Method, device and system for controlling temperature and humidity in a computer room

Abstract

The present invention relates to the field of temperature measurement technology, and specifically to a method, device, and system for controlling temperature and humidity in a computer room, comprising: obtaining temperature data of each temperature sensor and the power of its monitored equipment, and predicting the predicted power of the monitored equipment; performing a hysteresis correlation analysis on the temperature data of each temperature sensor and the power of its monitored equipment, as well as the power covariance relationship with each control unit of the heat dissipation device, thereby obtaining the heat release weight of the monitored equipment of each temperature sensor; obtaining the device feedback moment of each temperature sensor at the current moment; analyzing the heat accumulation effect of the monitored equipment, and obtaining the temperature at the device feedback moment in combination with the heat release weight, thereby achieving temperature control at the current moment. The present invention aims to solve the time lag problem of the existing computer room temperature and humidity control method based on real-time temperature control, and to achieve computer room temperature control by obtaining the temperature of the equipment when it can be adjusted in the future.

Description

Technical Field

[0001] The present invention relates to the technical field of temperature measurement, and in particular to a method, device and system for controlling temperature and humidity in a computer room. Background Art

[0002] Customs commodity inspection requires very high standards for professional technical site management. With the rapid development of information technology, data centers and computer rooms have become critical infrastructure supporting customs commodity inspection operations. High-density hardware such as servers, storage devices, and network equipment generates significant heat during continuous operation, requiring efficient air conditioning systems to manage heat and ensure stable operation within an appropriate temperature and humidity environment. In recent years, the demand for environmental control within computer rooms has increased, particularly in the pursuit of high efficiency, low carbon emissions, and energy conservation and emission reduction. Precise temperature and humidity control has become a key component of green data center construction. The implementation of intelligent temperature and humidity control technology not only helps extend equipment life and reduce energy consumption, but also improves the operational efficiency and reliability of the entire data center.

[0003] The existing temperature and humidity control system in customs and commodity inspection rooms primarily relies on temperature and humidity sensors to monitor temperature and humidity at fixed locations, but ignores the dynamic impact of server load on local temperature. Because server load fluctuations at different points in time lead to uneven heat dissipation, a single environmental sensor and a fixed temperature and humidity control strategy struggle to respond promptly, easily leading to localized overheating or wasted cooling. Furthermore, the system lacks the ability to integrate and analyze real-time server loads. Consequently, temperature and humidity control is performed after both heat dissipated by the load and temperature and humidity measured by the temperature sensors are detected, resulting in a lag in the temperature sensor's measurement results. Summary of the Invention

[0004] The present invention provides a method, device and system for controlling temperature and humidity in a machine room to solve the existing problems.

[0005] The present invention provides a method, device and system for controlling temperature and humidity in a computer room using the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a method for controlling temperature and humidity in a computer room, the method comprising the following steps:

[0007] Obtain the temperature data of each temperature sensor in the computer room at each historical moment and the power and load data of the monitored equipment. Based on the power and load data of the monitored equipment, predict the predicted power of each monitored equipment obtained from each temperature sensor at each moment after the current moment. Obtain the power of each control unit of the heat dissipation equipment at each historical moment; where each control unit corresponds to a number of temperature sensors.

[0008] The hysteresis correlation analysis is used to analyze the covariance between the temperature data of each temperature sensor and the power of the monitored device, thereby obtaining the relative heat dissipation lag time of each temperature sensor and the heat release weight of the monitored device of each temperature sensor.

[0009] The hysteresis correlation analysis is used to determine the covariance between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor, and the cooling response time of each temperature sensor at the current moment is obtained;

[0010] The device feedback time of each temperature sensor at the current moment is obtained by combining the relative heat dissipation lag time and cooling response time of each temperature sensor;

[0011] Analyze the thermal accumulation effect of the predicted power of the monitored device from the current time to the device feedback time of each temperature sensor, and combine the heat release weight of the monitored device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback time;

[0012] The temperature of each temperature sensor at the time of device feedback is used to achieve temperature control at the current moment.

[0013] Preferably, the specific method of obtaining the heat dissipation relative lag time includes:

[0014] The average of the standard deviations of the power of all regulated devices within the preset time series window of each temperature sensor at each historical moment is recorded as the power fluctuation degree of the regulated devices of the temperature sensor at each historical moment;

[0015] Dividing all moments according to the power fluctuation degree to obtain a plurality of first time periods for each temperature sensor;

[0016] The hysteresis correlation analysis is performed on the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device in each first time period to obtain the relative hysteresis time of heat dissipation of each temperature sensor.

[0017] Preferably, the specific method of obtaining the first time period includes:

[0018] Record any temperature sensor as the target temperature sensor and any historical moment as the target moment;

[0019] If the power fluctuation degree of the supervised device of the target temperature sensor at the target time is greater than or equal to a preset ratio of the maximum power fluctuation degree of the supervised device of the target temperature sensor, the target time is recorded as a first split point of the target temperature sensor;

[0020] The time period between every two adjacent first dividing points is recorded as the first time period of the target temperature sensor.

[0021] Preferably, the hysteresis correlation analysis of the covariance relationship between the temperature data of each temperature sensor and the power of the supervised device in each first time period to obtain the relative heat dissipation lag time of each temperature sensor includes:

[0022] Recording any first time period of the target temperature sensor as a target first time period of the target temperature sensor;

[0023] Recording a sequence of temperature data of the target temperature sensor in the target first time period as a temperature data sequence of the target temperature sensor in the target first time period;

[0024] Record the sequence consisting of the average power values ​​of all supervised devices of the target temperature sensor in the target first time period as the power sequence of the target temperature sensor in the target first time period;

[0025] Initialize the offset number n=0, shift each sequence value in the temperature data sequence of the target temperature sensor in the target first time period backward by the offset number of bits, and obtain the nth offset temperature data sequence of the target temperature sensor in the target first time period;

[0026] For the number of misalignments n+1, obtaining the n+1th misalignment temperature data sequence of the target temperature sensor in the target first time period;

[0027] After obtaining the Pearson correlation coefficient between the power sequence of the target temperature sensor in the target first time period and each misaligned temperature data sequence of the target temperature sensor in the target first time period, the number of misalignments of the misaligned temperature data sequence with the largest Pearson correlation coefficient is recorded as the first lag time of the target temperature sensor in the target first time period;

[0028] The first lag time of the target temperature sensor in all first time periods is recorded as the heat dissipation relative lag time of the target temperature sensor.

[0029] Preferably, the specific method of obtaining the heat release weight includes:

[0030] Obtain the first-order difference of the power mean of all regulated devices of each temperature sensor at each historical moment;

[0031] Obtain the first-order difference of the temperature data of each temperature sensor at each historical moment;

[0032] The heat release weight of the regulated device of the temperature sensor is obtained, wherein the heat release weight is inversely proportional to the first-order difference of the power mean and is directly proportional to the first-order difference of the temperature data of each temperature sensor at a relative lag time of backward heat release at each historical moment.

[0033] Preferably, the specific method of obtaining the refrigeration response time includes:

[0034] Obtain the relative cooling lag time of each temperature sensor;

[0035] The current moment and the cooling relative lag time are added to obtain the cooling response moment of each temperature sensor at the current moment.

[0036] Preferably, the specific method of obtaining the device feedback time includes:

[0037] The difference between the cooling response time of each temperature sensor at the current moment and the relative heat dissipation lag time of each temperature sensor at the current moment is recorded as the device feedback time of each temperature sensor at the current moment.

[0038] Preferably, the specific method of obtaining the temperature of each temperature sensor at the device feedback moment includes:

[0039] The cumulative sum of the absolute values ​​of the predicted power increments of each regulated device of each temperature sensor from the current moment to the device feedback moment of the temperature sensor at the current moment is recorded as the thermal accumulation coefficient of the regulated device of each temperature sensor;

[0040] The temperature increase of each temperature sensor at the device feedback moment is obtained; the temperature increase is directly proportional to the heat accumulation coefficient of the monitored device of each temperature sensor and the heat release weight of the monitored device of each temperature sensor.

[0041] In a second aspect, the present invention further proposes a temperature and humidity control system in a computer room, the system comprising:

[0042] Data acquisition module: used to obtain temperature data of each temperature sensor in the computer room at each historical moment and the power and load data of the monitored equipment. Based on the power and load data of the monitored equipment, it predicts the predicted power of each monitored equipment obtained from each temperature sensor at each moment after the current moment; it also obtains the power of each control unit of the heat dissipation equipment at each historical moment; where each control unit corresponds to a number of temperature sensors;

[0043] Device feedback moment temperature acquisition module: This module is used to perform hysteresis correlation analysis on the covariance between the temperature data of each temperature sensor and the power of the monitored device, obtain the relative heat dissipation lag time of each temperature sensor, and then obtain the heat release weight of the monitored device for each temperature sensor;

[0044] The hysteresis correlation analysis is used to determine the covariance between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor, and the cooling response time of each temperature sensor at the current moment is obtained;

[0045] The device feedback time of each temperature sensor at the current moment is obtained by combining the relative heat dissipation lag time and cooling response time of each temperature sensor;

[0046] Analyze the thermal accumulation effect of the predicted power of the monitored device from the current time to the device feedback time of each temperature sensor, and combine the heat release weight of the monitored device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback time;

[0047] Temperature control module: used to use the temperature of each temperature sensor at the time of device feedback to achieve temperature control at the current moment.

[0048] In the third aspect, the present invention also proposes a temperature and humidity control device in a computer room, including a heat dissipation device, the heat dissipation device including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above method when executing the computer program.

[0049] The beneficial effects of the technical solution of the present invention are as follows: the present invention obtains the temperature data of each temperature sensor in the computer room at each historical moment and the power and load data of the supervised equipment, and predicts the predicted power of each supervised equipment of each temperature sensor at each moment after the current moment according to the power and load data of the supervised equipment; obtains the power of each control unit of the heat dissipation equipment at each historical moment; wherein, each control unit corresponds to a number of temperature sensors; performs a hysteresis correlation analysis on the covariance relationship between the temperature data of each temperature sensor and the power of its supervised equipment to obtain the relative lag time of heat dissipation of each temperature sensor; obtains the lag time length after the supervised equipment of each temperature sensor receives the temperature change after generating heat; and then obtains the heat release weight of the supervised equipment of each temperature sensor; represents the relationship between the temperature and power of the supervised equipment of each temperature sensor; hysteresis The covariance relationship between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor is analyzed to obtain the cooling response time of each temperature sensor at the current moment; the lag time length of the temperature sensor receiving the temperature change after the control unit of the heat dissipation device is adjusted is obtained; the relative heat dissipation lag time of each temperature sensor and the cooling response time are combined to obtain the device feedback time of each temperature sensor at the current moment; the time when the regulated device can be affected by adjusting the heat dissipation device from the current moment is expressed; the thermal accumulation effect of the predicted power of the regulated device from the current moment to the device feedback moment is analyzed, and the heat release weight of the regulated device of each temperature sensor is combined to obtain the temperature of each temperature sensor at the device feedback moment; the temperature control at the current moment is realized by using the temperature of each temperature sensor at the device feedback moment. The present invention aims to solve the time lag problem of the existing computer room temperature and humidity control method based on real-time temperature control, and realize the temperature control of the customs commodity inspection computer room by obtaining the temperature of the device when it can be adjusted in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 The present invention is a flow chart of the steps of a method for controlling temperature and humidity in a computer room. DETAILED DESCRIPTION

[0052] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method, device, and system for controlling temperature and humidity in a computer room, including its specific implementation, structure, features, and effectiveness. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0053] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0054] The following describes in detail a method, device and system for controlling temperature and humidity in a computer room provided by the present invention with reference to the accompanying drawings.

[0055] First, see Figure 1 , which shows a flow chart of a method for controlling temperature and humidity in a computer room according to an embodiment of the present invention, the method comprising the following steps:

[0056] Step S001: Obtain temperature data of each temperature sensor in the computer room at each historical moment and power and load data of its supervised equipment; predict the predicted power of each supervised equipment obtained from each temperature sensor at each moment after the current moment based on the power and load data of the supervised equipment; obtain the power of each control unit of the heat dissipation equipment at each historical moment; wherein each control unit corresponds to a number of temperature sensors.

[0057] The temperature and humidity control of traditional computer rooms usually uses temperature sensors to monitor the temperature and humidity of the computer room. If the temperature is higher than a constant temperature, the power of the heat dissipation device is increased to control the temperature and humidity in the computer room. However, in this process, the power of the electronic equipment such as the server in the computer room changes first, and there is almost a resistive resistor in the computer room, causing all electrical energy to be converted into heat energy. After the server and other electronic equipment gradually release heat by consuming electricity, it is detected by the temperature sensor after a period of time. After the temperature change is detected, the heat is fed back to the heat dissipation device. The heat dissipation device needs to adjust its power, and the cold air takes some time to reach the electronic equipment such as the server that dissipates heat, resulting in a lag in the temperature and humidity control. Therefore, this embodiment proposes a method for controlling temperature and humidity in a computer room to solve the above problems.

[0058] First, temperature sensors need to be installed in the computer room to obtain temperature data. In order to achieve temperature detection at multiple points, ensure uniform temperature distribution in the computer room and avoid local overheating or heat island effect, temperature sensors need to be deployed in key locations of the computer room, such as in the middle and top of the rack, hot aisles, cold aisles, air outlets, exhaust vents, etc., and collect temperature change data in real time.

[0059] Specifically, each cabinet in the computer room is divided into several three-dimensional monitoring areas in equal proportion. A high-precision digital temperature sensor is installed in the center of each three-dimensional monitoring area to measure the temperature data of the temperature sensor at each historical moment.

[0060] Furthermore, electronic devices such as servers in the three-dimensional monitoring area to which each temperature sensor belongs are recorded as the monitored devices of each temperature sensor. While measuring the temperature data of the temperature sensor at each historical moment, the power of each monitored device of each temperature sensor at each moment is obtained.

[0061] It should be noted that, as an example, this embodiment uses a Sensirion SHT85 temperature sensor to collect temperature data at each moment, and the time interval between two adjacent moments is described as 5 seconds; the sampling time range of this embodiment is described as 12 hours.

[0062] It should be noted that inside the computer room, according to the cooling load requirements and air flow organization of different areas, the heat dissipation equipment adopts in-row air conditioning (In-Row Cooling). After the cold air is generated from the heat dissipation equipment, it enters the underground of the computer room through the cold channel buried under the computer room. The cooling power, air supply volume and humidity settings of the air conditioner in the corresponding computer room area are dynamically adjusted through the control units of the heat dissipation equipment to achieve heat absorption of the computer room. The generated hot air is guided to the return air outlet of the air conditioner or to the outdoors through the exhaust vents or heat recovery system above the computer room.

[0063] Specifically, the control module of each cabinet area in the computer room is used as a control unit. Each control unit corresponds to several temperature sensors in the cabinet above it. By monitoring the cooling power of the air conditioner controlled by each control unit, the power of each control unit of the heat dissipation equipment at each historical moment is obtained.

[0064] Considering that the temperature changes in the computer room are mainly due to the dynamic heating caused by the differences in the operating load of the servers in the computer room at different time points, it is necessary to introduce a server load monitoring module to obtain the key load data of the servers in each rack, such as CPU utilization, memory usage, and power consumption. These load data are strictly related to power. Therefore, the relationship between load data and power at each historical moment can be used to build a power prediction model for each monitored device.

[0065] Specifically, the load data of each monitored device of each temperature sensor at each moment is obtained in real time through the IPMI (Intelligent Platform Management Interface) interface.

[0066] Furthermore, the power and load data of each regulated device of each temperature sensor at all historical moments are used as a training set, and a long short-term memory artificial neural network (LSTM) is trained to obtain a power prediction model for each regulated device of each temperature sensor. The current load data of each regulated device of each temperature sensor is input into the power prediction model to obtain the predicted power of each regulated device of each temperature sensor at each moment after the current moment.

[0067] It should be noted that the loss function of the power prediction model during training uses the mean square error loss function; the time range for predicting power is 30 moments after the current moment.

[0068] Step S002: hysteresis correlation analysis is performed on the covariance relationship between the temperature data of each temperature sensor and the power of its supervised device to obtain the relative heat dissipation hysteresis time of each temperature sensor, and then obtain the heat release weight of the supervised device of each temperature sensor.

[0069] It should be noted that the purpose of this embodiment is to control the temperature and humidity in the computer room by predicting the power of the monitored device at a specific time after the current time, and the time when the power at the specific time is monitored by the temperature sensor is equivalent to the time when the power of the cooling system is adjusted at the current time and detected by the temperature sensor; however, in this process, it is first necessary to obtain the relative lag time of the temperature data change of each temperature sensor compared to the power change of the monitored device.

[0070] When making future predictions based on server load in the data center, it takes time for heat to transfer from heat sources (server chips, fan exhaust) to temperature sensors. Therefore, there's a time lag between temperature changes and the power changes of the monitored devices. If the data center's temperature and humidity control system relies solely on real-time temperature data, it will be delayed and unable to respond promptly to hotspots. Therefore, it's necessary to first determine the relative heat dissipation lag between each temperature sensor's temperature changes and the power changes of the monitored devices.

[0071] It should be further explained that since the heat generated by the monitored equipment is mainly related to its instantaneous power, and the monitored equipment is usually a resistive load, all of its electrical energy consumption will be released in the form of heat. For the heat generation of different areas in the computer room, the heating situation in each area can be approximately represented by the sum of the instantaneous power of the monitored equipment of each temperature sensor.

[0072] Preferably, the specific steps of performing hysteresis correlation analysis on the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device to obtain the relative hysteresis time of heat dissipation of each temperature sensor are as follows:

[0073] According to the power of the supervised device of each temperature sensor, the power fluctuation degree of the supervised device of each temperature sensor at each historical moment is obtained;

[0074] Dividing all moments according to the power fluctuation degree to obtain a plurality of first time periods for each temperature sensor;

[0075] The hysteresis correlation analysis is performed on the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device in each first time period to obtain the relative hysteresis time of heat dissipation of each temperature sensor.

[0076] Specifically, according to the power of the supervised device of each temperature sensor, a specific method for obtaining the power fluctuation degree of the supervised device of each temperature sensor at each historical moment is as follows:

[0077] The average of the standard deviations of the power of all regulated devices within the preset time series window of each temperature sensor at each historical moment is recorded as the power fluctuation degree of the regulated devices of the temperature sensor at each historical moment;

[0078] It should be noted that the time series window described in this embodiment is described by taking the time series interval consisting of 30 moments after each historical moment as an example.

[0079] Furthermore, a specific method of obtaining a plurality of first time periods of each temperature sensor by dividing all moments according to the power fluctuation degree is as follows:

[0080] Record any temperature sensor as the target temperature sensor and any historical moment as the target moment;

[0081] If the power fluctuation degree of the supervised device of the target temperature sensor at the target time is greater than or equal to a preset ratio of the maximum power fluctuation degree of the supervised device of the target temperature sensor, the target time is recorded as a first split point of the target temperature sensor;

[0082] It should be noted that, in this embodiment, 0.2 times of the maximum power fluctuation degree of the supervised device of the target temperature sensor is used as the basis for obtaining the first segmentation point.

[0083] All first segmentation points of the target temperature sensor are acquired; and a time period between every two adjacent first segmentation points is recorded as a first time period of the target temperature sensor.

[0084] It should be noted that the first moment in history and the last moment in history are both recorded as the first division point.

[0085] Furthermore, the hysteresis correlation analysis is performed on the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device in each first time period. The specific method for obtaining the relative hysteresis time of heat dissipation of each temperature sensor is as follows:

[0086] Recording any first time period of the target temperature sensor as a target first time period of the target temperature sensor;

[0087] Recording a sequence of temperature data of the target temperature sensor in the target first time period as a temperature data sequence of the target temperature sensor in the target first time period;

[0088] Record the sequence consisting of the average power values ​​of all supervised devices of the target temperature sensor in the target first time period as the power sequence of the target temperature sensor in the target first time period;

[0089] Initialize the offset number n=0, shift each sequence value in the temperature data sequence of the target temperature sensor in the target first time period backward by the offset number of bits, and obtain the nth offset temperature data sequence of the target temperature sensor in the target first time period;

[0090] For the number of misalignments n+1, obtaining the n+1th misalignment temperature data sequence of the target temperature sensor in the target first time period;

[0091] Similarly, each misaligned temperature data sequence of the target temperature sensor in the target first time period is obtained;

[0092] It should be noted that, during the above shifting, each time a position is shifted backward, a value of 0 is added to the first vacant sequence position, and the last sequence value in the original temperature data sequence is discarded.

[0093] After obtaining the power sequence of the target temperature sensor in the target first time period and the Pearson correlation coefficient of each misaligned temperature data sequence of the target temperature sensor in the target first time period, the number of misalignments of the misaligned temperature data sequence with the largest Pearson correlation coefficient is recorded as the first lag time of the target temperature sensor in the target first time period.

[0094] Similarly, the first lag time of the target temperature sensor in all first time periods is obtained;

[0095] The first lag time of the target temperature sensor in all first time periods is recorded as the heat dissipation relative lag time of the target temperature sensor.

[0096] It should be noted that different temperature sensors have different types of monitored devices, resulting in different heat generation. Therefore, the heat release weight of each temperature sensor's monitored device is calculated by comparing the power change amplitude of all monitored devices of each temperature sensor with the temperature data change amplitude during the relative heat dissipation lag time. This weight is used to indicate the impact of power changes on the temperature of the monitored devices of that temperature sensor.

[0097] Preferably, the specific steps of obtaining the heat release weight of the supervised device of each temperature sensor include:

[0098] Obtain the first-order difference of the power mean of all regulated devices of each temperature sensor at each historical moment;

[0099] Obtain the first-order difference of the temperature data of each temperature sensor at each historical moment;

[0100] The heat release weight of the regulated device of the temperature sensor is obtained, wherein the heat release weight is inversely proportional to the first-order difference of the power mean and is directly proportional to the first-order difference of the temperature data of each temperature sensor at a relative lag time of backward heat release at each historical moment.

[0101] As an example of the present invention, a specific method for obtaining the heat release weight of the supervised device of the temperature sensor is as follows:

[0102] The first-order difference of the power mean of all supervised devices of the i-th temperature sensor at time t is recorded as ;

[0103] The relative heat dissipation lag time of the i-th temperature sensor is recorded as ;

[0104] Move the tth moment backward At the moment, the first-order difference of the temperature data of the i-th temperature sensor is recorded as ;

[0105] Heat release weight of the regulated device of the i-th temperature sensor The calculation method is:

[0106]

[0107] in, is the total number of historical moments.

[0108] Step S003: hysteresis correlation analysis is performed on the covariance relationship between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor to obtain the cooling response time of each temperature sensor at the current moment; the relative heat dissipation lag time and the cooling response time of each temperature sensor are combined to obtain the device feedback time of each temperature sensor at the current moment.

[0109] After the power of each control unit in the heat dissipation device is adjusted, it takes some time for the cool air to reach the temperature sensor. Therefore, when traditionally performing real-time control based on the temperature data collected by the temperature sensor, there is a time lag between the heat dissipation device and the temperature sensor. Therefore, this embodiment uses lag correlation analysis to analyze the covariance between the power of each control unit in the heat dissipation device and the temperature data from the corresponding temperature sensor to determine the cooling response time of each temperature sensor at the current moment.

[0110] Preferably, the specific steps of performing a hysteresis correlation analysis on the covariance relationship between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor to obtain the cooling response time of each temperature sensor at the current moment include:

[0111] After replacing the power of the monitored device of each temperature sensor with the power of each control unit of the heat dissipation device corresponding to each temperature sensor, the relative cooling lag time of each temperature sensor is obtained by using the method of obtaining the relative heat dissipation lag time of each temperature sensor;

[0112] It should be noted that since each control unit corresponds to multiple temperature sensors, and each temperature sensor corresponds to a single control unit, the mean of the standard deviations of the power of all regulated devices of the above temperature sensors is replaced by the standard deviation of the power of each temperature sensor corresponding to a single control unit.

[0113] The current moment and the cooling relative lag time are added to obtain the cooling response moment of each temperature sensor at the current moment.

[0114] It should be noted that since the monitored device is very close to the temperature sensor, the time it takes for the heat from the monitored device to reach the temperature sensor is shorter. Correspondingly, after the control unit of the heat dissipation device performs power regulation, the cold air needs to reach the monitored device before reaching the temperature sensor, so it takes longer. The cooling response time indicates the time when the temperature sensor can monitor the temperature change if the power regulation of the control unit is performed at the current moment, which is the cooling response time.

[0115] It is further necessary to explain that after obtaining the cooling response moment of each temperature sensor at the current moment, since it is known that the temperature change collected by each temperature sensor is caused by the heat generated by the monitored device of the temperature sensor at the device feedback moment of the relative lag time of heat dissipation forward from the cooling response moment, this embodiment obtains the moment when the monitored device generates heat when the power regulation of the control unit of the heat dissipation device at the current moment produces a cooling effect based on the relative lag time of heat dissipation and the cooling response moment.

[0116] Preferably, the specific steps of obtaining the device feedback time of each temperature sensor at the current moment by comprehensively considering the relative heat dissipation lag time and the cooling response time of each temperature sensor are:

[0117] The difference between the cooling response time of each temperature sensor at the current moment and the relative heat dissipation lag time of each temperature sensor at the current moment is recorded as the device feedback time of each temperature sensor at the current moment.

[0118] Step S004: Analyze the thermal accumulation effect of the predicted power of the supervised device from the current moment to the device feedback moment of each temperature sensor, combine the heat release weight of the supervised device of each temperature sensor, and obtain the temperature of each temperature sensor at the device feedback moment; use the temperature of each temperature sensor at the device feedback moment to achieve temperature control at the current moment.

[0119] It should be noted that when controlling the temperature in the computer room, it is necessary to obtain the temperature at each moment, and then use the control unit of the cooling system to dynamically adjust the power of each control unit at different moments based on the difference between the temperature value at each moment and the preset value. Therefore, when controlling the temperature in the computer room, it is necessary to first obtain the temperature at the moment being controlled.

[0120] Preferably, the specific steps of analyzing the thermal accumulation effect of the predicted power of the supervised device from the current moment to the device feedback moment of each temperature sensor and combining the heat release weight of the supervised device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback moment include:

[0121] The cumulative sum of the absolute values ​​of the predicted power increments of each regulated device of each temperature sensor from the current moment to the device feedback moment of the temperature sensor at the current moment is recorded as the thermal accumulation coefficient of the regulated device of each temperature sensor;

[0122] Obtaining a temperature increase of each temperature sensor at the device feedback moment; the temperature increase is directly proportional to the heat accumulation coefficient of the monitored device of each temperature sensor and the heat release weight of the monitored device of each temperature sensor;

[0123] The temperature of each temperature sensor at the device feedback moment is obtained according to the temperature data of each temperature sensor at the current moment and the temperature increment of each temperature sensor at the device feedback moment.

[0124] Specifically, the specific method for obtaining the temperature increase of each temperature sensor at the device feedback moment is:

[0125] The product of the heat accumulation coefficient of the monitored equipment of each temperature sensor and the heat release weight is recorded as the temperature increase of each temperature sensor at the equipment feedback moment.

[0126] Furthermore, based on the temperature data of each temperature sensor at the current moment and the temperature increment of each temperature sensor at the device feedback moment, the specific method for obtaining the temperature of each temperature sensor at the device feedback moment is:

[0127] The sum of the temperature data of each temperature sensor at the current moment and the temperature increase of each temperature sensor at the device feedback moment is recorded as the temperature of each temperature sensor at the device feedback moment.

[0128] It should be noted that after obtaining the temperature of all temperature sensors at their device feedback time, that is, the power regulation of each control unit of the heat dissipation device at the current moment, the moment when the supervised device generates heat that can be cooled is achieved. Therefore, the temperature control at the current moment is achieved through the temperature of each temperature sensor at the device feedback moment.

[0129] Specifically, the average temperature of all temperature sensors corresponding to each control unit of the heat dissipation device at the device feedback moment is obtained, and recorded as the temperature to be controlled of each control unit of the heat dissipation device at the current moment;

[0130] The temperature to be controlled of each control unit of the heat dissipation device at the current moment and the preset constant room temperature are input into each control unit of the heat dissipation device, and the temperature inside the room is controlled through the existing PID control method of the heat dissipation device.

[0131] Furthermore, by deploying high-precision humidity sensors (such as the Sensirion SHT85) to monitor relative humidity at various locations within the computer room in real time, and combining them with the humidification and dehumidification functions of the cooling equipment, the humidity control strategy is dynamically adjusted in real time. The system maintains the relative humidity in the computer room within a range of 45%-55%, which strikes a balance between preventing condensation and reducing the risk of static electricity, and is suitable for most data center environments. When server loads are high and temperatures rise, the system automatically reduces humidification and increases dehumidification power appropriately to prevent over-humidification and condensation. When loads are low and temperatures drop, the system increases humidification appropriately to prevent static electricity accumulation caused by low humidity.

[0132] In a second aspect, another embodiment of the present invention provides a temperature and humidity control system in a computer room, the system comprising:

[0133] Data acquisition module: used to obtain temperature data of each temperature sensor in the computer room at each historical moment and the power and load data of the monitored equipment. Based on the power and load data of the monitored equipment, it predicts the predicted power of each monitored equipment obtained from each temperature sensor at each moment after the current moment; it also obtains the power of each control unit of the heat dissipation equipment at each historical moment; where each control unit corresponds to a number of temperature sensors;

[0134] Device feedback moment temperature acquisition module: This module is used to perform hysteresis correlation analysis on the covariance between the temperature data of each temperature sensor and the power of the monitored device, obtain the relative heat dissipation lag time of each temperature sensor, and then obtain the heat release weight of the monitored device for each temperature sensor;

[0135] The hysteresis correlation analysis is used to determine the covariance between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor, and the cooling response time of each temperature sensor at the current moment is obtained;

[0136] The device feedback time of each temperature sensor at the current moment is obtained by combining the relative heat dissipation lag time and cooling response time of each temperature sensor;

[0137] Analyze the thermal accumulation effect of the predicted power of the monitored device from the current time to the device feedback time of each temperature sensor, and combine the heat release weight of the monitored device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback time;

[0138] Temperature control module: used to use the temperature of each temperature sensor at the time of device feedback to achieve temperature control at the current moment.

[0139] In the second aspect, the present invention also proposes a temperature and humidity control device in a computer room, including a heat dissipation device, the heat dissipation device including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above method when executing the computer program.

[0140] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for controlling temperature and humidity in a computer room, characterized in that: The method comprises the following steps: Obtaining temperature data from each temperature sensor in the computer room at each historical moment, as well as power and load data from its monitored devices, and predicting the predicted power of each monitored device at each moment after the current moment based on the power and load data from the monitored devices; obtaining the power of each control unit of a heat dissipation device at each historical moment, where the heat dissipation device is used to absorb heat from the computer room; wherein each control unit corresponds to a number of temperature sensors; The hysteresis correlation analysis is used to analyze the covariance between the temperature data of each temperature sensor and the power of the monitored device, thereby obtaining the relative heat dissipation lag time of each temperature sensor and the heat release weight of the monitored device of each temperature sensor. The hysteresis correlation analysis is used to determine the covariance between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor, and the cooling response time of each temperature sensor at the current moment is obtained; The device feedback time of each temperature sensor at the current moment is obtained by combining the relative heat dissipation lag time and cooling response time of each temperature sensor; Analyze the thermal accumulation effect of the predicted power of the monitored device from the current time to the device feedback time of each temperature sensor, and combine the heat release weight of the monitored device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback time; The temperature of each temperature sensor at the time of device feedback is used to achieve temperature control at the current moment.

2. A method for controlling temperature and humidity in a computer room according to claim 1, characterized in that: The specific method for obtaining the heat dissipation relative lag time includes: The average of the standard deviations of the power of all regulated devices within the preset time series window of each temperature sensor at each historical moment is recorded as the power fluctuation degree of the regulated devices of the temperature sensor at each historical moment; Dividing all moments according to the power fluctuation degree to obtain a plurality of first time periods for each temperature sensor; The hysteresis correlation analysis is performed on the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device in each first time period to obtain the relative hysteresis time of heat dissipation of each temperature sensor.

3. A method for controlling temperature and humidity in a computer room according to claim 2, characterized in that: The specific method for obtaining the first time period includes: Record any temperature sensor as the target temperature sensor and any historical moment as the target moment; If the power fluctuation degree of the supervised device of the target temperature sensor at the target time is greater than or equal to a preset ratio of the maximum power fluctuation degree of the supervised device of the target temperature sensor, the target time is recorded as a first split point of the target temperature sensor; The time period between every two adjacent first dividing points is recorded as the first time period of the target temperature sensor.

4. A method for controlling temperature and humidity in a computer room according to claim 2, characterized in that: The hysteresis correlation analysis of the covariance relationship between the temperature data of each temperature sensor and the power of the monitored device in each first time period to obtain the relative heat dissipation lag time of each temperature sensor includes: Recording any first time period of the target temperature sensor as a target first time period of the target temperature sensor; Recording a sequence of temperature data of the target temperature sensor in the target first time period as a temperature data sequence of the target temperature sensor in the target first time period; Record the sequence consisting of the average power values ​​of all supervised devices of the target temperature sensor in the target first time period as the power sequence of the target temperature sensor in the target first time period; Initialize the offset number n=0, shift each sequence value in the temperature data sequence of the target temperature sensor in the target first time period backward by the offset number of bits, and obtain the nth offset temperature data sequence of the target temperature sensor in the target first time period; For the number of misalignments n+1, obtaining the n+1th misalignment temperature data sequence of the target temperature sensor in the target first time period; After obtaining the Pearson correlation coefficient between the power sequence of the target temperature sensor in the target first time period and each misaligned temperature data sequence of the target temperature sensor in the target first time period, the number of misalignments of the misaligned temperature data sequence with the largest Pearson correlation coefficient is recorded as the first lag time of the target temperature sensor in the target first time period; The first lag time of the target temperature sensor in all first time periods is recorded as the heat dissipation relative lag time of the target temperature sensor.

5. The method for controlling temperature and humidity in a computer room according to claim 1, wherein: The specific method of obtaining the heat release weight includes: Obtain the first-order difference of the power mean of all regulated devices of each temperature sensor at each historical moment; Obtain the first-order difference of the temperature data of each temperature sensor at each historical moment; The heat release weight of the regulated device of the temperature sensor is obtained, wherein the heat release weight is inversely proportional to the first-order difference of the power mean and is directly proportional to the first-order difference of the temperature data of each temperature sensor at a relative lag time of backward heat release at each historical moment.

6. A method for controlling temperature and humidity in a computer room according to claim 1, characterized in that: The specific method of obtaining the cooling response time includes: Obtain the relative cooling lag time of each temperature sensor; The current moment and the cooling relative lag time are added to obtain the cooling response moment of each temperature sensor at the current moment.

7. The method for controlling temperature and humidity in a computer room according to claim 1, wherein: The specific method for obtaining the device feedback time includes: The difference between the cooling response time of each temperature sensor at the current moment and the relative heat dissipation lag time of each temperature sensor at the current moment is recorded as the device feedback time of each temperature sensor at the current moment.

8. The method for controlling temperature and humidity in a computer room according to claim 1, wherein: The specific method for obtaining the temperature of each temperature sensor at the device feedback moment includes: The cumulative sum of the absolute values ​​of the predicted power increments of each regulated device of each temperature sensor from the current moment to the device feedback moment of the temperature sensor at the current moment is recorded as the thermal accumulation coefficient of the regulated device of each temperature sensor; The temperature increase of each temperature sensor at the device feedback moment is obtained; the temperature increase is directly proportional to the heat accumulation coefficient of the monitored device of each temperature sensor and the heat release weight of the monitored device of each temperature sensor.

9. A temperature and humidity control system in a machine room, characterized in that: The system includes: Data acquisition module: used to obtain temperature data of each temperature sensor in the computer room at each historical moment and the power and load data of the monitored equipment. Based on the power and load data of the monitored equipment, it predicts the predicted power of each monitored equipment obtained from each temperature sensor at each moment after the current moment. It also obtains the power of each control unit of the heat dissipation device at each historical moment. The heat dissipation device is used to absorb heat from the computer room. Each control unit corresponds to several temperature sensors. Device feedback moment temperature acquisition module: This module is used to perform hysteresis correlation analysis on the covariance between the temperature data of each temperature sensor and the power of the monitored device, obtain the relative heat dissipation lag time of each temperature sensor, and then obtain the heat release weight of the monitored device for each temperature sensor; The hysteresis correlation analysis is used to determine the covariance between the power of each control unit of the heat dissipation device and the temperature data of the corresponding temperature sensor, and the cooling response time of each temperature sensor at the current moment is obtained; The device feedback time of each temperature sensor at the current moment is obtained by combining the relative heat dissipation lag time and cooling response time of each temperature sensor; Analyze the thermal accumulation effect of the predicted power of the monitored device from the current time to the device feedback time of each temperature sensor, and combine the heat release weight of the monitored device of each temperature sensor to obtain the temperature of each temperature sensor at the device feedback time; Temperature control module: used to use the temperature of each temperature sensor at the time of device feedback to achieve temperature control at the current moment.

10. A temperature and humidity control device in a computer room, including a heat dissipation device, characterized in that: The heat dissipation device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the temperature and humidity control method in a computer room as described in any one of claims 1 to 8 are implemented.

Images