used as a regulation and control parameter to guide the flow control of the CDU, when the temperature of a certain device is too high, the CDU controls the flow of the whole tank system to increase until the temperature meets the requirement, so that the flow of the whole system is greatly increased, the waste of the flow is caused, and the internal devices are excessively flushed. Reducing the service life.
In embodiment 2 of the present invention, a neural network is operated in a baseboard management controller to predict traffic, and the neural network is schematically illustrated; the neural network comprises an input layer, a hidden layer and an output layer; the input layer is used for receiving the real-time temperature variation of each part to be cooled; the output layer is used for outputting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated; the hidden layer is positioned between the input layer and the output layer, and the target function is that each part to be cooled meets the temperature of a control point and the energy consumption of the system is the lowest, so that the real-time temperature variation is converted into the rotating speed of the heat dissipation pump corresponding to each part to be cooled.
Neural networks can learn and store a large number of input-output pattern mappings without prior disclosure of mathematical equations describing such mappings. The learning rule is that a gradient descent method is used, and the weight and the threshold value of the network are continuously adjusted through back propagation, so that the error square sum of the network is minimum.
The weight is the probability of implementation of this path. The connection path between neurons in each layer and adjacent layers.
The threshold value is a critical value, and when the external stimulation reaches a certain threshold value, the neuron is stimulated to influence the next neuron. On each neuron. (input layer neurons do not)
The initial values of the weights and thresholds of the neural network are generally randomly generated.
The neural network uses the BP algorithm. The BP algorithm consists of two processes, forward computation of the data stream (forward propagation) and back propagation of the error signal.
The idea of the BP algorithm is as follows: each neuron of the input layer is responsible for receiving input temperature information from the outside and transmitting the input temperature information to each neuron of the middle layer; the middle layer is an internal information processing layer and is responsible for information transformation, and can be designed into a single hidden layer or a multi-hidden layer structure according to the requirement of information change capability; the information of each neuron transmitted to the output layer by the last hidden layer is further processed to complete the forward propagation processing process of one learning, and the output layer outputs the information processing result, namely the rotating speed value, to the outside. When the actual output does not match the desired output, the error back-propagation phase is entered. The error is corrected through the output layer according to the error gradient descending mode, and the weight values of all layers are reversely transmitted to the hidden layer and the input layer by layer.
The repeated information forward propagation and error backward propagation process is a process of continuously adjusting weights of all layers and a process of learning and training the neural network, and the process is carried out until the error output by the network is reduced to an acceptable degree or preset learning times.
The detailed process for predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the neural network operation comprises the following steps:
if the temperature of any part to be cooled exceeds 10%, the temperature of other parts is lower than 10%, and the flow of the system is unchanged, increasing the rotating speed of a cooling pump corresponding to the heated part to be cooled to increase the local flow;
when the temperature rise of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is increased by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is increased by 30%;
if the temperature of any part to be cooled is more than 10%, the temperature of other parts is less than 10%, and the flow of the system is unchanged, the rotating speed of a heat-radiating pump corresponding to the part to be cooled is increased, and the local flow is reduced;
when the cooling of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is reduced by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is reduced by 30%.
The invention provides an immersion liquid cooling heat dissipation regulation and control optimization device, which further comprises a CDU (refrigeration discharge unit), namely a cold liquid distribution device, wherein the method for calculating the flow change in a liquid cooling cabinet by the cold liquid distribution device comprises the following steps: q ═ cm Δ t; q is the system flow in the liquid cooling cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature variation.
In the data center provided in embodiment 3 of the present invention, the data center includes an immersion liquid cooling heat dissipation regulation and control optimization apparatus, and from the perspective of a server architecture, a secondary joint regulation of temperature control of the entire immersion system is implemented in an artificial intelligence learning manner, so as to reduce the PUE value of a machine room as much as possible, and achieve the purpose of lowest overall power consumption. Particularly, a heat dissipation pump is additionally arranged in front of a high-power-consumption part in a node of a traditional immersion system, so that the local flow speed is increased, and the over-reaction of the flow of the whole system is reduced. And secondly, the original temperature components of the replacement points are linearly controlled, the control is carried out in an artificial intelligence mode, all machines automatically learn, the optimization is continuously carried out in continuous summary learning, and a set of flow and rotating speed control logic of the system is calculated to achieve the purpose of controlling the optimal power consumption.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include elements inherent in the list. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element. In addition, parts of the above technical solutions provided in the embodiments of the present application, which are consistent with the implementation principles of corresponding technical solutions in the prior art, are not described in detail so as to avoid redundant description.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, the scope of the present invention is not limited thereto. Various modifications and alterations will occur to those skilled in the art based on the foregoing description. And are neither required nor exhaustive of all embodiments. On the basis of the technical scheme of the invention, various modifications or changes which can be made by a person skilled in the art without creative efforts are still within the protection scope of the invention.
Claims (10)
1. An optimization method for regulating and controlling immersion liquid cooling heat dissipation is characterized by comprising the following steps:
acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet, and calculating the real-time temperature variation of each part to be cooled;
and inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the operation of the neural network by taking each part to be dissipated meeting the temperature of a regulation point and the lowest energy consumption of the system as an objective function.
2. The method of claim 1, further comprising calculating a change in flow rate within the liquid-cooled cabinet based on a rotational speed of the heat sink pump associated with each component to be cooled.
3. The method of claim 2, wherein the method of calculating the change in flow rate in the liquid cooled cabinet comprises: q ═ cm Δ t;
wherein Q is the system flow in the liquid-cooled cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature variation.
4. The method of claim 1, wherein the neural network comprises an input layer, a hidden layer, and an output layer;
the input layer is used for receiving real-time temperature variation of each part to be cooled;
the output layer is used for outputting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated;
the hidden layer is located between the input layer and the output layer, the target function is that each part to be cooled meets the temperature of a control point and the energy consumption of the system is the lowest, and the hidden layer is used for converting the real-time temperature variation into the rotating speed of the heat dissipation pump corresponding to each part to be cooled.
5. The method of claim 4, wherein the method further comprises:
when the rotating speed of the heat dissipation pump corresponding to each part to be dissipated, which is output by the output layer of the neural network, does not accord with the expected rotating speed, the error between the output rotating speed and the expected rotating speed is corrected through the output layer in a mode of error gradient reduction, and the weight of each layer is reversely transmitted to the hidden layer and the input layer by layer;
and continuously adjusting the weights of all layers until the errors meet the heat dissipation requirement.
6. The method for optimizing regulation and control of immersion liquid cooling heat dissipation according to any one of claims 1 to 5, wherein the detailed process of predicting the rotating speed of the heat dissipation pump corresponding to each component to be heat dissipated in the fluid entering direction through the neural network operation comprises:
if the temperature of any part to be cooled exceeds 10%, the temperature of other parts is lower than 10%, and the flow of the system is unchanged, increasing the rotating speed of a cooling pump corresponding to the heated part to be cooled to increase the local flow;
when the temperature rise of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is increased by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is increased by 30%;
if the temperature of any part to be cooled is more than 10%, the temperature of other parts is less than 10%, and the flow of the system is unchanged, the rotating speed of a heat-radiating pump corresponding to the part to be cooled is increased, and the local flow is reduced;
when the cooling of all the heat dissipation components exceeds 20%, the overall flow in the liquid cooling cabinet is reduced by 20%, and the rotating speed of the heat dissipation pump corresponding to each component to be dissipated is reduced by 30%.
7. An immersion liquid cooling heat dissipation regulation and control optimization device is characterized by comprising a substrate management controller, parts to be heat dissipated and heat dissipation pumps corresponding to the parts to be heat dissipated in the fluid entering direction;
the substrate management controller is in communication connection with each part to be cooled; the system comprises a liquid cooling cabinet, a temperature sensor and a temperature controller, wherein the liquid cooling cabinet is used for acquiring the current temperature and the real-time temperature of each part to be cooled in the liquid cooling cabinet and calculating the real-time temperature variation of each part to be cooled; inputting the temperature variation into a neural network, and predicting the rotating speed of the heat dissipation pump corresponding to each part to be dissipated in the fluid entering direction through the neural network operation by taking each part to be dissipated meeting the temperature of a regulation point and the lowest system energy consumption as a target function;
the heat dissipation pump is used for adjusting the rotating speed calculated by the substrate management controller.
8. The apparatus of claim 7, further comprising a cold fluid distribution device; the substrate management controller is in communication connection with the cold liquid distribution device; and the cold liquid distribution device is used for controlling the output flow of the system according to the rotating speed predicted by the neural network.
9. The apparatus of claim 9, wherein the means for calculating the change in flow rate in the liquid-cooled cabinet comprises: q ═ cm Δ t; q is the system flow in the liquid cooling cabinet; c is the specific heat capacity of the refrigerant; Δ t: is the temperature variation.
10. A data center comprising an immersion liquid cooling regulation and control optimization device of any one of claims 7 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210268074.8A CN114625229A (en) | 2022-03-18 | 2022-03-18 | Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210268074.8A CN114625229A (en) | 2022-03-18 | 2022-03-18 | Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114625229A true CN114625229A (en) | 2022-06-14 |
Family
ID=81901801
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210268074.8A Pending CN114625229A (en) | 2022-03-18 | 2022-03-18 | Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114625229A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115344069A (en) * | 2022-07-28 | 2022-11-15 | 上海蓝色帛缔智能工程有限公司 | Liquid cooling control method and system for data center |
| CN117156828A (en) * | 2023-10-31 | 2023-12-01 | 联通(广东)产业互联网有限公司 | Data center heat dissipation system and method based on liquid cooling |
| CN117666738A (en) * | 2023-12-14 | 2024-03-08 | 南方电网大数据服务有限公司 | Liquid cooling server control method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190360711A1 (en) * | 2018-05-22 | 2019-11-28 | Seokyoung Systems | Method and device for controlling power supply to heating, ventilating, and air-conditioning (hvac) system for building based on target temperature |
| US20200100394A1 (en) * | 2018-09-21 | 2020-03-26 | Johnson Controls Technology Company | Optimized thermal control of data center |
| US20210018016A1 (en) * | 2019-07-18 | 2021-01-21 | Hyundai Motor Company | Deep learning-based cooling system temperature prediction apparatus according to physical causality and method therefor |
| CN112367809A (en) * | 2020-11-02 | 2021-02-12 | 中国电子科技集团公司第二十研究所 | Directional efficient heat dissipation device based on deep learning and installation monitoring method |
| CN114025567A (en) * | 2021-11-03 | 2022-02-08 | 成都新洲航空设备有限责任公司 | Helicopter liquid cooling control system and control method thereof |
| EP4053653A1 (en) * | 2021-03-05 | 2022-09-07 | Air Products and Chemicals, Inc. | Method and apparatus for monitoring operational characteristics of an industrial gas plant complex |
-
2022
- 2022-03-18 CN CN202210268074.8A patent/CN114625229A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190360711A1 (en) * | 2018-05-22 | 2019-11-28 | Seokyoung Systems | Method and device for controlling power supply to heating, ventilating, and air-conditioning (hvac) system for building based on target temperature |
| US20200100394A1 (en) * | 2018-09-21 | 2020-03-26 | Johnson Controls Technology Company | Optimized thermal control of data center |
| US20210018016A1 (en) * | 2019-07-18 | 2021-01-21 | Hyundai Motor Company | Deep learning-based cooling system temperature prediction apparatus according to physical causality and method therefor |
| CN112367809A (en) * | 2020-11-02 | 2021-02-12 | 中国电子科技集团公司第二十研究所 | Directional efficient heat dissipation device based on deep learning and installation monitoring method |
| EP4053653A1 (en) * | 2021-03-05 | 2022-09-07 | Air Products and Chemicals, Inc. | Method and apparatus for monitoring operational characteristics of an industrial gas plant complex |
| CN114025567A (en) * | 2021-11-03 | 2022-02-08 | 成都新洲航空设备有限责任公司 | Helicopter liquid cooling control system and control method thereof |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115344069A (en) * | 2022-07-28 | 2022-11-15 | 上海蓝色帛缔智能工程有限公司 | Liquid cooling control method and system for data center |
| CN117156828A (en) * | 2023-10-31 | 2023-12-01 | 联通(广东)产业互联网有限公司 | Data center heat dissipation system and method based on liquid cooling |
| CN117156828B (en) * | 2023-10-31 | 2024-02-02 | 联通(广东)产业互联网有限公司 | Data center heat dissipation system and method based on liquid cooling |
| CN117666738A (en) * | 2023-12-14 | 2024-03-08 | 南方电网大数据服务有限公司 | Liquid cooling server control method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114625229A (en) | Method and device for optimizing regulation and control of immersion liquid cooling heat dissipation and data center | |
| Zhang et al. | A survey on data center cooling systems: Technology, power consumption modeling and control strategy optimization | |
| CN111256294B (en) | Model prediction-based optimization control method for combined operation of water chilling unit | |
| CN109800066B (en) | Energy-saving scheduling method and system for data center | |
| CN111174375B (en) | Data center energy consumption minimization-oriented job scheduling and machine room air conditioner regulation and control method | |
| RU2270469C2 (en) | System for controlling modes of electric energy based systems | |
| CN110609474B (en) | Data center energy efficiency optimization method based on reinforcement learning | |
| CN113039506B (en) | Causal learning-based data center foundation structure optimization method | |
| CN113962142B (en) | Data center temperature prediction method and system based on two-section LSTM | |
| CN113739365A (en) | Central air-conditioning cold station group control energy-saving control method, device, equipment and storage medium | |
| CN113825356B (en) | Energy-saving control method and device for cold source system, electronic equipment and storage medium | |
| Fang et al. | A neural-network enhanced modeling method for real-time evaluation of the temperature distribution in a data center | |
| Choi et al. | Development of an adaptive artificial neural network model and optimal control algorithm for a data center cyber–physical system | |
| CN113299059A (en) | Data-driven road traffic control decision support method | |
| CN115632406B (en) | Reactive voltage control method and system based on digital-mechanism fusion driving modeling | |
| CN116914751B (en) | Intelligent power distribution control system | |
| CN112413762A (en) | Parameter optimization method and system for cooling water system of refrigerating room | |
| CN116501104A (en) | Temperature control device and method for liquid cooling heat dissipation system | |
| Zhang et al. | A two-layer optimal scheduling framework for energy savings in a data center for Cyber–Physical–Social Systems | |
| WO2020227983A1 (en) | Hybrid-learning neural network architecture | |
| CN117350165A (en) | 3D chip dynamic thermal management method based on deep reinforcement learning | |
| CN113064480A (en) | Multi-data center collaborative energy saving method based on multi-agent reinforcement learning | |
| CN114980666B (en) | Control method and system for liquid cooling heat dissipation system of data center | |
| CN114636212B (en) | GRNN-based running control method for multiple water chilling unit systems | |
| Kumar et al. | Effect of cooling systems on the energy efficiency of data centers: machine learning optimisation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220614 |
|
| RJ01 | Rejection of invention patent application after publication |