CN108990383B - Predictive control method for air conditioning system of data center - Google Patents
Predictive control method for air conditioning system of data center Download PDFInfo
- Publication number
- CN108990383B CN108990383B CN201810927800.6A CN201810927800A CN108990383B CN 108990383 B CN108990383 B CN 108990383B CN 201810927800 A CN201810927800 A CN 201810927800A CN 108990383 B CN108990383 B CN 108990383B
- Authority
- CN
- China
- Prior art keywords
- controller
- data center
- refrigerating device
- set value
- refrigeration
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20709—Modifications to facilitate cooling, ventilating, or heating for server racks or cabinets; for data centers, e.g. 19-inch computer racks
- H05K7/208—Liquid cooling with phase change
- H05K7/20827—Liquid cooling with phase change within rooms for removing heat from cabinets, e.g. air conditioning devices
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20709—Modifications to facilitate cooling, ventilating, or heating for server racks or cabinets; for data centers, e.g. 19-inch computer racks
- H05K7/20836—Thermal management, e.g. server temperature control
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Thermal Sciences (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses a predictive control method of a data center air conditioning system, which comprises the steps that a predictive controller carries out online rolling optimization on each parameter of a field controller by utilizing a predictive control optimization objective function, the optimized parameter is taken as a set value of a controlled variable and is transmitted to the field controller of a field control layer, and the field controller controls a refrigerating device based on the set value of the optimized controlled parameter. The prediction controller performs rolling optimization calculation on the optimal set value at each sampling moment based on the prediction result, and the field controller utilizes a PID algorithm to enable the controlled parameter to quickly track the optimal set value, so that the large time lag and strong coupling characteristics of the data center air-conditioning refrigeration station system are overcome, and the control precision is improved; the optimization objective function comprehensively considers the energy efficiency index PUE and the composite demand, so that the system can reduce the energy consumption of the air conditioning system, reduce the PUE and achieve the purpose of energy saving on the premise of meeting the refrigeration demand of a machine room.
Description
Technical Field
The invention belongs to the technical field of data center air conditioner control, and particularly relates to a prediction control method of a data center air conditioner system.
Background
With the continuous advancing development of high and new technologies of computer internet, the data center industry is facing the peak period of construction. In parallel, the high energy consumption problem of data centers has entered the field of view of everyone. Data show that the energy consumption of the data center accounts for about 20% of the energy consumption of the whole building industry, and the energy consumption of the air conditioning system is as high as 40% -50%, so that the high energy consumption becomes the largest brake which restricts the green and efficient development of the data center. In the situation that the PUE value of a data center is more than 80% and the energy efficiency is low at present, the optimization transformation and the energy-saving control strategy research based on the air conditioning system of the data center are generally concerned by the people in the industry. However, by only modifying the internal structure design of the air conditioning system, on one hand, the increase of the basic investment is brought, and compared with the economic benefit, the energy-saving benefit is not obvious; on the other hand, the data center has different requirements on climate, region and environment, and the applicability to actual engineering is not high. Therefore, only by adopting an optimization control strategy, the cooling requirement of the IT equipment of the machine room can be met, the energy consumption of the air conditioning system of the data center is reduced to the maximum extent, and the energy saving of the data center is realized in the true sense, but in the prior art, a data center air conditioning system-based optimization transformation and energy saving control system and method are not provided, for example, in the Chinese patent application No. 201611079041.X, a control method of the air conditioning system of the data center is disclosed, and the data center air conditioning system control method is characterized by further comprising a monitoring module, a software filtering module and an output module, wherein the monitoring module is used for transmitting a monitored signal to a controller after being filtered by the software filtering module; the software filtering module is used for filtering the signal; the controller is used for identifying and processing the signals and then sending control signals to each module of the data center air conditioning system through the output module; the output module is used for outputting a control signal of the controller. The purpose of energy saving cannot be achieved simply by effectively monitoring and adjusting the data center.
Therefore, how to solve the above problems becomes a focus of research by those skilled in the art.
Disclosure of Invention
The invention aims to provide a predictive control method for a data center air conditioning system, which can effectively overcome the defect that the data center air conditioning system cannot save energy.
The purpose of the invention is realized by the following technical scheme:
a prediction control method for a data center air conditioning system comprises the steps that a prediction controller carries out online optimization on each parameter of a field controller by using a prediction control optimization objective function, the parameter obtained by optimization is used as a set value of the field controller, the field controller of a field control layer controls a refrigerating device on the basis of the set value of a controlled parameter after optimization, and the optimization objective function is
Where M is the prediction time domain and t is the initial value in the prediction time domainAt the beginning of the run, PUEset(k) The set value of the PUE in the kth sampling period, and PUE (k) is the actual value of the PUE in the kth sampling period;
set point of air supply temperature, T, of data center for k-th sampling periodAS(k) Is the average value of the supply air temperature of the data center of the k sampling period, JaFor optimizing the objective function, α is a PUE penalty weighting factor, β is a blowing temperature penalty weighting factor, the refrigeration devices comprise an individual refrigeration device, a first combined refrigeration device and a second combined refrigeration device, the predictive controller adopts a multilayer neural network, and the control predictive control method comprises the following steps:
s1, under the working condition of summer, the independent refrigerating unit carries out independent refrigeration;
s2, under the working condition of the transition season, the first combined refrigerating device performs independent refrigeration;
s3, under the working condition of winter, the second combined refrigerating device carries out independent refrigeration;
the system comprises a prediction controller, a field controller, a first combined refrigerating device, a second combined refrigerating device, a third combined refrigerating device, a fourth combined refrigerating device and a fourth combined refrigerating device, wherein the prediction controller adopts a multilayer neural network, and when the single refrigerating device and the first combined refrigerating device work, the prediction controller transmits the water supply temperature of chilled water, the water supply; taking the outdoor wet bulb temperature, the load of a data center machine room, the PUE value, the air supply temperature at the current moment, the PUE set value at the next moment, the air supply temperature set value at the next moment and-1 related to a threshold value as the input of a field controller; when the second combined refrigerating device works, the input of the prediction controller is the same as that of the single refrigerating device and the first combined refrigerating device, and the output of the prediction controller is chilled water supply temperature, chilled water supply and return water pressure difference, cooling tower outlet water temperature and return air temperature difference.
Furthermore, the independent refrigerating device is a refrigerating unit, the first combined refrigerating device comprises the refrigerating unit, a plate heat exchanger and a cooling tower, and the second combined refrigerating device comprises the cooling tower and the plate heat exchanger.
Furthermore, the working condition of 6-8 months in a year is a summer working condition, the working condition of 12-2 months in a winter working condition, and the working conditions of other months in transition seasons.
Furthermore, the field controller controls the refrigerating device by a PID control method.
Furthermore, the predictive controller utilizes a predictive control strategy to set the set value of the controlled parameter of the field controller.
Compared with the prior art, the invention has the beneficial effects that:
aiming at a chilled water cooling type air conditioning system of a data center, the invention takes the deviation square sum of actual values and expected values of two parameters of the air supply temperature of a machine room and an energy efficiency index PUE (Power Usage efficiency) as a prediction control optimization performance index. The control system adopts a two-layer control structure, the upper optimization layer uses a predictive control algorithm to optimize the set value of the bottom layer field controller, the established predictive model is utilized, the rolling optimization strategy is adopted to calculate the optimal set value of the controlled parameter of the field controller, the bottom layer controller utilizes a PID algorithm to enable the controlled parameter to quickly track the optimal set value, the system equipment is enabled to operate in an energy-saving operation state, and the global optimization control of the data center air conditioning system aiming at meeting the time-varying cooling capacity requirement and saving energy is realized.
Drawings
FIG. 1 is a diagram of a neural network predictive model architecture for a data center air conditioning system during summer and transition season operating conditions;
FIG. 2 is a block diagram of a data center air conditioning system neural network prediction model under winter conditions;
FIG. 3 is a diagram of a neural network topology for a predictive controller in summer and transition season conditions;
FIG. 4 is a topological structure diagram of a neural network predictive controller under winter conditions;
FIG. 5 is a block diagram of a predictive control system.
Detailed Description
The invention will be further described with reference to specific embodiments and the accompanying drawings.
Example one
As shown in fig. 1 to 5, a predictive control method for an air conditioning system of a data center includes a control system and a controlled system, where the control system includes an upper optimization layer and a lower field control layer, the optimization layer is provided with a predictive control system, the predictive control system includes a predictive controller, the field control layer includes a field controller, the field controller is connected to a refrigeration device, and the refrigeration device includes a single refrigeration device; the refrigerating device comprises a refrigerating unit, a plate heat exchanger and a cooling tower combined refrigerating device; the refrigerating device comprises a cooling tower and a plate heat exchanger refrigerating device, the three refrigerating devices can independently run, the optimization layer is connected with a field control layer, the optimization layer optimizes the set value of the controlled parameter of a field control layer controller by using a prediction control strategy, the optimization target is to enable the PUE value and the air supply temperature of the system to reach the set value, and the purposes of saving energy and meeting the requirement of cooling capacity are achieved, the controlled parameter of the field control layer comprises chilled water supply temperature, cooling water supply temperature, chilled pump flow control parameter, cooling pump flow control parameter and air supply flow control parameter, wherein the flow control parameter can select different variables according to different control methods of the field control layer, can be pressure difference, temperature difference or frequency control, preferably chilled water supply and return water pressure difference, cooling water supply and return water temperature difference and air supply and return temperature difference, and the controller of the field control layer is based on the optimized set value of the controlled, and the refrigerating machine, the cooling tower, the freezing water pump and the cooling water pump are controlled by adopting a PID control method.
In this embodiment, the purpose of energy saving is achieved by controlling different refrigeration devices to switch operation.
Example two
As shown in fig. 1 to 5, a predictive control method for an air conditioning system of a data center includes a control system and a controlled system, where the control system includes an upper optimization layer and a lower field control layer, the optimization layer is provided with a predictive control system, the predictive control system includes a predictive controller, the field control layer includes a field controller, the field controller is connected to a refrigeration device, and the refrigeration device includes a single refrigeration device; the first combined refrigerating device comprises a refrigerating unit, a plate heat exchanger and a cooling tower combined refrigerating device; the second combined refrigerating device comprises a cooling tower and a plate heat exchanger refrigerating device, the three refrigerating devices can independently operate, the optimization layer is connected with a field control layer, the optimization layer optimizes the set value of the controlled parameter of a field control layer controller by using a prediction control strategy, the optimization target is to enable the PUE value and the air supply temperature of the system to reach the set value, and the purposes of saving energy and meeting the requirement of cold capacity are achieved, the controlled parameters of the field control layer comprise chilled water supply temperature, cooling water supply temperature, chilled pump flow control parameters, cooling pump flow control parameters and air supply flow control parameters, wherein the flow control parameters can select different variables according to different control methods of the field control layer, can be pressure difference, return air temperature or frequency control, preferably chilled water supply and return water pressure difference, cooling water supply and return water temperature difference and air supply temperature difference, and the controller of the field control layer is based on the optimized set value of the controlled parameter, the invention integrates the plate heat exchanger on the water side, so that the system can fully utilize the natural cold source for cooling when the outdoor wet bulb temperature is lower, and the running power consumption of the refrigerator is reduced. The invention divides the controlled system into three working conditions to operate independently, namely a summer working condition, a transition season working condition and a winter working condition. The refrigerating unit performs independent refrigeration under the working condition in summer; refrigerating unit + plate heat exchanger (water side economizer) + cooling tower are jointly refrigerated under the working condition of transition season; the cooling tower and the plate heat exchanger (water-side economizer) are completely and naturally cooled under the working condition of winter. The latter two working conditions can use natural cold source in transition season with lower outdoor temperature and winter to realize partial or complete natural cooling, so as to reduce the running time of the water chilling unit and reduce the annual running power consumption of the air conditioning system.
The operation time of the three working conditions is divided into 6-8 months (3624-5832 h) as the summer working condition, 12-2 months (8016-1416 h) as the winter working condition and other months merging into the transition season working condition according to the climate condition of the Beijing area within 8760 hours in a year.
In the embodiment, different refrigerating devices are selected according to different outdoor temperatures in different seasons, and a natural cold source can be utilized for refrigerating to save energy.
EXAMPLE III
As shown in fig. 1 to 5, a predictive control method for an air conditioning system of a data center includes a control system and a controlled system, where the control system includes an upper optimization layer and a lower field control layer, the optimization layer is provided with a predictive control system, the predictive control system includes a predictive controller, the field control layer includes a field controller, the field controller is connected to a refrigeration device, and the refrigeration device includes a single refrigeration device; the refrigerating device comprises a refrigerating unit, a plate heat exchanger and a cooling tower combined refrigerating device; the refrigerating device comprises a cooling tower and a plate heat exchanger refrigerating device, the three refrigerating devices can independently run, the optimization layer is connected with a field control layer, the optimization layer optimizes the set value of the controlled parameter of a field control layer controller by using a prediction control strategy, the optimization target is to enable the PUE value and the air supply temperature of the system to reach the set value, and the purposes of saving energy and meeting the requirement of cooling capacity are achieved, the controlled parameter of the field control layer comprises chilled water supply temperature, cooling water supply temperature, chilled pump flow control parameter, cooling pump flow control parameter and air supply flow control parameter, wherein the flow control parameter can select different variables according to different control methods of the field control layer, can be pressure difference, temperature difference or frequency control, preferably chilled water supply and return water pressure difference, cooling water supply and return water temperature difference and air supply and return temperature difference, and the controller of the field control layer is based on the optimized set value of the controlled, the invention integrates the plate heat exchanger on the water side, so that the system can fully utilize the natural cold source for cooling when the outdoor wet bulb temperature is lower, and the running power consumption of the refrigerator is reduced. The invention divides the controlled system into three working conditions to operate independently, namely a summer working condition, a transition season working condition and a winter working condition. The refrigerating unit performs independent refrigeration under the working condition in summer; refrigerating unit + plate heat exchanger (water side economizer) + cooling tower are jointly refrigerated under the working condition of transition season; the cooling tower + plate heat exchanger (water side economizer) is completely free to supply cold under the working condition of winter. The latter two working conditions can use natural cold source in transition season with lower outdoor temperature and winter to realize partial or complete natural cooling, so as to reduce the running time of the water chilling unit and the annual running power consumption of the air conditioning system. At this time, in summer and transitional season, the input parameters of the prediction model include the supply water temperature of chilled water, the supply return water pressure difference of chilled water, the outlet water temperature of the cooling tower, the supply return water temperature difference of cooling water, the supply air temperature, the return air temperature difference, the load (rate), the outdoor wet bulb temperature, and the PUE value of the system at the current time and-1 related to the threshold, the output parameters of the prediction model include the PUE value and the supply air temperature at the next time, the neural network prediction model structure of the data center air conditioning system in summer and transitional season is shown in fig. 1, in winter, the input parameters of the prediction model can reduce the control variable of the supply return water temperature difference of cooling water, and the neural network prediction model structure of the data center air conditioning system in winter is shown in fig. 2.
In this embodiment, different variables are selected by the flow control parameters in the input parameters of the prediction model, and each parameter of the field controller is set as an optimized parameter.
Example four
As shown in fig. 1 to 5, a predictive control method for an air conditioning system of a data center includes a control system and a controlled system, where the control system includes an upper optimization layer and a lower field control layer, the optimization layer is provided with a predictive control system, the predictive control system includes a predictive controller, the field control layer includes a field controller, the field controller is connected to a refrigeration device, and the refrigeration device includes a single refrigeration device; the refrigerating device comprises a refrigerating unit, a plate heat exchanger and a cooling tower combined refrigerating device; the refrigerating device comprises a cooling tower and a plate heat exchanger refrigerating device, the three refrigerating devices can independently run, the optimization layer is connected with a field control layer, the optimization layer optimizes the set value of the controlled parameter of a field control layer controller by using a prediction control strategy, the optimization target is to enable the PUE value and the air supply temperature of the system to reach the set value, and the purposes of saving energy and meeting the requirement of cooling capacity are achieved, the controlled parameter of the field control layer comprises chilled water supply temperature, cooling water supply temperature, chilled pump flow control parameter, cooling pump flow control parameter and air supply flow control parameter, wherein the flow control parameter can select different variables according to different control methods of the field control layer, can be pressure difference, temperature difference or frequency control, preferably chilled water supply and return water pressure difference, cooling water supply and return water temperature difference and air supply and return temperature difference, and the controller of the field control layer is based on the optimized set value of the controlled, the invention integrates the plate heat exchanger on the water side, so that the system can fully utilize the natural cold source for cooling when the outdoor wet bulb temperature is lower, and the running power consumption of the refrigerator is reduced. The invention divides the controlled system into three working conditions to operate independently, namely a summer working condition, a transition season working condition and a winter working condition. The refrigerating unit performs independent refrigeration under the working condition in summer; refrigerating unit + plate heat exchanger (water side economizer) + open cooling tower are jointly refrigerated under the working condition of transition season; the cooling tower + plate heat exchanger (water side economizer) is completely free to supply cold under the working condition of winter. The latter two working conditions can use natural cold sources in transition seasons and winter with lower outdoor temperature, realize partial or complete natural cooling, so as to reduce the running time of the water chilling unit and reduce the annual running power consumption of the air conditioning system, and the target function of the predictive control and optimization of the data center air conditioning system is as follows:
where k represents the current time, M is the prediction time domain, t is the initial time in the prediction time domain, PUEset(k) The set value of the PUE in the kth sampling period, and PUE (k) is the actual value of the PUE in the kth sampling period;
is a set value T of the air supply temperature of the data center machine room in the kth sampling periodAS(k) For the average value of the supply air temperature of the data center machine room in the kth sampling period, a plurality of prediction controllers are selectedThe number of input layer nodes is 5, the number of output layer nodes is 5, 8 hidden layer nodes are determined according to an empirical method and a trial-and-error method, a tan sig function is used as a hidden layer excitation function, a linear function is used as an output layer excitation function, output variables of the prediction controller under working conditions of summer and transition seasons are chilled water supply water temperature, chilled water supply and return water pressure difference, cooling tower outlet water temperature, cooling water supply and return water temperature difference and return air temperature difference at the current moment, the output variables are used as controller outputs, and the controller outputs the variables to be transmitted to a field controller as set values of controlled parameters; the outdoor wet bulb temperature, the refrigerating load (rate) of the air conditioner room, the PUE value, the air supply temperature at the current moment, the PUE set value at the next moment, the air supply temperature set value at the next moment and-1 related to the threshold are taken as the input of the controller. FIG. 3 is a neural network topology of a predictive controller during summer and transition season conditions. The input of the prediction controller under the working condition in winter is the same as that of summer and transition seasons, the output is chilled water supply temperature, chilled water supply and return water pressure difference, cooling tower outlet water temperature and return air temperature difference, the topological structure of the neural network prediction controller under the working condition in winter is shown in figure 4, the block diagram of the prediction control system is shown in figure 5, wherein x [ k ] is]Relevant state variable parameters of the data center air conditioning system at the moment k, namely the chilled water supply water temperature, the chilled water supply and return water pressure difference, the cooling tower outlet water temperature, the cooling water supply and return water temperature difference, the outdoor wet bulb temperature, the air supply temperature, the air return temperature difference, the load (rate) and the PUE value, wherein the load (rate) and the PUE value are calculated values according to the power consumption and the cooling capacity provided by the refrigeration station, and the balance of the load (rate) and the PUE value can be obtained through actual measurement;
predicting model output at the k +1 moment, namely a PUE value and air supply temperature at the next moment; x [ k +1]]The PUE set value at the next moment and the air supply temperature set value of the machine room at the next moment; u [ k ]]The optimal control quantity at the time k after the optimization of the neural network prediction controller is finished, namely the set value of the controlled parameter of the controller of the field control layer; u' [ k + i-1 ]]And the control quantity is calculated according to the weight of the predictive controller at the previous moment in the rolling optimization process, and the control quantity is not optimized according to the variable value of the system running state at the current moment.
The optimization steps of the neural network controller can be summarized as follows:
① initializing each connection weight of the neural network prediction controller of the data center air conditioning system, assigning as a smaller random number in the range of [ -1,1] and a threshold value (taking-1 in the description), then calculating the controller output, the prediction time domain M is 6, and the prediction period takes 5 minutes;
② acting x [ k ], x [ k +1] and-1 on the neural network prediction controller, and optimizing by using the neural network to obtain a control variable output u [ k ];
③ outputs the control quantity u k]And initial quantity of state x k]Transmitting to the controlled object to obtain the actual output x [ k +1] of the controlled system]Simultaneously transmitting the data to a prediction model to obtain the output of the prediction model
④ weight W of neural network structure of prediction controllerjiInvariably, let x [ k +1]]、x*[k+2]-1 is transmitted to the predictive controller to obtain a new control vector
Will be provided with
And x [ k +2 ]]Transmitting to neural network prediction model, calculating
Repeating the steps until the control output is
And the predicted output is
⑤ when k is less than N-M, λ k]＝λ[t1+M]When k is not less than N-M, λ k]＝λ[N]Repeatedly using ③ - ④ steps to back-out L agarange multiplier vector lambda k]And gamma k]I.e. in the order k ═ t1+M-1,…t1+2,t1+1 is derived forward:
wherein L (-) represents the optimization performance index, f (-) represents the prediction model, g (-) represents the neural network controller, u k in the derivation process]And x [ k ]]By using
And
replacement;
⑥ finally, using the calculated γ [ k ], the weights of the neural network structure of the predictive controller are modified by:
W＝W+ΔW
wherein W is a weight matrix of the neural network predictive controller of the air-conditioning system, mu is a weight update rate, and mu is 0.05;
⑦ and looping ④ - ⑥ steps to continuously modify the weight of the neural network predictive controller until Δ W is 0
⑧, k is k +1, loop ② - ⑦ steps, and control variable u (k) of the controlled system at each moment is obtained.
And repeating the operation when the next sampling period is up, and respectively calculating the values of the control quantity at the later moments until the control process is finished.
In the embodiment, the optimized objective function is utilized, the controlled parameters are quickly tracked to the optimal set value through repeated optimization processing, the system equipment is operated in an energy-saving operation state, and the global optimization control of the data center air conditioning system aiming at meeting the time-varying cooling capacity requirement and saving energy is realized.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (5)
1. A predictive control method for a data center air conditioning system is characterized by comprising the following steps: the method comprises the steps that a prediction controller utilizes a prediction control optimization objective function to carry out online optimization on each parameter of a field controller, the parameter obtained by optimization is used as a set value of the field controller, the field controller of a field control layer controls a refrigerating device based on the set value of a controlled parameter after optimization, and the optimization objective function is
Where M is the prediction time domain, t is the initial time in the prediction time domain, PUEset(k) The set value of the PUE in the kth sampling period, and PUE (k) is the actual value of the PUE in the kth sampling period; t isASset(k) Set point of air supply temperature, T, of data center for k-th sampling periodAS(k) Is the average value of the supply air temperature of the data center of the k sampling period, JaFor optimizing the objective function, α is a PUE penalty weighting factor, β is a blowing temperature penalty weighting factor, the refrigeration devices comprise an individual refrigeration device, a first combined refrigeration device and a second combined refrigeration device, the predictive controller adopts a multilayer neural network, and the control predictive control method comprises the following steps:
s1, under the working condition of summer, the independent refrigerating unit carries out independent refrigeration;
s2, under the working condition of the transition season, the first combined refrigerating device performs independent refrigeration;
s3, under the working condition of winter, the second combined refrigerating device carries out independent refrigeration;
the system comprises a prediction controller, a field controller, a first combined refrigerating device, a second combined refrigerating device, a third combined refrigerating device, a fourth combined refrigerating device and a fourth combined refrigerating device, wherein the prediction controller adopts a multilayer neural network, and when the single refrigerating device and the first combined refrigerating device work, the prediction controller transmits the water supply temperature of chilled water, the water supply; taking the outdoor wet bulb temperature, the load of a data center machine room, the PUE value, the air supply temperature at the current moment, the PUE set value at the next moment, the air supply temperature set value at the next moment and-1 related to a threshold value as the input of a field controller; when the second combined refrigerating device works, the input of the prediction controller is the same as that of the single refrigerating device and the first combined refrigerating device, and the output of the prediction controller is chilled water supply temperature, chilled water supply and return water pressure difference, cooling tower outlet water temperature and return air temperature difference.
2. The predictive control method for a data center air conditioning system as claimed in claim 1, wherein the individual refrigeration units are refrigeration units, the first combined refrigeration unit comprises a refrigeration unit, a plate heat exchanger and a cooling tower, and the second combined refrigeration unit comprises a cooling tower and a plate heat exchanger.
3. The predictive control method for a data center air conditioning system of claim 1, wherein months 6-8 of the year are summer conditions, months 12-2 are winter conditions, and other months are transition season conditions.
4. The predictive control method for a data center air conditioning system as claimed in claim 1, wherein said site controller controls the chiller plant using PID control.
5. The predictive control method for a data center air conditioning system as claimed in claim 1, wherein the predictive controller sets the controlled parameter set value of the site controller using a predictive control strategy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810927800.6A CN108990383B (en) | 2018-08-15 | 2018-08-15 | Predictive control method for air conditioning system of data center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810927800.6A CN108990383B (en) | 2018-08-15 | 2018-08-15 | Predictive control method for air conditioning system of data center |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108990383A CN108990383A (en) | 2018-12-11 |
| CN108990383B true CN108990383B (en) | 2020-08-04 |
Family
ID=64553480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810927800.6A Active CN108990383B (en) | 2018-08-15 | 2018-08-15 | Predictive control method for air conditioning system of data center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108990383B (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109890176B (en) * | 2019-03-01 | 2020-10-27 | 北京慧辰资道资讯股份有限公司 | Device for optimizing energy consumption efficiency of machine room based on artificial intelligence |
| CN109917656B (en) * | 2019-03-29 | 2022-03-01 | 重庆大学 | Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target |
| CN110008575B (en) * | 2019-03-29 | 2023-01-31 | 重庆大学 | Multi-parameter predictive control algorithm for switching process medium multi-temperature target set values of circulating cooling water system |
| CN110594983B (en) * | 2019-09-20 | 2021-04-02 | 东北大学 | Temperature control method suitable for small data center |
| CN110686382B (en) * | 2019-10-16 | 2021-02-09 | 广东美的暖通设备有限公司 | Air conditioner control method and device and computer readable storage medium |
| CN111144543A (en) * | 2019-12-30 | 2020-05-12 | 中国移动通信集团内蒙古有限公司 | Data center air conditioner tail end temperature control method, device and medium |
| CN111649465B (en) * | 2020-06-05 | 2022-04-08 | 哈尔滨工业大学 | Automatic control method and system for air conditioning equipment |
| CN111561772B (en) * | 2020-07-15 | 2020-10-02 | 上海有孚智数云创数字科技有限公司 | Cloud computing data center precision air conditioner energy-saving control method based on data analysis |
| CN112084661B (en) * | 2020-09-10 | 2022-07-05 | 中国船舶重工集团海装风电股份有限公司 | Wind turbine converter water cooling system cooling state assessment early warning method |
| CN112487353B (en) * | 2020-11-30 | 2022-08-09 | 广东美的制冷设备有限公司 | Data processing method and device, dehumidifier and storage medium |
| CN113283156B (en) * | 2021-03-29 | 2023-09-15 | 北京建筑大学 | Energy-saving control method for subway station air conditioning system based on deep reinforcement learning |
| CN114001443A (en) * | 2021-09-27 | 2022-02-01 | 中盈优创资讯科技有限公司 | Machine room energy saving method and device based on AI |
| CN114167964B (en) * | 2021-12-01 | 2024-05-24 | 上海德衡数据科技有限公司 | Wind side natural cooling method and system based on data center |
| CN114322209A (en) * | 2021-12-27 | 2022-04-12 | 上海美控智慧建筑有限公司 | Data processing method and device of ice storage air conditioning system and electronic equipment |
| CN114462171B (en) * | 2022-01-05 | 2024-05-24 | 浙江力德节能科技有限公司 | Calculation method for supply and return water temperature difference predicted value of central air conditioner on demand side |
| CN115623753B (en) * | 2022-10-27 | 2024-03-15 | 南京群顶科技股份有限公司 | Water-cooling energy-saving timing regulation and control method based on dynamic time table |
| CN115696871B (en) * | 2022-11-04 | 2023-06-13 | 中国电子工程设计院有限公司 | Data center water cooling system regulation and control method and device based on machine learning |
| CN115581056B (en) * | 2022-11-09 | 2023-03-21 | 宁波亮控信息科技有限公司 | Energy-saving prediction control method and system suitable for water cooling system of data center |
| CN117389146A (en) * | 2023-11-09 | 2024-01-12 | 北京建筑大学 | Nonlinear model predictive control system for cooling unit system of data center |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106483843A (en) * | 2015-08-31 | 2017-03-08 | 中兴通讯股份有限公司 | A kind of method and system realizing IDC power consumption control |
| CN107169562A (en) * | 2017-06-15 | 2017-09-15 | 郑州云海信息技术有限公司 | A kind of energy efficiency controlling method and device for data center |
| CN108089440A (en) * | 2017-12-06 | 2018-05-29 | 北京百度网讯科技有限公司 | Energy-saving control method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102620378B (en) * | 2011-01-27 | 2014-01-15 | 国际商业机器公司 | Method and system for data center energy saving controlling |
| US10001761B2 (en) * | 2014-12-30 | 2018-06-19 | Schneider Electric It Corporation | Power consumption model for cooling equipment |
-
2018
- 2018-08-15 CN CN201810927800.6A patent/CN108990383B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106483843A (en) * | 2015-08-31 | 2017-03-08 | 中兴通讯股份有限公司 | A kind of method and system realizing IDC power consumption control |
| CN107169562A (en) * | 2017-06-15 | 2017-09-15 | 郑州云海信息技术有限公司 | A kind of energy efficiency controlling method and device for data center |
| CN108089440A (en) * | 2017-12-06 | 2018-05-29 | 北京百度网讯科技有限公司 | Energy-saving control method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108990383A (en) | 2018-12-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108990383B (en) | Predictive control method for air conditioning system of data center | |
| CN110288164B (en) | Predictive control method for building air-conditioning refrigeration station system | |
| CN111256294B (en) | Model prediction-based optimization control method for combined operation of water chilling unit | |
| CN108413567B (en) | Central air conditioner cost-saving optimization method and system based on Internet of things | |
| CN103912966B (en) | A kind of earth source heat pump refrigeration system optimal control method | |
| CN110392515A (en) | A kind of Cooling and Heat Source equipment room energy-conserving control method and system based on historical data | |
| CN102980272B (en) | Air conditioner system energy saving optimization method based on load prediction | |
| CN110895016A (en) | Fuzzy self-adaptive based energy-saving group control method for central air-conditioning system | |
| CN101782261B (en) | Nonlinear self-adapting energy-saving control method for heating ventilation air-conditioning system | |
| CN201666640U (en) | Control system of central air conditioner based on wireless sensor network | |
| CN110410942A (en) | A kind of Cooling and Heat Source machine room energy-saving optimal control method and system | |
| CN116398994B (en) | Water chilling unit group control optimization method based on load prediction | |
| CN110486896B (en) | Cascade air conditioning system optimization control method based on water chilling unit energy consumption model | |
| CN110895029A (en) | Building load prediction method based on temperature of chilled water | |
| CN116629021A (en) | Energy consumption prediction and energy-saving control strategy research method for cooling system | |
| CN114942584A (en) | Efficient machine room energy-saving control method based on magnetic suspension water chilling unit load optimization | |
| CN116734404A (en) | Energy-saving optimization control system and optimization method for central air conditioning system | |
| CN201944984U (en) | Energy-saving optimization control system of refrigerator room | |
| CN118361827A (en) | Energy-saving optimal control method and system for water-cooling central air conditioner and network side server | |
| CN115903712A (en) | Energy-saving optimization method and optimization control system suitable for industrial refrigeration system | |
| CN107843036B (en) | Refrigerating unit most energy-saving control method and equipment based on big data management | |
| CN110848889A (en) | Method for evaluating operation energy efficiency of main unit of central air-conditioning system | |
| CN117930659A (en) | Efficient machine room global optimizing control method | |
| CN117366799A (en) | Control method of building cold source system | |
| CN116592546A (en) | Q-learning-based multi-target refrigeration side water supply temperature control optimization method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |