An AI model-based machine room air conditioner self-adaptive refrigeration adjustment method
By using an AI-based adaptive cooling regulation method and an LSTM prediction model to dynamically adjust air conditioning parameters, the problem of energy waste and equipment stability in traditional air conditioning regulation methods is solved, and efficient and stable control of the computer room environment is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 北京英沣特能源技术有限公司
- Filing Date
- 2026-01-21
- Publication Date
- 2026-06-02
AI Technical Summary
Traditional air conditioning cooling regulation methods lack dynamic adjustment capabilities and cannot flexibly adjust according to real-time environmental changes in the computer room, resulting in over- or under-cooling, which affects equipment stability and energy consumption.
An AI-based adaptive cooling regulation method is adopted, which uses an LSTM prediction model to analyze environmental data and dynamically adjust air conditioning parameters to achieve the lowest energy consumption and environmental stability, including initial and secondary parameter adjustments.
Reduce energy waste, improve the working efficiency and stability of electronic equipment in the computer room, maintain environmental stability by dynamically adjusting air conditioning parameters, and reduce the risk of equipment failure.
Smart Images
Figure CN121568372B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, and in particular to an adaptive cooling regulation method for computer room air conditioning based on an AI model. Background Technology
[0002] With the rapid development of information technology, the number and scale of various computer rooms, such as data centers and communication equipment rooms, have experienced explosive growth. As the core location for information storage, processing, and transmission, computer rooms house a large number of critical electronic devices such as servers and network equipment. These devices generate significant heat during operation. If this heat cannot be dissipated effectively and promptly, it can lead to overheating, affecting equipment performance and stability, and potentially causing equipment failures, resulting in serious consequences such as data loss and business interruption. Therefore, providing a stable and reliable cooling environment for computer rooms, ensuring that the electronic equipment inside operates within a suitable temperature range, has become a key factor in guaranteeing the normal and stable operation of computer rooms.
[0003] Traditional air conditioning cooling regulation methods typically use fixed parameter settings, meaning the air conditioning cooling system operates with fixed temperature, humidity, and other parameters. This fixed parameter setting method lacks dynamic adjustment capabilities and has significant limitations in practical applications. It cannot flexibly adjust according to changes in the real-time environment of the computer room, which may lead to over-cooling or under-cooling during certain operating periods. Over-cooling results in energy waste, while under-cooling causes uneven heating and cooling inside the computer room, affecting the operating efficiency and stability of the equipment.
[0004] Therefore, how to dynamically adjust the parameters of the computer room air conditioner according to the changes in real-time environmental parameters in the computer room, and improve the stability of the electronic equipment inside the computer room, has become an urgent problem to be solved. Summary of the Invention
[0005] In view of this, embodiments of the present invention provide an AI model-based adaptive cooling adjustment method for computer room air conditioning to solve the problem of how to dynamically adjust the parameter settings of computer room air conditioning according to changes in real-time environmental parameters in the computer room, thereby improving the stability of electronic equipment inside the computer room.
[0006] This invention provides an AI model-based adaptive cooling adjustment method for computer room air conditioning, which includes the following steps:
[0007] The environmental monitoring data of each moment in a preset historical period containing the current moment is obtained in the computer room to form an environmental monitoring data sequence. Based on the environmental monitoring data sequence, an LSTM prediction model is used to obtain an environmental prediction data sequence for a preset future period.
[0008] If any air conditioning parameter at the current moment meets the adjustment conditions, and the air conditioning parameter and the environmental monitoring data belong to the same data type, then based on the data fluctuation characteristics of the environmental prediction data sequence and the data difference between the environmental monitoring data at the current moment and the preset parameter value of the air conditioning parameter, the energy consumption required for the environmental monitoring data at the current moment to meet the standard under the preset parameter value is obtained.
[0009] The system obtains the energy consumption required to meet the environmental monitoring data at the current moment under each preset parameter value, filters the target energy consumption from all the required energy consumption, records the preset parameter value corresponding to the target energy consumption as the target parameter, adjusts the parameter value of any air conditioning parameter at the current moment to the target parameter, and obtains the real-time environmental monitoring data under the target parameter.
[0010] Based on the real-time environmental monitoring data, it is determined whether the target parameter meets the secondary adjustment conditions. If it does, the secondary parameter adjustment value is obtained based on the real-time environmental monitoring data, and the target parameter is adjusted in a secondary manner to complete the adaptive adjustment of the computer room air conditioner.
[0011] Preferably, the step of obtaining the energy consumption required for the environmental monitoring data to meet the preset parameter value at the current moment, based on the data fluctuation characteristics of the environmental prediction data sequence and the data difference between the current environmental monitoring data and the preset parameter value of any of the air conditioning parameters, includes:
[0012] The environmental prediction data sequence is fitted to obtain a first fitted line, and the slope of the first fitted line is obtained and denoted as the environmental change rate of the environmental prediction data sequence.
[0013] If the rate of change of the environment is greater than or equal to 0, then in the environmental monitoring data sequence, any environmental monitoring data with the same data value as the environmental monitoring data at the current moment is obtained and recorded as reference data. After the reference data, the first environmental monitoring data with the same data value as the preset parameter value is obtained and recorded as compliance reference data. The interval time between the reference data and the compliance reference data is obtained and recorded as the reference compliance time of the environmental monitoring data at the current moment.
[0014] If the rate of environmental change is less than 0, then the reference time to meet the standard is set to 0.
[0015] Calculate the sum of the environmental change rate and the constant 1, obtain the product of the reference compliance time and the sum, and get the time required for the environmental monitoring data at the current moment to meet the standard under the preset parameter value;
[0016] Based on the time required for the environmental monitoring data at the current moment to meet the standards under the preset parameter values, obtain the energy consumption required for the environmental monitoring data at the current moment to meet the standards under the preset parameter values.
[0017] Preferably, the step of obtaining the energy consumption required for the environmental monitoring data to meet the preset parameter values based on the time required to meet the standards under the current environmental monitoring data includes:
[0018] To obtain the cooling energy efficiency ratio of the air conditioner, based on the current environmental monitoring data and the preset parameter value, the cooling capacity of the air conditioner is obtained using the heat balance equation. The ratio of the cooling capacity to the cooling energy efficiency ratio is calculated to obtain the actual operating power of the air conditioner. The product of the actual operating power and the time required to meet the standard is obtained to obtain the energy consumption required to meet the standard under the preset parameter value based on the current environmental monitoring data.
[0019] Preferably, the step of screening the target energy consumption from all the energy consumption required to achieve the standard includes:
[0020] The minimum energy consumption required to achieve the target is obtained from all the energy consumption requirements, and is denoted as the target energy consumption.
[0021] Preferably, acquiring real-time environmental monitoring data under the target parameters includes:
[0022] Real-time environmental monitoring data after the current moment is acquired until the real-time environmental monitoring data is the same as the data value of the target parameter, thus obtaining the first real-time environmental monitoring data sequence under the target parameter. The time when the real-time environmental monitoring data with the same data value as the target parameter is located is recorded as the first target time.
[0023] Preferably, the step of determining whether the target parameter meets the secondary adjustment conditions based on the real-time environmental monitoring data includes:
[0024] The environmental change rate of the first real-time environmental monitoring data sequence is obtained and denoted as the first environmental change rate.
[0025] Based on the first real-time environmental monitoring data sequence, the environmental prediction data within a preset future time period after the first target time is obtained using the LSTM prediction model, forming a second environmental prediction data sequence, and the environmental change rate of the second environmental prediction data sequence is obtained and denoted as the second environmental change rate.
[0026] If the rate of change of the first environment is different from the rate of change of the second environment, then the target parameter is confirmed to meet the secondary adjustment conditions.
[0027] Preferably, obtaining the secondary parameter adjustment value based on the real-time environmental monitoring data includes:
[0028] If the first environmental change rate is greater than the second environmental change rate, then the difference between the first environmental change rate and the second environmental change rate is obtained to obtain the environmental change rate difference. The ratio of the environmental change rate difference to the second environmental change rate is rounded up to obtain the speed change amount. The difference between the parameter value of any air conditioning parameter at the current moment and the target parameter is obtained to obtain the first parameter adjustment value. The product of the speed change amount and the first parameter adjustment value is calculated to obtain the second parameter adjustment value.
[0029] If the rate of change of the first environment is less than or equal to the rate of change of the second environment, then the secondary parameter adjustment value is set to 0.
[0030] Preferably, the secondary adjustment of the target parameter includes:
[0031] The difference between the target parameter and the secondary parameter adjustment value is obtained to obtain the secondary target parameter, and the target parameter is reduced to the secondary target parameter.
[0032] Preferably, the adjustment conditions include:
[0033] The current environmental monitoring data is not within the preset standard environment range of the computer room, or at least one environmental prediction data in the environmental prediction data sequence is not within the preset standard environment range of the computer room.
[0034] Preferably, after the target parameter is adjusted a second time, the method further includes:
[0035] Real-time environmental monitoring data after the first target time is acquired until the real-time environmental monitoring data is the same as the data value of the secondary target parameter, and the second real-time environmental monitoring data sequence under the secondary target parameter is obtained. The time when the real-time environmental monitoring data with the same data value as the secondary target parameter is located is recorded as the second target time.
[0036] Based on the second real-time environmental monitoring data sequence, the LSTM prediction model is used to obtain environmental prediction data for a preset future period after the second target time, forming a third environmental prediction data sequence.
[0037] If the environmental prediction data in the third environmental prediction data sequence are all within the preset standard environment range of the computer room, then the secondary target parameter is adjusted to the maximum value within the preset standard environment range of the computer room.
[0038] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:
[0039] In this invention, the energy consumption required to achieve the target environmental monitoring data at any given moment under each preset parameter value of any air conditioning parameter is obtained. This is used to assess the energy consumption required to make the environmental monitoring data at the current moment reach the standard data required by the computer room under different preset parameter values. Thus, the preset parameter value corresponding to the minimum energy consumption is selected to adjust the air conditioning parameters at the current moment, so that the computer room air conditioning can complete the adjustment of the current environment with the lowest energy consumption, reducing unnecessary energy waste. When the target parameter meets the secondary adjustment conditions, the secondary parameter adjustment value is obtained, and the target parameter is adjusted a second time to reduce the deviation between the actual adjustment effect and the theoretical calculation result, so that the environmental data in the computer room is kept in a stable state, and the working efficiency and stability of the electronic equipment in the computer room are improved. Attached Figure Description
[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0041] Figure 1 This is a flowchart of an adaptive cooling adjustment method for computer room air conditioning based on an AI model, provided in Embodiment 1 of the present invention. Detailed Implementation
[0042] Embodiments of this disclosure are described in detail below, with examples of these embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting it.
[0043] It should be noted that the terms "first," "second," etc., used in this disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure.
[0044] To illustrate the technical solution of the present invention, specific embodiments are described below.
[0045] See Figure 1 This is a flowchart of a method for adaptive cooling adjustment of a computer room air conditioner based on an AI model, as provided in Embodiment 1 of the present invention. Figure 1 As shown, the method may include:
[0046] Step S101: Obtain environmental monitoring data for each moment in a preset historical time period containing the current moment in the computer room, form an environmental monitoring data sequence, and based on the environmental monitoring data sequence, use an LSTM prediction model to obtain an environmental prediction data sequence for a preset future time period.
[0047] In data centers, communication equipment rooms, and other types of computer rooms, a large number of critical electronic devices such as servers and network equipment are housed inside. These devices generate a lot of heat during operation, and their operation is unstable, meaning that the temperature inside the computer room may rise rapidly. Therefore, providing a stable and reliable cooling environment for the computer room to ensure that the electronic devices inside always operate within a suitable temperature range has become one of the key factors in ensuring the normal and stable operation of the computer room.
[0048] To maintain a stable environment inside the computer room, when various environmental data exceed the specified environmental parameters, it is necessary to adjust the computer room air conditioning parameters to restore the environmental data to the specified environment in a timely manner. The air conditioning parameters include temperature parameters, humidity parameters, etc. The adjustment method for each air conditioning parameter is the same. In this embodiment, the temperature parameter is used as an example to perform adaptive parameter adjustment analysis.
[0049] First, temperature monitoring data for each moment within a preset historical time period, including the current moment, is acquired in the computer room to form a temperature monitoring data sequence. In order to adjust the temperature in the computer room in advance and prevent sudden temperature increases, this embodiment uses an LSTM prediction model based on the temperature monitoring data sequence to obtain a temperature prediction data sequence for a preset future time period. Through the temperature prediction data sequence and the temperature monitoring data sequence, an adaptive adjustment method for air conditioning temperature parameters is analyzed.
[0050] Temperature monitoring data is acquired through temperature sensors. After acquiring the temperature monitoring data, it needs to be preprocessed to remove noise and outliers using the 3 sigma principle. In this embodiment, the preset historical time period is set to 24 hours, and the preset future time period is set to 1 hour after the current time. This is not a limitation and can be set according to the specific implementation scenario. The LSTM prediction model is used to obtain the temperature prediction data sequence within the preset future time period. Data preprocessing is an existing technology and will not be elaborated here. At the same time, as new data is continuously collected, the LSTM prediction model can be updated to improve the accuracy of the prediction.
[0051] Traditional air conditioning settings based on fixed parameters lack dynamic adjustment capabilities and cannot be flexibly adjusted in time when the temperature rises rapidly inside the computer room. This results in heat not being dissipated in a timely and effective manner, which in turn affects the performance and stability of the equipment and may even cause equipment failure, leading to serious consequences such as data loss and business interruption.
[0052] Therefore, in this embodiment, by analyzing temperature prediction data sequences and temperature monitoring data sequences, the minimum energy consumption required for the air conditioner in the computer room to meet the specified temperature requirements of the computer room under different temperature parameter settings is analyzed. Then, the temperature parameter data corresponding to the minimum energy consumption is obtained, and the temperature parameters of the air conditioner are adjusted for the first time. If the initial parameter adjustment does not meet the requirements, the temperature parameters of the air conditioner are adjusted for the second time. This reduces uneven heating and cooling and energy waste caused by unreasonable manual temperature parameter settings, improves the working efficiency and stability of electronic equipment in the computer room, and extends the service life of the air conditioning equipment.
[0053] Step S102: If any air conditioning parameter at the current moment meets the adjustment conditions, and the air conditioning parameter and the environmental monitoring data belong to the same data type, then based on the data fluctuation characteristics of the environmental prediction data sequence and the data difference between the environmental monitoring data at the current moment and the preset parameter value of the air conditioning parameter, the energy consumption required for the environmental monitoring data at the current moment to meet the standard under the preset parameter value is obtained.
[0054] When an air conditioner adjusts the temperature, the main energy consumption for cooling is compressor energy consumption (related to the cooling capacity). The greater the cooling capacity, the greater the energy consumption of the air conditioner compressor. Therefore, when the ambient temperature in the computer room differs significantly from the set temperature parameter (i.e., the cooling temperature) of the air conditioner, starting the cooling process can easily lead to excessive cooling capacity and excessive energy consumption of the air conditioner compressor. Air conditioner temperature data typically changes in 1-degree Celsius increments; therefore, in this embodiment, the unit of change for the air conditioner temperature data is 1 degree Celsius.
[0055] Therefore, when the temperature difference between the ambient temperature in the computer room and the temperature parameter set by the air conditioner is within 1 degree Celsius, the cooling capacity is minimal and the energy consumption of the air conditioner compressor is minimal. That is, the temperature parameter set by the air conditioner is 1 degree Celsius lower than the ambient temperature in the computer room each time. After the air conditioner starts cooling, the ambient temperature in the computer room will decrease with the cooling effect until it reaches the cooling temperature set by the air conditioner. At this time, the temperature parameter of the air conditioner can be lowered by another 1 degree Celsius until the ambient temperature in the computer room reaches the specified temperature of the computer room.
[0056] However, in real-world environments, the electronic equipment in a computer room is inherently unstable, meaning the temperature may rise rapidly. In such cases, the air conditioner needs to cool quickly, which involves setting the temperature parameters to a lower setting and increasing the cooling capacity to achieve rapid cooling. However, this increases the compressor's energy consumption. Therefore, when rapid cooling is required, the air conditioner's temperature parameters should first be adjusted to a low cooling setting, and then the cooling capacity should be gradually reduced as the ambient temperature decreases to achieve rapid cooling while minimizing energy consumption.
[0057] Because electronic equipment in the computer room continuously emits heat, causing the ambient temperature to rise (i.e., the ambient temperature in the computer room is fluctuating), the cooling time can be extended to accommodate the temperature rise caused by the continuous heat dissipation of electronic equipment. Since the energy consumption of the air conditioner compressor increases with the cooling capacity and operating time, in this embodiment, if the current temperature monitoring data is not within the preset standard temperature range of the computer room, or if at least one temperature prediction data point in the temperature prediction data sequence is not within the preset standard temperature range, the temperature parameters of the air conditioner in the computer room need to be adjusted. In this embodiment, based on the data fluctuation characteristics of the temperature prediction data sequence and the data difference between the current temperature monitoring data and each preset parameter value of the air conditioner temperature parameter, the energy consumption required to achieve the standard for the current temperature monitoring data at each preset parameter value is obtained, and then a suitable preset parameter value is selected to adjust the air conditioner temperature parameter at the current moment. The range of the preset parameter values for the air conditioner temperature parameter is from the lowest cooling temperature of the air conditioner to the current temperature monitoring data. The lowest cooling temperature of the air conditioner is typically 16 degrees Celsius, but this is not limited here and can be set according to the specific implementation scenario.
[0058] It should be noted that in this embodiment, the preset standard temperature range of the computer room is set to [Tmin, Tmax]. Since the "Data Center Design Specification" clearly states that the temperature range of Class A (fault-tolerant) and Class B (redundant) computer rooms is 23℃±1℃ (when powered on), and Class C (basic) is 18℃~27℃, no restrictions are imposed here, and the temperature can be set according to the specific implementation scenario.
[0059] The method for obtaining the energy consumption required to achieve the target temperature at any preset parameter value for the air conditioner is as follows:
[0060] (1) Obtain the time required for the temperature monitoring data at the current moment to reach the standard under any of the preset parameter values.
[0061] Specifically, the temperature prediction data sequence is fitted using the least squares method to obtain a first fitted line, and the slope of the first fitted line is obtained and denoted as the temperature change rate of the temperature prediction data sequence. The least squares method for data fitting is a prior art and will not be elaborated here.
[0062] If the rate of temperature change is greater than or equal to 0, then in the temperature monitoring data sequence, any temperature monitoring data with the same data value as the current temperature monitoring data is obtained and recorded as reference data. After the reference data, the first temperature monitoring data with the same data value as any preset parameter value is obtained and recorded as the standard reference data. The interval time between the reference data and the standard reference data is obtained and recorded as the reference standard time of the current temperature monitoring data.
[0063] If the rate of temperature change is less than 0, then the reference time to meet the standard is set to 0.
[0064] Calculate the sum of the temperature change rate and the constant 1, obtain the product of the reference compliance time and the sum, and get the time required for the current temperature monitoring data to meet the standard under any preset parameter value.
[0065] In one embodiment, taking the i-th preset parameter value of the air conditioning temperature parameter as an example, the formula for calculating the time required for the current temperature monitoring data to reach the standard under the i-th preset parameter value is as follows:
[0066]
[0067] in, The time required for the current temperature monitoring data to reach the target value at the i-th preset parameter value; The reference time to meet the standard is denoted as k; k represents the rate of temperature change in the temperature prediction data series.
[0068] It should be noted that when k is greater than 0, it indicates that the temperature forecast data shows an upward trend in the future. The larger k is, the faster the temperature forecast data will rise in the future, and in this case, more time will be needed to reduce the temperature in the computer room to the preset standard temperature range. The smaller the value of k, the better. When k is less than 0, it indicates that the temperature prediction data shows a downward trend in the future. In this case, there is no need to adjust the air conditioning cooling temperature to reduce the temperature in the computer room to the preset standard temperature range. Therefore, setting k is appropriate. .
[0069] (2) Based on the time required for the temperature monitoring data at the current moment to reach the standard under any preset parameter value, obtain the energy consumption required for the temperature monitoring data at the current moment to reach the standard under any preset parameter value.
[0070] Specifically, the cooling energy efficiency ratio of the air conditioner is obtained from the air conditioner's instruction manual. Based on the current temperature monitoring data and the preset parameter value, the cooling capacity of the air conditioner is obtained using the heat balance equation. Obtaining the cooling capacity of the air conditioner using the heat balance equation is existing technology and will not be elaborated here. The ratio of the cooling capacity to the cooling energy efficiency ratio is calculated to obtain the actual operating power of the air conditioner. The product of the actual operating power and the time required to reach the standard is obtained to obtain the energy consumption required to reach the standard under any preset parameter value based on the current temperature monitoring data.
[0071] In one embodiment, taking the i-th preset parameter value of the air conditioning temperature parameter as an example, the formula for calculating the energy consumption required to achieve the standard under the i-th preset parameter value of the current temperature monitoring data is as follows:
[0072]
[0073] in, The energy consumption required to achieve the target for the current temperature monitoring data under the i-th preset parameter value; Based on the current temperature monitoring data and the i-th preset parameter value, the cooling capacity of the air conditioner is obtained using the heat balance equation; EER is the cooling energy efficiency ratio of the air conditioner. The time required for the current temperature monitoring data to reach the standard under the i-th preset parameter value; the energy consumption calculation formula for air conditioners is existing technology and will not be elaborated here.
[0074] Thus, the energy consumption required to achieve the target temperature at any of the preset parameter values is obtained from the current temperature monitoring data.
[0075] Step S103: Obtain the energy consumption required to meet the environmental monitoring data at the current moment under each preset parameter value, filter the target energy consumption from all the required energy consumption, record the preset parameter value corresponding to the target energy consumption as the target parameter, adjust the parameter value of any air conditioning parameter at the current moment to the target parameter, and obtain the real-time environmental monitoring data under the target parameter.
[0076] Furthermore, following the method described above for obtaining the energy consumption required to achieve the target for the current temperature monitoring data at the i-th preset parameter value, the energy consumption required to achieve the target for the current temperature monitoring data at each preset parameter value is obtained. Among all the energy consumption required to achieve the target, the minimum energy consumption required to achieve the target is obtained and recorded as the target energy consumption. The parameter value of the air conditioning temperature parameter at the current time is then reduced to the target parameter.
[0077] For example, suppose the time required to reach the standard for the current temperature monitoring data at the first preset parameter value (the first preset parameter value is 1 degree Celsius lower than the current temperature monitoring data) is 30 seconds, and the energy consumption required to reach the standard is E1. The time required to reach the standard for the current temperature monitoring data at the second preset parameter value (the second preset parameter value is 2 degrees Celsius lower than the current temperature monitoring data) is 20 seconds, and the energy consumption required to reach the standard is E2. The time required to reach the standard for the current temperature monitoring data at the third preset parameter value (the third preset parameter value is 3 degrees Celsius lower than the current temperature monitoring data) is 10 seconds, and the energy consumption required to reach the standard is E3. Among these, E2 is the smallest. Therefore, the parameter value of the current air conditioning temperature parameter will be reduced by 2 degrees Celsius.
[0078] After adjusting the current air conditioning temperature parameter to the target parameter, real-time temperature monitoring data after the current time is acquired until the real-time temperature monitoring data is the same as the target parameter data value, thus obtaining the first real-time temperature monitoring data sequence under the target parameter. The time when the real-time ambient temperature data with the same target parameter data value is recorded as the first target time (assuming the current time is t1, after adjusting the current air conditioning parameter to the target parameter T1, the temperature monitoring data in the computer room at time t2 decreases to T1, at this time t2 is the first target time, and T1 is the real-time temperature monitoring data of the first target time).
[0079] At this point, the parameter values of the air conditioning temperature parameters at the current moment have been adjusted, and real-time environmental monitoring data under the target parameters has been obtained.
[0080] Step S104: Determine whether the target parameter meets the secondary adjustment conditions based on the real-time environmental monitoring data. If it does, obtain the secondary parameter adjustment value based on the real-time environmental monitoring data, and perform secondary adjustment on the target parameter to complete the adaptive adjustment of the computer room air conditioner.
[0081] When adjusting air conditioning temperature parameters based on temperature prediction data, the actual adjustment effect in the computer room may differ from the theoretical calculation value during actual operation. This could result in the air conditioning cooling effect being too low or too high. Therefore, it is necessary to determine whether the target parameters meet the conditions for secondary adjustment based on real-time environmental monitoring data under the target parameters.
[0082] The method for determining whether the target parameters meet the secondary adjustment conditions based on real-time environmental monitoring data under the target parameters is as follows:
[0083] Using the above-mentioned method for obtaining the temperature change rate of the temperature prediction data sequence, the temperature change rate of the first real-time temperature monitoring data sequence is obtained and denoted as the first temperature change rate.
[0084] Based on the first real-time temperature monitoring data sequence, the temperature prediction data within a preset future time period after the first target time is obtained using the LSTM prediction model, forming a second temperature prediction data sequence. The temperature change rate of the second temperature prediction data sequence is obtained using the above-mentioned method for obtaining the temperature change rate of the temperature prediction data sequence, and is denoted as the second temperature change rate.
[0085] If the first temperature change rate is different from the second temperature change rate, it indicates that the actual adjustment effect of the temperature data in the computer room during actual operation may be different from the theoretical calculation value. In this case, the air conditioning cooling effect may be too low or too high. Therefore, it is confirmed that the target parameter meets the secondary adjustment condition, that is, the parameter value of the target parameter, i.e., the air conditioning temperature parameter at the first target time, needs to be adjusted.
[0086] The method for adjusting the target parameters is as follows:
[0087] (1) Obtain the secondary parameter adjustment value based on the real-time environmental monitoring data under the target parameters.
[0088] Specifically, if the first temperature change rate is greater than the second temperature change rate, the difference between the first temperature change rate and the second temperature change rate is obtained to obtain the temperature change rate difference. The ratio of the temperature change rate difference to the second temperature change rate is rounded up to obtain the rate change amount. The difference between the parameter value of any air conditioning parameter at the current moment and the target parameter is obtained to obtain the first parameter adjustment value. The product of the rate change amount and the first parameter adjustment value is calculated to obtain the second parameter adjustment value.
[0089] If the rate of change of the first temperature is less than or equal to the rate of change of the second temperature, then the secondary parameter adjustment value is set to 0.
[0090] In one embodiment, the formula for calculating the secondary parameter adjustment value is:
[0091]
[0092] in, This is a secondary parameter adjustment value; The first rate of temperature change; This is the second rate of temperature change; This is a parameter adjustment value, that is, the difference between the current air conditioning temperature parameter value and the target parameter. The rounding up symbol.
[0093] It should be noted that, if If the target parameter's cooling effect is poor, it indicates that the air conditioning temperature needs to be lowered again and the cooling capacity increased for further cooling. This indicates that the target parameter has a good cooling effect and there is no need to adjust the air conditioner cooling temperature again.
[0094] (2) Adjust the target parameters a second time based on the second parameter adjustment value.
[0095] Specifically, the difference between the target parameter and the adjusted value of the secondary parameter is obtained to get the secondary target parameter, and the target parameter is reduced to the secondary target parameter. For example, assuming t2 is the first target time and T1 is the target parameter, the calculation... The secondary target parameter T2 is obtained, and the target parameter (the parameter value of the air conditioning temperature parameter at the first target time) T1 is reduced to T2, thus completing the secondary adjustment of the target parameter.
[0096] This completes one adaptive cooling adjustment of the computer room air conditioning system.
[0097] Furthermore, considering that the air conditioning in the computer room will increase the energy consumption when it is kept at a low cooling temperature for a long time, after the target parameter is adjusted a second time, if the real-time temperature monitoring data after the first target time tends to be stable, the parameter value of the air conditioning temperature parameter can be adjusted again to reduce the energy consumption of the air conditioning.
[0098] Specifically, real-time temperature monitoring data after the first target time is acquired until the real-time temperature monitoring data is the same as the data value of the secondary target parameter, and a second real-time temperature monitoring data sequence under the secondary target parameter is obtained. The time when the real-time environmental monitoring data with the same data value as the secondary target parameter is located is recorded as the second target time.
[0099] Based on the second real-time temperature monitoring data sequence, the LSTM prediction model is used to obtain the temperature prediction data for a preset future period after the second target time, forming a third temperature prediction data sequence.
[0100] If the temperature prediction data in the third temperature prediction data sequence are all within the preset standard temperature range of the computer room, the secondary target parameter is adjusted to the maximum value within the preset standard temperature range of the computer room.
[0101] For example, assuming the secondary target parameter is T2, the temperature in the computer room reaches T2 at the second target time t3 after the first target time t2. The temperature prediction data for the future preset time period after time t3 is obtained to form a third temperature prediction data sequence. If the temperature prediction data in the third temperature prediction data sequence are all within the preset standard temperature range of the computer room [Tmin, Tmax], then the parameter value of the air conditioning temperature parameter is adjusted to Tmax at time t3.
[0102] In summary, in this embodiment of the invention, the energy consumption required to achieve the target environmental monitoring data at any given time under each preset parameter value of any air conditioning parameter is obtained. This is used to evaluate the energy consumption required to make the environmental monitoring data at the current time reach the standard data required by the computer room under different preset parameter values. The preset parameter value corresponding to the minimum energy consumption is then selected to adjust the air conditioning parameters at the current time, enabling the computer room air conditioning to complete the adjustment of the current environment with the lowest energy consumption, reducing unnecessary energy waste. When the target parameter meets the conditions for secondary adjustment, the secondary parameter adjustment value is obtained, and the target parameter is adjusted a second time to reduce the deviation between the actual adjustment effect and the theoretical calculation result, keeping the environmental data in the computer room in a stable state and improving the working efficiency and stability of the electronic equipment in the computer room.
[0103] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.
Claims
1. A method for adaptive cooling regulation of computer room air conditioning based on an AI model, characterized in that, The AI-based adaptive cooling adjustment method for computer room air conditioning includes: The environmental monitoring data of each moment in a preset historical period containing the current moment is obtained in the computer room to form an environmental monitoring data sequence. Based on the environmental monitoring data sequence, an LSTM prediction model is used to obtain an environmental prediction data sequence for a preset future period. If any air conditioning parameter at the current moment meets the adjustment conditions, and the air conditioning parameter and the environmental monitoring data belong to the same data type, then based on the data fluctuation characteristics of the environmental prediction data sequence and the data difference between the environmental monitoring data at the current moment and the preset parameter value of the air conditioning parameter, the energy consumption required for the environmental monitoring data at the current moment to meet the standard under the preset parameter value is obtained. The system obtains the energy consumption required to meet the environmental monitoring data at the current moment under each preset parameter value, filters the target energy consumption from all the required energy consumption, records the preset parameter value corresponding to the target energy consumption as the target parameter, adjusts the parameter value of any air conditioning parameter at the current moment to the target parameter, and obtains the real-time environmental monitoring data under the target parameter. Based on the real-time environmental monitoring data, it is determined whether the target parameter meets the secondary adjustment conditions. If it does, the secondary parameter adjustment value is obtained based on the real-time environmental monitoring data, and the target parameter is adjusted in a secondary manner to complete the adaptive adjustment of the computer room air conditioner. The step of obtaining the energy consumption required for the environmental monitoring data to meet the preset parameter value at the current moment, based on the data fluctuation characteristics of the environmental prediction data sequence and the data difference between the current environmental monitoring data and the preset parameter value of any air conditioning parameter, includes: The environmental prediction data sequence is fitted to obtain a first fitted line, and the slope of the first fitted line is obtained and denoted as the environmental change rate of the environmental prediction data sequence. If the rate of change of the environment is greater than or equal to 0, then in the environmental monitoring data sequence, any environmental monitoring data with the same data value as the environmental monitoring data at the current moment is obtained and recorded as reference data. After the reference data, the first environmental monitoring data with the same data value as the preset parameter value is obtained and recorded as compliance reference data. The interval time between the reference data and the compliance reference data is obtained and recorded as the reference compliance time of the environmental monitoring data at the current moment. If the rate of environmental change is less than 0, then the reference time to meet the standard is set to 0. Calculate the sum of the environmental change rate and the constant 1, obtain the product of the reference compliance time and the sum, and get the time required for the environmental monitoring data at the current moment to meet the standard under the preset parameter value; Based on the time required for the environmental monitoring data at the current moment to meet the standards under the preset parameter values, obtain the energy consumption required for the environmental monitoring data at the current moment to meet the standards under the preset parameter values.
2. The adaptive cooling adjustment method for computer room air conditioning based on an AI model according to claim 1, characterized in that, The step of obtaining the energy consumption required for the environmental monitoring data to meet the preset parameter values based on the time required to meet the standards at the current moment, according to the environmental monitoring data at the current moment, includes: To obtain the cooling energy efficiency ratio of the air conditioner, based on the current environmental monitoring data and the preset parameter value, the cooling capacity of the air conditioner is obtained using the heat balance equation. The ratio of the cooling capacity to the cooling energy efficiency ratio is calculated to obtain the actual operating power of the air conditioner. The product of the actual operating power and the time required to meet the standard is obtained to obtain the energy consumption required to meet the standard under the preset parameter value based on the current environmental monitoring data.
3. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 1, characterized in that, The process of selecting target energy consumption from all energy consumption requirements for achieving the standard includes: The minimum energy consumption required to achieve the target is obtained from all the energy consumption requirements, and is denoted as the target energy consumption.
4. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 1, characterized in that, The acquisition of real-time environmental monitoring data under the target parameters includes: Real-time environmental monitoring data after the current moment is acquired until the real-time environmental monitoring data is the same as the data value of the target parameter, thus obtaining the first real-time environmental monitoring data sequence under the target parameter. The time when the real-time environmental monitoring data with the same data value as the target parameter is located is recorded as the first target time.
5. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 4, characterized in that, The step of determining whether the target parameter meets the secondary adjustment conditions based on the real-time environmental monitoring data includes: The environmental change rate of the first real-time environmental monitoring data sequence is obtained and denoted as the first environmental change rate. Based on the first real-time environmental monitoring data sequence, the environmental prediction data within a preset future time period after the first target time is obtained using the LSTM prediction model, forming a second environmental prediction data sequence, and the environmental change rate of the second environmental prediction data sequence is obtained and denoted as the second environmental change rate. If the rate of change of the first environment is different from the rate of change of the second environment, then the target parameter is confirmed to meet the secondary adjustment conditions.
6. The adaptive cooling adjustment method for computer room air conditioning based on an AI model according to claim 5, characterized in that, The step of obtaining secondary parameter adjustment values based on the real-time environmental monitoring data includes: If the first environmental change rate is greater than the second environmental change rate, then the difference between the first environmental change rate and the second environmental change rate is obtained to obtain the environmental change rate difference. The ratio of the environmental change rate difference to the second environmental change rate is rounded up to obtain the speed change amount. The difference between the parameter value of any air conditioning parameter at the current moment and the target parameter is obtained to obtain the first parameter adjustment value. The product of the speed change amount and the first parameter adjustment value is calculated to obtain the second parameter adjustment value. If the rate of change of the first environment is less than or equal to the rate of change of the second environment, then the secondary parameter adjustment value is set to 0.
7. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 4, characterized in that, The secondary adjustment of the target parameter includes: The difference between the target parameter and the secondary parameter adjustment value is obtained to obtain the secondary target parameter, and the target parameter is reduced to the secondary target parameter.
8. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 1, characterized in that, The adjustment conditions include: The current environmental monitoring data is not within the preset standard environment range of the computer room, or at least one environmental prediction data in the environmental prediction data sequence is not within the preset standard environment range of the computer room.
9. The adaptive cooling regulation method for computer room air conditioning based on an AI model according to claim 7, characterized in that, After the target parameters are adjusted a second time, the method further includes: Real-time environmental monitoring data after the first target time is acquired until the real-time environmental monitoring data is the same as the data value of the secondary target parameter, and the second real-time environmental monitoring data sequence under the secondary target parameter is obtained. The time when the real-time environmental monitoring data with the same data value as the secondary target parameter is located is recorded as the second target time. Based on the second real-time environmental monitoring data sequence, the LSTM prediction model is used to obtain environmental prediction data for a preset future period after the second target time, forming a third environmental prediction data sequence. If the environmental prediction data in the third environmental prediction data sequence are all within the preset standard environment range of the computer room, then the secondary target parameter is adjusted to the maximum value within the preset standard environment range of the computer room.