Multi-objective collaborative optimization operation method of computing network considering computing task timing scheduling

Abstract

The present invention relates to the field of computer technology, and more particularly to a method for multi-objective collaborative optimization of computing network energy that considers the timing scheduling of computing power tasks. With the goal of maximizing data center efficiency and minimizing carbon emissions, the present invention establishes a multi-objective collaborative optimization operation model for computing network energy that considers the adjustable timing of computing power tasks. The optimization operation of the computing network system and the optimization operation of the energy system interact with each other through computing power tasks and energy supply as coupling points and boundaries, achieving a global optimal solution through collaborative iteration of computing power and energy. This method overcomes the inconsistency of traditional independent optimization schemes, achieves a high degree of coupling between computing power tasks and renewable energy generation, improves the renewable energy absorption rate, reduces data center energy costs, and realizes a multi-level interaction of computing power resources, computing power tasks, load demand, and energy supply, achieving the optimal overall operation plan for the computing network energy.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a computing network multi-objective collaborative optimization operation method that takes into account the timing scheduling of computing tasks. Background Art

[0002] The coordinated optimization operation of computing network and energy includes two parts: the optimization operation of computing power network system and the optimization operation of energy supply system. The optimization of computing power network system is used to meet the needs of various computing power tasks and formulate the optimal computing power task allocation plan. The optimization operation of energy supply system is used to meet the energy needs of computing power system and formulate the optimal electricity, cooling, heat and other energy system supply plans to ensure the safe, economical and reliable use of energy in computing power network system.

[0003] Currently, optimization of computing network systems and energy supply systems is typically performed separately. During computing network system optimization, the optimal computing task allocation plan is determined based on computing task demand, taking into account economic efficiency, environmental protection, and energy consumption. However, the power supply type and equipment output curve characteristics of the energy supply system are not considered. During energy supply system optimization, the computing network system load demand is used as a boundary condition, taking into account economic efficiency and environmental protection, to determine the optimal energy system supply plan. However, during computing network optimization, there is no data interaction or coordination between the computing network system optimization and the energy supply system optimization.

[0004] In a single-node computing network, all computing tasks are solely handled by a single-node data center, and the data center's energy supply is provided by a combination of local integrated energy stations and external power grids, natural gas networks, and hydrogen networks. Scheduling computing tasks and energy systems in a single-node data center requires comprehensive consideration of the timing optimization of computing tasks and the impact of energy pricing policies, renewable energy output, and energy system equipment. This prevents independent operational strategies for computing tasks and the energy system, which can lead to a suboptimal match between computing demand and the energy supply system.

[0005] Based on this, there is an urgent need for a multi-objective collaborative optimization operation method for computing networks that takes into account the timing scheduling of computing tasks to solve the technical problem of how to enable single-node computing networks to achieve collaborative operation. Summary of the Invention

[0006] In order to solve the technical problem of how to enable a single-node computing network to achieve collaborative operation, an embodiment of the present invention provides a method for multi-objective collaborative optimization operation of a computing network that takes into account the timing scheduling of computing tasks.

[0007] In a first aspect, an embodiment of the present invention provides a method for multi-objective collaborative optimization of a computing network that considers the timing scheduling of computing tasks. The method is applied to a single-node computing network and includes:

[0008] Establishing a multi-objective collaborative optimization model for computing network energy; wherein the multi-objective collaborative optimization model for computing network energy includes an upper layer model and a lower layer model;

[0009] The upper-layer model takes maximum benefit and minimum carbon emission as objective functions, latency constraints, computing power scale constraints, and energy supply constraints as constraints, and task timing allocation as decision variables to determine the computing power task allocation plan, and sends the computing power task allocation plan to the lower-layer model;

[0010] The lower-level model takes maximum benefit and minimum carbon emissions as objective functions, power balance constraints, power exchange constraints between the energy system and the upper-level network, surplus power grid constraints, and green power ratio constraints as constraints, and equipment operation strategies as decision variables to determine the energy supply plan, carbon emissions, and operating benefits, and sends the energy supply plan, carbon emissions, and operating benefits to the upper-level model;

[0011] The upper-layer model and the lower-layer model are repeatedly iterated to determine the optimal solution; wherein the optimal solution includes computing power task timing optimization and energy supply curve.

[0012] In a second aspect, an embodiment of the present invention further provides a computing network multi-objective collaborative optimization operation device that considers computing power task timing scheduling, including:

[0013] A model building module is used to establish a multi-objective collaborative optimization model for computing network energy; wherein the multi-objective collaborative optimization model for computing network energy includes an upper model and a lower model;

[0014] A first data processing module is used for the upper-layer model to determine a computing power task allocation plan with maximum benefit and minimum carbon emission as objective functions, latency constraints, computing power scale constraints, and energy supply constraints as constraints, and task timing allocation as decision variables, and to send the computing power task allocation plan to the lower-layer model;

[0015] The second data processing module is used for the lower-level model to determine the energy supply plan, carbon emissions and operating income with the maximum benefit and the minimum carbon emissions as the objective function, the power balance constraint, the power exchange constraint between the energy system and the upper network, the surplus power grid constraint, and the green power ratio constraint as the constraint conditions, and the equipment operation strategy as the decision variable, and send the energy supply plan, the carbon emissions and the operating income to the upper-level model;

[0016] The third data processing module is used to repeatedly iterate the upper-level model and the lower-level model to determine the optimal solution; wherein the optimal solution includes computing power task timing optimization and energy supply curve.

[0017] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of the present invention is implemented.

[0018] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method described in any embodiment of the present invention.

[0019] An embodiment of the present invention provides a multi-objective collaborative optimization operation method for computing network energy that takes into account the timing scheduling of computing power tasks. The present invention takes maximizing the benefits of data centers and minimizing carbon emissions as the goal, and establishes a multi-objective collaborative optimization operation model for computing network energy that takes into account the adjustable timing of computing power tasks. The computing network system optimization operation and the energy system optimization operation use computing power tasks and energy supply as coupling points and boundary interactions, and achieves the global optimal solution through computing power-energy collaborative iteration. It can overcome the inconsistency problem of operation plans of traditional independent optimization operations, achieve a high degree of coupling between computing power tasks and renewable energy power generation, improve the renewable energy absorption rate, reduce the energy consumption cost of data centers, and realize multi-level interaction of computing power resources-computing power tasks-load demand-energy supply, so as to achieve the optimal overall operation plan of computing network energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 This is a flow chart of a multi-objective collaborative optimization operation method for computing network energy considering the timing scheduling of computing tasks provided by an embodiment of the present invention;

[0022] Figure 2 This is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0023] Figure 3 This is a structural diagram of a computing network multi-objective collaborative optimization operation device that takes into account the timing scheduling of computing tasks, provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] Please refer to Figure 1 The embodiment of the present invention provides a method for multi-objective collaborative optimization operation of a computing network that considers the timing scheduling of computing tasks. The method is applicable to a single-node computing network and includes:

[0026] Step 100: Establishing a multi-objective collaborative optimization model for network computing. The multi-objective collaborative optimization model for network computing includes an upper layer model and a lower layer model.

[0027] Step 102: The upper-level model uses maximizing revenue and minimizing carbon emissions as its objective function, latency constraints, computing capacity constraints, and energy supply constraints as its constraints, and task timing allocation as its decision variable to determine a computing capacity task allocation plan. The upper-level model then sends the computing capacity task allocation plan to the lower-level model.

[0028] Step 104: The lower-level model uses maximizing revenue and minimizing carbon emissions as its objective function, and uses power balance constraints, power exchange constraints between the energy system and the upper-level network, surplus power grid-connected constraints, and green power ratio constraints as constraints. It uses the equipment operation strategy as a decision variable to determine the energy supply plan, carbon emissions, and operating revenue. The energy supply plan, carbon emissions, and operating revenue are then sent to the upper-level model.

[0029] Step 106: The upper-layer model and the lower-layer model are repeatedly iterated to determine the optimal solution; wherein the optimal solution includes computing task timing optimization and energy supply curve.

[0030] In the embodiment of the present invention, the present invention takes the maximum benefit of the data center and the lowest carbon emissions as the goal, and establishes a multi-objective collaborative optimization operation model of computing network energy that takes into account the adjustable timing of computing power tasks. The optimization operation of the computing network system and the optimization operation of the energy system use computing power tasks and energy supply as coupling points and boundary interactions, and achieves the global optimal solution through computing power-energy collaborative iteration. It can overcome the inconsistency problem of the operation plan of traditional independent optimization operation, achieve a high degree of coupling between computing power tasks and renewable energy power generation, improve the renewable energy absorption rate, reduce the energy consumption cost of data centers, and realize multi-level interaction of computing power resources-computing power tasks-load demand-energy supply, so as to achieve the optimal overall operation plan of computing network energy.

[0031] Described below Figure 1 How to perform the steps shown.

[0032] Regarding step 102:

[0033] In one embodiment of the present invention, the objective function is to maximize revenue and minimize carbon emissions, including the following formula:

[0034] max f i (y,e)

[0035] mix f e (y, e)

[0036] f i (y,e)=f in -f c

[0037] f in =f cp_profit +S up +S le,dp +S sub +S c

[0038]

[0039]

[0040] Where, max f i (y,e) is the objective function with the largest benefit, mix f e (y, e) is the objective function of minimizing carbon emissions, f in is the total revenue of the data center, f c is the total cost of the data center, f cp_profit is the computing power income, S up The surplus electricity is the income from the grid connection, S le,dp is the deviation income of green power medium and long term trading, S sub is the power generation subsidy income, S c Carbon trading income Representative Request m The computing power requirements, stands for data center dc n The unit price of computing power resources, Represents the data center dc of the day n The percentage of computing power requests processed in the total number of requests, H is the total number of time periods in a day, C up,e,h is the unit price of surplus electricity on the grid during period h, P up,e,h is the Internet power in period h, S i,le,dp is the deviation benefit of the i-th green power medium- and long-term transaction, C le,i,h,dp The i-th green power medium- and long-term transaction deviation penalty unit price in period h, P legm,i,h,dpis the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term trading in period h, where H is the total number of periods in a day, and C wt,e,h 、C pv,e,h are the electricity subsidies for wind power generation and photovoltaic power generation in period h, P wt,e,h 、P pv,e,h are the wind power and photovoltaic power consumed in period h, C c is the carbon trading price, E com is the carbon quota, E co2 is carbon dioxide emissions, is the carbon dioxide emission coefficient generated by the consumption of electricity, is the carbon dioxide emission coefficient of natural gas, C op is the energy purchase cost, C om For energy equipment maintenance costs, The cost of purchasing thermal power from the public grid, is the mid- and long-term transaction cost of the i-th green electricity, The normal green electricity purchase cost, is the penalty cost for the deviation of the i-th green power medium- and long-term transaction, The cost of purchasing natural gas, is the cost of purchasing hydrogen, H is the total number of hours in a day, C e,h is the unit price of public grid electricity purchase in period h, P e,h is the amount of electricity purchased from the public grid during period h, C le,i,h is the purchase price of the i-th green power medium- and long-term transaction in period h, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, C ge,h is the unit price of green electricity purchased during period h, P ge,h is the constant green electricity purchase amount in period h, C le,i,h,dp is the penalty price for the deviation of the i-th green power medium- and long-term transaction in period h, P le,i,h,dp is the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term transaction in period h, C g,h is the unit price of natural gas purchased during period h, G g,h is the natural gas purchase amount in period h, is the unit price of hydrogen gas purchased during the h period, is the amount of hydrogen purchased in period h, i is the number of energy supply production equipment, Ω1 is the set of energy supply production equipment, The operation and maintenance cost per unit output energy of the i-th energy supply production equipment, is the output energy of the i-th energy supply production equipment in time period h, j is the energy storage device number, Ω2 is the energy storage device set, is the operation and maintenance cost per unit charge and discharge power of the j-th energy storage device, is the charge and discharge power of the jth energy storage device in time period h, is CO2 emissions, is the emission coefficient of carbon dioxide generated by consuming electricity, is the carbon dioxide emission coefficient produced by natural gas.

[0041] In one embodiment of the present invention, the delay constraint is

[0042]

[0043] The above formula describes that the time from the task request arriving at the task entry router to being scheduled to the target data center should be less than the maximum tolerable delay of the task, and the task processing time should be after the task arrives. Indicates request r i processing time, Indicates that the task request

[0044] r i The total delay is composed of three parts: transmission delay, propagation delay and calculation delay, which are expressed as follows:

[0045] D total =(D tran +D pro +D com )Y m，l

[0046] Among them, the first item is the transmission delay, that is, the sum of the task passing through the servers on the transmission path, which is related to the bandwidth of the network path; the second item is the propagation delay (the actual physical distance divided by the speed of light); and the third item is the calculation delay, which is related to the task size and offloading decision.

[0047] Computing power scale constraints

[0048] The computing power facilities on a single-node data center DC are further represented as:

[0049] Device=((f,sum)

[0050] Among them, f represents the computing power of this type of device, sum represents the total number of devices of this type of computing devices in the data center, let d vrepresents an instantiated computing power facility object, then the computing power facility set on the data center can be represented as Device = {d1, d2, ..., d v Due to the computing power limitations of the data center servers, the computing power requested by the data center should be smaller than the maximum computing power currently available.

[0051]

[0052] in, Represents the computing resources required by the tasks accumulated in the data center.

[0053] Energy supply constraints

[0054] The energy consumption of the data center in time slot t is defined as the energy consumption generated by the data center processing tasks in the time slot. This energy consumption should be less than the total energy consumption resources that the energy system can provide in the time slot, that is:

[0055]

[0056] Where Q(t) represents the energy supply curve at time slot t.

[0057] Regarding step 104:

[0058] In one embodiment of the present invention, the power balance constraint includes an electric power balance constraint, a thermal power balance constraint, a cooling power balance constraint, a natural gas power balance constraint, and a hydrogen power balance constraint. The power balance constraint is determined by the following formula:

[0059] P e,h +∑P le,i,h +P ge,h +∑P e,m,h +ΣP e,j,ES-dis,h =ΣP e,k,h +P eL,h +P cpeL,h +ΣP e,j,ES-ch,h

[0060] ΣQ q,i,h +∑Q q,j,HS-dis,h =∑Q q,k,h +Q qL,h +ΣQ q,j,HS-ch,h

[0061] ΣC c,i,h +∑C c,j,HS-dis,h =C cL,h +∑C c,j,HS-ch,h

[0062] G g,h +∑G g,i,GS-dis,h =ΣG g,j,h +ΣGg,i,GS-ch,h

[0063]

[0064] Where Ω4 is the set of energy supply and production equipment that generates electricity, Ω5 is the set of energy storage equipment, Ω6 is the set of energy supply and production equipment that consumes electricity, and P e,m,h P is the electric power generated by the mth energy supply production equipment that generates electric energy in time period h, e,j,ES-dis,h is the discharge power of the jth storage device in time period h, P e,k,h P is the power consumed by the kth energy supply device consuming electricity in time period h, eL,h is the other electrical load during period h, P e,j,ES-ch,h is the charging power of the jth storage device in time period h, P cpeL,h is the power load consumed by the computing equipment in period h, Ω7 is the set of energy supply and production equipment that generates heat energy, Ω8 is the set of heat storage equipment, Ω9 is the set of energy supply and production equipment that consumes heat energy, Q q,i,h The thermal power generated by the energy supply production equipment that generates heat energy in time period h, Q q,j,HS-dis,h is the heat release power of the jth heat storage device in time period h, Q q,k,h The heat load of the energy supply production equipment for the kth heat-consuming energy in time period h, Q qL,h is the heat load during period t, Q q,j,HS-ch,h is the charging power of the jth heat storage device in time period h, Ω 10 A collection of energy supply production equipment for generating cold energy, Ω 11 For cold storage equipment collection, C c,i,h is the cooling power generated by the energy supply production equipment that generates cooling energy in the i-th period h, C c,j,HS-dis,h is the cooling power of the j-th cold storage device in time period h, C cL,h is the cooling load in period h, C c,j,HS-ch,h is the cooling power of the j-th cold storage device in time period h, Ω3 is the set of energy supply production equipment consuming natural gas, Ω 12 G is the gas storage equipment collection. g,h The natural gas purchase power in period h, G g,i,GS-dis,h is the gas discharge power of the j-th gas storage device in period h, G g,j,h The natural gas power of the j-th energy supply production equipment consuming natural gas in period h, G g,i,GS-ch,h The filling power Ω of the j-th gas storage device in period h 14 Energy supply production equipment collection for consuming hydrogen, Ω 13 G is a hydrogen storage device collection. H2,h is the amount of hydrogen purchased during period h, is the hydrogen discharge power of the jth hydrogen storage device in period h, The hydrogen power of the energy supply production equipment for the jth hydrogen-consuming energy in period h, The hydrogen charging power of the jth hydrogen storage device in period h.

[0065] In one embodiment of the present invention, the power exchange constraint between the energy system and the upper-level network is determined by the following formula:

[0066]

[0067] Where, are the minimum and maximum power purchases from the public grid thermal power, Purchase gas power for minimum and maximum, Purchase hydrogen power for minimum and maximum respectively, are the minimum and maximum constant green electricity purchase power, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, P legm,i,h is the amount of green electricity provided in the i-th medium- and long-term transaction in period h.

[0068] In one embodiment of the present invention, the green power ratio constraint

[0069] GE rate ≥GE rate_nor

[0070] In the formula, GE rate_nor The lower limit of the proportion of green electricity.

[0071] Restrictions on surplus power access

[0072] Different regions have different requirements for the power of surplus power connected to the grid. Some regions guarantee full grid access, while others set limits on the power of surplus power connected to the grid for integrated power generation, grid-load-storage projects. Therefore, it is necessary to consider the constraints on the power of surplus power connected to the grid. The constraints on the power of surplus power connected to the grid must meet two requirements: (1) it must not exceed the maximum surplus power receiving power constraint of the upper-level network; and (2) it must not exceed a certain percentage of the installed capacity.

[0073]

[0074] P up,e,h ≤τ·S sum,ren

[0075] Where, P up,e,h is the online power consumption in period h (kW); is the maximum residual power receiving power of the upper network (kW); S sum,ren is the sum of the installed capacity of renewable energy power generation (including wind turbines and photovoltaics, kW); τ is the maximum coefficient (%) of surplus power connected to the grid to the installed capacity of renewable energy power generation.

[0076] Regarding step 106:

[0077] In one embodiment of the present invention, step 106 may specifically include:

[0078] Determine economic and environmental goals based on the upper and lower models;

[0079] Normalize the economic goals and environmental protection goals to obtain normalized results;

[0080] Based on the normalized results, the optimal solution is determined.

[0081] In one embodiment of the present invention, the normalized processing result includes a normalized value of the environmental target and a normalized value of the economic target, and the normalized processing result is determined by the following formula:

[0082]

[0083] ω1+ω2=1

[0084] In the formula, ω1 and ω2 are weight coefficients, is the normalized value of the economic objective of the i-th solution in the current iteration, is the normalized value of the environmental protection target of the i-th solution in the current iteration, max i (f i,c ) is the maximum value of the economic objective among all solutions in the current iteration, min i (f i,c ) is the minimum value of the economic objective among all solutions in the current iteration, max i (f i,e ) is the maximum value of the environmental protection target among all solutions in the current iteration, min i (f i,e ) is the minimum value of the environmental protection target among all solutions in the current iteration, and M is the number of solutions in the current iteration.

[0085] In the embodiment of the present invention, a non-dominated sorting genetic algorithm is used for solving the problem. During the iteration process, the economic goal and the environmental goal are added together by weighted values. Since the economic goal and the environmental goal have different dimensions, they cannot be added together directly. Therefore, the two goals need to be normalized.

[0086] In one embodiment of the present invention, the method adopted by the present invention is solved by the NSGA-II algorithm.

[0087] In this embodiment, in order to solve the multi-objective optimization problem, the present invention adopts the NSGA-II algorithm, which can effectively obtain the Pareto solution (optimal solution) by adding a fast non-dominated sorting algorithm, a crowding distance, and an elite retention strategy. In the non-dominated sorting genetic algorithm that can be run collaboratively by the computing network, the crowding degree comparison method is used to replace the shared function method in NSGA, which to a certain extent eliminates the algorithm performance being too dependent on the shared parameter σ. share The new algorithm maintains diversity among population members without requiring user-defined parameters. Furthermore, the algorithm network can collaboratively run the non-dominated sorting genetic algorithm, which employs methods with lower computational complexity.

[0088] In summary, the multi-objective collaborative optimization operation method of computing network energy considering the timing scheduling of computing power tasks takes into account the mutual influence of the adjustability of computing power task timing and the supply of energy system, and takes the maximum benefit of data center and the lowest carbon emission as multiple objectives. The multi-objective collaborative optimization operation model of computing network energy considering the adjustability of computing power task timing is established. The optimization operation of computing network system and the optimization operation of energy system are coupled with computing power tasks and energy supply as the boundary interaction, and the global optimal Pareto solution set is achieved through the collaborative iteration of computing power-energy. It can overcome the inconsistency problem of the operation plan of traditional independent optimization operation, realize the high coupling degree of computing power tasks and renewable energy power generation, improve the renewable energy absorption rate, reduce the energy consumption cost of data center, realize the multi-level interaction of computing power resources-computing power tasks-load demand-energy supply, and achieve the optimal overall operation plan of computing network energy.

[0089] like Figure 2 、 Figure 3 As shown, the embodiment of the present invention provides a multi-objective collaborative optimization operation device for computing network considering the timing scheduling of computing tasks. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 2 As shown in the figure, it is a hardware architecture diagram of an electronic device where a computing network multi-objective collaborative optimization operation device is located, which takes into account the timing scheduling of computing tasks provided by an embodiment of the present invention. Figure 3 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 3 As shown, as a device in a logical sense, it is formed by the CPU of the electronic device in which it is located reading the corresponding computer program in the non-volatile memory into the internal memory and running it.

[0090] like Figure 3 As shown, this embodiment provides a computing network multi-objective collaborative optimization operation device that considers computing power task timing scheduling, and the device includes:

[0091] The model building module 300 is used to establish a multi-objective collaborative optimization model for computing network energy; wherein the multi-objective collaborative optimization model for computing network energy includes an upper layer model and a lower layer model;

[0092] The first data processing module 302 is used for the upper-layer model to determine a computing task allocation plan based on maximizing benefits and minimizing carbon emissions as objective functions, latency constraints, computing power scale constraints, and energy supply constraints as constraints, and task timing allocation as decision variables, and to send the computing task allocation plan to the lower-layer model;

[0093] The second data processing module 304 is used for the lower-level model to determine the energy supply plan, carbon emissions, and operating income with the maximum benefit and the minimum carbon emissions as the objective function, the power balance constraint, the power exchange constraint between the energy system and the upper network, the surplus power grid constraint, and the green power ratio constraint as the constraint conditions, and the equipment operation strategy as the decision variable, and send the energy supply plan, the carbon emissions, and the operating income to the upper-level model;

[0094] The third data processing module 306 is used to repeatedly iterate the upper-layer model and the lower-layer model to determine the optimal solution; wherein the optimal solution includes computing power task timing optimization and energy supply curve.

[0095] In an embodiment of the present invention, the model building module 300 can be used to execute step 100 in the above method embodiment, the first data processing module 302 can be used to execute step 102 in the above method embodiment, the second data processing module 304 can be used to execute step 104 in the above method embodiment, and the third data processing module 306 can be used to execute step 106 in the above method embodiment.

[0096] In one embodiment of the present invention, the objective function of maximizing benefits and minimizing carbon emissions includes the following formula:

[0097] max f i (y,e)

[0098] mix f e (y, e)

[0099] f i (y,e)=f in -f c

[0100] f in =f cp_profit +S up +S le,dp +S sub +S c

[0101]

[0102]

[0103] Where, max f i (y,e) is the objective function with the maximum benefit, mix f e (y, e) is the objective function of minimizing carbon emissions, f in is the total revenue of the data center, f c is the total cost of the data center, f cp_profit is the computing power income, S up The surplus electricity is the income from the grid connection, S le,dp is the deviation income of green power medium and long term trading, S sub is the power generation subsidy income, S c Carbon trading income, of which C rm Representative Request m The computing power requirements, stands for data center dc n The unit price of computing power resources, Represents the data center dc of the day n The percentage of computing power requests processed in the total number of requests, H is the total number of time periods in a day, C up,e,h is the unit price of surplus electricity on the grid during period h, P up,e,h is the Internet power in period h, S i,le,dp is the deviation benefit of the i-th green power medium- and long-term transaction, C le,i,h,dp The i-th green power medium- and long-term transaction deviation penalty unit price in period h, P legm,i,h,dp is the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term trading in period h, where H is the total number of periods in a day, and C wt,e,h 、C pv,e,h are the electricity subsidies for wind power generation and photovoltaic power generation in period h, P wt,e,h 、P pv,e,h are the wind power and photovoltaic power consumed in period h, C c is the carbon trading price, E com is the carbon quota, E co2 is carbon dioxide emissions, is the carbon dioxide emission coefficient generated by the consumption of electricity, is the carbon dioxide emission coefficient of natural gas, C op is the energy purchase cost, C om For energy equipment maintenance costs, The cost of purchasing thermal power from the public grid, is the mid- and long-term transaction cost of the i-th green electricity, The normal green electricity purchase cost, is the penalty cost for the deviation of the i-th green power medium- and long-term transaction, The cost of purchasing natural gas, is the cost of purchasing hydrogen, H is the total number of hours in a day, C e,h is the unit price of public grid electricity purchase in period h, P e,h is the amount of electricity purchased from the public grid during period h, C le,i,h is the purchase price of the i-th green power medium- and long-term transaction in period h, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, C ge,h is the unit price of green electricity purchased during period h, P ge,h is the constant green electricity purchase amount in period h, C le,i,h,dp is the penalty price for the deviation of the i-th green power medium- and long-term transaction in period h, P le,i,h,dp is the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term transaction in period h, C g,h is the unit price of natural gas purchased during period h, G g,h is the natural gas purchase amount in period h, is the unit price of hydrogen gas purchased during the h period, G H2,h is the amount of hydrogen purchased in period h, i is the number of energy supply production equipment, Ω1 is the set of energy supply production equipment, The operation and maintenance cost per unit output energy of the i-th energy supply production equipment, is the output energy of the i-th energy supply production equipment in time period h, j is the energy storage device number, Ω2 is the energy storage device set, is the operation and maintenance cost per unit charge and discharge power of the j-th energy storage device, is the charge and discharge power of the jth energy storage device in time period h, is CO2 emissions, is the emission coefficient of carbon dioxide generated by consuming electricity, is the carbon dioxide emission coefficient produced by natural gas.

[0104] In one embodiment of the present invention, the third data processing unit repeatedly iterates the upper layer model and the lower layer model to determine the optimal solution, and is used to perform the following operations:

[0105] Determining economic goals and environmental protection goals based on the upper model and the lower model;

[0106] Normalizing the economic goal and the environmental protection goal to obtain a normalized result;

[0107] Based on the normalization result, the optimal solution is determined.

[0108] In one embodiment of the present invention, the normalized processing result includes a normalized value of the environmental target and a normalized value of the economic target, and the normalized processing result is determined by the following formula:

[0109]

[0110] ω1+ω2=1

[0111] In the formula, ω1 and ω2 are weight coefficients, is the normalized value of the economic objective of the i-th solution in the current iteration, is the normalized value of the environmental protection target of the current iteration i-th solution, max i (f i,c ) is the maximum value of the economic objective among all solutions in the current iteration, min i (f i,c ) is the minimum value of the economic objective among all solutions in the current iteration, max i (f i,e ) is the maximum value of the environmental protection target among all solutions in the current iteration, min i (f i,e ) is the minimum value of the environmental protection target among all solutions in the current iteration, and M is the number of solutions in the current iteration.

[0112] In one embodiment of the present invention, the power balance constraint includes an electric power balance constraint, a thermal power balance constraint, a cooling power balance constraint, a natural gas power balance constraint, and a hydrogen power balance constraint, and the power balance constraint is determined by the following formula:

[0113] P e,h +ΣP le,i,h +P ge,h +ΣP e,m,h +∑P e,j,ES-dis,h =∑P e,k,h +P eL,h +P cpeL,h +∑P e,j,ES-ch,h

[0114] ∑Q q,i,h +∑Q q,j,HS-dis,h =∑Q q,k,h +Q qL,h +∑Q q,j,HS-ch,h

[0115] ∑C c,i,h +∑Cc,j,HS-dis,h =C cL,h +∑C c,j,HS-ch,h

[0116] G g,h +∑G g,i,GS-dis,h =∑G g,j,h +∑G g,i,GS-ch,h

[0117]

[0118] Where Ω4 is the set of energy supply and production equipment that generates electricity, Ω5 is the set of energy storage equipment, Ω6 is the set of energy supply and production equipment that consumes electricity, and P e,m,h P is the electric power generated by the mth energy supply production equipment that generates electric energy in time period h, e,j,ES-dis,h is the discharge power of the jth storage device in time period h, P e,k,h P is the power consumed by the kth energy supply device consuming electricity in time period h, eL,h is the other electrical load during period h, P e,j,ES-ch,h is the charging power of the jth storage device in time period h, P cpeL,h is the power load consumed by the computing equipment in period h, Ω7 is the set of energy supply and production equipment that generates heat energy, Ω8 is the set of heat storage equipment, Ω9 is the set of energy supply and production equipment that consumes heat energy, Q q,i,h The thermal power generated by the energy supply production equipment that generates heat energy in time period h, Q q,j,HS-dis,h is the heat release power of the jth heat storage device in time period h, Q q,k,h The heat load of the energy supply production equipment for the kth heat-consuming energy in time period h, Q qL,h is the heat load during period t, Q q,j,HS-ch,h is the charging power of the jth heat storage device in time period h, Ω 10 A collection of energy supply production equipment for generating cold energy, Ω 11 For cold storage equipment collection, C c,i,h is the cooling power generated by the energy supply production equipment that generates cooling energy in the i-th period h, C c,j,HS-dis,h is the cooling power of the j-th cold storage device in time period h, C cL,h is the cooling load in period h, C c,j,HS-ch,h is the cooling power of the j-th cold storage device in time period h, Ω3 is the set of energy supply production equipment consuming natural gas, Ω 12 G is the gas storage equipment collection. g,h The natural gas purchase power in period h, G g,i,GS-dis,h is the gas discharge power of the j-th gas storage device in period h, G g,j,h The natural gas power of the j-th energy supply production equipment consuming natural gas in period h, G g,i,GS-ch,hThe filling power Ω of the j-th gas storage device in period h 14 Energy supply production equipment collection for consuming hydrogen, Ω 13 For hydrogen storage equipment collection, is the amount of hydrogen purchased during period h, is the hydrogen discharge power of the jth hydrogen storage device in period h, The hydrogen power of the energy supply production equipment for the jth hydrogen-consuming energy in period h, The hydrogen charging power of the jth hydrogen storage device in period h.

[0119] In one embodiment of the present invention, the power exchange constraint between the energy system and the upper-level network is determined by the following formula:

[0120]

[0121] 0≤P le,i,h ≤P legm,i,h

[0122] Where, are the minimum and maximum power purchases from the public grid thermal power, Purchase gas power for minimum and maximum, respectively, Purchase hydrogen power for minimum and maximum respectively, are the minimum and maximum long-term green electricity purchase power, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, P legm,i,h is the amount of green electricity provided in the i-th medium- and long-term transaction in period h.

[0123] In one embodiment of the present invention, the method is solved by the NSGA-II algorithm.

[0124] It is understandable that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a computing network multi-objective collaborative optimization operation device that takes into account the timing scheduling of computing tasks. In other embodiments of the present invention, a computing network multi-objective collaborative optimization operation device that takes into account the timing scheduling of computing tasks may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The components shown in the figure can be implemented in hardware, software, or a combination of software and hardware.

[0125] The information interaction, execution process, etc. between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For specific contents, please refer to the description in the embodiment of the method of the present invention and will not be repeated here.

[0126] An embodiment of the present invention also provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory. When the processor executes the computer program, a computing network multi-objective collaborative optimization operation method considering the timing scheduling of computing power tasks in any embodiment of the present invention is implemented.

[0127] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a computing network multi-objective collaborative optimization operation method that considers the timing scheduling of computing power tasks in any embodiment of the present invention.

[0128] Specifically, a system or device equipped with a storage medium can be provided, on which software program codes that implement the functions of any of the above-mentioned embodiments are stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program codes stored in the storage medium.

[0129] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute part of the present invention.

[0130] Examples of storage media for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, the program code can be downloaded from a server computer via a communication network.

[0131] In addition, it should be clear that the functions of any of the above embodiments can be achieved not only by executing the program code read by the computer, but also by enabling the operating system operating on the computer to complete part or all of the actual operations based on the instructions of the program code.

[0132] In addition, it can be understood that the program code read from the storage medium is written into the memory provided in the expansion board inserted into the computer or into the memory provided in the expansion module connected to the computer, and then based on the instructions of the program code, the CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above embodiments.

[0133] It should be noted that, in this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0134] Those skilled in the art will understand that all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk, etc. Various media that can store program codes.

[0135] 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multi-objective collaborative optimization operation method for computing network energy considering the timing scheduling of computing tasks, characterized by: The method is applied to a single-node computing network and includes: Establishing a multi-objective collaborative optimization model for computing network energy; wherein the multi-objective collaborative optimization model for computing network energy includes an upper layer model and a lower layer model; The upper-layer model takes maximum benefit and minimum carbon emission as objective functions, latency constraints, computing power scale constraints, and energy supply constraints as constraints, and task timing allocation as decision variables to determine the computing power task allocation plan, and sends the computing power task allocation plan to the lower-layer model; The lower-level model takes maximum benefit and minimum carbon emissions as objective functions, power balance constraints, power exchange constraints between the energy system and the upper-level network, surplus power grid constraints, and green power ratio constraints as constraints, and equipment operation strategies as decision variables to determine the energy supply plan, carbon emissions, and operating benefits, and sends the energy supply plan, carbon emissions, and operating benefits to the upper-level model; The upper-layer model and the lower-layer model are repeatedly iterated to determine an optimal solution; wherein the optimal solution includes computing power task timing optimization and energy supply curve; The upper model and the lower model are repeatedly iterated to determine the optimal solution, including: Determining economic goals and environmental protection goals based on the upper model and the lower model; Normalizing the economic goal and the environmental protection goal to obtain a normalized result; Determining the optimal solution based on the normalization processing result; The normalized processing result includes the normalized value of the environmental protection target and the normalized value of the economic target, and the normalized processing result is determined by the following formula: ω1+ω2=1 In the formula, ω1 and ω2 are weight coefficients, is the normalized value of the economic objective of the i-th solution in the current iteration, is the normalized value of the environmental protection target of the current iteration i-th solution, max i (f i,c ) is the maximum value of the economic objective among all solutions in the current iteration, min i (f i,c ) is the minimum value of the economic objective among all solutions in the current iteration, max i (f i,e ) is the maximum value of the environmental protection target among all solutions in the current iteration, min i (f i,e ) is the minimum value of the environmental protection target among all solutions in the current iteration, and M is the number of solutions in the current iteration.

2. The method according to claim 1, characterized in that The objective function of maximizing benefits and minimizing carbon emissions includes the following formula: max f i (y,e) mix f e (y,e) f i (y,e)=f in -f c f in =f cp_profit +S up +S le,dp +S sub +S c S c =C c ·(E nom -E co2 ) / 10000 f c =C op +C om f e (y,e)=E CO2 Where, max f i (y,e) is the objective function with the largest benefit, mixf e (y, e) is the objective function of minimizing carbon emissions, f in is the total revenue of the data center, f c is the total cost of the data center, f cp_profit is the computing power income, S up The surplus electricity is the income from the grid connection, S le,dp is the deviation income of green power medium and long term trading, S sub is the power generation subsidy income, S c For carbon trading income, Representative Request m The computing power requirements, stands for data center dc n The unit price of computing power resources, Represents the data center dc of the day n The percentage of computing power requests processed in the total number of requests, H is the total number of time periods in a day, C up,e,h is the unit price of surplus electricity on the grid during period h, P up,e,h is the Internet power in period h, S i,le,dp is the deviation benefit of the i-th green power medium- and long-term transaction, C le,i,h,dp The i-th green power medium- and long-term transaction deviation penalty unit price in period h, P legm,i,h,dp is the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term trading in period h, where H is the total number of periods in a day, and C wt,e,h 、C pv,e,h are the electricity subsidies for wind power generation and photovoltaic power generation in period h, P wt,e,h 、P pv,e,h are the wind power and photovoltaic power consumed in period h, C c is the carbon trading price, E nom is the carbon quota, E co2 is carbon dioxide emissions, is the carbon dioxide emission coefficient generated by the consumption of electricity, is the carbon dioxide emission coefficient of natural gas, C op is the energy purchase cost, C om For energy equipment maintenance costs, The cost of purchasing thermal power from the public grid, is the mid- and long-term transaction cost of the i-th green electricity, The normal green electricity purchase cost, is the penalty cost for the deviation of the i-th green power medium- and long-term transaction, The cost of purchasing natural gas, is the cost of purchasing hydrogen, H is the total number of hours in a day, C e,h is the unit price of public grid electricity purchase in period h, P e,h is the amount of electricity purchased from the public grid during period h, C le,i,h is the purchase price of the i-th green power medium- and long-term transaction in period h, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, C ge,h is the unit price of green electricity purchased during period h, P ge,h is the constant green electricity purchase amount in period h, C le,i,h,dp is the penalty price for the deviation of the i-th green power medium- and long-term transaction in period h, P le,i,h,dp is the deviation of the i-th green power medium- and long-term transaction caused by the power purchaser’s responsibility in period h, P lest,i,h The transaction volume of the i-th green power medium- and long-term transaction contract in period h, P legm,i,h The amount of green electricity provided by the i-th medium- and long-term transaction in period h, C g,h is the unit price of natural gas purchased during period h, G g,h is the natural gas purchase amount in period h, is the unit price of hydrogen purchased during the h period, is the amount of hydrogen purchased in period h, i is the number of energy supply production equipment, Ω1 is the set of energy supply production equipment, The operation and maintenance cost per unit output energy of the i-th energy supply production equipment, is the output energy of the i-th energy supply production equipment in time period h, j is the energy storage device number, Ω2 is the energy storage device set, is the operation and maintenance cost per unit charge and discharge power of the j-th energy storage device, is the charge and discharge power of the jth energy storage device in time period h, is CO2 emissions, is the emission coefficient of carbon dioxide generated by consuming electricity, is the carbon dioxide emission coefficient produced by natural gas.

3. The method according to claim 2, characterized in that The power balance constraints include electric power balance constraints, thermal power balance constraints, cooling power balance constraints, natural gas power balance constraints, and hydrogen power balance constraints. The power balance constraints are determined by the following formula: Where Ω4 is the set of energy supply and production equipment that generates electricity, Ω5 is the set of energy storage equipment, Ω6 is the set of energy supply and production equipment that consumes electricity, and P e,m,h P is the electric power generated by the mth energy supply production equipment that generates electric energy in time period h, e,j,ES-dis,h is the discharge power of the jth storage device in time period h, P e,k,h P is the power consumed by the kth energy supply device consuming electricity in time period h, eL,h is the other electrical load during period h, P e,j,ES-ch,h is the charging power of the jth energy storage device in time period h, P cpeL,h is the power load consumed by the computing equipment in period h, Ω7 is the set of energy supply and production equipment that generates heat energy, Ω8 is the set of heat storage equipment, Ω9 is the set of energy supply and production equipment that consumes heat energy, Q q,i,h The thermal power generated by the energy supply production equipment that generates heat energy in time period h, Q q,j,HS-dis,h is the heat release power of the jth heat storage device in time period h, Q q,k,h The heat load of the energy supply production equipment for the kth heat-consuming energy in time period h, Q qL,h is the heat load during period t, Q q,j,HS-ch,h is the charging power of the jth heat storage device in time period h, Ω 10 A collection of energy supply production equipment for generating cold energy, Ω 11 For cold storage equipment collection, C c,i,h is the cooling power generated by the energy supply production equipment that generates cooling energy in the i-th period h, C c,j,HS-dis,h is the cooling power of the j-th cold storage device in time period h, C cL,h is the cooling load in period h, C c,j,HS-ch,h is the cooling power of the j-th cold storage device in time period h, Ω3 is the set of energy supply production equipment consuming natural gas, and Ω 12 G is the gas storage equipment collection. g,h The natural gas purchase power in period h, G g,i,GS-dis,h is the gas discharge power of the j-th gas storage device in period h, G g,j,h The natural gas power of the j-th energy supply production equipment consuming natural gas in period h, G g,i,GS-ch,h The filling power of the j-th gas storage device in period h, Ω 14 Energy supply production equipment collection for consuming hydrogen, Ω 13 For hydrogen storage equipment collection, is the amount of hydrogen purchased during period h, is the hydrogen discharge power of the jth hydrogen storage device in period h, The hydrogen power of the energy supply production equipment for the jth hydrogen-consuming energy in period h, The hydrogen charging power of the jth hydrogen storage device in period h.

4. The method according to claim 3, characterized in that The power exchange constraint between the energy system and the upper network is determined by the following formula: 0≤P le,i,h ≤P legm,i,h Where, are the minimum and maximum power purchases from the public grid thermal power, Purchase gas power for minimum and maximum, respectively, Purchase hydrogen power for minimum and maximum respectively, are the minimum and maximum long-term green electricity purchase power, P le,i,h is the purchase amount of green electricity in the i-th medium- and long-term transaction in period h, P legm,i,h is the amount of green electricity provided in the i-th medium- and long-term transaction in period h.

5. The method according to any one of claims 1 to 4, characterized in that The method is solved by the NSGA-II algorithm.

6. A computing network multi-objective collaborative optimization operation device that considers the timing scheduling of computing tasks, characterized by: The method according to any one of claims 1 to 5, comprising: A model building module is used to establish a multi-objective collaborative optimization model for computing network energy; wherein the multi-objective collaborative optimization model for computing network energy includes an upper model and a lower model; A first data processing module is used for the upper-layer model to determine a computing power task allocation plan with maximum benefit and minimum carbon emission as objective functions, latency constraints, computing power scale constraints, and energy supply constraints as constraints, and task timing allocation as decision variables, and to send the computing power task allocation plan to the lower-layer model; The second data processing module is used for the lower-level model to determine the energy supply plan, carbon emissions and operating income with the maximum benefit and the minimum carbon emissions as the objective function, the power balance constraint, the power exchange constraint between the energy system and the upper network, the surplus power grid constraint, and the green power ratio constraint as the constraint conditions, and the equipment operation strategy as the decision variable, and send the energy supply plan, the carbon emissions and the operating income to the upper-level model; The third data processing module is used to repeatedly iterate the upper-level model and the lower-level model to determine the optimal solution; wherein the optimal solution includes computing power task timing optimization and energy supply curve.

7. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 5.

Images