CN115841320A - Inspection method for energy utilization efficiency of data center - Google Patents

Inspection method for energy utilization efficiency of data center Download PDF

Info

Publication number
CN115841320A
CN115841320A CN202211190833.XA CN202211190833A CN115841320A CN 115841320 A CN115841320 A CN 115841320A CN 202211190833 A CN202211190833 A CN 202211190833A CN 115841320 A CN115841320 A CN 115841320A
Authority
CN
China
Prior art keywords
unit
energy consumption
probability
time period
energy
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211190833.XA
Other languages
Chinese (zh)
Inventor
朱文进
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Telecom Digital Intelligence Technology Co Ltd
Original Assignee
China Telecom Digital Intelligence Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Telecom Digital Intelligence Technology Co Ltd filed Critical China Telecom Digital Intelligence Technology Co Ltd
Priority to CN202211190833.XA priority Critical patent/CN115841320A/en
Publication of CN115841320A publication Critical patent/CN115841320A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Air Conditioning Control Device (AREA)

Abstract

The application discloses a method of patrolling and examining to data center energy utilization efficiency, including S1: calculating the energy consumption ratio of each unit in the data center in the current time period in the whole data center; s2: analyzing, for each unit in the data center and under a Markov chain-based model, a change in energy consumption of each unit in a next time period; s3: and (3) predicting and calculating the energy utilization efficiency of the data center in the next time period by combining the data contents of the step (S1) and the step (S2). According to the method and the device, the macroscopic overall energy utilization efficiency of the data center can be observed from the PUE value calculated through prediction, and then the aging risk of which unit is analyzed in the energy consumption change calculation of each microscopic unit is higher, so that the targeted investigation is realized, and the device is prevented from aging, and unnecessary energy consumption and power consumption are caused.

Description

Inspection method for energy utilization efficiency of data center
Technical Field
The invention relates to the technical field of energy conservation detection, in particular to a polling method for the energy utilization efficiency of a data center.
Background
Under the current social environment, the application of computer technology has been developed rapidly, the power consumption of data centers is larger and larger, servers can be driven to operate by relying on electric energy completely, and meanwhile, in order to ensure stable development of digital economy in China, and the requirement of the data centers on the quality of the electric energy is high, double-circuit power supply is considered frequently to ensure that the data centers operate uninterruptedly all the year round. It is based on this situation that the power consumption of data centers in China is continuously increased. However, in a data center of a computer room, besides IT equipment units for core processing computation, there are other auxiliary supporting equipment, such as air conditioning units for cooling, units for lighting, etc., which also increase the energy consumption of a computer room, but as the lines of these equipment age, the energy consumption also increases gradually, further increasing the overall energy consumption of the data center, decreasing the energy utilization efficiency of the data center, and increasing the PUE value of the data center.
According to the method and the device, on the basis of historical data, the PUE value of the data center is analyzed and predicted, the overall energy utilization rate of the data center is conveniently controlled and judged by machine room operation and maintenance personnel, meanwhile, the energy consumption change condition of each unit of the data center at the subsequent time is analyzed and predicted, the machine room operation and maintenance personnel are assisted to judge and detect each device, and unnecessary energy consumption and power consumption are avoided due to the fact that the circuit of the device is aged. In the existing technical scheme, the scheme content such as analysis and prediction of the application is not adopted for the energy consumption of the data center.
Disclosure of Invention
The invention provides a routing inspection method aiming at the energy utilization efficiency of a data center aiming at the defects in the prior art so as to solve the problems in the background technology.
In order to achieve the purpose, the invention adopts the following technical scheme:
a patrol inspection method for energy utilization efficiency of a data center comprises the following steps:
s1: calculating the energy consumption ratio of each unit in the data center in the current time period in the whole data center;
s2: analyzing, for each unit in the data center and under a Markov chain-based model, a change in energy consumption of each unit in a next time period;
s3: and (3) predicting and calculating the energy utilization efficiency of the data center in the next time period by combining the data contents of the step (S1) and the step (S2) so as to detect and examine the devices of the data center in the machine room in a targeted manner and prevent the devices from aging to cause unnecessary energy consumption and power consumption.
In order to optimize the technical scheme, the specific measures adopted further comprise:
further, the specific content of step S1 is:
each unit in the data center comprises an air conditioning unit, an air heat transfer unit, a power distribution system unit and an IT equipment unit; the IT equipment unit is a data center operation core, and represents that the energy utilization rate of the data center is higher when the integral energy consumption ratio of the data center is higher;
for the air conditioning unit, the electric energy a consumed by the air conditioning unit in the last time period is obtained based on historical data statistics 1 Electric energy b consumed by the air-conditioning unit in the current time period 1 And the electric energy consumed by the air conditioning unit in the whole data center in the last time period is compared with X 1 And further calculating the electric quantity increasing and decreasing amplitude c of the air-conditioning unit in the current time period 1 ，c 1 ＝(b 1 -a 1 )/a 1 According to the increasing and decreasing amplitude c 1 Determining the electric energy ratio X consumed by the air conditioning unit in the data center in the current time period 1 ′，X 1 ′＝X 1 -(X 1 *c 1 ) (ii) a Wherein the last time period is also called the last time period, and the current time period is also called the current time period;
for the wind heat transfer unit, the electric energy a consumed by the wind heat transfer unit in the last time period is obtained based on historical data statistics 2 And the electric energy b consumed by the wind heat transfer unit in the current time period 2 And the electric energy ratio X consumed by the heat transfer units in the whole data center in the last time period 2 And further calculating the electric quantity increasing and decreasing amplitude c of the heat transfer unit in the current time period 2 ，c 2 ＝(b 2 -a 2 )/a 2 According to increasing or decreasing amplitude c 2 Determining the electric energy ratio X of the wind heat transfer unit in the data center in the current time period 2 ′，X 2 ′＝X 2 -(X 2 *c 2 )；
For the power distribution system unit, the electric energy a consumed by the power distribution system unit in the last time period is obtained based on historical data statistics 3 B electric energy consumed by the power distribution system unit in the current time period 3 Andthe power consumption ratio X of the power distribution system unit in the whole data center in the last time period 3 And further calculating the electric quantity increasing and decreasing amplitude c of the power distribution system unit in the current time period 3 ，c 3 ＝(b 3 -a 3 )/a 3 According to increasing or decreasing amplitude c 3 Determining the electric energy ratio X consumed by the power distribution system unit in the whole data center in the current time period 3 ′，X 3 ′＝X 3 -(X 3 *c 3 )；
For the IT equipment unit, the electric energy a consumed by the IT equipment unit in the last time period is obtained based on historical data statistics 4 B electric energy consumed by IT equipment unit in current time period 4 And the electric energy consumed by the IT equipment units in the whole data center in the last time period is compared with X 4 And further calculating the electric quantity increasing and decreasing amplitude c of the IT equipment unit in the current time period 4 ，c 4 ＝(b 4 -a 4 )/a 4 According to increasing or decreasing amplitude c 4 Determining the electric energy ratio X consumed by the IT equipment units in the data center in the current time period 4 ′，X 4 ′＝X 4 -(X 4 *c 4 )。
Further, the wind heat-transferring unit comprises a water cooler, a humidifier, lighting and auxiliary equipment, a switch or a generator; the power distribution system unit comprises a UPS and a PDU; the IT equipment unit comprises an IT equipment CPU and a process load.
Further, the specific content of step S2 is:
for an air conditioning unit, obtaining energy consumption transfer probability [ d ] of the air conditioning unit in the last period through historical data 1 、e 1 Energy consumption transfer probability [ d ] of air conditioning unit in this period 2 、e 2 Energy consumption transition probability [ d ] of units other than air conditioning unit 3 、e 3 Wherein d is 1 Representing the probability of energy consumption transfer of the air conditioning unit in the last period, e 1 Representing the energy consumption roll-out probability of the air conditioning unit; d 2 Representing the probability of energy transfer into the air conditioning unit during the time period, e 2 Representing the energy consumption roll-out probability of the air conditioning unit; d 3 Representing the probability of energy transfer from the unit other than the air conditioning unit to the air conditioning unit during the present time period, e 2 The energy consumption transfer probability that the energy consumption of other units except the air conditioning unit is not transferred to the air conditioning unit is represented;
and further calculating the energy consumption transfer probability [ d ] of the air conditioning unit in the next time period based on the Markov chain model 4 、e 4 H ]; wherein d is 4 Represents a probability of a transition of the energy consumption of the air conditioning unit into the next time period, i.e., a probability of an increase in the energy consumption of the air conditioning unit in the next time period, and d 4 ＝d 1 *d 2 +e 1 *d 3 ，e 4 Representing the probability of the air conditioning unit's energy consumption rolling out in the next time period, i.e. the probability of the air conditioning unit's energy consumption decreasing in the next time period, and e 4 ＝d 1 *e 2 +e 1 *e 3 ；
For the wind heat transfer unit, the energy consumption transfer probability of the wind heat transfer unit in the last period is obtained through historical data [ f 1 、g 1 Energy consumption transfer probability [ f ] of wind heat transfer unit in this period 2 、g 2 Energy consumption transition probability [ f ] of units other than wind heat transfer unit 3 、g 3 Therein f 1 Representing the energy consumption transfer probability, g, of the heat transfer unit over the last period of time 1 Representing the energy consumption transfer probability of the wind heat transfer unit; f. of 2 Representing the energy transfer probability, g, of the heat transfer unit during this time period 2 Representing the energy consumption transfer probability of the wind heat transfer unit; f. of 3 Represents the energy consumption transfer probability of other units except the wind heat transfer unit transferring energy consumption to the wind heat transfer unit in the period, g 3 The energy consumption transfer probability that the energy consumption of other units except the wind heat transfer unit is not transferred to the wind heat transfer unit is represented;
and further calculating the energy consumption transfer probability (f) of the wind heat transfer unit in the next time period based on the Markov chain model 4 、g 4 H ]; wherein f is 4 Representing the probability of the energy consumption of the heat transfer unit transitioning into the next time period, i.e. the probability of the energy consumption of the heat transfer unit increasing in the next time period, and f 4 ＝f 1 *f 2 +g 1 *f 3 ，g 4 Represents the energy consumption transfer-out probability of the heat transfer unit in the next time period, i.e., the probability of the heat transfer unit energy consumption being reduced in the next time period, and g 4 ＝f 1 *g 2 +g 1 *g 3 ；
Aiming at the power distribution system unit, the energy consumption transfer probability [ h ] of the power distribution system unit in the previous period is obtained through historical data 1 、i 1 Energy consumption transfer probability [ h ] of power distribution system unit in this period 2 、i 2 Energy consumption transition probability [ h ] of other units except power distribution system unit 3 、i 3 Wherein h is 1 Representing the probability, i, of the transition of energy consumption of the power distribution system unit in the up period 1 Representing the energy consumption roll-out probability of the power distribution system unit; h is 2 Representing the probability, i, of the transition of energy consumption of the power distribution system unit in this time period 2 Representing the energy consumption roll-out probability of the power distribution system unit; h is a total of 3 Indicating the probability of energy transfer, i, into the distribution system unit of energy consumption of units other than the distribution system unit during the period 3 The energy consumption transition probability that the energy consumption of other units except the power distribution system unit is not transferred to the power distribution system unit is represented;
and further calculating the energy consumption transfer probability [ h ] of the power distribution system unit in the next time period based on the Markov chain model 4 、i 4 H ]; wherein h is 4 Represents the probability of the energy consumption of the distribution system unit shifting into the next time period, i.e. the probability of the energy consumption of the distribution system unit increasing in the next time period, and h 4 ＝h 1 *h 2 +i 1 *h 3 ，i 4 Represents the energy consumption roll-out probability of the power distribution system unit in the next time period, i.e. the probability of the power distribution system unit energy consumption reduction in the next time period, and i 4 ＝h 1 *i 2 +i 1 *i 3 ；
Acquiring the energy consumption transfer probability [ j ] of the IT equipment unit in the last period through historical data aiming at the IT equipment unit 1 、k 1 Energy consumption transfer probability [ j ] of IT equipment unit in this period 2 、k 2 Energy consumption of units other than IT equipment unitsProbability of shift [ j 3 、k 3 Therein j 1 Representing the probability of energy consumption of the IT equipment unit, k, in the last period 1 Representing the energy consumption roll-out probability of the IT equipment unit; j is a function of 2 Representing the probability of energy consumption transfer, k, of the IT equipment unit in the present period 2 Representing the energy consumption roll-out probability of the IT equipment unit; j is a function of 3 Represents the energy consumption transition probability, k, of other units except the IT equipment unit to the IT equipment unit in the period 3 The energy consumption transfer probability that the energy consumption of other units except the IT equipment unit is not transferred to the IT equipment unit is represented;
and then calculating the energy consumption transfer probability [ j ] of the IT equipment unit in the next time period based on the Markov chain model 4 、k 4 H ]; wherein j is 4 Represents a transition probability of the energy consumption of the IT equipment unit in the next time slot, that is, a probability of an increase in the energy consumption of the IT equipment unit in the next time slot, and j 4 ＝j 1 *j 2 +k 1 *j 3 ，k 4 Represents the energy consumption transfer-out probability of the IT equipment unit in the next time period, i.e. the probability of the energy consumption reduction of the IT equipment unit in the next time period, and k 4 ＝j 1 *k 2 +k 1 *k 3 。
Further, the specific content of step S3 is:
aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the reduction of the energy utilization rate is as follows:
PUE 1 ＝X 1 ′*d 4 +X 2 ′*f 4 +X 3 ′*h 4 +X 4 ′*j 4
aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the increase of the energy utilization rate is as follows:
PUE 2 ＝X 1 ′*e 4 +X 2 ′*g 4 +X 3 ′*i 4 +X 4 ′*k 4
therefore, the operation and maintenance personnel detect and investigate the devices of the data center in the machine room according to the predicted increase or decrease of the energy utilization efficiency of the data center in the next time period and the energy consumption transfer probability of each unit in the step S2 in the next time period, so as to prevent the devices from aging and causing unnecessary energy consumption and electricity consumption.
The invention has the beneficial effects that: the method comprises the steps of analyzing and calculating the energy consumption change of each unit of the data center in the next time period, and predicting the PUE value change of the data center of the computer room according to the prediction result; according to the data, operation and maintenance personnel of the machine room can judge the aging degree of each unit device of the machine room in an auxiliary mode according to the prediction data so as to realize targeted inspection and detection and prevent unnecessary energy consumption and power consumption caused by aging of each unit device and circuit. Specifically, the energy utilization efficiency of the data center macroscopically overall can be observed from the calculated PUE value, and then which unit has higher aging risk is specifically analyzed from the microscopic energy consumption change calculation of each unit, so that the targeted investigation is realized.
Drawings
Fig. 1 is a schematic diagram of energy consumption in inspection of an upper period machine room in the embodiment of the invention.
Fig. 2 is a schematic diagram of the inspection energy consumption of the time-slot machine room in the embodiment of the invention.
Fig. 3 is a schematic diagram illustrating calculation of probability of occurrence of energy transfer of the air conditioner in a time period according to the embodiment of the invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings.
This application mainly includes: the system comprises a data center intelligent management module and an AI energy consumption analysis module.
1. The intelligent management module of the data center comprises an air conditioning unit, an air heat transfer unit, an IT equipment unit and a power distribution system unit.
1. Air conditioning unit
Firstly, the electric quantity, unit (kwh) consumed by the air conditioner in the current time period is obtained through polling, and secondly, historical energy consumption data are analyzed and calculated to obtain energy consumption data in the previous time period. The cooling in the machine room is mainly performed by the machine room air conditioner, so that the power consumption of the machine room air conditioner is reduced, and the PUE value of the machine room can be effectively reduced;
taking an air conditioner as an example: the total energy consumption is calculated by combining historical data statistics with the self power of the air conditioner, and the following results are obtained: the air conditioner in the machine room accounts for 15% of the total energy consumption of the machine room. The power consumption 1.28kwh consumed by the air conditioner in the hour is subtracted from the power consumption 1.6kwh consumed by the air conditioner in the previous period, and the ratio of the rising amplitude to the falling amplitude is 20 percent, namely 0.2, and the total power consumption is reduced by 0.32 (kwh) =0.2x 1.6 (kwh); wherein:
the calculation process of the 20% reduction of the electric quantity consumed in the period is as follows:
(present period-upper period)/upper period = positive number indicates an increase in energy consumption;
(this period-upper period)/upper period = negative number represents energy consumption reduction;
the time period of the air conditioner on the machine room accounts for 15% of the total power consumption of the machine room, and the energy consumption of the time period data center accounts for: 12%, the calculation process is as follows:
the upper period (energy consumption) is 0.15-0.15x 0.2 (energy consumption reduction in the period) =0.12;
further, referring to the embodiment in fig. 1 and 2, the total energy consumption ratio of all the precision air conditioners to the machine room in the last period is 15%, and the total energy consumption ratio of all the precision air conditioners to the machine room in the present period is 12%.
2. Wind heat transfer unit
Firstly, the wind heat transfer energy consumption is mainly based on data indexes monitored by a temperature and humidity sensor, and the hardware of the wind heat transfer energy consumption mainly comprises a water cooler, a humidifier, lighting and auxiliary equipment and a switching device/a generator; acquiring a temperature unit (DEG C) and a humidity unit (RH) of a sensor of each hardware in the current time period through polling;
secondly, acquiring historical sensor temperature and humidity data so as to obtain sensor temperature and humidity data in the last period of time; the wind heat transfer energy consumption index in the machine room mainly refers to the fact that collected data obtained by a temperature and humidity sensor of the machine room are monitored, compared with the previous time period, the amplitude is increased and decreased for analysis, and therefore the PUE value of the machine room can be effectively reduced by reducing the temperature and humidity of wind heat transfer hardware (including a water cooler, a humidifier, lighting and auxiliary equipment, a switching device/a generator);
taking a water chiller as an example: the time period energy consumption calculation is carried out according to historical data statistics on the types of wind heat transfer hardware: the water cooler accounts for 24% of the total power consumption of the machine room in the upper period of the machine room. The relative humidity of the water chiller in the hour is RH (48%) and the relative humidity RH (60%) of the water chiller in the last period are subtracted and divided by RH (60%) in the last period to obtain the rising and falling amplitude proportion of 20%, namely, 0.2, the total energy consumption of the water chiller is reduced by RH (20%) = RH (48%) -RH (60%)/RH (60%); wherein:
the calculation process of the amplitude reduction RH (20%) of the energy consumption of the water chiller in the period is as follows:
(present period-upper period)/upper period = positive number indicates an increase in energy consumption;
(this period-upper period)/upper period = negative number represents energy consumption reduction;
the energy consumption of the water chiller in the machine room at the time interval is 19.2 percent by combining the fact that the water chiller accounts for 24 percent of the total power consumption of the machine room at the time interval on the machine room, and the specific calculation formula is as follows:
the upper period (energy consumption) is 0.24-0.24X 0.2 (energy consumption reduction in the period) =0.192;
similarly, energy consumption ratios of the humidifier, the lighting and auxiliary equipment, the switching device/the generator, temperature and humidity and the like are calculated and accumulated;
further, please refer to the embodiment in fig. 1 and 2, the ratio of the total energy consumption of all wind heat removal to the machine room in the last period is 24% of the water chiller, 3% of the humidifier, 1% of the lighting and auxiliary equipment, 1% of the switching device/generator, and 1% =29%; the total energy consumption ratio of all wind heat transfer to the machine room in the period is 25.2%.
3. IT equipment unit
Firstly, data of a unit (DEG C) of the temperature of a CPU of the IT equipment and a unit (%) of the percentage of load energy consumption of a process-level server are obtained through polling in the period. Secondly, analyzing historical IT equipment energy consumption data to obtain energy consumption index data of the last period. The main components of the IT equipment energy consumption index in the machine room comprise CPU temperature and process level server load, so that the PUE value of the IT equipment in the machine room can be effectively reduced by reducing the 2 components;
taking the CPU temperature as an example: and counting the time interval energy consumption of the IT equipment through historical data to obtain: the average energy consumption of the CPU temperature of the computer room accounts for 30% of the total energy consumption of the computer room. The average degree 27 (DEG C) of the CPU temperature of the IT equipment at the hour is subtracted from the average degree 30 (DEG C) of the IT equipment at the previous period, and the average degree is divided by the average degree of the CPU temperature of the IT equipment at the previous period to obtain the ascending-descending amplitude ratio of 0.1. A decrease in total CPU temperature of 3 (° c) =0.1x 30 (° c); wherein:
the calculation process of the 10% reduction amplitude of the CPU temperature average degree unit (DEG C) in the time period is as follows:
(present period-upper period)/upper period = positive number represents CPU temperature rise (power consumption rise);
(present period-upper period)/upper period = negative number represents a decrease in CPU temperature (decrease in power consumption);
the time interval of the CPU temperature of the computer room accounts for 30% of the total power consumption of the computer room, the ratio of the CPU temperature power consumption of the computer room in the time interval is 0.27, and the specific calculation formula is as follows:
the upper period (energy consumption) is 0.3-0.3x 0.1 (the amplitude of the upper period is reduced by 10%) =0.27; similarly, calculating the load energy consumption ratio of the process level server and accumulating;
further, referring to the embodiment in fig. 1 and 2, the total energy consumption ratio of all IT equipment units in the last period is 47%; the total power consumption of all IT equipment units in this period is 42.3%.
4. Power distribution system unit
Firstly, the energy consumption data kilovolt-ampere, unit (KVA) of the power distribution system in the period is obtained through polling. And secondly, analyzing the energy consumption data of the historical power distribution system and calculating to obtain the energy consumption index of the previous period. The main components of the power distribution system in the machine room comprise a UPS and a PDU (Power distribution Unit), so that the power consumption of the UPS and the PDU of the machine room can be reduced, and the PUE value of the machine room can be effectively reduced;
taking the UPS as an example, the total power consumption is calculated by combining historical data statistics and the power of the UPS itself to obtain: the UPS of the machine room accounts for 6 percent of the total power consumption of the machine room. The power consumption 5KVA consumed by the UPS in the hour is subtracted from the power consumption 10 (KVA) consumed by the UPS in the previous period, and the power consumption 10 (KVA) consumed by the UPS in the previous period is divided to obtain the ascending-descending amplitude ratio of 0.5, and the power consumption of the UPS is reduced by 5 (KVA) =0.5X 10 (KVA); wherein:
the calculation process of the UPS energy consumption 5 (KVA) reduction by 50% in the period is as follows:
(this period-up period)/up period = positive number represents UPS power consumption increase;
(this period-up period)/up period = negative number represents UPS power consumption reduction;
the UPS energy consumption of the time interval data center is obtained by combining the time interval of the UPS of the machine room accounting for 6% of the total power consumption of the machine room: 0.03. the specific calculation formula is as follows:
the upper period (energy consumption) is 0.06-0.06X 0.5 (energy consumption reduction in the period) =0.03; similarly, calculating the PDU energy consumption ratio and accumulating;
further, referring to the embodiment in fig. 1 and 2, the total energy consumption of all the power distribution system units in the last period is 9%; the total energy consumption of all IT equipment units in the period is 5 percent.
2. And an AI energy consumption analysis module.
S1: intelligent patrol reports are generated by constructing sub-module models with different energy consumption to assist operation and maintenance personnel in judgment and detection;
1. an air conditioner unit model is built based on a Markov chain model, and energy consumption indexes of the current transfer and the probability of occurrence of next energy consumption abnormity are set by combining historical big data analysis, wherein the specific examples are as follows:
example 1: (three sets of data required for Markov chain model)
1) The probability of energy consumption transfer of the air conditioner in the previous period is (0.3, 0.7); wherein 0.3 represents the energy consumption transfer-in probability of the air conditioning unit in the previous period, and 0.7 represents the energy consumption transfer-out probability of the air conditioning unit in the previous period;
2) The probability of air conditioner energy consumption transfer in the current time period is 0.6 and 0.4; wherein 0.6 represents the energy consumption transfer-in probability of the air conditioning unit in the present time period, and 0.4 represents the energy consumption transfer-out probability of the air conditioning unit in the present time period;
3) The probabilities of transferring other energy consumption to the air conditioner in the period are 0.3 and 0.7; wherein 0.3 represents the energy consumption transfer probability that the energy consumption of other units except the air conditioning unit is transferred to the air conditioning unit in the period, and 0.7 represents the energy consumption transfer probability that the energy consumption of other units except the air conditioning unit is not transferred to the air conditioning unit;
referring to the attached drawing 3, the probability of occurrence of air conditioner energy consumption transfer in the following time period [ 0.39.0.61 ] is obtained through calculation, and the specific calculation process is as follows:
air conditioner energy consumption is shifted to the occurrence probability of 0.3x0.6+0.3x0.7=0.39 in the lower period
Lower time period air conditioner energy consumption conversion occurrence probability of 0.3x0.4+0.7x0.7=0.61
2. And (4) constructing a [ wind heat transfer unit model ] and analyzing by combining historical data to obtain the energy consumption transfer index and the next energy consumption abnormity occurrence probability. Same example 1
3. And (4) constructing an IT equipment unit model, and analyzing by combining historical data to obtain the energy consumption index of the transfer and the probability of the occurrence of the next energy consumption abnormity. Same example 1
4. And (4) constructing a [ power distribution system unit model ] and analyzing by combining historical data to obtain the energy consumption index of the current transfer and the next energy consumption abnormity occurrence probability. Same example 1
And generating partial contents of the intelligent patrol inspection report form through the contents.
And S2, updating the model operation result and the energy consumption transition probability of the next time period to the inspection report. And predicting the next energy consumption change probability through historical data and routing inspection data so as to obtain a PUE predicted value. Generating a complete intelligent inspection report;
referring to fig. 1 and 2:
the energy consumption of the infrastructure of the data center accounts for 53% in the last period, and the power consumption of the IT equipment accounts for 47%. The energy usage efficiency (PUE value) of a data center is: (53% + 47%)/47% =2.13
The energy consumption of infrastructure of the data center accounts for 42.2% in the period, and the power consumption of IT equipment is 42.3%. The energy usage efficiency (PUE value) of a data center is: (42.2% + 42.3%)/42.3% =1.99
From the above, it can be seen that:
aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the reduction of the energy utilization rate is as follows: (in this example, d for the sake of brevity 4 -j 4 All assume 0.39,e of air conditioning unit example 1 4 -k 4 All adopt 0.61 of air conditioning unit example 1), namely [ 0.39 0.61 ] calculated by application example 1
PUE 1 ＝(X 1 ′*d 4 +X 2 ′*f 4 +X 3 ′*h 4 +X 4 ′*j 4 )/(X 4 ′*j 4 )
＝(0.12*0.39+0.252*0.39+0.05*0.39+0.423*0.39)/(0.423*0.39)＝2；
Aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the increase of the energy utilization rate is as follows:
PUE 2 ＝(X 1 ′*e 4 +X 2 ′*g 4 +X 3 ′*i 4 +X 4 ′*k 4 )/(X 4 ′*k 4 )
＝(0.12*0.61+0.252*0.61+0.05*0.61+0.423*0.61)/(0.423*0.61)＝2
the energy utilization efficiency of the data center at the next time period can be predicted from the above content, namely the PUE value is calculated, it is noted that the predicted increase and decrease values are the same as the values adopted in the data center are 0.39.61, but the differences exist in the actual situation;
further, the intelligent patrol inspection report content and format are shown in table 1:
TABLE 1 complete intelligent patrol inspection report form
Figure BDA0003869298830000081
/>
Figure BDA0003869298830000091
And S3, comprehensively analyzing the next period inspection energy consumption result and the energy consumption result predicted by the intelligent model, assisting the data center energy-saving system in inspection at the current period, and simultaneously predicting the next period inspection energy consumption transfer probability. If the prediction probability is high, the focus needs to be paid attention, and if the probability is low, the target condition is healthy. Thereby helping the data center analyze the energy consumption situation and providing artificial intelligence based data reference.
To supplement the concept of PUE are:
1. PUE refers to the abbreviation of Power usageeffectivenesss, is the ratio of all energy consumed by a data center to the energy consumed by an IT load, and is one of the most basic and effective indexes for evaluating the energy efficiency of the data center. The specific calculation formula is as follows: PUE = total data center energy consumption/IT equipment unit energy consumption.
2. Impact of PUE: the closer the PUE value is to 1, the higher the degree of greening of a data center. When the above value exceeds 1, IT indicates that the data center requires additional power overhead to support IT loads. Thus, the higher the value of PUE, the lower the overall efficiency of the data center.
It should be noted that the terms "upper", "lower", "left", "right", "front", "back", etc. used in the present invention are for clarity of description only, and are not intended to limit the scope of the present invention, and the relative relationship between the terms and the terms is not limited by the technical contents of the essential changes.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.

Claims (5)

1. A patrol inspection method for energy utilization efficiency of a data center is characterized by comprising the following steps:
s1: calculating the energy consumption ratio of each unit in the data center in the current time period in the whole data center;
s2: analyzing, for each unit in the data center and under a Markov chain-based model, a change in energy consumption of each unit in a next time period;
s3: and (3) predicting and calculating the energy utilization efficiency of the data center in the next time period by combining the data contents of the step (S1) and the step (S2) so as to detect and examine the devices of the data center in the machine room in a targeted manner and prevent the devices from aging to cause unnecessary energy consumption and power consumption.
2. The inspection method for the energy utilization efficiency of the data center according to claim 1, wherein the specific content of the step S1 is as follows:
each unit in the data center comprises an air conditioning unit, an air heat transfer unit, a power distribution system unit and an IT equipment unit; the IT equipment unit is a data center operation core, and represents that the energy utilization rate of the data center is higher when the integral energy consumption ratio of the data center is higher;
for the air conditioning unit, acquiring the electric energy a consumed by the air conditioning unit in the last time period based on historical data statistics 1 Electric energy b consumed by the air-conditioning unit in the current time period 1 And the electric energy consumption ratio X of the air conditioning unit in the whole data center in the last time period 1 And further calculating the electric quantity increasing and decreasing amplitude c of the air-conditioning unit in the current time period 1 ，c 1 ＝(b 1 -a 1 )/a 1 According to the increasing and decreasing amplitude c 1 Determining the electric energy ratio X consumed by the air conditioning unit in the data center in the current time period 1 ′，X 1 ′＝X 1 -(X 1 *c 1 ) (ii) a Wherein the last time period is also called as a last time period, and the current time period is also called as a present time period;
for the wind heat transfer unit, the electric energy a consumed by the wind heat transfer unit in the last time period is obtained based on historical data statistics 2 B electric energy consumed by wind heat transfer unit in current time period 2 And the electric energy consumed by the heat transfer units in the whole data center in the last time period is X 2 And further calculating the electric quantity increasing and decreasing amplitude c of the heat transfer unit in the current time period 2 ，c 2 ＝(b 2 -a 2 )/a 2 According to the increasing and decreasing amplitude c 2 Determining the electric energy ratio X of the wind heat transfer unit in the data center in the current time period 2 ′，X 2 ′＝X 2 -(X 2 *c 2 )；
For the power distribution system unit, acquiring the electric energy a consumed by the power distribution system unit in the last time period based on historical data statistics 3 B electric energy consumed by the power distribution system unit in the current time period 3 And at the previous oneElectric energy ratio X consumed by power distribution system units in data center in time period 3 And further calculating the electric quantity increasing and decreasing amplitude c of the power distribution system unit in the current time period 3 ，c 3 ＝(b 3 -a 3 )/a 3 According to increasing or decreasing amplitude c 3 Determining the electric energy ratio X of the distribution system unit in the data center in the current time period 3 ′，X 3 ′＝X 3 -(X 3 *c 3 )；
For the IT equipment unit, the electric energy a consumed by the IT equipment unit in the last time period is obtained based on historical data statistics 4 B electric energy consumed by IT equipment unit in current time period 4 And the electric energy proportion X consumed by the IT equipment units in the whole data center in the last time period 4 And further calculating the electric quantity increasing and decreasing amplitude c of the IT equipment unit in the current time period 4 ，c 4 ＝(b 4 -a 4 )/a 4 According to increasing or decreasing amplitude c 4 Determining the electric energy ratio X consumed by the IT equipment units in the data center in the current time period 4 ′，X 4 ′＝X 4 -(X 4 *c 4 )。
3. The inspection method for the energy utilization efficiency of the data center according to claim 2, wherein the wind heat removal unit comprises a water cooler, a humidifier, lighting and auxiliary equipment, a switch or a generator; the power distribution system unit comprises a UPS and a PDU; the IT equipment unit comprises an IT equipment CPU and a process load.
4. The inspection method for the energy utilization efficiency of the data center according to claim 2, wherein the specific content of the step S2 is as follows:
for an air conditioning unit, obtaining energy consumption transfer probability [ d ] of the air conditioning unit in the last period through historical data 1 、e 1 Energy consumption transfer probability [ d ] of air conditioning unit in this period 2 、e 2 Energy consumption transition probability [ d ] of units other than air conditioning unit 3 、e 3 Wherein d is 1 Representing the probability of energy consumption transfer of the air conditioning unit in the last period, e 1 Representing the energy consumption roll-out probability of the air conditioning unit; d 2 Representing the probability of energy transfer into the air conditioning unit during the time period, e 2 Representing the energy consumption roll-out probability of the air conditioning unit; d 3 Representing the probability of energy transfer from the unit other than the air conditioning unit to the air conditioning unit during the present time period, e 2 The energy consumption transfer probability that the energy consumption of other units except the air conditioning unit is not transferred to the air conditioning unit is represented;
and further calculating the energy consumption transfer probability [ d ] of the air conditioning unit in the next time period based on the Markov chain model 4 、e 4 H ]; wherein d is 4 Representing the probability of the transition of the energy consumption of the air conditioning unit into the next time period, i.e. the probability of the energy consumption of the air conditioning unit increasing in the next time period, and d 4 ＝d 1 *d 2 +e 1 *d 3 ，e 4 Representing the probability of the air conditioning unit's energy consumption rolling out in the next time period, i.e. the probability of the air conditioning unit's energy consumption decreasing in the next time period, and e 4 ＝d 1 *e 2 +e 1 *e 3 ；
For the wind heat transfer unit, the energy consumption transfer probability of the wind heat transfer unit in the last period is obtained through historical data [ f 1 、g 1 Energy consumption transfer probability [ f ] of wind heat transfer unit in this period 2 、g 2 Energy consumption transition probability [ f ] of units other than wind heat transfer unit 3 、g 3 Wherein f is 1 Representing the energy consumption transfer probability, g, of the heat transfer unit over the last period of time 1 Representing the energy consumption transfer probability of the wind heat transfer unit; f. of 2 Representing the energy transfer probability, g, of the heat transfer unit during this time period 2 Representing the energy consumption transfer probability of the wind heat transfer unit; f. of 3 Representing the probability of energy transfer from unit energy consumption other than the wind heat transfer unit to the wind heat transfer unit in the period, g 3 The energy consumption transfer probability that the energy consumption of other units except the wind heat transfer unit is not transferred to the wind heat transfer unit is represented;
and then calculating the energy of the wind heat transfer unit in the next time period based on the Markov chain modelProbability of attrition transfer [ f 4 、g 4 H ]; wherein f is 4 Representing the probability of the energy consumption of the heat transfer unit transitioning into the next time period, i.e. the probability of the energy consumption of the heat transfer unit increasing in the next time period, and f 4 ＝f 1 *f 2 +g 1 *f 3 ，g 4 Represents the energy consumption transfer-out probability of the heat transfer unit in the next time period, i.e., the probability of the heat transfer unit energy consumption being reduced in the next time period, and g 4 ＝f 1 *g 2 +g 1 *g 3 ；
Aiming at the power distribution system unit, the energy consumption transfer probability [ h ] of the power distribution system unit in the last period is obtained through historical data 1 、i 1 Energy consumption transfer probability [ h ] of power distribution system unit in this period 2 、i 2 Energy consumption transition probability [ h ] of other units except power distribution system unit 3 、i 3 Wherein h is 1 Representing the probability, i, of the transition of energy consumption of the power distribution system unit in the up period 1 Representing the energy consumption roll-out probability of the power distribution system unit; h is 2 Representing the probability, i, of the transition of energy consumption of the power distribution system unit in this time period 2 Representing the energy consumption roll-out probability of the power distribution system unit; h is a total of 3 Indicating the probability of energy transfer, i, into the distribution system unit of energy consumption of units other than the distribution system unit during the period 3 The energy consumption transition probability that the energy consumption of other units except the power distribution system unit is not transferred to the power distribution system unit is represented;
and further calculating the energy consumption transfer probability [ h ] of the power distribution system unit in the next time period based on the Markov chain model 4 、i 4 H ]; wherein h is 4 Represents the probability of the energy consumption of the distribution system unit shifting into the next time period, i.e. the probability of the energy consumption of the distribution system unit increasing in the next time period, and h 4 ＝h 1 *h 2 +i 1 *h 3 ，i 4 Represents the energy consumption roll-out probability of the power distribution system unit in the next time period, i.e. the probability of the power distribution system unit energy consumption reduction in the next time period, and i 4 ＝h 1 *i 2 +i 1 *i 3 ；
Acquiring the energy consumption transfer probability [ j ] of the IT equipment unit in the last period through historical data aiming at the IT equipment unit 1 、k 1 Energy consumption transfer probability [ j ] of IT equipment unit in this period 2 、k 2 Energy consumption transition probability [ j ] of units other than IT equipment units 3 、k 3 Wherein j 1 Representing the probability of energy consumption of the IT equipment unit, k, in the last period 1 Representing the energy consumption roll-out probability of the IT equipment unit; j is a function of 2 Representing the probability of energy consumption transfer, k, of the IT equipment unit in the present period 2 Representing the energy consumption transfer-out probability of the IT equipment unit; j is a unit of a group 3 Represents the energy consumption transition probability, k, of other units except the IT equipment unit to the IT equipment unit in the period 3 The energy consumption transfer probability that the energy consumption of other units except the IT equipment unit is not transferred to the IT equipment unit is represented;
and then calculating the energy consumption transfer probability [ j ] of the IT equipment unit in the next time period based on the Markov chain model 4 、k 4 H ]; wherein j is 4 Represents the probability of the energy consumption of the IT equipment unit turning into the next time slot, i.e. the probability of the energy consumption of the IT equipment unit increasing in the next time slot, and j 4 ＝j 1 *j 2 +k 1 *j 3 ，k 4 Represents the energy consumption transfer-out probability of the IT equipment unit in the next time period, i.e. the probability of the energy consumption reduction of the IT equipment unit in the next time period, and k 4 ＝j 1 *k 2 +k 1 *k 3 。
5. The inspection method for the energy utilization efficiency of the data center according to claim 4, wherein the specific content of the step S3 is as follows:
aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the reduction of the energy utilization rate is as follows:
PUE 1 ＝X 1 ′*d 4 +X 2 ′*f 4 +X 3 ′*h 4 +X 4 ′*j 4
aiming at the energy utilization efficiency of the data center in the next time period, a calculation formula for predicting the increase of the energy utilization rate is as follows:
PUE 2 ＝X 1 ′*e 4 +X 2 ′*g 4 +X 3 ′*i 4 +X 4 ′*k 4
therefore, operation and maintenance personnel detect and investigate the devices of the data center in the machine room according to the predicted increase or decrease of the energy utilization efficiency of the data center in the next time period and the energy consumption transition probability of each unit in the step S2 in combination, so as to prevent the devices from aging and unnecessary energy consumption and electricity consumption.
CN202211190833.XA 2022-09-28 2022-09-28 Inspection method for energy utilization efficiency of data center Pending CN115841320A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211190833.XA CN115841320A (en) 2022-09-28 2022-09-28 Inspection method for energy utilization efficiency of data center

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211190833.XA CN115841320A (en) 2022-09-28 2022-09-28 Inspection method for energy utilization efficiency of data center

Publications (1)

Publication Number Publication Date
CN115841320A true CN115841320A (en) 2023-03-24

Family

ID=85575018

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211190833.XA Pending CN115841320A (en) 2022-09-28 2022-09-28 Inspection method for energy utilization efficiency of data center

Country Status (1)

Country Link
CN (1) CN115841320A (en)

Similar Documents

Publication Publication Date Title
CN105956752A (en) Energy efficiency evaluation and diagnosis analysis method used for green public building operation phase
CN103208085B (en) Analysis intelligence system for improving load rate and reducing maximum demand of power utilization of enterprises
CN107036238B (en) Intelligent energy-saving control method for dynamically predicting external air and load
CN1869533A (en) Energy-saving device of network type digital multivariable central air conditioning system
CN115164361B (en) Data center control method and device, electronic equipment and storage medium
CN109765951B (en) Electrical cabinet temperature and humidity regulation method, system and medium based on heterogeneous data analysis
CN111380155B (en) Real-time online energy utilization diagnosis method for comprehensive efficiency of cold source operation
CN116887569B (en) Data center energy consumption prediction and energy saving adjustment method, system and storage medium
CN114996093B (en) Intelligent inspection optimization method for energy-saving system of data center
CN115841320A (en) Inspection method for energy utilization efficiency of data center
Kumar et al. Power usage efficiency (PUE) optimization with counterpointing machine learning techniques for data center temperatures
CN111780355A (en) Method and device for monitoring running state of air conditioner
CN204438420U (en) The monitoring and evaluation system that central air conditioning refrigerating plant room runs
CN115062914A (en) Intelligent comprehensive energy-saving service platform and energy-saving service method
CN115392684A (en) Data machine room carbon emission monitoring system and method based on out-of-band mode
CN113064075B (en) Motor service life prediction method based on edge calculation and deep learning
CN113779759A (en) Data center real-time energy management method and system
CN109471381A (en) Energy efficiency of equipment integrated control method based on big data fusion
CN114282683A (en) Early warning method and system for photovoltaic power station assembly
CN111290354B (en) Modularized data center waste heat efficient utilization control management system based on cloud computing
CN111538778A (en) Network boundary strategy robustness visualization control platform in electric power information
CN102520357A (en) Method for analyzing endurance of storage battery
CN117495203B (en) Multi-target electric energy management method and system based on energy storage system
CN115355731B (en) Energy-saving control device and control method for cooling tower
CN115081955B (en) Layered distributed fault handling system for comprehensive energy supply system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination