A data center fan and control method thereof

Abstract

The present invention discloses a data center fan and a control method thereof, wherein the data center fan comprises a machine room module, a supply and return air duct, an air conditioning unit, an external interface and an external optimization controller; the machine room module is fixedly connected to the air conditioning unit through the supply and return air duct; the external interface is fixedly connected to the output end of the air conditioning unit; and the external interface is also communicatively connected to the external optimization controller. The present invention can reduce the energy consumption of the fan in the data center, improve energy utilization, and save energy and protect the environment when the temperature field distribution in the data center machine room meets specific requirements. Moreover, the automatic redistribution of loads between units in the event of a fault is realized, thereby improving the reliability of the data center. Combined with a real-time optimization algorithm, the temperature field distribution in the data center machine room is adjusted, while the fan maintains the minimum necessary power consumption.

Description

Technical Field

[0001] The present invention belongs to the field of thermal energy technology, and in particular relates to a data center fan and a control method thereof. Background Art

[0002] With the rapid development of cloud computing and big data technology, data centers are becoming larger and larger, and their power consumption is also increasing. Studies have shown that the energy consumption of data centers has accounted for about 2% of the national electricity consumption, and is on an upward trend. In addition, studies have also shown that in the energy consumption of data centers, in addition to IT equipment, the energy consumption of air conditioning systems accounts for about 70%, and air conditioning fans account for about 50% of the total energy consumption of air conditioning systems. The need to reduce the energy consumption of fans in data center air conditioning systems is very urgent. Summary of the invention

[0003] The purpose of the present invention is to solve the problem of reducing the energy consumption of a fan in an air conditioning system of a data center, and proposes a fan for a data center and a control method thereof.

[0004] The technical solution of the present invention is: a data center fan includes a computer room module, a supply and return air duct, an air conditioning unit, an external interface and an external optimization controller;

[0005] The computer room module is fixedly connected to the air conditioning unit through the supply and return air duct; the external interface is fixedly connected to the output end of the air conditioning unit; and the external interface is also communicatively connected to the external optimization controller.

[0006] Furthermore, the room module includes IT equipment cabinets, cold aisles, and hot aisles;

[0007] The cold channel and the hot channel are alternately fixedly connected in sequence; and the IT equipment cabinets are respectively fixedly arranged between the alternately fixedly connected cold channel and the hot channel.

[0008] Furthermore, temperature sensors and temperature sensor networks are provided on both the cold channel and the hot channel.

[0009] Furthermore, the temperature sensor is an NTC temperature sensor.

[0010] Furthermore, the temperature sensor network is a star topology.

[0011] The beneficial effects of the present invention are: when the temperature field distribution in the data center computer room meets specific requirements, the present invention can also reduce the energy consumption of the data center fan, improve energy utilization, save energy and protect the environment. Moreover, the present invention can realize the automatic redistribution of loads between units when a fault occurs, thereby improving the reliability of the data center.

[0012] Based on the above system, the present invention also proposes a control method for a fan in a data center, comprising the following steps:

[0013] S1: Calculate m different cold and hot channel temperature differences Δthc based on adjacent cold channels and hot channels n ;

[0014] S2: Different hot and cold channel temperature difference Δthc n Set the weight factor C xm ;

[0015] S3: Set the fault factor matrix G of the air conditioner xy ;

[0016] S4: Based on the temperature difference Δthc of m different hot and cold channels n , weight factor C xm and the failure factor matrix G xy Computed variable CCON x ;

[0017] S5: Set the fan control variable to CSET x ;

[0018] S6: Adjust the variable CCON by trial and error x and the control variable CSET x The difference is minimized to complete the control of the data center fan.

[0019] Furthermore, in step S1, the temperature difference Δthc between m different hot and cold channels is n The calculation formula is:

[0020] Δthc n =j -1 *∑CTh j *Thot nj -i -1 *∑CTc i *Tcold ni

[0021] Among them, Thot nj Indicates the temperature of the jth temperature sensor on the hot aisle of the nth row of cabinets, Tcold ni Indicates the temperature of the ith temperature sensor on the cold aisle of the nth row of cabinets, CTh j Indicates the weight factor of the hot channel temperature sensor. The value is greater than or equal to 0. CTc i Indicates the weight factor of the cold channel temperature sensor. The value is greater than or equal to 0.

[0022] The beneficial effect of the above further scheme is: in the present invention, the temperature of the hot channel is set to a certain value to achieve the purpose of controlling the temperature distribution of the cold channel supply air; the temperature of the cold channel can also be set to a certain value to achieve the purpose of controlling the temperature distribution of the hot channel.

[0023] Furthermore, in step S2, the weight factor C xm It is a matrix with x rows and m columns, where x represents the number of air conditioners and m represents the number of temperature differences between hot and cold channels.

[0024] Further, in step S3, if the value of x is the same as the value of y, they are taken as the same fixed value; if the value of x is different from the value of y, the value of y of the air conditioner in the shutdown or fault state is greater than 0, otherwise the value of y is 0, where y≤x.

[0025] Further, in step S4, the variable CCON x The calculation formula is:

[0026] CCON x =∑G xy *C xm *Δthc n

[0027] Among them, Δthc n represents the temperature difference between m different hot and cold channels, C xm represents the weight factor, G xy represents the failure factor matrix.

[0028] The beneficial effects of the present invention are:

[0029] (1) By increasing the weight factor of the temperature abnormality area, the limit is set using the optimization algorithm. When the limit is reached, the weight factor of the temperature abnormality itself is automatically increased, so that the temperature abnormality area can be quickly restored to normal and the fan power consumption is only increased slightly as necessary.

[0030] (2) Combined with real-time optimization algorithms, the temperature field distribution in the data center room can be adjusted while keeping the fan at the lowest necessary power consumption.

[0031] (3) When an air conditioner in a conventional data center fails and stops working, it may cause abnormal distribution of the temperature field, and each unit cannot perceive this abnormality. However, the present invention can perceive it in real time and adjust the load redistribution in time to quickly restore it to normal. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is the structural diagram of the fan in the data center;

[0033] Figure 2 is a flow chart of a control method for a fan in a data center;

[0034] In the figure, 1. computer room module; 2. supply and return air ducts; 3. air conditioning unit; 4. external interface; 5. external optimization controller; 6. IT equipment cabinet; 7. cold channel; 8. hot channel; 9. temperature sensor; 10. temperature sensor network. DETAILED DESCRIPTION

[0035] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

[0036] like Figure 1 As shown, the present invention provides a data center fan, including a machine room module 1, a supply and return air duct 2, an air conditioning unit 3, an external interface 4 and an external optimization controller 5;

[0037] The computer room module 1 is fixedly connected to the air conditioning unit 3 via the air supply and return duct 2; the external interface 4 is fixedly connected to the output end of the air conditioning unit 3; and the external interface 4 is also communicatively connected to the external optimization controller 5.

[0038] In the embodiment of the present invention, Figure 1 As shown, the computer room module 1 includes an IT equipment cabinet 6, a cold aisle 7 and a hot aisle 8;

[0039] The cold channel 7 and the hot channel 8 are alternately fixedly connected in sequence; the IT equipment cabinets 6 are respectively fixedly arranged between the alternately fixedly connected cold channel 7 and the hot channel 8.

[0040] In the embodiment of the present invention, Figure 1 As shown, both the cold channel 7 and the hot channel 8 are provided with temperature sensors 9 and a temperature sensor network 10 .

[0041] In the embodiment of the present invention, Figure 1 As shown, the temperature sensor 9 is an NTC temperature sensor. In the present invention, the temperature sensor is not in a specific form, and can be an analog sensor plus a collector, or a digital sensor.

[0042] In the embodiment of the present invention, Figure 1 As shown, the temperature sensor network 10 is a star topology. In the present invention, the sensor network is not specific, and can be a wireless network or a wired network; it can be a bus topology or a star topology.

[0043] Based on the above system, the present invention also proposes a control method for a fan in a data center, such as Figure 2 As shown, the following steps are included:

[0044] S1: Calculate m different cold and hot channel temperature differences Δthc based on adjacent cold channels and hot channels n ;

[0045] S2: Different hot and cold channel temperature difference Δthc n Set the weight factor C xm ;

[0046] S3: Set the fault factor matrix G of the air conditionerxy ;

[0047] S4: Based on the temperature difference Δthc of m different hot and cold channels n , weight factor C xm and the failure factor matrix G xy Computed variable CCON x ;

[0048] S5: Set the fan control variable to CSET x ;

[0049] S6: Adjust the variable CCON by trial and error x and the control variable CSET x The difference is minimized to complete the control of the data center fan.

[0050] In practical applications, there are many adjustment methods. This patent provides an adjustment method that can be achieved by adjusting the matrix values, that is, the trial and error method. By observing the impact of different matrix values ​​on the fan power consumption and temperature distribution, the values ​​of each matrix are continuously tested.

[0051] Alternatively, an optimization algorithm can be used: the values ​​of these matrices, supply air temperature, return air temperature, and chilled water temperature are used as the input of the multi-layer neural network, the standard deviation of the fan power consumption and supply air temperature are used as the output of the multi-layer neural network, the values ​​of each matrix setting are changed, and the neural network learns the mathematical relationship between the values ​​of each matrix and the standard deviation of the fan power consumption and supply air temperature, and the optimal matrix value is found through the particle swarm algorithm, and the maximum value of the supply and return air temperature difference is used as a constraint condition to minimize the fan power consumption and the standard deviation of the supply air temperature. When the output of the neural network is different, different optimization goals can be achieved.

[0052] The adjustment algorithm can be run in each unit controller, or the optimized matrices can be transmitted to each air conditioning unit controller through an external interface. The present invention also requires that the air conditioning unit sets the outlet air temperature to achieve the purpose of minimizing the fan power by adjusting the air supply temperature through water valve opening control or compressor frequency conversion control, or by optimizing the air supply temperature through manual experience adjustment or real-time calculation of the adjustment algorithm.

[0053] In the embodiment of the present invention, Figure 2 As shown, in step S1, the temperature difference Δthc between m different hot and cold channels is m The calculation formula is:

[0054] Δthc n =j -1 *∑CTh j *Thot nj -i -1 *∑CTc i *Tcold ni

[0055] Among them, Thot nj Indicates the temperature of the jth temperature sensor on the hot aisle of the nth row of cabinets, Tcold ni Indicates the temperature of the ith temperature sensor on the cold aisle of the nth row of cabinets, CTh j Indicates the weight factor of the hot channel temperature sensor. The value is greater than or equal to 0. CTc i Indicates the weight factor of the cold channel temperature sensor, and its value is greater than or equal to 0. In the present invention, the hot channel temperature is set to a certain value to achieve the purpose of controlling the cold channel supply air temperature distribution; the cold channel temperature can also be set to a certain value to achieve the purpose of controlling the hot channel temperature distribution.

[0056] In the embodiment of the present invention, Figure 2 As shown, in step S2, the weight factor C xm It is a matrix with x rows and m columns, where x represents the number of air conditioners and m represents the number of temperature differences between hot and cold channels.

[0057] In the embodiment of the present invention, Figure 2 As shown, in step S3, if the value of x is the same as the value of y, they are taken as the same fixed value; if the value of x is different from the value of y, the value of y of the air conditioner in the shutdown or fault state is greater than 0, otherwise the value of y is 0, wherein y≤x.

[0058] In the embodiment of the present invention, Figure 2 As shown, in step S4, variable CCON x The calculation formula is:

[0059] CCON x =∑G xy *C xm *Δthc n

[0060] Among them, Δthc n represents the temperature difference between m different hot and cold channels, C xm represents the weight factor, G xy represents the failure factor matrix.

[0061] The working principle and process of the present invention are as follows: The purpose of the present invention is to provide a data center fan and a control method thereof, wherein the data center fan includes a unit controller, a sensor network, a sensor and an external interface, and the control method allocates different cold and hot channel temperature difference weight factors and fault factors to different units, thereby reducing the energy consumption of the fan in the data center, improving energy utilization, and saving energy and protecting the environment while satisfying the temperature field distribution in the data center computer room to meet specific requirements. Moreover, the automatic redistribution of loads between units when a fault occurs is realized, thereby improving the reliability of the data center.

[0062] The beneficial effects of the present invention are:

[0063] (1) When the temperature field distribution in the data center room meets specific requirements, the present invention can also reduce the energy consumption of the data center fan, improve energy utilization, save energy and protect the environment. It also realizes automatic redistribution of loads between units when a fault occurs, thereby improving the reliability of the data center.

[0064] (2) By increasing the weight factor of the temperature abnormality area, the limit is set using the optimization algorithm. When the limit is reached, the weight factor of the temperature abnormality itself is automatically increased, so that the temperature abnormality area can be quickly restored to normal and the fan power consumption is only increased slightly as necessary.

[0065] (3) Combined with real-time optimization algorithms, the temperature field distribution in the data center room can be adjusted while keeping the fan at the lowest necessary power consumption.

[0066] (4) When an air conditioner in a conventional data center fails and stops working, it may cause abnormal distribution of the temperature field, and each unit cannot perceive this abnormality. However, the present invention can perceive it in real time and adjust the load redistribution in time to quickly restore it to normal.

[0067] Those skilled in the art will appreciate that the embodiments herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A control method for a data center fan, based on a data center fan, comprising a computer room module (1), a supply and return air duct (2), an air conditioning unit (3), an external interface (4) and an external optimization controller (5); the computer room module (1) is fixedly connected to the air conditioning unit (3) through the supply and return air duct (2); the external interface (4) is fixedly connected to the output end of the air conditioning unit (3); the external interface (4) is also communicatively connected to the external optimization controller (5); the computer room module (1) comprises an IT equipment cabinet (6), a cold channel (7) and a hot channel (8); the cold channel (7) and the hot channel (8) are fixedly connected alternately in sequence; the IT equipment cabinet (6) is fixedly arranged between the alternately fixedly connected cold channel (7) and the hot channel (8), the cold channel (7) and the hot channel (8) are both provided with a temperature sensor (9) and a temperature sensor network (10), the temperature sensor (9) is an NTC temperature sensor, and the temperature sensor network (10) is a star topology structure, characterized in that: The following steps are involved: S1: Calculate m different cold and hot channel temperature differences Δthc based on adjacent cold channels and hot channels n , and its calculation formula is: Δthc n =j -1 *∑CTh j *Thot nj -i -1 *∑CTc i *Tcold ni Among them, Thot nj Indicates the temperature of the jth temperature sensor on the hot aisle of the nth row of cabinets, Tcold ni Indicates the temperature of the ith temperature sensor on the cold aisle of the nth row of cabinets, CTh j Indicates the weight factor of the hot channel temperature sensor. The value is greater than or equal to 0. CTc i Indicates the weight factor of the cold channel temperature sensor, the value is greater than or equal to 0; S2: Different hot and cold channel temperature difference Δthc n Set the weight factor C xm ; Weight factor C xm is a matrix with x rows and m columns, where x represents the number of air conditioners and m represents the number of temperature differences between hot and cold channels; S3: Set the fault factor matrix G of the air conditioner xy ; If the value of x is the same as the value of y, then they are taken as the same fixed value; if the value of x is different from the value of y, then the value of y for the air conditioner in shutdown or fault state is greater than 0, otherwise the value of y is 0, where y≤x; S4: Based on the temperature difference Δthc of m different hot and cold channels n , weight factor C xm and the failure factor matrix G xy Computed variable CCON x , and its calculation formula is: CCON x =∑G xy *C xm *Δthc n ; S5: Set the fan control variable to CSET x ; S6: Adjust the variable CCON by trial and error x and the control variable CSET x The difference is minimized to complete the control of the data center fan.

Images