Machine room energy consumption integrated management system, method and computer readable medium

Abstract

This invention discloses a data center energy consumption integrated management system, method, and computer-readable medium. The system comprises a power data analysis module that analyzes the power consumption of both ICT and non-ICT equipment in the data center. A power anomaly analysis module analyzes the power consumption of these devices to determine the data center's power efficiency. Historical data on the data center's power efficiency is then used to train a time series prediction model to predict the data center's power efficiency for a future period. The power anomaly analysis module performs anomaly analysis on the predicted and actual power efficiency results to identify data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption. An alarm module then notifies the data center management to implement improvements to restore normal power consumption or reduce energy consumption.

Description

Technical Field

[0001] This invention relates to a data center energy consumption integrated management technology, and more particularly to an information and communication technology (ICT) data center energy consumption integrated management system, method, and computer-readable medium. Prior Technology

[0002] With increasing attention being paid to the concept of sustainable business operations based on ESG (Environment, Social, and Governance), the management of energy consumption and carbon emissions of ICT equipment in data centers has become quite important.

[0003] However, current technologies cannot analyze the power consumption of ICT and non-ICT equipment in a data center, nor can they analyze the power efficiency of the data center based on the power consumption of ICT and non-ICT equipment. Furthermore, they cannot use historical data on the power efficiency of the data center to train a time series prediction model to predict the power efficiency of the data center in the future. Moreover, they cannot perform anomaly analysis on the predicted and actual power efficiency of the data center to identify data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption.

[0004] Therefore, how to provide an innovative energy consumption management technology for data center to solve any of the above problems, or to provide related systems and methods, has become a major research topic for those skilled in the art. Summary of the Invention

[0005] The data center energy consumption integrated management system of the present invention includes: a power data analysis module, which analyzes the power consumption of at least one data center's ICT equipment and non-ICT equipment; and a power anomaly analysis module, which is connected to the power data analysis module. The power anomaly analysis module further analyzes the power consumption of the data center's ICT equipment and non-ICT equipment based on the power data analysis module's analysis to determine the data center's power efficiency. The power anomaly analysis module then uses historical data on the data center's power efficiency to train a time series prediction model, and finally uses the time series prediction model to predict the data center's future power consumption. The system uses power usage efficiency data over a period of time to predict the power usage efficiency of the data center. A power anomaly analysis module then performs anomaly analysis on the predicted and actual power usage efficiency of the data center to identify data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption. An alarm module, connected to the power anomaly analysis module, notifies the data center management to improve the data centers, ICT equipment, and / or non-ICT equipment identified by the power anomaly analysis module, thereby restoring normal power usage or reducing energy consumption.

[0006] The data center energy consumption integration management method of the present invention includes: an electricity data analysis module analyzing the power consumption of at least one data center's ICT equipment and non-ICT equipment; an electricity anomaly analysis module further analyzing the data center's power usage efficiency based on the power consumption of the ICT equipment and non-ICT equipment analyzed by the electricity data analysis module; the electricity anomaly analysis module training a time series prediction model using historical data of the data center's power usage efficiency; and using the time series prediction model to predict the data center's power usage efficiency for a future period to obtain the predicted power usage efficiency of the data center; the electricity anomaly analysis module performing anomaly analysis on the predicted power usage efficiency and the actual power usage efficiency of the data center to identify data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption; and an alarm module notifying the data center's management terminal to improve the data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption identified by the electricity anomaly analysis module to achieve normal power consumption or reduce energy consumption.

[0007] The computer-readable medium of the present invention is used in a computing device or computer, stores instructions, and executes the computer-readable medium through a processor and memory, so as to perform the above-mentioned data center energy consumption integration management method when the computer-readable medium is executed.

[0008] Therefore, the present invention provides an innovative data center energy consumption integrated management system, method and computer-readable medium, which uses a power data analysis module to automatically analyze the power consumption of data communication technology equipment and non-data communication technology equipment in the data center.

[0009] Furthermore, the power anomaly analysis module of this invention efficiently analyzes the power consumption efficiency of the computer room based on the power consumption of the ICT and non-ICT equipment. It also uses historical data on the power consumption efficiency of the computer room to accurately train a time series prediction model to predict the power consumption efficiency of the computer room in the future. It also performs anomaly analysis on the predicted and actual power consumption efficiency of the computer room to identify computer rooms, ICT equipment and / or non-ICT equipment with abnormal power consumption in real time.

[0010] In addition, the alarm module of this invention actively notifies the management terminal of the computer room to quickly improve the computer room, ICT equipment and / or non-ICT equipment with abnormal power consumption as analyzed by the power anomaly analysis module, thereby effectively achieving normal power consumption or reducing energy consumption.

[0011] To make the above-described features and advantages of the present invention more apparent and understandable, specific embodiments are described below in conjunction with the accompanying drawings. Additional features and advantages of the invention will be set forth in part in the following description, and such features and advantages will be partly apparent from the description or may be learned by practice of the invention. It should be understood that both the foregoing general description and the following detailed description are illustrative and explanatory, and are not intended to limit the scope of the invention. Simple Explanation of the Diagram

[0012] Figure 1 is a schematic diagram of the architecture of the data center energy consumption integrated management system of the present invention.

[0013] Figure 2 is a flowchart illustrating the data center energy consumption integration management method of the present invention.

[0014] Figure 3 is a flowchart illustrating the operation of the power data analysis module in the data center energy consumption integrated management system and method of the present invention.

[0015] Figure 4 is a flowchart illustrating the operation of the power anomaly analysis module in the data center energy consumption integrated management system and method of the present invention. Implementation

[0016] The following describes the implementation of the present invention through specific embodiments. Those skilled in the art can understand other advantages and effects of the present invention from the content disclosed in this specification, and can therefore implement or use it through other different specific equivalent embodiments.

[0017] Figure 1 is a schematic diagram of the architecture of the data center energy consumption integrated management system 1 of the present invention. In one embodiment, the data center energy consumption integrated management system 1 is an information and communication technology data center energy consumption integrated management system. As shown in Figure 1, the data center energy consumption integrated management system 1 may include an energy system interface module 10, a cloud database 20, a power data analysis module 30, a power anomaly analysis module 40, and an alarm module 50, which are interconnected and built on a cloud service platform 2 (such as Platform as a Service (PaaS)). It may also include at least one (such as multiple) data center management subsystems 60 and a billing management subsystem 70, and the power anomaly analysis module 40 may have a time series prediction model 41.

[0018] In one embodiment, the energy system interfacing module 10 can communicate with a cloud database 20, a power data analysis module 30, a data center management subsystem 60, a data center database 61, and a billing management subsystem 70, etc. The cloud database 20 can further communicate with the power data analysis module 30, etc. The power data analysis module 30 can further communicate with the power anomaly analysis module 40 and the alarm module 50, etc. The power anomaly analysis module 40 can further communicate with the alarm module 50, etc.

[0019] In one embodiment, the energy system interface module 10 may be an energy management interface module, an energy system interface, an energy system interface component, or energy system interface software (program). The cloud service platform 2, also known as a service cloud platform, and the cloud database 20 (or cloud database) may be a cloud data server, a cloud data storage device, or a cloud data hard drive. The power data analysis module 30 may be a power data analyzer (chip / circuit), power data analysis software (program), etc.; the power anomaly analysis module 40 may be a power anomaly analyzer (chip / circuit), power anomaly analysis software (program), etc.; and the alarm module 50 may be an event alarm module, an alarm device (chip / circuit), or alarm software (program), etc.

[0020] In one embodiment, the data center energy consumption integrated management system 1 is also referred to as an energy integrated data center management system, and ICT is also referred to as information and communication technology. The data center 80 can be an ICT equipment data center, and includes at least one (or multiple) ICT equipment 81 and at least one (or multiple) non-ICT equipment 82, etc. The ICT equipment 81 can be information equipment, communication equipment, switching mode rectifier (SMR), uninterruptible power supply (UPS), server, computer host, network equipment, etc. The non-ICT equipment 82 can be a main electricity meter 83, cooling equipment, lighting equipment, etc. The management terminal 90 can be an administrator (such as a data center administrator), maintenance personnel, a data center management terminal, and / or the electronic devices used (such as servers / communication devices / smartphones / computer hosts / tablets / laptops / personal computers), etc.

[0021] In this invention, "at least one" means one or more (e.g., one, two, or three or more), "plural" or "multiple" means two or more (e.g., two, three, four, or ten or more), and "communication connection" means mutual communication or connection through various forms such as data, signal, electrical, wired (e.g., wired network), or wireless (e.g., wireless network). However, this invention is not limited to those mentioned in the various embodiments.

[0022] The present invention can establish a data center energy consumption integrated management system 1 in the cloud environment of a cloud service platform 2 (such as Platform as a Service PaaS) to complete the construction of the energy system interface module 10, power data analysis module 30, power anomaly analysis module 40 and alarm module 50 of the data center energy consumption integrated management system 1. It can also use the cloud service platform 2 (such as Platform as a Service PaaS) to establish a cloud database 20.

[0023] This invention can use the energy system interface module 10 to develop various interface methods, such as message queue protocol, database interface method, or presentation layer state transition application interface (RESTful API), etc., to interface with the data center management subsystem 60, billing management subsystem 70, etc. through the energy system interface module 10, and then obtain various information such as the data center list of the data center management subsystem 60, the billing data of the billing management subsystem 70, the equipment list of the data center 80, and the power consumption data of the data center equipment (such as ICT equipment 81 and non-ICT equipment 82). Subsequently, the power data analysis module 30 analyzes (calculates) the power consumption of at least one computer room 80's computer room equipment (such as ICT equipment 81 and non-ICT equipment 82). Based on the power consumption of the ICT equipment 81 and non-ICT equipment 82 in the computer room 80 analyzed (calculated) by the power data analysis module 30, the power anomaly analysis module 40 further analyzes the power usage effectiveness (PUE) 84 of the computer room 80. Then, the power anomaly analysis module 40 uses the historical data of the power usage effectiveness (PUE) 84 of the computer room 80 to train the time series prediction model 41. Furthermore, the power anomaly analysis module 40 can define a power anomaly in the computer room 80 when the error (such as difference or error ratio) between the predicted and actual power usage efficiency (PUE) 84 of the computer room 80 exceeds the error threshold. The alarm module 50 then notifies the management terminal 90 of the computer room 80 to make improvements, thereby tracking and managing the computer room 80 with power anomalies.

[0024] In one embodiment, the error (such as difference or error ratio) between the predicted result and the actual result is notified to the alarm module 50, and the power usage efficiency (PUE) 84 and the error (such as difference or error ratio) are used as inputs by artificial intelligence technology to generate improvement values ​​or suggestions for the management terminal 90 to automatically make improvements.

[0025] In other words, this invention can deploy a data center energy consumption integrated management system 1 on a cloud service platform 2 (such as Platform as a Service, PaaS), that is, deploy a cloud database 20 on the cloud service platform 2 (such as Platform as a Service, PaaS) and develop a power data analysis module 30, a power anomaly analysis module 40 and an alarm module 50, and then through the energy system interface module 10, via message queue protocol, database interface method, or presentation layer state transition application interface (RESTful)... The system interface module 10 can access the data center management subsystem 60 and the billing management subsystem 70 through various interface methods such as API, so that the energy system interface module 10 can obtain the data center list of the data center management subsystem 60 (such as a detailed list including multiple data centers 80), the billing data of the billing management subsystem 70 (such as total electricity consumption), the equipment list of the data center 80 (such as a detailed list of data center equipment in multiple data centers 80), and the power consumption data of the data center equipment (such as ICT equipment 81 and non-ICT equipment 82).

[0026] Furthermore, with the increasing focus on the concept of sustainable management in enterprise ESG (Environmental, Social, and Governance), the energy consumption and carbon emission management of ICT equipment 81 in data center 80 has become quite important. Therefore, this invention proposes an integrated management architecture and method for energy consumption of ICT equipment. Through a cloud service platform 2 (such as Platform as a Service PaaS), a data center energy consumption integrated management system 1 is established to integrate the data center management subsystem 60. Then, the power anomaly analysis module 40 analyzes the power usage efficiency (PUE) 84 of data center 80 based on the power consumption data (power consumption information) of all data center equipment (such as ICT equipment 81 and non-ICT equipment 82). Next, the power anomaly analysis module 40 can use the historical data (historical information) of the accumulated power usage efficiency (PUE) 84 to train the time series prediction model 41, and then use the time series prediction model 41 to predict the power usage efficiency (PUE) 84 of the computer room 80 for a future period of time to obtain the prediction result of the power usage efficiency (PUE) 84 of the computer room 80. Then, the power anomaly analysis module 40 can perform anomaly analysis on the predicted results of the power usage efficiency (PUE) 84 of the computer room 80 and the actual results of the power usage efficiency (PUE) 84 of the computer room 80 to identify computer room 80, ICT equipment 81 and / or non-ICT equipment 82 with abnormal power consumption. The alarm module 50 then notifies the management terminal 90 of the computer room 80 to improve the computer room 80, ICT equipment 81 and / or non-ICT equipment 82 with abnormal power consumption identified by the power anomaly analysis module 40 (such as inspection / repair / update / replacement / removal, etc.) to achieve normal power consumption or reduce energy consumption. This is to avoid unnecessary power consumption caused by the anomalies of the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82), thereby achieving the goal of energy conservation and emission reduction, so as to meet the sustainable operation requirements of the enterprise's ESG (environmental, social and corporate governance).

[0027] Figure 2 is a flowchart illustrating the data center energy consumption integration management method of the present invention, and includes the technical contents of steps S11 to S15 of [1] to [3] below, and is described in conjunction with the data center energy consumption integration management system 1 shown in Figure 1. In one embodiment, the data center energy consumption integration management method is an information and communication technology data center energy consumption integration management method.

[0028] [1] Step S11: The energy system interfacing module 10 manages equipment data, electricity consumption data and billing data, and stores them in the cloud database 20. That is, the energy system interfacing module 10 can interfacing with various systems such as the data center management subsystem 60 and the billing management subsystem 70 through different interfacing methods. For example, the energy system interfacing module 10 can interfacing with (connect to) the data center database 61 of the data center management subsystem 60, so as to obtain the equipment data (equipment information) and electricity consumption data (such as daily electricity consumption data) of the data center management subsystem 60 (data center database 61) through the database connection thread. The billing management subsystem 70 can periodically transmit billing data (such as monthly electricity consumption, cost data and / or electricity consumption classification data) through the message queue protocol, so as to store (store) the equipment data and electricity consumption data of the data center management subsystem 60, the billing data of the billing management subsystem 70 and other data in the cloud database 20.

[0029] [2] Step S12: The power data analysis module 30 analyzes the power consumption of the equipment in the computer room and stores the power consumption data in the cloud database 20. That is, the energy system interface module 10 can transmit the power consumption data (such as daily power consumption data) of the equipment in the computer room (such as ICT equipment 81 and non-ICT equipment 82) to the power data analysis module 30, and then the power data analysis module 30 calculates the power consumption of the equipment in the computer room (such as ICT equipment 81 and non-ICT equipment 82) based on the power consumption data, so as to store the power consumption data of the equipment in the computer room (such as ICT equipment 81 and non-ICT equipment 82) in the cloud database 20.

[0030] [3] Steps S13 to S15: The power anomaly analysis module 40 performs predictive analysis of power consumption anomalies in the computer room 80 and / or the computer room equipment (see step S13) to determine whether the power consumption anomaly status of the computer room 80 and / or the computer room equipment has reached the anomaly standard (see step S14). If yes (the anomaly standard has been reached), the alarm module notifies the management terminal 90 of the computer room 80 to make improvements (see step S15). Otherwise, if no (the anomaly standard has not been reached), the process ends directly.

[0031] That is, the power anomaly analysis module 40 can establish a time series prediction model 41 to predict and analyze the real-time power consumption of the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82). Then, the alarm module 50 notifies the management terminal 90 of the power anomaly event of the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) for improvement, so that the management terminal 90 can effectively manage the power anomaly of the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82).

[0032] In one embodiment, the error (such as difference or error ratio) between the predicted result and the actual result is notified to the alarm module 50, and the power usage efficiency (PUE) 84 and the error (such as difference or error ratio) are used as inputs by artificial intelligence technology to generate improvement values ​​or suggestions for the management terminal 90 to automatically make improvements.

[0033] Figure 3 is a flowchart illustrating the operation of the power data analysis module 30 in the data center energy consumption integrated management system 1 and its method of the present invention, and includes the technical content of steps S21 to S30 of the following [1] to [4], and is described in conjunction with the data center energy consumption integrated management system 1 shown in Figure 1. In one embodiment, the data center energy consumption integrated management system 1 and its method are information and communication technology data center energy consumption integrated management systems and methods.

[0034] In order to address situations such as incomplete or abnormal data management of equipment data in the computer room management subsystem 60, the power data analysis module 30 (power data analysis method) of the present invention can perform the following steps S21 to S30 to achieve the goal (purpose) of effectively calculating the power consumption.

[0035] In one embodiment, the power data analysis module 30 can analyze the monitoring point names or monitoring point units of the equipment in the computer room 80 (such as ICT equipment 81 and non-ICT equipment 82) to analyze the equipment managed by the computer room 80. The power data analysis module 30 then extracts power consumption data that can be used to analyze the power consumption of the equipment in the computer room 80 (such as ICT equipment 81 and non-ICT equipment 82), performs summation or integration calculations, and notifies the management terminal 90 (such as the administrator) of the computer room 80 of any computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) that cannot be analyzed (cannot be calculated) for improvement / correction via the alarm module 50. Then, the accuracy is checked through the correlation between the power consumption of the billing management subsystem 70, the electricity meter master 83, and the computer room equipment (such as ICT equipment 81) to avoid calculation errors and achieve the goal of power consumption management.

[0036] [1] Step S21: The power data analysis module 30 can analyze whether the name of the monitoring point with power consumption or the monitoring point unit is kilowatt-hour (kWh; commonly known as one kilowatt-hour) or megawatt-hour (MWh). If not, proceed to step S22; otherwise, proceed directly to step S27.

[0037] That is, the power data analysis module 30 can filter out the ICT equipment 81 (such as switched-mode DC power supply equipment, uninterruptible power supply system (UPS)) and the main electricity meter 83 from the multiple computer room equipment managed by the computer room 80. The power data analysis module 30 can then analyze the ICT equipment 81 and the main electricity meter 83 with the name of the monitoring point or the monitoring point unit being kilowatt-hour (kWh=1000 watt-hour) or megawatt-hour (MWh=1000000 watt-hour) and then the power data analysis module 30 can confirm whether the calculation of the power consumption of the ICT equipment 81 and the main electricity meter 83 with kilowatt-hour (kWh) or megawatt-hour (MWh) has been completed.

[0038] [2] Steps S22 to S23: The power data analysis module 30 can analyze whether the name of the monitoring point with power consumption or the monitoring point unit is kilowatt (kW). If yes, proceed to step S23; otherwise, proceed to step S24.

[0039] That is, if the power data analysis module 30 cannot calculate the power consumption of the ICT equipment 81 and the main meter 83 with monitoring point names or monitoring point units of kilowatt-hours (kWh) or megawatt-hours (MWh), then the power data analysis module 30 can filter out the ICT equipment 81 and the main meter 83 with monitoring point names or monitoring point units of kilowatts (kW=1000 watts). Next, as shown in step S23, the power data analysis module 30 can calculate the power consumption of the ICT equipment 81 and the main meter 83 with kilowatts (kW) through a first integral calculation method (such as a first integral formula), and the first integral calculation method (such as a first integral formula) is: power consumption of the ICT equipment 81 and the main meter 83 with kilowatts (kW) = , where p(t) is a function of instantaneous power that varies with time (t).

[0040] [3] Steps S24 to S26: The power data analysis module 30 can analyze whether the monitoring point name or monitoring point unit has voltage and current of volts (V) and amperes (A). If yes, proceed to step S25, so that the power data analysis module 30 can calculate the power based on the voltage and current; otherwise, if no, proceed directly to step S26, so that the power data analysis module 30 can record the abnormal information and communication technology equipment 81 and the electricity meter master 83.

[0041] That is, if the methods provided in steps S21 to S22 above for kilowatt-hour (kWh), megawatt-hour (MWh), and kilowatt (kW) cannot produce results, the power data analysis module 30 can filter out the ICT equipment 81 and the main meter 83 with monitoring point names of voltage (V) and current (i) from the multiple computer room equipment managed by the computer room 80, or the ICT equipment 81 and the main meter 83 with monitoring point units of volt (V) and ampere (A). Next, as shown in step S23, the power data analysis module 30 can calculate the power consumption of the ICT device 81 and the main meter 83, which have voltage (V) and current (i), or volts (V) and amperes (A), through a second integral calculation method (such as a second integral formula). The second integral calculation method (such as a second integral formula) is: power consumption of the ICT device 81 and the main meter 83, which have voltage (V) and current (i), or volts (V) and amperes (A) = , where v(t) and i(t) are functions of voltage and current that change with time (t), respectively.

[0042] If the methods provided in steps S21, S22 and S24 above regarding kilowatt-hours (kWh), megawatt-hours (MWh), kilowatts (kW), volts (V), and amperes (A) are unable to analyze the electricity consumption of the ICT equipment 81 and the main electricity meter 83, then the alarm module 50 will notify the management terminal 90 (such as an administrator) to improve the ICT equipment 81 and the main electricity meter 83.

[0043] [4] Steps S27 to S30: The power data analysis module 30 can import a calculation correctness verification mechanism (see step S27) to check whether the calculated power consumption of the main meter 83 is normal or abnormal through the power consumption recorded by the billing management subsystem 70 (see step S28). Then, the power data analysis module 30 compares the power consumption of the ICT equipment 81 and the main meter 83 to obtain whether the comparison result is normal or abnormal (see step S29). If it is normal, the process ends directly; otherwise, if it is abnormal, the alarm module 50 notifies the management terminal 90 of the computer room 80 to make improvements (see step S30).

[0044] That is, the power data analysis module 30 can incorporate a calculation accuracy verification mechanism. The power data analysis module 30 compares the electricity consumption (e.g., electricity kilowatt-hours) recorded during the billing period by the billing management subsystem 70 with the sum of the electricity consumption of the main meter 83 during the system calculation period. Then, the power data analysis module 30 sums the electricity consumption of all equipment in the same computer room 80 (e.g., ICT equipment 81 and non-ICT equipment 82). The sum of the electricity consumption of all equipment in the same computer room 80 (e.g., ICT equipment 81 and non-ICT equipment 82) is compared with the electricity consumption of the main meter 83 of the computer room 80. ICT equipment 81 and / or non-ICT equipment 82 whose electricity consumption error with the main meter 83 exceeds the error threshold is identified as having abnormal power consumption. The alarm module 50 then notifies the management terminal 90 of the computer room 80 with abnormal power consumption for improvement.

[0045] Figure 4 is a flowchart illustrating the operation of the power anomaly analysis module 40 in the data center energy consumption integrated management system 1 and its method of the present invention, and includes the technical contents of steps S31 to S39 of [1] to [9] below, and is described together with reference to the data center energy consumption integrated management system 1 shown in Figure 1. In one embodiment, the data center energy consumption integrated management system 1 and its method are information and communication technology data center energy consumption integrated management systems and methods.

[0046] In one embodiment, the power anomaly analysis module 40 can analyze the power usage efficiency (PUE) 84 of the computer room 80 through the power consumption data of the information and communication technology equipment 81 and the main electricity meter 83, and use this data to train the time series prediction model 41 to predict the power usage efficiency (PUE) 84 of the computer room 80 for a future period of time. When the error between the predicted power usage efficiency (PUE) 84 and the actual result exceeds the error threshold, the power anomaly analysis module 40 can determine that the computer room 80 has an abnormal power consumption. Then, through the time series prediction model 41, it analyzes the power consumption status of the computer room equipment in this computer room 80 one by one, and then filters out the computer room equipment with large power consumption differences, so as to trigger the alarm module 50 to notify the management terminal 90 of the computer room 80 to make improvements.

[0047] [1] Step S31: The power anomaly analysis module 40 determines whether it can obtain the list of information and communication technology equipment 81 and electricity meter 83 in the computer room 80, and obtain (read) the power consumption data (such as power consumption / daily power consumption) of information and communication technology equipment 81 and electricity meter 83. If not (cannot obtain), proceed to step S32; otherwise, if yes (can obtain), proceed to step S34.

[0048] [2] Step S32: When the power consumption data (such as power consumption / daily power consumption) of the ICT device 81 and the main meter 83 cannot be obtained, the power anomaly analysis module 40 can determine whether the ICT device 81 and the main meter 83 exist. If they exist, proceed to step S33; otherwise, if they do not exist, proceed to step S39.

[0049] [3] Step S33: If the ICT device 81 or the main electricity meter 83 exists, the power anomaly analysis module 40 can determine whether it can retrieve the electricity data of the ICT device 81 or the main electricity meter 83 from the computer room management subsystem 60 and complete the calculation of electricity consumption. If yes, proceed to step S34; otherwise, proceed to step S39.

[0050] [4] Step S34: The power anomaly analysis module 40 can calculate the power usage efficiency (PUE) 84 of the computer room 80 and store the power usage efficiency (PUE) 84 data to the cloud database 20. For example, the formula for the power usage efficiency (PUE) 84 of the computer room 80 is: Power usage efficiency (PUE) = Total power consumption of the computer room / Total power consumption of ICT equipment. If the power anomaly analysis module 40 fails to obtain the power usage efficiency (PUE) 84 of the computer room 80, the power anomaly analysis module 40 can notify the management terminal 90 of the computer room 80 through the alarm module 50 to make improvements according to the situation (such as the absence of this ICT equipment 81, data loss).

[0051] [5] Step S35: When the power usage efficiency (PUE) 84 of the computer room 80 has accumulated enough samples and exceeded the quantity threshold, the power anomaly analysis module 40 can use the power usage efficiency (PUE) 84 of the computer room 80 to train the time series prediction model 41, so as to obtain the predicted value of the power usage efficiency (PUE) 84 of the computer room 80 through the time series prediction model 41.

[0052] [6] Step S36: The power anomaly analysis module 40 can determine whether the calculated power usage efficiency (PUE) 84 of the computer room 80 in the future exceeds the threshold value of the predicted power usage efficiency (PUE) 84 of the time series prediction model 41. If yes (representing an anomaly), proceed to step S37. Otherwise, if no (representing normal), end directly.

[0053] [7] Step S37: When the power anomaly analysis module 40 analyzes (determines) that the calculated value of the power usage efficiency (PUE) 84 of the computer room 80 in the future exceeds the threshold value of the predicted value of the power usage efficiency (PUE) 84 of the time series prediction model 41, the power anomaly analysis module 40 can train the time series prediction model 41 based on the historical power consumption of the ICT equipment 81 and the non-ICT equipment 82 of the computer room 80 through the time series training model.

[0054] [8] Step S38: The power anomaly analysis module 40 can use the time series prediction model 41 to predict whether the daily electricity consumption and predicted electricity consumption of the ICT equipment 81 and non-ICT equipment 82 (such as the main electricity meter 83) exceed the threshold value. If yes (representing an anomaly), proceed to step S39. Otherwise, if no (representing normal), end directly.

[0055] [9] Step S39: The power anomaly analysis module 40 can filter out the computer room equipment (such as ICT equipment 81 and / or non-ICT equipment 82) whose daily power consumption and predicted power consumption exceed the threshold value in the computer room 80 to determine the computer room 80 and / or computer room equipment with power anomalies. Then, the alarm module 50 notifies the management terminal 90 (such as the administrator) of the computer room 80 to make improvements, so that the management terminal 90 (such as the administrator) of the computer room 80 can detect the computer room equipment (such as ICT equipment 81 and / or non-ICT equipment 82) with power anomalies.

[0056] For example, the present invention uses a data center energy consumption integrated management system 1 as an example, which adds related services such as an energy system interfacing module 10 (such as an energy management interfacing module), a cloud database 20, a power data analysis module 30, a power anomaly analysis module 40, and an alarm module 50 to a cloud service platform 2 (such as a platform as a service PaaS).

[0057] For example, the energy system interfacing module 10 can use a database to interfacing with the data center management subsystem 60, etc., to obtain the data center list, equipment list of the data center 80, and power consumption data (such as daily power consumption data) of the data center equipment (such as ICT equipment 81 and non-ICT equipment 82) of the data center management subsystem 60. At the same time, the energy system interfacing module 10 can also use a message queueing protocol to listen for and receive the billing information periodically transmitted by the billing management subsystem 70, so that the energy system interfacing module 10 can analyze the electricity consumption, electricity cost, allocation ratio and / or electricity type of the billing information transmitted by the billing management subsystem 70.

[0058] When the energy system interfacing module 10 receives the electricity consumption data (power consumption) of the computer room 80 for the previous day, the energy system interfacing module 10 can start the power data analysis module 30 to analyze the power consumption of the computer room 80 (such as the first computer room). Relevant examples are shown in Table 1 below.

[0059] Table 1: List of equipment in computer room 80 (e.g., computer room 1 or computer room 2) (names of equipment in computer room), names of monitoring points and units of monitoring points.

[0060] In the embodiments in Table 1 above, the equipment in the computer room 80 (such as the first computer room) (such as ICT equipment 81 and non-ICT equipment 82) may include a main electricity meter 83, two switched-mode DC power supplies (SMRs) and two uninterruptible power supplies (UPSs). The main electricity meter 83 may be a non-ICT equipment 82, and the switched-mode DC power supplies (SMRs) and uninterruptible power supplies (UPSs) may be ICT equipment 81.

[0061] [1] The monitoring point name of the main meter 83 in the computer room 80 includes the electricity consumption, so the power data analysis module 30 can directly accumulate the daily electricity consumption of the main meter 83.

[0062] [2] The first switching DC power supply device (SMR) in the computer room 80 has paired voltage and current (i.e., paired output voltage and output current), so the power data analysis module 30 can calculate the power consumption using the formula shown in the second integral calculation method (such as the second integral formula). Calculate the daily power consumption (e.g., total daily power consumption) of the first switched DC power supply device (SMR).

[0063] [3] The first uninterruptible power supply (UPS) in the computer room 80 has the effective power of the monitoring point name, so the power data analysis module 30 can calculate the power consumption using the formula shown in the first integral calculation method (such as the first integral formula) = Calculate the daily power consumption of the first uninterruptible power supply (UPS).

[0064] [4] The monitoring point names of the second switched DC power supply (SMR) in the computer room 80 and [5] the second uninterruptible power supply (UPS) in the computer room do not have power consumption, effective power, and voltage and current cannot be paired (i.e. there is no paired output voltage and output current). Therefore, the power data analysis module 30 can notify the management terminal 90 of the computer room 80 through the alarm module 50 to check the monitoring point names of the second switched DC power supply (SMR) and the second uninterruptible power supply (UPS) in the computer room 80.

[0065] The power data analysis module 30 can periodically check the accuracy of the power consumption of the computer room 80, and sum and verify the power consumption during the billing cycle calculation period through the billing management subsystem 70 to improve the accuracy of the calculated power consumption. The power data analysis module 30 can also sum the power consumption of all computer room equipment (such as ICT equipment 81 and non-ICT equipment 82), and then compare the power consumption of ICT equipment 81 with that of the main electricity meter 83 to verify the rationality of the power consumption. Then, the power data analysis module 30 can notify the management terminal 90 of the computer room 80 through the alarm module 50 to make improvements if the analyzed (calculated) computer room 80 has any abnormalities.

[0066] The power anomaly analysis module 40 can perform power anomaly analysis on at least one (or multiple) computer rooms 80 to confirm the power consumption of each computer room 80. If there are computer room devices (such as ICT equipment 81 and non-ICT equipment 82) with missing power consumption data, it checks whether the computer room devices can be re-managed and whether the power consumption data can be recovered. If there is missing power consumption data, the power anomaly analysis module 40 can notify the management terminal 90 of the computer room 80 through the alarm module 50 for improvement.

[0067] The power anomaly analysis module 40 can calculate the power usage efficiency (PUE) 84 of the computer room 80 based on the power consumption of the computer room equipment (such as ICT equipment 81 and non-ICT equipment 82). For example, if the daily power consumption of the main meter 83 is 200 kWh, the power consumption of the uninterruptible power supply (UPS) is 90 kWh, and the power consumption of the switched-mode DC power supply (SMR) is 95 kWh, then the power usage efficiency (PUE) 84 of this computer room 80 is 200 / (90+95) = 1.081. When the cumulative number of power usage efficiency (PUE) 84 of this computer room 80 exceeds the threshold value of a predetermined time interval (such as four weeks), the power anomaly analysis module 40 can train the time series prediction model 41 to predict the power usage efficiency (PUE) 84 of the computer room 80 in the future (such as next week, next day, next month, etc.). If the error (such as difference or error ratio) between the predicted power usage efficiency (PUE) 84 of the computer room 80 by the time series prediction model 41 and the actual result exceeds the error threshold (such as 20%), then the power anomaly analysis module 40 can determine that the computer room 80 and / or the computer room equipment are experiencing power anomalies.

[0068] List 2 below is an example of a data center 80. The power anomaly analysis module 40 can use the historical data of power usage efficiency (PUE) 84 of data center 80 over a past period (such as four weeks) to train the time series prediction model 41 through a Long Short-Term Memory (LSTM) network, store the time series prediction model 41 in the cloud database 20, and then use the time series prediction model 41 to predict the value of power usage efficiency (PUE) 84 of data center 80 in a future period (such as next week, next day, next month, etc.).

[0069] Table 2: Historical data, forecast results (e.g., forecast values), actual results, and error ratios (e.g., error / difference between forecast and actual results) for Power Usage Effectiveness (PUE) 84 of computer room 80 (e.g., the first or second computer room).

[0070] When the error (such as difference or error ratio) between the preset result and the actual result of Power Usage Effectiveness (PUE) 84 exceeds the error threshold (such as 20%), as shown in Table 2 above for Thursday and Friday, the power anomaly analysis module 40 can obtain the power consumption data of the ICT equipment 81 or non-ICT equipment 82 managed by the computer room 80 for a period of time (such as one week). The power anomaly analysis module 40 then trains the time series prediction model 41 using the power consumption data for a period of time (such as one week) through a Long Short-Term Memory (LSTM) network. Then, the power anomaly analysis module 40 can predict the power consumption of the computer room 80 on Thursday and Friday of the current week through the time series prediction model 41, and select the ICT equipment 81 or non-ICT equipment 82 with the largest difference in power consumption as anomaly analysis devices. The alarm module 50 then notifies the management terminal 90 (such as the administrator) to check the various computer room devices such as the ICT equipment 81 or non-ICT equipment 82 with abnormal power consumption.

[0071] Furthermore, the present invention also provides a computer-readable medium for a data center energy consumption integration management method, which is applied in a computing device or computer having a processor and memory. The computer-readable medium stores instructions, and the computing device or computer can execute the computer-readable medium through the processor and memory to perform the content of the aforementioned data center energy consumption integration management method when the computer-readable medium is executed. In another embodiment, this computer-readable medium is a non-transitory computer-readable storage medium.

[0072] In one embodiment, the processor may be a central processing unit (CPU), graphics processing unit (GPU), microprocessor (MPU), microcontroller (MCU), artificial intelligence (AI) processor, etc., and the memory may be random access memory (RAM), read-only memory (ROM), flash memory, memory card, hard disk (such as cloud / network / external hard disk), optical disk, USB flash drive, database, etc., and the computing device or computer may be a computer, smartphone, tablet computer, personal computer, notebook computer, desktop computer, server (such as cloud / remote / network / AI server), etc.

[0073] In one embodiment, the modules can be software, hardware, or firmware; if they are hardware, they can be processing units, processors, computers, or servers with data processing and computing capabilities; if they are software or firmware, they can include instructions executable by processing units, processors, computers, or servers, and can be installed on the same hardware device or distributed across different multiple hardware devices.

[0074] In summary, the data center energy consumption integrated management system, method, and computer-readable medium of the present invention have at least the following features, advantages, or technical effects.

[0075] 1. The power data analysis module 30 of the present invention can analyze the computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) managed by the computer room 80. It analyzes the equipment using the monitoring point name or monitoring point unit, extracts data that can be used for power consumption analysis, performs summation or integration, and then notifies the management terminal 90 of the computer room 80 of the computer room for correction through the alarm module 50 for computer room equipment that cannot be analyzed (cannot be calculated). The correctness is checked through the correlation between the power consumption of the billing management subsystem 70, the electricity meter master meter 83 and the computer room equipment (such as ICT equipment 81) to avoid calculation errors and achieve the goal of power consumption management.

[0076] Second, the present invention can make good use of the power data analysis module 30 and the power anomaly analysis module 40 to complete the energy consumption management of the ICT equipment 81. It can also effectively connect various types of data center management subsystems 60 through the cloud service platform 2 (such as Platform as a Service PaaS). It can also quickly store the data center list of the data center management subsystem 60, the billing data of the billing management subsystem 70, the equipment list of the data center 80, and the power consumption data of the data center equipment (such as ICT equipment 81 and non-ICT equipment 82) to the cloud database 20.

[0077] Third, the power data analysis module 30 of the present invention can effectively analyze / calculate the power consumption of computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) through the monitoring point name and monitoring point unit. It can also import the power consumption of billing management subsystem 70, the correlation between the electricity meter master meter 83 and computer room equipment (such as ICT equipment 81) to verify the accuracy of the calculated power consumption.

[0078] IV. The power anomaly analysis module 40 of this invention can automatically analyze the power usage efficiency (PUE) 84 of the computer room 80 through the power consumption data of the information and communication technology equipment 81 and the main electricity meter 83, so as to efficiently train the time series prediction model 41 to predict the power usage efficiency (PUE) 84 of the computer room 80 for a future period of time. When the error between the predicted power usage efficiency (PUE) 84 and the actual result exceeds the error threshold, the power anomaly analysis module 40 can also determine that the computer room 80 has abnormal power consumption. Then, through the time series prediction model 41, it quickly analyzes the power consumption status of the computer room equipment in this computer room 80 one by one, and then immediately filters out the computer room equipment with large power consumption differences to trigger the alarm module 50 to notify the management terminal 90 of the computer room 80 for improvement.

[0079] V. The power anomaly analysis module 40 of the present invention can calculate the power usage efficiency (PUE) 84 of the computer room 80 based on the calculated power consumption. After the power usage efficiency (PUE) 84 has accumulated a sufficient number of samples (such as the number of samples exceeding the quantity threshold), the time series prediction model 41 is trained and completed. Then, the power usage efficiency (PUE) 84 of the computer room 80 in the future is predicted through the time series prediction model 41, so as to effectively determine whether the actual power usage efficiency (PUE) 84 of the computer room 80 may be abnormal.

[0080] VI. When the power usage efficiency (PUE) 84 of the computer room 80 is abnormal, the power anomaly analysis module 40 of the present invention can automatically analyze the computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) of the computer room 80 through the time series prediction model 41, so as to efficiently screen out the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) with abnormal power consumption. It can also notify the management terminal 90 of the computer room 80 (such as the administrator) through the alarm module 50 to improve the computer room 80 and / or computer room equipment with abnormal power consumption, thereby avoiding unnecessary power consumption of the computer room 80 and / or computer room equipment.

[0081] VII. The power anomaly analysis module 40 of the present invention can perform anomaly analysis based on the predicted power usage efficiency (PUE) 84 of the computer room 80 for a future period of time and the actual power usage efficiency (PUE) 84. It can also notify the management terminal 90 (such as the administrator) of the computer room 80 (such as the administrator) of the computer room 80 (such as the computer room 80 and / or computer room equipment (such as ICT equipment 81 and non-ICT equipment 82) with abnormal power consumption through the alarm module 50 so as to make improvements, so as to avoid unnecessary power consumption caused by the abnormality of the computer room 80 and / or computer room equipment, thereby achieving the goal of energy conservation and emission reduction, so as to meet the sustainable operation requirements of the enterprise's ESG (environmental / social / government) requirements.

[0082] The above embodiments are merely illustrative of the principles, features, and effects of the present invention and are not intended to limit the scope of implementation of the present invention. Anyone skilled in the art can modify and alter the above embodiments without departing from the spirit and scope of the present invention. Any equivalent changes and modifications made using the content disclosed in this invention should still be covered by the claims. Therefore, the scope of protection of this invention should be as listed in the claims.

[0083] 1: Data Center Energy Consumption Integrated Management System 2: Cloud service platform 10: Energy System Interconnection Module 20: Cloud-based database 30: Power Data Analysis Module 40: Power Anomaly Analysis Module 41: Time Series Forecasting Model 50: Alarm Module 60: Computer Room Management Subsystem 61: Computer Room Database 70: Billing Management Subsystem 80: Computer Room 81: Information and communication technology equipment 82: Non-Telecommunications Technology Equipment 83: Electricity Meter Master 84: Electricity usage efficiency 90: Management Terminal S11 to S15: Steps S21 to S30: Steps S31 to S39: Steps

Claims

1. A data center energy consumption integrated management system, comprising: A power data analysis module analyzes the power consumption of ICT and non-ICT equipment in at least one data center. A power anomaly analysis module is connected to the power data analysis module. Based on the power consumption data analyzed by the power data analysis module, the power anomaly analysis module further analyzes the power utilization efficiency of the data center. The power anomaly analysis module then trains a time series prediction model using historical data on the data center's power utilization efficiency. This time series prediction model is then used to predict the data center's power utilization efficiency for a future period, thus obtaining a prediction result for the data center's power utilization efficiency. The power anomaly analysis module performs anomaly analysis on the predicted and actual power usage efficiency of the data center to identify data center, ICT equipment, and / or non-ICT equipment with abnormal power usage; and an alarm module is connected to the power anomaly analysis module to notify the data center management so that improvements can be made to the data center, ICT equipment, and / or non-ICT equipment with abnormal power usage as identified by the power anomaly analysis module to achieve normal power usage or reduce energy consumption.

2. The data center energy consumption integrated management system as described in claim 1, wherein, The power data analysis module further uses the monitoring point names or monitoring point units of the ICT equipment and the non-ICT equipment to analyze the ICT equipment and the non-ICT equipment managed by the computer room. Then, the power data analysis module performs summation or integration calculations on the power consumption data that can be used to analyze the ICT equipment and the non-ICT equipment.

3. The data center energy consumption integrated management system as described in claim 1, wherein, The power data analysis module further filters out the information and communication technology equipment and the main electricity meter from the multiple equipment in the computer room under its management. The power data analysis module then analyzes the monitoring point names or monitoring point units in kilowatt-hours or megawatt-hours for the information and communication technology equipment and the main electricity meter.

4. The data center energy consumption integrated management system as described in claim 1, wherein, When the power data analysis module is unable to calculate the power consumption of the ICT equipment and the main meter with monitoring point names or monitoring point units of kilowatts per hour or megawatts per hour, the power data analysis module filters out the ICT equipment and the main meter with monitoring point names or monitoring point units of kilowatts, and then calculates the power consumption of the ICT equipment and the main meter with kilowatts through an integral calculation method.

5. The data center energy consumption integrated management system as described in request item 1, wherein, The power data analysis module further filters out the ICT devices and main meters with monitoring point names of voltage and current, or monitoring point units of volts and amperes, from the multiple equipment managed by the computer room. Then, the power data analysis module calculates the power consumption of the ICT devices and main meters with voltage and current, or volts and amperes, through an integral calculation method.

6. The data center energy consumption integrated management system as described in claim 1, wherein, The power data analysis module further sums up the power consumption of all ICT devices and non-ICT devices in the same computer room, and then compares the sum of the power consumption of all ICT devices and non-ICT devices with the power consumption of the computer room's main electricity meter. This allows ICT devices and / or non-ICT devices whose power consumption difference from the main electricity meter exceeds an error threshold to be identified as having abnormal power consumption.

7. The data center energy consumption integrated management system as described in claim 1 further includes a cloud database, so that the power anomaly analysis module can use historical data of the data center's power usage efficiency over a past period to train the time series prediction model through a long short-term memory network, and then the power anomaly analysis module can store the time series prediction model in the cloud database so that the power usage efficiency of the data center over a future period can be predicted through the time series prediction model.

8. The data center energy consumption integrated management system as described in claim 1 further includes an energy system interface module, a data center management subsystem, and a billing management subsystem. The energy system interface module interfaces with the data center management subsystem and the billing management subsystem through a message queue protocol, a database interface, or a presentation layer state transition application interface. The energy system interface module then obtains the data center list of the data center management subsystem, the billing data of the billing management subsystem, the equipment list of the data center, the power consumption data of the ICT equipment, and the power consumption data of the non-ICT equipment.

9. A method for integrated energy consumption management in a computer room, comprising: A power data analysis module analyzes the power consumption of ICT equipment and non-ICT equipment in at least one computer room; A power anomaly analysis module further analyzes the power consumption of the data center's ICT and non-ICT equipment based on the power data analysis module to determine the data center's power efficiency. The module then trains a time series prediction model using historical power efficiency data. This model predicts the data center's power efficiency for a future period, providing a forecast. The module then performs anomaly analysis on the predicted and actual power efficiency results to identify data centers, ICT equipment, and / or non-ICT equipment with abnormal power consumption. An alarm module notifies the data center's management to implement improvements to address these anomalies, achieving normal power consumption or reducing energy consumption.

10. The data center energy consumption integration management method as described in claim 9 further includes the power data analysis module using the monitoring point names or monitoring point units of the ICT equipment and the non-ICT equipment to analyze the ICT equipment and the non-ICT equipment under the management of the data center, and then the power data analysis module summing or integrating the power consumption data that can be used to analyze the power consumption of the ICT equipment and the non-ICT equipment.

11. The data center energy consumption integration management method as described in claim 9 further includes the power data analysis module filtering out the information and communication technology equipment and the main electricity meter from the multiple data center equipment managed by the data center, and the power data analysis module analyzing the information and communication technology equipment and the main electricity meter with the name of the monitoring point or the monitoring point unit of kilowatt-hour or megawatt-hour.

12. The data center energy consumption integration management method as described in claim 9 further includes, when the power data analysis module is unable to calculate the power consumption of the ICT equipment and the main electricity meter with the monitoring point name or the monitoring point unit being kilowatt-hour or megawatt-hour, the power data analysis module filters out the ICT equipment and the main electricity meter with the monitoring point name or the monitoring point unit being kilowatt, and then the power data analysis module calculates the power consumption of the ICT equipment and the main electricity meter with kilowatts through an integral calculation method.

13. The data center energy consumption integration management method as described in claim 9 further includes the power data analysis module filtering out the ICT equipment and the main electricity meter with monitoring point names of voltage and current, or the ICT equipment and the main electricity meter with monitoring point units of volts and amperes, from the plurality of data center equipment managed by the data center, and then the power data analysis module calculating the power consumption of the ICT equipment and the main electricity meter with voltage and current, or volts and amperes, through an integral calculation method.

14. The data center energy consumption integration management method as described in claim 9 further includes having the power data analysis module sum up the power consumption of all ICT devices and non-ICT devices in the same data center, and then comparing the sum of the power consumption of all ICT devices and non-ICT devices with the power consumption of the data center's main electricity meter, so that ICT devices and / or non-ICT devices whose power consumption difference with the main electricity meter exceeds the error threshold value are judged as having abnormal power consumption.

15. The data center energy consumption integration management method as described in claim 9 further includes the power anomaly analysis module using historical data of the data center's power usage efficiency over a past period to train the time series prediction model through a long short-term memory network, and then the power anomaly analysis module storing the time series prediction model in a cloud database so as to predict the value of the data center's power usage efficiency over a future period through the time series prediction model.

16. The data center energy consumption integration management method as described in claim 9 further includes an energy system interface module connecting a data center management subsystem and a billing management subsystem through a message queue protocol, a database interface, or a presentation layer state transition application interface, and then the energy system interface module obtaining the data center list of the data center management subsystem, the billing data of the billing management subsystem, the equipment list of the data center, the power consumption data of the ICT equipment, and the power consumption data of the non-ICT equipment.

17. A computer-readable medium, applied in a computing device or computer, storing instructions and executing the computer-readable medium via a processor and memory, so as to perform the data center energy consumption integration management method as described in any one of claims 9 to 16 when the computer-readable medium is executed.

Images