CN110659830A

CN110659830A - Multi-energy micro-grid planning method for comprehensive energy system

Info

Publication number: CN110659830A
Application number: CN201910907911.5A
Authority: CN
Inventors: 杨宇全; 秦丽杰; 张奇; 张建海; 李健; 李禹曈; 魏巍; 奚鹏飞; 邵鹏; 陈利; 方韶; 鲁丹丹; 李鹏程; 张文茹; 赵重; 白宗辉; 张金禄; 何玉龙; 张志朋; 武晓晶
Original assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd
Current assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd
Priority date: 2019-09-25
Filing date: 2019-09-25
Publication date: 2020-01-07

Abstract

The invention discloses a multi-energy micro-grid planning method for an integrated energy system, which comprises the following steps: 1) designing a multi-energy microgrid basic framework: interconnecting and integrating a traditional power system, a natural gas system and a thermodynamic system in an area; 2) constructing a multi-energy micro-grid multi-target collaborative planning model with the objective functions of minimum system planning operation cost, minimum wind and light abandonment and minimum pollutant emission, and realizing collaborative optimization planning of a plurality of energy subsystems; 3) the method comprises the steps of improving a traditional multi-target genetic algorithm by using Tent mapping improved chaotic optimization algorithm to form a chaotic optimization multi-target genetic algorithm, and solving a model by using the chaotic optimization multi-target genetic algorithm on the basis of verifying algorithm robustness. The invention effectively reduces the system cost while meeting the requirement of the multi-load, obviously improves the efficiency of resource allocation, has positive effect on the safe operation of a large power grid system, and has better application prospect.

Description

Multi-energy micro-grid planning method for comprehensive energy system

Technical Field

The invention relates to the fields of power industry, physical asset management, centralized control management of an information system and the like, in particular to a multi-energy micro-grid planning method for an integrated energy system.

Background

With the continuous development of economic society of China, the energy production and consumption modes are greatly changed, and the energy industry bears new missions of improving energy efficiency, guaranteeing energy safety, promoting new energy consumption, promoting environmental protection and the like. The construction of the comprehensive energy system is a necessary choice for building a clean, low-carbon, safe and efficient modern energy system in China.

At present, the research on the comprehensive energy system in China is still basically in the aspect of concept, system form and operation mode exploration, and the research on a system planning method is less. And the planning operation of the comprehensive energy system under the target form is directly researched, so that the feasibility is not provided at the modeling and solving level, and the direct and effective ground verification and application are difficult to realize at the relevant achievement stage.

Disclosure of Invention

The invention aims to solve the technical problems and provides a comprehensive energy system-oriented multi-energy microgrid planning method with a wide application prospect.

In order to solve the technical problems, the invention adopts the following technical scheme:

a multi-energy micro-grid planning method for an integrated energy system comprises the following steps:

1) designing a multi-energy microgrid basic framework: the multi-energy microgrid is used for interconnecting and integrating a traditional power system, a natural gas system and a thermodynamic system in a region, and can realize the cooperative utilization of multiple types of energy systems by taking the microgrid as a core, so that the multi-energy microgrid can meet the multiple types of energy requirements of users;

2) constructing a multi-energy micro-grid multi-target collaborative planning model: the multi-energy micro-grid multi-target collaborative planning model takes the minimum system planning operation cost, the minimum wind and light abandonment and the minimum pollutant emission as objective functions to realize collaborative optimization planning of a plurality of energy subsystems;

3) improving a model solving algorithm: the method comprises the steps of improving a traditional multi-target genetic algorithm by using Tent mapping improved chaotic optimization algorithm to form a chaotic optimization multi-target genetic algorithm, and solving a model by using the chaotic optimization multi-target genetic algorithm on the basis of verifying algorithm robustness.

Further, the objective function of the multi-energy micro-grid multi-target collaborative planning model comprises a system planning operation cost objective function, a system wind and light abandoning objective function and a system environment cost objective function.

Further, assuming that the planning period of the system is 20 years, each year includes 365 same scheduling days, so that the operation cost of the whole operation period of the system is completely discounted, the total operation cost of the system can be obtained, and the objective function of the planning operation cost of the system is expressed as follows:

wherein Y represents the Yth year of the planning period; r is interest rate, is used for reflecting the time value of capital, and takes 5 percent; c_plan(1-r)¹⁹Representing the residual value recovery at the end of the planning period, and taking the residual value recovery rate as 5 percent;

wherein, C_planRepresenting the planning cost of the system, and developing the following formula (2):

C_plan＝C_sup+C_l+C_es (2)

C_sup、C_l、C_esrespectively representing the cost of the energy supply equipment and the energy storage equipment; assuming that the cost of each energy supply device is a direct proportional function of its capacity, the cost of the energy transmission network is a direct proportional function of its length, and the cost of the energy storage device is a direct proportional function of its energy storage, the above costs can be further expanded as shown in equation (3):

in the formula (I), the compound is shown in the specification,

represents the capacity of the energy supply device d; c. C_dRepresents the cost per unit capacity of the energy supply device d; l is_lDenotes the length of the network l, c_lRepresents the cost per unit length of the network;

represents the capacity of the energy storage device s; c. C_sRepresents the cost per unit capacity of the energy storage device s;

C_operrepresents the system operating costIncluding the cost of purchasing mains electricity and the operating costs of fuel cells, distributed gas turbines and natural gas hot water boilers, the cost of purchasing gas from gas companies, and demand response costs, as shown in equation (4):

in the formula, q_gas,t，q_H2,tRespectively representing the electric output of the distributed gas turbine set and the fuel cell at the time t; h is_gbThe thermal output of the natural gas hot water boiler at the time t is shown; epsilon_H2Hydrogen consumption, V, representing the specific output of the fuel cell_H2,tRepresenting the residual hydrogen amount of the hydrogen storage tank at the time t; p is a radical of_gasAnd p_gbRespectively representing the unit power supply cost of the distributed gas turbine unit and the unit heat supply cost of the natural gas hot water boiler, p_H2Represents the hydrogen price; q. q.s_b,gridAnd q is_s,gridRespectively representing the electric quantity purchased from the main network and the electric quantity sold to the main network; p is a radical of_b,tAnd p_s,tRespectively showing real-time electricity prices of electricity purchasing and electricity selling to the main network at the time t; q. q.s_gal,tAnd P_gasRespectively representing the natural gas load and the natural gas price of the user terminal at the time t; mu.s_dr,i,tIs a variable from 0 to 1, which represents the interruptible load calling state of the user corresponding to the node i at the time t: when 0 is taken, the load of the user is not interrupted, and when 1 is taken, the load of the user is interrupted; c. C_dr,iRepresenting the fee that the invocations of the interruptible load resource of the user i need to pay to the user; at present, only interruptible loads are considered, and in the actual operation process, due to technical limitation, a multi-energy microgrid operator cannot directly interact with specific electric equipment inside a user so as to achieve the purpose of selectively cutting off electric equipment of the user, and therefore if a certain interruptible load resource needs to be called, all loads of the user are cut off.

Further, the wind abandoning and light abandoning at the time t is defined as the difference between the maximum electric quantity which can be generated by the system wind power and the photovoltaic set and the actual generated electric quantity under the natural condition at the time, and a system wind abandoning and light abandoning objective function is expressed as follows:

in the formula (I), the compound is shown in the specification,

and

respectively representing the upper output limits of distributed wind power generation and distributed photovoltaic power generation in a system at the moment t; p_wind,tAnd P_PV,tThe total output power of all distributed wind power and photovoltaic power generation in the system at the moment t is respectively.

Further, the system environment cost objective function is expressed as follows:

in the formula, P_gas,t、P_CHP,tAnd P_boil,tRespectively the output of the distributed gas turbine set, the coal-fired CHP and the natural gas hot water boiler at the moment t; omega_gas,NOx、ω_CHP,NOxAnd ω_boil,NOxThe nitrogen oxide emission coefficients of the distributed gas turbine unit, the coal-fired CHP and the natural gas hot water boiler are respectively; omega_gas,SO2、ω_CHP,SO2And ω_boil,SO2Sulfur dioxide emission coefficients of the distributed gas turbine unit, the coal-fired CHP and the natural gas hot water boiler are respectively; omega_CHP,dustIs the dust emission coefficient of the coal-fired CHP unit; c. C_NOx、c_SO2And c_dustRespectively representing the pollution control cost of nitrogen oxide, sulfur dioxide and dust.

Furthermore, the constraint conditions of the multi-energy micro-grid multi-target collaborative planning model comprise power balance constraint, electric heating power flow constraint, unit output constraint, cogeneration operation constraint, energy coupling constraint, energy storage operation constraint, clean energy installation constraint and demand response constraint.

Further, the key steps of the chaos optimization multi-objective genetic algorithm are as follows:

1) inputting basic parameters: inputting an initial network structure, a load curve and price information of the multi-energy micro-grid; determining the value of each constraint condition; inputting initial parameters of a chaos optimization multi-target genetic algorithm;

2) chaotic initialization: generating a plurality of times of initial solutions of the population number by using a Tent mapping-based chaotic optimization algorithm, and preferentially selecting the initial solutions as a primary generation father population of the multi-target genetic algorithm; the result of simple superposition of the dimensionless per unit values of the multiple targets is used as an index for evaluating the chaotic solution;

3) genetic manipulation and elite retention: carrying out genetic operation on the parent population according to a traditional multi-target genetic algorithm to generate a variant population; combining the parent population and the variant population, and preferentially selecting individuals from the parent population as a child population, namely a next generation parent population; keeping the number of each generation of father population unchanged;

4) periodic elite attenuation: when the iteration times reach integral multiples of 1/100 of the maximum times, a half of elite solutions are cut off in a roulette mode and replaced by solutions with the same quantity generated by a chaotic optimization algorithm; the mechanism of generating the solution is the same as that of the chaotic initialization stage, namely, the chaotic solution of multiple times is generated firstly, and then the solution entering the population is selected preferentially;

5) outputting an optimal solution: and when the iteration times reach the upper limit, outputting the parent population at the moment as a Pareto optimal solution set.

The invention discloses a multi-energy micro-grid planning method for an integrated energy system, which has the following effects in particular:

1. although additional equipment types need to be put into the multi-energy microgrid in the planning stage, from the viewpoint of optimization results, compared with the current planning mode, the planning idea of the multi-energy microgrid has significant advantages in the aspects of planning, operation and total cost, and the pollutant emission is less.

2. The co-planning and coupled operation of the various types of energy subsystems helps to fully consume the renewable energy.

3. The system effectively reduces the system cost while meeting the requirement of multi-element load by a flexible energy supply mode, obviously improves the efficiency of resource allocation, and also has positive effect on the safe operation of a large power grid system.

4. The method is characterized in that a Tent mapping based chaotic optimization algorithm is provided for improving an NSGA-II algorithm to realize multi-target model solution, the improved NSGA-II algorithm utilizes the Tent mapping based chaotic optimization algorithm to generate an initial solution, and meanwhile, a new chaotic solution is periodically utilized to replace part of elite solutions in an elite retention link of the algorithm, so that the solution efficiency is improved at the initial stage of the algorithm, the advantages of the elite solutions are weakened at the later stage of the algorithm, and the possibility that the algorithm jumps out of local optimization and continues to find the optimal solution in a larger space is improved.

5. The performance test result of the algorithm shows that the algorithm provided by the invention is superior to the traditional NSGA-II algorithm and the multi-target PSO algorithm, and has better application prospect when solving the multi-target optimization problem.

Drawings

Fig. 1 is a basic architecture diagram of a multi-energy microgrid;

FIG. 2 is a diagram of the operational feasible region of a typical CHP;

FIG. 3 is a flow chart of a chaos optimization NSGA-II algorithm;

FIG. 4 is a graph of test results without Gaussian noise;

FIG. 5 is a graph of test results for 0.01% Gaussian noise;

FIG. 6 is a graph of test results for 0.1% Gaussian noise;

FIG. 7 is a diagram of an improved Garver test system;

FIG. 8 is a graph of electrical, thermal, and gas loads for a region to be planned;

FIG. 9 is a time of use electricity price graph;

FIG. 10 is a graph of typical operating day light intensity and wind speed;

FIG. 11 is a Pareto optimal solution space diagram obtained by chaos optimization of NSGA-II algorithm;

FIG. 12 is a diagram showing the planning result of the scheme A according to the present invention.

Detailed Description

The invention is described in further detail below with reference to the following figures and detailed description:

the technology mainly comprises the following parts: a basic framework of a multi-energy microgrid in an area is provided; constructing a multi-energy micro-grid multi-target collaborative planning model; solving the model by using Tent mapping improved chaos optimization algorithm; a park to be planned in North China is taken as a research object to design an arithmetic example, and compared with the existing micro-grid planning method, the scientific and economic advantages of the model built in the arithmetic example are analyzed.

1. Basic architecture of multi-energy microgrid in area: the multi-energy micro-grid is formed by interconnecting and integrating a traditional power system, a natural gas system and a thermodynamic system in a region and takes the micro-grid as a core. The multi-energy microgrid can realize the cooperative utilization of multi-type energy systems, so that the multi-type energy requirements of users are met. The basic architecture of the multi-energy microgrid is shown in the attached figure 1 according to an energy supply and demand structure and energy flow direction.

2. The method comprises the following steps of (1) target functions of a multi-energy micro-grid multi-target collaborative planning model: the multi-objective function constructed in the method considers three aspects of minimum operation cost, minimum wind and light abandoning amount and minimum pollutant discharge amount of system planning.

(1) System planning and operation cost minimization

It is assumed herein that the planning period of the system is 20 years, each year comprising 365 identical scheduling days. Therefore, the total operation cost of the system can be obtained by converting the operation cost of the whole operation period of the system. In summary, the system planning operation cost objective function (1) is shown.

Wherein Y represents the Yth year of the planning period; r is interest rate, which is used to represent the time value of capital, which is taken as 5% herein. The last term in the above equation represents the recovery of the residual at the end of the planning period, taking the residual recovery as 5%.

C_plan＝C_sup+C_l+C_es (2)

C_sup、C_l、C_esrepresenting the cost of the energy supply device and the energy storage device, respectively.Assuming that the cost of each energy supply device is a direct proportional function of its capacity, the cost of the energy transmission network is a direct proportional function of its length, and the cost of the energy storage device is a direct proportional function of its energy storage (fixed input and output power), the above-mentioned costs can be further expanded as shown in equation (3):

in the formula (I), the compound is shown in the specification,represents the capacity of the energy supply device d; c. C_dRepresents the cost per unit capacity of the energy supply device d; l is_lDenotes the length of the network l, c_lRepresents the cost per unit length of the network;

represents the capacity of the energy storage device s; c. C_sRepresenting the cost per unit capacity of the energy storage device s.

C_operRepresenting the cost of system operation. Including the cost of purchasing mains electricity and operating costs of fuel cells, distributed gas turbines, and natural gas hot water boilers, the cost of purchasing gas from gas companies, and demand response costs. As shown in formula (4):

in the formula, q_gas,t，q_H2,tRespectively representing the electric output of the distributed gas turbine set and the fuel cell at the time t; h is_gbThe thermal output of the natural gas hot water boiler at the time t is shown; epsilon_H2Hydrogen consumption, V, representing the specific output of the fuel cell_H2,tIndicating the residual hydrogen amount of the hydrogen storage tank at time t. p is a radical of_gasAnd p_gbRespectively represents the unit power supply cost of the distributed gas turbine unit and the unit heat supply cost (mainly converted by the natural gas price) of the natural gas hot water boiler, p_H2Indicating the hydrogen price. q. q.s_b,gridAnd q is_s,gridRespectively indicate the power purchased from the main network andthe amount of electricity sold to the main network; p is a radical of_b,tAnd p_s,tAnd respectively showing the real-time electricity prices of electricity purchase and electricity sale to the main network at the time t. q. q.s_gal,tAnd P_gasRespectively representing the natural gas load and the natural gas price of the user terminal at the moment t. Mu.s_dr,i,tIs a variable from 0 to 1, which represents the interruptible load calling state of the user corresponding to the node i at the time t: when 0 is taken, the load of the user is not interrupted, and when 1 is taken, the load of the user is interrupted; c. C_dr,iRepresenting the cost, ten thousand dollars, that needs to be paid to the user to invoke the interruptible load resource of user i. At present, only interruptible loads are considered, and in the actual operation process, due to technical limitation, a multi-energy microgrid operator cannot directly interact with specific electric equipment inside a user, so that the purpose of selectively cutting off the electric equipment of the user is achieved. Thus, if a certain interruptible load resource needs to be invoked, the entire load of the user will be cut off.

(2) The system has least wind and light abandonment

The method allows wind curtailment and light curtailment to be allowed, but in order to improve the utilization efficiency of renewable energy sources as much as possible in optimization, the wind curtailment and the light curtailment are taken as one of the objective functions at least, the wind curtailment and the light curtailment at the time t are defined as the difference between the maximum electric quantity capable of being generated by the system wind power and the photovoltaic set and the actual generated electric quantity under the natural condition at the time, and the objective function with the minimum wind curtailment and the light curtailment of the system can be expressed as follows:

in the formula (I), the compound is shown in the specification,and

and respectively representing the output upper limit of the distributed wind power generation and the distributed photovoltaic power generation in the system at the moment t. P_wind,tAnd P_PV,tThe total output power of all distributed wind power and photovoltaic power generation in the system at the moment t is respectively. As mentioned above, the upper limit of output of distributed wind power generation and distributed photovoltaic power generation in the system at the time tThe specific physical characteristics are determined by the unit performance and the wind speed and the illumination intensity at the moment, and are given in the following.

(3) Minimum pollutant discharge amount of system

The burning of coal CHP will produce sulfur dioxide, nitrogen oxides and dust, and the distributed gas turbine and natural gas hot water boiler will produce nitrogen oxides. The influence of various pollutants on the environment is quantified and uniformized by introducing a discharge coefficient and treating pollution cost. The system environment cost objective function is shown as equation (6).

In the formula, P_gas,t、P_CHP,tAnd P_boil,tRespectively the output of the distributed gas turbine set, the coal-fired CHP and the natural gas hot water boiler at the moment t; omega_gas,NOx、ω_CHP,NOxAnd ω_boil,NOxThe nitrogen oxide emission coefficients of the distributed gas turbine unit, the coal-fired CHP and the natural gas hot water boiler are respectively; omega_gas,SO2、ω_CHP,SO2And ω_boil,SO2Sulfur dioxide emission coefficients of the distributed gas turbine unit, the coal-fired CHP and the natural gas hot water boiler are respectively; omega_CHP,dustIs the dust emission coefficient of the coal-fired CHP unit. c. C_NOx、c_SO2And c_dustRespectively representing the pollution control cost of nitrogen oxide, sulfur dioxide and dust. It should be noted that although the pollution discharge amount is converted by the pollution discharge cost, the pollutant discharge amount can be directly regarded as the environmental cost and added to the planned operation cost. However, such treatments may result in environmental benefits being covered by economic benefits. In order to specifically analyze the environmental benefits of the multi-energy microgrid and the relationship between the environmental benefits and the economic benefits and the clean energy consumption, the method still selects to set the converted pollutant emission amount as an objective function independently.

It should be particularly noted that, in view of the general normality of energy and power system planning, the model constructed herein is a multi-objective optimization model, that is, the system planning needs to ensure the minimum total cost of planning operation, the minimum amount of wind and light abandonment, and simultaneously ensure the minimum pollution emission on a typical scheduling day. However, the 3 objective functions are not independent but coupled to each other, and it is often impossible to achieve these objectives simultaneously, and only a relatively optimal solution can be found. The core ideas of the model and the calculation process herein are therefore: after multiple iterative computations, a plurality of groups of optimal solutions are obtained through solving, and a result with the minimum total planning operation cost is selected from the optimal solutions. The optimal solutions are calculation results which can relatively realize the optimization of the three objective functions in the solution set, possible results such as low planning operation cost, high abandoned wind and abandoned light, low pollution emission, high planning operation cost and the like are eliminated, and then a result meeting a certain requirement is selected as the optimal solution in the relatively optimal scheme. The planning problem of the multi-energy microgrid is mainly researched, so that after a plurality of groups of optimal results are obtained, the result with the relatively minimum planning operation cost is selected. Meanwhile, although the optimal scheme is selected by minimizing the planning operation cost, the realization of the relative optimization of the three objective functions is still the premise of selecting the final calculation result.

3. Constraint conditions of the multi-energy micro-grid multi-target collaborative planning model are as follows:

(1) power balance constraint

The output of the electric and thermal units at any time, the energy charging and supplying of various types of stored energy and the load are kept balanced. On the natural gas side, only the terminal gas load needs to be considered, since the natural gas consumption of the gas turbine is already converted by the unit cost. Line losses are not considered for the moment. In summary, the power balance constraint is shown in equations (7) - (8).

q_gas,t+q_CHP,t+q_PV,t+q_wind,t+q_H2,t+q_es,t+q_b,grid,t＝q_load,t+q_eb.t+q_e2h.t+q_s,grid,t-q_DR,t (7)

h_gb,t+h_eb,t+h_es,t＝h_load,t (8)

In the formula, q_gas,t、q_PV,t、q_wind,t、q_H2,t、q_es,tAnd q is_b,grid,tRespectively representing the electric power injected by a distributed gas turbine unit, fire coal CHP, photovoltaic power, wind power, a fuel cell, a storage battery and a main network in a system at the time t; q. q.s_load,t、q_eb,t、q_e2h,tAnd q is_s,grid,tRespectively representing the electric load, the electric water boiler load, the electric hydrogen load and the power injected into the main network at the time t; q. q.s_DR,tRepresenting the power curtailment of the demand response at time t. h is_gb,t、h_eb,t、h_es,tRespectively showing the thermal power of the natural gas hot water boiler, the electric water heating boiler and the heat storage; h is_load,；tIndicating the thermal load at time t. Since a plurality of nodes exist in the system, each of equations (7) and (8) is the sum of the corresponding physical quantities of the respective nodes.

(2) Electric heat current restraint

In the aspect of power flow, simplified single-phase direct current flow is adopted, MATPOWER software is used for simulation calculation, and specific formulas are not repeated. However, for safe and stable operation of the system, the current and the node voltage of each branch need to be checked after the load flow calculation is converged. As shown in formula (9):

in the formula, in the formula:Vand

the lower limit and the upper limit of the node voltage are respectively; v is used for voltage of node k at t_k,tAnd (4) showing. I is_l,tRepresents the current of branch i at t;

representing the current maximum capacity of branch i.

In terms of thermodynamic flow, reference is made herein to a heat network model, including node flow balance constraints and thermal energy-flow constraints. As shown in equation (10).

In the formula (f)_,i,tThe flow between the node i and the ith coal-fired CHP system at the moment t; f. of_ij,tThe flow rate of each thermal power pipe section connected with the node i at the moment t. c is the specific heat capacity of water, T_sAnd T_tRespectively the feed water temperature and the return water temperature. By integrating the equation (10), a node heat balance equation can be obtained, as shown in equation (11):

H_ex,i,t+∑H_ij,t＝0 (11)

(3) unit output constraint

The output of various power generation and heat supply units is in a rated range, and the distributed gas unit also has climbing restraint as shown in formula (12).

In the formula, v_downAnd v_upThe maximum downhill and uphill speeds of the gas turbine unit are respectively, and the meanings of the other variables are the same as the previous. Needs to be explained; the lower limit of the output of the distributed photovoltaic and wind power at the time t is 0, the upper limit is a function of the local illumination intensity, the wind speed and the rated power of the corresponding unit, and the specific form of the function is not repeated here. Upper limit or rated power of each unit and in formula (3)

The same is true. The operation mode of the coal CHP for fixing the power by heat requires that the thermoelectric ratio of the system is in a certain interval, and is special; while fuel cells, electrical hydrogen placement and storage are considered herein as energy coupling units, relevant constraints and physical models are given later.

(4) Cogeneration operation constraints

The CHP is constrained by the feasible domain during operation, as shown in fig. 2. The closed area enclosed by FIJK is a feasible area of CHP, and the following constraints are required to be met:

in the formula: psi and

for the parameters of the line segment composed by the feasible region, L is equal to {1, 2.Respectively representing the power supply power of the unit at the fixed points F, K, I and J of the feasible region,

respectively represent the heat supply power of the unit at the feasible region fixed points F, K, I and J.

(5) Restraint of energy coupling

The energy coupling loop in the system comprises thermal-electric coupling (electric water heating boiler), hydrogen-electric coupling (fuel cell/electric hydrogen storage/hydrogen storage), gas-thermal coupling (natural gas hot water boiler) and gas-electric coupling (distributed gas turbine unit). However, since natural gas is only used as fuel for the hot water boiler and the gas turbine unit, respectively, and is not coupled with the terminal natural gas load, and the constraint of the upper limit of natural gas supply in unit time is not involved, only the thermo-electric and hydrogen-electric coupling is considered at the present stage.

The thermo-electric coupling constraint is shown as equation (14).

q_eb,t＝η_ebh_eb,t (14)

In the formula eta_ebThe comprehensive efficiency of the heat-electricity conversion of the electric water boiler is shown.

The hydrogen-electric coupling constraint is shown in equation (15).

In the formula eta_e2h,tAnd η_H2Thermal efficiencies of the electro-hydrogen and fuel cell power generation, respectively; c. C_H2Is the heating value of hydrogen; Δ V_e2h,tAnd Δ V_H2The hydrogen amount generated by the electric hydrogen placing and the hydrogen consumed by the fuel cell for power generation are respectively; n is_e2hIs the number of the electric hydrogen placing devices,

is the maximum rate at which a single electrical hydrogen plant produces hydrogen;is the upper limit of the capacity of the hydrogen storage tank.

(6) Energy storage operation restraint

The energy storage comprises a storage battery and a heat storage pool, and for the sake of convenience in subsequent solving, the physical models of the storage battery and the heat storage pool are considered to be consistent in other structures except for different specific parameters. The energy storage comprises three operation states of energy charging, energy supply and stability.

In the state of energy charging:

under the energy supply state:

in a steady state:

S_t+1＝S_t(1-ε) (18)

in the formula, S_t+1Is the residual energy of the stored energy at the end of the t +1 time interval, PC SOC, t +1 and PD SOC, t +1 is the charging and energy supply power of the stored energy at the time interval, eta_cAnd η_DRespectively charge and discharge efficiency of energy storage; ε is the loss rate. Further considering the life and safety of the stored energy, the following constraints should be added:

firstly, energy charging and energy supply rate constraint: the charging and energy supply rate is positive and does not exceed the upper limit.

In the formula (I), the compound is shown in the specification,

and

respectively, the upper limit of the charging and energizing rates.

Secondly, energy charging rate constraint: the remaining energy of the stored energy should be kept within a range that is neither too close to a full state nor too close to a completely empty state.

In the formula, B_SOC,tIs the remaining capacity S of the stored energy at the end of the t period_tAccount for the total quantity of electricity

The proportion of (a) to (b), i.e. the charging rate of the stored energy;B_SOC and

respectively, the upper and lower limits of the energy storage state of charge.

(7) Clean energy installation restraint

Because the multi-energy microgrid is developed by a microgrid, the national requirement on the installed capacity of clean energy of the grid-connected microgrid is met. Distributed wind power, photovoltaic and fuel cell installed capacities are required here to be not less than 50% of the typical daily maximum electrical load of the system. Namely:

q_PV+q_wind+q_H2≥50％×maxD_t (21)

in the formula, q_PV、q_windAnd q is_H2Respectively representing distributed wind power, photovoltaic and fuel cell installed capacities; d_tRepresenting the electrical load of the system at a typical time of day t.

(8) Demand response constraints

Since the interruptible load considered herein does not have the requirement of load translation by the same amount, the demand response constraint here is relatively simple: if a load node is powered off for multiple times in a day, the principle goes against the principle of high-quality power supply service, so that the requirement that a load node is powered off at most once in a day can be realized by limiting the virtual variables as follows:

4. solving the model: the objective function constructed herein has three sub-objectives, which is essentially a multi-objective optimization problem. Therefore, the NSGA-II algorithm widely applied in the field of multi-objective optimization is introduced, and the chaotic optimization algorithm is applied to carry out adaptive improvement on the NSGA-II algorithm so as to realize model solution.

(1) The multi-target genetic algorithm comprises the following steps: the core of multi-objective optimization is to coordinate the relationship among the objective functions and find out the optimal solution set which enables the objective functions to reach the satisfied values as much as possible. The NSGA-II is a classical multi-target genetic algorithm, a rapid non-dominated sorting method and an elite strategy are introduced, the crowdedness is defined to replace the sharing of adaptive values, the computational complexity of the algorithm is reduced, and the computational efficiency is improved. If there are N sub-targets, NSGA-II defines the individual i crowdedness as:

in the formula (f)_k(i +1) and f_k(i-1) the optimized values at k target for i +1 and i-1 individuals, respectively. The detailed calculation process of the conventional NSGA-II algorithm is not described in detail herein.

(2) The improved chaos optimization algorithm comprises the following steps: the basic idea of chaos optimization is as follows: and mapping the optimized variable into a value interval of a chaotic variable space through a chaotic mapping rule, searching for optimization by utilizing the ergodicity and regularity of the chaotic variable, and finally linearly converting the obtained optimized solution into the optimized space. The chaotic algorithm usually adopts Logistic mapping, but recent related research shows that Tent mapping has better chaotic characteristics than Logistic mapping after being improved. The improved Tent map is shown in equation (24).

Wherein the content of the first and second substances,

the calculation process of the conventional chaotic algorithm is not described in detail herein.

(3) Chaos optimization multi-objective genetic algorithm (chaos optimization NSGA-II algorithm): although NSGA-II has gained wide application in the field of multi-objective optimization, it still has significant local optimization problems. This is because the elite strategy makes the locally optimal solution have too great a genetic advantage in later iterations of the algorithm, thereby limiting the algorithm from further searching to wider or more optimal regions. Meanwhile, the initialization process of the algorithm completely adopts random values, so that a multi-peak function may have a search blind area. Therefore, the chaos optimization NSGA-II algorithm is provided, the chaos optimization algorithm based on Tent mapping is embedded into the main process of the NSGA-II algorithm, and the initial process and the elite retention process of the algorithm are improved, so that the initial optimization efficiency of the algorithm can be improved; on the other hand, the genetic advantage of the elite strategy can be weakened in the later stage of the algorithm, so that the algorithm can jump out of the local optimum and the probability of obtaining the global optimum solution is increased. The flow of the chaos optimization NSGA-II algorithm is shown in the attached figure 3.

The key steps of the algorithm are explained in detail below:

first, basic parameters are input. Inputting an initial network structure, a load curve and price information of the multi-energy micro-grid; determining the value of each constraint condition; inputting initial parameters of the chaos optimization NSGA-II algorithm and the like.

Second, chaotic initialization. And generating a plurality of times of initial solutions of the population quantity by using a Tent mapping-based chaotic optimization algorithm, and preferentially selecting the initial solutions as the initial generation father population of the NSGA-II algorithm. And the result obtained by simply superposing the dimensionless per unit values of the targets is used as an index for evaluating the chaotic solution.

Third, genetic manipulation and elite retention. And (4) carrying out genetic operation on the parent population according to the traditional NSGA-II algorithm to generate a variant population. And combining the parent population and the variant population, and preferentially selecting individuals from the parent population as the sub-population, namely the next generation parent population. Keeping the number of each generation parent population unchanged.

Fourth, periodic elite weakening. When the number of iterations reaches an integer multiple (rounded) of 1/100, which is the maximum number, half of the elite solutions are discarded in the roulette manner and replaced with an equivalent solution generated by a chaotic optimization algorithm. The mechanism of generating the solution is the same as that of the chaotic initialization stage, namely, the chaotic solution of multiple times is generated firstly, and then the solution entering the population is selected preferentially.

And fifthly, outputting the optimal solution. And when the iteration times reach the upper limit, outputting the parent population at the moment as a Pareto optimal solution set.

The robustness of the above algorithm is tested herein below by the classical DTLZ-3 test function, which is shown as equation (26).

The test results are shown in fig. 4 to fig. 6, the optimal result of the test of the conventional DTLZ-3 is a spherical surface, but in order to solve the model of the text, the optimal curved surface is a saddle surface after the algorithm is modified to a certain extent. Fig. 4 shows the test results without gaussian noise, fig. 5 shows the test results with 0.01% gaussian noise, and fig. 6 shows the test results with 0.1% gaussian noise. From the test results of adding gaussian noise twice, as shown in the figure, after 0.01% gaussian noise interference is added, the test result of the improved NSGA-II algorithm is basically consistent with the noise-free result except for individual points, and the curved surface of the result has part of points raised. After the Gaussian noise is increased to 0.1%, part of calculation results float, but the consistency with the original results is still kept, and the curved surface of the test result is basically kept unchanged. Therefore, the improved NSGA-II algorithm has better performance and stronger robustness in solving the multi-objective optimization problem through all three test results.

5. Example analysis

(1) Example system

The method is characterized in that a Garver test system is improved based on the data of the comprehensive energy system test points in a certain plan in North China, and the performance of a model is verified by designing an example, wherein the initial grid structure of the system is shown in figure 7.

The system comprises a 7-node heat supply network system and a 6-node power test system. The natural gas network system comprises 2 CHP nodes (configured with machine sets) and 4 heat load nodes, data of the heat load nodes are marked in the graph, and system basic data of the power system are shown in tables 1 and 2.

Table 1 Garver test system transmission line data

TABLE 2 load node data for Garver test System

Electrical load node	Node electrical load ratio%	Thermal load node	Node thermal load ratio/%
				1	10	1	30
2	20	3	25
				3	15	4	25
4	15	6	20
				5	30
6	10

It should be noted that there is a specific load of a given node in the basic Garver test system, but in order to match the example with the park to be planned on the system scale, the total load of the system is shown in fig. 8, and the total load is distributed to each node in the node system according to the load proportion of the original system. Here, since the natural gas pipeline network is not planned for a while, the natural gas load is not distributed, and is simply handled as a part of the cost.

The local time-of-use electricity price curve is shown in fig. 9, and the electricity price is sold to the main network to be the same.

The wind speed and the illumination intensity of a typical day are obtained by constructing an ARIMA time series model according to historical data provided by a local meteorological office, the specific process is not repeated here, and the prediction result is shown in FIG. 10. The investment and operating costs of each device in the multi-energy microgrid are shown in table 2.

TABLE 3 System Primary device parameters

The parameter values of the chaos optimization NSGA-II algorithm are set as follows:

first, each generation of population comprises 200 individuals;

secondly, the chaos optimization process always generates a chaos solution with three times of the needed vector for preferential selection;

thirdly, the variation rate in the NSGA-II algorithm is 0.8, and the crossing rate is 0.2;

fourth, the maximum number of iterations t_max1000. And inputting the parameters into MatlabR2016a software, and solving and analyzing the model by self-programming.

To analyze the superiority of the model built herein, the example gives 2 alternative planning options:

scheme A: the protocol resulting from the modeling is constructed herein.

Scheme B: and each subsystem is independently planned without considering system coupling and energy conversion links. This is also the mainstream mode of current campus planning of the same kind.

And substituting the data into the model to solve the planning model. The programming tool is Matlab R2016a, the OPF power flow calculation tool loaded by the platform is adopted in the OPF power flow calculation, and the main parameters of the operation server are as follows: a CPU: E5-2650V 310 core 20 threads; RAM: 64G; and OS: MS Windows10 Pro.

(2) Comparison analysis of optimized results

The Pareto solution set of the multi-objective optimization of the multi-energy microgrid is obtained by solving the optimization model and is shown in the attached figure 11. As can be seen from the figure, there is no optimization result that simultaneously satisfies low optimization operation cost, low wind and light abandonment power and low pollutant emission, and it can be seen that there is a contradiction relationship between the three objective functions constructed herein, and it is difficult to obtain the optimum at the same time. This is mainly because: 1) if the economic cost of the system needs to be reduced, the output of a coal-fired cogeneration unit needs to be used for replacing the output of a gas-fired unit as much as possible, and meanwhile, a distributed wind power and photovoltaic unit needs not to be invested as much as possible. 2) In order to reduce pollutant emission, wind power, photovoltaic power generation and gas turbine units should output power as much as possible, which will significantly increase the system planning cost, so that no low-cost solution exists in the low pollutant emission area. Therefore, the multi-objective optimization thought adopted in the model built by the method avoids the blindness of integrating a plurality of objective functions into a single-objective optimization thought in a manual weighting mode. In actual planning, a decision maker can select a preference scheme as a planning scheme according to the actual requirements of a project, so that support is provided for practically realizing multi-objective optimization in economic, technical and environmental aspects. According to the previous thought, the result with the lowest total running cost of Pareto solution centralized planning is selected as a planning scheme, and the next analysis is carried out.

The form of the Pareto optimal solution space of the scheme B is similar to that of the figure 11, and is omitted here, but the Pareto front edge of the scheme B is closer to the outer side of paper than the form in the figure 11, which shows that the optimization results of various indexes are inferior to the scheme A. Further according to the method for selecting the optimal solution from the solution space given above, the configuration results of each unit of the system of the schemes a and B are obtained as shown in table 4. The net rack planning result of the scheme a is shown in fig. 12, and the unit configuration condition of each node is shown in table 5.

Table 4 configuration results of multi-energy microgrid system

Type of unit	Parameter(s)	Scheme A	Scheme B
				Distributed gas turbine	Installed capacity/MW	1511	1787
Distributed photovoltaic power generation	Installed capacity/MW	1273	1198
				Distributed wind power	Installed capacity/MW	1008	973
Fuel cell	Installed capacity/kW	120	-
				Electricity storage	Maximum power storage/MWh	563	1206
Heat storage	Maximum heat storage capacity/MWh	3116	0
				Storing hydrogen	Hydrogen storage volume/Nm 3	2018	-
Electrically powered hydrogen generator	Number of equipment/equipment	6	-
				Natural gas hot water boiler	Maximum heating power/MW	855	3600
Electric water-heating boiler	Maximum heating power/MW	1141	-

TABLE 5 scheme A Each node unit configuration

The cost and pollutant emission data for the resulting planning for both types of processes are shown in table 6.

TABLE 6 target function evaluation of each scenario

Comparing table 4 and table 6, it can be seen that compared with the current planning mode (scenario B), scenario a has significant advantages in planning the total cost of operation, and also has less pollutant emissions and is more desirable for clean energy consumption. Therefore, the system planning scheme obtained by the multi-energy micro-grid multi-target collaborative optimization model built by the method is superior to the planning scheme of the existing micro-grid.

(3) Algorithm performance analysis

In order to prove the superiority of the Tent mapping chaos-based improved NSGA-II algorithm, the traditional NSGA-II algorithm and the multi-target PSO algorithm which is widely used in the multi-target field are also used for calculating and solving the model established in the text, and the solving results of the algorithms are shown in the table 7.

TABLE 7 calculation results of the respective algorithms

As can be seen from Table 7, the three-target optimization results obtained by the Tent mapping chaos-based improved NSGA-II algorithm provided by the invention are significantly superior to those of the traditional NSGA-II algorithm and the multi-target PSO algorithm, and the calculation results of the last two algorithms are not very different. The improved NSGA-II algorithm can be distributed in the whole search space in the operation process by introducing a chaos optimization mechanism, so that the defect that the traditional algorithm is easy to fall into local optimum is effectively avoided, and the probability of obtaining the global optimum solution by the algorithm is improved. Considering that the multi-target PSO algorithm of the NSGA-II algorithm has more outstanding optimization performance when solving the multi-target optimization algorithm, the NSGA-II algorithm based on Tent mapping chaos improvement can be considered to have better performance and better application space when solving the multi-target optimization problem compared with most of the existing algorithms.

In summary, the present invention is not limited to the above-mentioned embodiments, and those skilled in the art can propose other embodiments within the technical teaching of the present invention, but these embodiments are included in the scope of the present invention.

Claims

1. A multi-energy microgrid planning method oriented to an integrated energy system is characterized by comprising the following steps:

2. The integrated energy system-oriented multi-energy microgrid planning method of claim 1, wherein the objective functions of the multi-energy microgrid multi-target collaborative planning model comprise a system planning operation cost objective function, a system wind and light abandoning objective function and a system environment cost objective function.

3. The integrated energy system-oriented multi-energy microgrid planning method of claim 2, characterized in that a planning period of the system is assumed to be 20 years, each year includes 365 same scheduling days, so that the total operation cost of the system can be obtained by reflecting all the operation costs of the whole operation period of the system, and the system planning operation cost objective function is expressed as follows:

in the formula, Y represents a gaugeYear Y of the installments; r is interest rate, is used for reflecting the time value of capital, and takes 5 percent; c_plan(1-r)¹⁹Representing the residual value recovery at the end of the planning period, and taking the residual value recovery rate as 5 percent;

C_plan＝C_sup+C_l+C_es (2)

C_operrepresents the system operating costs, including the cost of purchasing mains electricity and operating costs of fuel cells, distributed gas turbines, and natural gas hot water boilers, the cost of purchasing gas from gas companies, and demand response costs, as shown in equation (4):

4. The integrated energy system-oriented multi-energy microgrid planning method of claim 2, characterized in that the wind curtailment at the time t is defined as the difference between the maximum electric quantity generated by the system wind power and photovoltaic set and the actual generated electric quantity under the natural condition at the time, and the objective function of the wind curtailment and the curtailment of the system is expressed as follows:

in the formula (I), the compound is shown in the specification,

and

5. The integrated energy system-oriented multi-energy microgrid planning method of claim 2, characterized in that a system environment cost objective function is expressed as follows:

6. The integrated energy system-oriented multi-energy microgrid planning method of claim 1, wherein the constraint conditions of the multi-energy microgrid multi-target collaborative planning model include a power balance constraint, an electric heating power flow constraint, a unit output constraint, a cogeneration operation constraint, an energy coupling constraint, an energy storage operation constraint, a clean energy installation constraint and a demand response constraint.

7. The comprehensive energy system-oriented multi-energy microgrid planning method of claim 1, characterized in that the key steps of the chaotic optimization multi-objective genetic algorithm are as follows: