A control method and system for minimizing the power cost of a data center

Abstract

The present invention discloses a control method and system for minimizing the power cost of a data center, belonging to the technical field of data centers, the method comprising: dividing the working time of a UPS into multiple time periods, each time period corresponding to a certain duration T; calculating the total power consumption required by IT equipment and heat dissipation equipment in each time period, analyzing the total cost corresponding to three different UPS control strategies respectively adopted in each time period according to the calculation result of the total power consumption, and determining the UPS control strategy for each time period with the lowest total cost; and controlling the UPS to charge, discharge or stand still in each time period based on the UPS control strategy for each time period with the lowest total cost. The present invention divides the main load of the data center into IT equipment and heat dissipation equipment, and models the power consumption of each device respectively, so as to realize accurate control of the energy consumption of the data center and minimize the power cost of the data center.

Description

Technical Field

[0001] The present invention belongs to the technical field of data centers, and in particular relates to a control method and system for minimizing the power cost of a data center. Background Art

[0002] Internet Data Center (IDC) is the physical carrier of massive data. In order to meet users' increasing demand for computing, storage and network resources, the number of data centers is increasing and their scale is getting larger and larger, which also leads to a sharp increase in the energy consumption cost of these large-scale data centers.

[0003] The energy consumption of data centers mainly includes the energy consumption of IT equipment (for example, servers, network switches, etc.) and the energy consumption of heat dissipation equipment (for example, air conditioners, heat exchangers, etc.). How to reduce the power consumption of data centers is a current research hotspot in data centers.

[0004] At present, the main approach is to optimize the energy consumption of IT equipment and the energy consumption of cooling equipment. For example, based on the difference in electricity costs of data centers in different regions, computing tasks can be distributed to data centers in regions with lower electricity costs; or, cooling equipment with lower energy consumption (higher cooling efficiency) can be used, or even data centers can be built in locations with lower natural temperatures (such as plateaus) to reduce cooling needs. However, there are few solutions to directly optimize the power supply equipment in data centers.

[0005] Existing solutions mainly focus on load balancing of IT equipment and hardware improvements of cooling equipment. There are few solutions that directly optimize the power supply equipment of the data center. In addition, existing solutions can only roughly control the power supply equipment of the data center, making it difficult to accurately minimize electricity costs.

[0006] Prior art CN113131584A discloses a data center battery charging and discharging optimization control method and device, the method comprising: obtaining parameter information of a data center battery, the parameter information comprising at least one or more of the following: remaining power, battery capacity, and charging and discharging current; obtaining environmental information, the environmental information comprising real-time electricity prices and workload of IT equipment in a data center; based on the parameter information and the environmental information, using a deep reinforcement learning algorithm to control the charging and discharging of the data center battery.

[0007] The inadequacy of the existing technical documents is that the solution only considers the power consumption of IT equipment in the data center, while ignoring the heat dissipation equipment. The energy consumption of heat dissipation equipment can reach about 30% of the total energy consumption of the data center, which cannot be ignored. Since the energy consumption of heat dissipation equipment is generally assumed to be relatively stable, it is easy to ignore the heat dissipation equipment when dynamically optimizing the data center. In addition, the solution does not consider the life cost of the UPS itself, resulting in the energy cost being 0 when the UPS is powered in a certain period of time. It is easy to prematurely conclude that the UPS is the optimal action for a certain period of time during reinforcement learning, resulting in insufficient remaining power of the UPS during subsequent peak electricity prices, or causing the UPS to be charged and discharged multiple times, affecting the service life of the UPS. Summary of the invention

[0008] In order to solve the deficiencies in the prior art, the present invention provides a control method and system for minimizing the power cost of a data center.

[0009] The present invention adopts the following technical solution.

[0010] A first aspect of the present invention provides a control method for minimizing the power cost of a data center, comprising the following steps:

[0011] Divide the UPS working time into multiple time periods, each time period corresponds to a certain duration T;

[0012] Calculate the total power consumption required by IT equipment and cooling equipment in each time period. Based on the calculation results of the total power consumption, analyze the total costs corresponding to the three different UPS control strategies in each time period, and determine the UPS control strategy for each time period with the lowest total cost.

[0013] The UPS control strategy for each period with the lowest total cost is used to control the UPS to charge, discharge or standby in each period.

[0014] Optionally, the three different UPS control strategies include UPS charging, UPS discharging and UPS standing, and the three control strategies correspond to three actions respectively;

[0015] UPS charging means that the mains electricity charges the UPS, and the mains electricity supplies power to IT equipment and cooling equipment, and the mains electricity also supplies power to the UPS;

[0016] UPS discharge refers to the UPS supplying power to IT equipment and cooling equipment;

[0017] UPS static means that the UPS is neither charging nor discharging, and the mains power is used to power IT equipment and cooling equipment.

[0018] Optionally, calculating the total power consumption required by IT equipment and cooling equipment in each time period includes:

[0019] Get the power consumption of IT equipment in each time period:

[0020] In the i-th time period, if the power consumption of the IT equipment in the data center is , the power consumption of IT equipment for:

[0021] ;

[0022] Get the power consumption of the heat dissipation equipment in each time period:

[0023] Assume that the starting time point of the i-th time period is , the end time is , if at the start time , the initial actual temperature of the IT equipment is , and in the i-th time period, the temperature of the IT equipment needs to be reduced to the desired temperature , then at the end time of the i-th time period , the temperature of IT equipment should ideally be:

[0024] ;

[0025] In the i-th time period, the cooling device reduces the temperature of the IT equipment from Down to The power consumption required is ;

[0026] Get the total power consumption required by IT equipment and cooling equipment in each time period:

[0027] The total power consumption required by IT equipment and cooling equipment in the i-th time period , a total of:

[0028] .

[0029] Optionally, the heat dissipation device reduces the temperature of the IT equipment from Down to The power consumption required is include:

[0030] Get the power consumption of the cooling device It also includes predicting the power consumption of cooling equipment by fitting the relationship between expected temperature and power consumption.

[0031] Optionally, in the method, the total cost also includes the life cost of the UPS:

[0032] The charging efficiency of UPS is And the discharge efficiency is The value range is between 0 and 1;

[0033] If during a charging process or a discharging process, the UPS power change value is , then every time the UPS is charged or discharged Q The amount of electricity consumed, the lifetime cost It can be expressed as:

[0034] .

[0035] Optionally, among the total costs corresponding to the three different UPS control strategies, the total cost of UPS charging includes:

[0036] In the i-th time period, when the UPS is charging, the amount of electricity provided by the mains to the UPS is , the amount of electricity provided by the mains to IT equipment and cooling equipment is , the amount of electricity provided by the city power is:

[0037] ;

[0038] The charging efficiency of UPS is , then the amount of electricity that can be charged by UPS in time T is ;

[0039] The total cost in the i-th time period is include:

[0040] Power provided by the city electricity Cost, and UPS charging power Lifetime cost, total cost of UPS charging It can be expressed as:

[0041] ,

[0042] in, is the electricity price in the i-th time period.

[0043] Optionally, among the total costs corresponding to the three different UPS control strategies, the total cost when the UPS is discharging includes:

[0044] In the i-th time period, when the UPS is discharging, the amount of electricity provided by the mains is , then the total cost in the i-th time period is Including the life cost of UPS, the total cost when UPS is discharged It can be expressed as:

[0045] .

[0046] Optionally, among the total costs corresponding to the three different UPS control strategies, the total cost when the UPS is at rest is:

[0047] In the i-th time period, when the UPS is idle, the total cost in the i-th time period is include:

[0048] Power provided by the city electricity The cost includes the electricity provided by the mains The total cost It is expressed as:

[0049] .

[0050] Optionally, the UPS control strategy for each time period based on the lowest total cost also includes deciding whether to be powered by the mains in each time period, and whether to charge the UPS when the mains needs to be powered.

[0051] A second aspect of the present invention provides a control system for minimizing the power cost of a data center. Based on the control method for minimizing the power cost of a data center described in the first aspect of the present invention, the system includes:

[0052] Time period division module, power consumption calculation module, total cost calculation module and control module;

[0053] The time division module is used to divide the UPS working time into multiple time periods; the power consumption calculation module is used to calculate the total power consumption required by IT equipment and cooling equipment in each time period; the total cost calculation module is used to analyze the total cost corresponding to three different UPS control strategies in each time period, and determine the UPS control strategy for each time period with the lowest total cost; the control module controls the UPS to charge, discharge or stand still in each time period based on the UPS control strategy for each time period with the lowest total cost determined by the total cost calculation module.

[0054] The beneficial effect of the present invention is that, compared with the prior art,

[0055] 1. The present invention divides the main loads of the data center into IT equipment and heat dissipation equipment, and models the power consumption of each device respectively, so as to achieve accurate control of the energy consumption of the data center and minimize the electricity cost of the data center.

[0056] 2. The control strategies for each time period are divided into three types. Algorithms such as reinforcement learning can be used to easily solve the actions in each time period.

[0057] 3. By fitting the expected temperature and power consumption of the heat dissipation equipment, the power consumption of the heat dissipation equipment in different time periods can be predicted; and by configuring the expected temperature, the heat can be dissipated as much as possible during the electricity price trough, thereby optimizing the power consumption cost of the heat dissipation equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flow chart of the method of the present invention;

[0059] Figure 2 It is a graph of the dynamic changes of electricity prices over time;

[0060] Figure 3 The power consumption of the present invention With expected temperature A relationship between. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0062] The power supply equipment of the data center mainly includes diesel generators, UPS (Uninterruptible Power Supply), etc. When the city power is cut off or fails, the UPS can continuously provide short-term power to the data center, and then start the backup power supply such as diesel generators, and the diesel generators will provide stable power supply. With the development of UPS technology, its energy storage capacity has gradually increased, and it can provide power supply for data centers for a longer time. Based on the charge and discharge control of UPS, the energy consumption of the data center can be further reduced.

[0063] One current solution is to charge the UPS in the data center when the electricity price is low, and to use the UPS to power the data center when the electricity price is high. This solution can only roughly control the power supply equipment in the data center and it is difficult to accurately minimize the power consumption.

[0064] In embodiment 1, the present invention provides a control method for minimizing the power cost of a data center, such as Figure 1 As shown, the following steps are included:

[0065] Step 1, divide the working time of the UPS into multiple time periods, each time period corresponds to a certain duration T;

[0066] In a further preferred but non-limiting embodiment, in step 1, the number and duration T of the time periods are determined based on the electricity price variation cycle.

[0067] In a specific embodiment, in order to adapt to changes in electricity prices, a day is divided into multiple time periods based on the electricity price change cycle. For example, if the electricity price change cycle is 15 minutes, each time period is a 15-minute period, that is, T=15 minutes, and a day can be divided into 96 time periods in total.

[0068] Step 2: Calculate the total power consumption required by IT equipment and cooling equipment in each time period. According to the calculation results of the total power consumption, analyze the total costs corresponding to the three different UPS control strategies in each time period, and determine the UPS control strategy for each time period with the lowest total cost.

[0069] In a further preferred but non-limiting embodiment, in step 2, the three different UPS control strategies include UPS charging, UPS discharging and UPS static, and the three control strategies correspond to three actions respectively, so the action in each time period can be obtained based on reinforcement learning (RL), so as to realize the control strategy in each time period when the sum of the electric energy cost is minimized, and the charging and discharging of the UPS can be controlled based on the control strategy;

[0070] UPS charging means that the mains charges the UPS, that is, the mains supplies power to IT equipment and cooling equipment, and the mains also supplies power to the UPS;

[0071] UPS discharge refers to the UPS supplying power to IT equipment and cooling equipment;

[0072] UPS static means that the UPS is neither charging nor discharging, and the mains power is used to power IT equipment and cooling equipment.

[0073] In a further preferred but non-limiting embodiment, in step 2, calculating the total power consumption required by the IT equipment and the cooling equipment in each time period specifically includes:

[0074] Step 2.1, obtain the power consumption of IT equipment in each time period; in the i-th time period, if the power consumption of IT equipment in the data center is , the power consumption of IT equipment for ;

[0075] Among them, the power consumption of IT equipment It can be obtained based on historical data from a data center, or predicted based on a prediction model or other methods;

[0076] Step 2.2, obtaining the power consumption of the heat dissipation equipment in each time period;

[0077] Assume that the starting time point of the i-th time period is , the end time is , if at the start time , the initial actual temperature of the IT equipment is , and in the i-th time period, if the temperature of the IT equipment needs to be reduced to the desired temperature , then at the end time of the i-th time period , the temperature of IT equipment should ideally be ;

[0078] In the i-th time period, the cooling device reduces the temperature of the IT equipment from Down to The power consumption required is , and then the power is Provide explanation;

[0079] In a further preferred but non-limiting embodiment, in step 2.2, the power consumption of the heat dissipation device is obtained. It also includes predicting the power consumption of the cooling device by fitting the relationship between the expected temperature and power consumption, including:

[0080] In order to reduce the cost of cooling equipment as much as possible, When the temperature is low, the AC power will be used to dissipate heat for the heat dissipation device first, and the temperature of the heat dissipation device (for example, the coolant in the heat dissipation device) will be reduced as much as possible, so that when the UPS is used to supply power later, the heat dissipation device can consume less electricity, and the heat dissipation can be effectively achieved by using the heat dissipation device with a lower temperature.

[0081] Specifically, when the UPS is discharged, the power required by the cooling device can be Set to minimum , and the minimum value Can be compared with the start time point The actual temperature Regarding the actual temperature The lower the minimum The lower the initial actual temperature, the With minimum The corresponding table between them is used to determine the appropriate minimum value by looking up the table. ;

[0082] When the UPS is at rest or charging, the cooling device needs to be powered by the mains, so the temperature of the cooling device can be reduced as much as possible. However, the more the cooling device reduces the temperature, the higher the corresponding power consumption. Therefore, it is necessary to select the appropriate power consumption. .

[0083] The present invention pre-constructs a power consumption relationship for cooling the heat dissipation device from the actual temperature to the desired temperature at each initial actual temperature; for a certain actual temperature, the lower the desired temperature, the higher the power consumption, and the two are basically exponentially related; therefore, the heat dissipation device is cooled from the initial actual temperature to the desired temperature. Reduce to desired temperature , the power consumption required It can be expressed as: ; Wherein, a, b, c are corresponding coefficients, and the coefficients a, b, c can be solved by fitting;

[0084] Specifically, the actual temperature can be obtained Under the condition of Drop to different desired temperatures The power consumption at that time is then fitted by the least square method and other methods to obtain the power consumption With expected temperature The relationship between. Figure 3 A functional relationship between the two is shown, with the horizontal axis representing the desired temperature (°C) and the vertical axis representing the required power (MWh). After fitting, a=16.8948, b=10.3284, and c=1.8769 are determined;

[0085] To reduce the power consumption of cooling equipment, the expected temperature Real-time electricity prices There is a positive correlation between them, that is, the real-time electricity price The lower the temperature, the lower the cost of reducing the temperature of the cooling device (the lower the temperature of the cooling device, the lower the temperature of the IT equipment). Therefore, the temperature of the cooling device can be reduced as much as possible. It is also relatively low.

[0086] The expected temperature Mainly used to predict the power consumption of cooling equipment in various time periods , at the end time , the temperature of IT equipment Need to be actually measured; the temperature It is also the starting temperature of the i+1th time period, and the actual measured temperature shall prevail in the future.

[0087] Step 2.3, obtaining the total power consumption required by IT equipment and cooling equipment in each time period;

[0088] The total power consumption required by IT equipment and cooling equipment in the i-th time period , a total of .

[0089] In a further preferred but non-limiting embodiment, in step 2, the total cost also includes the life cost of the UPS, specifically:

[0090] The charging and discharging process of UPS involves the conversion between electrical energy and chemical energy, but due to the existence of losses, the energy conversion is not 100%. Assume that the charging efficiency of UPS is , the discharge efficiency is , both efficiencies are values ​​between 0 and 1;

[0091] The cost of a new UPS is , during the life of the UPS, it can withstand Cycle charge and discharge cycles, then the life cost of the UPS in each charge and discharge cycle is ;

[0092] If the life cost of the UPS charging process and the life cost of the discharging process are the same, then the life cost of the charging process or the discharging process Both , that is ;

[0093] Furthermore, if during a charging process or a discharging process, the UPS power change value is (When fully charged, can be the rated capacity of the UPS), then for every Q of electricity charged (or discharged) by the UPS, the lifetime cost consumed is Can be simplified to: .

[0094] In a further preferred but non-limiting embodiment, the total costs corresponding to the three different UPS control strategies in step 2 are specifically:

[0095] (1) In the i-th time period, when the UPS is charging,

[0096] The amount of electricity provided by the mains to the UPS is , the amount of electricity provided by the mains to IT equipment and cooling equipment is , the power provided by the city electricity is ;

[0097] Since the charging efficiency of UPS is , then the amount of electricity that can be charged by UPS in time T is ;

[0098] The total cost in the i-th time period is Including: Power provided by the city electricity Cost, and UPS charging power The total life cost It can be expressed as:

[0099] .

[0100] in, is the electricity price in the i-th time period, the unit can be ¥ / MWh, “¥” is the unit of currency;

[0101] For example, within two days (48 hours), the dynamic change diagram of a certain electricity price can be as follows: Figure 2 As shown, the horizontal axis is time (unit: h), and the vertical axis is the real-time electricity price, unit: ¥ / MWh;

[0102] (2) In the i-th time period, when the UPS is discharging, the amount of electricity provided by the mains is , then the total cost in the i-th time period is Includes: UPS life cost, so the total cost It can be expressed as: ;

[0103] (3) In the i-th time period, when the UPS is at rest,

[0104] The total cost in the i-th time period is Including: Power provided by the city electricity The cost of electricity provided by the city electricity The total cost is It can be expressed as: .

[0105] In a further preferred but non-limiting embodiment, in step 2, determining the UPS control strategy for each time period with the lowest total cost specifically includes:

[0106] The problem of minimizing the power cost of a data center can be expressed as minimizing the sum of the total costs of each time period within multiple time periods, such as within a day, that is, , n is the number of time periods, such as the number of time periods in a day, for example n = 96. This problem is actually a dynamic programming problem.

[0107] Step 3: Control the UPS to charge, discharge or standby in each time period based on the UPS control strategy for each time period with the lowest total cost.

[0108] Specifically, based on the UPS control strategy for each time period with the lowest total cost, it is decided whether to be powered by the mains in each time period, and whether to charge the UPS when the mains is needed.

[0109] In embodiment 2, the present invention provides a control system for minimizing the power cost of a data center. Based on the control method for minimizing the power cost of a data center described in embodiment 1 of the present invention, the system includes:

[0110] Time period division module, power consumption calculation module, total cost calculation module and control module;

[0111] The time division module is used to divide the UPS working time into multiple time periods; the power consumption calculation module is used to calculate the total power consumption required by IT equipment and cooling equipment in each time period; the total cost calculation module is used to analyze the total cost corresponding to three different UPS control strategies in each time period, and determine the UPS control strategy for each time period with the lowest total cost; the control module controls the UPS to charge, discharge or stand still in each time period based on the UPS control strategy for each time period with the lowest total cost determined by the total cost calculation module.

[0112] In summary, the present invention takes into account the power consumption of IT equipment and heat dissipation equipment at the same time. The power consumption of the two accounts for about 90% of the total power consumption of the data center, which can more accurately reflect the actual power consumption of the data center; and, by utilizing the characteristic that the power consumption of the heat dissipation equipment can be reduced when the IT equipment is at a low temperature, the power consumption of the heat dissipation equipment is predicted based on the expected temperature, and optimized in combination with the city electricity price, so that the power of the heat dissipation equipment can be increased when the electricity price is low, so as to reduce the temperature of the IT equipment as much as possible, so that when the electricity price rises later, the heat dissipation equipment can operate at low power, and the optimization of the heat dissipation equipment can be achieved. In addition, the UPS life cost is introduced, so that the cost when the UPS is powered is non-zero, and there is also a life cost when the UPS is charged, so that when making decisions based on reinforcement learning, the decision action of being powered by the UPS will not be easily obtained. While ensuring the overall low energy consumption, the number of UPS charge and discharge times can also be reduced, which can increase the service life of the UPS.

[0113] The technical solution provided by the present invention can more accurately reflect the actual power consumption of the data center, optimize the heat dissipation equipment, and increase the service life of the UPS.

[0114] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A control method for minimizing the power cost of a data center, characterized in that: The steps include: The UPS working time is divided into multiple time periods, each time period corresponds to a certain duration T, and the number of time periods and duration T are dynamically adjusted according to the electricity price change cycle; Calculating the total power consumption required by the IT equipment and the heat dissipation equipment in each time period, wherein calculating the power consumption of the heat dissipation equipment includes constructing an exponential model by fitting the relationship between the expected temperature of the IT equipment and the power consumption required by the heat dissipation equipment to reduce the temperature of the IT equipment to the expected temperature, and dynamically predicting the power consumption of the heat dissipation equipment; According to the calculation results of total power consumption, the total costs corresponding to the three different UPS control strategies in each time period are analyzed. The calculation of total power consumption includes: Get the power consumption of IT equipment in each time period: In the i-th time period, if the power consumption of the IT equipment in the data center is P IT ( i ), the power consumption of IT equipment Q IT ( i )for: Q IT ( i )= P IT ( i )×T； Get the power consumption of the heat dissipation equipment in each time period: Assume that the starting time point of the i-th time period is t i , the end time is t i+1 , if at the start time t i , the initial actual temperature of the IT equipment is Tem ( t i ), and in the i-th time period, the temperature of the IT equipment needs to be reduced to the desired temperature Tem E , then at the end time of the i-th time period t i+1 , the temperature of IT equipment should ideally be: Tem ( t i+1 )= Tem E ； In the i-th time period, the cooling device reduces the temperature of the IT equipment from Tem ( t i ) down to Tem E The power consumption required is Q COOL ( i ); Get the total power consumption required by IT equipment and cooling equipment in each time period: The total power consumption required by IT equipment and cooling equipment in the i-th time period Q IDC ( i ), a total of: Q IDC ( i )= Q IT ( i )+ Q COOL ( i ) The total cost includes the mains power supply cost and the UPS life cost. The UPS life cost calculation includes: in a charging process or a discharging process, the UPS power change value is Q max , then every time the UPS is charged or discharged Q The amount of electricity consumed, the lifetime cost C Q It can be expressed as: in, C Bat It indicates the life cost of a single charging or discharging process of the UPS. It is the quantitative value of the life loss cost corresponding to a single charge / discharge after the total cost of the UPS is evenly distributed to the charging and discharging stages according to its life cycle. Determine the UPS control strategy for each time period with the lowest total cost; The UPS control strategy for each period with the lowest total cost is used to control the UPS to charge, discharge or standby in each period.

2. A control method for minimizing the power cost of a data center according to claim 1, characterized in that: The three different UPS control strategies include UPS charging, UPS discharging and UPS static, and the three control strategies correspond to three actions respectively; UPS charging means that the mains electricity charges the UPS, and the mains electricity supplies power to IT equipment and cooling equipment, and the mains electricity also supplies power to the UPS; UPS discharge refers to the UPS supplying power to IT equipment and cooling equipment; UPS static means that the UPS is neither charging nor discharging, and the mains power is used to power IT equipment and cooling equipment.

3. A control method for minimizing the power cost of a data center according to claim 2, characterized in that: The heat dissipation device reduces the temperature of the IT equipment from Tem ( t i ) down to Tem E The power consumption required is Q COOL ( i )include: Get the power consumption of the cooling device Q COOL ( i ) also includes predicting the power consumption of cooling equipment by fitting the relationship between expected temperature and power consumption.

4. A control method for minimizing the power cost of a data center according to any one of claims 1 to 3, characterized in that: Among the total costs corresponding to the three different UPS control strategies, the total cost of UPS charging includes: In the i-th time period, when the UPS is charging, the amount of electricity provided by the mains to the UPS is R ( i ), the amount of electricity provided by the mains to IT equipment and cooling equipment is Q IDC ( i ), the amount of electricity provided by the city power is: Q ( i )= Q IDC ( i )+ R ( i )； The charging efficiency of UPS is , then the amount of electricity that can be charged by UPS in time T is ; The total cost in the i-th time period is C ( i )include: Power provided by the city electricity Q ( i ) and the amount of electricity charged by the UPS Lifetime cost, total cost of UPS charging C ( i ) can be expressed as: ， in, is the electricity price in the i-th time period.

5. A control method for minimizing the power cost of a data center according to any one of claims 1 to 3, characterized in that: Among the total costs corresponding to the three different UPS control strategies, the total cost of UPS discharge includes: In the i-th time period, when the UPS is discharging, the amount of electricity provided by the mains is Q ( i )=0, then the total cost in the i-th time period C ( i ) includes the life cost of the UPS and the total cost of the UPS when it is discharged C ( i ) can be expressed as: 。 6. A control method for minimizing the power cost of a data center according to any one of claims 1 to 3, characterized in that: Among the total costs corresponding to the three different UPS control strategies, the total cost when the UPS is at rest is: In the i-th time period, when the UPS is idle, the total cost in the i-th time period is C ( i )include: Power provided by the city electricity Q ( i ) includes the cost of electricity provided by the mains Q IDC ( i ) cost, the total cost C ( i ) is expressed as: 。 7. A control method for minimizing the power cost of a data center according to claim 1, characterized in that: The UPS control strategy for each time period based on the lowest total cost also includes deciding whether to be powered by the mains in each time period, and whether to charge the UPS when the mains is needed.

8. A control system for minimizing the power cost of a data center, based on a control method for minimizing the power cost of a data center as claimed in any one of claims 1 to 7, characterized in that: The system includes: Time period division module, power consumption calculation module, total cost calculation module and control module; The time division module is used to divide the UPS working time into multiple time periods; the power consumption calculation module is used to calculate the total power consumption required by IT equipment and cooling equipment in each time period; the total cost calculation module is used to analyze the total cost corresponding to three different UPS control strategies in each time period, and determine the UPS control strategy for each time period with the lowest total cost; the control module controls the UPS to charge, discharge or stand still in each time period based on the UPS control strategy for each time period with the lowest total cost determined by the total cost calculation module.

Images