Machine room energy-saving control method and system

Abstract

The invention relates to the technical field of air conditioning control, in particular to a machine room energy-saving control method and system, and the method comprises the steps: dynamically adjusting the starting number of servers according to the actual load of the servers in a machine room; respectively adjusting the power of the air supply system of each area according to the server starting number and the indoor and outdoor temperature difference; predicting a future power curve of the air supply system through the collected historical server operation data and the power data of the corresponding air supply system; adjusting an adjusting strategy of the air supply system according to the power curve. Therefore, continuous improvement and a better energy-saving effect are achieved. In conclusion, the method not only pays attention to immediate response, but also emphasizes long-term optimization, and aims at remarkably reducing the operating cost of the machine room through an intelligent means.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning control, and in particular to a method and system for energy-saving control of a computer room. Background Art

[0002] The computer room is a key component of modern information technology infrastructure. With the continuous expansion of data center scale and the increase of operating load, energy consumption and heat dissipation issues have become important challenges. Energy saving in computer rooms not only helps to reduce operating costs, but also reduces the impact on the environment.

[0003] The existing energy-saving control of computer rooms generally controls the air conditioning system as a whole to adjust parameters such as the temperature in the computer room. This solution cannot be flexibly adjusted in real time according to usage, thereby increasing losses. Summary of the invention

[0004] The purpose of the present invention is to provide a computer room energy-saving control method and system, aiming to focus on immediate response and achieve significant reduction in computer room operating costs through intelligent means.

[0005] To achieve the above object, the present invention provides a method for energy-saving control of a computer room, comprising dynamically adjusting the number of servers turned on according to the actual load of the servers in the computer room;

[0006] Adjust the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor areas;

[0007] Predict the future power curve of the air supply system by collecting historical server operation data and corresponding air supply system power data;

[0008] Adjust the air supply system adjustment strategy according to the power curve.

[0009] Wherein, after adjusting the adjustment strategy of the air supply system according to the power curve, the method further includes: setting energy-saving strategy parameters, correcting the adjustment strategy, and displaying the working status.

[0010] The specific steps of dynamically adjusting the number of enabled servers according to the actual load of the servers in the computer room include:

[0011] Collect real-time load data of the server;

[0012] Set the upper and lower load thresholds;

[0013] When the real-time load data exceeds the upper threshold, the server is started; when the real-time load data is less than the lower threshold, the server is shut down.

[0014] The real-time load data of the server includes CPU usage, memory usage and disk read and write status.

[0015] The specific steps of adjusting the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor areas include:

[0016] Get indoor and outdoor temperature data and calculate the temperature difference;

[0017] Calculate the power adjustment level based on the temperature difference;

[0018] Calculate the required air volume based on the number of servers and power adjustment levels in each area;

[0019] Use control algorithms to adjust the air supply system power according to the required air volume and control it by zones;

[0020] Document the tuning process and evaluate the temperature regulation effect of the tuning strategy.

[0021] The specific steps of calculating the power adjustment level according to the temperature difference include:

[0022] Define the temperature range;

[0023] Allocate power adjustment levels according to temperature difference ranges.

[0024] The specific steps of calculating the required air volume in combination with the number of servers and the power adjustment level in each area include:

[0025] Count the number of servers currently open in each region;

[0026] Get the heat generated by each server when running at full load and adjust the heat according to the current load;

[0027] Add up the heat generated by each server in the area to get the total heat generated by the area;

[0028] Get the number of air changes required per hour;

[0029] Calculates air volume based on heat output and air exchange rate.

[0030] The specific steps of predicting the future power curve of the air supply system by using the collected historical server operation data and the corresponding power data of the air supply system include:

[0031] Collect historical server operation data and corresponding air supply system power data;

[0032] Clean and organize the collected data;

[0033] Extract identification features from raw data;

[0034] Use historical data to train a linear regression model based on the identified features;

[0035] Use the test set to evaluate the generalization ability of the linear regression model;

[0036] The trained model is used to predict future time points to obtain the power curve of the future air supply system.

[0037] In the second aspect, the present invention also provides an energy-saving control system for a computer room, including a server control module, a power calculation module, a curve prediction module and an adjustment module; the server control module is used to dynamically adjust the number of servers turned on according to the actual load of the servers in the computer room; the power calculation module is used to adjust the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor; the curve prediction module is used to predict the future power curve of the air supply system through the collected historical server operation data and the corresponding power data of the air supply system; the adjustment module is used to adjust the adjustment strategy of the air supply system according to the power curve.

[0038] A computer room energy-saving control method and system of the present invention dynamically adjusts the number of servers powered on according to the actual workload of the servers in the computer room. When it is detected that some parts of the server cluster have not reached their maximum performance utilization, the system can automatically reduce the number of these parts powered on, thereby saving power consumption. This dynamic adjustment not only relies on real-time monitoring, but also requires a set of efficient algorithms to determine when to increase or decrease the number of servers, and which servers to increase or decrease, so as to ensure that the service quality is not affected. Based on the difference between the adjusted number of server startups and the ambient temperature inside and outside the computer room, the operating power of the air supply system in each area is further adjusted. This process involves precise temperature management and airflow control to ensure that the cold is effectively delivered to the location where cooling is required, while avoiding energy waste caused by excessive cooling. In addition, it is also necessary to consider the changes in the distribution of heat sources inside the computer room in order to flexibly adjust the air supply strategy. Using the historically accumulated server operation data (such as CPU usage, memory occupancy, etc.) and the corresponding air supply system power data, machine learning or other prediction technologies are used to establish a model to predict the power curve required by the air supply system in the future. This step is critical for planning cooling resources in advance, helping managers prepare before peak demand arrives, while also reducing unnecessary energy consumption in a timely manner when demand decreases. Based on the predicted power curve, a more sophisticated air supply system adjustment strategy can be developed. These strategies may include, but are not limited to, changing the fan speed, adjusting the damper opening, or reconfiguring the air flow path. Throughout the process, it is important to maintain continuous monitoring of the actual operating conditions, and to promptly revise the prediction model and adjust the plan according to the actual situation, so as to achieve continuous improvement and better energy saving effects. In summary, this approach not only focuses on immediate response, but also emphasizes long-term optimization, aiming to significantly reduce the operating costs of the computer room through intelligent means. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0040] Figure 1 It is a flow chart of a computer room energy-saving control method of the present invention.

[0041] Figure 2 It is a flow chart of the present invention for dynamically adjusting the number of servers turned on according to the actual load of the servers in the computer room.

[0042] Figure 3This is a flow chart of the present invention for adjusting the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor.

[0043] Figure 4 It is a flow chart of calculating the power adjustment level according to the temperature difference of the present invention.

[0044] Figure 5 It is a flow chart of the present invention for calculating the required air volume in combination with the number of servers in each area and the power adjustment level.

[0045] Figure 6 It is a flow chart of the present invention for predicting the future power curve of the air supply system by collecting historical server operation data and corresponding power data of the air supply system.

[0046] Figure 7 It is a flow chart of the present invention for cleaning and arranging the collected data. DETAILED DESCRIPTION

[0047] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0048] First embodiment

[0049] See also Figure 1 to Figure 7 The present invention provides a computer room energy-saving control method, comprising:

[0050] S101 dynamically adjusts the number of servers enabled according to the actual load of the servers in the computer room;

[0051] The specific steps include:

[0052] S201 collects real-time load data of the server;

[0053] The real-time load data of the server includes CPU usage, memory usage, and disk read and write status.

[0054] Use various monitoring tools or API interfaces to obtain real-time server load information. This information usually includes but is not limited to: CPU usage (percentage); memory usage (ratio of used memory to total memory); disk read and write rate (amount of data read or written per second); network traffic (upload / download speed); current number of active connections, etc.

[0055] S202 sets a load upper threshold and a load lower threshold;

[0056] Set reasonable load thresholds based on historical data and business needs. These thresholds should reflect the maximum load that the server can withstand under normal working conditions and the level of resource savings that can be achieved under lower load conditions. For example:

[0057] The CPU usage reaches 70% as the upper threshold; the memory usage reaches 80% as the upper threshold; the disk read and write rate exceeds 10MB / s as the upper threshold; the CPU usage is less than 30% as the lower threshold; the memory usage is less than 40% as the lower threshold; the disk read and write rate is less than 1MB / s as the lower threshold

[0058] These thresholds can be adjusted based on actual conditions to ensure that the server can provide sufficient resources under high load conditions and save power and other resources under low load conditions.

[0059] S203: When the real-time load data exceeds the upper threshold, the server is started; when the real-time load data is less than the lower threshold, the server is shut down.

[0060] When the load data exceeds the set upper threshold, the system should automatically start additional servers to distribute the load. Conversely, if the load data continues to be below the set lower threshold, some servers can be safely shut down to reduce energy consumption and costs.

[0061] When implementing this strategy, you should also consider the following points:

[0062] When starting or shutting down a server, ensure a smooth transition to avoid impacting running applications. Before starting a server, check if there are servers in standby mode that can be quickly activated to reduce the time it takes to start a new server. Before shutting down a server, migrate or terminate the tasks and services running on it. The effectiveness of thresholds should be evaluated regularly and adjusted based on actual conditions.

[0063] In this way, the number of servers can be dynamically adjusted according to the actual load conditions, thereby effectively managing the resource utilization of the data center while also reducing costs and energy consumption.

[0064] S102 adjusts the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor areas;

[0065] The specific steps include:

[0066] S301 obtains indoor and outdoor temperature data and calculates the temperature difference;

[0067] In this step, real-time temperature data needs to be collected through temperature sensors installed indoors and outdoors. This data can be obtained through Internet of Things (IoT) devices or dedicated environmental monitoring systems. The data obtained should include outdoor temperature and the temperature of each indoor area.

[0068] Next, calculate the temperature difference between the outdoor and indoor areas for each zone:

[0069] Temperature difference = outdoor temperature - indoor temperature.

[0070] S302 calculates the power adjustment level according to the temperature difference;

[0071] The specific steps include:

[0072] S401 defines the temperature difference range;

[0073] Based on historical data and experience, define a series of temperature difference intervals, for example: the temperature difference is less than or equal to 2℃

[0074] , the temperature difference is between 2℃ and 5℃, and the temperature difference is greater than 5℃.

[0075] S402 allocates power adjustment levels according to the temperature difference intervals.

[0076] Assign a power adjustment level to each temperature range. The power adjustment level can be expressed as a percentage increase or decrease in the power of the air supply system. For example:

[0077] The temperature difference is less than or equal to 2℃, corresponding to the power adjustment level +10%

[0078] The temperature difference between 2℃ and 5℃ corresponds to the power adjustment level +20%

[0079] The temperature difference is greater than 5℃, corresponding to the power adjustment level +30%

[0080] S303 calculates the required air volume based on the number of servers and the power adjustment level in each area;

[0081] The specific steps include:

[0082] S501 counts the number of currently enabled servers in each region;

[0083] The number of servers currently enabled in each region is obtained through the monitoring system. This data can be obtained through the server management system or dedicated software.

[0084] S502 obtains the heat generated by each server when it is running at full load and adjusts the heat generated according to the current load condition;

[0085] Get the heat output of the server at full load (usually in watts W) according to the server model. Then, adjust the heat output according to the current server load (such as CPU usage). For example, if the full load heat output of a server is 200W, its heat output under the current load may be 50% of the full load heat output, that is, 100W.

[0086] S503: summing up the heat generated by each server in the area to obtain the total heat generated by the area;

[0087] Add the adjusted heat values ​​of all servers in the region to get the total heat value of the region.

[0088] S504 obtains the number of ventilation changes required per hour;

[0089] Determine the number of air changes required per hour based on the data center's standards or specific requirements. This number is usually fixed, such as 10 air changes per hour.

[0090] S505 calculates the air volume based on the heat value and the ventilation frequency.

[0091] Calculate the required air volume using the total heat output and ventilation times. The air volume calculation formula is as follows:

[0092] Required air volume = total heat generation / number of air changes per hour × air heat capacity ratio

[0093] Among them, the air heat capacity ratio is a constant, approximately 1.01kJ / (kg·K).

[0094] S304 uses a control algorithm to adjust the power of the air supply system according to the required air volume and performs zone control;

[0095] Use a pre-set control algorithm (such as a PID controller) to adjust the power of the air supply system to meet the calculated required air volume. At the same time, adjust the power of the air supply system according to different areas to ensure that the temperature of each area remains within the ideal range.

[0096] S305 records the adjustment process and evaluates the temperature regulation effect of the adjustment strategy.

[0097] Record the actions and results of each adjustment, and regularly evaluate the effect of the adjustment strategy to determine whether the algorithm or parameters need to be optimized. This can be done by analyzing temperature change trends and changes in server load to ensure system stability and energy efficiency.

[0098] S103 predicts the future power curve of the air supply system by using the collected historical server operation data and the corresponding power data of the air supply system;

[0099] The specific steps include:

[0100] S601 collects historical server operation data and corresponding air supply system power data;

[0101] In this step, historical data needs to be collected, including but not limited to:

[0102] The operating status of the server (such as CPU utilization, memory utilization, disk I / O, etc.), the power consumption of the air supply system, environmental parameters such as ambient temperature and humidity, and other variables that may affect the power of the air supply system (such as external temperature, internal layout of the data center, etc.).

[0103] S602 cleans and organizes the collected data;

[0104] The specific steps include:

[0105] S701 obtains server operation data and air supply system power data;

[0106] Obtain historical data from data sources, which may involve exporting data from a database, making API calls, or parsing log files.

[0107] S702 finds missing values ​​in the data;

[0108] By checking each data record, find any missing fields or values. This can be automated using programming languages ​​such as Python's Pandas library.

[0109] S703 deletes records containing missing values;

[0110] For those records with missing values, decisions can be made to remove these records, especially when the missing values ​​are large or cannot be reasonably filled.

[0111] S704 finds outliers in the data and deletes them.

[0112] Identify outliers through statistical methods (such as Z-score, IQR, etc.) or visualization methods (such as box plots) and decide whether to remove them.

[0113] S603 extracts identification features from the original data;

[0114] In this step, key features that are helpful for prediction need to be extracted from the raw data, such as time series features (such as date, time, day of the week, etc.), server load features (such as average CPU usage, peak CPU usage, etc.), and environmental condition features (such as indoor and outdoor temperature difference, humidity, etc.).

[0115] S604 uses historical data to train a linear regression model based on the identification features;

[0116] Based on the extracted features, a linear regression model is trained using historical data. The linear regression model can be used to establish the relationship between the air supply system power and related features. When training the model, the data needs to be divided into a training set and a validation set to ensure that the model does not overfit.

[0117] S605 uses the test set to evaluate the generalization ability of the linear regression model;

[0118] Use test set data that has not been used in training to evaluate the generalization ability of the model. The performance of the model can be measured by calculating the error between the predicted value and the actual value (such as mean square error MSE, mean absolute error MAE, etc.).

[0119] S606 uses the trained model to predict future time points to obtain a power curve of the future air supply system.

[0120] Once the model is trained and validated by the test set, it can be used to predict the power of the air supply system in a certain time period in the future. Specifically, the characteristic data of the future time period (such as the expected server load, expected environmental conditions, etc.) can be input, and the model will output the corresponding power prediction value, thus forming a predicted power curve.

[0121] S104 adjusts the adjustment strategy of the air supply system according to the power curve.

[0122] According to the real-time monitored power curve changes, the operating parameters of the air supply system, such as fan speed and guide vane opening, are dynamically adjusted to achieve the goals of optimizing energy efficiency, ensuring stable system operation, and improving indoor air quality. This adjustment strategy can not only effectively respond to load changes, but also reduce unnecessary energy waste and improve the overall energy efficiency ratio.

[0123] S105 sets energy-saving strategy parameters, modifies the adjustment strategy, and displays the working status.

[0124] Specify energy-saving goals, such as reducing energy consumption by a percentage or achieving a specific energy efficiency ratio.

[0125] Record the actual power curve data of the air supply system under different working conditions, including but not limited to environmental parameters such as temperature, humidity, air flow rate, and electrical parameters such as the working current and voltage of the equipment.

[0126] Use data analysis tools to analyze existing power curves to identify peak energy consumption periods and potential energy savings.

[0127] According to the analysis results, new energy-saving parameters are set, such as adjusting the threshold of fan speed, changing the control logic of guide vane angle, etc.

[0128] Based on the above parameters, a new control logic or algorithm is formulated to guide the air supply system to automatically adjust its working state under different conditions.

[0129] In actual operation, the system performance indicators are continuously monitored, and the adjustment strategy is corrected based on the actual effect feedback, such as fine-tuning the response speed or sensitivity through the PID controller.

[0130] Apply the new adjustment strategy to the actual air supply system and conduct a test run for a period of time to observe its impact on energy saving.

[0131] Use the monitoring interface or dashboard to display the current working status of the system, including but not limited to power consumption, environmental parameters, equipment health status, etc. Update the displayed data in real time so that operators can understand the system operation status in a timely manner and make necessary intervention measures. Review the energy-saving effect regularly and adjust the energy-saving strategy parameters as needed to ensure the long-term stability and efficiency of the system.

[0132] Through the above steps, the energy-saving strategy parameters can be effectively set, and the adjustment strategy of the air supply system can be continuously optimized, while ensuring that the operator can grasp the working status of the system at any time.

[0133] Second embodiment

[0134] The present invention also provides an energy-saving control system for a computer room, comprising a server control module, a power calculation module, a curve prediction module and an adjustment module; the server control module is used to dynamically adjust the number of servers turned on according to the actual load of the servers in the computer room; the power calculation module is used to adjust the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor; the curve prediction module is used to predict the power curve of the future air supply system through the collected historical server operation data and the corresponding power data of the air supply system; the adjustment module is used to adjust the adjustment strategy of the air supply system according to the power curve.

[0135] In this embodiment, the server control module monitors the actual load of each server in the computer room and dynamically adjusts the number of servers turned on based on the load. When it detects that some servers are in a low load state, the module will automatically shut down these servers to reduce unnecessary energy consumption. Conversely, during high load periods, it will increase the number of servers turned on to ensure that business needs are met while maintaining efficient energy use.

[0136] The power calculation module calculates the power demand of the air supply system based on the number of servers currently turned on and the temperature difference between indoor and outdoor. It can adjust the power output of the air supply system in each area in real time according to the working status of the server and changes in external environmental conditions, thereby achieving precise temperature control and energy saving.

[0137] The curve prediction module analyzes historical server operation data (such as load, processing task volume, etc.) and the corresponding air supply system power data, and uses advanced prediction algorithms to calculate the power demand curve of the future air supply system. This helps to plan the operation mode of the air supply system in advance and avoid energy waste or overheating risks caused by sudden load changes.

[0138] Based on the information provided by the above three modules, the adjustment module is responsible for formulating specific adjustment strategies and performing corresponding operations to control the air supply system. It will dynamically adjust parameters such as air supply volume and fan speed according to the predicted power curve to maintain the optimal state of the internal environment of the computer room. In addition, it can continuously optimize the adjustment strategy according to the actual situation to achieve higher energy saving effects.

[0139] This system is designed to achieve efficient and energy-saving operation of the computer room through intelligent management and refined control, while ensuring the normal operation of the server and unaffected service quality.

[0140] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for energy-saving control of a computer room, It is characterized in that Including: dynamically adjusting the number of servers enabled according to the actual load of the servers in the computer room; Adjust the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor areas; Predict the future power curve of the air supply system by collecting historical server operation data and corresponding air supply system power data; Adjust the air supply system adjustment strategy according to the power curve.

2. A computer room energy-saving control method as claimed in claim 1, characterized in that: After adjusting the adjustment strategy of the air supply system according to the power curve, the method further includes: setting energy-saving strategy parameters, correcting the adjustment strategy, and displaying the working status.

3. A computer room energy-saving control method as claimed in claim 2, characterized in that: The specific steps of dynamically adjusting the number of servers enabled according to the actual load of the servers in the computer room include: Collect real-time load data of the server; Set the upper and lower load thresholds; When the real-time load data exceeds the upper threshold, the server is started; when the real-time load data is less than the lower threshold, the server is shut down.

4. A computer room energy-saving control method as claimed in claim 3, characterized in that: The real-time load data of the server includes CPU usage, memory usage, and disk read and write status.

5. A computer room energy-saving control method as claimed in claim 4, characterized in that: The specific steps of adjusting the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor include: Get indoor and outdoor temperature data and calculate the temperature difference; Calculate the power adjustment level based on the temperature difference; Calculate the required air volume based on the number of servers and power adjustment levels in each area; Use control algorithms to adjust the air supply system power according to the required air volume and control it by zones; Document the tuning process and evaluate the temperature regulation effect of the tuning strategy.

6. A computer room energy-saving control method as claimed in claim 5, characterized in that: The specific steps of calculating the power adjustment level according to the temperature difference include: Define the temperature range; Allocate power adjustment levels according to temperature difference ranges.

7. A computer room energy-saving control method as claimed in claim 6, characterized in that: The specific steps of calculating the required air volume in combination with the number of servers and the power adjustment level in each area include: Count the number of servers currently open in each region; Get the heat generated by each server when running at full load and adjust the heat according to the current load; Add up the heat generated by each server in the area to get the total heat generated by the area; Get the number of air changes required per hour; Calculates air volume based on heat output and air exchange rate.

8. A computer room energy-saving control method as claimed in claim 7, characterized in that: The specific steps of predicting the future power curve of the air supply system by using the collected historical server operation data and the corresponding power data of the air supply system include: Collect historical server operation data and corresponding air supply system power data; Clean and organize the collected data; Extract identification features from raw data; Use historical data to train a linear regression model based on the identified features; Use the test set to evaluate the generalization ability of the linear regression model; The trained model is used to predict future time points to obtain the power curve of the future air supply system.

9. A computer room energy-saving control system, using a computer room energy-saving control method according to any one of claims 1 to 8, characterized in that: It includes a server control module, a power calculation module, a curve prediction module and an adjustment module; The server control module is used to dynamically adjust the number of servers enabled according to the actual load of the servers in the computer room; The power calculation module is used to adjust the power of the air supply system in each area according to the number of servers turned on and the temperature difference between indoor and outdoor; The curve prediction module is used to predict the future power curve of the air supply system through the collected historical server operation data and the corresponding power data of the air supply system; The adjustment module is used to adjust the adjustment strategy of the air supply system according to the power curve.

Images