CN118228878B

CN118228878B - Double-layer collaborative optimization method, system, equipment and medium applied to data center co-production system

Info

Publication number: CN118228878B
Application number: CN202410422848.7A
Authority: CN
Inventors: 张晓烽; 傅昂; 孙小琴; 李�杰; 王蒙
Original assignee: Changsha University of Science and Technology
Current assignee: Changsha University of Science and Technology
Priority date: 2024-04-09
Filing date: 2024-04-09
Publication date: 2024-09-06
Anticipated expiration: 2044-04-09
Also published as: CN118228878A

Abstract

The invention relates to a double-layer collaborative optimization method, a system, equipment and a medium applied to a data center co-production system, wherein the method comprises the following steps: optimizing upper layer configuration parameters comprising the capacity of the internal combustion engine, the area of the photovoltaic panel, the capacity of the heat storage tank and the area of the heat collector by using a non-dominant ranking genetic algorithm; optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge of an electrolytic tank and the proportion of power supply of a hydrogen fuel cell to the total power supply of an internal combustion engine by utilizing a multi-target wolf optimization algorithm; optimizing the optimized upper configuration parameters and lower proportion parameters through a decision algorithm, and solving the optimal capacity configuration and optimal operation parameters. In consideration of the interaction between configuration optimization and actual operation parameters, the method introduces the concept of collaborative optimization, and ensures the flexibility of the co-production system and the reasonable configuration of resources.

Description

Double-layer collaborative optimization method, system, equipment and medium applied to data center co-production system

Technical Field

The invention relates to the technical field of energy supply, in particular to a double-layer collaborative optimization method, a double-layer collaborative optimization system, double-layer collaborative optimization equipment and double-layer collaborative optimization media applied to a data center co-production system.

Background

Green data centers that reduce energy consumption by employing renewable energy and energy saving technologies have become an important research area worldwide, and in the existing data center energy consumption composition, IT equipment and cooling systems consume about 86% of the total power consumption of the data center.

On one hand, IT is important to grasp accurate data center energy consumption data to perform accurate green planning and deployment, the existing data center energy consumption simulation mode mainly uses methods such as a data center server power consumption model, a constant heat load density simulation data center IT equipment load of unit area, actual monitoring data and the like, and because of unique internal load characteristics of the data center, input parameters required in the simulation process are difficult to collect, so that the existing modeling method is difficult or incapable of effectively optimizing the data center design, and changes of the IT equipment operation state are considered in the energy system design to become important points and difficulties of the data center energy consumption simulation. On the other hand, for the specific energy consumption composition of the data center, a reasonable energy supply mode is used as a focus of the data center for reducing energy consumption and carbon emission, and the combined heat and power generation system for coupling renewable energy sources is widely applied to energy management of the data center.

In the existing research, most of the energy storage modes of the data center are electrochemical energy storage, and the data center has the characteristics of high energy consumption and stable requirement, so that the searching of alternative low-carbon supplementary energy sources is very important; while most research has focused on comparing natural cooling with mechanical cooling, no further discussion of other natural cooling schemes has been made.

Therefore, the configuration and operation scheme of the data center co-production system are at a distance from high availability and practicality, and further improvement and optimization are required to achieve better service for the data center.

Disclosure of Invention

First, the technical problem to be solved

In view of the above-mentioned shortcomings and drawbacks of the prior art, the present invention provides a dual-layer collaborative optimization method, system, device and medium applied to a data center co-production system, which solves the technical problems that the system configuration and operation scheme of the data center co-production system do not have high availability and practicality.

(II) technical scheme

In order to achieve the above purpose, the main technical scheme adopted by the invention comprises the following steps:

In a first aspect, an embodiment of the present invention provides a dual-layer collaborative optimization method applied to a data center co-production system, including: optimizing upper configuration parameters including the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector by using a non-dominant ranking genetic algorithm and taking the energy utilization rate, the economic cost and the environmental benefit as optimization targets; optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of electric consumption of an electrolytic tank to the total electric consumption of an electrolytic tank and the total electric charge of a storage battery and the proportion of electric power supply of a hydrogen fuel cell to the total electric power supply of a hydrogen fuel cell and an internal combustion engine by using a multi-target gray wolf optimization algorithm with energy utilization rate and economic cost as optimization targets; optimizing the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solving the optimal capacity configuration and the optimal operation parameters.

Optionally, the data center co-production system includes: the heat supply assembly comprises a heat collector for collecting solar heat and a heat storage tank connected with the heat collector; a hydrogen fuel assembly comprising an electrolyzer for receiving power from a photovoltaic panel and/or a power grid and a hydrogen storage tank connected to the electrolyzer; the electric load assembly comprises a photovoltaic panel, an energy storage battery for receiving the photovoltaic panel and/or supplying power, a hydrogen fuel cell for receiving the air supply of the hydrogen storage tank and an internal combustion engine, wherein the photovoltaic panel, a power grid, the energy storage battery, the hydrogen fuel cell and the internal combustion engine are all used as electric load sources of the data center; the cold load assembly comprises an electric refrigerator for receiving power supplied by a photovoltaic panel and/or a power grid, a heat exchanger for receiving a natural cold source and an absorption refrigerator for receiving heat supplied by at least one of a hydrogen fuel cell, a heat storage tank and an internal combustion engine, wherein the electric refrigerator, the heat exchanger and the absorption refrigerator are used as cold load sources of a data center.

Optionally, optimizing upper layer configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity, and collector area using a non-dominant ranking genetic algorithm with energy utilization, economic cost, and environmental benefit as optimization objectives includes: taking upper layer configuration parameters comprising the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector as optimization variables; initializing a population, performing non-dominant ranking on solutions in the population, calculating the crowding degree of each solution, selecting a certain number of individuals as parents according to the non-dominant ranking and the crowding degree, performing crossover and mutation based on the parents to update the population, performing non-dominant ranking and crowding degree calculation again on the updated population, selecting a certain number of individuals as the next-generation population according to the non-dominant ranking and crowding degree calculation result again, and performing termination condition judgment operation, and searching to obtain an optimal solution set comprising a group of solutions with optimal energy utilization rate, economic cost and environmental benefit; and (3) inputting an optimal solution set comprising a group of solutions with optimal energy utilization rate, economic cost and environmental benefit as an optimal upper-layer configuration parameter into the TOPSIS decision algorithm to participate in optimizing.

Optionally, optimizing the lower layer proportion parameters including the proportion of the absorption refrigeration capacity to the total refrigeration capacity, the proportion of the natural cooling refrigeration capacity to the total refrigeration capacity, the proportion of the power consumption of the electrolytic tank to the total amount of power consumption of the electrolytic tank and the total amount of power charge of the storage battery, and the proportion of the power supply of the hydrogen fuel cell to the total power supply of the hydrogen fuel cell and the internal combustion engine by using a multi-target wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets comprises: taking lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge amount of an electrolytic tank and the proportion of power supply quantity of a hydrogen fuel cell to the total power supply quantity of a hydrogen fuel cell and an internal combustion engine as optimization variables; initializing a wolf population, evaluating the fitness of each individual in the population, determining behavior parameters of the wolves, selecting leading wolves according to the fitness values, carrying out a wolf search process based on the behavior parameters of the wolves and the leading wolves, gradually optimizing the values of lower parameters and judging termination conditions by updating the positions and the fitness values of the wolf individuals, and obtaining an optimal solution set comprising a group of solutions with optimal energy utilization rate and economic cost; and (3) inputting an optimal solution set comprising a group of solutions with optimal energy utilization rate and economic cost as an optimal lower-layer proportion parameter into a TOPSIS decision algorithm to participate in optimizing.

Optionally, optimizing the optimized upper layer configuration parameter and lower layer proportion parameter by using a TOPSIS decision algorithm, and solving the optimal capacity configuration and the optimal operation parameter includes: based on a TOPSIS decision algorithm, performing Pareto front solution on the optimized upper configuration parameters and the optimized lower proportion parameters, and finding out upper capacity configuration and lower operation parameters according to the minimum normalized Euclidean distance; inputting the upper capacity configuration obtained by TOPSIS optimizing to a lower layer for optimizing lower layer proportion parameters based on a multi-objective gray wolf optimizing algorithm, and simultaneously inputting lower layer operation parameters to an upper layer for optimizing upper layer configuration parameters based on a non-dominant sorting genetic algorithm for iterative loop, so as to obtain optimal capacity configuration and optimal operation parameters meeting system configuration and operation parameters;

Wherein,

Normalized S _norm,d for pareto front solution is expressed as:

Wherein d represents the d objective function, S _p,d is the pareto front solution, and S _best,d is the best pareto front solution;

Normalized Euclidean distance Dist _p, expressed as:

Wherein Dist _p is the p-th normalized Euclidean distance, p and l respectively represent the number of optimal solutions and the number of objective functions, and finally Is the ideal value for the d-th objective function.

Optionally, optimizing upper configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity and collector area with energy utilization, economic cost and environmental benefit as optimization targets by using a non-dominant ranking genetic algorithm; or optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total power consumption of an electrolytic tank and the total power consumption of a storage battery and the proportion of power supply of a hydrogen fuel cell to the total power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target gray wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets, and constructing the following model:

the generated energy of the photovoltaic panel is as follows:

E_mn,pv＝A_pvI_bη_pv×10^-3；

Wherein E _mn,pv is the generated energy of the photovoltaic panel, A _pv is the receiver area covered by the photovoltaic panel, I _b is the direct solar radiation intensity, and eta _pv is the electrical efficiency of the photovoltaic module;

the electricity generation amount of the hydrogen fuel cell is as follows:

E_mn,fc＝H_mn,fcη_fc,e；

Wherein E _mn,fc is the electricity generation amount of the hydrogen fuel cell, H _mn,fc is the hydrogen consumption amount of the hydrogen fuel cell, and eta _fc,e is the electrical efficiency of the hydrogen fuel cell;

the heat generation amount of the hydrogen fuel cell is:

Q_mn,fc＝H_mn,fcη_fc,h；

Wherein Q _mn,fc is the heat generation amount of the hydrogen fuel cell, and η _fc,h is the heat efficiency of the hydrogen fuel cell;

The power consumption of the electrolytic water of the electrolytic tank is as follows:

Wherein E _mn,el is the power consumption of the water for electrolysis in the electrolytic cell, H _mn,el is the hydrogen yield of the electrolytic cell, and eta _el is the conversion efficiency of the electrolytic cell;

the hydrogen storage amount of the hydrogen storage tank is as follows:

Wherein H _mn,hst is the hydrogen storage amount of the hydrogen storage tank, H _mn-1,hst is the hydrogen storage amount of the hydrogen storage tank at the previous moment, and eta _hst,c and eta _hst,d are the hydrogen storage efficiency and the hydrogen release efficiency respectively;

The model of the battery is expressed as:

Wherein E _mn,es is the electric energy stored in the storage battery, E _mn-1,es is the electric energy stored in the storage battery at the previous moment, E _mn,esc is the electric energy stored in the storage battery at the current moment, E _mn,esd is the discharge amount of the storage battery, eta _es,c is the electric energy storage efficiency, and eta _es,d is the discharge efficiency;

The power generation amount of the internal combustion engine is:

Wherein E _mn,ice represents the power generation amount of the internal combustion engine, E _r,ice represents the rated power of the system, and lambda represents a key parameter for determining whether the internal combustion engine is started; e _mn,dc represents the electrical demand of the data center, E _mn,a represents the additional power of the system, f _mn,ice represents the part load rate of the internal combustion engine, expressed as:

The fuel consumption of the internal combustion engine is:

Where F _mn,ice is the fuel consumption of the internal combustion engine, η _mn,ice is the electrical efficiency of the internal combustion engine, expressed as:

The heat generation amount of the internal combustion engine is:

Q_mn,ice＝F_mn,ice(1-η_mn,ice)；

Wherein Q _mn,ice is the heat generation amount of the internal combustion engine;

the power consumption of the power grid is as follows:

E_mn,grid＝E_mn,dc+E_mn,a+E_mn,el-E_mn,ice-E_mn,pv-E_mn,fc；

wherein E _mn,grid is the power consumption of the power grid;

The fuel consumption of the power grid is:

wherein F _mn,grid is the fuel consumption of the power grid, and eta _grid and eta _e respectively represent the transmission efficiency of the power grid and the electric efficiency of the power grid;

The thermal energy storage model of the heat storage tank is as follows:

Wherein Q _mn,tst represents the amount of heat contained in the heat storage tank, Q _mn-1,tst represents the amount of heat contained in the heat storage tank at the previous time, Q _mn,tst.out represents the amount of heat released by the amount of heat stored in the heat storage tank, Q _mn,tst.in represents the amount of heat stored in the heat storage tank, η _hs,in represents the heat storage efficiency, and η _hs,out represents the heat release efficiency;

The mathematical model of the absorption refrigerator is as follows:

Wherein Q _mn,ac is the heat absorbed by the absorption refrigerator, C _mn,ac is the refrigerating capacity of the absorption refrigerator, and COP _mn,ac is the refrigerating performance coefficient of the absorption refrigerator;

the heat collection amount of the heat collector is as follows:

Q_mn,st(t)＝A_stI_mn,bη_st×10^-3；

Wherein Q _mn,st (t) is the heat collection amount of the heat collector, A _st is the area of the heat collector, I _mn,b is the direct solar radiation intensity, and eta _st is the heat efficiency of the heat collector;

The cooling efficiency of natural cooling is as follows:

In the formula, xi is the cooling efficiency of natural cooling, t _r,in is the return air temperature, t _s,out is the supply air temperature, and t _amb,in is the temperature of ambient air, namely the inlet air temperature;

the electric balance of the data center co-production system is as follows:

E_mn,ice+E_mn,grid+E_mn,pv+E_mn,fc＝E_mn,dc+E_mn,a+E_mn,el；

The heat balance of the data center co-production system is as follows:

wherein COP _ac represents the energy efficiency ratio of the absorption refrigerator, and C _ac represents the cooling capacity of the absorption refrigerator;

The total amount of fuel consumption of the data center co-production system is:

wherein F _pg is the total amount of fuel consumption of the data center co-production system;

the total power consumption of the reference system set by the data center co-production system is as follows:

Wherein E _sp,grid is the electric quantity from a power grid, E _dc is the electric load of a data center, C _dc is the cold load of a data center machine room, and COP _echill represents the energy efficiency ratio of an electric refrigerator;

the fuel consumption from the grid is:

Wherein F _sp,grid is the fuel consumption of the power grid;

The primary energy consumption of the reference system is as follows:

Wherein F _sp is the primary energy consumption of the reference system.

Optionally, the data center cogeneration system is configured to execute a refrigeration demand regulation strategy based on power supply drive suitable for temperatures no less than 12 ℃, or a power optimization configuration strategy based on refrigeration demand direction suitable for temperatures less than 12 ℃;

the refrigeration demand regulation strategy based on electric power supply driving comprises:

Acquiring the total power consumption requirement E _need of the co-production system of the data center, and judging whether the generated energy E _mn,pv of the photovoltaic panel is smaller than the total power consumption requirement E _need;

If the generated energy of the photovoltaic panel is not less than the total electricity consumption requirement E _need, storing the electric quantity E _mn,el in a hydrogen storage mode or storing the electric quantity E _mn,esc in a direct storage battery electricity storage mode by producing hydrogen through water electrolysis by using the redundant electric quantity except the total electricity consumption requirement of the generated energy of the photovoltaic panel;

If the generated energy of the photovoltaic panel is smaller than the total power consumption requirement E _need, judging whether the total amount of partial generated energy after the generated energy E _mn,fc of the hydrogen fuel cell and/or the generated energy E _mn,ice of the internal combustion engine is added is not smaller than the total power consumption requirement;

When the total amount of partial power generation is smaller than the total power consumption requirement E _need, further adding power E _mn,grid of the power grid for supplying;

Acquiring a total cooling requirement C _need of the data center co-production system, and obtaining required heat Q _ac,need for driving an absorption refrigerator and cooling capacity Q _free provided by natural cooling according to the total cooling requirement C _need;

collecting the heat generation quantity Q _mn,fc of the hydrogen fuel cell and/or the heat generation quantity Q _mn,ice of the internal combustion engine and using the heat generation quantity Q _mn,ice to drive the absorption refrigerator, and judging whether the heat generation quantity is smaller than the required heat for driving the absorption refrigerator;

If the collected generated heat is not less than the required heat for driving the absorption chiller, storing the remaining heat Q _hst,in except for the heat for driving the absorption chiller into the heat storage tank;

If the collected generated heat is less than the required heat for driving the absorption refrigerator, invoking corresponding heat Q _hst,out from the heat storage tank;

Wherein,

If the generated energy of the photovoltaic panel is not less than the total electricity demand E _need, the electricity E _mn,el is stored in a hydrogen storage mode by water electrolysis to produce hydrogen:

E_mn,el＝K₁(E_need-E_mn,pv)；

If the generated energy of the photovoltaic panel is not less than the total electricity consumption requirement E _need, the electricity quantity E _mn,esc is stored in the form of direct storage battery electricity storage:

E_mn,esc＝(1-K₁)(E_need-E_mn,pv)；

Wherein K ₁ is a proportionality coefficient stored in a hydrogen storage form by hydrogen production by water electrolysis;

The required heat Q _ac,need for driving the absorption chiller is:

wherein Q _ac,need is the required heat for driving the absorption refrigerator, K ₂ is the proportion of absorption refrigeration, Q _free is the cold energy provided by natural cooling, eta _free is the natural cooling efficiency, and C _need is the total cold requirement of the data center co-production system;

And, the power optimization configuration strategy based on refrigeration demand direction comprises:

Acquiring a total cold requirement C _need of the data center co-production system, and obtaining required heat Q _ac,need for driving an absorption refrigerator and refrigerating capacity C _chill of an electric refrigerator according to the total cold requirement;

Judging whether the heat collection quantity Q _mn,st of the heat collector is larger than the required heat quantity Q _ac,need for driving the absorption refrigerator;

if the heat collection quantity Q _mn,st of the heat collector is not smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator, the redundant heat is stored in a heat storage tank, and the heat storage quantity of the heat storage tank is recorded as Q _mn,tst,in;

If the heat collection quantity Q _mn,st of the heat collector is smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator, judging whether the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank are smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator;

If the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank are smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator, the difference value between the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank and the required heat quantity is recorded as Q _need1;

Providing the heat of the difference Q _need1 by delivering the heat of production Q _mn,fc of the hydrogen fuel cell and/or the heat of production Q _mn,ice of the internal combustion engine according to the engine start state;

Acquiring the total electricity demand E _need of the data center cogeneration system, and judging whether the generated energy E _mn,pv of the photovoltaic panel is smaller than the total electricity demand E _need;

If the generated energy E _mn,pv of the photovoltaic panel is not smaller than the total power consumption requirement E _need, calculating the residual power of the power supply of the photovoltaic panel;

If the generated energy E _mn,pv of the photovoltaic panel is smaller than the total electricity consumption requirement E _need, acquiring the generated energy of a hydrogen fuel cell and/or an internal combustion engine which participate in heat supply;

judging whether the total amount of partial power generation including the power generation amount of the hydrogen fuel cell and/or the internal combustion engine participating in heat supply and the power generation amount of the photovoltaic panel is larger than the total power consumption requirement E _need;

When the total amount of partial power generation of the power generation amount of the hydrogen fuel cell and/or the internal combustion engine which participate in heat supply and the power generation amount of the photovoltaic panel is smaller than the total power consumption requirement E _need, the electric power E _mn,grid which is further added into the power grid is supplied;

Wherein,

The required heat Q _ac,need for driving the absorption chiller is:

Q_ac,need＝C_ac,need/1.2

C_ac,need＝K₃C_need；

the refrigerating capacity of the electric refrigerator is as follows:

C_chill＝(1-K₃)C_need

Wherein, C _ac,need represents the refrigerating capacity of the absorption refrigerator under the condition of meeting the cold load demand preferentially, K ₃ represents the proportion of the absorption refrigerator, and C _chill represents the refrigerating capacity of the electric refrigerator;

In providing the heat of the difference Q _need1, the generated heat Q _mn,fc of the hydrogen fuel cell is:

Q_mn,fc＝K₄Q_need1；

In providing the heat of the difference Q _need1, the heat generation amount Q _mn,ice of the internal combustion engine is:

Q_mn,ice＝(1-K₄)Q_need1；

where K ₄ represents the proportion of the heat supplied by the hydrogen fuel cell.

In a second aspect, an embodiment of the present invention provides a dual-layer optimization system for a co-production system of a data center, including: an upper layer optimization model configured to optimize upper layer configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity, and collector area using a non-dominant ranking genetic algorithm with energy utilization, economic cost, and environmental benefit as optimization targets; the lower layer optimization module is configured to optimize lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge of a storage battery and the proportion of power supply of a hydrogen fuel cell to the total power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets; the interaction optimization module is configured to optimize the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solve the optimal capacity configuration and the optimal operation parameters.

In a third aspect, an embodiment of the present invention provides an apparatus, including: at least one database; and a memory communicatively coupled to the at least one database; wherein the memory stores instructions executable by the at least one database to enable the at least one database to perform a dual-tier co-optimization method of a data center co-production system as described above.

In a fourth aspect, embodiments of the present invention provide a computer readable medium having stored thereon computer executable instructions that when executed by a processor implement a dual-tier co-optimization method for a data center co-production system as described above.

(III) beneficial effects

The beneficial effects of the invention are as follows: considering the interaction between the system configuration optimization and the actual operation parameters, the invention introduces the concept of a double-layer collaborative optimization model, and the upper layer adopts a non-dominant ordering genetic algorithm (NSGAII) to optimize the configuration capacity; the lower layer adopts a multi-objective gray wolf optimization algorithm (MOGWO) to optimize the proportion parameters in the operation process, meanwhile, the upper layer and the lower layer output and then adopt a TOPSIS decision algorithm to perform optimization, and the interaction connection between the two layers is formed through the mutual transmission between the decided optimal solutions. Through the operation, the flexibility of the data center co-production system and the reasonable configuration of resources are ensured.

Drawings

FIG. 1 is a schematic flow chart of a double-layer collaborative optimization method of a data center co-production system provided by an embodiment of the invention;

FIG. 2 is a schematic logic diagram of a method for double-layer collaborative optimization of a data center co-production system according to an embodiment of the present invention;

Fig. 3 is a schematic diagram of a data center co-production system according to a dual-layer collaborative optimization method of the data center co-production system according to an embodiment of the present invention;

fig. 4 is a schematic diagram of a specific flow including a genetic algorithm and a wolf algorithm of a double-layer collaborative optimization method of a data center co-production system according to an embodiment of the present invention

Fig. 5 is a schematic flow chart of a refrigeration demand regulation strategy based on power supply driving of a double-layer collaborative optimization method of a data center co-production system according to an embodiment of the present invention;

fig. 6 is a schematic flow chart of a power optimization configuration strategy based on refrigeration requirement guidance of a double-layer collaborative optimization method of a data center co-production system according to an embodiment of the present invention.

Detailed Description

The invention will be better explained for understanding by referring to the following detailed description of the embodiments in conjunction with the accompanying drawings.

As shown in fig. 1 and fig. 2, an embodiment of the present invention provides a dual-layer collaborative optimization method applied to a co-production system of a data center, including: optimizing upper configuration parameters including the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector by using a non-dominant ranking genetic algorithm and taking the energy utilization rate, the economic cost and the environmental benefit as optimization targets; optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of electric consumption of an electrolytic tank to the total electric consumption of an electrolytic tank and the total electric charge of a storage battery and the proportion of electric power supply of a hydrogen fuel cell to the total electric power supply of a hydrogen fuel cell and an internal combustion engine by using a multi-target gray wolf optimization algorithm with energy utilization rate and economic cost as optimization targets; optimizing the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solving the optimal capacity configuration and the optimal operation parameters.

Considering the interaction between the system configuration optimization and the actual operation parameters, the invention introduces the concept of a double-layer collaborative optimization model, and the upper layer adopts a non-dominant ordering genetic algorithm (NSGAII) to optimize the configuration capacity; the lower layer adopts a multi-objective gray wolf optimization algorithm (MOGWO) to optimize the proportion parameters in the operation process, meanwhile, the upper layer and the lower layer output and then adopt a TOPSIS decision algorithm to perform optimization, and the interaction connection between the two layers is formed through the mutual transmission between the decided optimal solutions. Through the operation, the flexibility of the data center co-production system and the reasonable configuration of resources are ensured.

Compared with a reference system, the system has the advantages that the primary energy saving rate of the system after double-layer collaborative optimization is 46.33%, the total annual cost saving rate is 38.25%, the carbon dioxide emission reduction rate is 49.08%, the grid independence is 56.07%, the total annual operation cost is 3.41×10 ⁵ $, the carbon emission intensity is 0.62, the system after optimization is remarkably improved in the aspects of energy utilization efficiency, economy and environmental protection, meanwhile, the carbon emission intensity of the system is 0.65 according to the carbon emission evaluation standard of the data center, the standard of the excellent data center is achieved, and powerful support is provided for sustainable development of the data center.

In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

Referring to fig. 3, the data center co-production system constructed in the present invention includes: the heat supply assembly comprises a heat collector for collecting solar heat and a heat storage tank connected with the heat collector; a hydrogen fuel assembly comprising an electrolyzer for receiving power from a photovoltaic panel and/or a power grid and a hydrogen storage tank connected to the electrolyzer; the electric load assembly comprises a photovoltaic panel, an energy storage battery for receiving the photovoltaic panel and/or supplying power, a hydrogen fuel cell for receiving the air supply of the hydrogen storage tank and an internal combustion engine, wherein the photovoltaic panel, a power grid, the energy storage battery, the hydrogen fuel cell and the internal combustion engine are all used as electric load sources of the data center; the cold load assembly comprises an electric refrigerator for receiving power supplied by a photovoltaic panel and/or a power grid, a heat exchanger for receiving a natural cold source and an absorption refrigerator for receiving heat supplied by at least one of a hydrogen fuel cell, a heat storage tank and an internal combustion engine, wherein the electric refrigerator, the heat exchanger and the absorption refrigerator are used as cold load sources of a data center.

In the data center co-production system, when the energy is sufficient, the solar photovoltaic power generation system is preferentially used for generating electricity, meanwhile, the gas internal combustion engine is used for generating electricity and the hydrogen fuel cell is used for supplementing, so that the total electricity requirement of the system is met, the redundant electricity is stored in the energy storage battery or in a hydrogen storage mode, and the insufficient electricity is provided by the power grid; under the condition that natural cold sources can be utilized, natural cooling is preferably adopted for cooling, waste heat generated by a gas internal combustion engine and a hydrogen fuel cell is provided for an absorption refrigerator to meet the cooling requirement, and the redundant heat is stored in a heat reservoir. Meanwhile, the system is provided with a reference system, the reference system provides the electric quantity required by the data center and the power consumption of the electric refrigerator through the public power grid, and the electric refrigerator meets the cooling capacity required by the data center.

The co-production system provided by the invention can completely meet the overall electric load and cold load demands of the data center, and meanwhile, under the condition of ensuring uninterrupted energy supply of the data center, the system cost is greatly reduced by utilizing the internal waste heat of the system to drive absorption refrigeration, and the system has high availability.

Specifically, the double-layer collaborative optimization method applied to the data center co-production system provided by the embodiment of the invention comprises the following steps:

s1, optimizing upper configuration parameters including the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector by using a non-dominant ranking genetic algorithm and taking the energy utilization rate, the economic cost and the environmental benefit as optimization targets.

Further, referring to fig. 4, step S1 includes:

and S11, establishing a mathematical model, determining an optimization target and an optimization variable, and taking upper configuration parameters comprising the capacity of the internal combustion engine, the area of the photovoltaic panel, the capacity of the heat storage tank and the area of the heat collector as the optimization variable.

S12, the introduced non-dominant sorting genetic algorithm specifically comprises the following steps: initializing a population, randomly generating a group of initial solutions as the population, and evaluating each solution; non-dominant ranking, determining a non-dominant rank for each solution; calculating the crowding degree of each solution for measuring the distribution density of the solutions in the target space; selecting a certain number of individuals as parents according to non-dominant ranking and crowding degree; performing cross operation on the selected parent individuals to generate new offspring individuals; carrying out mutation operation on the offspring individuals, and introducing new variant individuals; combining parent individuals and offspring individuals, and updating the population; non-dominant sorting and crowding degree calculation are carried out on the updated population; selecting a certain number of individuals as a next generation population according to the non-dominant ranking and the crowding degree; judging termination conditions, namely judging whether the termination conditions are met, if the maximum iteration times are reached or the solution is converged to a stable solution; and outputting an optimization result, and outputting an optimal solution set, namely a group of solutions with optimal energy utilization rate, economic cost and environmental benefit.

S13, inputting an optimal solution set comprising a group of solutions with optimal energy utilization rate and economic cost as an optimized lower-layer proportion parameter into a TOPSIS decision algorithm to participate in optimizing.

Furthermore, the devices mainly considered by the upper layer comprise the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector, the four configuration parameters are used as optimization variables of the upper layer, and the energy, economic and environmental benefits of the system are comprehensively considered, wherein the three optimization targets are as follows:

The maximized primary energy saving ratio PECR is:

wherein F _pg and F _system are primary energy consumption of the reference system and the integrated energy system, respectively.

The maximum annual total cost savings ACER is:

Where AC _pg and AC _system are the total annual costs of the reference system and the integrated energy system, respectively.

The maximized carbon dioxide emission reduction rate CDESR is:

CDE＝V_gasK_gas+E_gridK_grid；

wherein CDE _pg is the carbon dioxide emission amount of a reference system, kg; CDE _system is the carbon dioxide emission of the integrated energy system, kg, CDE is the carbon dioxide emission, V _gas is the amount of natural gas used, kg; k _gas、K_grid is the carbon dioxide conversion coefficient of natural gas and the carbon dioxide conversion coefficient of commercial power, and the values are 1.97kg/m ³ and 0.968kg/kWh respectively.

S2, optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total amount of power consumption of an electrolytic tank and the total amount of battery charge and the proportion of power supply of a hydrogen fuel cell to the total amount of power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target gray wolf optimization algorithm with energy utilization rate and economic cost as optimization targets.

Further, referring to fig. 4, step S2 includes:

S21, establishing a mathematical model, determining an optimization target and an optimization variable, and taking lower layer proportion parameters comprising the proportion of absorption refrigerating capacity to total refrigerating capacity, the proportion of natural cooling refrigerating capacity to total refrigerating capacity, the proportion of power consumption of an electrolytic tank to the total charge of an electrolytic tank and the total charge of a storage battery and the proportion of power supply of a hydrogen fuel cell to the total power supply of the hydrogen fuel cell and an internal combustion engine as the optimization variables.

S22, specifically comprising the following steps of: initializing a wolf population, randomly generating a group of initial wolf individuals, evaluating each individual, and calculating the energy utilization rate and economic cost of each individual. Marking three individuals with the best fitness as alpha, beta and sigma, wherein the rest wolves are omega, MOGWO, and the optimization process is mainly guided to be completed by three best solutions in each generation of population; designing a fitness function, wherein the fitness function is determined according to the weight of the absorption refrigeration capacity, the natural cooling refrigeration capacity, the power consumption of the electrolytic tank and the power supply capacity of the hydrogen fuel cell and the association relation of the weight and the fitness function, comprehensively considering the energy utilization rate, the economic cost and the optimization target of the carbon utilization rate, and evaluating the fitness of each individual; determining behavior parameters of the wolves, including parameters such as jump step length, search range and the like, and guiding the search behavior of the wolves; selecting a certain number of wolf individuals as leading wolves according to the value of the fitness function, and guiding the searching process; carrying out a wolf searching process, and gradually optimizing the value of the lower-layer parameter by updating the position and the fitness value of the wolf individual; judging termination conditions, such as reaching the maximum iteration times or converging to a stable solution, and determining whether to terminate the search process; and outputting an optimization result, and outputting an optimal solution set, namely a solution set with optimal energy utilization rate and economic cost, wherein the solution set comprises values of the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge of an electrolytic tank and a storage battery, and the proportion of power supply of a hydrogen fuel cell to the total power supply of the hydrogen fuel cell and an internal combustion engine.

S23, inputting an optimal solution set comprising a group of solutions with optimal energy utilization rate and economic cost as an optimized lower-layer proportion parameter into a TOPSIS decision algorithm to participate in optimizing.

Furthermore, the lower layer mainly considers the influence of the output proportion of the operation equipment on the system in the operation process, takes four operation parameters of the proportion of the absorption refrigeration capacity to the total refrigeration capacity, the proportion of the natural cooling refrigeration capacity to the total refrigeration capacity, the proportion of the power consumption of the electrolytic tank to the total charge of the electrolytic tank and the total power consumption of the storage battery and the proportion of the power supply of the fuel cell to the total power supply of the fuel cell and the internal combustion engine as the optimization variables of the upper layer, and comprehensively considers the energy and economic benefits of the system, wherein the model of the optimization targets is as follows:

Carbon Usage (CUE) measures greenhouse gas emissions associated with a data center using a ratio of carbon emissions to data center equipment energy consumption, minimizing carbon usage as:

Wherein CUE is carbon dioxide emission amount generated by data center load per megawatt hour, kgCO ₂/kWh;CDE_system is total emission amount of CO ₂, kg; e _IT represents the amount of power consumed by the data center IT load.

Minimizing data center system operating costs, the annual operating cost of the co-production system, OC _pg, is described as:

Wherein C _gas is the price of natural gas, $/kW, and C _e is the price of electricity, $/kW.

The independence of the energy system is represented by a grid integration level GI, and the minimum grid integration level is as follows:

Where E _dc represents the total power consumption of the data center, kW.

And S3, optimizing the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solving the optimal capacity configuration and the optimal operation parameters.

Further, referring to fig. 4, step S3 includes:

S31, based on a TOPSIS decision algorithm, performing Pareto front solving on the optimized upper-layer configuration parameters and the optimized lower-layer proportion parameters, and finding out upper-layer capacity configuration and lower-layer operation parameters according to the minimum normalized Euclidean distance.

Wherein,

Normalized S _norm,d for pareto front solution is expressed as:

Normalized Euclidean distance Dist _p, expressed as:

S32, inputting the upper capacity configuration obtained by TOPSIS optimization into a lower layer for optimizing lower layer proportion parameters based on a multi-objective gray wolf optimization algorithm, and simultaneously inputting lower layer operation parameters into an upper layer for optimizing upper layer configuration parameters based on a non-dominant ordering genetic algorithm for iterative loop, so as to obtain the optimal capacity configuration and the optimal operation parameters meeting the system configuration and operation parameters.

Further, prior to step S1 or S2, the following mathematical model is constructed for the data center co-production system:

the generated energy of the photovoltaic panel is as follows:

E_mn,pv＝A_pvI_bη_pv×10^-3；

Wherein E _mn,pv is the power generation of the photovoltaic panel, A _pv is the receiver area covered by the photovoltaic panel, m ²,I_b is the direct solar radiation intensity, and W/m ²,η_pv is the electrical efficiency of the photovoltaic module.

The electricity generation amount of the hydrogen fuel cell is as follows:

E_mn,fc＝H_mn,fcη_fc,e；

Where E _mn,fc is the power generation amount of the hydrogen fuel cell, kW, H _mn,fc is the hydrogen consumption amount of the hydrogen fuel cell, kW, η _fc,e is the electrical efficiency of the hydrogen fuel cell.

The heat generation amount of the hydrogen fuel cell is:

Q_mn,fc＝H_mn,fcη_fc,h；

Where Q _mn,fc is the heat generation capacity of the hydrogen fuel cell, kW, η _fc,h is the thermal efficiency of the hydrogen fuel cell,%.

Wherein H _mn,el is hydrogen yield of the electrolytic tank, kW, E _mn,el is power consumption of water electrolyzed by the electrolytic tank, kW, eta _el is conversion efficiency of the electrolytic tank.

The hydrogen storage amount of the hydrogen storage tank is as follows:

Wherein H _mn,hst is the hydrogen storage amount of the hydrogen storage tank, kW, H _mn-1,hst is the hydrogen storage amount of the hydrogen storage tank at the previous moment, kW, eta _hst,c and eta _hst,d are the hydrogen storage efficiency and the hydrogen release efficiency respectively.

The storage battery realizes the mutual conversion of chemical energy and electric energy by utilizing reversible chemical reaction, and the model of the storage battery is expressed as:

Wherein E _mn,es is the electric quantity contained in the storage battery, KW, E _mn-1,es is the electric quantity contained in the storage battery at the previous moment, KW, E _mn,esc is the electric quantity of the storage battery at the current moment, KW, E _mn,esd is the discharge quantity of the storage battery, KW, eta _es,c is the electric storage efficiency,%, eta _es,d is the discharge efficiency,%.

The power generation amount of the internal combustion engine is:

Wherein E _mn,ice represents the power generation amount of the internal combustion engine, kW, E _r,ice represents the rated power amount of the system, kW and lambda represents key parameters for determining whether the internal combustion engine is started or not; e _mn,a represents additional power to the system, kW, E _mn,dc represents data center electrical load, kW.

The partial load factor of the internal combustion engine is:

where f _mn,ice denotes a part load ratio of the internal combustion engine.

The fuel consumption of the internal combustion engine is:

where F _mn,ice is the fuel consumption of the internal combustion engine, η _mn,ice is the electrical efficiency of the internal combustion engine;

The heat generation amount of the internal combustion engine is:

Q_mn,ice＝F_mn,ice(1-η_mn,ice)；

Wherein Q _mn,ice is the heat generation amount of the internal combustion engine.

The power consumption of the power grid is as follows:

E_mn,grid＝E_mn,dc+E_mn,a+E_mn,el-E_mn,ice-E_mn,pv-E_mn,fc；

wherein E _mn,grid is the power consumption of the power grid.

The fuel consumption F _mn,grid of the grid is:

Where F _mn,grid is the fuel consumption of the grid, η _grid and η _e represent the transmission efficiency of the grid and the electrical efficiency of the grid, respectively.

The thermal energy storage model of the heat storage tank is as follows:

Wherein Q _mn,tst represents the amount of heat contained in the heat storage tank, kW; q _mn-1,tst represents the heat contained in the heat storage tank at the previous time, kW; η _hs,in denotes heat storage efficiency, η _hs,out denotes heat release efficiency,%; q _mn,tst.out represents the heat release amount of the heat storage tank, kW; q _mn,tst.in represents the heat absorption capacity of the heat storage tank, kW.

The mathematical model of the absorption refrigerator is as follows:

Wherein Q _mn,ac is the amount of heat absorbed by the absorption refrigerator, kW, C _mn,ac is the cooling capacity of the absorption refrigerator, kW, and COP _mn,ac is the cooling performance coefficient of the absorption refrigerator.

According to the data provided in the national standard "vacuum tube solar collector" GB/T17581-2021, for a medium temperature vacuum tube solar collector, the working temperature is 100-150 ℃, the medium temperature efficiency of the collector is not lower than 0.45, and the heat collection quantity Q _mn,st of the collector is:

Q_mn,st＝A_stI_mn,bη_st×10^-3；

Wherein Q _mn,st is the heat collection amount of the heat collector, A _st is the heat collector area, m ²,I_mn,b is the direct solar radiation intensity, and W/m ²,η_st is the heat efficiency of the heat collector.

The system of the present invention does not directly enter the data center with outdoor low temperature air, but indirectly cools outside air through the heat exchanger by using natural air flow, thereby reducing the temperature of air from the data center. The cooled air passes through the data center again to form natural air flow circulation, thereby achieving the effect of natural cooling, and the cooling efficiency of natural cooling is as follows:

In the formula, xi is the cooling efficiency of natural cooling, t _r,in is the return air temperature, t _s,out is the supply air temperature, and t _amb,in is the ambient air temperature, namely the inlet air temperature;

the electric balance of the data center co-production system is as follows:

E_mn,ice+E_mn,grid+E_mn,pv+E_mn,fc＝E_mn,dc+E_mn,a+E_mn,el；

wherein E _mn,ice represents the power generation amount of the internal combustion engine, kW, E _mn,pv represents the power generation amount of the photovoltaic, kW; e _mn,dc represents data center electrical load, kW, E _mn,a represents additional power to the system, kW.

The heat balance of the data center co-production system is as follows:

Wherein Q _mn,ice represents the heat generation amount of the internal combustion engine, KW; q _mn,st represents the heat collection amount of the heat collector, KW; q _mn,fc represents the heat generation amount of the fuel cell, KW; COP _mn,ac represents the energy efficiency ratio of the absorption chiller,%; c _mn,ar,c denotes the cooling capacity of the absorption refrigerator, KW.

Where F _pg is the total amount of fuel consumption for the data center co-production system.

Next, the data center co-production system further includes a reference system, the parameter system mathematical model including:

The total power consumption of the reference system is:

Where E _sp,grid is the amount of electricity from the grid, kW, E _dc is the electrical load of the data center, kW, C _mn,dc is the cooling load of the data center room, kW, and COP _echill represents the energy efficiency ratio of the electric refrigerator.

The fuel consumption from the grid is:

where F _sp,grid is the fuel consumption of the grid, η _grid and η _e represent the transmission efficiency of the grid and the electrical efficiency of the grid, respectively.

The primary energy consumption F _sp of the reference system is as follows:

Wherein F _sp is the primary energy consumption of the reference system, kW.

In turn, the data center cogeneration system is configured to execute a refrigeration demand regulation strategy based on power supply drive suitable for temperatures no less than 12 ℃ or a power optimization configuration strategy based on refrigeration demand direction suitable for temperatures less than 12 ℃.

As shown in fig. 5 and 6, the present invention sets the following two optimization strategies based on whether natural cooling is adopted:

Referring to fig. 5, the refrigeration demand regulation strategy based on electric power supply driving is specifically:

And acquiring the total power consumption requirement E _need of the data center co-production system, and judging whether the generated energy E _mn,pv of the photovoltaic panel is smaller than the total power consumption requirement E _need.

If the generated energy of the photovoltaic panel is not less than the total electricity demand E _need, storing E _mn,el in a hydrogen storage form or storing E _mn,esc in a direct storage battery storage form by hydrogen production through water electrolysis of surplus electric energy except the total electricity demand of the generated energy of the photovoltaic panel, which can be expressed as:

E_mn,el＝K₁(E_need-E_mn,pv)；

E_mn,estc＝(1-K₁)(E_need-E_mn,pv)；

wherein K ₁ represents a proportionality coefficient stored in the form of hydrogen storage by hydrogen production by electrolysis of water;

If the generated energy of the photovoltaic panel is smaller than the total power consumption requirement E _need, judging whether the total amount of partial generated energy after the generated energy E _mn,fc of the hydrogen fuel cell and/or the generated energy E _mn,ice of the internal combustion engine is added is not smaller than the total power consumption requirement.

And when the total amount of partial power generation is smaller than the total power consumption requirement E _need, further adding the electric power E _mn,grid of the electric network to supply.

The total cooling demand C _need of the data center co-production system is obtained, and the required heat Q _ac,need for driving the absorption chiller and the cooling capacity Q _free provided by natural cooling are obtained according to the total cooling demand C _need, which is expressed as:

Where Q _ac,need represents the heat required to drive the absorption chiller, K ₂ represents the proportion of absorption refrigeration, Q _free represents the amount of cooling provided by natural cooling, η _free represents the efficiency of natural cooling, and C _need is the total cooling demand of the data center cogeneration system.

If the collected heat generation amount is smaller than the required heat amount for driving the absorption refrigerator, the corresponding heat amount Q _hst,out is called from the heat storage tank.

And, referring to fig. 6, the power optimization configuration strategy based on refrigeration demand guidance is:

The total cold demand C _need of the data center co-production system is obtained, and then the required heat Q _ac,need for driving the absorption refrigerator and the refrigerating capacity C _chill of the electric refrigerator are obtained, which are expressed as:

C_ac,need＝K₃C_need；

C_chill＝(1-K₃)C_need；

Q_ac,need＝C_ac,need/1.2；

where C _ac,need represents the cooling capacity of the absorption refrigerator in the case where the cooling load demand is preferentially satisfied, K ₃ represents the proportion of the absorption refrigeration, and C _chill represents the cooling capacity of the electric refrigerator.

It is determined whether the heat collection amount Q _mn,st of the heat collector is greater than the required amount Q _ac,need for driving the absorption refrigerator.

If the heat collection amount Q _mn,st of the heat collector is not less than the required amount Q _ac,need for driving the absorption refrigerator, the surplus heat is stored in the heat storage tank, and the heat storage amount of the heat storage tank is denoted as Q _mn,tst,in.

If the heat collection quantity Q _mn,st of the heat collector is smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator, judging whether the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank are smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator.

If the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank are smaller than the required heat quantity Q _ac,need for driving the absorption refrigerator, the difference value between the heat collection quantity Q _mn,st of the heat collector and the heat storage quantity Q _mn,tst,in of the heat storage tank and the required heat quantity is recorded as Q _need1.

Depending on the engine start state, the difference Q _need1 is provided by the hydrogen fuel cell heat generation amount Q _mn,fc and/or the engine heat generation amount Q _mn,ice, and the fuel cell and engine electric quantity are expressed as:

Q_mn,fc＝K₄Q_need1；

Q_mn,ice＝(1-K₄)Q_need1；

Where K ₄ represents the proportion of the fuel cell heat supply amount.

And acquiring the total electricity demand E _need of the data center cogeneration system, and judging whether the generated energy E _mn,pv of the photovoltaic panel is smaller than the total electricity demand E _need.

And if the generated energy E _mn,pv of the photovoltaic panel is not smaller than the total power consumption requirement E _need, calculating the residual power of the power supply of the photovoltaic panel.

And if the generated energy E _mn,pv of the photovoltaic panel is smaller than the total electricity consumption requirement E _need, acquiring the generated energy of the hydrogen fuel cell and/or the internal combustion engine which participate in heat supply.

And judging whether the total amount of partial power generation including the power generation amount of the hydrogen fuel cell and/or the internal combustion engine participating in heat supply and the power generation amount of the photovoltaic panel is larger than the total power consumption requirement E _need.

When the total amount of partial power generation of the power generation amount of the hydrogen fuel cell and/or the internal combustion engine participating in heat supply and the power generation amount of the photovoltaic panel is smaller than the total power consumption requirement E _need, the electric power E _mn,grid further added into the electric network is supplied.

In a specific embodiment, the basic parameters of the double-layer collaborative optimization method of the co-production system of the data center provided by the invention are set as follows:

table 1 characteristic parameters of co-production system and reference system

Table 2 economic parameters of the system

TABLE 3 Capacity optimization results for Co-production System Equipment

Table 4 annual performance index of co-production system after optimization

Then, the embodiment of the invention also provides a double-layer optimization system of the data center co-production system, which is applied to the data center co-production system to execute the double-layer collaborative optimization method of the data center co-production system, and the double-layer optimization system comprises the following steps: and the upper layer optimization model is configured to optimize upper layer configuration parameters comprising the capacity of the internal combustion engine, the area of the photovoltaic panel, the capacity of the heat storage tank and the area of the heat collector by using a non-dominant ranking genetic algorithm with the energy utilization rate, the economic cost and the environmental benefit as optimization targets. The lower layer optimization module is configured to optimize lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of electricity consumption of an electrolytic tank to the total charge of a storage battery and the proportion of electricity consumption of a hydrogen fuel cell to the total charge of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets. The interaction optimization module is configured to optimize the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solve the optimal capacity configuration and the optimal operation parameters.

And, an embodiment of the present invention provides an apparatus, including: at least one database; and a memory communicatively coupled to the at least one database; wherein the memory stores instructions executable by the at least one database to enable the at least one database to perform a dual-tier co-optimization method of a data center co-production system as described above.

Furthermore, embodiments of the present invention provide a computer readable medium having stored thereon computer executable instructions that when executed by a processor implement a two-tier co-optimization method for a data center co-production system as described above.

It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art upon learning the basic inventive concepts. Therefore, the appended claims should be construed to include preferred embodiments and all such variations and modifications as fall within the scope of the invention.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, the present invention should also include such modifications and variations provided that they come within the scope of the following claims and their equivalents.

Claims

1. The double-layer collaborative optimization method applied to the data center co-production system is characterized by comprising the following steps of:

optimizing upper configuration parameters including the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector by using a non-dominant ranking genetic algorithm and taking the energy utilization rate, the economic cost and the environmental benefit as optimization targets;

Optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of electric consumption of an electrolytic tank to the total electric consumption of an electrolytic tank and the total electric charge of a storage battery and the proportion of electric power supply of a hydrogen fuel cell to the total electric power supply of a hydrogen fuel cell and an internal combustion engine by using a multi-target gray wolf optimization algorithm with energy utilization rate and economic cost as optimization targets;

Optimizing the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solving the optimal capacity configuration and optimal operation parameters;

Wherein,

Optimizing upper layer configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity, and collector area using a non-dominant ranking genetic algorithm with energy utilization, economic cost, and environmental benefit as optimization targets includes:

Taking upper layer configuration parameters comprising the capacity of an internal combustion engine, the area of a photovoltaic panel, the capacity of a heat storage tank and the area of a heat collector as optimization variables;

Initializing a population, performing non-dominant ranking on solutions in the population, calculating the crowding degree of each solution, selecting a certain number of individuals as parents according to the non-dominant ranking and the crowding degree, performing crossover and mutation based on the parents to update the population, performing non-dominant ranking and crowding degree calculation again on the updated population, selecting a certain number of individuals as the next-generation population according to the non-dominant ranking and crowding degree calculation result again, and performing termination condition judgment operation, and searching to obtain an optimal solution set comprising a group of solutions with optimal energy utilization rate, economic cost and environmental benefit;

Inputting an optimal solution set comprising a group of solutions with optimal energy utilization rate, economic cost and environmental benefit as an optimal upper-layer configuration parameter to a TOPSIS decision algorithm to participate in optimizing;

optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of electric consumption of an electrolytic tank to the total electric consumption of an electrolytic tank and the total electric charge of a storage battery and the proportion of electric power supply of a hydrogen fuel cell to the total electric power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target wolf optimization algorithm with energy utilization rate and economic cost as optimization targets comprises:

Taking lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge amount of an electrolytic tank and the proportion of power supply quantity of a hydrogen fuel cell to the total power supply quantity of a hydrogen fuel cell and an internal combustion engine as optimization variables;

Initializing a wolf population, evaluating the fitness of each individual in the population, determining behavior parameters of the wolves, selecting leading wolves according to the fitness values, carrying out a wolf search process based on the behavior parameters of the wolves and the leading wolves, gradually optimizing the values of lower parameters and judging termination conditions by updating the positions and the fitness values of the wolf individuals, and obtaining an optimal solution set comprising a group of solutions with optimal energy utilization rate and economic cost;

inputting an optimal solution set containing a group of solutions with optimal energy utilization rate and economic cost as an optimized lower-layer proportion parameter into a TOPSIS decision algorithm to participate in optimizing;

optimizing the optimized upper layer configuration parameters and lower layer proportion parameters through a TOPSIS decision algorithm, and solving the optimal capacity configuration and optimal operation parameters comprises the following steps:

Based on a TOPSIS decision algorithm, performing Pareto front solution on the optimized upper configuration parameters and the optimized lower proportion parameters, and finding out upper capacity configuration and lower operation parameters according to the minimum normalized Euclidean distance;

And inputting the upper capacity configuration obtained by TOPSIS optimization to a lower layer for optimizing lower proportion parameters based on a multi-objective gray wolf optimization algorithm, and simultaneously inputting lower operation parameters to an upper layer for optimizing upper configuration parameters based on a non-dominant ordering genetic algorithm for iterative loop, thereby obtaining the optimal capacity configuration and the optimal operation parameters which meet the system configuration and the operation parameters.

2. The method of double-layer co-optimization of a data center co-production system of claim 1, wherein the data center co-production system comprises:

The heat supply assembly comprises a heat collector for collecting solar heat and a heat storage tank connected with the heat collector;

a hydrogen fuel assembly comprising an electrolyzer for receiving power from a photovoltaic panel and/or a power grid and a hydrogen storage tank connected to the electrolyzer;

The electric load assembly comprises a photovoltaic panel, an energy storage battery for receiving the photovoltaic panel and/or supplying power, a hydrogen fuel cell for receiving the air supply of the hydrogen storage tank and an internal combustion engine, wherein the photovoltaic panel, a power grid, the energy storage battery, the hydrogen fuel cell and the internal combustion engine are all used as electric load sources of the data center;

The cold load assembly comprises an electric refrigerator for receiving power supplied by a photovoltaic panel and/or a power grid, a heat exchanger for receiving a natural cold source and an absorption refrigerator for receiving heat supplied by at least one of a hydrogen fuel cell, a heat storage tank and an internal combustion engine, wherein the electric refrigerator, the heat exchanger and the absorption refrigerator are used as cold load sources of a data center.

3. The method of claim 1, wherein the two-tier co-optimization of the data center co-production system,

Normalized S _norm,d for pareto front solution is expressed as:

Normalized Euclidean distance Dist _p, expressed as:

4. The method for double-layer collaborative optimization of a data center co-production system according to claim 2, wherein upper layer configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity and collector area are optimized with energy utilization, economic cost and environmental benefit as optimization targets by using a non-dominant ranking genetic algorithm; or optimizing lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total power consumption of an electrolytic tank and the total power consumption of a storage battery and the proportion of power supply of a hydrogen fuel cell to the total power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target gray wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets, and constructing the following model:

the generated energy of the photovoltaic panel is as follows:

E_mn,pv＝A_pvI_bη_pv×10^-3；

the electricity generation amount of the hydrogen fuel cell is as follows:

E_mn,fc＝H_mn,fcη_fc,e；

the heat generation amount of the hydrogen fuel cell is:

Q_mn,fc＝H_mn,fcη_fc,h；

the hydrogen storage amount of the hydrogen storage tank is as follows:

The model of the battery is expressed as:

The power generation amount of the internal combustion engine is:

The fuel consumption of the internal combustion engine is:

The heat generation amount of the internal combustion engine is:

Q_mn,ice＝F_mn,ice(1-η_mn,ice)；

the power consumption of the power grid is as follows:

E_mn,grid＝E_mn,dc+E_mn,a+E_mn,el-E_mn,ice-E_mn,pv-E_mn,fc；

wherein E _mn,grid is the power consumption of the power grid;

The fuel consumption of the power grid is:

The thermal energy storage model of the heat storage tank is as follows:

The mathematical model of the absorption refrigerator is as follows:

the heat collection amount of the heat collector is as follows:

Q_mn,st(t)＝A_stI_mn,bη_st×10^-3；

The cooling efficiency of natural cooling is as follows:

the electric balance of the data center co-production system is as follows:

E_mn,ice+E_mn,grid+E_mn,pv+E_mn,fc＝E_mn,dc+E_mn,a+E_mn,el；

The heat balance of the data center co-production system is as follows:

the fuel consumption from the grid is:

Wherein F _sp,grid is the fuel consumption of the power grid;

The primary energy consumption of the reference system is as follows:

Wherein F _sp is the primary energy consumption of the reference system.

5. The method of double-layer collaborative optimization of a data center co-production system of claim 4, wherein the data center co-production system is configured to execute a refrigeration demand regulation strategy based on power supply drive suitable for temperatures no less than 12 ℃ or a power optimization configuration strategy based on refrigeration demand guidance suitable for temperatures less than 12 ℃;

Wherein,

E_mn,el＝K₁(E_need-E_mn,pv)；

E_mn,esc＝(1-K₁)(E_need-E_mn,pv)；

The required heat Q _ac,need for driving the absorption chiller is:

Wherein,

The required heat Q _ac,need for driving the absorption chiller is:

Q_ac,need＝C_ac,need/1.2

C_ac,need＝K₃C_need；

the refrigerating capacity of the electric refrigerator is as follows:

C_chill＝(1-K₃)C_need

Q_mn,fc＝K₄Q_need1；

Q_mn,ice＝(1-K₄)Q_need1；

6. A dual-layer optimization system for a data center co-production system, comprising:

An upper layer optimization model configured to optimize upper layer configuration parameters including internal combustion engine capacity, photovoltaic panel area, heat storage tank capacity, and collector area using a non-dominant ranking genetic algorithm with energy utilization, economic cost, and environmental benefit as optimization targets;

the lower layer optimization module is configured to optimize lower layer proportion parameters comprising the proportion of absorption refrigeration capacity to total refrigeration capacity, the proportion of natural cooling refrigeration capacity to total refrigeration capacity, the proportion of power consumption of an electrolytic tank to the total charge of a storage battery and the proportion of power supply of a hydrogen fuel cell to the total power supply of a hydrogen fuel cell and an internal combustion engine by utilizing a multi-target wolf optimization algorithm with the energy utilization rate and the economic cost as optimization targets;

The interaction optimization module is configured to optimize the optimized upper-layer configuration parameters and lower-layer proportion parameters through a TOPSIS decision algorithm, and solve the optimal capacity configuration and the optimal operation parameters;

Wherein,

7. An apparatus, comprising: at least one database; and a memory communicatively coupled to the at least one database; wherein the memory stores instructions executable by the at least one database to enable the at least one database to perform the dual-layer co-optimization method of the data center co-production system of any one of claims 1-5.

8. A computer readable medium having stored thereon computer executable instructions which when executed by a processor implement a method of double-layer co-optimization of a data center co-production system according to any of claims 1-5.