WO2018122635A1

WO2018122635A1 - Method and system for regulating temperature of data center

Info

Publication number: WO2018122635A1
Application number: PCT/IB2017/055825
Authority: WO
Inventors: Nareshkumar NANDOLA; Shrikant Bhat; Niket KAISARE; Wilhelm Wiese
Original assignee: Abb Schweiz Ag
Priority date: 2016-12-28
Filing date: 2017-09-26
Publication date: 2018-07-05

Abstract

The present disclosure discloses a method and a system for regulating temperature in a data center. The method comprises identifying one or more target locations in the data center using a heat map of the data center and temperature measurements at various locations in the data center. The method further comprises generating a set point for each of the one or more target locations. In addition the method comprises selecting a control model from a model bank, corresponding to each of the one or more target locations for controlling temperature of the one or more target locations based on the set point.

Description

METHOD AND SYSTEM FOR REGULATING TEMPERATURE OF DATA

CENTER

TECHNICAL FIELD

The present disclosure relates to temperature control techniques in a data center. Particularly, but not specifically, the present disclosure relates to controlling temperature at target locations in a data center.

BACKGROUND

A data center has varying temperature due to high computation and processing of data. Generally, temperature regulation of the data center has always been challenging due to highly distributed nature of heat pattern and non-uniformity of heat generation at various locations of the data center. Further, number of sensors in a data center are limited, and hence temperature cannot be measured at every location of the data center. As a result, traditional cooling systems calculate average of temperature measurements received from limited number of sensors. Thus, specific location where data cooling or heating is required cannot be determined using traditional systems.

Some conventional systems use offline modeling techniques to locate specific sensors which have detected temperature deviation from a set point. Further, the conventional systems identify locations of the sensors and control temperature at locations of the sensors. Thus, temperature of target locations where temperature has deviated from the set point is not regulated. As a result, the target locations are under cooled or over cooled.

SUMMARY

In an embodiment, the present disclosure relates to a method for regulating temperature in a data center. The method comprises generating a heat map of the data center using one or more inputs. Further, one or more target locations in the data center are identified using the heat map. Thereafter, a set point for each of the one or more target locations is generated. In one embodiment, the set point is generated by selecting a control model from a model bank, corresponding to each of the one or more target locations. Lastly, temperature of the one or more target locations is controlled using the selected control model based on the set point.

In an embodiment, the present disclosure relates to a system for regulating temperature of a data center. The system comprises a plurality of sensors for measuring temperature of the data center at various locations and a controller configured to generate a heat map of the data center using one or more inputs from the plurality of sensors and one or more components of the data center. Further, one or more target locations in the data center are identified using the heat map. Then, a set point for each of the one or more target locations is generated. In one embodiment, the set point is generated by selecting a control model. Thereafter, a control model from a model bank corresponding to each of the one or more target locations is selected to control temperature of the one or more target locations using the selected control model based on the set point.

The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.

BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS

The novel features and characteristic of the disclosure are set forth in the appended claims. The disclosure itself, however, as well as a best mode of use, further objectives and advantages thereof, will best be understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying figures. One or more embodiments are now described, by way of example only, with reference to the accompanying figures wherein like reference numerals represent like elements and in which:

Figure 1A is illustrative of an exemplary environment for regulating temperature of a data center, in accordance with embodiments of the present disclosure;

Figure IB is an exemplary illustration of target locations in a data center, in accordance with embodiments of the present disclosure;

Figure 1C is an exemplary illustration of connectivity between a controller of a data center and a server and display unit, in accordance with embodiments of the present disclosure;

Figure ID is an exemplary illustration of a display device displaying hotspots details, in accordance with some embodiments of the present disclosure; Figure 2 is an exemplary representation of internal architecture of a controller used for regulating temperature of a data center, in accordance with embodiments of the present disclosure;

Figure 3 is a flow chart illustrative of method steps for regulating temperature of a data center, in accordance with embodiments of the present disclosure;

Figure 4 is a hybrid diagram illustrative of generating heat map of a data center, in accordance with embodiments of the present disclosure;

Figure 5 is a hybrid diagram illustrative of generating a control model bank, in accordance with embodiments of the present disclosure;

Figure 6A is illustrative of an exemplary block diagram for cooling the data center racks using robotic cooling jets, in accordance with some embodiments of the present disclosure;

Figure 6B is illustrative of a graph indicating temperature distribution in a data center, in accordance with some embodiments of the present disclosure; and

Figure 7 is illustrative of an exemplary block diagram for cooling the data center racks using flow channels, in accordance with some embodiments of the present disclosure.

It should be appreciated by those skilled in the art that any block diagrams herein represent conceptual views of illustrative systems embodying the principles of the present subject matter. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in computer readable medium and executed by a computer or processor, whether or not such computer or processor is explicitly shown.

DETAILED DESCRIPTION

In the present document, the word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any embodiment or implementation of the present subject matter described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. While the disclosure is susceptible to various modifications and alternative forms, specific embodiment thereof has been shown by way of example in the drawings and will be described in detail below. It should be understood, however that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternative falling within the scope of the disclosure.

The terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a setup, device or method that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such setup or device or method. In other words, one or more elements in a system or apparatus proceeded by "comprises... a" does not, without more constraints, preclude the existence of other elements or additional elements in the system or apparatus.

Embodiments of the present disclosure relate to a method and a system for regulating temperature of a data center. The system comprises a plurality of sensors for measuring temperature of the data center at various locations. Further, a model is used to generate a heat map of the data center. The heat map is indicative of temperature variation of the data center at the various locations. Using the model, target locations are identified where temperature has to be regulated. In one embodiment, regulation of temperature includes increasing the temperature and decreasing the temperature from current temperature reading. The system uses control mechanisms to identify a control model applicable to corresponding target locations. Using the corresponding control models, values of manipulated variables are calculated for regulating temperature of the target locations.

Figure 1A shows an exemplary block diagram of a system for regulating temperature of target locations in a data center. The system comprises a controller 100, a plurality of sensors 101, and a plurality of actuators 102. In an embodiment, the controller is one of a multivariate controller, a Model Predictive Controller (MPC), an Internal Model Controller (IMC), one or more Proportional Integral Derivative (PID) controllers and the like.

The plurality of sensors 101 are installed in a data center 103 to measure temperature at various locations. The plurality of sensors 101 are configured to take temperature measurements at predefined intervals of time. Then, the measured temperature is provided as input to the controller 100. The controller 100 also receives inputs such as IT load and ambient temperature as inputs (e.g. with programs running on the corresponding devices or from a server connected thereto). The measured temperature, the ambient temperature and the IT load (e.g. of servers, storages etc. of the data center) can be denoted individually / jointly as one or more inputs henceforth in the present disclosure. The controller 100 uses the one or more inputs (e.g. temperature from sensors) to generate a heat map of the data center 103. Here, the controller has computational resources (e.g. with a server (not illustrated in Figure 1 A)) to carry out computations as required for control. The heat map is indicative of temperature variation and temperature values of various locations of the data center 103. Further, target locations are identified using the heat map. The target locations indicate locations in the data center 103 where temperature regulation may be required. Further, the controller 100 identifies a control model from a model bank for regulating temperature at the target locations.

In an embodiment, the heat map may be generated using existing heat map generation models, for example Computational Fluid Dynamics (CFD) model or the like. Also, any models that can generate a heat map of a data center 103 may be used in accordance with aspects of the present disclosure.

In an embodiment, the heat map generation and the model bank generation may take place either online or offline. In an embodiment, the heat map generation model and the plurality of control models may be parametrized models.

In an embodiment, the models may be implemented by at least one of a Distributed Control System (DCS), a Personal Computer (PC), and a cloud controller (or virtual controller).

In an embodiment, each sensor of the plurality of sensors is a temperature sensor. In another embodiment, the plurality of sensors includes, but are not limited to, temperature sensors, flow sensors (air flow), and humidity sensors. The plurality of sensors may be distributed across various locations such as at air vents, at bottom, top and middle of a rack, in between racks and so forth.

In an embodiment, the target locations may be certain locations in the data center 103 with maximum temperature or certain locations having a range of temperatures. For example, locations with top 5 temperature values may be considered as target locations. In an embodiment, the target locations may be locations in the data center 103 with temperature value above a predefined temperature threshold. The one or more target locations having temperature value above a temperature threshold are referred as hotspots in the present disclosure. In a similar manner, locations having temperature value below a temperature threshold are referred as a coldspots in the present disclosure

In an embodiment, the model bank comprises a plurality of control models, each corresponding to a target location. The control model is defined for a particular target location for regulating temperature at that target location. In an embodiment, the control model defines one or more manipulated variables associated with a target location. Also, the manipulated variables may be provided with weights based on at least one of cost, ease of accessibility, performance, and location. Here, the manipulated variables may indicate data related to the one or more actuators 102.

In an embodiment, the one or more actuators 102 include, but are not limited to, roof tiles, fan, air conditioning system, robotic cooling systems, Computer Room Air Handler (CRAH) systems, flow channels and the like. In an embodiment, the CRAH systems use fans, cooling coils, water chiller units and the like to regulate temperature of the data center 103.

Figure IB shows a data center 103 with identified one or more target locations 104. As shown in the Figure IB, the one or more target locations 104 may be on racks of the data center 103, or any other location in the data center 103. For illustration, Figure IB, indicates the one or more target locations 104 on the racks of the data center 103.

In an embodiment, the model bank is on a data storage unit for example a memory, a database and the like. The model bank may be accessed by the controller using wired or wireless interface. The wired interface may be via Ethernet and the like. Likewise, wireless interface may include but are not limited to Wireless Fidelity (Wi-Fi), Bluetooth and the like. The connection between the controller 100 and a display device 105 and a server 106 is shown in Figure 1C. In an embodiment, the server 106 is connected to a database. The controller may request model bank data from the server 106, and the server 106 may fetch the data from the database and provide it to the controller 100. In accordance with the embodiment, model generation (e.g. heat map generation) is performed at the server (such as in case of CFD type of model), and the generated model (e.g. heat map) or information derived from the generated model (e.g. specific hotspot locations, heat profiles etc.), is periodically updated/available with the controller 100. Figure ID shows an exemplary diagram of the display device 105. The display device 105 may display hotspots data to a user. Also, hotspots control related data like control models and actuators available for regulating temperature at the hotspots may be shown by the display device 105. The display device may show any other data related to regulation of temperature at the target locations 104.

Figure 2 illustrates internal architecture of a controller 100 for regulating temperature in the data center 103, in accordance with embodiments of the present disclosure. The controller may include at least one Central Processing Unit ("CPU" or "processor") (not shown in figure) and a memory 202 storing instructions executable by the at least one processor. The processor may comprise at least one data processor for executing program components for executing user or system-generated requests. The memory 202 is communicatively coupled to the processor. The controller 100 further comprises an Input/Output (I/O) interface 201. The I/O interface 201 is coupled with the processor through which an input signal or/and an output signal is communicated.

In an embodiment, data 203 may be stored within the memory 202. The data 203 may include, for example, target location temperature 204, location identifier 205, control calculations 206 and other data 207.

In an embodiment, the target location temperature 204 indicates temperature at each of the one or more target locations. The temperature at each of the one or more target locations 104 is determined using the heat map.

In an embodiment, a location identifier 205 is associated with the hotspots. The location identifier 205 helps to identify control models associated with the hotspots. In one embodiment, the model bank has several models, for example 20 models to 20 grids of data center. In case of a hotspot, only a particular model (or set of models) is needed which relates temperature of hotspot location with manipulated variables (e.g. airflow, supply air temperature, IT load etc.).

The location identifier provides location of each of the one or more hotspots. The location may be already be known (e.g. provided as input), or determined using at least one of Global Positioning System (GPS), Internet Protocol (IP address) or any other means for locating the location of the one or more hotspots.

In an embodiment, the control calculations 206 are used to calculate values of manipulated variables for the temperature of the hotspots to reach the set point value. Executing the specific model (identified as described above) for a hotspot can involve calculating how temperature at each grid point will change with time (e.g. in next half hour / one hour). This is based on mapping between the current temperature at each grid point (identified with the model 1), and manipulated variable (e.g. airflow, set temperature etc.).

In an embodiment, the other data 207 may include temperature values of a part of or entire data center 103. As an example, the heat map can be for a rack, or for a group of racks or for the entire data center. Other data 207 may further include temperature variation at various locations in the data center 103. The current temperature and / or temperature variation values may be used to predict temperature of a location in the data center 103. Hotspot prediction can involve comparing heat map with threshold values (e.g. temperature threshold set by user, expert, and /or derived from historic data). This prediction can be for a time period (e.g. for a cycle / half-cycle, or Is, 2s ... 30s etc.). Consider that expected temperature values and expected variation in the corresponding values at a location is known. The same can be provided as input for simulation of future values (e.g. simulation carried out with extrapolation using the time trend of these input parameters).

Thus, measured temperature values can be used to determine locations in the data center at which temperature values are likely to deviate from expected values (e.g. based on simulation). Thus, the one or more target locations 104 may be predicted using the proposed system.

In an embodiment, the data 203 in the memory 202 is processed by modules 208 of the controller 100. As used herein, the term module refers to an application specific integrated circuit (ASIC), an electronic circuit, a field-programmable gate arrays (FPGA), Programmable System-on-Chip (PSoC), a combinational logic circuit, and/or other suitable components that provide the described functionality. The modules 208 when configured with the functionality defined in the present disclosure may result in a novel hardware. In one implementation, the modules 208 may include, for example, a target location recognizer 209, a set point generator 210, a control model selector 211, manipulated variable value generator 212 and other modules 213. It will be appreciated that such aforementioned modules 209 may be represented as a single module or a combination of different modules.

In an embodiment, the target location recognizer 209 recognizes the one or more target locations 104 in the data center 103 using the heat map. Here, the target location recognizer 209 receives data of the heat map from the heat map generation model. Further, the target location recognizer 209 compares temperature values at each location in the data center 103 with predefined temperature threshold. Such comparison may involve predicting values for a time period / cycle as described above. The target location recognizer 209 flags all the locations where temperature value deviates from the predefined temperature threshold. For example, an upper temperature threshold and a lower temperature threshold may be set for each location in the data center 103. For example, for a particular location, an upper temperature threshold may be 24° C and lower temperature threshold may be 21° C. Thus, all locations where temperature values are more than (or predicted to be more than) the upper temperature threshold and all the locations lower than (or predicted to be lower than) the lower temperature threshold are considered as one or more target locations 104.

In an embodiment, when none of the locations in the data center 103 deviates from the temperature threshold, then the locations having temperature values close to the temperature threshold may be considered as one or more target locations 104.

In an embodiment, the set point generator 210 generates a set point for each of the target locations 104. The set point is a nominal temperature value at which the data center 103 should be operated. The nominal temperature value can be the desired value for temperature at the location (e.g. set by user), or expected temperature under normal operating conditions at the location. Thus, the set point is generated according to predefined values for locations, or calculated based on history data. The set point for each of the target locations is used to calculate amount of regulation required for the measured temperature value to meet the set point value.

In an embodiment, the control model selector 211 selects a control model from the model bank, corresponding to the one or more target locations 104. The selection is made based on the location identifier associated with each of the one or more target locations 104. In one embodiment, the models in model bank are associated with plurality of locations of the data center. Accordingly, one or more corresponding control models can be identified for each of the one or more target location 104. The control model for each target location is used to regulate temperature at the corresponding target location. Thus, the control model is mapped with the target location using the location identifier. The model bank comprises of a plurality of control models, each control model corresponding to a location. In an embodiment, when a control model is selected, associated manipulated variables and the one or more actuators 102 required for regulating temperature of the corresponding target locations are inherently selected.

In an embodiment, the manipulated variable (MV) value generator 212 generates values for the manipulated variables associated with the selected control model. The value of a manipulated variable can be generated by the controller as a result of its control related computation. In a digital control system, the manipulated variable is sent out as a digital value to the actuator, while in an analog system we will have analogue signal transmitted to the actuator. The MV values are generated such that the temperature at the one or more target locations 104 is regulated to reach the set point temperature value. The MV values indicate amount of regulation and / or time period of regulation for the measured temperature value to meet the set point value. In an embodiment, a fan speed of 600 rpm is set for controlling a fan at a particular location. In another embodiment, in addition to fan speed, a time duration of 10 minutes is set.

In an embodiment, the other modules 213 may include a notification module. The notification module may indicate a user operating the data center 103 regarding temperature of various locations in the data center 103.

Figure 3 shows a flow chart illustrating a method for regulating temperature in the data center 103, in accordance with embodiments of the present disclosure.

The illustrated operations of Figure 3 show certain events occurring in a certain order. In alternative embodiments, certain operations may be performed in a different order, modified or removed. Moreover, steps may be added to the above described logic and still conform to the described embodiments. Further, operations described herein may occur sequentially or certain operations may be processed in parallel. Yet further, operations may be performed by a single processing unit or by distributed processing units. As illustrated in Figure 3, the method 300 may comprise one or more steps for regulating temperature of the data center 103, in accordance with some embodiments of the present disclosure. The method 300 may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform particular functions or implement particular abstract data types.

The order in which the method 300 is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the spirit and scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.

At step 301, the heat map of the data center 103 is generated. The heat map indicates temperature at various locations of the data center 103 and temperature variation in the data center 103 (i.e. expected temperature and variations across different locations). Figure 4 shows a hybrid diagram for generating a heat map.

Referring now to Figure 4, a heat map generation model is shown in accordance with an embodiment. This model may be generated offline or at a server and provided to the controller (e.g. periodically updated). Here, a relationship between temperature at each location of the data center 103, present IT load and ambient temperature is identified. In offline model generation, the temperature data 401 and IT load 402 are provided to a random signal generator 403. The temperature data 401 includes, but is not limited to, measured temperature at various locations and ambient temperature of corresponding locations. IT load may be the computational load (or operating level) of servers, storages etc. of the data center, that may affect the temperature of the equipment and surrounding.

In an embodiment, the random signal generator 403 generates random signals for temperature value or range corresponding to every location in the data center 103. The random signal may be of the form, but not limited to, pseudo random binary sequence, multi sine sequence, multi-step sequence and the like. Each combination of the random signals is provided to a heat map generation model 404 for generating detailed simulation results. The simulations are stored in the memory 202 for analysis. The simulations may be stored in at least one of a tabular from, a simple linear or non-linear model, polytope, and the like. The analysis of the simulations comprises approximating temperature of the data center 103 at each location precisely with measured temperature 401, IT load 402 and ambient temperature 401. The output of the heat map generation model 404 may be represented as a function of measured temperature, ambient temperature and the IT load 402, as given below:

T(X, y, Z) = F( Tmeas , Tamb , lT|o; l) (1)

Where

T(x, y, z) : temperature at location x, y, z;

Tmeas : measured temperature;

Tamb : ambient temperature; and

ITioad : ΓΓ load of the data center.

At step 302, the one or more target locations 104 are identified (predicted) using the heat map. The target location recognizer 209 is used to identify the one or more target locations 104. Here, the locations in the data center 103 where the temperature value deviates (or expected to deviate) from the temperature threshold are flagged by the target location recognizer 209. The flagged locations are identified as target locations.

Referring back to Figure 3, at step 303, a set point is generated (e.g. with set point generator 210) for each of the one or more target locations 104.

At step 304, the control model selector 211 selects a control model corresponding to the one or more target locations 104 from the model bank. The selection is done based on location identifier associated with each of the one or more target locations 104.

Referring to Figure 5, a hybrid model illustrative of generation of a model bank is detailed. Similar to heat map modelling, the model bank may be generated offline and used as needed in real-time. This hybrid model generation involves identifying how temperature at each location of the data center 103 varies with time.

Block 501 represents manipulated variables. The manipulated variables are the variables which can be varied in order to keep the value of a corresponding control variable at a desired level. The manipulated variables can be varied ahead of time based on reference. For example, based on the set point, the manipulated variables corresponding to a target location 104 can be varied such that measured temperature at that target location meets the set point temperature value. Further, the manipulated variables can be adjusted for the measured temperature to remain constant at the set point value. The manipulated variables indicate data related to the one or more actuators 102.

A range of manipulated variables 501, IT load 402 and location identifier 205 are provided as input to the random signal generator 403. Further, the random signal generator generates random sequences. For each combination of random sequence, the control model generator 502 generates a control model. Then, each of the generated control model is stored in the model bank 503. The model bank 503 may be database associated with the controller 100. In an embodiment, the control model generator 502 may be a CFD model or the like. The control model may be a control relevant dynamic model resultant of a simulation of the combination of the random sequences. The dynamic model may be linear or non-linear model that may predict temperature profiles for a given target location for a predefined time period. The output of the model bank generator may be equationally represented as:

T(x, y, / )/ dt = G[To(x, y, z), u(t), ITioad(t)] (2)

Where,

T(x, y, z)/ dt : rate of change of temperature at location x, y, z with time;

To(x, y, z) : Initial temperature at the location x, y, z; and

U(t) : manipulated variables of a control model corresponding to location x, y, z;

Referring back to Figure 3, at step 305, the manipulated variable value generator 212 generates values related to the manipulated variables for controlling the temperature at the one or more target locations 104. The controlling is based on the set point (or set points for the different locations). The temperature is controlled such that the measured temperature at the one or more target locations 104 reach the set point value(s). The values of the manipulated variables may relate to amount of actuation required for the measured temperature to reach the set point value, and optionally the time period of regulation. The values are then provided to the one or more actuators 102 for acting on the target locations 104.

Table 1 shows an example for control model corresponding to location identifier. Let the location identifier of the one or more target locations be as shown in the table. Location Identifier Control Model ID Actuators Manipulated

variables

50 2 Fan, AC, Roof tiles State 1, state 2, mode 1, mode 2, mode 3

34 13 Fan, Roof tiles State 1, state 2, mode 1, mode 2

3 7 Robotic cooling State 1, state 2, system, AC mode 1, mode 2, mode 3, mode 4

8 10 Fan State 1, state 2, mode 1, mode 2

16 1 Fan, Ac State 1, state 2, mode 1, mode 2, mode 3

TABLE 1

The corresponding control models are shown in column 2 of the table, and the manipulated variables of each control model are shown in column 4 of the table. Each manipulated variable is associated with corresponding one or more actuators 102. The manipulated variables may indicate operating states and operating modes of the one or more actuators 102. In an embodiment, the operating states may include but are not limited to "ON" state and "OFF" state. As an example, a fan speed of 600 rpm may be set for controlling a fan at a particular location. In addition to fan speed, a time duration of 10 minutes can be set. The operating modes may relate to, but are not limited to, speed, temperature, timer, and air flow distribution. For example, when attending the target location 104 having location identifier numbered 16, the control model numbered 1 is selected from the model bank. When the control model is selected, associated manipulated variables and available one or more actuators 102 are selected as well. The one or more actuators 102 are fan and AC.

In an embodiment, the invention provides a system for regulating temperature at the one or more target locations with robotic cooling devices (jets). A robotic cooling device is a device capable of manoeuvring across various locations in the data center. As an example, the robotic cooling jet may have fans (or other cooling device) movable vertically or horizontally, relative to a rack(s) in the data center. For example, the fans may be mounted on rails (or provided as drones), and may be configured such that the fans can move vertically / horizontally. Further, such fans (or cooling devices) can be controlled with the controller of the invention. Figure 6A shows an exemplary block diagram illustrating cooling of data center racks using robotic cooling jets. As an example, let us consider a data center comprises a first server rack 601, a second server rack 602 and a robotic cooling jet 603. The Figure 6A further shows roof 604, a floor 605 and a Computer Room Air Conditioning Unit (CRAC) 606 of a data center.

The robotic cooling jet 603 may direct cool air on one or more hot spots that are identified in the first server rack 601 and the second server rack 602. The cool air is directed based on at least one of IT load, Infra-Red (IR) based sensing, wireless sensors, and the like. Directed distribution of the cool air may ensure that a uniform temperature is maintained across the first server rack 601 and the second server rack 602. Such directed distribution can take into account temperature of different locations to avoid overcooling already cool locations. Thus, the use of robotic cooling jet 603 may avoid overcooling and save energy. The temperature spread as per the robotic cooling jet 603 is as shown in Figure 6B. In an embodiment, the reduction in temperature spread may also result in maintaining a higher overall temperature and energy savings.

The CRAC 606 is a device that monitors and maintains the temperature, air distribution and humidity in a data center.

In one of the embodiment, the number of robotic cooling jets 603 and area over which the robotic cooling jets 603 move may be based on at least one of criticality of the one or more hotspots formation and the response time required from robots.

Figure 7 shows an exemplary block diagram illustrating cooling of data center racks, using flow channels (robotic cooling devices). A flow channel may have a conduit or a pipe to receive cool / hot air, and circulate and feed the air through several openings (like 702). Thus, cool / hot air flow can be directed to one or more areas (relative to the openings) to regulate temperature therein. The flow channels 701 may be used to direct cool air to the one or more hot spots by controlling openings 702 of the flow channels 701.

In another embodiment, information on provisioning of the robotic cooling jets 603 or the flow channels 701 may be used as an input while scheduling IT loads for cooling or scheduling of IT loads. In another embodiment, the control of the robotic cooling jet 603 may be performed either independently or in conjunction with the conventional CRAC control.

Thus, the invention provides for predicting temperature variations at different locations in the data center, and accordingly initiating control for the specific locations.

REFERRAL NUMERALS:

405 Heat map

501 Manipulated variables

502 Control model generator

503 Model bank

601 First server rack

602 Second server rack

603 Robotic cooling jets

604 Data center roof

605 Data center tiles

606 CRAC

701 Flow channels

702 openings

Claims

1. A method for regulating temperature of a data center, the method comprising:

identifying, by a controller, one or more target locations in the data center using a heat map of the data center and one or more inputs, wherein the one or more inputs comprises measurements received from a plurality of sensors of the data center, wherein the one or more target locations are predicted using the heat map and the one or more inputs;

generating, by the controller, a set point for each of the one or more target locations; selecting, by the controller, a control model from a model bank, corresponding to each of the one or more target locations based on the location of the corresponding target location; and

controlling, by the controller, temperature of the one or more target locations using the selected control model based on the set point.

2. The method as claimed in claim 1, wherein the heat map is generated using a heat map generation model, and wherein the heat map is used to predict the one or more target locations.

3. The method as claimed in claim 1, wherein the one or more inputs comprises at least one of measured temperature at the one or more target locations, ambient temperature and IT load.

4. The method as claimed in claim 1, wherein the one or more target locations are identified as hotspots when temperature of the one or more target location deviates from a temperature threshold, wherein the identification is performed at predefined time intervals.

5. The method as claimed in claim 1, wherein each of the one or more target locations are associated with a location identifier, wherein the location identifier is used for selecting the control model from the model bank, the model bank comprises a plurality of control models, each of the plurality of control models corresponds to a location among a plurality of locations in the data center, and comprises one or more manipulated variables related to temperature of the one or more target locations.

6. The method as claimed in claim 1, wherein the step of controlling comprises updating values of the one or more manipulated variables for the measured temperature required to reach the set point, wherein the updated values of the one or more manipulated variables are provided to an actuator for controlling the temperature of the data center, wherein each of the one or more manipulated variables are provided a weight based on one or more parameters.

7. A system for regulating temperature of a data center, comprising:

a plurality of sensors for measuring temperature of the data center at predetermined locations;

a controller configured to:

identify one or more target locations in the data center using the heat map and the measurements received from the plurality of sensors;

generate a set point for each of the one or more target locations;

select a control model from a model bank corresponding to each of the one or more target locations; and

control temperature of the one or more target locations with one or more actuators, using the selected control model based on the set point for each target location.

8. The system as claimed in claim 7, wherein each of the one or more target locations are associated with a location identifier, wherein the location identifier is used for selecting the control model from the model bank, the control model comprises one or more manipulated variables related to temperature of the one or more target locations.

9. The system as claimed in claim 8, wherein the model bank comprises a plurality of control models, each of the plurality of control models corresponds to a target location among the one or more target locations.

10. The system as claimed in claim 9, wherein each of the plurality of control models are associated with an actuator for controlling the temperature of the data center.