Movable air port guide pipe system of data center machine room and working method

Abstract

The invention belongs to the field of data center thermal management, and discloses a movable tuyere guide pipe system of a data center machine room and a working method.By collecting temperature information of different positions in real time and constructing temperature field distribution, local hot spot areas can be accurately recognized, and the working efficiency is improved. And the position and the air supply direction of the movable tuyere guide pipe are automatically adjusted according to the hot spot distribution. And in a refrigeration control strategy, a hot spot region sorting and grading refrigeration mechanism is adopted. And according to the degree of the temperature exceeding the threshold value and the duration time, the system automatically determines the cooling priority, and refrigeration equipment is sequentially started to directionally cool the hot spot area. Compared with a global unified cooling mode of a traditional system, the method has the advantages that fine distribution of refrigeration resources is achieved, excessive refrigeration and energy waste of a low-temperature area are avoided, the refrigeration power can be intensively used for the most needed area, and therefore the overall energy consumption is remarkably reduced.

Description

Technical Field

[0001] This invention belongs to the field of data center thermal management, specifically relating to a movable air duct system for data center computer rooms and its working method. Background Technology

[0002] As data center equipment continues to expand in scale and computing tasks become increasingly demanding, the problem of heat accumulation within the server room is becoming more and more prominent. Existing cooling systems are struggling to keep up with the rapid increase in computing power, leading to frequent hotspots, uneven airflow distribution, and a significant decrease in cooling efficiency during operation. Against this backdrop, how to rationally adjust airflow to ensure that cool air is precisely applied to the high-heat areas of the equipment has become crucial for improving the overall cooling performance of data centers.

[0003] Traditional cooling methods, such as downdraft or updraft systems, have several shortcomings. Air vents are typically fixed in specific locations, making it difficult to adjust them flexibly according to real-time load and heat distribution within the server room. Even with oscillating louvers to guide airflow, significant losses occur during air delivery to hot spots, resulting in low cooling efficiency. Furthermore, because cool air cannot be precisely delivered to hot spots, administrators are often forced to lower the overall server room temperature to maintain stable equipment operation. This not only increases the risk of condensation but also leads to substantial energy waste. In addition, the relatively fixed internal structure of server rooms means that adjusting vent locations or changing air delivery strategies often involves high modification costs and lengthy implementation periods.

[0004] In summary, existing technologies for data center cooling suffer from three main problems: first, the air supply method lacks flexibility and cannot dynamically adapt to load changes; second, the cooling efficiency is low, resulting in energy waste and the risk of condensation; and third, the fixed structure of the data center makes it difficult to adjust the air supply system quickly and cost-effectively. These problems make designing a flexible and adjustable system that can optimize the air supply path in real time according to changes in heat sources a pressing technical challenge. Summary of the Invention

[0005] The purpose of this invention is to overcome the problems of fixed air supply methods that are difficult to adjust flexibly, low cooling efficiency and serious energy waste, and to provide a movable air outlet duct system and working method for data center computer rooms.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a movable air outlet duct system for a data center computer room, including a refrigeration device connected to a static pressure box, a guide rail guiding mechanism provided on the top of the data center computer room, an air outlet provided on the guide rail guiding mechanism, the static pressure box connected to the air outlet through an air supply pipe, and a number of temperature sensors provided inside the data center computer room, all temperature sensors connected to a control unit, and the control unit connected to the refrigeration device and the guide rail guiding mechanism. The control unit is used to acquire data from the temperature sensor, identify hot spots based on the temperature sensor data, and control the movement of the guide rail mechanism and the cooling power of the refrigeration equipment based on the hot spots.

[0007] A further improvement of the present invention is that the guide rail guiding mechanism includes a first guide rail, which consists of two guide rails respectively arranged on both sides of the top of the data center server room, and a second guide rail is arranged between the two guide rails, with the air outlet arranged on the second guide rail.

[0008] A further improvement of the present invention is that the second guide rail and the air outlet are both driven by corresponding synchronous belts, the synchronous belts are connected to corresponding drive motors, and all drive motors are connected to a control unit, which is used to control the start, stop, direction and speed of the drive motors.

[0009] A further improvement of the present invention is that a guide vane is provided on the air outlet.

[0010] A further improvement of the present invention is that the air supply duct is a corrugated pipe.

[0011] A method for operating a movable air duct system for a data center computer room includes the following steps: Acquire temperature information from different locations within the data center server room, preprocess the temperature information to obtain the required temperature information; The preprocessed temperature information is interpolated to obtain temperature field data for each area within the data center. The temperature field data is processed based on a preset temperature threshold to obtain hotspot areas with temperatures higher than the temperature threshold, and the hotspot areas are sorted according to preset rules. According to the order of the hot spots, the cooling equipment is activated to assist in cooling the hot spots in turn until the temperature of all hot spots is reduced below the temperature threshold.

[0012] A further improvement of this invention lies in the method of acquiring temperature information from different locations within a data center server room, preprocessing the temperature information, and obtaining the required temperature information, as follows: Temperature sensors are deployed at different locations within the data center server room to obtain the coordinate information of all temperature sensors, forming a set of coordinates for the temperature sensors. The temperature information from all temperature sensors is acquired, first-order linear calibration is performed on the temperature information, and the drift of the temperature sensors is corrected to obtain the corrected temperature information.

[0013] A further improvement of this invention lies in the following method for interpolating the preprocessed temperature information to obtain temperature field data for each area within the data center: The corrected temperature information is traversed, and noise reduction and outlier removal are performed. The selected monitoring plane is discretized into a regular grid. Based on the correspondence between measuring points and the grid, the grid nodes are interpolated using an inverse distance weighting method to reconstruct the temperature field data. A further improvement of this invention lies in processing the temperature field data based on a preset temperature threshold to obtain hotspot regions with temperatures exceeding the threshold. The specific method for sorting these hotspot regions according to preset rules is as follows: Preset the main temperature threshold and hysteresis threshold, binarize the temperature field data, and obtain the hotspot set; The area, peak temperature, temperature excess, and centroid data of each hotspot are obtained, wherein the temperature excess is the integral amount of the portion of the temperature exceeding the main temperature threshold within the hotspot area. The area, peak temperature, temperature excess, and centroid data of each hotspot are assigned corresponding weight values ​​and weighted calculations are performed to finally obtain the ranking of the hotspot areas.

[0014] Energy balance is performed on the equivalent controlled area of ​​the hot spot area ranked first to obtain the required air flow rate, and the guide rail guiding mechanism (3) is controlled to move the air outlet (4) above the hot spot area. The cooling equipment (1) is started according to the required air flow rate to cool down the hot spot area ranked first until the temperature of the hot spot area ranked first drops below the temperature threshold. Repeat the above steps for the remaining hotspot areas in the order described until the temperature of all hotspot areas drops below the temperature threshold.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention achieves dynamic adjustment of airflow direction and position through the design of a guide rail mechanism and movable air outlets. The air outlets can move along the guide rails on the top of the computer room, precisely delivering cool air to hot spots by combining real-time temperature monitoring data, avoiding overcooling of non-hot spots and thus improving cooling efficiency. The combined design of the plenum chamber and air duct ensures that the airflow output from the cooling equipment is evenly distributed and balanced before entering the computer room, avoiding uneven airflow distribution and insufficient or excessive cooling in some areas. By buffering airflow fluctuations through the plenum chamber, the system maintains stable air pressure and smooth airflow under different loads and airflow conditions, thereby improving overall airflow quality and cooling uniformity. Multiple temperature sensors in this invention form a real-time temperature monitoring network. The control unit can automatically identify hot spots in the computer room based on the temperature data fed back by the sensors and control the guide rail mechanism and cooling equipment to make corresponding adjustments. This intelligent control method enables the system to dynamically respond to load changes within the computer room, possessing adaptive adjustment capabilities and significantly improving the intelligence level and response speed of the cooling system. The power regulation of the cooling equipment is automatically completed by the control unit, which flexibly adjusts the output power according to the number of hot spots and temperature changes, avoiding the energy waste of continuous high-power operation in traditional systems, and realizing on-demand cooling and energy efficiency optimization. In summary, this invention, with its modular structure and flexible control method, is suitable for data centers of different sizes and layouts. Through the organic combination of intelligent identification, dynamic air supply, and energy consumption linkage management, the system has significant comprehensive advantages in reducing operating energy consumption, improving cooling efficiency, extending equipment life, and ensuring stable operation of the computer room.

[0016] This invention, by collecting temperature information from different locations in real time and constructing a temperature field distribution, can accurately identify local hotspot areas and automatically adjust the position and airflow direction of movable air ducts based on the hotspot distribution. In terms of cooling control strategy, a hotspot area sorting and graded cooling mechanism is adopted. Based on the degree and duration of temperature exceeding the threshold, the system automatically determines the cooling priority and sequentially activates cooling equipment to provide targeted cooling to the hotspot areas. Compared to the traditional method of global uniform cooling, this method achieves refined allocation of cooling resources, avoiding over-cooling and energy waste in low-temperature areas, and concentrating cooling power on the areas most in need, thereby significantly reducing overall energy consumption. Through temperature data preprocessing and dynamic analysis, sensing errors and environmental interference can be effectively eliminated, improving the accuracy of temperature field monitoring. Through continuous temperature data updates, the system can reflect changes in heat flow within the computer room in real time, immediately adjusting the airflow direction and intensity when new hotspots appear to maintain a balanced temperature within the computer room and prevent performance degradation or malfunctions caused by equipment overheating. In summary, this invention is significantly superior to traditional solutions in terms of airflow flexibility, cooling accuracy, and energy consumption control, ensuring that data center equipment operates in a more stable temperature environment while reducing energy consumption. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the air outlet structure of the present invention; Figure 3 This is a schematic diagram of the guide rail guiding mechanism in this invention; Figure 4 This is a flowchart of the present invention; Among them, 1. Refrigeration equipment; 2. Static pressure box; 3. Guide rail guiding mechanism; 4. Air outlet; 5. Air supply duct; 6. Guide louvers; 3-1. First guide rail; 3-2. Second guide rail. Detailed Implementation

[0018] To further understand the content of this invention, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0019] See Figure 1 A movable air duct system for a data center computer room includes a cooling unit 1 connected to a static pressure box 2. A guide rail mechanism 3 is installed on the top of the data center computer room, and an air outlet 4 is installed on the guide rail mechanism 3. The static pressure box 2 is connected to the air outlet 4 through an air supply pipe 5. Several temperature sensors are installed inside the data center computer room, and all temperature sensors are connected to a control unit. The control unit is connected to the cooling unit 1 and the guide rail mechanism 3. The control unit is used to acquire data from the temperature sensors, identify hot spots based on the temperature sensor data, and control the movement of the guide rail mechanism 3 and the cooling power of the cooling unit 1 based on the hot spots.

[0020] See Figure 2 and Figure 3 The guide rail guiding mechanism 3 includes a first guide rail 3-1, which consists of two rails respectively positioned on either side of the top of the data center server room. A second guide rail 3-2 is positioned between the two rails, and an air outlet 4 is located on the second guide rail 3-2. Both the second guide rail 3-2 and the air outlet 4 are driven by corresponding synchronous belts, which are connected to corresponding drive motors. All drive motors are connected to a control unit, which controls the start, stop, direction, and speed of the drive motors. The air outlet 4 is equipped with guide vanes 6. The air supply duct 5 is a corrugated pipe.

[0021] After system startup, the control unit first wakes up all temperature sensors. Each temperature sensor collects temperature data for its area in real time and uploads it to the control unit. The control unit filters, averages, and interpolates the temperature data to obtain a complete two-dimensional temperature distribution field within the server room, thus accurately reflecting the current thermal environment status of the server room. The control unit compares the temperature field data with set temperature thresholds to identify hotspot areas where the temperature exceeds the threshold. If multiple hotspot areas exist, the system prioritizes them based on the degree of temperature exceeding limits, server rack density, or power load to determine the cooling order.

[0022] Based on the coordinates of the hotspot area within the computer room, the control unit sends movement commands to the guide rail mechanism. The motor on the first guide rail 3-1 drives the second guide rail 3-2 to move longitudinally along the computer room; the motor on the second guide rail drives the air outlet 4 to move laterally; together, they achieve precise two-dimensional positioning of the air outlet at the top of the computer room. The movement of the guide rails and air outlet is executed by a synchronous belt drive mechanism, and the start, stop, direction, and speed of the drive motor are precisely controlled by the control unit, enabling the air outlet to be moved quickly and smoothly above the target hotspot.

[0023] Once the air outlet is positioned, the control unit dynamically adjusts the cooling power of the refrigeration unit 1 based on the temperature deviation in the hot spot area. After being pressurized by the static pressure box 2, the cooled air is delivered to the air outlet 4 through the corrugated air duct 5, and then directionally ejected through the guide louvers 6 at the air outlet. The guide louvers can adjust the air outlet angle to specifically enhance the coverage of cold air in the hot spot area and improve local cooling efficiency.

[0024] During the cooling process, temperature sensors near the hotspot area continuously provide real-time temperature feedback. The control unit judges the cooling effect based on the temperature change curve. When the temperature in the hotspot area drops below a set threshold, it automatically reduces or shuts down the cooling power in that area and re-identifies other hotspot areas based on the new temperature distribution, continuing the cycle. Through this dynamic closed-loop control, the system can achieve automatic hotspot identification, automatic air outlet positioning, and adaptive adjustment of cooling power.

[0025] This invention effectively reduces overall energy consumption while ensuring uniform temperature across all server racks. The air outlet location and cooling capacity are dynamically allocated based on real-time load changes, avoiding over-cooling and energy waste under fixed airflow methods, thus achieving intelligent distribution and efficient utilization of cooling capacity in the data center.

[0026] See Figure 4 A method for operating a movable air duct system for a data center computer room includes the following steps: S1: Obtain temperature information from different locations within the data center server room, preprocess the temperature information, and obtain the required temperature information.

[0027] S2 performs interpolation on the preprocessed temperature information to obtain temperature field data for each area within the data center.

[0028] S3, the control unit processes the temperature field data based on a preset temperature threshold to obtain hotspot areas with temperatures higher than the temperature threshold, and sorts the hotspot areas according to preset rules.

[0029] S4, the control unit starts the cooling equipment 1 to assist in cooling the hot spots in sequence according to the sorting results of the hot spots, until the temperature of all hot spots is reduced to below the temperature threshold.

[0030] Example 1: The system deploys temperature probes (e.g., NTC / PT100 or wireless temperature tags) on a selected monitoring plane (such as the top plane of the cabinet) to form a coordinate set. :

[0031] Preset sampling period And the unified timestamp NTP.

[0032] Acquire temperature information from all temperature sensors and perform first-order linear calibration on the temperature information:

[0033] in, This is the actual temperature. The temperature value measured by the temperature sensor. This represents the linear proportional deviation of the sensor readings. This is a constant error caused by the overall temperature of the sensor being too high or too low.

[0034] The calibrated temperature information is iterated through, and outliers are removed using median filtering: In the window inside, if If it is an anomaly, then time-domain substitution estimation (EWMA) smoothing is used instead:

[0035] in, This represents the temperature measurement value of the temperature sensor at a specific moment. The median temperature within the window. This represents the median absolute deviation. This is the tolerance factor. For the current moment The smoothed temperature value, For the current moment The original measured temperature value, This is the smoothing coefficient.

[0036] Discretize the selected plane into a regular grid. , There are 1 grid, with the center of the grid being 1. Reconstructing the temperature field from point temperature measurements :

[0037]

[0038] For sensors Position Planar weighting coefficients, The number of sensors that can effectively participate in interpolation. For the first The planar coordinates of a temperature sensor, For distance smoothing term, This is the decay index.

[0039] Preset temperature master threshold With hysteresis threshold This is used to avoid controlling jitter.

[0040] Binarize the temperature field data:

[0041] Denoising is achieved by performing morphological opening and closing operations on the plane, followed by labeling connected components by 8-neighborhood to obtain a set of hotspots. .

[0042] Get the area of ​​each hotspot Peak temperature Excessive temperature and center of mass data, This represents the area of ​​a single grid point.

[0043]

[0044]

[0045]

[0046]

[0047] Area of ​​each hotspot Peak temperature Excessive temperature and center of mass The data is assigned corresponding weight values, and the final ranking of hotspot areas is obtained. :

[0048] in, This is the temperature normalization factor. As the heat normalization factor, This represents the path distance between the current position of the air outlet and the centroid of the hot spot within the two-dimensional guide plane. This represents the current coordinates of the wind vent.

[0049] Hot topics Target cooling Energy balance is performed on the equivalent controlled area (the control area above the hotspot) to approximate the required air mass flow rate.

[0050]

[0051] in, This represents the equivalent heat dissipation power of the hotspot. , For supply air temperature, For the temperature near the hotspot, if we only consider the deceleration rate of the temperature difference, a first-order room temperature model can be used:

[0052] in, To represent the equivalent heat capacity of the hot spot region calculated per unit thickness, For mass flow rate.

[0053] According to the sorted queue [ Control the air outlet to move to Above, turn on the corresponding fan speed to the set point, when the hot spot Temperature field satisfies And after 2–5 minutes, it transitions to a hotspot. If there are multiple air outlets, allocate them according to local optimality, avoid moving across them, until the temperature of all hot spots drops below the temperature threshold.

[0054] Example 2: The existing data center server room was selected for renovation. The server room has an area of ​​304.7m² and a ceiling height of 5.3 meters, and is divided into an air conditioning room and an equipment room. The air conditioning room is equipped with 6 chilled water precision air conditioners, operating in a 4-in-operation and 2-out-of-service configuration. Each air conditioner has a maximum air volume of 9 m³ / s, a minimum air supply temperature of 21℃, and a relative humidity of 50%.

[0055] The equipment room contains 8 rows of server racks (15 racks per row and one power distribution cabinet), using standard 47U racks measuring 600mm × 1100mm × 2200mm. Each IT device in the rack has a heat dissipation capacity of 200W, with a single rack power limit of 4kW, and a total power consumption of 202kW. The server room adopts an underfloor return airflow system, with a raised floor height of 800mm.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A movable air duct system for a data center computer room, characterized in that, Includes a refrigeration device (1), which is connected to a static pressure box (2). A guide rail mechanism (3) is installed on the top of the data center computer room. An air outlet (4) is installed on the guide rail mechanism (3). The static pressure box (2) is connected to the air outlet (4) through an air supply pipe (5). Several temperature sensors are installed in the data center computer room. All temperature sensors are connected to a control unit. The control unit is connected to the refrigeration device (1) and the guide rail mechanism (3). The control unit is used to acquire data from the temperature sensor, identify hot spots based on the data from the temperature sensor, and control the movement of the guide rail guiding mechanism (3) and the cooling power of the cooling equipment (1) based on the hot spots.

2. The movable air duct system for a data center computer room according to claim 1, characterized in that, The guide rail guiding mechanism (3) includes a first guide rail (3-1), which consists of two guide rails. The two guide rails are respectively set on both sides of the top of the data center computer room. A second guide rail (3-2) is set between the two guide rails. The air outlet (4) is set on the second guide rail (3-2).

3. A movable air duct system for a data center computer room according to claim 1 or 2, characterized in that, The second guide rail (3-2) and the air outlet (4) are both driven by corresponding synchronous belts. The synchronous belts are connected to the corresponding drive motors. All drive motors are connected to the control unit, which is used to control the start, stop, direction and speed of the drive motors.

4. A movable air duct system for a data center computer room according to claim 1 or 2, characterized in that, The air outlet (4) is equipped with guide vanes (6).

5. A movable air duct system for a data center computer room according to claim 1, characterized in that, The air supply duct (5) is a corrugated pipe.

6. A method for operating the movable air duct system for a data center computer room as described in claim 1, characterized in that, Includes the following steps: Acquire temperature information from different locations within the data center server room, preprocess the temperature information to obtain the required temperature information; The preprocessed temperature information is interpolated to obtain temperature field data for each area within the data center. The temperature field data is processed based on a preset temperature threshold to obtain hotspot areas with temperatures higher than the temperature threshold, and the hotspot areas are sorted according to preset rules. According to the order of the hot spots, start the cooling equipment (1) to cool down the hot spots in sequence until the temperature of all hot spots drops below the temperature threshold.

7. The working method of a movable air duct system for a data center computer room according to claim 6, characterized in that, The specific method for obtaining temperature information from different locations within a data center and preprocessing this information to obtain the required temperature information is as follows: Temperature sensors are deployed at different locations within the data center server room to obtain the coordinate information of all temperature sensors, forming a set of coordinates for the temperature sensors. The temperature information from all temperature sensors is acquired, first-order linear calibration is performed on the temperature information, and the drift of the temperature sensors is corrected to obtain the corrected temperature information.

8. The working method of a movable air duct system for a data center computer room according to claim 6, characterized in that, The specific method for interpolating the preprocessed temperature information to obtain temperature field data for each area within the data center is as follows: The corrected temperature information is traversed to perform noise reduction and outlier removal; Discretize the selected monitoring plane into a regular grid; Based on the correspondence between measurement points and grids, the inverse distance weighting method is used to interpolate the grid nodes to reconstruct the temperature field data.

9. The working method of a movable air duct system for a data center computer room according to claim 6, characterized in that, The temperature field data is processed based on a preset temperature threshold to obtain hotspot regions with temperatures higher than the threshold. The specific method for sorting these hotspot regions according to preset rules is as follows: Preset the main temperature threshold and hysteresis threshold, binarize the temperature field data, and obtain the hotspot set; The area, peak temperature, temperature excess, and centroid data of each hotspot are obtained, wherein the temperature excess is the integral amount of the portion of the temperature exceeding the main temperature threshold within the hotspot area. The area, peak temperature, temperature excess, and centroid data of each hotspot are assigned corresponding weight values ​​and weighted calculations are performed to finally obtain the ranking of the hotspot areas.

10. The working method of a movable air duct system for a data center computer room according to claim 6, characterized in that, According to the order of the hot spots, the cooling equipment (1) is started to cool down the hot spots in sequence until the temperature of all hot spots is reduced to below the temperature threshold. The specific method is as follows: Energy balance is performed on the equivalent controlled area of ​​the hot spot area ranked first to obtain the required air flow rate, and the guide rail guiding mechanism (3) is controlled to move the air outlet (4) above the hot spot area. The cooling equipment (1) is started according to the required air flow rate to cool down the hot spot area ranked first until the temperature of the hot spot area ranked first drops below the temperature threshold. Repeat the above steps for the remaining hotspot areas in the order described until the temperature of all hotspot areas drops below the temperature threshold.

Images