Data center energy efficiency monitoring method and system based on digital twinning

Abstract

The invention relates to the technical field of energy efficiency management, in particular to a data center energy efficiency monitoring method and system based on digital twinning, and the method comprises the following steps: building a digital twinning model of a data center based on the layout and position information of equipment, collecting and synchronizing the environment temperature and humidity, and the operation state data of a server and cooling equipment in real time; and generating a monitoring data synchronization record. According to the method, real-time mapping and monitoring of energy use and equipment operation states in a virtual environment are realized by constructing a digital twin model of a data center and synchronizing operation data in real time, and cooling requirements are predicted and energy consumption of cooling equipment is calculated by analyzing energy consumption data of a server under different loads and combining real-time temperature data. The method improves the understanding capability of the relationship between the energy consumption and the load of the data center, combines the monitoring of the real-time energy efficiency data, timely discovers and solves the problem of abnormal energy consumption, and reduces the risk of accidental shutdown and the related maintenance cost.

Description

Technical Field

[0001] The present invention relates to the field of energy efficiency management technology, and in particular to a data center energy efficiency monitoring method and system based on digital twins. Background Art

[0002] The field of energy efficiency management technology focuses on improving the efficiency of energy use through a variety of monitoring, control, and optimization methods, including the application of a variety of technical means and management strategies to reduce energy consumption, reduce costs, and mitigate environmental impacts. It uses a variety of data analysis tools, sensor technologies, and automation technologies, combined with real-time data monitoring and historical data analysis, to identify the causes of excessive energy consumption, adjust equipment operating parameters, and reduce electricity costs and carbon footprints through intelligent scheduling and equipment optimization. It optimizes energy efficiency and is applied to a variety of energy use environments in industry, commerce, and residences to support sustainable development goals.

[0003] Among them, the data center energy efficiency monitoring method focuses on using a variety of methods to detect the actual operation of the data center, aiming to monitor and optimize the energy use of the data center in real time, and to achieve decision support for energy efficiency optimization, fault prediction, maintenance planning and system upgrades by analyzing the energy efficiency performance of the data center under a variety of operating conditions. It also tests and implements energy management strategies without affecting the actual operation of the data center, optimizes equipment operating parameters, and achieves the purpose of reducing energy consumption, improving energy efficiency, reducing maintenance costs and extending equipment life.

[0004] Traditional data center energy efficiency monitoring methods rely on traditional data monitoring and post-processing, lack the ability to respond to dynamic changes in the data center in real time, and cannot provide sufficient data support for real-time decision support when the data center load changes frequently. When processing large-scale energy data, it is difficult to achieve efficient data synchronization and real-time analysis, which limits the immediate implementation of energy efficiency optimization measures. There is a lack of effective prediction tools and real-time monitoring systems, and insufficient performance in fault prediction and system upgrade decision support, resulting in higher operation and maintenance costs and energy waste for data centers. The inability to adjust the cooling system in real time leads to over-cooling or under-cooling, affecting equipment performance and increasing energy expenses. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and propose a data center energy efficiency monitoring method and system based on digital twins.

[0006] In order to achieve the above objectives, the present invention adopts the following technical solution, a data center energy efficiency monitoring method based on digital twin, comprising the following steps:

[0007] S1: Based on the equipment layout and location information, a digital twin model of the data center is constructed to collect and synchronize the environmental temperature and humidity, server and cooling equipment operating status data in real time, and generate synchronous records of monitoring data;

[0008] S2: Analyze the energy consumption data and load data of the server by using the synchronous recording of the monitoring data, identify the energy consumption of the server under various load conditions, evaluate the relationship between the server load and heat generation in combination with the real-time temperature data, and generate energy consumption correlation analysis results;

[0009] S3: Based on the energy consumption correlation analysis results, simulate the air flow and temperature distribution inside the data center in real time, consider the external climate conditions and internal temperature and humidity data, predict the cooling demand and calculate the energy consumption of the cooling equipment, and generate the cooling equipment energy consumption data;

[0010] S4: using the cooling equipment energy consumption data to predict the energy demand of the data center under various load conditions according to the server load, and by comparing with the actual energy consumption data, calibrating the parameters of the prediction model to generate a data center energy consumption model;

[0011] S5: According to the data center energy consumption model, by performing time series analysis on the energy consumption data of the data center, analyzing the change trend of energy demand, and generating energy consumption trend prediction information;

[0012] S6: Based on the energy consumption trend forecast information, by real-time comparison and analysis of the data processing requirements and energy consumption data of the data center, abnormal fluctuations in the energy consumption data are identified, and the time and cause of the abnormal energy consumption are analyzed to generate real-time energy efficiency monitoring records.

[0013] As a further solution of the present invention, the monitoring data synchronous record includes ambient temperature and humidity data, server operation status information, and power consumption data of the cooling equipment. The energy consumption correlation analysis results include energy consumption under multiple server configurations and workloads, correlation analysis results of server heat generation and real-time ambient temperature, and correlation evaluation information between energy consumption data and server performance indicators. The cooling equipment energy consumption data includes internal air flow patterns, predicted temperature distribution diagrams, and cooling demand change information. The data center energy consumption model includes predicted values ​​of energy demand under multiple load states, comparative analysis results of energy prediction values ​​and actual consumption data, and optimization and calibration records of prediction model parameters. The energy consumption trend prediction information includes seasonal and periodic change trends of data center energy consumption, energy consumption peak and valley time point prediction information, and energy consumption change pattern analysis results. The real-time energy efficiency monitoring records include abnormal energy consumption fluctuation points, abnormal occurrence time and duration, and abnormal fluctuation cause analysis results.

[0014] As a further solution of the present invention, based on the equipment layout and location information, a digital twin model of the data center is constructed, and the operating status data of the ambient temperature and humidity, servers and cooling equipment are collected and synchronized in real time. The steps of generating synchronous records of monitoring data are specifically as follows:

[0015] S101: Based on the equipment layout and location information, a three-dimensional digital twin model of the data center is constructed by analyzing the location information of multiple servers, cooling equipment, air conditioning outlets, and sensors in the data center to generate a spatial data mapping record;

[0016] S102: Based on the spatial data mapping record, collect temperature and humidity data of multiple locations in the data center in real time, record server and cooling equipment operation information, sort them in combination with time data, and generate a real-time monitoring data set;

[0017] S103: Based on the real-time monitoring data set, synchronize the monitoring data to the digital twin model in real time, update the model parameters in real time to reflect the real-time status of the data center, and generate a monitoring data synchronization record.

[0018] As a further solution of the present invention, the energy consumption data and load data of the server are analyzed by synchronously recording the monitoring data, the energy consumption of the server under various load conditions is identified, and the relationship between the server load and the heat generation is evaluated in combination with the real-time temperature data. The steps of generating the energy consumption correlation analysis result are specifically as follows:

[0019] S201: extracting the energy consumption and load data of the server during operation based on the synchronous recording of the monitoring data, and generating server energy consumption and load information;

[0020] S202: Based on the server energy consumption load information, by analyzing the energy consumption changes of the server device under various operating states, evaluating the impact of load changes on energy consumption, calculating a correlation coefficient, and generating a correlation analysis result;

[0021] S203: Based on the correlation analysis result, using real-time temperature data, evaluating the relationship between server load and heat generation, combining energy efficiency changes under various operating states, and generating energy consumption correlation analysis results.

[0022] As a further solution of the present invention, the specific formula for calculating the correlation coefficient is:

[0023]

[0024] Among them, x i Represents the load data for a single data point, y i Represents the corresponding energy consumption data, and They represent the sample averages of load data and energy consumption data respectively, and r represents the correlation coefficient between load and energy consumption.

[0025] As a further solution of the present invention, based on the energy consumption correlation analysis results, the air flow and temperature distribution inside the data center are simulated in real time, the external climate conditions and the internal temperature and humidity data are considered, the cooling demand is predicted and the energy consumption of the cooling equipment is calculated, and the steps of generating the cooling equipment energy consumption data are specifically as follows:

[0026] S301: Based on the energy consumption correlation analysis result and according to the server load data, the operating status and heat generation of the equipment in the data center are predicted in real time, and air flow and temperature distribution are simulated to generate temperature distribution simulation data;

[0027] S302: Based on the temperature distribution simulation data, considering real-time external climate conditions and internal environmental data, evaluating the cooling demand of the data center, and generating a cooling demand prediction result;

[0028] S303: Utilizing the cooling demand prediction result, by analyzing the response efficiency and working performance of the cooling equipment, the energy consumption demand of the cooling equipment is calculated in real time to generate cooling equipment energy consumption data.

[0029] As a further solution of the present invention, the energy consumption data of the cooling equipment is used to predict the energy demand of the data center under various load conditions according to the server load, and the parameters of the prediction model are calibrated by comparing with the actual energy consumption data. The steps of generating the data center energy consumption model are specifically as follows:

[0030] S401: extracting the server load, ambient temperature and energy consumption data of the data center based on the cooling equipment energy consumption data, and generating a real-time monitoring data set;

[0031] S402: Analyze the energy demand of the data center under various load conditions by using the real-time monitoring data set according to the relationship between the load level and the energy consumption of the server and the relationship between the server load and the energy consumption of the cooling equipment, and generate an energy demand analysis result;

[0032] S403: According to the energy demand analysis result, by comparing the predicted value with the actual energy consumption data, calibrating the parameters of the prediction model to reflect the actual energy consumption status, and generating a data center energy consumption model.

[0033] As a further solution of the present invention, according to the data center energy consumption model, by performing time series analysis on the energy consumption data of the data center, analyzing the changing trend of energy demand, and generating energy consumption trend prediction information, the specific steps are:

[0034] S501: Based on the data center energy consumption model, format and standardize the energy consumption data of the data center, and sort them using time information to generate a time series data set;

[0035] S502: Based on the time series data set, evaluate the seasonal changes and periodic patterns of the energy consumption of the data center through time series analysis, and generate periodic energy consumption analysis results;

[0036] S503: Based on the periodic energy consumption analysis result, predict the change trend and periodic fluctuation of the energy demand of the data center, and generate energy consumption trend prediction information.

[0037] As a further solution of the present invention, according to the energy consumption trend prediction information, by real-time comparison and analysis of the data processing requirements and energy consumption data of the data center, identifying abnormal fluctuations in the energy consumption data, and analyzing the time and cause of the abnormal energy consumption, the steps of generating real-time energy efficiency monitoring records are specifically as follows:

[0038] S601: Based on the energy consumption trend prediction information, according to the relationship between the data center load data and the energy consumption level, compare the energy consumption data and the data processing requirements in real time, analyze the consistency of data changes, and generate real-time energy consumption comparison data;

[0039] S602: Analyze the real-time energy consumption comparison data, detect and identify abnormal energy consumption fluctuations in the data, record the time when the abnormality occurs and the duration, and generate an abnormal energy consumption detection record;

[0040] S603: Based on the abnormal energy consumption detection record, analyze and identify the causes of abnormal energy consumption, including equipment failure, network attack, external environmental impact, and cooling equipment working configuration error, and generate real-time energy efficiency monitoring records.

[0041] A data center energy efficiency monitoring system based on digital twins, the data center energy efficiency monitoring system based on digital twins is used to execute the above-mentioned data center energy efficiency monitoring method based on digital twins, the system comprising:

[0042] The digital twin model construction module builds a three-dimensional digital twin model of the data center based on the equipment layout and location information, and synchronizes the operating data of the environment temperature and humidity, servers, and cooling equipment in real time to generate synchronous records of monitoring data;

[0043] The load data correlation evaluation module uses the monitoring data to record synchronously, analyzes the relationship between the energy consumption and load of the server in real time, combines the temperature data, evaluates the relationship between the load and the heat generation, and generates energy consumption correlation analysis results;

[0044] The temperature and cooling demand prediction module predicts the cooling demand of the data center based on the energy consumption correlation analysis result, by simulating the air flow and temperature distribution inside the data center, and calculates the energy consumption of the cooling equipment to generate cooling equipment energy consumption data;

[0045] The energy consumption trend prediction module uses the energy consumption data of the cooling equipment, combines the real-time load data of the server and the energy efficiency level of the cooling equipment, analyzes the energy demand of the data center, predicts the change trend of energy consumption, and calibrates the energy consumption prediction model through actual consumption data to generate energy consumption trend prediction information;

[0046] The abnormal energy consumption monitoring module utilizes the energy consumption trend prediction information to identify and analyze abnormal fluctuations in the energy consumption data by comparing the data processing requirements with the energy consumption data in real time, evaluates the causes of abnormal energy consumption, and generates real-time energy efficiency monitoring records.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are:

[0048] In the present invention, by constructing a digital twin model of the data center and synchronizing the operating data in real time, real-time mapping and monitoring of energy usage and equipment operating status in a virtual environment are achieved. By analyzing the energy consumption data of the server under different loads and combining the real-time temperature data to predict the cooling demand and calculate the energy consumption of the cooling equipment, the ability to understand the relationship between the energy consumption and load of the data center is improved. Combined with the monitoring of real-time energy efficiency data, abnormal energy consumption problems can be discovered and resolved in a timely manner, reducing the risk of unexpected downtime and related maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0050] Figure 2 This is a detailed flow chart of S1 of the present invention;

[0051] Figure 3 This is a detailed flow chart of S2 of the present invention;

[0052] Figure 4 This is a detailed flow chart of S3 of the present invention;

[0053] Figure 5 This is a detailed flow chart of S4 of the present invention;

[0054] Figure 6 This is a detailed flow chart of S5 of the present invention;

[0055] Figure 7 This is a detailed flow chart of S6 of the present invention;

[0056] Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0058] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0059] See also Figure 1 The present invention provides a technical solution, a data center energy efficiency monitoring method based on digital twins, comprising the following steps:

[0060] S1: Based on the equipment layout and location information, a digital twin model of the data center is constructed to collect and synchronize the environmental temperature and humidity, server and cooling equipment operating status data in real time, and generate synchronous records of monitoring data;

[0061] S2: Analyze the energy consumption data and load data of the server using the synchronous recording of monitoring data, identify the server energy consumption under various load conditions, and evaluate the relationship between server load and heat generation in combination with real-time temperature data to generate energy consumption correlation analysis results;

[0062] S3: Based on the energy consumption correlation analysis results, the air flow and temperature distribution inside the data center are simulated in real time, the external climate conditions and internal temperature and humidity data are considered, the cooling demand is predicted and the energy consumption of the cooling equipment is calculated, and the energy consumption data of the cooling equipment is generated;

[0063] S4: Use the cooling equipment energy consumption data to predict the energy demand of the data center under various load conditions according to the server load, and calibrate the parameters of the prediction model by comparing with the actual energy consumption data to generate a data center energy consumption model;

[0064] S5: Based on the data center energy consumption model, by performing time series analysis on the data center energy consumption data, analyzing the changing trend of energy demand, and generating energy consumption trend forecast information;

[0065] S6: Based on the energy consumption trend forecast information, by real-time comparison and analysis of the data processing requirements and energy consumption data of the data center, abnormal fluctuations in the energy consumption data are identified, and the time and cause of abnormal energy consumption are analyzed to generate real-time energy efficiency monitoring records.

[0066] The monitoring data synchronously recorded include environmental temperature and humidity data, server operation status information, and power consumption data of cooling equipment. The energy consumption correlation analysis results include energy consumption under various server configurations and workloads, correlation analysis results between server heat generation and real-time ambient temperature, and correlation evaluation information between energy consumption data and server performance indicators. The cooling equipment energy consumption data includes internal air flow patterns, predicted temperature distribution diagrams, and cooling demand change information. The data center energy consumption model includes predicted values ​​of energy demand under various load states, comparative analysis results of energy prediction values ​​and actual consumption data, and optimization and calibration records of prediction model parameters. Energy consumption trend prediction information includes seasonal and periodic change trends of data center energy consumption, energy consumption peak and valley time point prediction information, and energy consumption change pattern analysis results. Real-time energy efficiency monitoring records include abnormal energy consumption fluctuation points, abnormal occurrence time and duration, and abnormal fluctuation cause analysis results.

[0067] See also Figure 2 , based on the equipment layout and location information, a digital twin model of the data center is built to collect and synchronize the environmental temperature and humidity, server and cooling equipment operating status data in real time. The specific steps for generating monitoring data synchronization records are as follows:

[0068] S101: Based on the equipment layout and location information, a three-dimensional digital twin model of the data center is constructed by analyzing the location information of multiple servers, cooling equipment, air conditioning outlets, and sensors in the data center to generate a spatial data mapping record;

[0069] In sub-step S101, the spatial layout of the data center is obtained through measurement tools, including the location of the server, the arrangement of the cooling equipment, the air-conditioning outlets and the installation locations of the sensors. The target data is used as an input parameter to create a spatial data mapping record of the model. The location of each device and sensor in the model is consistent with the actual physical location. Computer-aided design software is used to convert the target location information into the coordinates of the three-dimensional model, which is visualized through Autodesk Revit. Each change will be updated to the model in real time to ensure the real-time and accuracy of the model data. The completed digital twin model can accurately reflect the physical state of the data center in a virtual environment, providing basic data for subsequent simulation and monitoring.

[0070] S102: Based on the spatial data mapping record, collect temperature and humidity data at multiple locations in the data center in real time, record the operation information of the server and cooling equipment, sort them in combination with the time data, and generate a real-time monitoring data set;

[0071] In sub-step S102, based on the constructed digital twin model, environmental monitoring of the data center is implemented. Multi-point temperature and humidity sensors are used to collect environmental data at key locations, including air temperature and relative humidity. Wireless sensor network technology is used to transmit data to the central monitoring system in real time. The data is sorted and stored according to data timestamps to generate a real-time monitoring data set. Data processing uses data cleaning and normalization methods to eliminate the influence of outliers and noise, ensure the accuracy and reliability of the data, and extract patterns and trends of environmental changes from real-time data through cluster analysis for more in-depth analysis and application.

[0072] S103: Based on the real-time monitoring data set, synchronize the monitoring data to the digital twin model in real time, update the model parameters in real time to reflect the real-time status of the data center, and generate a synchronization record of the monitoring data;

[0073] In sub-step S103, after the real-time monitoring data set is obtained, the data set is synchronized to the digital twin model in real time. During the synchronization process, API calling technology is used to ensure the real-time transmission of data from the collection point to the model. The model parameters are dynamically adjusted as the data is updated, such as the changes in air-conditioning adjustment parameters caused by changes in temperature and humidity. MATLAB Simulink is used to perform dynamic simulation of the model during the process. The simulation process takes into account the changes in real-time data to ensure that the model can accurately reflect the operating status of the data center. Scripts are used to automatically update model parameters to reduce manual intervention and improve system response speed and processing efficiency. The generated monitoring data synchronization record records in detail the history of all data updates and parameter adjustments, providing a basis for system maintenance and troubleshooting.

[0074] See also Figure 3 ,Using the synchronous recording of monitoring data, the energy consumption data and load data of the server are analyzed, the server energy consumption under various load conditions is identified, and the relationship between server load and heat generation is evaluated in combination with real-time temperature data. The specific steps for generating energy consumption correlation analysis results are as follows:

[0075] S201: extracting energy consumption and load data of the server during operation based on synchronous recording of monitoring data, and generating server energy consumption and load information;

[0076] In sub-step S201, a database management system is used to extract the energy consumption and load data of the server from the synchronous records of the monitoring data. The data includes key performance indicators such as the server's power consumption, CPU and memory usage. During the data extraction process, SQL query statements are used to accurately select the required data fields and pre-process the data, such as removing invalid or erroneous data and standardizing the data format to ensure data quality and provide accurate input for analysis. Through the target processing step, detailed server energy consumption and load information is generated. The information records the energy consumption and operating status of each server at different time points, providing basic data for subsequent energy consumption analysis.

[0077] S202: Based on the server energy consumption load information, by analyzing the energy consumption changes of the server device under various operating states, evaluating the impact of the load change on the energy consumption, calculating the correlation coefficient, and generating a correlation analysis result;

[0078] The specific formula for calculating the correlation coefficient is:

[0079]

[0080] Among them, x i Represents the load data for a single data point, y i Represents the corresponding energy consumption data, and They represent the sample averages of load data and energy consumption data respectively, and r represents the correlation coefficient between load and energy consumption.

[0081] formula:

[0082]

[0083] Detailed explanation of the formula and the process of formula calculation and derivation:

[0084] The formula is used to calculate the Pearson correlation coefficient between load data and energy consumption data. The coefficient is used to quantify the linear relationship between the two and evaluate how load changes affect energy consumption;

[0085] Parameter meaning and setting value:

[0086] x i Represents the load data of a single data point. Set the load data to the real-time CPU usage of the server. Assume that the four monitored data are 20%, 25%, 30%, and 35% respectively;

[0087] y i Indicates the corresponding energy consumption data. Set the energy consumption data to the real-time power consumption of the server. Assume that the monitored power is 200W, 250W, 300W, and 350W respectively.

[0088] and They are x i and i The average value, load average Average energy consumption

[0089] r is the calculated correlation coefficient, which expresses the strength of the correlation between load and energy consumption;

[0090] Formula calculation process:

[0091] Compute the sum of the deviation products:

[0092]

[0093] Calculate the sum of squares of load and energy consumption deviations:

[0094]

[0095]

[0096] Calculate r:

[0097]

[0098] The result r=1 shows that there is a positive correlation between load and energy consumption, which means that as the server load increases, energy consumption also increases accordingly. The results can help data center managers understand and predict energy consumption changes and optimize energy use and cost efficiency.

[0099] S203: Based on the correlation analysis results, using the real-time temperature data, evaluating the relationship between the server load and heat generation, combining the energy efficiency changes under various operating states, and generating energy consumption correlation analysis results;

[0100] In sub-step S203, based on the correlation analysis results, the relationship between server load and heat generation is evaluated using real-time monitored temperature data. Dynamic simulation is performed using thermodynamic simulation software to examine how the server load affects its heat generation and the thermal environment of the entire data center under different ambient temperatures. The simulation uses parameter settings based on actual operating data, such as the server's thermal output and the responsiveness of the cooling system. The temperature distribution diagram obtained through simulation allows us to intuitively see how heat is distributed and how cooling demand changes under high load conditions. Combined with actual energy efficiency data, energy consumption correlation analysis results are generated. The results reflect the direct correlation between load and temperature, and reveal the potential space for optimizing cooling strategies.

[0101] See also Figure 4Based on the energy consumption correlation analysis results, the air flow and temperature distribution inside the data center are simulated in real time, the external climate conditions and internal temperature and humidity data are considered, the cooling demand is predicted and the energy consumption of the cooling equipment is calculated. The specific steps for generating the cooling equipment energy consumption data are as follows:

[0102] S301: Based on the energy consumption correlation analysis results and server load data, the operating status and heat generation of the equipment in the data center are predicted in real time, and air flow and temperature distribution are simulated to generate temperature distribution simulation data;

[0103] In sub-step S301, based on the results of energy consumption correlation analysis, the support vector machine model in machine learning technology is used to predict the operating status and heat generation of data center equipment according to historical server load data and environmental variable data. The prediction model analyzes the power usage and temperature monitoring records of the equipment inside the data center, and uses the Gaussian kernel function to process nonlinear relationships to ensure that the model can accurately reflect the dynamic changes between the equipment status and temperature. Then, computational fluid dynamics simulation software is used to simulate the air flow and temperature distribution in the data center. Boundary conditions and initial states are set during the simulation process, such as equipment layout and air duct design. The distribution of air flow and temperature field is updated in real time through iterative calculation. The simulation provides detailed temperature distribution simulation data for subsequent cooling demand assessment.

[0104] S302: Based on the temperature distribution simulation data, considering the real-time external climate conditions and internal environmental data, evaluating the cooling demand of the data center, and generating a cooling demand prediction result;

[0105] In the above content, based on the temperature distribution simulation data, combined with the real-time external climate conditions and internal environmental data, the cooling demand of the data center is evaluated according to the formula Q = c p VρΔT calculates cooling demand;

[0106] Where Q represents the required cooling energy, c p represents the specific heat capacity of air, V represents the volume of air inside the data center, ρ represents the air density, and ΔT represents the temperature difference of the air;

[0107] Detailed explanation of the formula and the process of formula calculation and derivation:

[0108] Assuming that the temperature inside the data center needs to be maintained at T1 = 22°C, and the real-time temperature is T2 = 35°C, calculate ΔT = T2-T1 = 35-22 = 13°C, and the specific heat capacity of air c p =1.005kJ / kg·K, the air volume inside the data center V=1000m3, the air density ρ=1.225kg / m3, substitute the values ​​into the formula:

[0109] Q=1.005kJ / kg·K×1000m3 ×1.225kg / m 3 ×13K=16272.75kJ

[0110] The result of 16272.75 kJ indicates that an additional 16272.75 kJ of cooling energy is needed to reduce the temperature inside the data center to 22 degrees. The formula is used to evaluate the actual cooling system requirements of the data center to ensure effective management of the temperature and energy consumption of the data center.

[0111] S303: using the cooling demand prediction result, by analyzing the response efficiency and working performance of the cooling equipment, calculating the energy consumption demand of the cooling equipment in real time, and generating cooling equipment energy consumption data;

[0112] In sub-step S303, the cooling demand prediction results obtained from step S302 are used to continue analyzing the response efficiency and working effectiveness of the cooling equipment, and the energy consumption demand of the cooling equipment is calculated in real time. During the calculation process, the thermal balance analysis method is used, combined with the performance parameters of the cooling equipment, including the cooling coefficient and the energy consumption rate, and the energy consumption of each device under the predicted cooling demand is estimated through the heat load calculation formula. Taking into account the dynamic adjustment capability of the system, the analysis helps optimize the operating settings of the cooling system, such as adjusting the fan speed and cooling water flow of the cooling tower. The generated cooling equipment energy consumption data reflects the energy efficiency of the equipment under current and predicted conditions, providing a scientific basis for energy management and cost control.

[0113] See also Figure 5 ,Using the energy consumption data of cooling equipment, the energy demand of the data center under various load conditions is predicted according to the server load, and the parameters of the prediction model are calibrated by comparing with the actual energy consumption data. The specific steps of generating the data center energy consumption model are as follows:

[0114] S401: extracting the server load, ambient temperature and energy consumption data of the data center based on the energy consumption data of the cooling equipment, and generating a real-time monitoring data set;

[0115] In sub-step S401, based on the energy consumption data of the cooling equipment, the core monitoring parameters of the data center are collected, including the real-time load of the server, ambient temperature and energy consumption data. The data acquisition system is used in the process. The system is equipped with temperature sensors and energy consumption meters. The data is transmitted to the central monitoring system in real time through the network interface. The system uses a time series database to store and manage the collected data to ensure the integrity and real-time nature of the data. The data is pre-processed, including filtering, denoising and outlier processing, to ensure data quality and provide accurate input for subsequent energy demand analysis and model calibration. The generated real-time monitoring data set contains the timestamp data and corresponding monitoring values ​​of each monitoring point. The target data is crucial for in-depth analysis and decision support.

[0116] S402: using the real-time monitoring data set, analyzing the energy demand of the data center under various load states according to the relationship between the load level and the energy consumption of the server, and combining the relationship between the server load and the energy consumption of the cooling equipment, and generating an energy demand analysis result;

[0117] In sub-step S402, based on the real-time monitoring data set, statistical analysis techniques and machine learning methods, such as the random forest algorithm, are applied to process and analyze the energy consumption patterns and server load fluctuations in the data set. The relationship between the energy demand of the data center and the server operating status is determined through target analysis. Taking into account the interdependence between server load and cooling equipment energy consumption, multivariate analysis is performed to identify the main influencing factors and energy consumption drivers. The analysis results help understand the energy utilization efficiency of the data center under different operating conditions and generate energy demand analysis results. The target results provide a basis for operational optimization for data center management, ensuring that energy efficiency and cost-effectiveness are maximized while maintaining service quality.

[0118] S403: According to the energy demand analysis result, by comparing the predicted value with the actual energy consumption data, calibrating the parameters of the prediction model to reflect the actual energy consumption status, and generating a data center energy consumption model;

[0119] In sub-step S403, based on the results of the energy demand analysis, the energy consumption prediction model of the data center is fine-tuned using data-driven model calibration technology. The adjustment is performed by comparing the model prediction value with the actual energy consumption data obtained from the real-time monitoring data set. Regression analysis and error correction methods, including Kalman filters, are applied to optimize the parameter settings of the model to ensure that the model accurately reflects the energy consumption in actual operations. In the process, seasonal changes and abnormal fluctuations in the data are taken into account, and the model is adjusted to adapt to target changes. The generated data center energy consumption model provides a real-time response and prediction capability for data center operations, making energy efficiency management more accurate and efficient, and providing a powerful tool for reducing energy waste and costs.

[0120] See also Figure 6 ,According to the data center energy consumption model, by performing time series analysis on the data center energy consumption data, analyzing the changing trend of energy demand, the steps of generating energy consumption trend forecast information are as follows:

[0121] S501: Based on the data center energy consumption model, the energy consumption data of the data center is formatted and standardized, and sorted using time information to generate a time series data set;

[0122] In sub-step S501, based on the data center energy consumption model, the collected energy consumption data is formatted and standardized, and data sorting technology is used to ensure that all data items meet the analysis requirements, such as converting the data formats of different devices and sensor sources into a unified format, applying normalization processing to standardize various energy consumption data, eliminating dimensional effects and improving the accuracy of subsequent analysis, and using the database management system to sort and index data by timestamp, quickly extracting data from a specific time period for analysis, laying the foundation for building an efficient time series data set. The generated time series data set contains energy consumption data collected from multiple time points, and the target data set will be directly used in subsequent energy consumption trend analysis and pattern recognition.

[0123] S502: Based on the time series data set, evaluate the seasonal changes and periodic patterns of the energy consumption of the data center through time series analysis, and generate periodic energy consumption analysis results;

[0124] In sub-step S502, based on the generated time series data set, time series analysis techniques such as seasonal decomposition and periodic analysis methods, such as the autoregressive moving average model, are used to process the energy consumption data of the data center. The target technology helps to reveal the seasonal changes and periodic patterns in the data, decompose the time series data into trend, seasonal and random components, evaluate the contribution and change characteristics of each component, identify the patterns of energy use in a specific season or cycle, and predict possible future change trends. The generated periodic energy consumption analysis results provide a scientific basis for data center management, adjust energy supply and optimize energy consumption strategies.

[0125] S503: Based on the periodic energy consumption analysis result, predict the change trend and periodic fluctuation of the energy demand of the data center, and generate energy consumption trend prediction information;

[0126] In sub-step S503, based on the results of periodic energy consumption analysis, the exponential smoothing method and seasonal autoregressive integrated moving average model are used to predict the changing trend and periodic fluctuations of data center energy demand. The target model combines the seasonal pattern of historical energy consumption data and statistical principles to achieve accurate prediction of future energy demand, including predicting the changes in energy consumption of data centers in different seasons or operating cycles, so that data center operators can make strategic adjustments in advance to cope with expected changes in energy demand. The generated energy consumption trend forecast information provides important support for real-time energy efficiency management and long-term energy planning, and enhances the ability of data centers to cope with fluctuations in energy consumption.

[0127] See also Figure 7 According to the energy consumption trend forecast information, by real-time comparison and analysis of the data processing requirements and energy consumption data of the data center, abnormal fluctuations in the energy consumption data are identified, and the time and cause of abnormal energy consumption are analyzed. The specific steps for generating real-time energy efficiency monitoring records are as follows:

[0128] S601: Based on the energy consumption trend prediction information and the relationship between the data center load data and the energy consumption level, the energy consumption data and the data processing requirements are compared in real time, the consistency of the data changes is analyzed, and real-time energy consumption comparison data is generated;

[0129] In sub-step S601, based on the energy consumption trend forecast information, data fusion technology is used to comprehensively analyze the real-time load data and energy consumption level of the data center. The real-time monitoring data includes the CPU utilization rate, memory usage and device power consumption of the server. The target data is processed through an advanced data analysis platform, and regression analysis and correlation analysis methods are used to compare the relationship between energy consumption data and data processing requirements. Time series analysis is introduced in the process to consider the time dependence of the data to ensure the timeliness and accuracy of the analysis results. Through this comparative analysis, energy utilization efficiency is effectively monitored and possible energy-saving improvement areas are identified. The generated real-time energy consumption comparison data provides instant feedback on data center energy efficiency optimization.

[0130] S602: Analyze the real-time energy consumption comparison data, detect and identify abnormal energy consumption fluctuations in the data, record the time and duration of the abnormality, and generate abnormal energy consumption detection records;

[0131] In the above content, analyze the real-time energy consumption comparison data, detect and identify abnormal energy consumption fluctuations in the data, and follow the formula Calculate standard scores for energy consumption data and identify abnormal fluctuations;

[0132] In the formula, Z represents the standard score, X represents the real-time observed energy consumption value, μ represents the expected value of energy consumption or the historical average energy consumption, and σ represents the standard deviation of energy consumption data;

[0133] Detailed explanation of the formula and the process of formula calculation and derivation:

[0134] Assume that historical data shows that the average energy consumption of the data center is μ = 250 kW, the standard deviation is σ = 15 kW, and the current observed energy consumption is X = 290 kW.

[0135] Calculate the standard score Z:

[0136]

[0137] The result Z≈2.67 indicates that the current energy consumption value is about 2.67 standard deviations higher than the average value. The result is used to detect and identify energy consumption data that deviates significantly from the norm in real time and take necessary diagnostic or corrective measures.

[0138] S603: Based on the abnormal energy consumption detection record, analyze and identify the causes of abnormal energy consumption, including equipment failure, network attack, external environment impact, and cooling equipment working configuration error, and generate real-time energy efficiency monitoring records;

[0139] In sub-step S603, based on the abnormal energy consumption detection records, data analysis and fault diagnosis technology are used to analyze and determine the root cause of the abnormal energy consumption. The process involves extracting key information from system logs, equipment performance records and network monitoring data, and using causal analysis and root cause analysis methods to explore the background and environmental factors of the abnormality. Factors considered include equipment failure, network security incidents, adverse environmental impacts or configuration errors of the cooling system. Target analysis helps to determine the key variables and improvement points that affect the energy efficiency of the data center. The generated real-time energy efficiency monitoring records not only provide a basis for operational decisions for data center management, but also provide data support for future risk prevention and system optimization.

[0140] See also Figure 8 , a data center energy efficiency monitoring system based on digital twins, the data center energy efficiency monitoring system based on digital twins is used to execute the above-mentioned data center energy efficiency monitoring method based on digital twins, the system includes:

[0141] The digital twin model construction module builds a three-dimensional digital twin model of the data center based on the equipment layout and location information, and synchronizes the operating data of the environmental temperature and humidity, servers, and cooling equipment in real time to generate synchronous records of monitoring data;

[0142] The load data correlation evaluation module uses the monitoring data to record synchronously, analyzes the relationship between the server's energy consumption and load in real time, combines the temperature data, evaluates the relationship between the load and the heat generation, and generates energy consumption correlation analysis results;

[0143] The temperature and cooling demand prediction module predicts the cooling demand of the data center based on the energy consumption correlation analysis results, by simulating the air flow and temperature distribution inside the data center, and calculates the energy consumption of the cooling equipment to generate the cooling equipment energy consumption data;

[0144] The energy consumption trend prediction module uses the energy consumption data of cooling equipment, combined with the real-time load data of the server and the energy efficiency level of the cooling equipment, to analyze the energy demand of the data center, predict the trend of energy consumption, and calibrate the energy consumption prediction model through actual consumption data to generate energy consumption trend prediction information;

[0145] The abnormal energy consumption monitoring module uses energy consumption trend forecast information to identify and analyze abnormal fluctuations in energy consumption data by comparing data processing requirements with energy consumption data in real time, evaluate the causes of abnormal energy consumption, and generate real-time energy efficiency monitoring records.

[0146] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A data center energy efficiency monitoring method based on digital twins, characterized in that: The following steps are involved: Based on the equipment layout and location information, a digital twin model of the data center is built to collect and synchronize the environmental temperature and humidity, server and cooling equipment operating status data in real time, and generate synchronous records of monitoring data; Utilizing the synchronous recording of the monitoring data, analyzing the energy consumption data and load data of the server, identifying the energy consumption of the server under various load conditions, and combining the real-time temperature data to evaluate the relationship between the server load and heat generation, generating energy consumption correlation analysis results; Based on the energy consumption correlation analysis results, the air flow and temperature distribution inside the data center are simulated in real time, the external climate conditions and internal temperature and humidity data are considered, the cooling demand is predicted and the energy consumption of the cooling equipment is calculated, and the cooling equipment energy consumption data is generated.

2. The data center energy efficiency monitoring method based on digital twin according to claim 1 is characterized in that: Also includes: Using the cooling equipment energy consumption data, predicting the energy demand of the data center under various load conditions according to the server load, and calibrating the parameters of the prediction model by comparing with the actual energy consumption data to generate a data center energy consumption model; According to the data center energy consumption model, by performing time series analysis on the energy consumption data of the data center, analyzing the changing trend of energy demand, and generating energy consumption trend prediction information; According to the energy consumption trend forecast information, by real-time comparison and analysis of the data processing requirements and energy consumption data of the data center, abnormal fluctuations in the energy consumption data are identified, and the time and cause of the abnormal energy consumption are analyzed to generate real-time energy efficiency monitoring records.

3. The data center energy efficiency monitoring method based on digital twin according to claim 2 is characterized in that: The monitoring data synchronous record includes environmental temperature and humidity data, server operation status information, and power consumption data of cooling equipment. The energy consumption correlation analysis results include energy consumption under various server configurations and workloads, correlation analysis results between server heat generation and real-time ambient temperature, and correlation evaluation information between energy consumption data and server performance indicators. The cooling equipment energy consumption data includes internal air flow patterns, predicted temperature distribution diagrams, and cooling demand change information. The data center energy consumption model includes predicted values ​​of energy demand under various load states, comparative analysis results of energy prediction values ​​and actual consumption data, and optimization and calibration records of prediction model parameters. The energy consumption trend prediction information includes seasonal and periodic change trends of data center energy consumption, energy consumption peak and valley time point prediction information, and energy consumption change pattern analysis results. The real-time energy efficiency monitoring record includes abnormal energy consumption fluctuation points, abnormal occurrence time and duration, and abnormal fluctuation cause analysis results.

4. A data center energy efficiency monitoring system based on digital twins, characterized in that: According to the data center energy efficiency monitoring method based on digital twins according to any one of claims 1 to 3, the system comprises: The digital twin model construction module builds a three-dimensional digital twin model of the data center based on the equipment layout and location information, and synchronizes the operating data of the environment temperature and humidity, servers, and cooling equipment in real time to generate synchronous records of monitoring data; The load data correlation evaluation module uses the monitoring data to record synchronously, analyzes the relationship between the energy consumption and load of the server in real time, combines the temperature data, evaluates the relationship between the load and the heat generation, and generates energy consumption correlation analysis results; The temperature and cooling demand prediction module predicts the cooling demand of the data center based on the energy consumption correlation analysis result, by simulating the air flow and temperature distribution inside the data center, and calculates the energy consumption of the cooling equipment to generate cooling equipment energy consumption data.

5. The data center energy efficiency monitoring system according to claim 4, characterized in that: Also includes: The energy consumption trend prediction module uses the energy consumption data of the cooling equipment, combines the real-time load data of the server and the energy efficiency level of the cooling equipment, analyzes the energy demand of the data center, predicts the change trend of energy consumption, and calibrates the energy consumption prediction model through actual consumption data to generate energy consumption trend prediction information; The abnormal energy consumption monitoring module utilizes the energy consumption trend prediction information to identify and analyze abnormal fluctuations in the energy consumption data by comparing the data processing requirements with the energy consumption data in real time, evaluates the causes of abnormal energy consumption, and generates real-time energy efficiency monitoring records.

Images