Data center energy efficiency online monitoring method based on neural network

Abstract

The invention discloses a data center energy efficiency online monitoring method based on a neural network, relates to the field of center energy management, and solves the problem that the electric energy utilization rate of a data center cannot be efficiently monitored. Comprising the steps that IT power consumption data of IT equipment in a data center, real-time lighting power consumption of lighting equipment, refrigeration power consumption data of refrigeration equipment and refrigeration influence data of the data center are collected; iT theoretical power consumption corresponding to the IT devices and refrigerating capacity corresponding to the refrigeration devices are calculated; abnormal data processing is carried out on the power consumption data of the IT equipment, the refrigeration equipment and the lighting equipment in the data center; calculating the electric energy utilization efficiency of the data center according to the power consumption conditions of the IT equipment, the refrigeration equipment and the lighting equipment corresponding to the data center, and then calculating the PUE predicted value of the next day; and the user terminal receives the electric energy utilization rate of the data center and the PUE predicted value of the next day. According to the invention, the electric energy utilization rate of the data center is efficiently monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of data center energy management, and in particular to an online monitoring method for data center energy efficiency based on a neural network. Background Art

[0002] Data centers are IT infrastructures used to maintain thousands or even tens of thousands of servers. They consist of hardware entities such as power systems, cooling systems, servers, and software services such as virtual machine applications. Data centers are key infrastructures in the modern information society, supporting a variety of online services, data processing, and storage needs, and play an important role in ensuring the stable operation of the digital economy. With the rapid development of information technologies such as cloud computing, big data, and artificial intelligence, and the digital transformation of industries, data centers are increasingly evolving towards high power density, and power efficiency is currently the most widely used indicator for measuring the energy efficiency of data centers.

[0003] However, at present, when evaluating the power utilization rate of data centers, the power consumption data corresponding to IT equipment, refrigeration equipment and lighting equipment in the data center are too scattered, forcing existing operation and maintenance personnel to integrate power data by manual meter reading and then perform statistical calculations of indicators such as PUE, which results in extremely low management efficiency. To this end, the present invention proposes a data center energy efficiency online monitoring method based on neural network. Summary of the invention

[0004] The purpose of the present invention is to propose a method for online monitoring of data center energy efficiency based on a neural network, so as to solve the problem in the above background technology that the power utilization rate of the data center cannot be efficiently monitored.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A method for online monitoring of data center energy efficiency based on a neural network, comprising: Step S1, collecting IT power consumption data of multiple IT devices corresponding to the data center, real-time lighting power consumption of multiple lighting devices, refrigeration power consumption data corresponding to multiple refrigeration devices, and refrigeration impact data corresponding to the data center; Step S2, calculating IT theoretical power consumption corresponding to the plurality of IT devices and cooling capacity corresponding to the plurality of cooling devices; Step S3, performing abnormal data processing on the power consumption data of IT equipment, refrigeration equipment and lighting equipment in the data center; Step S4, calculating the power utilization efficiency of the data center according to the power consumption of multiple IT equipment, refrigeration equipment and lighting equipment corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through the PUE prediction model; Step S5, the user terminal receives the power utilization rate corresponding to the data center and the power utilization rate forecast value for the next day. The user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center.

[0006] Furthermore, IT power consumption data includes the rated power, real-time IT power consumption and working hours corresponding to multiple IT devices; refrigeration power consumption data includes the rated power and real-time refrigeration power consumption corresponding to multiple refrigeration equipment; refrigeration impact data includes the floor space, indoor temperature and outdoor temperature corresponding to the data center.

[0007] Furthermore, the step S2 includes the following sub-steps: Step S21, obtaining the rated power IEDi and the working time ISCi of the multiple IT devices corresponding to the data center; wherein i is the number of the multiple IT devices, i=1, 2, ..., y, y is a positive integer; Step S22, the total IT theoretical power consumption ITL of the multiple IT devices corresponding to the data center is calculated by a formula, and the specific formula is: .

[0008] Furthermore, the step S2 further includes the following sub-steps: Step S23, obtaining the floor area MJ corresponding to the data center, the indoor temperature value, and the rated electric power ZEDj corresponding to the plurality of refrigeration equipment, where j is the serial number of the plurality of refrigeration equipment, j=1, 2, ..., z, and z is a positive integer; Step S24, the cooling capacity ZLL corresponding to the multiple cooling devices in the data center is calculated by a formula, and the specific formula is: ; The unit of ZLL is KW, GLX is the power factor, which indicates the utilization efficiency of electric energy; DFR is the heating coefficient of electric energy, that is, the efficiency of converting electric energy into heat energy; HFR is the ambient heat value, which is determined by the temperature difference between the indoor temperature value and the outdoor temperature value.

[0009] Furthermore, the step S3 includes the following sub-steps: Step S31, obtaining the historical power consumption LHDid corresponding to the plurality of IT devices in the previous month, wherein d is the number of the day in the previous month, the upper limit of d is x, x=28, 29, 30 or 31; the historical power consumption is the daily power consumption of the corresponding IT device in the previous month; Step S32, summing up the historical power consumption corresponding to all IT devices in the previous month to obtain the average value, and then dividing it by the number of days corresponding to the previous month to obtain the historical daily power consumption average value JHD of the IT devices; Step S33, calculating the standard deviation BZCi of the power consumption corresponding to the multiple IT devices by using the formula, the specific formula is: .

[0010] Furthermore, the step S3 also includes the following sub-steps: Step S34, obtain the first IT power consumption limit value by subtracting the minimum value BZCmin corresponding to the standard deviation BZCi from the historical daily power consumption mean value; obtain the second IT power consumption limit value by subtracting the maximum value BZCmax corresponding to the standard deviation BZCi from the historical daily power consumption mean value, and obtain the IT power consumption limit interval with the first IT power consumption limit value as the left endpoint and the second IT power consumption limit value as the right endpoint; Step S35, obtaining the IT real-time power consumption corresponding to the plurality of IT devices, comparing the IT real-time power consumption with the IT power consumption limit interval, if the IT real-time power consumption is outside the IT power consumption limit interval, recording the corresponding IT real-time power consumption as an abnormal IT power consumption value; if the IT real-time power consumption is within the IT power consumption limit interval, recording the corresponding IT real-time power consumption as a normal IT power consumption value; Step S36, discarding the abnormal IT power consumption value, and then adding and averaging all the retained normal IT power consumption values ​​as the updated value of the discarded abnormal IT power consumption, to obtain the updated real-time IT power consumption after the abnormal data is processed; Step S37, and so on, perform abnormal data processing on the power consumption data of multiple refrigeration devices and multiple lighting devices to obtain updated real-time power consumption of refrigeration and updated real-time power consumption of lighting respectively.

[0011] Furthermore, the step S4 includes the following sub-steps: Step S41, obtaining IT update real-time power consumption of multiple IT devices, refrigeration update real-time power consumption of multiple refrigeration devices, and lighting update real-time power consumption of multiple lighting devices; Step S42, adding and summing the IT updated real-time power consumption of multiple IT devices to obtain the total IT device power consumption ITZ, adding and summing the refrigeration updated real-time power consumption of multiple refrigeration devices to obtain the total refrigeration device power consumption ZLZ, adding and summing the lighting updated real-time power consumption of multiple lighting devices to obtain the total lighting device power consumption ZMZ of the multiple lighting devices; Step S43, obtaining the IT theoretical total power consumption ITL corresponding to the multiple IT devices and the cooling capacity ZLL corresponding to the multiple cooling devices.

[0012] Furthermore, the step S4 further includes the following sub-steps: Step S44, calculating the power utilization efficiency corresponding to the data center; Step S45, constructing a neural network original input data table; Step S46, inputting the original input data table of the neural network into the PUE prediction model to obtain the PUE prediction value for the next day; Step S47, sending the power utilization rate corresponding to the data center and the predicted value of the power utilization rate for the next day to the user terminal.

[0013] Furthermore, the calculation process of the electric energy utilization efficiency is specifically as follows: In step S441, the present invention defines the power utilization efficiency of the data center as the ratio of the total power consumption of the data center to the total power consumption of IT equipment in the data center. The specific power utilization efficiency PUE can be expanded as: PUE=(ITZ+ZLZ+ZMZ+QTZ) / ITZ; where QTZ is the total power consumption of other equipment in the data center; Step S442: In order to simultaneously reflect the energy consumption of the refrigeration equipment and further indicate the refrigeration effect of the refrigeration equipment, the concept of energy efficiency ratio (EER) is introduced. The energy efficiency ratio indicates the refrigeration capacity per unit power of the refrigeration equipment. Specifically, the energy efficiency ratio (EER) is the ratio of the refrigeration capacity corresponding to the refrigeration equipment to the refrigeration power. Step S443: Based on step S442, the power utilization efficiency PUE of the data center is updated to: ; In the formula, EER is the energy efficiency ratio of the refrigeration equipment, D1 is the proportion of the total power consumption of IT equipment in the total power consumption of the data center, D2 is the proportion of the total power consumption of the refrigeration equipment in the total power consumption of the data center; HDZ is the total power consumption, HDZ=ITZ+ZLZ+ZMZ+QTZ.

[0014] Furthermore, the PUE prediction model is a three-layer fully connected neural network model, the number of nodes in the input layer is the number of data in the original input data table of the neural network, and the two hidden layers contain 64 nodes and 32 nodes respectively; the output layer has only one node, namely the PUE prediction value; the ReLU activation function is used between layers to fit the nonlinear function relationship.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: The present invention first collects IT power consumption data of multiple IT devices corresponding to a data center, refrigeration power consumption data of multiple refrigeration devices, and real-time lighting power consumption of multiple lighting devices; then calculates the IT theoretical power consumption corresponding to the multiple IT devices and the cooling capacity corresponding to the multiple refrigeration devices; further, performs abnormal data processing on the power consumption data of the IT equipment, refrigeration equipment, and lighting equipment in the data center; after the abnormal data processing is completed, the power utilization efficiency of the data center is calculated according to the power consumption of the multiple IT devices, refrigeration equipment, and lighting equipment corresponding to the data center, and then calculates the PUE prediction value of the data center for the next day through the PUE prediction model; finally, the power utilization rate corresponding to the data center and the power utilization rate prediction value for the next day are received through the user terminal, and the user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center; the present invention realizes efficient monitoring of the power utilization rate of the data center. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0017] Figure 1 is a flow chart of the method of the present invention; Figure 2 It is a structural schematic diagram of the data acquisition box in the present invention; Figure 3 Schematic diagram of the PUE prediction model in the present invention; Figure 4 It is a schematic diagram of the structure of the computer device in the present invention. DETAILED DESCRIPTION

[0018] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Embodiment 1: See also Figure 1-Figure 3 As shown, the technical solution provided by the present invention is: a method for online monitoring of energy efficiency of a data center based on a neural network, wherein a data center is an infrastructure for maintaining thousands or even tens of thousands of IT devices, and is composed of hardware entities such as power systems, refrigeration equipment, IT equipment, and lighting equipment, and software services such as virtual machine applications; In this embodiment, the data center energy efficiency online monitoring method is specifically as follows: Step S1, collecting IT power consumption data of multiple IT devices corresponding to the data center, real-time lighting power consumption of multiple lighting devices, refrigeration power consumption data corresponding to multiple refrigeration devices, and refrigeration impact data corresponding to the data center; Among them, IT power consumption data includes the rated power, real-time IT power consumption and working hours corresponding to multiple IT devices; refrigeration power consumption data includes the rated power and real-time refrigeration power consumption corresponding to multiple refrigeration devices; refrigeration impact data includes the floor space, indoor temperature and outdoor temperature corresponding to the data center; the data is collected through the data collection box of the present invention, and the IT power consumption data, refrigeration power consumption data and real-time lighting power consumption are all power consumption data of the data center on the day; The data acquisition box of the present invention specifically comprises: 1 data acquisition gateway, 1 multi-channel communication converter, 1 power control panel, 1 calculation unit and 1 digital display screen; Data acquisition gateway: It consists of a WiFi unit, a storage unit and a mobile power supply, and is used to establish data connection with the user terminal; Power control panel: responsible for controlling the mobile power switch and real-time display of power; Calculation unit: calculates the power utilization rate of the data center and predicts the future power utilization rate of the data center through a neural network model; Multi-channel communication converter: communicates with on-site electric meters and other equipment, and collects power consumption data corresponding to the data center through multiple communication channels; Digital display screen: receives data transmitted by the calculation unit and displays it.

[0020] Step S2, calculating IT theoretical power consumption corresponding to the plurality of IT devices and cooling capacity corresponding to the plurality of cooling devices; In this embodiment, step S2 includes the following sub-steps: Step S21, obtaining the rated power IEDi and the working time ISCi of the multiple IT devices corresponding to the data center; wherein i is the number of the multiple IT devices, i=1, 2, ..., y, y is a positive integer; Step S22, the total IT theoretical power consumption ITL of the multiple IT devices corresponding to the data center is calculated by a formula, and the specific formula is: ; Step S23, obtaining the floor area MJ, indoor temperature value and rated electric power ZEDj corresponding to the data center, where i is the serial number of the refrigeration equipment, j=1, 2, ..., z, and z is a positive integer; Step S24, the cooling capacity ZLL corresponding to the multiple cooling devices in the data center is calculated by a formula, and the specific formula is: ; The unit of ZLL is KW, GLX is the power factor, which indicates the utilization efficiency of electric energy; DFR is the heating coefficient of electric energy, that is, the efficiency of converting electric energy into heat energy; HFR is the ambient heat value, which is determined by the temperature difference between the indoor temperature value and the outdoor temperature value.

[0021] Step S3, performing abnormal data processing on the power consumption data of IT equipment, refrigeration equipment and lighting equipment in the data center; In this embodiment, step S3 includes the following sub-steps: Step S31, obtaining the historical power consumption LHDid corresponding to the plurality of IT devices in the previous month, wherein d is the number of the day in the previous month, the upper limit of d is x, x=28, 29, 30 or 31; the historical power consumption is the daily power consumption of the corresponding IT device in the previous month; Step S32, summing up the historical power consumption corresponding to all IT devices in the previous month to obtain the average value, and then dividing it by the number of days corresponding to the previous month to obtain the historical daily power consumption average value JHD of the IT devices; Step S33, calculating the standard deviation BZCi of the power consumption corresponding to the multiple IT devices by using the formula, the specific formula is: ; Step S34, obtain the first IT power consumption limit value by subtracting the minimum value BZCmin corresponding to the standard deviation BZCi from the historical daily power consumption mean value; obtain the second IT power consumption limit value by subtracting the maximum value BZCmax corresponding to the standard deviation BZCi from the historical daily power consumption mean value, and obtain the IT power consumption limit interval with the first IT power consumption limit value as the left endpoint and the second IT power consumption limit value as the right endpoint; Step S35, obtaining the IT real-time power consumption corresponding to the plurality of IT devices, comparing the IT real-time power consumption with the IT power consumption limit interval, if the IT real-time power consumption is outside the IT power consumption limit interval, recording the corresponding IT real-time power consumption as an abnormal IT power consumption value; if the IT real-time power consumption is within the IT power consumption limit interval, recording the corresponding IT real-time power consumption as a normal IT power consumption value; Step S36, discarding the abnormal IT power consumption value, and then adding and averaging all the retained normal IT power consumption values ​​as the updated value of the discarded abnormal IT power consumption, to obtain the updated real-time IT power consumption after the abnormal data is processed; For example, if a set of IT real-time power consumption includes six sets of data: A, B, C, D, E and F; wherein B and C are abnormal IT power consumption values, and A, D, E and F are normal IT power consumption values; Then discard B and C, and then calculate the average value of A, D, E, and F as the update of B and C. The real-time power consumption of IT update is: A, (A+D+E+F) / 4, (A+D+E+F) / 4, D, E, and F; Step S37, and so on, perform abnormal data processing on the power consumption data of multiple refrigeration devices and multiple lighting devices to obtain updated real-time power consumption of refrigeration and updated real-time power consumption of lighting respectively.

[0022] Step S4, calculating the power utilization efficiency of the data center according to the power consumption of multiple IT equipment, refrigeration equipment and lighting equipment corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through the PUE prediction model; In this embodiment, step S4 includes the following sub-steps: Step S41, obtaining IT update real-time power consumption of multiple IT devices, refrigeration update real-time power consumption of multiple refrigeration devices, and lighting update real-time power consumption of multiple lighting devices; Step S42, adding and summing the IT updated real-time power consumption of multiple IT devices to obtain the total IT device power consumption ITZ, adding and summing the refrigeration updated real-time power consumption of multiple refrigeration devices to obtain the total refrigeration device power consumption ZLZ, adding and summing the lighting updated real-time power consumption of multiple lighting devices to obtain the total lighting device power consumption ZMZ of the multiple lighting devices; Step S43, obtaining the IT theoretical total power consumption ITL corresponding to the multiple IT devices and the cooling capacity ZLL corresponding to the multiple cooling devices; Step S44, calculating the power utilization efficiency corresponding to the data center; Specifically, the calculation process of the electric energy utilization efficiency is as follows: In step S441, the present invention defines the power utilization efficiency of the data center as the ratio of the total power consumption of the data center to the total power consumption of IT equipment in the data center. The specific power utilization efficiency PUE can be expanded as: PUE=(ITZ+ZLZ+ZMZ+QTZ) / ITZ; where QTZ is the total power consumption of other equipment in the data center; Step S442: In order to simultaneously reflect the energy consumption of the refrigeration equipment and further indicate the refrigeration effect of the refrigeration equipment, the concept of energy efficiency ratio (EER) is introduced. The energy efficiency ratio indicates the refrigeration capacity per unit power of the refrigeration equipment. Specifically, the energy efficiency ratio (EER) is the ratio of the refrigeration capacity corresponding to the refrigeration equipment to the refrigeration power. Step S443: Based on step S442, the power utilization efficiency PUE of the data center is updated to: ; Where, EER is the energy efficiency ratio of the cooling equipment, D1 is the proportion of the total power consumption of IT equipment in the total power consumption of the data center, D2 is the proportion of the total power consumption of the cooling equipment in the total power consumption of the data center; HDZ is the total power consumption, HDZ=ITZ+ZLZ+ZMZ+QTZ; Step S45, constructing the neural network original input data table: Data Center IT Equipment Refrigeration equipment Lighting Data Center Power IT equipment power Refrigeration equipment power Lighting equipment power Data center power rating IT equipment rated power Rated power of refrigeration equipment Rated power of lighting equipment IT equipment usage time Refrigeration equipment usage time Lighting equipment usage time The IT equipment power in the table is obtained by dividing the total power consumption of the IT equipment by the IT equipment usage time. The same is true for the cooling equipment power and lighting equipment power. The IT equipment usage time, cooling equipment usage time and lighting equipment usage time are estimated values ​​for the next day and are filled in by the user. Step S46, inputting the original input data table of the neural network into the PUE prediction model to obtain the PUE prediction value for the next day; The PUE prediction model is a three-layer fully connected neural network model, such as Figure 3 As shown, the number of nodes in the input layer is the number of data in the original input data table of the neural network, which is 11 in this embodiment; the two hidden layers contain 64 nodes and 32 nodes respectively; the output layer has only one node, namely the PUE prediction value; the ReLU activation function is used between the layers to fit the nonlinear function relationship; Step S47, sending the power utilization rate corresponding to the data center and the predicted value of the power utilization rate for the next day to the user terminal; Step S5, the user terminal receives the power utilization rate corresponding to the data center and the power utilization rate forecast value for the next day. The user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center.

[0023] In this application, if corresponding calculation formulas appear, the above calculation formulas are all dimensionless and take their numerical calculations. The weight coefficients, proportional coefficients and other coefficients in the formulas are set to a result value obtained by quantifying each parameter. The size of the weight coefficient and the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the result value.

[0024] Embodiment 2: Figure 4The structural diagram of a computer device is illustrated, and the computer device may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The processor may call the logic instructions in the memory to execute a method for online monitoring of data center energy efficiency based on a neural network, and the method includes: collecting IT power consumption data corresponding to multiple IT devices in the data center, real-time lighting power consumption of multiple lighting devices, cooling power consumption data corresponding to multiple refrigeration devices, and cooling impact data corresponding to the data center; calculating IT theoretical power consumption corresponding to multiple IT devices and cooling capacity corresponding to multiple refrigeration devices; performing abnormal data processing on power consumption data of IT devices, refrigeration devices, and lighting devices in the data center; calculating the power utilization efficiency of the data center based on the power consumption of multiple IT devices, refrigeration devices, and lighting devices corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through the PUE prediction model; the user terminal receives the power utilization rate corresponding to the data center and the power utilization rate prediction value for the next day, and the user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center.

[0025] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0026] On the other hand, the present application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute a neural network-based data center energy efficiency online monitoring method provided by the above methods, the method including: collecting IT power consumption data of multiple IT devices corresponding to the data center, real-time lighting power consumption of multiple lighting devices, refrigeration power consumption data corresponding to multiple refrigeration devices, and refrigeration impact data corresponding to the data center; calculating the IT theoretical power consumption corresponding to multiple IT devices and the cooling capacity corresponding to multiple refrigeration devices; performing abnormal data processing on the power consumption data of IT equipment, refrigeration equipment and lighting equipment in the data center; calculating the power utilization efficiency of the data center based on the power consumption of multiple IT devices, refrigeration equipment and lighting equipment corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through the PUE prediction model; the user terminal receives the power utilization rate corresponding to the data center and the power utilization rate prediction value for the next day, and the user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center.

[0027] On the other hand, the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the above-mentioned method for online monitoring of data center energy efficiency based on a neural network, the method comprising: collecting IT power consumption data of multiple IT devices corresponding to the data center, real-time lighting power consumption of multiple lighting devices, refrigeration power consumption data corresponding to multiple refrigeration devices, and refrigeration impact data corresponding to the data center; calculating IT theoretical power consumption corresponding to multiple IT devices and cooling capacity corresponding to multiple refrigeration devices; performing abnormal data processing on power consumption data of IT equipment, refrigeration equipment and lighting equipment in the data center; calculating the power utilization efficiency of the data center based on the power consumption of multiple IT devices, refrigeration equipment and lighting equipment corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through a PUE prediction model; a user terminal receives the power utilization rate corresponding to the data center and the power utilization rate prediction value for the next day, and the user learns about the power utilization status of the data terminal by receiving the power utilization rate of the data center.

[0028] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0029] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0030] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for online monitoring of data center energy efficiency based on neural network, characterized in that: include: Step S1, collecting IT power consumption data of multiple IT devices corresponding to the data center, real-time lighting power consumption of multiple lighting devices, refrigeration power consumption data corresponding to multiple refrigeration devices, and refrigeration impact data corresponding to the data center; Step S2, calculating IT theoretical power consumption corresponding to the plurality of IT devices and cooling capacity corresponding to the plurality of cooling devices; Step S3, performing abnormal data processing on the power consumption data of IT equipment, refrigeration equipment and lighting equipment in the data center; Step S4, calculating the power utilization efficiency of the data center according to the power consumption of multiple IT equipment, refrigeration equipment and lighting equipment corresponding to the data center, and then calculating the PUE prediction value of the data center for the next day through the PUE prediction model; Step S5, the user terminal receives the power utilization rate corresponding to the data center and the power utilization rate forecast value for the next day. The user learns the power utilization status of the data terminal by receiving the power utilization rate of the data center.

2. According to claim 1, a method for online monitoring of data center energy efficiency based on neural network is characterized in that: The IT power consumption data includes the rated power, real-time IT power consumption and working hours corresponding to multiple IT devices; the refrigeration power consumption data includes the rated power and real-time refrigeration power consumption corresponding to multiple refrigeration equipment; the refrigeration impact data includes the floor area, indoor temperature and outdoor temperature corresponding to the data center.

3. The method for online monitoring of data center energy efficiency based on neural network according to claim 2, characterized in that: The step S2 includes the following sub-steps: Step S21, obtaining the rated power IEDi and the working time ISCi of the multiple IT devices corresponding to the data center; wherein i is the number of the multiple IT devices, i=1, 2, ..., y, y is a positive integer; Step S22, the total IT theoretical power consumption ITL of the multiple IT devices corresponding to the data center is calculated by a formula, and the specific formula is: 。 4. The method for online monitoring of data center energy efficiency based on neural network according to claim 3 is characterized in that: The step S2 further comprises the following sub-steps: Step S23, obtaining the floor area MJ, indoor temperature value, outdoor temperature value and rated power ZEDj corresponding to the data center, where j is the serial number of the multiple refrigeration devices, j=1, 2, ..., z, and z is a positive integer; Step S24, the cooling capacity ZLL corresponding to the multiple cooling devices in the data center is calculated by a formula, and the specific formula is: ； The unit of ZLL is KW, GLX is the power factor, which indicates the utilization efficiency of electric energy; DFR is the heating coefficient of electric energy, that is, the efficiency of converting electric energy into heat energy; HFR is the ambient heat value, which is determined by the temperature difference between the indoor temperature value and the outdoor temperature value.

5. The method for online monitoring of data center energy efficiency based on neural network according to claim 1, characterized in that: The step S3 includes the following sub-steps: Step S31, obtaining the historical power consumption LHDid corresponding to the plurality of IT devices in the previous month, wherein d is the number of the day in the previous month; the historical power consumption is the daily power consumption of the corresponding IT devices in the previous month; Step S32, summing up the historical power consumption corresponding to all IT devices in the previous month to obtain the average value, and then dividing it by the number of days corresponding to the previous month to obtain the historical daily power consumption average value JHD of the IT devices; Step S33, calculating the standard deviation BZCi of the power consumption corresponding to the multiple IT devices by using the formula, the specific formula is: 。 6. The method for online monitoring of data center energy efficiency based on neural network according to claim 5, characterized in that: The step S3 further comprises the following sub-steps: Step S34, obtain the first IT power consumption limit value by subtracting the minimum value BZCmin corresponding to the standard deviation BZCi from the historical daily power consumption mean value; obtain the second IT power consumption limit value by subtracting the maximum value BZCmax corresponding to the standard deviation BZCi from the historical daily power consumption mean value, and obtain the IT power consumption limit interval with the first IT power consumption limit value as the left endpoint and the second IT power consumption limit value as the right endpoint; Step S35, obtaining the IT real-time power consumption corresponding to the plurality of IT devices, comparing the IT real-time power consumption with the IT power consumption limit interval, if the IT real-time power consumption is outside the IT power consumption limit interval, recording the corresponding IT real-time power consumption as an abnormal IT power consumption value; if the IT real-time power consumption is within the IT power consumption limit interval, recording the corresponding IT real-time power consumption as a normal IT power consumption value; Step S36, discarding the abnormal IT power consumption value, and then adding and averaging all the retained normal IT power consumption values ​​as the updated value of the discarded abnormal IT power consumption, to obtain the updated real-time IT power consumption after the abnormal data is processed; Step S37, and so on, perform abnormal data processing on the power consumption data of multiple refrigeration devices and multiple lighting devices to obtain updated real-time power consumption of refrigeration and updated real-time power consumption of lighting respectively.

7. The method for online monitoring of data center energy efficiency based on neural network according to claim 6, characterized in that: The step S4 includes the following sub-steps: Step S41, obtaining IT update real-time power consumption of multiple IT devices, refrigeration update real-time power consumption of multiple refrigeration devices, and lighting update real-time power consumption of multiple lighting devices; Step S42, adding and summing the IT updated real-time power consumption of multiple IT devices to obtain the total IT device power consumption ITZ, adding and summing the refrigeration updated real-time power consumption of multiple refrigeration devices to obtain the total refrigeration device power consumption ZLZ, adding and summing the lighting updated real-time power consumption of multiple lighting devices to obtain the total lighting device power consumption ZMZ of the multiple lighting devices; Step S43, obtaining the IT theoretical total power consumption ITL corresponding to the multiple IT devices and the cooling capacity ZLL corresponding to the multiple cooling devices.

8. The method for online monitoring of data center energy efficiency based on neural network according to claim 7, characterized in that: The step S4 also includes the following sub-steps: Step S44, calculating the power utilization efficiency corresponding to the data center; Step S45, constructing a neural network original input data table; Step S46, inputting the original input data table of the neural network into the PUE prediction model to obtain the PUE prediction value for the next day; Step S47, sending the power utilization rate corresponding to the data center and the predicted value of the power utilization rate for the next day to the user terminal.

9. The method for online monitoring of data center energy efficiency based on neural network according to claim 7, characterized in that: The calculation process of the electric energy utilization efficiency is specifically as follows: In step S441, the present invention defines the power utilization efficiency of the data center as the ratio of the total power consumption of the data center to the total power consumption of IT equipment in the data center. The specific power utilization efficiency PUE can be expanded as: PUE=(ITZ+ZLZ+ZMZ+QTZ) / ITZ; where QTZ is the total power consumption of other equipment in the data center; Step S442: In order to simultaneously reflect the energy consumption of the refrigeration equipment and further indicate the refrigeration effect of the refrigeration equipment, the concept of energy efficiency ratio (EER) is introduced. The energy efficiency ratio indicates the refrigeration capacity per unit power of the refrigeration equipment. Specifically, the energy efficiency ratio (EER) is the ratio of the refrigeration capacity corresponding to the refrigeration equipment to the refrigeration power. Step S443: Based on step S442, the power utilization efficiency PUE of the data center is updated to: ； In the formula, EER is the energy efficiency ratio of the refrigeration equipment, D1 is the proportion of the total power consumption of IT equipment in the total power consumption of the data center, D2 is the proportion of the total power consumption of the refrigeration equipment in the total power consumption of the data center; HDZ is the total power consumption, HDZ=ITZ+ZLZ+ZMZ+QTZ.

10. The method for online monitoring of data center energy efficiency based on neural network according to claim 8, characterized in that: The PUE prediction model is a three-layer fully connected neural network model. The number of nodes in the input layer is the number of data in the original input data table of the neural network. The two hidden layers contain 64 nodes and 32 nodes respectively. The output layer has only one node, namely the PUE prediction value. The ReLU activation function is used between layers to fit the nonlinear function relationship.

Images