A Robust Planning Method for a Comprehensive Park Energy System

Abstract

The present invention relates to a robust planning method for a park integrated energy system, which includes the steps of: constructing a multi-stage planning model considering the construction timing of the PIES and determining the device digital models of energy production, energy conversion, and energy storage devices in the PIES based on modeling the typical physical structure of the PIES; introducing stepped carbon trading in the PIES, determining the free carbon emission allowance of the PIES, and constructing a calculation model for the stepped carbon trading cost; using a box uncertainty set to describe the source-load uncertainty of the PIES and establishing a robust multi-stage planning model of the PIES considering stepped carbon trading and source-load uncertainty based on the multi-stage planning model; performing dual and linearization processing on the robust multi-stage planning model of the PIES to obtain a mixed-integer linear programming model, and solving the mixed-integer linear programming model to obtain a robust planning scheme for the PIES. Considering the planning problem of the PIES construction timing and source-load uncertainty factors can give full play to the advantages of the PIES.

Description

Technical Field

[0001] The present invention relates to the field of integrated energy system planning for parks, and particularly to a robust planning method for an integrated energy system of a park. Background Art

[0002] With the gradual expansion of the global carbon trading market, in order to promote the energy industry to play a leading role in energy conservation and emission reduction, introducing a carbon trading mechanism in the applications of IES (Integrated Energy System) and PIES (Park-level Integrated Energy System) has become an important way to build a green and low-carbon energy system. IES is an energy production, supply and sales integrated system that covers the input and output of various energy sources such as electricity, heat, and gas, combines technologies such as combined heat and power (CHP), power to gas (P2G), and energy storage, realizes the mutual conversion and unified scheduling of different energy sources in the same system, and improves energy efficiency while meeting energy demands. And PIES is a micro integrated energy system for terminal energy users such as buildings, communities, industrial parks or towns. As a coupling object and supply object of various energy forms on the terminal user side, PIES can improve the consumption of renewable energy such as photovoltaic (PV) and wind energy and the energy utilization efficiency, and is the main form of energy consumption on the terminal user side in the future new energy system. Carbon trading is the trading of carbon emission rights, which endows the emission rights of greenhouse gases including carbon dioxide with commodity attributes for the purpose of controlling the total carbon emissions.

[0003] Currently, when introducing a carbon trading mechanism based on PIES, since various loads in the park gradually increase with the park's investment promotion plan, if the PIES is planned and built once only according to the predicted maximum load after the park is built, there will be adverse situations such as over-investment in the initial operation stage, resulting in waste of resources, and the diverse load demands that increase in the later operation stage cannot be met. In addition, there is a strong uncertainty between renewable energy power generation and the diverse loads in the park. When the actual renewable energy output and load deviate far from the predicted values considered in the plan, it will not only greatly affect the feasibility of the PIES planning scheme, but also seriously affect the operation stability of the PIES in severe cases.

[0004] Therefore, how to give full play to the advantages of PIES during the planning process of PIES, and at the same time, on the premise of scientific and reasonable planning, ensure the economic and low-carbon operation of PIES is an urgent problem to be solved. Summary of the Invention

[0005] An embodiment of the present invention provides a robust planning method and device for a park integrated energy system, which can consider the uncertainty factors of the source and load in the PIES, give full play to the advantages of the PIES, and address the planning problem considering the construction time sequence of the PIES during the introduction of the carbon trading mechanism.

[0006] An embodiment of the present invention provides a robust planning method for a park integrated energy system, characterized by including the steps of:

[0007] Construct a multi-stage planning model considering the construction time sequence of the PIES and determine the device digital models of the energy production, energy conversion, and energy storage devices in the PIES based on the modeling of the typical physical structure of the PIES. Among them, the multi-stage planning model is used to decide the candidate device configurations for each planning stage according to the device digital models and the relevance between stages;

[0008] Introduce stepped carbon trading in the PIES, determine the free carbon emission quota of the PIES, and construct a calculation model for the stepped carbon trading cost;

[0009] Use a box uncertainty set to describe the source and load uncertainty of the PIES and establish a robust multi-stage planning model of the PIES considering stepped carbon trading and source and load uncertainty based on the multi-stage planning model;

[0010] Perform dual and linearization processing on the robust multi-stage planning model of the PIES to obtain a mixed-integer linear programming model, and solve the mixed-integer linear programming model to obtain a robust planning scheme for the PIES.

[0011] The beneficial effects brought by the technical solution provided by the present invention include:

[0012] An embodiment of the present invention provides a robust planning method for a park integrated energy system, which takes into account the construction time sequence of the PIES and the uncertainty factors of the source and load in the PIES in the planning problem of the PIES, and considers the synchronous implementation of the construction of the PIES and the development process of the park, which is of great significance for improving the economy of the PIES planning scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0014] Figure 1 It is a schematic flow chart of the robust planning method for the park integrated energy system provided by the embodiment of the present invention;

[0015] Figure 2 A schematic diagram of the physical architecture of a campus integrated energy system provided by an embodiment of the present invention;

[0016] Figure 3 A schematic diagram of a multi-stage planning model provided by an embodiment of the present invention;

[0017] Figure 4 A schematic diagram of a tiered carbon trading mechanism provided by an embodiment of the present invention;

[0018] Figure 5 A schematic diagram of a box uncertainty set of source-load uncertainty variables provided by an embodiment of the present invention;

[0019] Figure 6 A flowchart of a solution based on the C&CG algorithm provided in an embodiment of the present invention;

[0020] Figure 7 The experimental data curve of load and photovoltaic output under a typical day provided by the embodiment of the present invention

[0021] Figure 8 A schematic diagram of actual values of source load simulation on a typical day in the transition season provided by an embodiment of the present invention;

[0022] Figure 9 This is a graph showing the convergence results of the C&CG algorithm provided in an embodiment of the present invention;

[0023] Figure 10 A schematic diagram of the full life cycle cost under different uncertain adjustment parameter values provided by an embodiment of the present invention;

[0024] Figure 11 A schematic diagram of optimization results under different uncertain adjustment parameter values provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0026] like Figure 1 As shown, an embodiment of the present invention provides a robust planning method for a campus integrated energy system, which includes the following steps:

[0027] S100: Build a multi-stage planning model considering the construction timeliness of PIES, and determine the device digital models of energy production, energy conversion, and energy storage devices in PIES based on the modeling of the typical physical architecture of PIES. Among them, the multi-stage planning model is used to decide the candidate device configurations for each planning stage according to the device digital models and the relevance between stages.

[0028] S200: Introduce stepped carbon trading in PIES, determine the free carbon emission quota of PIES, and build a calculation model for stepped carbon trading fees.

[0029] S300: Use a box-type uncertainty set to describe the source-load uncertainty of PIES, and establish a robust multi-stage planning model of PIES considering stepped carbon trading and source-load uncertainty based on the multi-stage planning model.

[0030] S400: Perform dual and linearization processing on the robust multi-stage planning model of PIES to obtain a mixed-integer linear programming model, and solve the mixed-integer linear programming model to obtain the robust planning scheme of PIES.

[0031] It should be noted that the source-load uncertainty includes source uncertainty and load uncertainty. The source uncertainty may include the output of photovoltaic (PV) devices, and the load uncertainty may include electricity, heat, and gas loads.

[0032] In the embodiments of the present invention, the construction timeliness of PIES and the uncertainty factors of the source-load in PIES are considered in the planning problem of PIES. Considering that the construction of PIES is implemented synchronously with the development process of the park is of great significance for improving the economy of the PIES planning scheme.

[0033] In some embodiments, building the multi-stage planning model considering the construction timeliness of PIES in S100 includes the steps of:

[0034] S110: Take the shortest service life of the device to be configured as the value of the planning period of the multi-stage planning model.

[0035] S120: Determine the number of planning stages of the multi-stage planning model according to the change of the maximum load level within the planning period.

[0036] S130: Set constraint conditions so that the solutions of the previous planning stages are associated with the subsequent planning stages, and the decision results of the current planning stage are related to all previous planning stages.

[0037] S140: Take the minimum full life cycle cost as the goal and combine the constraint conditions to decide the candidate device configurations for each planning stage.

[0038] It should be noted that the full life cycle cost includes the investment cost and operation and maintenance cost during the full life cycle.

[0039] Regarding the construction timing of PIES, the embodiments of the present invention adopt a multi-stage planning method, such as Figure 3 shown, P sets the planning period N as the shortest service life of the equipment to be configured, and the number of planning stages K of the multi-stage planning model is determined by the change of the maximum load level during the planning period; the planning model fully considers the relevance between stages, and the planning scheme of the previous stage is connected to the subsequent planning stage through multi-stage constraints, and the decision result of the current planning stage is related to all previous stages; and the candidate equipment configuration of each planning stage is determined with the goal of minimizing the full life cycle cost, where the full life cycle cost includes the investment cost and operation and maintenance cost in the initial investment stage.

[0040] Preferably, the set of planning schemes for the current stage is based on meeting the load growth demand of the current stage, and the candidate equipment for the current stage is configured based on the equipment investment and construction conditions of all previous stages until the equipment configuration of all planning stages is determined;

[0041] such as Figure 2 shown, in some embodiments, in S100, the equipment digital models of the energy production, energy conversion, and energy storage equipment in PIES are determined based on the modeling of the typical physical structure of PIES, including the steps:

[0042] S101: Equivalent PIES to a two-port network covering the input and output of various energy forms based on the energy hub EH model;

[0043] S102: Determine the energy production, energy conversion, and energy storage equipment of PIES. Among them, the energy production equipment is a photovoltaic PV device, the energy conversion equipment includes a power-to-gas P2G device, an electric boiler EB device, a combined heat and power CHP device, and a gas boiler GB device, and the energy storage equipment includes an electricity storage ES device, a thermal storage TS device, and a gas storage GS device; among them, the photovoltaic PV device can be abbreviated as the PV device later, the power-to-gas P2G (Power to Gas) device is abbreviated as the P2G device later, the electric boiler EB (Electric Boiler) device is abbreviated as the EB device later, the combined heat and power CHP (Combined Heat and Power) device is abbreviated as the CHP device later, the gas boiler GB (Gas Boiler) device is abbreviated as the GB device later, the electricity storage ES (Electricity Storage) device is abbreviated as the ES device later, the thermal storage TS (Thermal Storage) device is abbreviated as the TS device later, and the gas storage GS (Gas Storage) device is abbreviated as the GS device later;

[0044] S103: Numerically model the energy production, energy conversion, and energy storage devices of the PIES and obtain the digital models of each device. The digital models of each device include:

[0045] The mathematical model of the PV device, which is expressed as:

[0046] P PV (t) ≤ P PV ,

[0047] where P PV (t) represents the output electric power of the PV device at time t, represents the upper limit of the output electric power of the PV device, which is determined by the rated capacity of the device;

[0048] The mathematical model of the P2G device, which is expressed as:

[0049]

[0050] where G P2G (t), g P2G (t) respectively represent the natural gas power output and the natural gas flow rate output by the P2G device at time t, η P2G represents the energy conversion efficiency of the P2G device, P P2G (t) represents the electric power input to the P2G device at time t; Q represents the lower heating value of natural gas, which can take the value of 9.97 kWh / m 3 , represents the upper limit of the electric power input to the P2G device, which is determined by the rated capacity of the device;

[0051] The mathematical model of the EB device, which is expressed as:

[0052]

[0053] where H EB (t) represents the thermal power output by the EB device at time t, P EB (t) represents the electric power input to the EB device at time t, η EB represents the energy conversion efficiency of the EB device, represents the upper limit of the electric power input to the EB device, which is determined by the rated capacity of the device;

[0054] The mathematical model of the CHP device, which is expressed as:

[0055]

[0056] where H CHP (t), η CHP,h respectively represent the thermal power output by the CHP device at time t and the thermal conversion efficiency of the device, PCHP (t) and η CHP,e respectively represent the electric power output by the CHP device at time t and the electric conversion efficiency of the device. G CHP (t) represents the natural gas power input to the CHP device at time t. Its unit of value is the same as that of the electric power, and it is converted according to the lower heating value Q of natural gas. represents the upper limit of the natural gas power input to the CHP device, which is determined by the rated capacity of the device. ΔG CHP and respectively represent the maximum and minimum ramping rates of the CHP device;

[0057] The mathematical model of the GB device is expressed as:

[0058]

[0059] where H GB (t) represents the thermal power output by the GB device at time t; η GB represents the energy conversion efficiency of the GB device. G GB (t) represents the natural gas power input to the GB device at time t. Its unit of value is the same as that of the electric power, and it is converted according to the lower heating value Q of natural gas. represents the upper limit of the natural gas power input to the GB device, which is determined by the rated capacity of the GB device. ΔG GB and respectively represent the maximum and minimum ramping rates of the GB device;

[0060] The mathematical models of the ES, TS, and HS devices are expressed as:

[0061]

[0062] where s ∈ Ω, and Ω = {ES, TS, GS} represents the set of three types of energy storage devices, ES, TS, and GS. P s cha (t), respectively represent the charging power magnitude and the charging power upper limit of the s-th type of energy storage device at time t. P s dis (t), respectively represent the discharging power magnitude and the discharging power upper limit of the s-th type of energy storage device at time t. is a 0-1 variable. When takes the value of 1, it means that the s-th type of energy storage is charging at time t. takes the value of 0, it means that the s-th type of energy storage is discharging at time t. D s (t) represents the state of charge of the s-th type of energy storage device at time t. respectively represent the charging and discharging efficiencies of the s-th type of energy storage device, Δt is the operation step length, and the value is 1 h. D s respectively represent the s maximum and minimum energy values of the s-th type of energy storage device, D s (t0), D s (t T ) respectively represent the energy states of the s-th type of energy storage device at the beginning and end of the scheduling period.

[0063] In some embodiments, S200 includes the steps of:

[0064] S210: Determine the initial free carbon emission quota of PIES for the carbon emission sources of PIES by using the baseline method. The carbon emission sources of PIES include PIES purchased electricity, CHP equipment, and GB equipment, and the initial free carbon emission quota is:

[0065]

[0066] wherein, E * represents the free carbon emission quota of PIES, and respectively represent the free carbon emission quotas corresponding to PIES purchased electricity, CHP equipment, and GB equipment. T represents the settlement period of carbon trading fees (the settlement period of carbon trading fees is one year), and T = 8760, represents the free carbon emission quota per unit of electricity purchased, P PN (t) represents the power purchase rate of PIES from the superior power grid at time t, represents the free carbon emission amount per unit of heat, represents the conversion coefficient of CHP equipment power generation to heat supply;

[0067] S220: Determine the actual carbon emissions of PIES, and the actual carbon emissions are expressed as:

[0068]

[0069] wherein, E represents the total actual carbon emissions of PIES, E PN 、E CHP and E GB respectively represent the actual carbon emissions corresponding to PIES purchased electricity, CHP equipment, and GB equipment. E P2G represents the CO2 consumption of the P2G equipment, β g represents the carbon consumption coefficient, which represents the amount of CO2 required for the P2G equipment to achieve unit power conversion;

[0070] It should be noted that among them, E PN 、E CHPand E GB The calculation methods of the three are basically the same as those corresponding to calculating the free carbon emission quota in step S210 ( and ), and the difference lies in the carbon emission coefficients β e corresponding to the purchased electricity and β h corresponding to the heat, which are different from the values of the free quota coefficient .

[0071] S230: Construct a calculation model for the stepped carbon trading fee, and the calculation model for the stepped carbon trading fee is expressed as:

[0072]

[0073] Among them, as Figure 4 shown, C co represents the stepped carbon trading fee of PIES, c represents the carbon trading benchmark price, α represents the price growth coefficient, d represents the price interval length, and the number of price intervals corresponds to the number of segments of C co .

[0074] As Figure 5 shown, in some embodiments, in S300, a box - type uncertainty set is used to describe the source - load uncertainty of PIES, including the steps:

[0075] S310: Use a box - type uncertainty set to describe the source uncertainty and load uncertainty of PIES, and it is expressed as:

[0076]

[0077] Among them, U represents the uncertainty set, u represents the set of uncertain variables within the planning period, which is composed of the actual PV output P PV,n (t), the actual electrical load P L,n (t), the actual heat load H L,n (t), and the actual gas load G L,n (t) at each moment of each year. The superscript ^ represents the predicted value, Δu represents the maximum error fluctuation range allowed for each uncertain variable within the planning period, and δ PV , δ PL , δ GL , δ GL respectively represent the allowable fluctuation deviation coefficients of PV output, electrical load, heat load, and gas load;

[0078] S320: Define the uncertainty adjustment parameter Γ of the uncertainty set U as:

[0079]

[0080]

[0081]

[0082]

[0083] Among them, Γ PV 、Γ PL 、Γ HL 、Γ GL respectively represent the uncertain adjustment parameters of PV output, electrical load, heat load, and gas load, and respectively characterize the maximum number of years in the planning period during which the actual values of PV output, electrical load, heat load, and gas load are allowed to deviate from the predicted values; and

[0084] For the PV output, if z PV,n = 0, it means that the actual PV output value at all times in the nth year of the planning period is the predicted value, and z PV,n = 1 means that the actual PV output value in the nth year of the planning period deviates from the predicted value;

[0085] For the electrical load, if z PL,n = 0, it means that the actual electrical load value at all times in the nth year of the planning period is the predicted value, and z PL,n = 1, which means that the actual electrical load value in the nth year of the planning period deviates from the predicted value;

[0086] For the heat load, if z HL,n = 0, it means that the actual heat load value at all times in the nth year of the planning period is the predicted value, and z HL,n = 1, which means that the actual heat load value in the nth year of the planning period deviates from the predicted value;

[0087] For the gas load, if z GL,n = 0, it means that the actual gas load value at all times in the nth year of the planning period is the predicted value, and z GL,n = 1, which means that the actual gas load value in the nth year of the planning period deviates from the predicted value.

[0088] In some embodiments, in S300, a PIES robust multi-stage planning model considering stepped carbon trading and source-load uncertainty is established based on the multi-stage planning model, including the steps of:

[0089] S330: Divide the multi-stage planning model into a first-stage problem model and a second-stage problem model, where the first-stage problem model is used to make decisions on the investment of equipment, and the second-stage problem model is used to make decisions on the operation of equipment;

[0090] S340: Set the objective functions for the first-stage problem model and the second-stage problem model respectively, and the objective functions consider minimizing the life-cycle cost, and the life-cycle cost consists of investment cost, equipment salvage value, operation cost, and maintenance cost. The objective function can be expressed as:

[0091]

[0092] Among them, f1(·) represents the objective function of the first-stage problem model, which consists of the investment cost C inv,k , equipment salvage value C R . f2(·) represents the objective function of the second-stage problem model, which consists of the operation cost C ope,n , maintenance cost C main,n , and carbon trading cost C co,n . x and y represent the optimization variable sets of the first-stage problem model and the second-stage problem model respectively. Ξ(x,u) represents the feasible region of y after making decisions x and u. K represents the number of planning stages, k represents the k-th planning stage, N represents the number of years in the planning period, n represents the n-th year in the planning period, and n k represents that the first year of the k-th planning stage is the n k -th year of the planning period. γ represents the discount rate;

[0093] The calculation method of the investment cost C inv,k is as follows:

[0094]

[0095] Among them, c PV represents the investment cost per unit capacity of PV, W PV,k represents the capacity configuration of PV equipment in the k-th planning stage. Π = {P2G, EB, CHP, GB} and Ω = {ES, TS, GS} represent the sets of energy conversion equipment and energy storage equipment respectively. c i , c s represent the investment cost per unit capacity of the i-th type of energy conversion equipment and the s-th type of energy storage equipment respectively. W i,k , W s,k represent the capacity configurations of the i-th type of energy conversion equipment and the s-th type of energy storage equipment in the k-th planning stage respectively;

[0096] The calculation method of the equipment salvage value C R is as follows:

[0097]

[0098] Among them, c dep,j represents the average annual depreciation cost of the equipment whose service life has not ended at the end of the j-th planning period. C inv,jDenote the sum of the investment costs of the j-th type of equipment in all stages, δ j Denote the net salvage value rate of the j-th type of equipment, Y j Denote the lifespan of the j-th type of equipment, N j Denote the number of years of use of the j-th type of equipment from its construction to the end of the planning period;

[0099] The operating cost C ope,n is calculated as follows:

[0100]

[0101] where c e (t), c g (t) denote the electricity price and gas price at time t respectively; P PN,n (t) denotes the power purchase amount of PIES from the superior power grid at time t of the n-th year, G GN,n (t) denotes the gas purchase amount of PIES from the superior gas grid at time t of the n-th year;

[0102] The maintenance cost C main,n is calculated as follows:

[0103]

[0104] where O represents the unit maintenance cost matrix of various types of equipment, P n (t) denotes the output power matrix of various types of equipment at time t of the n-th year, P n T (t) denotes the transpose of matrix P n (t), P PV,n (t) denotes the electric power output by the PV equipment at time t of the n-th year, respectively denote the charging power and discharging power of the gas storage device at time t of the n-th year;

[0105] where the carbon trading cost C CO,n is calculated in the same way as the stepped carbon trading cost C co ;

[0106] S350: Set the constraints at the investment level for the problem model in the first stage, and the constraints at the investment level are expressed as:

[0107]

[0108] where, W i max and W s max respectively denote the maximum capacity values allowed for the construction of PV equipment, the i-th type of energy conversion equipment, and the s-th type of energy storage equipment;

[0109] S360: Set the operation - level constraints for the problem model in the second stage. The operation - level constraints include equipment operation constraints, electric - heat - gas balance constraints, PIES's power exchange constraints with the upper - level power grid and gas grid, and carbon trading constraints. Among them,

[0110] The equipment operation constraints include the equipment digital models of energy production, energy conversion, and energy storage equipment in PIES and multi - stage constraint conditions. The multi - stage constraint conditions are:

[0111]

[0112]

[0113]

[0114] Among them, P PV,n (t) represents the electric power output by the PV equipment at time t in the nth year. ∑W PV represents the cumulative investment capacity of the PV equipment in the current k planning stages. n ∈ k means that the nth year of the planning period belongs to the kth planning stage. P i,n (t) represents the power input into the ith type of energy - conversion equipment at time t in the nth year. ∑W i represents the cumulative investment capacity of the ith type of energy - conversion equipment in the current k planning stages, represents the maximum allowable state of charge of the sth type of energy - storage equipment in the nth year. ∑W s represents the cumulative investment capacity of the sth type of energy - storage equipment in the current k planning stages;

[0115] The electric - heat - gas balance constraints include electric - power balance constraints, heat - power balance, and gas - power balance. And

[0116] The electric - power balance constraint is expressed as:

[0117]

[0118] The heat - power balance is expressed as:

[0119]

[0120] The gas - power balance is expressed as:

[0121]

[0122] Among them, P CHP,n (t) represents the electric power output by the CHP equipment at time t in the nth year. P P2G,n (t) represents the electric power input into the P2G equipment at time t in the nth year. P EB,n(t) represents the electric power input to the EB device at time t in the nth year, P L,n (t), H L,n (t) and G L,n (t) respectively represent the electric, heat, and gas load powers of the PIES at time t in the nth year. And respectively represent the charging powers of the ES, TS, and GS devices at time t in the nth year. And respectively represent the discharging powers of the ES, TS, and GS devices at time t in the nth year; H EB,n (t), H CHP,n (t) and H GB,n (t) respectively represent the heat powers output by the EB, CHP, and GB devices at time t in the nth year, G P2G,n (t) represents the natural gas power output by the P2G device at time t in the nth year, G CHP,n (t), G GB,n (t) respectively represent the natural gas powers input to the CHP and GB devices at time t in the nth year;

[0123] The power exchange constraints between the PIES and the superior power grid and natural gas grid are expressed as:

[0124]

[0125] Among them, respectively represent the minimum and maximum values of the allowable externally purchased electric power of the PIES, respectively represent the minimum and maximum values of the allowable externally purchased natural gas power;

[0126] It can be understood that the robust multi-stage planning model of the PIES described in S300 is composed of a first-stage problem model and a second-stage problem model, and is determined based on the objective function, the constraints at the investment level, and the constraints at the operation level.

[0127] In some embodiments, the robust multi-stage planning model of the PIES in S400 is dualized and linearized to obtain a mixed-integer linear programming model, including the steps of:

[0128] S410: Set And add an auxiliary variable β s (t), and replace the bilinear terms And in the device digital model with the auxiliary variable β s (t), and add a first constraint condition, and the first constraint condition is:

[0129]

[0130] Where M is a positive real number greater than a preset threshold, which can be set to 10 5 .

[0131] S420: For the carbon trading amount at the m endpoints in the piecewise linear function (i.e., the calculation model of the step carbon trading fee), the values a1, a2, ..., a m And the tiered carbon trading fee values b1, b2, ..., b m , introduce m-1 auxiliary variables (0-1 variables) x1, x2, ..., x m-1 and m auxiliary continuous variables y1,y2,…,y m , and satisfy the second constraint, which is:

[0132]

[0133] And for any carbon trading amount a∈[a1,a m ], and its carbon trading fee b is expressed as:

[0134]

[0135] It is understandable that S410 and S420 are bilinear terms for energy storage charge and discharge constraints. and and the tiered carbon trading fee C co and carbon trading volume EE * The piecewise linear function relationship is linearized based on the Big-M method and the concept of piecewise linear function.

[0136] S430: Based on the first constraint and the second constraint, the PIES robust multi-stage planning model is expressed as:

[0137]

[0138] in,

[0139] x is the decision variable of the first-stage problem model, which is composed of the candidate equipment investment capacity in each planning stage: photovoltaic W PV,k 、Various energy conversion equipment W i,k 、Various energy storage devices W s,k , the charge and discharge state variables of energy storage equipment at each moment each year and auxiliary variables of the piecewise function of annual carbon trading fees Composition, W PV,k 、W i,k 、W s,k is an integer variable, When the value is 1, it means that the s-th type of energy storage is charging at time t in the n-th year, and when the value is 0, it means discharging. is a vector matrix, which contains m - 1 auxiliary variables, and among them When the value is 1, it means that the carbon trading volume in the n-th year is in the interval [a m-1 , a m , and when the value is 0, it means that the carbon trading volume is in the interval [a1, a m-1 ).

[0140] y is the decision variable of the second-stage problem model, which is composed of the purchased electricity P PN,n (t), the purchased natural gas G GN,n (t), the actual PV output P PV,n (t), the input power P i,n (t) of various energy conversion devices, the charging and discharging power of various energy storage devices the state of charge D s,n (t) of various energy storage devices, the actual electro-thermal-gas load P L,n (t), H L,n (t), G L,n (t) and the auxiliary variable of the annual carbon trading cost piecewise function and each variable is a continuous variable. A, B, C, D, E, F, G, H, I, c, d, e, f all represent coefficient matrices and are independent of the decision variables.

[0141] S440: Dualize the second-stage problem model in the PIES robust multi-stage planning model, and the dual process includes:

[0142] S441: Set the variable x * of the first-stage problem model, so that x * is a constant for the max-min problem of the second-stage problem model of the PIES robust multi-stage planning model, and represent the max-min problem as:

[0143]

[0144] Among them, λ1, λ2, π represent the set of dual variables corresponding to the constraint conditions (the s.t. part of the above formula) of the max-min problem, and λ1, λ2, π are all continuous variables;

[0145] S442: By using the strong duality theory, dualize the inner-layer min problem into a max problem, and convert the max-min problem into a single-layer max problem. The single-layer max problem is expressed as:

[0146]

[0147] S443: For the bilinear term u T π in the single-layer max problem, change the uncertainty set U, and represent the changed uncertainty set U as:

[0148]

[0149] where Δu·X represents the dot product, and X consists of 0-1 variables and can be expressed as:

[0150]

[0151] where x PV,n , x PL,n , x HL,n , x GL,n are all 0-1 variables. When their values are 1, they respectively represent that the actual output of PV takes the lower boundary of the fluctuation range at all times in the nth year of the planning period, and the actual electricity, heat, and gas loads take the upper boundary of the fluctuation range at all times in the nth year of the planning period;

[0152] It can be understood that when x PV,n , x PL,n , x HL,n , x GL,n these 0-1 variables take the value of 0, they respectively represent that the actual output of PV and the actual electricity, heat, and gas loads take their predicted values at all times in the nth year of the planning period without fluctuations.

[0153] By changing the uncertainty set U, the original bilinear term u T π becomes There is still a bilinear term X·π with the product of 0-1 variables and continuous variables; therefore, further enter the linearization process in S444.

[0154] S444: Based on the Big-M method, linearize the bilinear term X·π, and the linearization process includes:

[0155] Introduce an auxiliary variable β u , and replace all X·π in the single-layer max problem with β u , and the single-layer max problem is equivalent to:

[0156]

[0157] where Γ represents the set of uncertainty adjustment parameters.

[0158] It can be understood that in S400, finally, the second-stage problem model determined by the equivalent single-layer max problem is used as the mixed-integer linear programming model.

[0159] In some embodiments, in S400, the C&CG algorithm is adopted and a commercial solver is called to solve the mixed-integer linear programming model to obtain a robust planning scheme for PIES. As Figure 6 shown, the solution process of the C&CG algorithm is as follows:

[0160] S01: Initialize the parameters of the C&CG algorithm, including setting the iteration number κ = 0, the set of iteration numbers for generating the optimal cut (if an optimal cut needs to be generated after solving the sub-problem in the κ-th iteration, then Θ = Θ ∪ {κ + 1}, otherwise Θ remains unchanged), the upper bound UB = +∞, the lower bound LB = -∞, the allowable error coefficient ε = 0.001, and a set of initial values X of the uncertain variables are given * , and determine

[0161] S02: Solve the master problem according to u * and update the decision result: x * = x * {κ}, y * = y * {κ}, and update the lower bound LB = A T x * {κ}+η * {κ}, and the master problem is expressed as:

[0162]

[0163] where η represents the sub-problem auxiliary variable, and y{·} represents the new two-stage variable introduced by the cut set; x * {·} is the set of solutions obtained after solving the master problem in step S02, not a new variable.

[0164] S03: Solve the sub-problem according to x * (at this time, it is the second-stage problem after linearization and dual processing), let θ(x * ) represent the optimal solution of the sub-problem objective function, and judge whether the sub-problem has a feasible solution under the current x * . If there is no feasible solution, then let θ(x * ) = +∞, and u * remains unchanged; if there is a feasible solution, then update the decision result: X * = X * {κ}, u * = u * {κ}, π * = π * {κ}; and update the upper bound after the judgment as: UB = min{UB, A T x * + θ(x *)};

[0165] S04: If the upper bound and the lower bound satisfy the convergence condition |UB - LB| / LB ≤ ε, then the solution results x*, y * , u* are obtained and the C&CG algorithm loop is exited. Otherwise, a new set of decision variables y{κ + 1} for the second-stage problem model is generated, and feasible cuts FCs and optimal cuts OCs are generated, expressed as:

[0166]

[0167] S05: If the sub-problem in step S03 has no feasible solution, then let κ = κ + 1, and only the feasible cuts FCs are fed back to the master problem, and the feasible cuts FCs are added to the constraint conditions of the master problem and then return to step S02; if the sub-problem in step S03 has a feasible solution, then let Θ = Θ ∪ {κ + 1}, κ = κ + x, and then the feasible cuts FCs and the optimal cuts OCs are fed back to the master problem, and the feasible cuts FCs and the optimal cuts OCs are added to the constraint conditions of the master problem and then return to step S02.

[0168] It can be understood that the C&CG algorithm decomposes the original problem into a master problem and a sub-problem. The master problem corresponds to the first-stage problem, and the sub-problem corresponds to the second-stage problem.

[0169] In some embodiments, after obtaining the robust planning scheme of the PIES, the collaborative optimization effect of new energy and multi-energy storage in the PIES is evaluated, specifically including:

[0170] An integrated energy efficiency evaluation model of the PIES is established, and the integrated energy efficiency evaluation model is expressed as:

[0171]

[0172] where η PIES represents the integrated energy efficiency of the PIES, P L , H L , G L respectively represent the electricity, heat, and gas loads of the PIES, respectively represent the charging powers of the ES device, the TS device, and the GS device, respectively represent the charging efficiencies of the ES device, the TS device, and the GS device, respectively represent the discharging powers of the ES device, the TS device, and the GS device, P PN , G GN respectively represent the electricity purchase power and the gas purchase power of the PIES;

[0173] Based on the integrated energy efficiency evaluation model, the collaborative optimization effect of new energy and multi-energy storage is evaluated.

[0174] Such asFigure 7 , 8 As shown in 8 , in a specific embodiment, the uncertain adjustment parameters of the source and load are both 5, and the fluctuation coefficients describing the fluctuation ranges of the uncertain sets are 0.2, 0.1, 0.1, and 0.1 respectively; the investment parameters and operation parameters of each planned candidate device are shown in Table 1; among various types of devices, the ES device has the shortest service life of 15 years, so it is determined as the planning cycle number of the multi-stage planning. The 15-year planning cycle is divided into 3 planning stages, and the equipment investment and construction in each stage are carried out at the beginning of the 1st year, the beginning of the 4th year, and the beginning of the 9th year respectively. The maximum load levels in each planning stage are shown in Table 2; economic parameters such as time-of-use electricity price, natural gas price, discount rate, and equipment net salvage rate are shown in Table 3; the relevant parameters of the carbon trading mechanism are shown in Table 4.

[0175] Table 1 Parameters of Various Candidate Devices

[0176]

[0177] Table 2 Maximum Load in Each Planning Stage

[0178]

[0179]

[0180] Table 3 Planning Economic Parameters

[0181]

[0182] Table 4 Carbon Trading Parameters

[0183]

[0184] To verify the effectiveness of solving in the Matlab environment by invoking the Gurobi solver in the embodiment of the present invention, the following four scenarios are selected for comparative analysis:

[0185] Scenario Z1: Considering the stepped carbon trading mechanism, the uncertain adjustment parameters of PV output, electricity / heat / gas load are 0, 0, 0, 0 respectively, that is, without considering uncertainty, considering the construction timeliness of the park integrated energy system, and conducting multi-stage planning according to the predicted values;

[0186] Scenario Z2: Considering the stepped carbon trading mechanism, considering the source-load uncertainty, without considering the park construction timeliness, and conducting robust single-stage planning;

[0187] Scenario Z3: Considering the traditional carbon trading mechanism, considering the source-load uncertainty, considering the park construction timeliness, and conducting robust multi-stage planning;

[0188] Scenario Z4: Considering the stepped carbon trading mechanism, considering the source-load uncertainty, considering the park construction timeliness, and conducting robust multi-stage planning.

[0189] The convergence process of the C&CG algorithm under Scenario Z4 is as follows Figure 9 shown.

[0190] From Figure 9 the results shown in it, it can be seen that the algorithm converges after the 4th generation of iteration. At this time, the relative error |UB - LB| / LB between the upper and lower bounds is 0.00096, which meets the convergence condition |UB - LB| / LB ≤ 0.001 of the algorithm. Its convergence speed is relatively fast, and the effectiveness of the C&CG algorithm for solving the constructed robust multi-stage planning model can be verified.

[0191] The planning schemes for the four scenarios are shown in Table 5, and the optimization results for the four scenarios are shown in Table 6. Generally speaking, P2G equipment and GS equipment are not planned in all four scenarios. The comparative analysis of each scenario is as follows

[0192] Table 5 Planning Schemes for Each Scenario (Unit: kW)

[0193]

[0194] Table 6 Optimization Results for Each Scenario

[0195]

[0196] (1) Comparative Analysis between Scenario Z1 and Scenario Z4

[0197] Both Scenario Z1 and Scenario Z4 adopt the stepped carbon trading mechanism and the multi-stage planning method. The difference is that Scenario Z1 is a deterministic planning, while Scenario Z4 is a robust planning. In terms of the planning scheme, the configurations of EB, TS, GB, and CHP in Scenario Z1 and Scenario Z4 are not very different, but the configurations of ES and PV in Scenario Z4 are higher than those in Scenario Z1. Therefore, the investment cost and equipment salvage value of Scenario Z4 are correspondingly higher than those of Scenario Z1. At the same time, considering the source-load uncertainty, in the worst-case scenario, Scenario Z4 will inevitably purchase more electricity and natural gas to cope with the reduction of PV output and the increase of load. As a result, the operating cost, carbon emissions, and carbon trading cost of Scenario Z4 will also be correspondingly higher than those of Scenario Z1.

[0198] (2) Comparative Analysis between Scenario Z2 and Scenario Z4

[0199] Both Scenario Z2 and Scenario Z4 consider the stepped carbon trading mechanism and robust planning. The difference is that Scenario Z2 is a single-stage planning, while Scenario Z4 is a multi-stage planning. In terms of the planning scheme, the configurations of EB, ES, and TS in Scenario Z2 are slightly higher than those in Scenario Z4. There are significant differences in the configurations of GB, CHP, and PV between Scenario Z2 and Scenario Z4. Generally speaking, the capacity configuration in Scenario Z4 is higher than that in Scenario Z2. Therefore, the investment cost, equipment salvage value, and equipment maintenance in Scenario Z4 will be correspondingly higher. However, the life cycle cost in Scenario Z2 is 866,600 yuan higher than that in Scenario Z4, with an increase rate of 2.37%. Moreover, the carbon emissions in Scenario Z2 are also 3,621.17 tons more than those in Scenario Z4, with an increase rate of 9.08%. It can be seen that the multi-stage planning scheme not only has better economy, but also has relatively better low-carbon performance.

[0200] (3) Comparative analysis between Scenario Z3 and Scenario Z4

[0201] Both Scenario Z3 and Scenario Z4 are robust multi-stage models. The difference is that Scenario Z3 adopts the traditional carbon trading mechanism with a fixed trading price, while Scenario Z4 adopts the stepped carbon trading mechanism. In terms of the planning scheme, the configurations of EB and TS in the two scenarios are close. The configurations of ES and GB in Scenario Z3 are higher than those in Scenario Z4, while the configurations of CHP and PV are lower. Since Scenario Z4 configures higher-capacity CHP and PV, its investment cost and life cycle cost are both higher than those in Scenario Z3. The increase rate of the life cycle cost is 0.42%. However, the carbon trading cost and carbon emissions in Scenario Z3 are 163,200 yuan and 2,041.57 tons higher than those in Scenario Z4 respectively, with increase rates of 19.95% and 5.12% respectively, both of which are greater than the increase rate of the life cycle cost. Thus, it can be seen the advantage of the stepped carbon trading mechanism over the traditional carbon trading mechanism in balancing economy and low-carbon performance.

[0202] The most severe scenario obtained by solving Scenario Z4 is as follows: the PV output takes the lower boundary of the value fluctuation range in the 4th, 5th, 6th, 9th, and 10th years of the planning period; the electric load takes the upper boundary of the value fluctuation range in the 4th, 5th, 6th, 9th, and 10th years of the planning period; the heat load takes the upper boundary of the value fluctuation range in the 1st, 4th, 5th, 9th, and 10th years of the planning period; the gas load takes the upper boundary of the value fluctuation range in the 4th, 5th, 6th, 7th, and 9th years of the planning period. In order to explore the effectiveness of the most severe scenario, the present invention considers the stepped carbon trading mechanism and the multi-stage planning method, and further sets the following scenarios and conducts deterministic planning according to the following scenarios:

[0203] Case1: Randomly select the number of planning years when the PV output takes the lower boundary, and the values of the three types of loads are the same as those in the most severe scenario;

[0204] Case2: Randomly select the number of planning years when the electric load takes the upper boundary, and the PV output and the heat / gas load values are the same as those in the most severe scenario;

[0205] Case 3: Randomly select the number of planning years with the heat load taking the upper boundary value. The PV output and the electricity / gas load values are the same as those in the worst-case scenario.

[0206] Case 4: Randomly select the number of planning years with the gas load taking the upper boundary value. The PV output and the electricity / heat load values are the same as those in the worst-case scenario.

[0207] The optimization results for each case are shown in Table 7. The life-cycle costs, carbon trading costs, and carbon trading volumes under Case 1–Case 4 are all lower than those in Case Z4. This optimization result can verify the effectiveness of the worst-case scenario found by the robust planning model.

[0208] Table 7 Optimization results for each case

[0209]

[0210] In addition, for Case Z1 and Case Z4, according to the optimization results shown in Table 7, although the life-cycle cost and carbon emissions of Case Z1 using deterministic planning are lower than those of Case Z4 using robust planning, such a result does not necessarily mean that the planning scheme of Case Z1 is better than that of Case Z4. During actual operation, if the PV output fluctuates below the predicted value and the electricity / heat / gas load fluctuates above the predicted value, the PIES under Case Z1 cannot operate according to the deterministic plan. It is necessary to increase the purchased electricity and natural gas, and in severe cases, load shedding is required to ensure the stable operation of the system. From this perspective, the robust planning scheme has better robustness and the ability to cope with fluctuations.

[0211] Therefore, the present invention further conducts simulation operation analysis on the planning schemes of Case Z1 and Case Z4. Taking the typical day in the transition season of the 9th year of the planning period as an example, according to the Figure 8 five groups of actual PV output and actual electricity / heat / gas load as shown, the system operation conditions under the deterministic planning scheme and the robust planning scheme are simulated. The operation comparison under each group of simulated actual data is shown in Table 8.

[0212] Table 8 Simulation operation comparison based on the typical day in the transition season (costs not discounted)

[0213]

[0214] According to Tables 6 and 7, the life cycle cost, investment cost, operation cost, and carbon emissions of Scenario Z4 with robust planning are all higher than those of Scenario Z1 with deterministic planning. However, according to Table 8, when simulating the actual operation of PIES, the operation cost and carbon emissions of Scenario Z4 are lower than those of Scenario Z1. This is because the robust planning scheme can cope with the most adverse scenarios of PV output and electricity / heat / gas loads. The actual PV output and electricity / heat / gas loads must be better than the most adverse scenarios. At this time, the externally purchased electricity and natural gas of the system will surely decrease accordingly, and thus the operation cost and carbon emissions will also be lower. The deterministic planning scheme corresponds to the predicted values of PV output and electricity / heat / gas loads, and the capacity configuration of PV is also lower than that of the robust planning scheme. When the actual values fluctuate and deviate from the predicted values, the system under the deterministic planning scheme cannot cope with the fluctuations according to the predicted values and can only purchase more electricity and natural gas externally compared with the robust planning scheme. Therefore, the operation cost and carbon emissions of Scenario Z1 are higher than those of Scenario Z4. In addition, the above analysis of the simulated operation does not consider the costs of curtailment of light, load shedding, and electricity sales revenue. In actual operation, considering the additional penalty costs of curtailment of light and load shedding under the deterministic planning scheme and the electricity sales revenue under the robust planning scheme, the operation costs of the two types of planning schemes will differ even more. It can be seen from this that the robust planning scheme has stronger robustness and stronger ability to resist the fluctuations of uncertain variables.

[0215] The uncertain adjustment parameter determines the maximum number of years in which the PV output and electricity / heat / gas loads may fluctuate. The larger its value, the worse the fluctuations of the source-load uncertain variables, and the more conservative the obtained planning scheme. The present invention further explores the influence of the uncertain adjustment parameter on the PIES planning scheme. The life cycle costs under different uncertain adjustment parameters are as Figure 10 shown, where Γ PV is the uncertain adjustment parameter of the PV output, and Γ<X load is the load uncertain adjustment parameter. The results shown in Figure 10 are based on the same values of the uncertain adjustment parameters for the three types of electricity / heat / gas loads.

[0216] As can be seen from Figure 10 , the life cycle of the system increases with the increase of the uncertain adjustment parameter. This is because the increase of the uncertain adjustment parameter means that the number of years of the lower boundary of the PV output value and the upper boundary of the electricity / heat / gas load value within the planning period is increased, and the range of the most adverse scenarios that can be taken becomes larger. In order to cope with more adverse scenarios, it is necessary to increase the capacity configuration of PIES, and more electricity and natural gas need to be purchased during system operation. Therefore, the life cycle cost also increases accordingly.

[0217] In addition, for the case of considering source-load uncertainty simultaneously and considering source and load uncertainties separately, the present invention further explores the influence of different uncertain factors:

[0218] Uncertain regulation parameter value scheme 0: Γ PV = 0, Γ load = 0;

[0219] Uncertain regulation parameter value scheme 1: Γ PV = 5, Γ load = 0;

[0220] Uncertain regulation parameter value scheme 2: Γ PV = 0, Γ load = 5;

[0221] Uncertain regulation parameter value scheme 3: Γ PV = 5, Γ load = 5.

[0222] The planning schemes and optimization results under different source-load uncertain regulation parameter value schemes are shown in Table 9 and Figure 11 . From Table 9 and Figure 11 it can be seen that compared with scheme 0 without considering uncertainty, scheme 1 that only considers the uncertainty of PV output has more PV planned than scheme 0, and the configurations of the other equipment are the same. The life-cycle cost and carbon emissions of scheme 1 are 1.35% and 0.51% higher than those of scheme 0 respectively; compared with scheme 3 that considers both source-load uncertainty, the planning scheme of scheme 3 tends to plan more ES, PV, and CHP and less GB, and the configurations of the other equipment are close under the two schemes. Therefore, the investment cost under scheme 3 is higher than that of value scheme 1. At the same time, because value scheme 3 considers more load uncertainty, it will purchase more electricity and natural gas to cope with load fluctuations, so its operating cost, carbon trading cost, and carbon emissions are also relatively high. The life-cycle cost and carbon emissions of scheme 3 are 4.16% and 2.78% higher than those of scheme 1 respectively.

[0223] Table 9 Planning schemes under different uncertain regulation parameters

[0224]

[0225] Compared with Scenario 0 that does not consider uncertainty, for Scenario 2 that only considers load uncertainty, except for the configuration of CHP being lower than that of Scenario 0, the configurations of the remaining equipment in Scenario 2 are higher than those in Scenario 0. Therefore, its investment cost will be correspondingly higher. At the same time, to cope with load fluctuations, Scenario 2 will purchase more electricity and natural gas externally, resulting in higher operating costs and carbon emissions. The life-cycle cost and carbon emissions of Scenario 2 are 3.96% and 1.62% higher than those of Scenario 0 respectively. For Scenario 2 that only considers load uncertainty and Scenario 3 that considers both source and load uncertainties, the differences between the two in the planning scenarios are as follows: the configurations of EB, TS, and GB in Scenario 2 are higher than those in Scenario 3, and the configurations of the remaining equipment are lower than those in Scenario 3. The differences in various costs and carbon emissions between the two are also smaller than the differences between Scenario 1 and Scenario 3. The life-cycle cost and carbon emissions of Scenario 3 are only 1.55% and 1.65% higher than those of Scenario 2 respectively.

[0226] The present invention further conducts a comparative analysis on the comprehensive energy efficiency and life-cycle cost of PIES to evaluate the collaborative optimization effect of new energy and diversified energy storage. The comparison of the life-cycle cost and the comprehensive energy efficiency on a typical day in the transition season of the 15th year of the planning period under different configurations of PV equipment and energy storage equipment is shown in Table 10.

[0227] Table 10 Comparison of life-cycle cost and comprehensive energy efficiency under different equipment configurations

[0228]

[0229] As can be seen from Table 10, compared with not considering the participation of PV and energy storage in the planning, when only considering the participation of PV in the planning, the life-cycle cost is reduced and the comprehensive energy efficiency is improved; when only considering the participation of energy storage in the planning, although the life-cycle cost is reduced to some extent, its comprehensive energy efficiency is not improved because the charge-discharge efficiency of the energy storage is less than 100%. When considering the participation of both PV and energy storage in the planning, both economy and energy efficiency are improved, and the energy storage device can effectively relieve the energy supply pressure. Thus, it can be shown that the collaborative optimization of renewable energy power generation equipment and energy storage equipment participating in the planning is effective in improving the economy and energy efficiency of the planning scenario.

[0230] Further analysis of the above results:

[0231] (1) The stepped carbon trading mechanism has a better carbon control effect than the traditional carbon trading mechanism, and the multi-stage planning method has better economy and low carbon characteristics than the single-stage planning method;

[0232] (2) The C&CG algorithm is used to solve the proposed two-stage robust planning model. The algorithm converges after 4 iterations, and the convergence speed is relatively fast. By solving the model, the obtained worst-case scenario is verified to be effective, and the PIES planning scenario can cope with the worst-case scenario of uncertainty.

[0233] (3) Although the robust planning scheme has a relatively higher life-cycle cost compared with the deterministic planning scheme, when simulating the actual operation of PIES, the actual operation cost and actual carbon emissions under the deterministic planning scheme are both higher than those of the robust planning scheme, and the robust planning scheme has stronger robustness and the ability to cope with fluctuations;

[0234] (4) The larger the value of the uncertain adjustment parameter, the greater the life-cycle cost of the obtained PIES planning scheme. The planning scheme enhances its conservatism by sacrificing economy. By adjusting the value of the uncertain adjustment parameter, the conservatism of the planning scheme can be flexibly adjusted;

[0235] (5) By comparing the comprehensive energy efficiency and life-cycle cost of PIES, the collaborative optimization effect of new energy and diversified energy storage can be illustrated.

[0236] It should be noted that in the present invention, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the said element.

[0237] The above are only specific embodiments of the present invention, which enable those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features claimed herein.

Claims

1. A robust planning method for a comprehensive energy system in a park, characterized in that, It includes the steps of: Constructing a multi-stage planning model considering the construction timeliness of PIES, and determining the device digital models of energy production, energy conversion, and energy storage devices in PIES based on the modeling of the typical physical structure of PIES. Among them, the multi-stage planning model is used to decide the candidate device configurations for each planning stage according to the device digital models and the relevance between stages; Introducing stepped carbon trading in PIES, determining the free carbon emission quota of PIES, and constructing a calculation model for stepped carbon trading fees; Describing the source-load uncertainty of PIES using a box-type uncertainty set, and establishing a robust multi-stage planning model of PIES considering stepped carbon trading and source-load uncertainty based on the multi-stage planning model; Performing dual and linearization processing on the robust multi-stage planning model of PIES to obtain a mixed-integer linear programming model, and solving the mixed-integer linear programming model to obtain the robust planning scheme of PIES.

2. The robust planning method for a comprehensive park energy system according to claim 1, characterized in that, The construction of the multi-stage planning model considering the construction timeliness of PIES includes the steps of: Taking the shortest service life of the devices to be configured as the value of the planning period of the multi-stage planning model; Determining the number of planning stages of the multi-stage planning model according to the change of the maximum load level within the planning period; Setting constraint conditions to make the schemes of the previous planning stages associated with the subsequent planning stages and the decision results of the current planning stage related to all previous planning stages; Taking the minimum full life cycle cost as the goal and combining the constraint conditions to decide the candidate device configurations for each planning stage.

3. The robust planning method for a comprehensive park energy system according to claim 2, characterized in that The step of taking the minimum full life cycle cost as the goal and combining the constraint conditions to decide the candidate device configurations for each planning stage includes the steps of: Making the scheme of the current planning stage meet the load growth demand of the current planning stage; Configuring the candidate devices for the current planning stage based on the equipment investment and construction conditions of all previous planning stages until the candidate device configurations for all planning stages are decided.

4. The robust planning method for a comprehensive energy system in a park according to claim 3, characterized in that The step of determining the device digital models of energy production, energy conversion, and energy storage devices in PIES based on the modeling of the typical physical structure of PIES includes the steps of: Equivalently representing PIES as a two-port network covering multiple energy form inputs and outputs based on the energy hub EH model; Determining the energy production, energy conversion, and energy storage devices in PIES. Among them, the energy production device is a photovoltaic PV device, the energy conversion devices include power-to-gas P2G devices, electric boilers EB devices, combined heat and power CHP devices, and gas boilers GB devices, and the energy storage devices include electrical energy storage ES devices, thermal energy storage TS devices, and gas storage GS devices; Performing digital modeling on the energy production, energy conversion, and energy storage devices in PIES and obtaining the digital models of each device, and the digital models of each device include: The mathematical model of the PV device, which is expressed as: Among them, P PV (t) represents the output electric power of the PV device at time t, represents the upper limit of the output electric power of the PV device, which is determined by the rated capacity of the device; The mathematical model of the P2G device, which is expressed as: Among them, G P2G (t), g P2G (t) respectively represent the natural gas power output by the P2G device and the natural gas flow rate output at time t, η P2G represents the energy conversion efficiency of the P2G device, P P2G (t) represents the electric power input to the P2G device at time t; Q represents the lower heating value of natural gas, represents the upper limit of the electric power input to the P2G device, which is determined by the rated capacity of the device; The mathematical model of the EB device, which is expressed as: Among them, H EB (t) represents the thermal power output by the EB device at time t, P EB (t) represents the electric power input to the EB device at time t, η EB represents the energy conversion efficiency of the EB device, represents the upper limit of the electric power input to the EB device, which is determined by the rated capacity of the device; The mathematical model of the CHP device, which is expressed as: Among them, H CHP (t), η CHP,h respectively represent the thermal power output by the CHP device at time t and the thermal conversion efficiency of the device, P CHP (t), η CHP,e respectively represent the electric power output by the CHP device at time t and the electric conversion efficiency of the device, G CHP (t) represents the natural gas power input to the CHP device at time t, and its unit of value is the same as that of the electric power, and it is converted according to the low calorific value Q of natural gas, represents the upper limit of the natural gas power input to the CHP device, which is determined by the rated capacity of the device, Δ G CHP 、 respectively represent the maximum and minimum ramp rates of the CHP device; The mathematical model of the GB device, which is expressed as: Among them, H GB (t) represents the thermal power output by the GB device at time t; η GB represents the energy conversion efficiency of the GB device, G GB (t) represents the natural gas power input to the GB device at time t, and its unit of value is the same as that of the electric power, which is converted according to the lower calorific value Q of natural gas, represents the upper limit of the natural gas power input to the GB device, which is determined by the rated capacity of the GB device, Δ G GB 、 respectively represent the maximum and minimum ramp rates of the GB device; The mathematical models of the ES, TS, and HS devices, which are expressed as: Among them, \(s\in\Omega\), where \(\Omega = \{ES, TS, GS\}\) represents the set of three types of energy storage devices, namely ES, TS, and GS. They respectively represent the charging power magnitude and the upper limit of the charging power of the \(s\)-th type of energy storage device at time \(t\). They respectively represent the discharging power magnitude and the upper limit of the discharging power of the \(s\)-th type of energy storage device at time \(t\). is a 0-1 variable. When takes the value of 1, it means that the \(s\)-th type of energy storage is charging at time \(t\). When it takes the value of 0, it means that the \(s\)-th type of energy storage is discharging at time \(t\), \(D\) s (t) represents the state of charge of the \(s\)-th type of energy storage device at time \(t\). They respectively represent the charging and discharging efficiencies of the \(s\)-th type of energy storage device. \(\Delta t\) is the operation step length, and its value is 1h. D s They respectively represent the maximum and minimum state of charge values of the \(s\)-th type of energy storage device, \(D\) s (t0), \(D\) s (t T ) respectively represent the state of charge of the \(s\)-th type of energy storage device at the beginning and end of the scheduling period.

5. The robust planning method for a comprehensive energy system in a park according to claim 4, characterized in that Introducing stepped carbon trading in PIES, determining the free carbon emission allowance of PIES, and constructing a calculation model for stepped carbon trading fees, including the steps: Using the baseline method to determine the initial free carbon emission allowance of PIES for the carbon emission sources of PIES. The carbon emission sources of PIES include PIES purchased electricity, CHP equipment, and GB equipment, and the initial free carbon emission allowance is: Among them, E * represents the free carbon emission quota of PIES, and respectively represent the free carbon emission quotas corresponding to the externally purchased electricity, CHP equipment, and GB equipment of PIES. T represents the settlement period of the carbon trading fee, and T = 8760, represents the free carbon emission quota per unit of electricity purchase, P PN (t) represents the power purchase rate of PIES from the superior power grid at time t, represents the free carbon emission amount per unit of heat, represents the conversion coefficient of the power generation of the CHP equipment to the heat supply; Determining the actual carbon emissions of PIES, and the actual carbon emissions are expressed as: Among them, E represents the total actual carbon emissions of PIES, E PN , E CHP and E GB respectively represent the actual carbon emissions corresponding to the externally purchased electricity, CHP equipment, and GB equipment of PIES, E P2G represents the CO2 consumption of the P2G equipment, and β g represents the carbon consumption coefficient; Constructing a calculation model for stepped carbon trading fees, and the calculation model for stepped carbon trading fees is expressed as: Among them, C co represents the stepped carbon trading fee of PIES, c represents the carbon trading benchmark price, α represents the price growth coefficient, d represents the price range length, and the number of price ranges corresponds to the number of segments of C co respectively.

6. The robust planning method for a park integrated energy system according to claim 5, characterized in that Describing the source-load uncertainty of PIES using a box uncertainty set, including the steps: Describing the source uncertainty and load uncertainty of PIES using a box uncertainty set, and it is expressed as: Among them, \(U\) represents the uncertainty set, and \(u\) represents the set of uncertain variables within the planning period, which consists of the actual PV output \(P PV,n (t)\), the actual electrical load \(P L,n (t)\), the actual heat load \(H L,n (t)\), and the actual gas load \(G L,n (t)\) at each moment of each year. The superscript \(\hat{}\) represents the predicted value, \(\Delta u\) represents the maximum error fluctuation range allowed for each uncertain variable within the planning period, and \(\delta PV \), \(\delta PL \), \(\delta GL \), \(\delta GL represent the allowable fluctuation deviation coefficients of PV output, electrical load, heat load, and gas load respectively; Defining the uncertainty adjustment parameter Γ of the uncertainty set U as: Among them, Γ PV , Γ PL , Γ HL , Γ GL respectively represent the uncertain adjustment parameters of PV output, electrical load, heat load, and gas load, and respectively characterize the maximum number of years in the planning period during which the actual values of PV output, electrical load, heat load, and gas load are allowed to deviate from the predicted values.

7. The robust planning method for a comprehensive park energy system according to claim 6, characterized in that, Establishing a robust multi-stage planning model of PIES considering stepped carbon trading and source-load uncertainty based on the multi-stage planning model, including the steps: Dividing the multi-stage planning model into a first-stage problem model and a second-stage problem model. The first-stage problem model is used to make decisions on equipment investment, and the second-stage problem model is used to make decisions on equipment operation; Setting objective functions for the first-stage problem model and the second-stage problem model respectively, and the objective functions are expressed as: Among them, f1(·) represents the objective function of the first-stage problem model, which consists of the investment cost C inv,k , the equipment salvage value C R . f2(·) represents the objective function of the second-stage problem model, which consists of the operating cost C ope,n , the maintenance cost C main,n , and the carbon trading cost C co,n . x and y respectively represent the sets of optimization variables of the first-stage problem model and the second-stage problem model. Ξ(x,u) represents the feasible region of y after making decisions x and u. K represents the number of planning stages, k represents the k-th planning stage, N represents the number of years in the planning period, n represents the n-th year in the planning period, and n k represents that the first year of the k-th planning stage is the n k -th year of the planning period. γ represents the discount rate; The investment cost C inv,k is calculated as follows: Among them, c PV represents the investment and construction cost per unit capacity of PV, and W PV,k represents the capacity configuration of PV equipment in the k-th planning stage. Π = {P2G, EB, CHP, GB} and Ω = {ES, TS, GS} respectively represent the sets of energy conversion equipment and energy storage equipment. c i and c s respectively represent the investment and construction cost per unit capacity of the i-th type of energy conversion equipment and the s-th type of energy storage equipment, and W i,k and W s,k respectively represent the capacity configuration of the i-th type of energy conversion equipment and the s-th type of energy storage equipment in the k-th planning stage; The residual value C of the device R is calculated as follows: Among them, c dep,j represents the annual average depreciation expense of the equipment whose service life has not ended at the end of the j-th planning period, C inv,j represents the sum of the investment expenses of the j-th type of equipment in all stages, δ j represents the net salvage value rate of the j-th type of equipment, Y j represents the service life of the j-th type of equipment, N j represents the number of years of use of the j-th type of equipment from the start of construction to the end of the planning period; The operating cost C ope,n is calculated as follows: Among them, c e (t), c g (t) represent the electricity price and gas price at time t respectively; P PN,n (t) represents the power purchase of PIES from the superior power grid at time t in the nth year, and G GN,n (t) represents the gas purchase power of PIES from the superior gas grid at time t in the nth year; The maintenance cost C main,n is calculated as follows: Among them, O represents the unit maintenance cost matrix of various devices, and P n (t) represents the output power matrix of various devices at time t in the nth year, and P n T (t) represents the matrix P n (t) is the transpose of P PV,n (t) represents the electric power output by the PV device at time t in the nth year, respectively represent the charging power and discharging power of the gas storage device at time t in the nth year; Among them, the carbon trading fee C CO,n has the same calculation method as that of the stepped carbon trading fee C co ; Setting investment-level constraints for the first-stage problem model, and the investment-level constraints are expressed as: Among them, W i max and W s max respectively represent the maximum capacity values allowed for the construction of PV devices, the i-th type of energy conversion device, and the s-th type of energy storage device; Setting operation-level constraints for the second-stage problem model, and the operation-level constraints include equipment operation constraints, electro-thermal-gas balance constraints, power exchange constraints between PIES and the superior power grid and natural gas network, and carbon trading constraints. Among them, The equipment operation constraints include the equipment digital models and multi-stage constraint conditions of energy production, energy conversion, and energy storage equipment in PIES. The multi-stage constraint conditions are: Among them, P PV,n (t) represents the electric power output by the PV device at time t in the nth year, and ∑W PV represents the cumulative investment capacity of the PV device in the current k planning stages. n ∈ k indicates that the nth year of the planning period belongs to the kth planning stage. P i,n (t) represents the power input to the ith type of energy conversion device at time t in the nth year, and ∑W i represents the cumulative investment capacity of the ith type of energy conversion device in the current k planning stages. W s,n represents the maximum allowable charge state of the sth type of energy storage device in the nth year, and ∑W s represents the cumulative investment capacity of the sth type of energy storage device in the current k planning stages; The electro-thermal-gas balance constraints include electric power balance constraints, heat power balance, and gas power balance, and The electric power balance constraint is expressed as: The heat power balance is expressed as: The gas power balance is expressed as: Among them, P CHP,n (t) represents the electric power output by the CHP device at time t in the nth year, P P2G,n (t) represents the electric power input to the P2G device at time t in the nth year, P EB,n (t) represents the electric power input to the EB device at time t in the nth year, P L,n (t), H L,n (t) and G L,n (t) respectively represent the electric, heat, and gas load powers of the PIES at time t in the nth year, and respectively represent the charging powers of the ES, TS, and GS devices at time t in the nth year, and respectively represent the discharging powers of the ES, TS, and GS devices at time t in the nth year; H EB,n (t), H CHP,n (t) and H GB,n (t) respectively represent the heat powers output by the EB, CHP, and GB devices at time t in the nth year, G P2G,n (t) represents the natural gas power output by the P2G device at time t in the nth year, G CHP,n (t), G GB,n (t) respectively represent the natural gas powers input to the CHP and GB devices at time t in the nth year; The power exchange constraints between PIES and the superior power grid and natural gas network are expressed as: Among them, respectively represent the minimum and maximum values of the power of externally purchased electricity allowed by PIES, respectively represent the minimum and maximum values of the power of externally purchased natural gas allowed; The robust multi-stage planning model of PIES consists of a first-stage problem model and a second-stage problem model, and is determined based on the objective function, the investment-level constraints, and the operation-level constraints.

8. The robust planning method for a comprehensive park energy system according to claim 7, characterized in that Performing dual and linearization processing on the robust multi-stage planning model of PIES to obtain a mixed-integer linear programming model, including the steps: Settings And add an auxiliary variable β s (t), and replace the bilinear terms in the digital model of the device and with the auxiliary variable β s (t), and add a first constraint condition, where the first constraint condition is: Among them, M is a positive real number greater than a preset threshold; For the carbon trading volume values \(a_1, a_2, \cdots, a\) at \(m\) endpoints in the calculation model of the stepped carbon trading fee m and the stepped carbon trading fee values \(b_1, b_2, \cdots, b\) m , introduce \(m - 1\) auxiliary variables \(x_1, x_2, \cdots, x\) m-1 and \(m\) auxiliary continuous variables \(y_1, y_2, \cdots, y\) m , and satisfy the second constraint condition, where the second constraint condition is: and for any carbon trading volume \(a\in[a_1,a m \), its carbon trading fee \(b\) is expressed as: Based on the first constraint condition and the second constraint condition, expressing the robust multi-stage planning model of PIES as: Among them, x is the decision variable of the problem model in the first stage, which consists of the candidate device investment and construction capacity in each planning stage: photovoltaic W PV,k , various energy conversion devices W i,k , various energy storage devices W s,k , the charging and discharging energy state variables of the energy storage devices at each moment of each year and the auxiliary variables of the piecewise function of the annual carbon trading cost . W PV,k , W i,k , W s,k are integer variables. When taking the value of 1, it means that the s-th type of energy storage is charging at the t-th moment in the n-th year, and when taking the value of 0, it means discharging. is a vector matrix, which contains m - 1 auxiliary variables, and among them When taking the value of 1, it means that the carbon trading volume in the n-th year is in the interval [a m-1 , a m , and when taking the value of 0, it means that the carbon trading volume is in the interval [a1, a m-1 ). y is the decision variable of the second-stage problem model, which is composed of the purchased electricity P PN,n (t) at each moment of each year, the purchased natural gas G GN,n (t), the actual PV output P PV,n (t), the input power P i,n (t) of various energy conversion devices, the charging and discharging power of various energy storage devices the state of charge D s,n (t) of various energy storage devices, the actual electro-thermal-gas load P L,n (t), H L,n (t), G L,n (t) and the auxiliary variable of the annual carbon trading cost piecewise function ; A, B, C, D, E, F, G, H, I, c, d, e, f all represent coefficient matrices; Performing duality on the second-stage problem model in the robust multi-stage planning model of PIES, and the duality process includes: Set the first-stage problem model variable x * , such that x * is a constant for the max-min problem of the second-stage problem model of the PIES robust multi-stage planning model; Expressing the max-min problem as: Among them, λ1, λ2, and π represent the set of dual variables corresponding to the constraint conditions of the max-min problem, and λ1, λ2, and π are all continuous variables; Convert the max-min problem into a single-layer max problem, which is expressed as: For the bilinear term u in the single-layer max problem T π, change the uncertainty set U, and denote the changed uncertainty set U as: where, Δu·X represents dot product, and X is expressed as: where x PV,n , x PL,n , x HL,n , x GL,n are all 0-1 variables, and when their values are 1, they respectively represent that the actual output of PV takes the lower boundary of the fluctuation range at all times in the nth year of the planning period, and the actual electricity, heat, and gas loads take the upper boundary of the fluctuation range at all times in the nth year of the planning period; Linearize the bilinear term X·π, and the linearization process includes: Introduce an auxiliary variable β u , replace all X·π in the single-layer max problem with β u , and equivalent the single-layer max problem to: where, Γ represents a set of uncertain adjustment parameters; Determine the second-stage problem model determined by the equivalent single-layer max problem as the mixed-integer linear programming model.

9. The robust planning method for a comprehensive energy system in a park according to claim 8, characterized in that Solving the mixed-integer linear programming model to obtain the robust planning scheme of PIES includes the steps: S01: Initialize the parameters of the C&CG algorithm, including setting the iteration number κ = 0 and the set of iteration numbers for generating the optimal cut The upper bound UB = +∞, the lower bound LB = -∞, the allowable error coefficient ε = 0.001, and an initial value X of a set of uncertain variables is given * , determine S02: Solve according to u * Solve the first-stage problem and update the decision result: x * = x * {κ}, y * = y * {κ}, and update the lower bound LB = A T x * {κ}+η * {κ}, and represent the first-stage problem as: where, η represents the auxiliary variable of the second-stage problem, and y{·} represents the new variable introduced into the second-stage problem model by the cut set; S03: Solve according to x * Solve the second-stage problem. Let θ(x * ) denote the optimal solution of the objective function of the second-stage problem model, and determine whether the second-stage problem model has a feasible solution under the current x * . If there is no feasible solution, let θ(x * ) = +∞, and u * remains unchanged; if there is a feasible solution, update the decision result: X * = X * {κ}, u * = u * {κ}, π * = π * {κ}; and update the upper bound after the judgment: UB = min{UB, A T x * + θ(x * )}; S04: If the upper bound and the lower bound satisfy the convergence condition |UB - LB| / LB ≤ ε, then the solution results x * , y * , u * and exit the C&CG algorithm loop; otherwise, generate a new set of decision variables y{κ+1} for the second-stage problem model and generate feasible cuts FCs and optimal cuts OCs, expressed as: S05: If there is no feasible solution to the second-stage problem in step S03, let κ = κ + 1, and only feedback the feasible cut FCs to the first-stage problem, and add the feasible cut FCs to the constraint conditions of the first-stage problem model and then return to step S02; If there is a feasible solution to the second-stage problem in step S03, let Θ = Θ ∪ {κ + 1}, κ = κ + 1, then feedback the feasible cut FCs and the optimal cut OCs to the first-stage problem, and add the feasible cut FCs and the optimal cut OCs to the constraint conditions of the first-stage problem model and then return to step S02.

10. A robust planning method for a comprehensive park energy system according to claim 9, characterized in that, After obtaining the robust planning scheme of PIES, evaluate the collaborative optimization effect of new energy and multiple energy storages in PIES, specifically including: Establish a comprehensive energy efficiency evaluation model of PIES, which is expressed as: Among them, η PIES represents the comprehensive energy efficiency of PIES, P L , H L , G L respectively represent the electric, heat, and gas loads of PIES, respectively represent the charging powers of ES devices, TS devices, and GS devices, respectively represent the charging efficiencies of ES devices, TS devices, and GS devices, respectively represent the discharging powers of ES devices, TS devices, and GS devices, P PN , G GN respectively represent the electricity purchase power and gas purchase power of PIES; Evaluate the collaborative optimization effect of new energy and multiple energy storages based on the comprehensive energy efficiency evaluation model.

Images