Micromodule data center air conditioner group control method and system

Abstract

The invention provides a micromodule data center air conditioner group control method and system. The method comprises the steps that a clustering algorithm is adopted to divide a plurality of cabinets into a plurality of cabinet groups, wherein each air conditioner corresponds to one cabinet group. And determining a group target of each cabinet group according to the average temperature of the air inlet or the average temperature of the air outlet of each cabinet group. And the air conditioner corresponding to each cabinet group performs area control aiming at the group target. According to the area control target completion condition of the air conditioner corresponding to the cabinet group and / or the operation condition of the air conditioner, whether the air conditioner needs refrigeration cooperation or not is judged; if yes, refrigeration cooperation requests are sent to the air conditioners corresponding to the other cabinet groups, and the air conditioners receiving the refrigeration cooperation requests conduct refrigeration cooperation operation. According to the method, the defects of independent operation, frequent start and stop, excessive check-in and large idle electric energy consumption of each air conditioner in a traditional control mode are overcome by adopting a refrigeration cooperative control mode, the overall operation efficiency of the micromodule is remarkably improved, and the energy consumption of the micromodule air conditioner is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a micro-module data center air conditioning group control method and system. Background Art

[0002] With the rapid development of Internet services, more and more enterprises and institutions are establishing data centers to support the sustainable development of their various businesses. Among them, the micro-module data center, which forms an efficient management by placing the inter-row air conditioner and the cabinet in a closed hot and cold channel, has gradually become a hot spot in data center construction. The micro-module data center has a full module architecture including refrigeration module, power supply and distribution module, intelligent monitoring and lighting. The high integration can achieve rapid deployment and flexible expansion, and is suitable for different application environments. The micro-module data center adopts the closed cold channel method to improve the utilization rate of cold capacity and the energy efficiency of the refrigeration system. The cold air sent by the precision air conditioner enters the closed cold channel, is sucked in by the cabinet fan to cool the load, and then is sent back to the precision air conditioner through the hot channel for circulation. Due to the significant increase in the power density of the cabinet, the heat generation is concentrated, which is easy to cause local overheating or overcooling. It can be seen that it is of great significance to build a data center with high reliability, easy maintenance, and green energy saving. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a micro-module data center air conditioning group control method and system.

[0004] According to a first aspect, the present invention provides a method for group controlling air conditioners in a micromodule data center, wherein the micromodule data center includes multiple air conditioners and multiple cabinets; the method includes: dividing the multiple cabinets into multiple cabinet groups, wherein each cabinet group includes at least one cabinet, and each air conditioner corresponds to a cabinet group; determining a group target of each cabinet group according to an average air inlet temperature or an average air outlet temperature of each cabinet group; the air conditioner corresponding to each cabinet group performs regional control based on the group target; for each cabinet group, judging whether the air conditioner needs cooling coordination according to the completion of the regional control target of the air conditioner corresponding to the cabinet group and / or the operation status of the air conditioner; if necessary, sending a cooling coordination request to the air conditioners corresponding to other cabinet groups, and the air conditioner receiving the cooling coordination request performs cooling coordination operation.

[0005] Preferably, whether the air conditioner needs cooling coordination is determined based on the completion of the area control target of the air conditioner corresponding to the cabinet group and / or the operation status of the air conditioner, including: within a preset first set time period, if the air conditioner corresponding to the cabinet group fails to control the target temperature in the group target within the set temperature boundary conditions, it is determined that the air conditioner needs cooling coordination; otherwise, it is determined that the air conditioner does not need cooling coordination.

[0006] Preferably, whether the air conditioner needs refrigeration coordination is judged according to the completion of the area control target of the air conditioner corresponding to the cabinet group and / or the operating status of the air conditioner, including: determining the refrigeration demand of the air conditioner, and determining the required value of the compressor operating frequency of the air conditioner according to the refrigeration demand; if the required value of the compressor operating frequency of the air conditioner is within a preset operating range, the air conditioner does not need refrigeration coordination; if the required value of the compressor operating frequency of the air conditioner exceeds the maximum value of the preset operating range, the air conditioner needs refrigeration coordination.

[0007] Preferably, determining the cooling demand of the air conditioner includes: determining the cooling demand of the air conditioner according to the cooling temperature measurement value of the air conditioner corresponding to the cabinet group, the cooling temperature setting value of the air conditioner, the temperature control dead zone of the air conditioner and the temperature proportional band.

[0008] Preferably, the sending of cooling collaboration request to the air conditioners corresponding to other cabinet groups includes: calculating the Euclidean distance between the air conditioner and each of the air conditioners corresponding to the other cabinet groups; screening out the minimum value among the Euclidean distances, and taking the air conditioners corresponding to the other cabinet groups corresponding to the minimum Euclidean distance as cooling collaboration objects; and sending the cooling collaboration request to the cooling collaboration objects.

[0009] Preferably, the air conditioner that receives the refrigeration coordination request performs a refrigeration coordination operation, including: the refrigeration coordination request includes a coordinated control set temperature; the higher temperature between the coordinated control set temperature and the target temperature of the air conditioner that receives the refrigeration coordination request is used as the new target temperature of the air conditioner that receives the refrigeration coordination request; the air conditioner that receives the refrigeration coordination request performs a refrigeration coordination operation according to the new target temperature.

[0010] Preferably, the cabinets are divided into a plurality of cabinet groups according to the positions of the air conditioners and the cabinets, including: taking all the air conditioners as the cabinet group centers and the number of air conditioners as the k value of the clustering algorithm; calculating the Euclidean distance from each cabinet to the center of each cabinet group according to the positions of the air conditioners and the cabinets; and using a clustering algorithm to assign each cabinet to the cabinet group center closest to it according to the Euclidean distance from each cabinet to the center of each cabinet group, thereby forming a cabinet group.

[0011] Preferably, a clustering algorithm is used to allocate each cabinet to the cabinet group center that is closest to it according to the Euclidean distance from each cabinet to the center of each cabinet group, and also includes: if the Euclidean distance between the cabinet and the centers of multiple cabinet groups is the same, the cabinet is used as an intermediate cabinet, and the intermediate cabinet is simultaneously allocated to the centers of the multiple cabinet groups; after the clustering algorithm is completed, the total load of the cabinets in the multiple cabinet groups corresponding to the centers of the multiple cabinet groups is calculated; and according to the total load, the intermediate cabinet is allocated to the cabinet group with the smallest total load.

[0012] Preferably, the micromodule data center also includes multiple loads; the air conditioner corresponding to each cabinet group performs regional control based on the group target, including: within a preset second set time period at the beginning of the load startup, based on the trained predictive control model, according to the power of the load, the ambient temperature of the cabinets in the cabinet group, the air outlet temperature of the cabinet, and the air inlet temperature of the cabinet, output the cooling temperature setting value of the air conditioner; during the period of outputting the cooling temperature setting value of the air conditioner using the predictive control model, if it is detected that the air outlet temperature of the cabinet exceeds the preset first temperature range or the air inlet temperature of the cabinet exceeds the preset second temperature range, then the cooling temperature setting value of the air conditioner is calculated by an adaptive control method; wherein the adaptive control method includes calculating the cooling temperature setting value of the air conditioner according to the first temperature range, the second temperature range, the power of the load, the measured value of the cooling temperature of the air conditioner, and the air outlet temperature of the cabinet or the air inlet temperature of the cabinet.

[0013] Preferably, it also includes: after the second set time period, calculating the fluctuation rate of the load power of the micromodule data center; if the fluctuation rate exceeds the preset fluctuation rate threshold, using the predictive control model to output the cooling temperature setting value of the air conditioner; if the fluctuation rate does not exceed the preset fluctuation rate threshold, using the adaptive control method to calculate the cooling temperature setting value of the air conditioner.

[0014] Preferably, it also includes: in the return air control mode, the measured value of the cooling temperature of the air conditioner is the measured value of the return air temperature, and the cooling temperature setting value of the air conditioner is the return air temperature setting value; in the supply air control mode, the measured value of the cooling temperature of the air conditioner is the measured value of the supply air temperature, and the cooling temperature setting value of the air conditioner is the supply air temperature setting value.

[0015] According to the second aspect, the present invention provides a micro-module data center air conditioning group control system, including: a temperature acquisition module, a power acquisition module and a temperature control module; the temperature acquisition module is used to collect the air outlet temperature of the cabinet, the air inlet temperature of the cabinet, the measured value of the cooling temperature of the air conditioner, and the cabinet ambient temperature; the power acquisition module is used to collect load power; the temperature control module is used to execute any of the air conditioning group control methods described in the first aspect.

[0016] The present invention effectively overcomes the shortcomings of traditional control, such as independent operation of each air conditioner, frequent start and stop, and excessive idle power consumption due to excessive check-in. It adopts refrigeration collaborative control to timely adjust the control target of each air conditioner in the air-conditioning group, which significantly improves the overall operating efficiency of the micromodule, has better temperature balance and robustness, and effectively reduces the energy consumption of the micromodule air conditioner. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flowchart of a micro-module data center air conditioning group control method according to an embodiment of the present invention.

[0018] Figure 2 It is a flowchart of an air-conditioning group collaborative control method according to an embodiment of the present invention.

[0019] Figure 3 The figure is a schematic diagram of a collaborative control logic flow according to an embodiment of the present invention.

[0020] Figure 4 The figure is a schematic diagram of clustering cabinet groups based on air conditioner locations according to an embodiment of the present invention.

[0021] Figure 5 The diagram is a schematic diagram of key data trends of a micromodule data center during collaborative control optimization operation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments herein are only used for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to those of ordinary skill in the art that these specific details are not necessarily used to implement the present invention. In other examples, in order to avoid confusing the present invention, known procedures, materials or methods are not specifically described.

[0023] With the in-depth development of information technology, micro-module data centers have gradually become a hot spot for data center construction. Through a highly integrated design, micro-module data centers can provide power supply, power distribution, thermal management, environmental monitoring, system connection and distributed management in one system, providing users with an integrated integrated solution. The characteristic of micro-module data centers is that the inter-row air conditioners and cabinets are placed in a closed hot and cold channel, pursuing lower energy usage efficiency (PUE, Power Usage Effectiveness).

[0024] To ensure the safe and stable operation of the equipment in the computer room, multiple air conditioners need to operate redundantly. Conventional computer room air conditioners operate independently and lack effective coordination with each other, which easily leads to frequent start and stop of air conditioning equipment, excessive idleness of on-duty machines, lack of overall scheduling of individual air conditioners, and failure to adjust the number of on-duty machines in time when the load changes. At the beginning of the load startup, the temperature in the computer room gradually rises. When the temperature threshold is reached, the air conditioner starts to cool, and there is a phenomenon of cooling lag. Under low load rate operation, there is a waste of cooling capacity and competitive operation of air conditioners in parallel. Under uneven load operation, the heat load density is non-uniform. The temperature of the precision air conditioner is set based on the temperature of the hotspot monitoring, which will cause a large waste of energy.

[0025] like Figure 1 As shown, a micro-module data center air conditioning group control method provided according to an embodiment of the present invention includes: step S101 uses a clustering algorithm to divide multiple cabinets into multiple cabinet groups according to the air conditioner position and the cabinet position, wherein each air conditioner corresponds to a cabinet group. Steps S102 to S104 are executed in one control cycle: step S102 determines the group target of each cabinet group according to the average air inlet temperature or the average air outlet temperature of each cabinet group. Step S103 performs regional control on the air conditioner corresponding to each cabinet group for the group target. Step S104 When the group target cannot be achieved within the set time period in the regional control, and / or the operating frequency of the air conditioner corresponding to the cabinet group when executing the regional control is not within the preset efficient operating range, it is necessary to find neighboring air conditioners to perform collaborative control while executing the regional control. In this way, after calculating the cooling demand of each air conditioner for the cabinet group and the air conditioner setting parameters in the current control cycle, the regional control and collaborative control algorithms can be executed, and a cascade group control method with multiple machines in parallel can be adopted to dynamically adjust the number of air conditioners in operation, and collaboratively control the cooling capacity distribution of the air-conditioning equipment so that the air-conditioning equipment operates in a high-efficiency range, thereby realizing the air-conditioning energy efficiency optimization control of the micro-module data center.

[0026] The following is a detailed description of each step of the embodiment of the present invention:

[0027] For step S101, since multiple air conditioners need to work together in the micro-module data center, the cabinets served by each air conditioner are first determined. According to the channel type (single channel or multi-channel) of the micro-module data center, the cabinet layout and load data, the temperature sensor and the air conditioning data, etc., the simulation results of the cooling design and heat dissipation analysis are used to guide the clustering of cabinet groups and assign them to the corresponding air conditioners. Taking a single-channel micro-module data center as an example, the micro-module data center contains 7 cabinets, 2 air-cooled rack air conditioners, and 28 servers. The simulation software is used to simulate and analyze the airflow, heat transfer and energy transfer processes of the data center, and the physical and boundary conditions are set to calculate the temperature and air volume distribution of the data center, so as to predict and evaluate the thermal characteristics of the data center and provide guidance for the design of the cooling system. From the heat flow simulation results, it is found that in a single-channel micro-module data center, the air supply distance of the air conditioner affects the cooling efficiency. The closer the cabinet is to the air conditioner, the better the heat dissipation effect. After airflow organization and air conditioning design simulation, the rack air conditioner takes its own position as the center, and assigns the cabinet to the air conditioner with the closest cooling distance, so as to achieve airflow optimization and energy-saving configuration. Therefore, according to an embodiment of the present invention, multiple cabinets may be clustered into a cabinet group for cooling design, and one air conditioner corresponds to one cabinet group.

[0028] For a given data set of n cabinets, the number of clusters (groups) k is given according to the number of air conditioners, the group to which each cabinet belongs is initialized, and each cabinet is assigned to the group with the closest air conditioner according to the distance. The steps to implement the cabinet group clustering algorithm include:

[0029] (1) Selection of the number of groups k

[0030] The selection of k is generally determined according to actual needs. In one embodiment of the present invention, the value of k is given according to the number of air conditioners running in the micro-module data center.

[0031] (2) The distance from each cabinet to the center of the cluster

[0032] Given a cabinet dataset X, which contains n cabinets {X1,X2,…,X n}, where each cabinet has m-dimensional location attributes (such as deployment location and load power information). The goal of the cabinet group clustering algorithm is to cluster n cabinets according to the cabinet to air conditioner C j The minimum Euclidean distance is clustered into the specified cabinet group. First, k cluster centers {C1, C2, …, C k}, and then calculate each cabinet X according to formula (1) i To each cluster center C j The Euclidean distance.

[0033]

[0034] Among them, X i represents the i-th cabinet (1≤i≤n), C j represents the jth cluster center, i.e. the jth air conditioner (1≤j≤k), X it represents the t-th attribute of the i-th cabinet (1≤t≤m), C jt represents the tth attribute of the jth cluster center. In the embodiment of the present invention, the position attribute is used in formula (1).

[0035] (3) Group division

[0036] After defining the cluster center of the cabinet group using the air conditioner location information, each cabinet is assigned to the air conditioner that is closest to it. According to formulas (2) and (3), the distances from each cabinet to each air conditioner are compared in turn, and the cabinet X i Assign to the closest air conditioner C r The corresponding cluster group R r {X1,X2,…,X i}.

[0037] L r =min(L(X i ,C)) (2)

[0038] R r ={X1,X2,…,X i} (3)

[0039] Among them, L r Indicates cabinet X i With air conditioning C r The minimum distance between the cabinets X i Clustered to air conditioner C r Cabinet group R r (1≤r≤k), and then get the air conditioning group {C1,C2,…,C k} corresponding to the cabinet group {R1, R2, …, R k}.

[0040] In one embodiment, according to the groups divided by the deployment location, it may happen that the distance between the cabinet and multiple air conditioners is the same. For example, if the distance between the cabinet and two air conditioners is the same, then the cabinet can be used as an intermediate cabinet and the intermediate cabinet can be assigned to both groups at the same time. After the clustering algorithm is completed, the total cabinet load in each cabinet group is calculated according to formula (4), and the total cabinet load in the cabinet group is used as the condition for updating the cabinet clustering group. By calculating the total cabinet load of the two groups after distance clustering, the intermediate cabinet is assigned to the group with the smaller total load, and the group {R1′, R2′, …, R k ′}.

[0041]

[0042] Among them, P j Represents group R j Total cabinet load power, X rp Indicates the load power of the rth cabinet in the group.

[0043] For step S102, it is necessary to determine the group target of each cabinet group, as follows:

[0044] (1) Group Target

[0045] The cabinet group is taken as the air conditioning and cooling control object. According to the cabinet group clustering results, the cabinet group R is classified according to formula (5). j The average temperature of the temperature sensor at the cabinet air inlet (cold aisle temperature) and the average temperature of the temperature sensor at the air outlet (hot aisle temperature) are used to determine the air conditioner C j Area Control TargetS j,R , so that it can be maintained within a preset adaptive range.

[0046] S j,R =f(T inlet ,T outlet ) (5)

[0047] Among them, T inlet and T outlet The values ​​may be the measured values ​​of the temperature sensors deployed at the air inlet and air outlet of the cabinet, or the measured temperature at the air supply point and the measured temperature at the return air point of the air conditioner.

[0048] (2) Temperature boundaries and system goals

[0049] According to one embodiment of the present invention, T inlet and T outlet The temperature must be controlled within the preset temperature boundary, that is, T inlet Controlled within the preset first temperature range, T outlet Control within the preset second temperature range. According to the requirements of the Data Center Design Specification, the first temperature range can be 18℃~27℃, and the second temperature range can be 4℃~40℃. The temperature boundary can be adjusted according to the actual needs of the user.

[0050] For step S103, in each cabinet group, the air conditioner corresponding to the cabinet group needs the group target to perform regional control. Regional control refers to the perception of the environmental status of the cabinet group in the data center and the change of IT load, and the use of temperature field adaptive control and temperature field predictive control to establish a temperature control model, and adjust the air conditioning refrigeration temperature with the goal of optimizing energy efficiency. The following will introduce the control method of temperature field adaptive control, temperature field predictive control and the fusion of the two.

[0051] (1) Temperature field adaptive control

[0052] The air conditioner performs temperature field adaptive control tasks at the minute level, uses the data center temperature boundary as the temperature setting target for energy efficiency optimization, and always keeps close to the temperature boundary of the cabinet group, meeting the temperature control needs of the cabinet group with less air conditioning cooling.

[0053] S at =f(T inlet ,T outlet ,T cool ,T hot ) (6)

[0054] As shown in formula (6), T cool and T hot are the energy-saving target temperatures at the cabinet air inlet and outlet under adaptive control, and the output is the cooling temperature set value S of the air conditioner under adaptive control. at , S at In supply air control mode, it is the supply air temperature set point, and in return air control mode, it is the return air temperature set point.

[0055] The following introduces the temperature field adaptive control method under the return air control and supply air control modes (hereinafter referred to as the adaptive control method).

[0056] Return air temperature control refers to the use of the temperature value collected by the temperature sensor on the return air side of the unit to participate in the control, compare the return air temperature of the air conditioner with the set return air set temperature, calculate the cooling demand to control the unit's capacity output and the on-demand actions of other components to adjust T inlet and T outlet , so that both meet the set temperature boundary conditions. The cooling demand is related to the return air temperature, return air set temperature, temperature dead zone, and temperature proportional band of the air conditioner, namely:

[0057] Cooling demand = f(return air temperature, return air set temperature, temperature dead zone, temperature proportional band) (7)

[0058] From the above formula, we can know that the cooling output of precision air conditioner can be adjusted by controlling the return air set temperature. Therefore, by dynamically adjusting the return air set temperature, the cooling capacity can be dynamically controlled. The principle of adaptive dynamic adjustment of return air set temperature algorithm is as follows:

[0059] T return,set (t+1)=T′ return,set (t)+k p (T cool_max -T inlet (t)) (8)

[0060] T return,set (t+1)=T′ return,set (t)+k p (T hot_max -T outlet (t)) (9)

[0061] Among them, T cool_max T is the temperature boundary condition of the data center. inlet The upper limit of the range, T hot_max The temperature boundary condition of the data center is T outlet The upper limit of the range, T outlet (t) is the cabinet outlet temperature at time t, T inlet (t) is the cabinet air inlet temperature at time t, T′ return,set (t) is the return air set temperature at time t, T return,set (t+1) is the calculated return air set temperature at time t+1; k p It is the preset temperature proportional band.

[0062] With temperature boundary conditions: load power P dev >0, the first temperature range is 18℃ <T inlet <27℃ and the second temperature range is 5℃ <T outlet <40℃, load power P dev =0, the first temperature range and the second temperature range are 5°C <T inlet =T outlet <40℃, the adjustment range of the return air setting temperature of the air conditioner

[0063] 16℃ <T return,set Taking <40℃ as an example, the adaptive control method in return air control mode is explained.

[0064] The sensors collect the data center operation data at the minute level, and then adjust the air conditioning return air set temperature every minute according to the cooling demand under different operating conditions, as follows:

[0065] 1) Collect the working condition information at the current time t, including: load power P at time tdev (t), T outlet (t), T inlet (t) and T′ return,set (t).

[0066] Take the average of the multiple cabinet outlet temperatures between time t-1 and time t as T outlet (t); take the average of the air inlet temperatures of multiple cabinets between time t-1 and time t as T inlet (t). In addition, since T′ return,set The value of (t) may be calculated or directly set, so when calculating the return air set temperature at time t+1, it is necessary to read the return air set temperature running at time t from the database or from the communication module of the air conditioner, instead of directly using the return air set temperature calculated at time t.

[0067] 2) In P dev (t)>0, that is, in the server running scenario:

[0068] If T inlet (t)<27℃, then calculate the return air set temperature T at time t+1 according to the following method return,set (t+1):

[0069] T return,set (t+1)=T′ return,set (t)+k p (40℃-T outlet (t))

[0070] According to the above temperature boundary conditions, T hot_max =40℃ is used as the high temperature boundary of the cabinet air outlet, and the return air set temperature is dynamically adjusted to control the cooling demand to meet T outlet <40℃ environmental requirements;

[0071] Otherwise, the return air set temperature at time t+1 is T return,set (t+1) is calculated as:

[0072] T return,set (t+1)=T′ return,set (t)+k p (27℃-T inlet (t))

[0073] According to the above temperature boundary conditions, T cool_max = 27℃ as the cooling optimization target of the cabinet air inlet temperature, and dynamically adjust it by controlling the return air set temperature to meet T inlet Environmental requirement: <27℃.

[0074] In P dev(t) = 0, that is, in the server shutdown scenario, the return air set temperature T at time t+1 return,set (t+1) is calculated as:

[0075] T return,set (t+1)=T′ return,set (t)+k p (40℃-T outlet (t))

[0076] According to the above temperature boundary conditions, T hot_max =40℃ is used as the high temperature boundary of the data center, and the return air set temperature is dynamically adjusted to meet T outlet <40℃ environment requirements, try to increase the return air set temperature to reduce cooling energy consumption. And when the load is shut down, the air conditioner has no cooling output.

[0077] 3) After calculating the return air set temperature T at time t+1 return,set After (t+1), it is also necessary to determine whether the calculated value is within the adjustment range of the return air set temperature of 16°C. <T return,set <40℃:

[0078] If T return,set (t+1)>40℃, then let T return,set (t+1) = 40°C;

[0079] If T return,set (t+1)<16℃, then let T return,set (t+1) = 16°C;

[0080] That is, the calculation result exceeds the return air setting temperature range, and the maximum or minimum return air setting value is output as the result.

[0081] 4) Based on the return air temperature of the air conditioner at time t and the return air setting temperature T at time t+1 obtained by the above calculation return,set (t+1), temperature dead zone, temperature proportional band k p The cooling demand at time t+1 is calculated according to formula (7), and then the cooling operation is performed according to the calculated cooling demand to adjust T inlet and T outletMake both meet the requirements of temperature boundary conditions. Among them, the return air temperature of the air conditioner is collected by the air conditioner's built-in sensor, indicating the mixed air temperature of the return air channel (i.e., the hot channel). The temperature dead zone represents the error between the real-time value of the allowable return air temperature and the return air set temperature, that is, the error between the two is within the temperature dead zone, which can be regarded as having reached a regulation balance; the temperature dead zone can usually be set to -0.5°C ~ 0.5°C. The temperature dead zone and the temperature proportional band can be fixed values ​​or values ​​set at time t. The above calculation of the cooling demand based on the return air set temperature using formula (7) is a well-known technology in the field, and the present invention will not be repeated here.

[0082] 5) Return to step 1) and continue to dynamically adjust the return air set temperature at the next moment.

[0083] Supply air temperature control refers to the use of the temperature value collected by the temperature sensor on the supply air side of the unit to participate in the control, compare the real-time value of the supply air side temperature value with the set supply air set temperature, and control the unit's capacity output and the on-demand actions of other components through the calculated cooling demand to adjust T inlet and T outlet , so that both meet the temperature boundary conditions. The cooling demand is related to the air supply temperature, air supply set temperature, temperature dead zone, and temperature proportional band of the air conditioner, namely:

[0084] Cooling demand = f(supply air temperature, supply air set temperature, temperature dead zone, temperature proportional band) (10)

[0085] From the above formula, we can know that the cooling output of precision air conditioner can be adjusted by controlling the supply air set temperature. Therefore, the dynamic adjustment of the supply air set temperature can also dynamically control the cooling capacity. The principle of the algorithm for adaptive dynamic adjustment of the supply air set temperature is as follows:

[0086] T supply,set (t+1)=T′ supply,set (t)+k p (T cool_max -T inlet (t)) (11)

[0087] T supply,set (t+1)=T′ supply,set (t)+k p (T hot_max -T outlet (t)) (12)

[0088] Among them, T′ supply,set (t) is the supply air set temperature at time t, T supply,set (t+1) is the calculated supply air set temperature at time t+1.

[0089] Still based on the temperature boundary condition: load power Pdev >0, the first temperature range is 18℃ <T inlet <27℃ and the second temperature range is 5℃ <T outlet <40℃, load power P dev =0, the first temperature range and the second temperature range are 5°C <T inlet =T outlet <40℃, the adjustment range of air supply setting temperature is 13℃ <T supply,set Taking <25℃ as an example, the adaptive control method under air supply control is explained.

[0090] The sensors collect data on the data center operation at the minute level, and then adjust the air supply set temperature of the air conditioner every minute according to the cooling demand under different operating conditions, as follows:

[0091] 6) Collect the working condition information at the current time t, including: load power P at time t dev (t), T outlet (t), T inlet (t) and T′ supply,set (t).

[0092] Take the average of the multiple cabinet outlet temperatures between time t-1 and time t as T outlet (t); take the average of the air inlet temperatures of multiple cabinets between time t-1 and time t as T inlet (t). In addition, since T′ supply,set (t) may be calculated or directly set, so when calculating the supply air set temperature at time t+1, it is necessary to read the supply air set temperature running at time t from the database or from the communication module of the air conditioner, instead of directly using the supply air set temperature calculated at time t.

[0093] 7) In P dev (t)>0, that is, in the server running scenario:

[0094] If T outlet (t)<40℃, the air supply set temperature T at time t+1 is calculated as follows supply,set (t+1):

[0095] T supply,set (t+1)=T′ supply,set (t)+k p (27℃-T inlet (t))

[0096] According to the above temperature boundary conditions, T cool_max = 27℃ as the cooling optimization target of the cabinet air inlet temperature, and dynamically adjust it by controlling the air supply set temperature to meet Tinlet Environmental requirements of <27℃;

[0097] Otherwise, the air supply setting temperature at time t+1 is T supply,set (t+1) is calculated as:

[0098] T supply,set (t+1)=T′ supply,set (t)+k p (40℃-T outlet (t))

[0099] According to the above temperature boundary conditions, T hot_max =40℃ as the high temperature boundary of the cabinet air outlet, dynamically adjust the air supply set temperature to meet the cooling demand to meet T outlet <40℃ environmental requirement.

[0100] In P dev (t) = 0, that is, in the server downtime scenario:

[0101] If T outlet (t)>40℃, the air supply setting temperature T at time t+1 is calculated as follows supply,set (t+1):

[0102] T supply,set (t+1)=T′ supply,set (t)+k p (40℃-T outlet (t))

[0103] According to the above temperature boundary conditions, T hot_max =40℃ as the high temperature boundary of the cabinet air outlet, and dynamically adjust the air supply set temperature to meet T outlet <40℃ environment requirements, try to increase the air supply set temperature to reduce cooling energy consumption. And when the load is shut down, the air conditioner has no cooling output.

[0104] 8) Calculate the air supply set temperature T at time t+1 supply,set After (t+1), it is also necessary to determine whether the calculated value is within the adjustment range of the air supply set temperature of 13°C. <T supply,set <25℃:

[0105] If T supply,set (t+1)>25℃, then let T supply,set (t+1) = 25°C;

[0106] If T supply,set (t+1)<13℃, then let T supply,set (t+1) = 13°C;

[0107] That is, the calculation result exceeds the set air supply temperature range and is output with the maximum or minimum air supply set value as the result.

[0108] 9) Based on the air supply temperature of the air conditioner at time t, the calculated air supply setting temperature T at time t+1 supply,set (t+1), temperature dead zone, temperature proportional band k p The cooling demand at time t+1 is calculated according to formula (10), and then the cooling operation is performed according to the calculated cooling demand to adjust T inlet and T outlet Make both meet the requirements of temperature boundary conditions. Among them, the air supply temperature of the air conditioner is collected by the air conditioner's own sensor, indicating the cooling temperature of the air supply channel (i.e., the cold channel). The temperature dead zone represents the error between the real-time value of the allowable supply air temperature and the supply air set temperature, that is, the error between the two is within the temperature dead zone, which can be regarded as having reached the balance of regulation; the temperature dead zone can usually be set to -0.5℃~0.5℃. The temperature dead zone and the temperature proportional band can be fixed values ​​or values ​​set at time t. The above calculation of the cooling demand based on the supply air set temperature using formula (10) is a well-known technology in the field, and the present invention will not be repeated here.

[0109] 10) Return to step 6) and continue to dynamically adjust the air supply set temperature at the next moment.

[0110] From the above, it can be seen that the adaptive temperature control method can enable the air conditioner to execute adaptive control targets at the minute level. Under the constraint of meeting the temperature boundary conditions of the data center, the air conditioner cooling temperature can be adjusted autonomously according to different operating environments, so that the cooling capacity and air volume output are adjusted according to the actual power of the load, and a higher set temperature and less air conditioning cooling are always maintained to meet the load operating environment, which greatly improves the matching degree with the actual load and reduces the energy usage efficiency (PUE) of the computer room, thereby achieving the purpose of reducing energy consumption.

[0111] (2) Temperature field prediction control

[0112] Temperature field predictive control is based on the temperature field adaptive control method. It further analyzes energy-saving data, analyzes air-conditioning setting parameters according to the cabinet operating environment and IT load changes, builds a temperature field self-learning neural network model, and quickly infers and sets reasonable temperatures for the current and future periods of time, so as to synchronize the demand for load heat dissipation and air-conditioning cooling, thereby achieving the effect of saving refrigeration energy consumption.

[0113] S pt =α1T Env +α2P dev +α3T outlet +α4T inlet +b (13)

[0114] Among them, α1 is the influence coefficient of the cabinet ambient temperature, α2 is the influence coefficient of the load power, α3 is the influence coefficient of the cabinet outlet temperature, α4 is the influence coefficient of the cabinet inlet temperature, and b is a bias term, which covers other influencing factors. According to one embodiment of the present invention, a trained three-layer fully connected neural network model is used to calculate the cabinet ambient temperature T Env 、Load power P dev , Cabinet air outlet temperature T outlet , Cabinet air inlet temperature T inlet , output the cooling temperature setting value S of the air conditioner under the temperature field prediction control pt , S pt In the air supply control mode, it is the supply air temperature setting value, and in the return air control mode, it is the return air temperature setting value. Specifically, obtain the current load power P of the cabinet group dev , Cabinet ambient temperature T env 、T inlet and T outlet , using the above prediction model, according to the current load power, cabinet ambient temperature, T inlet and T outlet , output the air conditioning cooling temperature setting value at the next moment, that is, the return air temperature setting value (under return air control) or the supply air temperature setting value (under supply air control); based on the air conditioning cooling temperature at the next moment, calculate the cooling demand according to formula (7) or formula (10). The training data of the prediction model includes: the load power P collected at the historical moment at the preset time interval and preset time window dev , Cabinet ambient temperature T Env , Cabinet air inlet temperature T inlet , Cabinet air outlet temperature T outlet and time series data of the measured values ​​of the air conditioning cooling temperature, which is generated during the historical temperature control process of the data center. Compared with the adaptive control method, when the load power fluctuation exceeds the preset value, the temperature control method of the embodiment of the present invention can adjust the air conditioning cooling temperature based on historical experience to quickly adapt to the current load power changes, find a reasonable air conditioning cooling temperature, and make the cooling capacity close to the heat generated by the load. Since the heat and cold balance is completed faster than the adaptive control method, the energy consumption will also be lower, achieving energy saving effect.

[0115] (3) Model Fusion

[0116] According to an embodiment of the present invention, during a preset second set time period (i.e., the second set time period, which may be 10 minutes or 20 minutes or may be set by the user according to actual conditions) at the beginning of IT load startup, the cooling temperature setting value of the air conditioner is output based on the trained prediction control model. During the period of outputting the cooling temperature setting value of the air conditioner using the prediction control model, if it is detected that the air outlet temperature of the cabinet exceeds the preset first temperature range or the air inlet temperature of the cabinet exceeds the preset second temperature range, the cooling temperature setting value of the air conditioner is calculated using an adaptive control method.

[0117] In some embodiments, as shown in formula (14), when a large load rate change is detected (for example, the minute fluctuation rate exceeds 10%), the air conditioner executes the temperature field prediction control target, sets a reasonable temperature suitable for the current and future period of time, and continuously detects whether the temperature field is out of bounds (that is, detects whether the air outlet temperature of the cabinet exceeds the preset first temperature range, or whether the air inlet temperature of the cabinet exceeds the preset second temperature range). If it is out of bounds, the temperature field adaptive control target is executed; otherwise, the temperature field prediction control target continues to be executed during this time period.

[0118]

[0119] Among them, S is the cooling temperature setting value of the air conditioner, lr is the load rate change before and after two minutes, and V is the fluctuation rate.

[0120] For step S104, since the cooling load of the micro-module data center is constantly changing, while the cabinet group air conditioner performs regional control, a multi-machine parallel cascade group control method can be used to dynamically adjust the operating status of the air conditioner in the data center and coordinately control the cooling capacity distribution of the air conditioner equipment, that is, find neighboring air conditioners to perform coordinated control so that the air conditioner equipment operates in a high-efficiency range. Figure 2 As shown, for any cabinet group, step S201 can determine whether the air conditioner needs cooling coordination according to the target control completion status of the air conditioner corresponding to the cabinet group and / or the operation status of the air conditioner during the execution of the zone control. Step S202 If necessary, a cooling coordination request is sent to the air conditioners corresponding to other cabinet groups. Step S203 The air conditioner that receives the cooling coordination request performs cooling coordination operation.

[0121] The following will be combined Figure 3 The specific process of collaborative control is introduced. Figure 3 As shown, the situations where cooling coordination is needed may include the need for coordinated cooling and coordinated optimization. Regardless of the above situations, after determining that cooling coordination is needed, it is necessary to determine the demand allocation within the air conditioning group of the data center and adjust the setting parameters of the relevant air conditioning.

[0122] (1) Collaborative cooling

[0123] According to an embodiment of the present invention, judging whether it belongs to collaborative cooling includes: within a preset first set time period, if the air conditioner corresponding to the cabinet group does not control the target temperature in the group target within the set temperature boundary condition (for example, the air outlet temperature of the cabinet exceeds the preset first temperature range, or the air inlet temperature of the cabinet exceeds the preset second temperature range), then judging that the air conditioner needs cooling collaboration; otherwise, judging that the air conditioner does not need cooling collaboration. The first set time period can be 10 minutes, 20 minutes, 30 minutes or set according to actual needs.

[0124] (2) Collaborative Optimization

[0125] According to one embodiment of the present invention, the air conditioning refrigeration demand D can be calculated according to formula (7) or formula (10), and the operating frequency of the variable frequency compressor can be adjusted to make the refrigeration output meet the current refrigeration demand. Determining whether collaborative optimization is needed includes: determining the demand value of the air conditioning compressor operating frequency according to the refrigeration demand, if the demand value of the air conditioning compressor operating frequency is within the preset operating range, the air conditioning does not need refrigeration collaboration; if the demand value of the air conditioning compressor operating frequency exceeds the maximum value of the preset operating range or is lower than the minimum value of the operating range, the air conditioning needs refrigeration collaboration. In some embodiments, according to the operating characteristics of the precision air conditioning variable frequency compressor, if the frequency operates in the high-efficiency range, the air conditioning can be efficiently cooled. On the contrary, if the frequency exceeds the maximum output, the air conditioner will not be able to meet the cooling demand of the computer room, and a collaborative demand for insufficient refrigeration can be issued; if the frequency is lower than the high-efficiency range or higher than the high-efficiency range, the air conditioner issues a collaborative demand for improved refrigeration efficiency.

[0126] f=γD (15)

[0127]

[0128] Where f is the compressor operating frequency. γ is the conversion factor between cooling demand and operating frequency, which is usually an empirical value or the performance value of the air conditioning equipment.

[0129] According to the variable frequency energy efficiency adjustment logic, if the compressor operates within the preset high efficiency operating range (i.e. the energy efficiency lower limit fe min and energy efficiency ceiling max The cooling energy efficiency is relatively better, and the cooling demand that meets the current high-efficiency operation can continue to operate. cd ; If the frequency exceeds the upper limit of energy efficiency or is lower than the lower limit of energy efficiency, cooling coordination is required.

[0130] (3) Cooling synergy

[0131] According to one embodiment of the present invention, cooling coordination includes sending a cooling coordination request D to a neighboring air conditioner. cr, the cooling collaborative request includes collaborative control of the set temperature. The air conditioner that receives the cooling collaborative request performs a cooling collaborative operation based on the collaborative control of the set temperature. Specifically, it includes neighbor calculation and demand allocation:

[0132] 1) Nearest Neighbor Calculation

[0133] The goal of air conditioner nearest neighbor calculation is to select the cooling collaborative control object based on the minimum Euclidean distance between k air conditioners as the neighbor value. Given an air conditioner dataset C, which contains air conditioners {C1, C2, …, C k}, the air conditioner nearest neighbor algorithm first needs to calculate each air conditioner C j The Euclidean distance with the air conditioner group C′ is used to obtain the nearest air conditioner according to the minimum distance principle.

[0134]

[0135]

[0136] N j,n =min(N(C j ,C′)) (19)

[0137] As shown in formulas (17) to (19), C j represents the jth air conditioner (1≤j≤k), C′ is C excluding C j Air conditioning group, N(C j ,C l ) indicates air conditioner C j and air conditioner C in air conditioner group C′ l The Euclidean distance, N j,n Indicates air conditioner C j and air conditioner C in air conditioner group C′ n The minimum Euclidean distance between them. Therefore, C n Air Conditioner C j Neighbors air conditioning.

[0138] 2) Demand allocation

[0139] Given the cabinet group target of the air conditioner itself and the cabinet group target of the neighboring air conditioners, the coordination request of the neighboring air conditioners is converted into its own coordination target, and the control target of the air conditioner is adjusted as needed to control the cooling output.

[0140] S j,C =max(S j,R ,S n,R ) (20)

[0141]

[0142] As shown in formulas (20) and (21), S j,RAir Conditioner C j The corresponding cabinet group target, S n,R It is the neighbor air conditioner C n The cabinet group target, S j,C Represents air conditioner C j The collaborative goal of D n,cr For neighbor air conditioner C n The cooling coordination request issued is output as air conditioner C j The collaborative control setting value S j,set . Set the collaborative control temperature (i.e. the neighboring air conditioner C n Cabinet group target S n,R ) and the target temperature of the air conditioner that receives the cooling cooperation request (air conditioner C j The corresponding cabinet group target S j,R ), the temperature with the higher temperature value is used as the new target temperature S of the air conditioner that receives the cooling cooperation request. j,C The neighboring air conditioner receives the cooling coordination request (i.e. D n,cr =1), the neighboring air conditioner performs cooling cooperative operation according to the new target temperature. The neighboring air conditioner does not receive the cooling cooperative request (i.e., D n,cr =0), the neighboring air conditioner performs cooling operation according to its own group target.

[0143] Based on the above group control method, an embodiment of the present invention further provides a micro-module data center air conditioning group control system, including: a temperature acquisition module, a power acquisition module and a temperature control module. The temperature acquisition module is used to collect the air outlet temperature of the cabinet, the air inlet temperature of the cabinet, the measured value of the cooling temperature of the air conditioner, and the cabinet ambient temperature. The power acquisition module is used to collect load power. The temperature control module is used to execute any of the air conditioning group control methods in the above embodiments.

[0144] Next, we can verify the energy-saving effect of collaborative control energy efficiency optimization by analyzing the micro-module cooling design and airflow organization optimization and deploying a standard single-channel micro-module data center. The single-channel micro-module data center contains 7 cabinets RACK_1~RACK_7, 2 air-cooled rack 7.5KW air conditioners, 28 0.5KW simulation servers, 9 temperature sensors and other equipment. Figure 4As shown, with the rack position of each air conditioner as the center, the distance L between each rack position and each air conditioner is counted. According to the principle of minimum distance L, the air conditioners are divided into cooling groups. If the air conditioners are at equal distances, they are allocated to the air conditioner with the smaller total cabinet load. According to the group control grouping strategy, the regional control target of the cabinet group corresponding to the air conditioner is calculated. C1 is mainly responsible for the cabinet loads RACK_1, RACK_2, and RACK_3 in the cooling-related area. The average temperature of sensors T_inlet_1, T_inlet_2, and T_inlet_3 is used as the air inlet temperature T of the cabinet group. inlet , the temperature of sensor T_outlet_1 is used as the outlet temperature T of the cabinet group outlet , calculate the target S of regional control 1,R In a data center with closed hot and cold aisles, the T inlet 、T outlet The average values ​​can be used as the cold channel temperature and hot channel temperature of the micromodule.

[0145] Adjust the load power of the simulated server in the micro-module data center to make the cabinet cooling load change continuously. The simulated servers in cabinets RACK_1 to RACK_7 adopt load start, load balancing, load unevenness, load fluctuation and other operation modes respectively, set the system goals of cold channel target 27℃ and hot channel target 40℃, and use the cascade group control method of multiple machines in parallel. The collaborative control of the air conditioning group control energy efficiency optimization algorithm is adopted to collaboratively control the cooling capacity distribution of the air conditioning equipment, and record the cold and hot channel temperature, outdoor temperature, server load power and system PUE value of the micro-module during the air conditioning group control optimization period. Key data of the micro-module during the collaborative control optimization operation are as follows: Figure 5 shown.

[0146] The micro-module data center air conditioning group control system monitors the changes in cabinet group temperature and IT load. At the beginning of the load startup, the temperature field prediction control model is used to set a reasonable temperature suitable for the current and future period of time, so that the air conditioning group can output cooling in advance, achieve synchronous adjustment of heat dissipation and cooling, and solve the problem of delayed cooling. In the load fluctuation stage, the temperature field adaptive control model is used to always keep close to the temperature boundary of the cabinet group to meet the system goals of 27°C for the cold channel and 40°C for the hot channel. Under the load-balanced and uneven load operating conditions, the collaborative control model uses a cascade group control method of parallel air conditioners to adjust the cooling distribution, so that the air conditioning refrigeration is stably supplied and the system PUE is maintained in the high-efficiency range.

[0147] In order to further verify the energy-saving effect of the air conditioning group control energy efficiency optimization method, grid tests with multiple temperature conditions and multiple load rates were carried out. Temperature sensors were deployed on the air conditioner outdoor unit to monitor the outdoor temperature conditions. Under the conditions of 35℃, 25℃, 15℃, and 5℃, servers with load rates of 25%, 50%, and 75% were turned on under load balancing and uneven load, and the system PUE of 12 sets of data was recorded. If the conventional method is used to set a fixed supply air temperature experience value of 22℃, or the return air temperature experience value of 32℃ as a comparison, under similar operating conditions, the coordinated control of the air conditioning group control method has achieved different degrees of energy efficiency optimization. The system PUE is converted into air conditioning energy saving effect, and the energy efficiency optimization test data is shown in Table 1.

[0148] Table 1 Energy saving effect of coordinated control of air conditioner group control

[0149]

[0150] The collaborative control method can dynamically calculate the cooling demand and control the cooling capacity distribution of the air conditioner according to the changes in the load power of the cabinet group. For application scenarios with large fluctuations in load operating power, the energy efficiency optimization effect is better. Taking the test case of a load rate of 25% as an example, the collaborative control method does not turn on the second air conditioner while meeting the cooling efficiency of a single air conditioner, which saves 21% energy overall; at a load rate of 50%, from the overall test of load balancing and uneven load, an 18% energy saving effect was achieved; at an overall load rate of 75%, both air conditioners turned on a larger cooling output, and the collaborative control method achieved an energy saving effect of about 7% in a limited space where the target did not change much.

[0151] The micro-module data center air conditioning group control system of the present invention solves the problems of uneven heating and cooling of independent air conditioning in the computer room, large fluctuations in ambient temperature, short circuit of air flow, frequent start and stop of equipment, and idle power consumption of on-duty air conditioning through the overall control of all air conditioning in the computer room. The system collects data of each air conditioner through the communication port, and adds an external network temperature and humidity sensor to each air conditioning area node to collect the overall temperature field data of the computer room. The group control system performs comprehensive calculation and analysis on the air conditioning equipment parameters and temperature field distribution of the entire computer room according to the collected data, and overall dispatches the operation of all air conditioners in the computer room, balances and controls the temperature field of the computer room, and improves the overall operation efficiency, thereby achieving energy saving. The cabinet groups of the present invention are clustered and allocated to the corresponding air conditioners, and then the energy efficiency optimization algorithm of the air conditioning group control of the collaborative control is used to calculate the cooling demand of each air conditioner for the cabinet group in the current control cycle, dynamically adjust the number of air conditioners in parallel, and collaboratively control the cooling capacity distribution of the air conditioning equipment, so as to realize the on-demand distribution of cooling capacity supply, stable supply of air conditioning refrigeration, and efficient and energy-saving system operation, which optimizes the operating environment of the micro-module as a whole and makes product maintenance more convenient.

[0152] The embodiments of the present invention include sequential exemplary steps, but these steps do not have to be performed in the order shown. Performing these steps in different orders is within the scope of the present disclosure. Within the spirit and scope of the embodiments of the present invention, these steps can be added, replaced, changed in order and / or omitted as appropriate.

[0153] It should be noted that although the above describes the various steps in a specific order, it does not mean that the various steps must be executed in the above specific order. In fact, some of these steps can be executed concurrently or even in a different order as long as the required functions can be achieved.

[0154] The present invention may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present invention.

[0155] Computer readable storage medium can be a tangible device that holds and stores instructions used by an instruction execution device. Computer readable storage medium can include, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a protruding structure in a groove on which instructions are stored, and any suitable combination thereof.

[0156] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for controlling air conditioners in a micro-module data center, wherein the micro-module data center includes a plurality of air conditioners and a plurality of cabinets; the method includes: Dividing the plurality of cabinets into a plurality of cabinet groups, wherein each cabinet group includes at least one cabinet, and each air conditioner corresponds to a cabinet group; Determine the group target of each cabinet group based on the average air inlet temperature or the average air outlet temperature of each cabinet group; The air conditioner corresponding to each cabinet group performs zone control based on the group target; For each cabinet group, determine whether the air conditioner corresponding to the cabinet group needs cooling coordination based on the completion of the area control target of the air conditioner and / or the operating status of the air conditioner; if necessary, send a cooling coordination request to the air conditioners corresponding to other cabinet groups, and the air conditioner that receives the cooling coordination request performs cooling coordination operation.

2. The method according to claim 1, characterized in that According to the completion status of the regional control target of the air conditioner corresponding to the cabinet group and / or the operation status of the air conditioner, judging whether the air conditioner needs cooling coordination includes: During the preset first set time period, If the air conditioner corresponding to the cabinet group fails to control the target temperature in the group target within the set temperature boundary condition, it is determined that the air conditioner needs cooling coordination; Otherwise, it is determined that the air conditioner does not require cooling coordination.

3. The method according to claim 1, characterized in that According to the completion status of the regional control target of the air conditioner corresponding to the cabinet group and / or the operation status of the air conditioner, judging whether the air conditioner needs cooling coordination includes: Determining a refrigeration demand of the air conditioner, and determining a required value of an operating frequency of a compressor of the air conditioner according to the refrigeration demand; If the required value of the compressor operating frequency of the air conditioner is within the preset operating range, the air conditioner does not need cooling coordination; If the required value of the compressor operating frequency of the air conditioner exceeds the maximum value of the preset operating range, the air conditioner requires cooling coordination.

4. The method according to claim 3, characterized in that: Determining a cooling demand of the air conditioner, including: The cooling demand of the air conditioner is determined according to the cooling temperature measurement value of the air conditioner corresponding to the cabinet group, the cooling temperature setting value of the air conditioner, the temperature control dead zone of the air conditioner and the temperature proportional band.

5. The method according to claim 4, characterized in that Also includes: In the return air control mode, the measured value of the cooling temperature of the air conditioner is the measured value of the return air temperature, and the set value of the cooling temperature of the air conditioner is the set value of the return air temperature; In the air supply control mode, the measured value of the cooling temperature of the air conditioner is the measured value of the supply air temperature, and the set value of the cooling temperature of the air conditioner is the set value of the supply air temperature.

6. The method according to claim 1, characterized in that The sending of a cooling coordination request to air conditioners corresponding to other cabinet groups includes: Calculating the Euclidean distance between the air conditioner and each air conditioner corresponding to the other cabinet groups; Filter out the minimum value of the Euclidean distance, and use the air conditioner corresponding to the other cabinet group corresponding to the minimum Euclidean distance as a cooling coordination object; The cooling coordination request is sent to the cooling coordination object.

7. The method according to claim 1, characterized in that The air conditioner receiving the cooling cooperative request performs a cooling cooperative operation, including: The cooling coordination request includes coordinated control of a set temperature; The cooperative control setting temperature and the target temperature of the air conditioner that has received the cooling cooperative request have a higher temperature value as a new target temperature of the air conditioner that has received the cooling cooperative request; The air conditioner receiving the cooling cooperative request performs a cooling cooperative operation according to the new target temperature.

8. The method according to claim 1, characterized in that According to the positions of the air conditioners and the cabinets, the cabinets are divided into a plurality of cabinet groups, including: All air conditioners are regarded as the center of the cabinet group, and the number of air conditioners is used as the k value of the clustering algorithm; Calculate the Euclidean distance from each cabinet to the center of each cabinet group according to the position of the air conditioner and the position of the cabinet; A clustering algorithm is used to allocate each cabinet to the cabinet group center closest to it according to the Euclidean distance from each cabinet to the center of each cabinet group, so as to form a cabinet group.

9. The method according to claim 7, characterized in that: Using a clustering algorithm to allocate each cabinet to the cabinet group center closest to it according to the Euclidean distance from each cabinet to the center of each cabinet group, further comprising: If the Euclidean distances between the cabinet and the centers of multiple cabinet groups are the same, then Using the cabinet as an intermediate cabinet, and simultaneously allocating the intermediate cabinet to the centers of the plurality of cabinet groups; After the clustering algorithm is completed, the total load of the cabinets in the multiple cabinet groups corresponding to the centers of the multiple cabinet groups is calculated; According to the total load, the intermediate cabinet is allocated to a cabinet group with the smallest total load.

10. The method according to claim 1, characterized in that The micro-module data center also includes multiple loads; The air conditioner corresponding to each cabinet group performs zone control based on the group target, including: In a preset second set time period at the beginning of the load startup, based on the trained predictive control model, output a cooling temperature setting value of the air conditioner according to the power of the load, the ambient temperature of the cabinets in the cabinet group, the air outlet temperature of the cabinet, and the air inlet temperature of the cabinet; During the period of outputting the cooling temperature setting value of the air conditioner using the predictive control model, if it is detected that the air outlet temperature of the cabinet exceeds a preset first temperature range or the air inlet temperature of the cabinet exceeds a preset second temperature range, the cooling temperature setting value of the air conditioner is calculated using an adaptive control method; Among them, the adaptive control method includes calculating the cooling temperature setting value of the air conditioner based on the first temperature range, the second temperature range, the power of the load, the measured value of the cooling temperature of the air conditioner, and the air outlet temperature of the cabinet or the air inlet temperature of the cabinet.

11. The method according to claim 10, characterized in that Also includes: After the second set time period, calculating the fluctuation rate of the load power of the micromodule data center; If the fluctuation rate exceeds a preset fluctuation rate threshold, outputting a cooling temperature setting value of the air conditioner using the predictive control model; If the fluctuation rate does not exceed the preset fluctuation rate threshold, the cooling temperature setting value of the air conditioner is calculated using the adaptive control method.

12. A micro-module data center air conditioning group control system, characterized in that: include: Temperature acquisition module, power acquisition module and temperature control module; The temperature acquisition module is used to collect the cabinet's air outlet temperature, the cabinet's air inlet temperature, the measured value of the air conditioner's cooling temperature, and the cabinet's ambient temperature; The power collection module is used to collect load power; The temperature control module is used to execute the air conditioning group control method as described in any one of claims 1 to 11.

Images