Power reduction control device, power reduction control method, power reduction control system, and program

The power consumption reduction control device optimizes data center power usage by considering equipment-specific conditions and using machine learning to determine efficient server and air conditioner control, resulting in reduced overall power consumption.

JP7589834B2Active Publication Date: 2024-11-26NIPPON TELEGRAPH & TELEPHONE CORP
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2023565667
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-12-06
Publication Date
2024-11-26
Estimated Expiration
2041-12-06

AI Technical Summary

Technical Problem

Existing methods for optimizing data center power consumption fail to account for individual equipment conditions such as air conditioning equipment location, airflow, and server layout, leading to suboptimal power efficiency.

Method used

A power consumption reduction control device that considers equipment-specific conditions by dividing data centers into placement and air conditioning control areas, using machine learning to optimize server and air conditioner control based on external factors, and determining optimal placement patterns to minimize total power consumption.

Benefits of technology

The device effectively reduces data center power consumption by accounting for varying equipment conditions, achieving optimal server and air conditioning power efficiency through intelligent resource placement and control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007589834000001
    Figure 0007589834000001
  • Figure 0007589834000002
    Figure 0007589834000002
  • Figure 0007589834000003
    Figure 0007589834000003
Patent Text Reader

Abstract

An electric power amount reduction control device (100) comprises: an external factor acquisition unit (211) that acquires an external factor for air conditioning control; a situation determination unit (212) that determines a situation category based on the external factor; a control value search unit (220) that calculates an air conditioning control value for each situation category; an air conditioning control execution unit (230) that uses the air conditioning control value to control an air conditioner; a correspondence information generation unit (250) that generates control value-electric power amount correspondence information for each situation category; a layout pattern calculation unit (310) that calculates a virtual resource layout pattern; a server consumption electric power amount estimation unit (320) that estimates the server consumption electric power amount for each layout pattern; and a layout pattern determination unit (330) that calculates a total amount of server consumption electric power and air conditioning consumption electric power, for each layout pattern, and determines the layout pattern for which the total amount is the smallest.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a power reduction control device, a power reduction control method, a power reduction control system, and a program for reducing the amount of power consumption in a data center (hereinafter sometimes referred to as "DC"). [Background technology]

[0002] Air conditioning in data centers (DCs) accounts for a large proportion of the power consumption, and as the number and size of DCs increases, there is a demand to reduce the amount of power consumed by air conditioning. In addition, the amount of data processed in DCs is increasing year by year, making it necessary to improve the power consumption efficiency of the DC as a whole (the amount of power consumed by the entire DC for a certain amount of data processing).

[0003] The technology described in Non-Patent Document 1 has been disclosed as a technology for optimizing the power consumption of an entire DC by taking into consideration the power consumption of air conditioning and the power consumption of a server (IT device). In the datacenter air conditioning-linked IT load placement optimization method described in Non-Patent Document 1, future load trends of IT equipment are predicted by collecting operation information and monitoring information of IT equipment in a datacenter, and power increases of air conditioning facilities corresponding to power increases of IT equipment are calculated. Then, an optimization problem is solved to minimize the objective function, which is the power consumption of the datacenter, so that the load concentration rate of IT equipment is increased over time, that is, the number of operating IT equipment is reduced. In this way, the placement of IT loads (virtual machines) on IT equipment that minimizes the power consumption of the datacenter is calculated. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Jun Okitsu and 4 others, "Optimization of IT Load Placement in tandem with Air Conditioning for Eco-Friendly Data Centers," FIT (Forum on Information Technology) 2010, 9th Forum on Information Science and Technology, RC-009 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the technology described in Non-Patent Document 1, a general-purpose rule-based standard that is not dependent on equipment conditions that differ for each DC is adopted in the air conditioning power model used to calculate the power of the air conditioning equipment. Therefore, it was difficult to perform optimization to reduce the total power consumption of the DC, taking into account individual equipment conditions such as the location of the air conditioning equipment, air flow, server configuration in the DC, and thermal cooling efficiency.

[0006] The present invention was made in consideration of these points, and its objective is to reduce the total power consumption of a data center, consisting of server power consumption and air conditioning power consumption, in response to the different equipment conditions of each data center. [Means for solving the problem]

[0007] The power amount reduction control device according to the present invention is a power amount reduction control device that controls a plurality of servers and a plurality of air conditioners that a data center has, and the data center has, on its floor, a plurality of placement control zones in which a group of servers on which virtual resources are placed are placed, and a plurality of air conditioning control zones that are areas in which the effect of air conditioning control by the plurality of air conditioners is measured and that are located on either the inlet side or the outlet side of the group of servers, and the power amount reduction control device includes an external factor acquisition unit that acquires information on external factors related to air conditioning control, including a floor average temperature calculated from an average value of temperatures measured in the plurality of air conditioning control zones, an outside temperature that is the temperature outside the data center, and a server power consumption amount for each of the placement control zones that is a predicted amount when the virtual resources are placed in the servers, a Situation determination unit that divides the value of each of the external factors into a predetermined range width, defines a combination of the ranges divided for each of the external factors as a Situation classification, and determines which Situation classification the acquired information on the external factor belongs to, and a Situation determination unit that determines the Situation classification to which the acquired information on the external factor belongs, and ... a control value search unit that calculates air conditioning control values ​​including at least a target temperature to be set, an air conditioning control execution unit that executes control of the multiple air conditioners using the calculated air conditioning control values, a correspondence information generation unit that acquires the air conditioning control values ​​of the multiple air conditioners and the air conditioning power consumption of the multiple air conditioners when control is performed using the air conditioning control values, and generates control value-power correspondence information that associates the air conditioning power consumption with the air conditioning control values ​​of the multiple air conditioners for each of the Situation classifications, a placement pattern calculation unit that calculates a placement pattern in which the virtual resource to be newly placed is placed on one of the multiple servers, a server power consumption estimation unit that estimates the server power consumption of a server group belonging to each of the placement control zones for each of the calculated placement patterns, and acquires a judgment result of the Situation classification using the server power consumption of each of the placement control zones from the Situation judgment unit, and acquires the air conditioning control values ​​and the air conditioning power consumption of the multiple air conditioners in the Situation classification for each placement pattern by referring to the control value-power correspondence information, and totals the server power consumption for each of the placement control zones for each placement pattern,and an arrangement pattern determination unit that calculates a sum of the total server power consumption, which is the sum of the total server power consumption and the air conditioning power consumption, and determines the arrangement pattern in which the calculated sum is the smallest as the arrangement pattern in which the virtual resources are arranged. Effect of the Invention

[0008] According to the present invention, it is possible to reduce the total power consumption of a data center, which is made up of server power consumption and air conditioning power consumption, in response to the different facility conditions of each data center. [Brief description of the drawings]

[0009] [Figure 1] 1 is a diagram showing an overall configuration of a power amount reduction control system including a power amount reduction control device according to an embodiment of the present invention; [Diagram 2] 1 is a functional block diagram showing a configuration example of a power amount reduction control device according to an embodiment of the present invention; [Diagram 3] FIG. 2 is a diagram for explaining a situation classification according to the present embodiment. [Figure 4] FIG. 13 is a diagram for explaining a reward (temperature reward) according to the present embodiment. [Diagram 5] FIG. 13 is a diagram for explaining a learning phase in which air-conditioning control values ​​that satisfy rewards in each situation classification are searched for. [Figure 6] 13 is a diagram illustrating an operation phase in which degenerate operation is performed and more optimal air-conditioning control values ​​are searched for. FIG. [Figure 7] 10 is a flowchart showing the flow of an allocation pattern determination process for determining an allocation destination of a virtual resource, which is executed by a power amount reduction control device according to the present embodiment. [Figure 8] FIG. 2 is a hardware configuration diagram showing an example of a computer that realizes the functions of the power amount reduction control device according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Next, an embodiment of the present invention (hereinafter, referred to as "the present embodiment") will be described. FIG. 1 is a diagram showing the overall configuration of a power amount reduction control system 1 including a power amount reduction control device 100 according to this embodiment.

[0011] As shown in Figure 1, the power consumption reduction control system 1 is configured to include a data center (DC10) having multiple servers 3 and multiple air conditioners 2, and a power consumption reduction control device 100 that is communicatively connected to the multiple servers 3 and multiple air conditioners 2 in the DC10. The power amount reduction control device 100 may be provided inside the DC 10, or may be provided at a location separate from the DC 10 and control a plurality of DCs 10.

[0012] This power consumption reduction control device 100 may acquire status information and transmit air conditioning control information for air conditioners 2 (air conditioners "1," "2," and "3" in FIG. 1) installed within DC 10 via an air conditioning management device not shown, or may be directly connected to communicate with each air conditioner 2 without going through the air conditioning management device. In addition, the power consumption reduction control device 100 may acquire status information and transmit control information of the servers 3 provided in the DC 10 via a server management device (not shown), or may be directly connected to communicate with each server 3.

[0013] In the DC 10 of this embodiment, as shown in Fig. 1, all the servers 3 accommodated are divided into areas in which a plurality of servers 3 are arranged, and are controlled as "placement control areas". These placement control areas 30 are areas that accommodate a group of servers in which virtual resources are arranged. Fig. 1 shows an example in which placement control areas "1" to "6" are provided.

[0014] DC10 will be described assuming that a virtualization platform is constructed and operated. Known open source virtualization platforms include OpenStack (registered trademark), which is software for constructing cloud environments, and Kubernetes (registered trademark), which is software for operating and managing containerized workloads and services. OpenStack is primarily used for managing and operating physical machines and virtual machines (VMs). Kubernetes is primarily used for managing and operating containers. In this specification, an application (consisting of one or more containers, one or more VMs, etc.) virtualized in a virtualization platform is called a virtual resource. In Kubernetes, the smallest execution unit of an application is a Pod consisting of one or more containers.

[0015] In this embodiment, an "air conditioning controlled area" is provided in correspondence with the placement control area 30 of the server group, as shown in Fig. 1. The air conditioning controlled area 20 is a collective area in which the room temperature effect of air conditioning control is measured, and faces either the intake side or the exhaust side of the server 3. The air blown from the air conditioner 2 is blown out from the air conditioning controlled areas 20 on the intake side (air conditioning controlled areas "3," "4," "7," and "8" in FIG. 1) via piping installed, for example, under the floor of the DC 10. Then, air whose temperature has increased due to the heat of each server is taken in from the intake of the piping installed in the air conditioning controlled areas 20 on the exhaust side (air conditioning controlled areas "1," "2," "5," and "6" in FIG. 1), and an airflow is generated that returns to the air conditioner 2.

[0016] A plurality of sensors (temperature sensors, etc.) are installed in each of the air-conditioning controlled areas 20. In addition, sensors (temperature sensors, etc.) are also installed outside the DC 10. Information obtained from these sensors (sensor information) can be acquired by the power reduction control device 100 via a communication line or the like.

[0017] The power amount reduction control device 100 according to this embodiment predicts the power consumption (server power consumption) of the allocation control area 30 in each allocation pattern of the virtual resource to the server 3 based on schedule information for the generation and deletion of the virtual resource. Then, based on the power consumption of each allocation control area 30, the power amount reduction control device 100 searches for a control value (air conditioning control value) of each air conditioner 2 that satisfies a reward (temperature reward), and determines, as the allocation destination of the virtual resource, an allocation pattern that minimizes the total amount of air conditioning power consumption and server power consumption when the air conditioner 2 is controlled with the searched air conditioning control value. In addition, in the operation stage (operation phase), the power amount reduction control device 100 searches for an air conditioning control value with the lowest power consumption within a range that satisfies the reward by degenerate operation, and further reduces the power consumption (details will be described later).

[0018] <Power reduction control device> FIG. 2 is a functional block diagram showing an example of the configuration of the power amount reduction control device 100 according to this embodiment. The power consumption reduction control device 100 predicts the power consumption (server power consumption) of the placement control area 30 for each placement pattern of the virtual resources to the server 3, and searches for the control value (air conditioning control value) of the air conditioner 2 that satisfies the reward (temperature reward), thereby determining, as the placement destination for the virtual resources, the placement pattern that minimizes the total amount of server power consumption and air conditioning power consumption. The power amount reduction control device 100 calculates the power consumption (server power consumption) of the allocation control area 30 in each allocation pattern of the virtual resources to the server 3, for example, by using a learning model generated by machine learning such as a neural network. The power amount reduction control device 100 also calculates the optimal control value of the air conditioner 2 for the calculated power consumption (server power consumption) by using, for example, a learning model generated through a learning phase and an operation phase. Furthermore, in the operation phase, the power amount reduction control device 100 searches for the air conditioning control value with the lowest power consumption within a range that satisfies the reward, by performing degenerate operation. As a result, the power amount reduction control device 100 can determine the optimal allocation of the virtual servers, taking into account the server power consumption and the air conditioning power consumption, and reduce the total power consumption in the DC 10. The power consumption reduction control device 100 is configured by a computer including a control unit, an input / output unit, and a storage unit (all of which are not shown).

[0019] The input unit inputs and outputs information between each device (each air conditioner 2 and each server 3) in the DC 10. This input / output unit is composed of a communication interface that transmits and receives information via a communication line, and an input / output interface that inputs and outputs information between an input device such as a keyboard (not shown) and an output device such as a monitor.

[0020] The storage unit is composed of a hard disk, a flash memory, a RAM (Random Access Memory), and the like. This storage unit temporarily stores programs for executing the functions of the control unit and information necessary for the processing of the control unit. This storage unit also stores operation history information 201, air conditioning control learning data 202, air conditioning control learning model 203, control value-power amount correspondence information 204, and server power amount learning model 301 as information for determining the control value (air conditioning control value) of the air conditioner 2 (details will be described later).

[0021] The control unit is responsible for all the processes executed by the power amount reduction control device 100, and includes an air conditioning control unit 200 and a server control unit 300, as shown in FIG.

[0022] The air conditioning control unit 200 regards the average temperature of the floors in DC10 before control (floor average temperature), the outside temperature, and the server power consumption for each placement control area 30 as Situation components, and calculates the optimal air conditioning control value that satisfies the reward (temperature reward) in each Situation through a learning phase and an operation phase. This air conditioning control unit 200 includes a situation recognition unit 210 , a control value search unit 220 , an air conditioning control execution unit 230 , a remuneration calculation unit 240 , and a corresponding information generation unit 250 .

[0023] The situation recognition unit 210 acquires information on external factors, which are parameter elements constituting the Situation classification. Then, the situation recognition unit 210 divides each external factor into a plurality of ranges, defines a combination of each range region as one Situation, and determines a Situation classification indicating which Situation the external factor belongs to based on the acquired information on the external factor. The situation recognition unit 210 includes an external factor acquisition unit 211 and a situation determination unit 212 .

[0024] The external factor acquisition unit 211 acquires information on the measurement results of external factors. Here, the external factors are elements that affect the increase or decrease in the amount of power consumed by air conditioning, and refer to parameter elements that constitute the Situation classification. Here, the external factors are (1) the average floor temperature in the DC 10 before control, (2) the outside temperature, and (3) the amount of power consumed by servers in each placement control area 30.

[0025] (1) The external factor acquisition unit 211 calculates the average floor temperature in DC 10 before control as follows: The external factor acquisition unit 211 calculates the average value of temperatures acquired from the temperature sensors in the air-conditioning controlled zone 20, and calculates the average temperature for each air-conditioning controlled zone 20. The external factor acquisition unit 211 then averages the calculated average temperatures for each air-conditioning controlled zone 20 for the entire floor, and sets the obtained temperature as the floor average temperature.

[0026] (2) The external temperature is information obtained from a temperature sensor installed outside the DC10. (3) The server power consumption amount for each placement control area 30 is information calculated by the server control unit 300 (details will be described later). When the external factor acquisition unit 211 acquires the information on the external factors, it outputs the information to the situation determination unit 212 .

[0027] The situation determination unit 212 determines to which situation classification the information acquired by the external factor acquisition unit 211 belongs. Each external factor is divided into multiple ranges between the minimum and maximum values ​​according to the characteristics of the external factor. A combination of the ranges into which each external factor is divided is defined as one Situation. The following will be explained with reference to FIG. 3.

[0028] As shown in FIG. 3, each external factor is defined as a "factor" and a range to be divided is defined (hereinafter referred to as "division definition"). For example, the external factor of "factor1" shown in the Situation classification information 52 in FIG. 3 is "average floor temperature," and the division definition is "0-48 degrees divided into 6." The external factor of "factor2" is "outside temperature," and the division definition is "0-48 degrees divided into 6." The external factor of "factor3" is "server power consumption in placement control area 1," and the division definition is "0-200W divided into 20." Similarly, the external factor of "factor8" is "server power consumption in placement control area 6," and the division definition is "0-200W divided into 20."

[0029] Here, it is assumed that the information on the external factors acquired by the Situation determination unit 212 is the external factor information 51 shown in Fig. 3. In this case, since the value of "factor1" (floor average temperature) is "25", the Situation determination unit 212 determines that the "range" is included in the "24-32 range" (24 degrees or more and less than 32 degrees) and sets the "factor range identifier" to "factor1-4". This "factor range identifier" is information that identifies the range to which the temperature belongs, for example, by dividing 0-48 degrees into 6, with 0 degrees or more and less than 8 degrees being "factor1-1", 8 degrees or more and less than 16 degrees being "factor1-2", and 16 degrees or more and less than 24 degrees being "factor1-3". The same applies to the other "factors".

[0030] The situation determination unit 212 combines the information of the "factor range identifier" of the external factors to form a "Situation classification", and determines that the situation classification is "factor1-4_factor2-4_factor3-4_factor4-4_factor5-5_factor6-5_factor7-4_factor8-4". In this manner, the situation determination unit 212 determines the "situation classification" based on the acquired information on the external factors.

[0031] Returning to Fig. 2, the control value searcher 220 calculates optimal control values ​​for each air conditioner 2 in the situation indicated by the determined Situation classification. The control values ​​(air conditioning control values) of the air conditioners 2 are parameters for controlling the air conditioners 2, and include at least temperature (target temperature), and may also include air volume, air direction, etc. In this embodiment, the parameters of the air conditioning control values ​​will be described as being the target temperature and air volume. The calculation of the air conditioning control value by the control value search unit 220 can be performed using a method that uses past performance data or a predetermined rule-based calculation method, but in this embodiment, an example is described in which a learning model (air conditioning control learning model 203) is constructed and the calculation is performed. The control value search unit 220 includes a control value generation unit 221 , a learning model management unit 222 , and a reduction operation unit 223 .

[0032] In the learning phase, the control value generation unit 221 randomly generates air conditioning control values ​​(target temperature, air volume, etc.) for each Situation classification up to a predetermined number of times (N times). At this time, the control value generation unit 221 may generate air conditioning control values ​​by turning off the power of any one of the multiple air conditioners 2. The control value generation unit 221 then outputs the randomly generated air conditioning control value to the air conditioning control execution unit 230. When the air conditioning control execution unit 230 acquires the randomly generated air conditioning control value, it executes control of the air conditioner 2. Then, for each Situation classification, the air conditioning control execution unit 230 stores, as operation history information 201, the external factor information, the air conditioning control value, the reward (area reward) obtained by the reward calculation unit 240 described below when that control is executed, and information on the amount of air conditioning power consumption.

[0033] Furthermore, when the degenerate operation is completed in the operation phase, the control value generator 221 generates an air conditioning control value by referring to control value vs. power amount correspondence information 204 (described in detail later).

[0034] When a predetermined number of times (N times) is reached in the learning phase, the learning model management unit 222 refers to the operation history information 201, and for each Situation classification, imports external factor information, air-conditioning control values, and rewards (area rewards) to generate air-conditioning control learning data 202. Then, the learning model management unit 222 uses the generated air-conditioning control learning data 202 to perform machine learning so as to maximize the reward (so that the reward approaches 100%), thereby generating an air-conditioning control learning model 203 for each Situation classification.

[0035] Furthermore, after a predetermined number of times (N times) in the learning phase, the learning model management unit 222 outputs air conditioning control values ​​(target temperature, air volume, etc.) by inputting external factor information into the air conditioning control learning model 203 for each Situation classification. The learning model management unit 222 then outputs the air conditioning control values ​​calculated by the air conditioning control learning model 203 to the air conditioning control execution unit 230. When the air conditioning control execution unit 230 acquires the air conditioning control value calculated by the air conditioning control learning model 203, it executes control of the air conditioner 2. Then, for each Situation classification, the air conditioning control execution unit 230 stores, as operation history information 201, the external factor information, the air conditioning control value, the reward (area reward) obtained by the reward calculation unit 240 described below when executing that control, and information on air conditioning power consumption.

[0036] The learning model management unit 222 references the operation history information 201, imports the newly added information, that is, external factor information, air-conditioning control value, and reward (area reward), as air-conditioning control learning data 202, and updates the air-conditioning control learning model 203 for each Situation classification. Then, after a predetermined number of times (N times), the learning model management unit 222 repeats predicting the air-conditioning control value based on the new Situation information. In addition, when the air-conditioning control learning model 203 for the corresponding Situation classification satisfies a condition based on a predetermined reward described below, the learning model management unit 222 ends the learning phase and transitions to the operation phase.

[0037] After the air-conditioning control learning model 203 for each Situation classification transitions from the learning phase to the operation phase, that is, after it converges to an air-conditioning control value that satisfies a predetermined reward in a certain Situation classification, the degenerate operation unit 223 determines whether there is an air-conditioning control value with a lower control cost (lower power consumption cost) that satisfies the reward from the converged value by gradually executing a trial search for air-conditioning control values ​​in the lower cost direction. Note that information on the air-conditioning control values ​​that satisfy the predetermined reward is stored in control value-power amount correspondence information 204 described later (details will be described later).

[0038] For example, the degenerate operation unit 223 further divides Xn (e.g., air volume) for a control value (X1, X2, ..., Xn) of a certain air conditioner 2 into M stages (M is an integer equal to or greater than 2) between Xnmin-Xnmax (minimum value to maximum value). Then, the degenerate operation unit 223 gradually lowers Xn (air volume) in the direction of lower air conditioning control cost (lower power consumption). This gradual reduction in the control value may be tried in the order of air conditioners "1" → "2" → "3", etc., or air conditioners "1", "2", and "3" may be tried all at once, and a predetermined logic for lowering the control value is set in advance. When degenerate operation is performed with respect to the target temperature among the air-conditioning control values, a process is performed to set the target temperature, for example, by raising it by 1 degree so that the air-conditioning control cost is reduced.

[0039] Furthermore, after failing the reward in Z+1 degenerations (Z is an integer equal to or greater than 0), the degeneration operation unit 223 may set the air conditioning control value from the previous Z degenerations as the air conditioning control value at the time of completion of the degeneration. As another example, after failing the reward after the Z+1th degeneration, the degeneration operation unit 223 may execute a one-step rollback to the previous Zth degeneration, and recheck whether the reward is met again after the Zth degeneration. This is to prevent disturbance. Then, if the recheck after the Zth degeneration passes, the degeneration operation unit 223 sets the control value as the air conditioning control value at the time of completion of the degeneration. If the recheck after the Zth degeneration fails, the degeneration operation unit 223 repeats the process of going back one step. Furthermore, the degeneration operation unit 223 may set the air conditioning control value at the time of completion of degeneration after the recheck when rolling back one step passes k times (a predetermined number of times) in a row. Here, if the degeneration operation unit 223 fails to pass k times in a row, it can go back one step and repeat the calculation of the reward, and determine the control value at the stage after passing the reward k times in a row as the control value with the lowest control cost.

[0040] The air conditioning control execution unit 230 controls each air conditioner 2 based on information about the air conditioning control values ​​(target temperature, air volume, etc.) in a certain situation calculated by the control value search unit 220. The air conditioning control execution unit 230 divides the time required from the start of control to reach a predetermined target temperature by a predetermined number of steps, which is defined as one turn, and sets the target temperature for each turn. The time required from the start of control to reach the predetermined target temperature uses information obtained in advance from past performance data of the air conditioner 2, etc. The target temperature for each turn uses, for example, a value obtained by equally dividing the temperature from the "temperature before control" to the final "target temperature" by the number of control operations. The air conditioning control execution unit 230 outputs instruction information to each air conditioner 2 based on the air conditioning control value set for each turn.

[0041] The air conditioning control execution unit 230 stores, for each Situation classification, external factor information, the air conditioning control values ​​for each turn (target temperature, air volume, etc. for each turn), the reward (area reward) obtained by the reward calculation unit 240 described below when executing the control, and information on air conditioning power consumption as operation history information 201.

[0042] The reward calculation unit 240 calculates a reward (temperature reward) as an index for evaluating the result of executing control using the air conditioning control value calculated by the control value search unit 220. The reward calculation unit 240 then determines whether the control result satisfies a predetermined reward, that is, whether the air conditioning control value satisfies a predetermined reward condition. The remuneration calculation unit 240 includes an area remuneration calculation unit 241 and an overall remuneration calculation unit 242 .

[0043] The zone reward calculation unit 241 defines two types of rewards, a high temperature alert reward and a low temperature alert reward, for each air-conditioning controlled zone 20, and calculates the reward for the control result for each turn. The high temperature alert reward is applied when the temperature before control is higher than the target temperature, i.e., when the room temperature is high and the temperature is controlled to be lowered. The low temperature alert reward is applied when the temperature before control is lower than the target temperature, i.e., when the room temperature is too low and the temperature is controlled to be higher.

[0044] The area reward calculation unit 241 uses the following indicators when calculating the reward (area reward). (Indicator 1) Calculate reward based on the difference between the target temperature for each turn and the current actual temperature. (Index 2) Focus only on the most recent turn and calculate the reward based on the temperature change in the most recent turn. The temperature used here is, for example, the average temperature of the temperature sensors designated in each air-conditioning controlled area 20.

[0045] (Index 1) calculates the difference between the "temperature after turn control" and the "target temperature for each turn." If the "temperature after turn control" is equal to or lower than the "target temperature per turn," the reward is "100%." ​​In addition, the reward is "-10%" for every degree above the "target temperature per turn." Note that this reward is not limited to the above values ​​and can be set arbitrarily.

[0046] For example, as shown in Figure 4, assume that the temperature at the start of control is 48 degrees, and the target temperature for the first turn is 41 degrees. In this case, if the temperature after the first turn control is 42 degrees, the reward will be "90%" since it is 1 degree higher than the target temperature for each turn. Note that Figure 4 shows an example of a high temperature alert reward, but in the case of a low temperature alert reward, if the temperature is, for example, 1 degree lower than the target temperature for each turn, the reward is calculated in the same way, with "-10%".

[0047] In (metric 2), the reward is calculated based on the temperature change in the most recent turn. Then, if the temperature before control is higher than the target temperature, a reward (high temperature alert reward) is calculated as shown in the following (Equation 1). "Temperature lowered this time" / "Temperature that should have been lowered originally" x 100% (Formula 1) For example, if the temperature was lowered by 8 degrees and the temperature that should have been lowered was 10 degrees, the reward is calculated as "80%." If the reward calculated using (Formula 1) is higher than 100%, the reward will be 100%. In addition, when calculating the low temperature vigilance reward, the reward is calculated using the following (Equation 2). "The temperature that rose this time" / "The temperature that should have risen" x 100% (Equation 2)

[0048] The area reward calculation unit 241 may evaluate the control result using only (index 1), or may evaluate the control result using both (index 1) and (index 2). When evaluating using both (index 1) and (index 2), even if the target temperature for each turn is not reached in (index 1) and the reward is not 100%, the control result of the air conditioning control value can be evaluated with a higher reward for the larger the value of "current temperature drop" in (index 2) (the larger the temperature change). The information on the reward (area reward) calculated by the area reward calculation unit 241 is stored in the operation history information 201 by the air conditioning control execution unit 230 for each Situation classification, together with external factor information, air conditioning control values ​​for each turn (target temperature, air volume, etc. for each turn), and air conditioning power consumption information.

[0049] Returning to FIG. 2, the overall remuneration calculation unit 242 calculates a remuneration for the entire floor (overall remuneration) using the remuneration for each air-conditioning controlled zone 20 calculated by the zone remuneration calculation unit 241. The overall reward calculation unit 242 subtracts the reward for the "warning zone" from the sum of the rewards (zone rewards) for each air-conditioning controlled zone 20 (total of maximum rewards) to calculate the overall reward so that 100 points is the maximum score. Furthermore, the overall remuneration calculation unit 242 determines whether the air conditioning control for the entire floor is successful or not depending on whether the calculated overall remuneration is equal to or greater than a predetermined threshold value (pass threshold value) related to the overall remuneration.

[0050] Here, the alert area is defined as follows, for example. If the area reward is 90% or more, it is considered "safe", if the area reward is between 85% and 90%, it is considered "caution", and if the area reward is less than 85%, it is considered "alert". Then, the total reward is calculated using the following (Equation 3). (Total Zone Rewards - Number of Warning Zones x 10 - Number of Alert Zones x 30) / Number of Climate Control Zones ...(Formula 3)

[0051] The overall reward calculation unit 242 sets the pass threshold for the entire area so that, for example, if the caution zones account for 20% of the total and the alert zones account for 40% of the total, the area will fail. Specifically, if the number of air-conditioning controlled zones is "8", the caution zones are "-10" points, and the alert zones are "-30" points, the pass threshold for the entire floor is calculated as follows: (8 x 100 - 8 x 0.2 x 10 - 8 x 0.4 x 30) / 8 = 688 / 8 = 86 (%) Therefore, if the calculated total remuneration is equal to or greater than a predetermined pass threshold (86%), the total remuneration calculation unit 242 judges the application as a pass, and if the calculated total remuneration is less than the predetermined pass threshold, the total remuneration calculation unit 242 judges the application as a fail. If the total reward calculation unit 242 determines that the turn has failed, the processing for the next turn is cancelled. In addition, if the overall reward calculated by the overall reward calculation unit 242 for the air conditioning control learning model 203 of the corresponding Situation classification exceeds a predetermined pass threshold (86%) and is judged to be passed, the learning model management unit 222 terminates the learning phase and transitions to the operation phase.

[0052] In addition, the overall reward calculation unit 242 may be configured to determine that the temperature is finally passed when both the high temperature alert reward and the low temperature alert reward are passed. For example, if the initial temperature before control of the air conditioner 2 is higher than the target temperature and control is performed using an air conditioning control value based on the high temperature alert reward, the temperature may exceed a predetermined pass threshold, but the temperature may be controlled to be too low beyond the target temperature. In this case, excessive air conditioning power consumption is consumed. Therefore, if the temperature before control is lower than the target temperature, control is performed based on the low temperature alert reward until it is determined to be passed. In this way, by the overall reward calculation unit 242 determining that both the high temperature alert reward and the low temperature alert reward are passed, an air conditioning control value that further reduces the amount of air conditioning power consumption can be obtained.

[0053] For controls that the overall remuneration calculation unit 242 has determined to be successful in terms of overall remuneration, the correspondence information generation unit 250 acquires, for each Situation classification, the air conditioning control values ​​(target temperature, air volume, etc.) for each turn and the power consumption of the air conditioner 2 when that turn was executed, to generate control value-power consumption correspondence information 204. Note that the air conditioning power consumption of the air conditioner 2 is the total power consumption of each air conditioner 2 measured by a power consumption measurement means (not shown) that monitors the air conditioner 2.

[0054] Furthermore, when the degenerate operation is completed by the degenerate operation unit 223, the correspondence information generation unit 250 updates the control value-power amount correspondence information 204 with the air conditioning control value that will result in a lower power consumption cost at the time the degenerate operation is completed and the amount of air conditioning power consumption when that air conditioning control value is executed.

[0055] The server control unit 300 predicts the server power consumption of each server 3 in possible placement patterns (multiple placement patterns) based on schedule information for creation and deletion of virtual resources, and calculates the server power consumption for each placement control area 30. The server control unit 300 also acquires information on the air conditioning power consumption due to the execution of air conditioning control at the time of completion of degenerate operation in the operation phase of the corresponding Situation classification in each placement pattern, calculates the total value of the server power consumption and the air conditioning power consumption in each placement pattern, determines the placement pattern with the smallest value, and executes the placement of virtual resources. The server control unit 300 includes an arrangement pattern calculation unit 310 , a server power consumption estimation unit 320 , and an arrangement pattern determination unit 330 .

[0056] At the start of each control turn, the allocation pattern calculation unit 310 acquires schedule information for the generation and deletion of virtual resources, and obtains virtual resource amount information (e.g., the number of CPU cores) to be newly allocated. Then, based on the most recent resource usage status (e.g., CPU usage rate), the allocation pattern calculation unit 310 calculates an allocation pattern in which new virtual resources are allocated to each server 3. After allocating the virtual resources to each server 3, the allocation pattern calculation unit 310 ensures that the resource occupation amount in each server 3 is equal to or less than the server capacity (upper limit value) x a predetermined threshold value.

[0057] The server power consumption estimation unit 320 predicts the power consumption of each server 3 by utilizing a learning model (server power consumption learning model 301) for each placement pattern calculated by the placement pattern calculation unit 310. Then, the server power consumption estimation unit 320 calculates the total server power consumption for each placement control area 30 in each placement pattern based on the server placement configuration for each placement control area 30.

[0058] Specifically, the server power consumption estimation unit 320 predicts the power consumption of each server 3 for each deployment pattern using a learning model (server power consumption learning model 301) that uses the server 3's air intake temperature and resource usage information (e.g., CPU usage rate, memory usage rate, etc.) as input data and the server power consumption as output data. The server power consumption learning model 301 may be generated in advance using the intake port temperature, the resource usage of the server 3, and the server power consumption information, which is the result information at that time, as learning data.

[0059] In addition, the server power consumption estimation unit 320 calculates the server power consumption for each placement control area 30 by adding up the server power consumption of each server 3 in the placement control area 30 based on the server placement configuration for each placement control area 30. The server power consumption estimation unit 320 outputs the calculated server power consumption for each placement control area 30 to the external factor acquisition unit 211 of the air conditioning control unit 200 in the learning phase and the operation phase.

[0060] The placement pattern determination unit 330 totals the server power consumption for each placement control area 30 for each placement pattern of the virtual resources, calculates the sum of the total server power consumption, which is the sum value, and the air conditioning power consumption obtained from the control value-power consumption correspondence information 204, and determines the placement pattern with the smallest total amount as the placement pattern in which to actually place the virtual resources.

[0061] Specifically, in the operation phase, when degenerate operation for each Situation classification is completed and the optimal air conditioning control value that reduces the amount of power consumption is stored in the control value-power amount correspondence information 204, the placement pattern determination unit 330 executes the following processing. The placement pattern determination unit 330 first determines the current situation classification by inquiring of the situation determination unit 212 of the air conditioning control unit 200. Then, the placement pattern determination unit 330 references the control value-power amount correspondence information 204 for that situation classification, and acquires, for each placement pattern, information on the air conditioning control values ​​(target temperature, air volume, etc.) and the amount of air conditioning power consumption when the air conditioner 2 executes those air conditioning control values.

[0062] The placement pattern determination unit 330 totals the server power consumption in the placement control area 30 for each placement pattern of the virtual resources, calculates the sum of the total server power consumption and the acquired air conditioning power consumption, and determines the placement pattern with the smallest total amount as the placement pattern in which the virtual resources will actually be placed. Then, the placement pattern determination unit 330 executes placement of the virtual servers according to the determined placement pattern via a server management device (not shown) and executes control by the air conditioning control unit 200 based on the air conditioning control values ​​(target temperature, air volume, etc.) of the air conditioner 2 in the Situation.

[0063] <Processing flow> Next, the flow of processing executed by the power amount reduction control device 100 according to this embodiment will be described. Here, we will explain the learning phase (see Figure 5), which searches for air conditioning control values ​​that satisfy the reward for each situation classification, the operation phase (see Figure 6), which performs degenerate operation and searches for more optimal air conditioning control values, and the placement pattern determination process (see Figure 7), which determines where to place virtual resources.

[0064] <Learning Phase> FIG. 5 is a diagram illustrating the learning phase in which air-conditioning control values ​​that satisfy rewards are searched for in each Situation classification. In the learning phase, the power amount reduction control device 100 performs a search by repeating the PDCA cycle until an air-conditioning control value that satisfies the reward (temperature reward) for each Situation classification is found.

[0065] First, in step A1, the situation recognition unit 210 (external factor acquisition unit 211) of the power amount reduction control device 100 acquires information on external factors (floor average temperature, outside temperature, server power consumption for each placement control area 30). Then, the situation recognition unit 210 (the situation determination unit 212) determines to which situation classification the acquired information on the external factor belongs. It should be noted that step A1 corresponds to the Plan step of the PDCA cycle.

[0066] Next, in step A2 (P (Plan): plan), control value searcher 220 calculates an air conditioning control value for each Situation classification. Note that control value generator 221 of control value searcher 220 randomly searches for air conditioning control values ​​up to N times. Then, when N times is reached, learning model manager 222 of control value searcher 220 causes air conditioning control learning model 203 to learn learning data (external factor information and reward information associated with air conditioning control values). As a result, when N times is exceeded, learning model manager 222 inputs situation information into air conditioning control learning model 203, thereby causing information on air conditioning control values ​​to be output.

[0067] Next, in step A3 (D(Do): Execute), the air conditioning control execution unit 230 executes actual control of each air conditioner 2 based on the air conditioning control values ​​(target temperature, air volume, etc.) calculated by the control value search unit 220.

[0068] Next, in step A4 (C(Check):Evaluation), the reward calculation unit 240 calculates rewards (area reward and total reward) that are the results of executing the control of the air-conditioning control value. Here, if the calculated total reward is less than the pass / fail threshold and the reward calculation unit 240 judges that the turn has failed, the reward calculation unit 240 does not execute the processing of the subsequent turns.

[0069] In step A4, if the overall reward is pass, in the next step A5 (A (Action): Improve), the control value search unit 220 (learning model management unit 222) updates the air conditioning control learning model 203 for each situation classification based on the learning data (air conditioning control learning data 202) that associates rewards with external factor information and air conditioning control values.

[0070] In the learning phase, a search is performed by repeating this PDCA cycle until a control value for which the reward (overall reward) is determined to be successful is found for each Situation classification. When an air-conditioning control value that satisfies the reward (overall reward) is found, the correspondence information generator 250 generates control value-power-amount correspondence information 204 for each Situation classification, which stores the air-conditioning control value (target temperature, air volume, etc.) for each turn and the total power consumption (air-conditioning power consumption) of the air conditioners 2 when that turn is performed. This transitions the corresponding Situation classification from the learning phase to the operation phase.

[0071] <Operation Phase> FIG. 6 is a diagram illustrating an operation phase in which degenerate operation is performed and more optimal air-conditioning control values ​​are searched for. In the operation phase, in the initial state (a state in which degenerate operation is not yet complete), the control value search unit 220 (degenerate operation unit 223) processes the control value by one step from the control value with the smallest air-conditioning control cost that satisfies the current reward, and runs the PDCA cycle to find a control value that cannot be lowered any further, and then completes the degenerate operation. Then, if the corresponding Situation classification is degenerate operation completion, the air-conditioning control value at the time of completion of degenerate operation is called up, and air-conditioning control is executed.

[0072] First, in step B1 (P (Plan): plan), the situation recognition unit 210 (external factor acquisition unit 211) of the power reduction control device 100 acquires information on external factors (floor average temperature, outside temperature, server power consumption for each placement control area). Then, the situation recognition unit 210 (the situation determination unit 212) determines to which situation classification the acquired information on the external factor belongs.

[0073] Next, in step B2 (P (Plan): plan), if the degenerate operation is incomplete, the control value search unit 220 (degenerate operation unit 223) refers to the control value-power correspondence information 204 in which air conditioning control values ​​that satisfy a specified reward are stored, and generates an air conditioning control value by lowering the control value by one step from the air conditioning control value with the lowest cost that satisfies the current reward.

[0074] Next, in step B3 (D (Do): Execute), the air conditioning control execution unit 230 executes control of each air conditioner 2 based on the control value (target temperature, air volume, etc.) calculated by the control value search unit 220, i.e., the control value lowered by one step.

[0075] Next, in step B4 (C (Check): Evaluation), the reward calculation unit 240 calculates rewards (area reward and overall reward) that are the result of executing control of the air-conditioning control value. If the overall reward is acceptable, the degenerate operation unit 223 further lowers the control value by one step (step B2) and causes the air-conditioning control execution unit 230 to control the air conditioner 2 in step B3 (A (Action): Improvement). Then, the degenerate operation unit 223 determines the control value one step before the control value for which the reward was not acceptable as the control value with the lowest control cost.

[0076] In addition, after failing the reward after the Z+1th degeneration, the degeneration operation unit 223 may execute a one-step rollback to the previous Zth degeneration, and recheck whether the reward is met again after the Zth degeneration. If the recheck after the Zth degeneration passes, the degeneration operation unit 223 sets the control value as the air-conditioning control value at the time of completion of the degeneration. If the recheck after the Zth degeneration fails, the degeneration operation unit 223 repeats the process of going back one step, and sets the control value at the reward-passing stage as the air-conditioning control value at the time of completion of the degeneration.

[0077] When the control value with the lowest control cost has been determined (when degenerate operation has been completed), the correspondence information generation unit 250 updates the control value-power amount correspondence information 204 with the air conditioning control value that will result in the lowest power consumption cost when degenerate operation is completed and the amount of air conditioning power consumption when that air conditioning control value is executed.

[0078] When the degenerate operation is completed and the process returns to step B1, the situation recognition unit 210 (external factor acquisition unit 211) acquires information on external factors (average floor temperature, outside temperature, and server power consumption for each placement control area), and the situation determination unit 212 determines to which situation classification the acquired information on external factors belongs.

[0079] Then, in step B5, if the degenerate operation for that Situation classification is complete, control value searcher 220 (control value generator 221) generates an air conditioning control value based on the updated control value / power amount correspondence information 204. Next, the process proceeds to step B6, where the air conditioning control execution unit 230 executes control of the air conditioner 2 based on the air conditioning control values ​​in the state in which the degenerate operation has been completed.

[0080] By performing this processing for each Situation classification, the power consumption reduction control device 100 can generate, for each Situation classification, information on the air conditioning control value that can further reduce air conditioning power consumption and the air conditioning power consumption at that time as control value power consumption correspondence information 204.

[0081] <Placement pattern determination process that determines the placement destination of virtual resources> FIG. 7 is a flowchart showing the flow of an allocation pattern determination process for determining the allocation destination of a virtual resource, which is executed by the power amount reduction control device 100 according to this embodiment.

[0082] First, the server control unit 300 (allocation pattern calculation unit 310) of the power amount reduction control device 100 acquires schedule information for generating or deleting virtual resources at the start of each control turn. Then, the allocation pattern calculation unit 310 calculates an allocation pattern in which a new virtual resource is added (or deleted) to the current server resource usage status (step S101).

[0083] Next, the server power consumption estimation unit 320 of the server control unit 300 predicts the server power consumption of each server 3 for each placement pattern calculated by the placement pattern calculation unit 310, for example, by using the server power consumption learning model 301. Then, the server power consumption estimation unit 320 sums up the server power consumption of the servers 3 in each placement control area 30 for each placement pattern, and calculates the server power consumption for each placement control area 30 (step S102).

[0084] Next, the situation recognition unit 210 (external factor acquisition unit 211) of the air conditioning control unit 200 acquires information on external factors (floor average temperature, outside temperature, server power consumption for each placement control area) for each placement pattern. Then, the situation determination unit 212 determines the situation classification (step S103).

[0085] Next, the placement pattern determination unit 330 refers to the control value-power correspondence information 204 in the determined Situation classification, and obtains, for each placement pattern, information on the air conditioning control value when the degenerate operation of the operation phase is completed and the air conditioning power consumption when that control is executed (step S104).

[0086] Next, the placement pattern determination unit 330 sums up the server power consumption for each placement control area 30 in each placement pattern, calculates the sum of the total server power consumption and the air conditioning power consumption, and determines the placement pattern with the smallest sum (step S105). Then, information on the virtual server placement pattern determined by the placement pattern determination unit 330 is transmitted to a server management device (not shown), thereby causing actual placement control of the virtual servers to be performed.

[0087] In this way, the power reduction control device 100 can reduce the total power consumption of the data center, which is made up of the server power consumption and the air conditioning power consumption, in response to the facility conditions that differ for each data center (DC) 10.

[0088] <Hardware configuration> The power amount reduction control device 100 according to this embodiment is realized by a computer 900 having a configuration as shown in FIG. 8 is a hardware configuration diagram showing an example of a computer 900 that realizes the functions of the power amount reduction control device 100 according to this embodiment. The computer 900 has a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM 903, an HDD (Hard Disk Drive) 904, an input / output I / F (Interface) 905, a communication I / F 906, and a media I / F 907.

[0089] The CPU 901 operates based on a program stored in the ROM 902 or the HDD 904, and performs control by the control unit. The ROM 902 stores a boot program executed by the CPU 901 when the computer 900 is started up, programs related to the hardware of the computer 900, and the like.

[0090] The CPU 901 controls an input device 910 such as a mouse or a keyboard, and an output device 911 such as a display or a printer, via an input / output I / F 905. The CPU 901 acquires data from the input device 910 via the input / output I / F 905, and outputs generated data to the output device 911. Note that a GPU (Graphics Processing Unit) or the like may be used as a processor together with the CPU 901.

[0091] The HDD 904 stores programs executed by the CPU 901 and data used by the programs, etc. The communication I / F 906 receives data from other devices via a communication network (e.g., NW (Network) 920) and outputs the data to the CPU 901, and also transmits data generated by the CPU 901 to other devices via the communication network.

[0092] The media I / F 907 reads a program or data stored in the recording medium 912, and outputs it to the CPU 901 via the RAM 903. The CPU 901 loads a program related to a target process from the recording medium 912 onto the RAM 903 via the media I / F 907, and executes the loaded program. The recording medium 912 is an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase change rewritable Disc), a magneto-optical recording medium such as an MO (Magneto Optical disk), a magnetic recording medium, a semiconductor memory, or the like.

[0093] For example, when the computer 900 functions as the power amount reduction control device 100 of the present invention, the CPU 901 of the computer 900 executes a program loaded onto the RAM 903 to realize the functions of the power amount reduction control device 100. Furthermore, the HDD 904 stores data in the RAM 903. The CPU 901 reads and executes a program relating to a target process from a recording medium 912. Additionally, the CPU 901 may read a program relating to a target process from another device via a communication network (NW 920).

[0094] <Effects> The effects of the power amount reduction control device 100 according to the present invention will be described below. The power consumption reduction control device according to the present invention is a power consumption reduction control device 100 that controls a plurality of servers 3 and a plurality of air conditioners 2 in a data center 10, and a plurality of placement control areas 30 in which a group of servers in which virtual resources are placed are placed on the floor of the data center 10, and a plurality of air conditioning control areas 20 that are areas for measuring the effect of air conditioning control by the plurality of air conditioners 2 and are located on either the intake side or the exhaust side of the group of servers are set, and the power consumption reduction control device 100 calculates a flow rate from an average value of temperatures measured in the plurality of air conditioning control areas 20. an external factor acquisition unit 211 for acquiring information on external factors related to air conditioning control, including an average air temperature, an external temperature which is the temperature outside the data center 10, and a server power consumption amount for each placement control area which is a predicted amount when a virtual resource is placed on the server 3; a situation determination unit 212 for dividing the value of each external factor into a predetermined range width, combining the ranges divided for each external factor, defining them as a situation classification, and determining which situation classification the acquired information on the external factor belongs to; a control value search unit 220 that calculates air conditioning control values ​​including at least a target temperature to be set for the multiple air conditioners 2; an air conditioning control execution unit 230 that executes control of the multiple air conditioners 2 using the calculated air conditioning control values; a correspondence information generation unit 250 that acquires the air conditioning control values ​​of the multiple air conditioners 2 and the air conditioning power consumption of the multiple air conditioners 2 when control is performed using the air conditioning control values, and generates control value-power amount correspondence information 204 that associates the air conditioning control values ​​of the multiple air conditioners 2 with the air conditioning power consumption for each Situation classification; a server power consumption estimation unit 320 that estimates the server power consumption of the server group belonging to each of the placement control areas 30 for each calculated placement pattern; a situation classification determination result obtained from a situation determination unit 212 using the server power consumption of each placement control area 30, and referring to the control value-power consumption correspondence information 204, the air conditioning control values ​​and air conditioning power consumption of the multiple air conditioners 2 in the situation classification are obtained for each placement pattern, and for each placement pattern,and an arrangement pattern determination unit 330 that sums up the server power consumption for each arrangement control area 30, calculates the sum of the total server power consumption and the air conditioning power consumption, and determines the arrangement pattern with the smallest calculated total amount as the arrangement pattern for allocating the virtual resources.

[0095] In this way, when allocating new virtual resources, the power consumption reduction control device 100 can take into account the server power consumption and air conditioning power consumption and determine the optimal virtual resource allocation pattern that reduces the power consumption of the entire data center 10. The power reduction control device 100 can determine the placement of virtual servers in the placement controlled areas 30 and air conditioning controlled areas 20 set in the data center 10, taking into consideration different facility conditions (airflow, server placement configuration, air conditioning placement position, thermal cooling efficiency, etc.) and temperature differences between the air conditioning controlled areas 20 due to heat accumulation. Therefore, it is possible to realize reductions in power consumption suited to the respective conditions of each data center 10.

[0096] The power consumption reduction control device 100 also includes a reward calculation unit 240 that calculates a reward for evaluating the results of the air conditioning control execution unit 230 controlling the multiple air conditioners 2 with air conditioning control values, using the target temperature as an index, and the control value search unit 220 includes a learning model management unit 222 that generates an air conditioning control learning model 203 for each situation classification, which outputs air conditioning control values ​​for the multiple air conditioners 2 when information on external factors is input, by machine learning using information on external factors, the air conditioning control values ​​of the multiple air conditioners 2, and the reward as learning data so as to maximize the reward.

[0097] In this way, the power amount reduction control device 100 generates the air-conditioning control learning model 203 for each Situation classification, thereby making it possible to generate optimal air-conditioning control values ​​for the air conditioner 2 for each Situation classification.

[0098] The power consumption reduction control device 100 also includes a reward calculation unit 240 that calculates a reward for evaluating the results of the air conditioning control execution unit 230 controlling multiple air conditioners 2 with air conditioning control values, using a target temperature as an index, and a degenerate operation unit 223. When the reward satisfies a predetermined reward condition, the correspondence information generation unit 250 generates control value-to-power consumption correspondence information 204 for the Situation classification. The degenerate operation unit 223 performs degenerate operation to search for an air conditioning control value with a lower power consumption cost within a range that satisfies the predetermined reward condition, for each Situation classification. The correspondence information generation unit 250 updates the control value-to-power consumption correspondence information 204 for each Situation classification using the air conditioning control value searched for by the degenerate operation unit 223.

[0099] In this way, the power reduction control device 100 can search for an air-conditioning control value with a lower power consumption cost by degenerate operation within a range that satisfies a predetermined remuneration condition. This makes it possible to further reduce the power consumption of the entire data center 10.

[0100] In addition, in the power consumption reduction control device 100, the degeneracy operation unit 223 defines the air conditioning control value into M (an integer equal to or greater than 2) stages between the minimum value and the maximum value, and when the calculated reward is a pass that satisfies a predetermined reward condition, the process of gradually lowering the air conditioning control value in a direction that reduces the amount of power consumption is repeated, and the control value one stage before the air conditioning control value that fails because the calculated reward does not satisfy the predetermined reward condition is determined to be the air conditioning control value with the smallest power consumption cost.

[0101] In this way, the power amount reduction control device 100 can search for an air conditioning control value that has a smaller power consumption cost within a range that satisfies a predetermined remuneration condition.

[0102] In addition, in the power consumption reduction control device 100, the degeneracy operation unit 223 defines the air conditioning control value into M (an integer equal to or greater than 2) stages between the minimum value and the maximum value, and if the calculated reward is a pass that satisfies a specified reward condition, repeats a process of gradually lowering the air conditioning control value in a direction that reduces the amount of power consumption, and rechecks whether the reward is satisfied with the control value one stage before the air conditioning control value that failed the calculated reward because it did not satisfy the specified reward condition, and if the recheck fails, repeatedly rechecks whether the reward is satisfied with the control value one stage before, and determines the air conditioning control value that passed the recheck as the air conditioning control value with the lowest power consumption cost.

[0103] In this way, the power amount reduction control device 100 can search for an air conditioning control value with a smaller power consumption cost within a range that satisfies a predetermined reward condition, while preventing disturbances.

[0104] The present invention is not limited to the above-described embodiment, and many modifications can be made by those having ordinary skill in the art within the technical scope of the present invention. [Explanation of symbols]

[0105] 1. Power reduction control system 2 Air conditioner 3 Server 10 Data Center (DC) 20 Climate Controlled Area 30 Placement Control Area 100 Electric power reduction control device 200 Air conditioning control unit 201 Operational History Information 202 Air conditioning control learning data 203 Air Conditioning Control Learning Model 204 Control value power amount correspondence information 210 Situational Awareness Division 211 External factor acquisition part 212 Situation judgment section 220 Control value search unit 221 Control value generator 222 Learning Model Management Department 223 Degraded Operation Unit 230 Air Conditioning Control Execution Unit 240 Remuneration Calculation Department 241 Area Remuneration Calculation Department 242 Total Remuneration Calculation Department 250 Correspondence information generation unit 300 Server control unit 301 Server Power Consumption Learning Model 310 Placement pattern calculation unit 320 Server Power Consumption Estimation Unit 330 Placement pattern determination unit

Claims

1. A power consumption reduction control device that controls a plurality of servers and a plurality of air conditioners in a data center, A floor of the data center is provided with a plurality of placement control areas in which a group of servers on which virtual resources are placed is placed, and a plurality of air conditioning control areas which are areas for measuring the effect of air conditioning control by the plurality of air conditioners and are located on either the intake side or the exhaust side of the group of servers; The power reduction control device includes: an external factor acquisition unit that acquires information about external factors related to air conditioning control, including an average floor temperature calculated from an average value of temperatures measured in a plurality of the air conditioning control zones, an external temperature that is the temperature outside the data center, and server power consumption for each of the placement control zones that is a predicted amount when the virtual resource is placed on the server; a situation determination unit that divides the value of each of the external factors into a predetermined range width, defines a combination of the divided ranges for each of the external factors as a situation classification, and determines to which situation classification the acquired information of the external factor belongs; a control value search unit that calculates air conditioning control values, including at least a target temperature, to be set for the plurality of air conditioners for each of the determined situation classifications; an air-conditioning control execution unit that executes control of the plurality of air conditioners using the calculated air-conditioning control value; a correspondence information generating unit that acquires air-conditioning control values ​​of the plurality of air conditioners and air-conditioning power consumption of the plurality of air conditioners when control is performed using the air-conditioning control values, and generates control value-power correspondence information that associates the air-conditioning power consumption with the air-conditioning control values ​​of the plurality of air conditioners for each of the Situation classifications; an allocation pattern calculation unit that calculates an allocation pattern for allocating the virtual resource to be newly allocated to one of the plurality of servers; a server power consumption estimation unit that estimates server power consumption of a server group that belongs to each of the placement control areas for each of the calculated placement patterns; obtaining, from the Situation determination unit, a determination result of a Situation classification using the server power consumption of each of the placement control areas, and obtaining, for each placement pattern, air conditioning control values ​​and the air conditioning power consumption of the plurality of air conditioners in the Situation classification by referring to the control value-power amount correspondence information; an arrangement pattern determination unit that sums up the server power consumption for each of the arrangement control areas for each of the arrangement patterns, calculates a sum of the total server power consumption, which is the sum, and the air conditioning power consumption, and determines the arrangement pattern having the smallest calculated sum as the arrangement pattern for allocating the virtual resources; A power reduction control device comprising:

2. The power reduction control device includes: a reward calculation unit that calculates a reward for evaluating a result of the air-conditioning control execution unit controlling the plurality of air conditioners using the air-conditioning control value, using the target temperature as an index; The control value search unit a learning model management unit that generates an air-conditioning control learning model for each of the situation classifications, the learning model managing unit inputting information about the external factors, the air-conditioning control values ​​of the plurality of air conditioners, and the reward, and performing machine learning so as to maximize the reward, the learning model managing unit outputting air-conditioning control values ​​of the plurality of air conditioners when information about the external factors is input; The power reduction control device according to claim 1 .

3. The power reduction control device includes: a reward calculation unit that calculates a reward for evaluating a result of the air-conditioning control execution unit controlling the plurality of air conditioners using the air-conditioning control value, using the target temperature as an index; and a degenerate operation unit, the correspondence information generation unit generates the control value-power correspondence information for the Situation classification when the reward satisfies a predetermined reward condition, the degenerate operation unit performs a degenerate operation to search for an air-conditioning control value having a lower power consumption cost within a range that satisfies the predetermined reward condition, for each of the situation classifications; the correspondence information generation unit updates the control value-power amount correspondence information for each of the situation classifications by using the air-conditioning control values ​​searched for by the degenerate operation unit; The power reduction control device according to claim 1 .

4. The degenerate operating unit is The air conditioning control value is defined in M ​​stages (M is an integer equal to or greater than 2) between a minimum value and a maximum value, When the calculated remuneration is a pass that satisfies a predetermined remuneration condition, a process of gradually lowering the air-conditioning control value in a direction that reduces the amount of power consumption is repeated, and the control value one step before the air-conditioning control value that fails because the calculated remuneration does not satisfy the predetermined remuneration condition is determined to be the air-conditioning control value with the smallest power consumption cost; 4. The power reduction control device according to claim 3,

5. The degenerate operating unit is The air conditioning control value is defined in M ​​stages (M is an integer equal to or greater than 2) between a minimum value and a maximum value, If the calculated reward is a pass that satisfies a predetermined reward condition, a process of gradually lowering the air conditioning control value in a direction that reduces the amount of power consumption is repeated, and a control value one step before the air conditioning control value that failed the calculated reward because it did not satisfy the predetermined reward condition is rechecked to see if it satisfies the reward, and if the recheck fails, a further recheck is repeated to see if the control value one step before satisfies the reward, and the air conditioning control value that passed the recheck is determined to be the air conditioning control value with the smallest power consumption cost; 4. The power reduction control device according to claim 3,

6. A power reduction control method for a power reduction control device that controls a plurality of servers and a plurality of air conditioners in a data center, comprising: A floor of the data center is provided with a plurality of placement control areas in which a group of servers on which virtual resources are placed is placed, and a plurality of air conditioning control areas which are areas for measuring the effect of air conditioning control by the plurality of air conditioners and are located on either the intake side or the exhaust side of the group of servers; The power reduction control device includes: acquiring information on external factors related to air conditioning control, including an average floor temperature calculated from an average value of temperatures measured in a plurality of the air conditioning control zones, an outside temperature which is the temperature outside the data center, and server power consumption for each placement control zone which is a predicted amount when the virtual resource is placed on the server; a step of dividing the value of each of the external factors into a predetermined range width, combining the divided ranges for each of the external factors to define a situation classification, and determining to which situation classification the acquired information of the external factor belongs; calculating air conditioning control values, including at least a target temperature, to be set for the plurality of air conditioners for each of the determined situation classifications; executing control of the plurality of air conditioners using the calculated air conditioning control values; acquiring air-conditioning control values ​​of the plurality of air conditioners and air-conditioning power consumption of the plurality of air conditioners when control is performed according to the air-conditioning control values, and generating control value-power correspondence information that associates the air-conditioning power consumption with the air-conditioning control values ​​of the plurality of air conditioners for each of the Situation classifications; calculating an allocation pattern for allocating the virtual resource to be newly allocated to one of the plurality of servers; estimating, for each of the calculated placement patterns, server power consumption amounts of a server group that belongs to each of the placement control areas; acquiring a determination result of a situation classification using the server power consumption of each of the placement control areas, and acquiring the air conditioning control values ​​and the air conditioning power consumption of the plurality of air conditioners in the situation classification for each placement pattern by referring to the control value-power amount correspondence information; a step of summing up the server power consumption for each of the placement control areas in each placement pattern, calculating a sum of the total server power consumption, which is the sum of the server power consumption and the air conditioning power consumption, and determining the placement pattern in which the calculated sum is the smallest as the placement pattern in which the virtual resources are placed; A power reduction control method comprising the steps of:

7. A power reduction control system including a data center having a plurality of servers and a plurality of air conditioners, and a power reduction control device that controls the amount of power consumption of the data center, A floor of the data center is provided with a plurality of placement control areas in which a group of servers on which virtual resources are placed is placed, and a plurality of air conditioning control areas which are areas for measuring the effect of air conditioning control by the plurality of air conditioners and are located on either the intake side or the exhaust side of the group of servers; The power reduction control device includes: an external factor acquisition unit that acquires information about external factors related to air conditioning control, including an average floor temperature calculated from an average value of temperatures measured in a plurality of the air conditioning control zones, an external temperature that is the temperature outside the data center, and server power consumption for each of the placement control zones that is a predicted amount when the virtual resource is placed on the server; a situation determination unit that divides the value of each of the external factors into a predetermined range width, defines a combination of the divided ranges for each of the external factors as a situation classification, and determines to which situation classification the acquired information of the external factor belongs; a control value search unit that calculates air conditioning control values, including at least a target temperature, to be set for the plurality of air conditioners for each of the determined situation classifications; an air-conditioning control execution unit that executes control of the plurality of air conditioners using the calculated air-conditioning control value; a correspondence information generating unit that acquires air-conditioning control values ​​of the plurality of air conditioners and air-conditioning power consumption of the plurality of air conditioners when control is performed using the air-conditioning control values, and generates control value-power correspondence information that associates the air-conditioning power consumption with the air-conditioning control values ​​of the plurality of air conditioners for each of the Situation classifications; an allocation pattern calculation unit that calculates an allocation pattern for allocating the virtual resource to be newly allocated to one of the plurality of servers; a server power consumption estimation unit that estimates server power consumption of a server group that belongs to each of the placement control areas for each of the calculated placement patterns; obtaining, from the Situation determination unit, a determination result of a Situation classification using the server power consumption of each of the placement control areas, and obtaining, for each placement pattern, air conditioning control values ​​and the air conditioning power consumption of the plurality of air conditioners in the Situation classification by referring to the control value-power amount correspondence information; an arrangement pattern determination unit that sums up the server power consumption for each of the arrangement control areas for each of the arrangement patterns, calculates a sum of the total server power consumption, which is the sum, and the air conditioning power consumption, and determines the arrangement pattern having the smallest calculated sum as the arrangement pattern for allocating the virtual resources; A power consumption reduction control system comprising:

8. A program for causing a computer to function as the power reduction control device according to any one of claims 1 to 5.

Citation Information

Patent Citations

  • Information processing system and power saving control method thereof, and device

    JP2013092951A

  • Information apparatus and method of optimizing cooling efficiency of air conditioner

    US20120116595A1

  • Electronic computer system and virtual machine deployment method

    WO2013042615A1

  • Deployment search program, deployment search method, and deployment search device

    WO2017168664A1

  • Air conditioning control device, power consumption prediction device, learning device, and air conditioning control method

    WO2021192201A1