Risk-avoidance park integrated energy system multi-time scale scheduling method and device

Abstract

The invention discloses a risk-avoidance park integrated energy system multi-time scale scheduling method and device, and the method comprises the steps: carrying out the modeling of a park integrated energy system through setting a modeling strategy, and obtaining a park integrated energy system constraint model; generating a day-ahead scene through a scene generation strategy and a scene reduction strategy; generating an intra-day scene according to the prediction data of the day-ahead scene through a scene tree strategy; generating a real-time scene according to the intraday scene through a scene tree strategy; according to the park integrated energy system constraint model, respectively constructing a day-ahead three-stage stochastic optimization model, an intra-day two-stage stochastic rolling optimization model and a real-time rolling optimization model considering CVaR; and respectively solving the model by setting a solver to obtain a day-ahead stage scheduling strategy, an intra-day stage scheduling strategy and a real-time stage scheduling strategy of the park integrated energy system. According to the method, economic and reliable scheduling under multiple time scales can be realized, and the risk avoidance capability is realized while the economy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy system operation optimization, and in particular to a risk-avoiding multi-time scale scheduling method and device for a park integrated energy system. Background Art

[0002] As the energy crisis and environmental degradation intensify, the disadvantages of the inefficiency of traditional energy systems have gradually become apparent. PIES has a higher comprehensive energy utilization rate than traditional single user-side energy systems by organically coordinating and optimizing the user-side utilization of different energy sources such as electricity, natural gas, and heating.

[0003] At present, domestic and foreign scholars have conducted a lot of research on PIES multi-time scale optimization scheduling, but most of the research adopts a two-stage optimization scheduling model, which means that the economic efficiency of multi-time scale scheduling still has room for further improvement. In addition, PIES involves multiple uncertainties such as renewable energy output, multi-energy load and energy price. A large number of studies use stochastic programming methods to consider the impact of uncertain factors, which makes the scheduling plan often too optimistic and lacks risk aversion ability.

[0004] Therefore, how to invent a risk-averse PIES multi-time scale optimization scheduling method that can improve economic efficiency while having risk avoidance capabilities has become an urgent problem to be solved. Summary of the invention

[0005] To this end, the present invention provides a risk-avoiding multi-time-scale scheduling method and device for a park integrated energy system, which can achieve economical and reliable scheduling under multiple time scales, and has risk avoidance capabilities while improving economy.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a multi-time scale scheduling method for a risk-averse park integrated energy system, comprising:

[0007] By setting the modeling strategy, the renewable energy output equipment, energy coupling link, energy storage equipment and bus structure in the park's integrated energy system are modeled to obtain the park's integrated energy system constraint model;

[0008] Generate a day-ahead scenario through scenario generation strategy and scenario reduction strategy; generate an intraday scenario based on the forecast data of the day-ahead scenario through scenario tree strategy; generate a real-time scenario based on the intraday scenario through scenario tree strategy;

[0009] According to the comprehensive energy system constraint model of the park, a three-stage stochastic optimization model for the day before, a two-stage stochastic rolling optimization model for the day, and a real-time rolling optimization model are constructed respectively.

[0010] By setting the solver to solve the three-stage day-ahead random optimization model taking into account CVaR, the two-stage intraday random rolling optimization model and the real-time rolling optimization model respectively, the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park's integrated energy system are obtained.

[0011] As a preferred scheme for a risk-averse multi-time scale scheduling method for a park integrated energy system, in the process of modeling the renewable energy output equipment in the park integrated energy system by setting a modeling strategy, the constraint model expression of the renewable energy output equipment is:

[0012]

[0013] In the formula, The actual output of PV; is the maximum output of PV;

[0014] In the process of modeling the energy coupling link in the park comprehensive energy system by setting the modeling strategy, the energy coupling link includes: cogeneration unit, ground source heat pump, electric boiler and gas boiler; the constraint model expression of the energy coupling link is:

[0015]

[0016] In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; It is the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices;

[0017] In the process of modeling the energy storage equipment in the park comprehensive energy system by setting the modeling strategy, the energy storage equipment includes: power storage equipment and heat storage equipment; the constraint model expression of the energy storage equipment is:

[0018]

[0019] In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment;

[0020] In the process of modeling the busbar structure in the park integrated energy system by setting the modeling strategy, the constraint model expression of the busbar structure is:

[0021]

[0022] In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

[0023] As a preferred scheme for the multi-time scale scheduling method of the integrated energy system of a risk-averse park, in the process of generating the day-ahead scenario through the scenario generation strategy and the scenario reduction strategy, according to the distribution information of photovoltaic output, electric load and electricity price, the scenario generation strategy is used to generate relevant samples; according to the relevant samples, the scenario reduction strategy is used to cluster and obtain typical scenarios; the scenario generation strategy is a scenario generation strategy based on CD-LHS; and the scenario reduction strategy is a scenario reduction strategy based on GMM clustering.

[0024] As a preferred scheme for a risk-averse multi-time scale scheduling method for a park integrated energy system, the expression of the three-stage stochastic optimization model taking into account CVaR is:

[0025]

[0026] In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient;

[0027] The expression of the intraday two-stage stochastic rolling optimization model is:

[0028]

[0029] In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set;

[0030] The expression of the real-time rolling optimization model is:

[0031]

[0032] In the formula, A collection of time periods for the intraday scheduling model.

[0033] As a preferred scheme for a risk-averse multi-time scale scheduling method for a park integrated energy system, the expression of the day-ahead scheduling strategy is:

[0034]

[0035] In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP;

[0036] The expression of the intraday stage scheduling strategy is:

[0037]

[0038] In the formula, It is a two-stage scheduling strategy within the day; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power;

[0039] The expression of the real-time stage scheduling strategy is:

[0040]

[0041] In the formula, It is a real-time rolling scheduling strategy; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

[0042] The present invention also provides a risk-averse multi-time scale scheduling device for a park integrated energy system, based on the above risk-averse multi-time scale scheduling method for a park integrated energy system, comprising:

[0043] The park comprehensive energy system modeling module is used to model the renewable energy output equipment, energy coupling links, energy storage equipment and busbar structure in the park comprehensive energy system by setting modeling strategies to obtain the park comprehensive energy system constraint model;

[0044] A scenario generation module, used to generate a day-ahead scenario through scenario generation strategy and scenario reduction strategy; generate an intraday scenario based on the forecast data of the day-ahead scenario through scenario tree strategy; generate a real-time scenario based on the intraday scenario through scenario tree strategy;

[0045] An optimization model construction module, for constructing a three-stage stochastic optimization model for the day before, a two-stage stochastic rolling optimization model for the day, and a real-time rolling optimization model taking into account CVaR according to the constraint model of the park comprehensive energy system;

[0046] The scheduling strategy acquisition module is used to solve the three-stage random optimization model taking into account CVaR, the two-stage random rolling optimization model within the day, and the real-time rolling optimization model by setting a solver to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy, and the real-time stage scheduling strategy of the park's comprehensive energy system.

[0047] As a preferred solution for a multi-time scale scheduling device of a risk-averse park integrated energy system, in the park integrated energy system modeling module, in the process of modeling the renewable energy output equipment in the park integrated energy system by setting a modeling strategy, the constraint model expression of the renewable energy output equipment is:

[0048]

[0049] In the formula, The actual output of PV; is the maximum output of PV;

[0050] In the process of modeling the energy coupling link in the park comprehensive energy system by setting the modeling strategy, the energy coupling link includes: cogeneration unit, ground source heat pump, electric boiler and gas boiler; the constraint model expression of the energy coupling link is:

[0051]

[0052] In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; It is the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices;

[0053] In the process of modeling the energy storage equipment in the park comprehensive energy system by setting the modeling strategy, the energy storage equipment includes: power storage equipment and heat storage equipment; the constraint model expression of the energy storage equipment is:

[0054]

[0055] In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment;

[0056] In the process of modeling the busbar structure in the park integrated energy system by setting the modeling strategy, the constraint model expression of the busbar structure is:

[0057]

[0058] In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

[0059] As a preferred solution for a risk-averse multi-time-scale scheduling device for a park integrated energy system, in the scenario generation module, in the process of generating the day-ahead scenario through the scenario generation strategy and the scenario reduction strategy, a correlated sample is generated according to the distribution information of photovoltaic output, electric load and electricity price through the scenario generation strategy; based on the correlated samples, a typical scenario is clustered through the scenario reduction strategy; the scenario generation strategy is a scenario generation strategy based on CD-LHS; and the scenario reduction strategy is a scenario reduction strategy based on GMM clustering.

[0060] As a preferred solution for a multi-time scale scheduling device for a risk-averse park integrated energy system, in the optimization model construction module, the expression of the three-stage stochastic optimization model taking into account CVaR is:

[0061]

[0062] In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient;

[0063] The expression of the intraday two-stage stochastic rolling optimization model is:

[0064]

[0065] In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set;

[0066] The expression of the real-time rolling optimization model is:

[0067]

[0068] In the formula, A collection of time periods for the intraday scheduling model.

[0069] As a preferred solution for a multi-time scale scheduling device for a risk-averse park integrated energy system, in the scheduling strategy acquisition module, the expression of the day-ahead scheduling strategy is:

[0070]

[0071] In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP;

[0072] The expression of the intraday stage scheduling strategy is:

[0073]

[0074] In the formula, It is a two-stage scheduling strategy within the day; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power;

[0075] The expression of the real-time stage scheduling strategy is:

[0076]

[0077] In the formula, It is a real-time rolling scheduling strategy; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

[0078] The present invention has the following advantages: by setting a modeling strategy, the renewable energy output equipment, energy coupling links, energy storage equipment and bus structure in the park comprehensive energy system are modeled to obtain the park comprehensive energy system constraint model; by using a scenario generation strategy and a scenario reduction strategy, a day-ahead scenario is generated; by using a scenario tree strategy, an intraday scenario is generated according to the predicted data of the day-ahead scenario; by using a scenario tree strategy, a real-time scenario is generated according to the intraday scenario; according to the park comprehensive energy system constraint model, a day-ahead three-stage random optimization model, an intraday two-stage random rolling optimization model and a real-time rolling optimization model taking into account CVaR are respectively constructed; by setting a solver, the day-ahead three-stage random optimization model taking into account CVaR, the intraday two-stage random rolling optimization model and the real-time rolling optimization model are respectively solved to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park comprehensive energy system. The present invention first models the renewable energy output equipment, energy coupling equipment, energy storage equipment and park structure involved in PIES; then, the Cholesky decomposition-based Latin hypercubic sampling method (CD-LHS) and Gaussian mixture model (GMM) clustering method are used to generate and reduce scenarios, and typical scenarios with correlation are obtained; on this basis, a three-stage random optimization model of the day before, a two-stage random rolling optimization model within the day and a real-time rolling optimization model are constructed taking into account the conditional value at risk (CVaR), so as to realize economic and reliable scheduling under multiple time scales. The present invention can have risk avoidance capabilities while improving economic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0080] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0081] Figure 1 A schematic flow chart of a multi-time scale scheduling method for a risk-averse park integrated energy system provided in Example 1 of the present invention;

[0082] Figure 2 A schematic diagram of the structure of a typical electrical-thermal coupling PIES in a risk-averse multi-time-scale scheduling method for a park integrated energy system provided in Example 1 of the present invention;

[0083] Figure 3 A schematic diagram of a three-stage scheduling scenario tree generation process in a risk-averse multi-time-scale scheduling method for a park integrated energy system provided in Example 1 of the present invention;

[0084] Figure 4 A schematic diagram of a multi-time-scale optimization scheduling framework in a risk-averse multi-time-scale scheduling method for a park integrated energy system provided in Example 1 of the present invention;

[0085] Figure 5 A schematic diagram of typical day data in winter in a possible embodiment provided in Embodiment 1 of the present invention;

[0086] Figure 6 A schematic diagram of PIES electric power operation in a possible embodiment provided in Embodiment 1 of the present invention;

[0087] Figure 7 This is a schematic diagram of the thermal power operation of PIES in a possible embodiment provided in Embodiment 1 of the present invention;

[0088] Figure 8 This is a schematic diagram of the architecture of a multi-time scale scheduling device for a risk-avoiding integrated energy system in a park provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0089] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. 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 creative work are within the scope of protection of the present invention.

[0090] Example 1

[0091] See also Figure 1 Embodiment 1 of the present invention provides a risk-averse multi-time scale scheduling method for a park integrated energy system, comprising the following steps:

[0092] S1. Model the renewable energy output equipment, energy coupling links, energy storage equipment and busbar structure in the park's integrated energy system by setting a modeling strategy to obtain a constraint model of the park's integrated energy system;

[0093] S2. Generate a day-ahead scenario through a scenario generation strategy and a scenario reduction strategy; generate an intraday scenario based on the predicted data of the day-ahead scenario through a scenario tree strategy; generate a real-time scenario based on the intraday scenario through a scenario tree strategy;

[0094] S3. According to the park comprehensive energy system constraint model, a three-stage stochastic optimization model for the day before, a two-stage stochastic rolling optimization model for the day, and a real-time rolling optimization model are constructed respectively taking into account CVaR;

[0095] S4. By setting a solver, the three-stage day-ahead random optimization model taking into account CVaR, the two-stage intraday random rolling optimization model and the real-time rolling optimization model are solved respectively to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park's comprehensive energy system.

[0096] In this embodiment, in step S1, the renewable energy output equipment, energy coupling link, energy storage equipment and bus structure in the park comprehensive energy system are modeled by setting a modeling strategy to obtain a constraint model of the park comprehensive energy system;

[0097] Specifically, the typical electrical and thermal coupling PIES structure is as follows: Figure 2 As shown, it includes modules such as renewable energy output equipment, energy coupling links, energy storage equipment and busbar structure.

[0098] The renewable energy output equipment in PIES is photovoltaic power generation equipment (PV), and the constraint model expression of the renewable energy output equipment is:

[0099]

[0100] In the formula, The actual output of PV; is the maximum output of PV; formula (1) is the upper and lower limit constraints of PV output.

[0101] The energy coupling link in PIES includes combined heat and power (CHP), ground source heat pump (HP), electric boiler (EB) and gas boiler (GB). The constraint model expression of the energy coupling link is:

[0102]

[0103] In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; It is the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices;

[0104] Formula (2) is the upper and lower limit constraints of CHP output; Formula (3) is the upper and lower limit constraints of energy coupling equipment output; Formula (4) is the energy coupling equipment ramp rate constraint; Formula (5) is the CHP gas-to-electricity conversion constraint; Formula (6) is the CHP gas-to-heat conversion constraint; Formula (7) is the HP electric-to-heat conversion constraint; Formula (8) is the EB electric-to-heat conversion constraint; Formula (9) is the GB gas-to-heat conversion constraint.

[0105] The energy storage equipment in PIES includes electric storage equipment (ES) and heat storage equipment (HS). The constraint model expression of the energy storage equipment is:

[0106]

[0107] In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment;

[0108] Formulas (10)-(12) are the charging and discharging constraints of the energy storage device; Formula (13) is the state of charge (SOC) balance constraint of the energy storage device; Formula (14) is the upper and lower limit constraints of the SOC of the energy storage device; Formula (15) is the initial and final SOC constraint of the energy storage device.

[0109] Since PIES is small in scale, its topological structure is ignored and the busbar structure is used for modeling. The constraint model expression of the busbar structure is:

[0110]

[0111] In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

[0112] Formula (17) is the power balance constraint of the electric bus; Formula (18) is the power balance constraint of the thermal bus; Formula (19) is the power balance constraint of the natural gas bus.

[0113] In this embodiment, in step S2, a day-ahead scenario is generated through a scenario generation strategy and a scenario reduction strategy; an intraday scenario is generated based on the predicted data of the day-ahead scenario through a scenario tree strategy; a real-time scenario is generated based on the intraday scenario through a scenario tree strategy;

[0114] Specifically, the present invention adopts the scenario method to model the integrated environment. Photovoltaic output obeys Beta distribution, electric heat load obeys normal distribution, and electricity price obeys normal distribution. There is a positive correlation between photovoltaic output, electric load, and electricity price, and the three are negatively correlated with thermal load. According to the distribution information, samples with correlation are generated by the CD-LHS method. The number of scenarios obtained by scenario generation is large. In order to avoid excessive calculation, the GMM clustering method is used to cluster the samples to obtain typical scenarios.

[0115] Among them, the scene generation strategy based on CD-LHS is:

[0116] Suppose we want to sample M N-dimensional scenes, then the initial sample X0| M×N It can be expressed as:

[0117]

[0118] In the formula, It is the inverse function of the probability density function of the uncertain factors in the scene; U m,n is a random variable and satisfies U m,n ~U(0,1);R n It is a sequence of natural numbers 1, 2, ..., M randomly shuffled.

[0119] For the initial sample correlation coefficient matrix C0| N×N and target correlation coefficient matrix C| N×N Performing Cholesky decomposition yields:

[0120]

[0121] C=LU=U T U (21)

[0122] Where U0 is the upper triangular matrix after C0 decomposition; L0 is the lower triangular matrix after C0 decomposition; U is the upper triangular matrix after C decomposition; L is the lower triangular matrix after C decomposition.

[0123] Through the decomposed upper triangular matrices U0 and U, the initial sample can be transformed into a sample with the target correlation coefficient X| M×N :

[0124]

[0125] Among them, the scene reduction strategy based on GMM clustering is:

[0126] Suppose we want to reduce M N-dimensional scenes to K, then the probability density function of the Gaussian function corresponding to each scene after the reduction is:

[0127]

[0128] In the formula, γ i =[μ i ,Σ i ,p i ] is the scene parameter; μ i is the scene mean; Σ i is the scene variance diagonal matrix; p i is the scene probability.

[0129] The reduced K scenes can form a GMM:

[0130]

[0131] Where, γ={γ1,γ2,...,γ K} are GMM parameters.

[0132] For the sample before reduction X = {x1, x2, ..., x M}, its likelihood function can be expressed as:

[0133]

[0134] By finding the parameters that maximize the likelihood function, we can get the GMM parameters. This problem can usually be solved using the expectation maximization (EM) method.

[0135] In this embodiment, unlike the traditional two-stage scheduling method, since the three stages of day-ahead, intraday and real-time need to be considered, it is impossible to obtain a typical scenario through only one scenario generation and scenario reduction. Therefore, the day-ahead scenario is obtained through the scenario tree method; secondly, the intraday scenario is generated based on the day-ahead forecast data; and then the real-time scenario is generated based on the intraday scenario, such as Figure 3 As shown. If M is generated ID Intraday scenarios, each intraday forecast scenario generates M RT For real-time scenarios, the number of day-ahead scenarios is:

[0136] M DA =M ID M RT (27)

[0137] Where M DA is the number of day-ahead scenarios.

[0138] In this embodiment, in step S3, according to the park integrated energy system constraint model, a three-stage random optimization model for the day before, a two-stage random rolling optimization model for the day, and a real-time rolling optimization model taking into account CVaR are respectively constructed;

[0139] Specifically, the multi-time scale stochastic optimization scheduling model considering CVaR consists of three optimization models: a three-order stochastic optimization model considering CVaR, a two-stage stochastic rolling optimization model within the day, and a real-time rolling optimization model, such as Figure 4As shown in the figure. In the day-ahead dispatch process, the forecast errors of photovoltaic output, electric heat load and electricity market price are large, which may cause the dispatch strategy to be too optimistic and the dispatch cost to be too high. Therefore, CVaR is used to avoid the dispatch risk. In the intraday dispatch process, the intraday forecast error is small, and the dispatch risk is also reduced. In order to obtain a more economical intraday market power purchase plan, a random rolling optimization model is adopted. In the real-time dispatch process, in order to ensure the balance of real-time power, a deterministic rolling optimization model is adopted to obtain the output plan as soon as possible.

[0140] The three-stage stochastic optimization model taking into account CVaR is:

[0141] Assume that the time scale of the day-ahead scheduling model is Δt DA , and its optimization range is the entire scheduling day. Then the time period set of the day-ahead scheduling model can be expressed as:

[0142]

[0143] Through scenario generation and scenario reduction, the day-ahead scenario set can be expressed as:

[0144]

[0145] In the formula, The scene is in the recent scene set; It is the maximum output value of photovoltaic in MPPT mode; is the electrical load power; is the heat load power; is the intraday market electricity price; The real-time market electricity price.

[0146] In the day-ahead scheduling model, only the day-ahead scheduling strategy is actually executed, and the intraday and real-time scheduling strategies are not executed. Therefore, each scenario corresponds to a different intraday scheduling strategy. and real-time scheduling strategies Different daily scheduling costs are generated accordingly and real-time scheduling costs Based on the above settings, the three-stage stochastic optimization model for the day-ahead considering CVaR can be expressed as:

[0147]

[0148]

[0149] Formula (1)-(18),,

[0150] In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient; is the day-ahead electricity price; is the intraday electricity price; is the real-time electricity price; p Gas is the gas purchase price; p Heat is the heat purchase price; p Mai,j is the unit power maintenance cost of equipment j; p Mai,j is the single startup cost of equipment j; Depreciation cost for energy storage; is the startup cost of equipment j; is the discharge depth of ES; is the investment cost per unit capacity of ES; ES is the ES depreciation rate; a, b, c are the ES cycle life parameters.

[0151] Formula (32) is the expression for the day-ahead cost; Formula (33) is the expression for the intraday cost; Formula (34) is the expression for the real-time cost; Formulas (35)-(36) are the expressions for the CHP startup cost; Formula (37) is the expression for the ES depreciation cost; Formula (38) is the expression for the CVaR calculation; Formula (39) is the non-foreseeable constraint for day-ahead scheduling.

[0152] Among them, the intraday two-stage random rolling optimization model is:

[0153] The intraday optimization model is rolled out with the actual operation of the system, and the intraday power purchase strategy is given a few minutes before each clearing. Assume that its time scale is Δt ID , the optimization range is l ID Δt ID . Rolling optimization every Δt ID The optimization is performed once every time, and its optimization range can be divided into control domain and prediction domain: the control domain is Δt after each optimization moment. ID The intraday strategy in the control domain will be executed in this optimization; the prediction domain is the (l ID -1)Δt ID The intraday strategy in the prediction domain is not executed during the time period, but is executed in the subsequent rolling process. If the intraday optimization model is called at time τ, the set of optimized time periods can be expressed as:

[0154]

[0155] At time τ, the intraday scene set can be obtained through scene generation and scene reduction methods:

[0156]

[0157] In the intraday optimization model, the real-time strategy is not executed. Therefore, each scenario corresponds to a different real-time strategy. Different real-time costs are generated accordingly Based on the above settings, the intraday two-stage stochastic optimization model can be expressed as:

[0158]

[0159] Formula (1)-(18), (32)-(37),

[0160] In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set.

[0161] Wherein, the real-time rolling optimization model is:

[0162] The rolling process of the real-time optimization model is similar to that of the intraday optimization model. Let its time scale be Δt RT , the optimization range is l RT Δt RT If the real-time optimization model is called at time τ, the set of optimization time periods can be expressed as:

[0163]

[0164] Real-time scenarios are obtained through real-time measurement and market information:

[0165]

[0166] Based on the above settings, the real-time rolling optimization model can be expressed as:

[0167]

[0168] Formula (1)-(18), (32)-(37),

[0169] In the formula, A collection of time periods for the intraday scheduling model.

[0170] In this embodiment, in step S4, the solver is set to solve the three-stage day-ahead random optimization model taking into account CVaR, the two-stage intraday random rolling optimization model and the real-time rolling optimization model, respectively, to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park integrated energy system.

[0171] Specifically, the three-stage stochastic optimization model taking CVaR into account, the two-stage stochastic rolling optimization model within the day, and the real-time rolling optimization model can all be converted into MILP problems, which can be solved by various MILP solvers.

[0172] Among them, in the three-stage dispatch model: the day-ahead stage decision variables include CHP start and stop status, energy storage equipment output and day-ahead market power purchase; the intraday stage decision variables include unit output and intraday market power purchase; the real-time stage decision variables include unit output adjustment, photovoltaic output, real-time market power purchase, gas purchase and heat purchase. Therefore, the expression of the day-ahead stage dispatch strategy is:

[0173]

[0174] In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP;

[0175] The expression of the intraday stage scheduling strategy is:

[0176]

[0177] In the formula, It is a two-stage scheduling strategy within the day; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power;

[0178] The expression of the real-time stage scheduling strategy is:

[0179]

[0180] In the formula, It is a real-time rolling scheduling strategy; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

[0181] In a possible embodiment, a scheduling method for a certain electricity-gas-heat coupled PIES in northern China is provided as follows:

[0182] Taking a certain electric-gas-heat coupled PIES in northern China as an example, the capacity configuration of various equipment in PIES is: PV capacity 650kW, CHP capacity 350kW, HP capacity 300kW, EB capacity 100kW, GB capacity 100kW, ES capacity 500kW·h, HS capacity 600kW·h, maximum power of electric load 550kW, maximum power of thermal load 800kW. The park power market refers to the three-layer architecture of the day-ahead-intraday-real-time of the European power market. The average intraday electricity price is 1.2 times the day-ahead electricity price, and the average real-time electricity price is 1.8 times the day-ahead electricity price. The park adopts a fixed gas price of 0.28 yuan / kW; a fixed heat price of 0.6 yuan / kW. CVaR confidence β=0.90. The day-ahead scheduling time scale is 1h, and the optimization range is 24h; the intraday scheduling time scale is 15min, and the optimization range is 2h; the real-time scheduling time scale is 5min, and the optimization range is 30min.

[0183] The day-ahead data of PV output, electrical load, thermal load, and electricity price are based on typical winter day data, such as Figure 5 As shown in the figure, the intraday data is used to generate 100 sets of data through the CD-LHS method to verify the intraday scenario. The correlation coefficient between PV output and electric load is 0.80, the correlation coefficient between PV output and thermal load is -0.50, the correlation coefficient between PV output and electricity price is 0.40, the correlation coefficient between electric load and thermal load is -0.70, the correlation coefficient between electric load and electricity price is 0.60, and the correlation coefficient between thermal load and electricity price is -0.30.

[0184] After calculation, the PIES power operation status can be obtained as follows: Figure 6 As shown, the thermal power operation is as follows Figure 7 As shown. By analyzing the scheduling process, we can draw the following conclusions:

[0185] 1. The equipment EB was in shutdown state throughout the whole process and did not participate in the scheduling process. This is because EB converts electrical energy into heat energy at a low efficiency, and the comprehensive energy utilization rate is low, so it is abandoned by the scheduling method.

[0186] Second, at 3:00 a.m., the electricity price is relatively low and the electricity demand is small, so the equipment HP and GB cooperate to supply energy; at 22:00-23:00 at night, the heat demand is reduced, and the equipment CHP and GB cooperate to supply energy; in the rest of the time, the load demand is large, and the equipment CHP, HP, and GB cooperate to complete the energy supply. The above process shows that the scheduling method of the present invention can flexibly mobilize equipment to cooperate according to demand and cost, and complete the economical and reliable supply of multi-energy loads.

[0187] 3. Energy storage devices ES and HS can effectively smooth out the peaks of electric and thermal loads, which helps other components complete system integration and can effectively reduce system integration costs. The output time of renewable energy power generation equipment PV is more consistent with the demand for electric and thermal loads, and can provide clean electricity for the system.

[0188] 4. The purchase of electricity by the industrial park is mainly carried out during the day-ahead dispatch process. The purpose of purchasing electricity during this period is to meet the energy demand of the industrial park. The purpose of purchasing electricity during intraday dispatch and real-time dispatch is to cope with forecast errors and fluctuations in uncertainty factors.

[0189] This embodiment sets four solutions for comparison:

[0190] Solution 1: Use the three-stage risk-averse stochastic optimization model described in this paper for scheduling;

[0191] Solution 2: Use the traditional two-stage risk-averse stochastic optimization model for scheduling;

[0192] Solution 3: Use a three-stage stochastic optimization model without considering risk aversion for scheduling;

[0193] Solution 4: Use a three-stage robust optimization model for scheduling;

[0194] After calculation, the daily operating cost, CVaR of daily operating cost, day-ahead scheduling cost, intraday operating cost and CVaR of intraday operating cost of each scheme can be obtained, as shown in Table 1.

[0195] Cost item / yuan Solution 1 Solution 2 Solution 3 Solution 4 Daily operating cost / yuan 10930 10974 10934 11270 CVaR of daily operating cost / yuan 11080 11125 11092 11324 Day-ahead dispatch cost / yuan 5401 5533 5322 7311 Intraday scheduling cost / yuan 5529 5441 5612 3959 CVaR of intraday dispatch cost / yuan 5679 5592 5770 4013

[0196] Table 1 Comparison of scheme data

[0197] By comparing the daily operating costs under various scenarios, it can be seen that the daily operating costs of Scheme 1 are basically the same as those of Scheme 3, 0.40% lower than those of Scheme 2, and 3.11% lower than those of Scheme 4; the daily operating CVaR of Scheme 1 is 0.41% lower than those of Scheme 2, 0.11% lower than those of Scheme 3, and 2.20% lower than those of Scheme 4. This shows that Scheme 1 is more economical than other schemes and has lower cost tail risk.

[0198] The day-ahead dispatch cost of Scheme 2 is 2.44% higher than that of Scheme 1, and the intraday dispatch cost is 1.59% higher than that of Scheme 1. This is because Scheme 2 is a two-stage dispatch method. Compared with the three-stage dispatch method, it ignores the intraday dispatch process, so it is impossible to avoid the additional costs and risks brought by the fluctuation of uncertain factors through the intraday market, so its dispatch cost and dispatch risk are too high.

[0199] The day-ahead operating cost of Scheme 3 is 1.46% lower than that of Scheme 1, and the intraday operating cost is 1.50% higher than that of Scheme 1. This is because Scheme 2 did not consider risk aversion during the day-ahead dispatch process and purchased less electricity during the day-ahead period. In the intraday period, because its plan is too optimistic, it needs to bear higher intraday operating costs, making its dispatch risk higher than that of Scheme 1.

[0200] The day-ahead dispatching cost of Scheme 4 is 35.36% higher than that of Scheme 1, and the intraday dispatching cost is 28.40% lower than that of Scheme 1. This is because Scheme 4 adopts a robust dispatching method, so in the day-ahead dispatching stage, Scheme 4 will consider the worst intraday scenario, so it needs to pay a high cost. Since a more conservative strategy is adopted in the day-ahead dispatching, the intraday dispatching cost of Scheme 4 is lower, but this still makes the total dispatching cost too high, making the economic efficiency of this scheme poor.

[0201] In summary, the present invention models the renewable energy output equipment, energy coupling links, energy storage equipment and bus structure in the park comprehensive energy system by setting a modeling strategy to obtain the park comprehensive energy system constraint model; generates a day-ahead scenario through a scenario generation strategy and a scenario reduction strategy; generates an intraday scenario according to the predicted data of the day-ahead scenario through a scenario tree strategy; generates a real-time scenario according to the intraday scenario through a scenario tree strategy; and constructs a day-ahead three-stage random optimization model, an intraday two-stage random rolling optimization model and a real-time rolling optimization model taking into account CVaR according to the park comprehensive energy system constraint model respectively according to the park comprehensive energy system constraint model; and solves the day-ahead three-stage random optimization model, the intraday two-stage random rolling optimization model and the real-time rolling optimization model taking into account CVaR respectively by setting a solver to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park comprehensive energy system. The present invention first models the renewable energy output equipment, energy coupling equipment, energy storage equipment and park structure involved in PIES; then, the Cholesky decomposition-based Latin hypercubic sampling method (CD-LHS) and Gaussian mixture model (GMM) clustering method are used to generate and reduce scenarios, and typical scenarios with correlation are obtained; on this basis, a three-stage random optimization model of the day before considering the conditional value at risk (CVaR), a two-stage random rolling optimization model within the day and a real-time rolling optimization model are constructed to achieve economic and reliable scheduling under multiple time scales. The present invention can have risk avoidance capabilities while improving economic efficiency.

[0202] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0203] It should be noted that some embodiments of the present disclosure are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0204] Example 2

[0205] See also Figure 8 Embodiment 2 of the present invention further provides a risk-averse multi-time scale scheduling device for a park integrated energy system, comprising:

[0206] The park comprehensive energy system modeling module 001 is used to model the renewable energy output equipment, energy coupling links, energy storage equipment and busbar structure in the park comprehensive energy system by setting modeling strategies to obtain the park comprehensive energy system constraint model;

[0207] The scenario generation module 002 is used to generate a day-ahead scenario through a scenario generation strategy and a scenario reduction strategy; generate an intraday scenario based on the forecast data of the day-ahead scenario through a scenario tree strategy; and generate a real-time scenario based on the intraday scenario through a scenario tree strategy;

[0208] The optimization model construction module 003 is used to construct a three-stage random optimization model for the day before, a two-stage random rolling optimization model for the day, and a real-time rolling optimization model taking into account CVaR according to the constraint model of the park comprehensive energy system;

[0209] The scheduling strategy acquisition module 004 is used to solve the three-stage random optimization model taking into account CVaR, the two-stage random rolling optimization model within the day, and the real-time rolling optimization model by setting a solver to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy, and the real-time stage scheduling strategy of the park's integrated energy system.

[0210] In this embodiment, in the park comprehensive energy system modeling module 001, in the process of modeling the renewable energy output equipment in the park comprehensive energy system by setting the modeling strategy, the constraint model expression of the renewable energy output equipment is:

[0211]

[0212] In the formula, The actual output of PV; is the maximum output of PV;

[0213] In the process of modeling the energy coupling link in the park comprehensive energy system by setting the modeling strategy, the energy coupling link includes: cogeneration unit, ground source heat pump, electric boiler and gas boiler; the constraint model expression of the energy coupling link is:

[0214]

[0215]

[0216] In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; It is the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices;

[0217] In the process of modeling the energy storage equipment in the park comprehensive energy system by setting the modeling strategy, the energy storage equipment includes: power storage equipment and heat storage equipment; the constraint model expression of the energy storage equipment is:

[0218]

[0219] In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment;

[0220] In the process of modeling the busbar structure in the park integrated energy system by setting the modeling strategy, the constraint model expression of the busbar structure is:

[0221]

[0222] In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

[0223] In this embodiment, in the scenario generation module 002, in the process of generating the day-ahead scenario through the scenario generation strategy and the scenario reduction strategy, a correlated sample is generated according to the distribution information of photovoltaic output, electric load and electricity price through the scenario generation strategy; based on the correlated samples, a typical scenario is clustered through the scenario reduction strategy; the scenario generation strategy is a scenario generation strategy based on CD-LHS; and the scenario reduction strategy is a scenario reduction strategy based on GMM clustering.

[0224] In this embodiment, in the optimization model construction module 003, the expression of the three-stage stochastic optimization model taking into account CVaR is:

[0225]

[0226] In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient;

[0227] The expression of the intraday two-stage stochastic rolling optimization model is:

[0228]

[0229] In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set;

[0230] The expression of the real-time rolling optimization model is:

[0231]

[0232] In the formula, A collection of time periods for the intraday scheduling model.

[0233] In this embodiment, in the scheduling strategy acquisition module 004, the expression of the day-ahead scheduling strategy is:

[0234]

[0235] In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP;

[0236] The expression of the intraday stage scheduling strategy is:

[0237]

[0238] In the formula, It is a two-stage scheduling strategy within the day; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power;

[0239] The expression of the real-time stage scheduling strategy is:

[0240]

[0241] In the formula, It is a real-time rolling scheduling strategy; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

[0242] It should be noted that the information interaction, execution process and other contents between the modules of the above-mentioned system are based on the same concept as the method embodiment in Example 1 of the present application, and the technical effects they bring are the same as those of the method embodiment of the present application. For specific contents, please refer to the description in the method embodiment shown above in the present application, and will not be repeated here.

[0243] Example 3

[0244] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code for a risk-avoiding multi-time-scale scheduling method for a park integrated energy system is stored. The program code includes instructions for executing embodiment 1 or any possible implementation method of a risk-avoiding multi-time-scale scheduling method for a park integrated energy system.

[0245] The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0246] Example 4

[0247] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0248] The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute a risk-avoiding multi-time-scale scheduling method for a park integrated energy system in accordance with Example 1 or any possible implementation thereof.

[0249] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor implemented by reading software codes stored in a memory. The memory can be integrated in the processor or can be located outside the processor and exist independently.

[0250] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable systems. The computer instructions can be stored in a computer-readable storage medium, or transmitted from a computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center.

[0251] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing system, they can be concentrated on a single computing system, or distributed on a network composed of multiple computing systems, and optionally, they can be implemented by a program code executable by a computing system, so that they can be stored in a storage system and executed by the computing system, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0252] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A risk-averse multi-time scale scheduling method for a park integrated energy system, characterized in that: include: By setting the modeling strategy, the renewable energy output equipment, energy coupling link, energy storage equipment and bus structure in the park's integrated energy system are modeled to obtain the park's integrated energy system constraint model; Generate day-ahead scenarios through scenario generation and scenario reduction strategies; Generate intraday scenarios based on the forecast data of the day-ahead scenarios through a scenario tree strategy; Generate a real-time scenario based on the intraday scenario through a scenario tree strategy; According to the comprehensive energy system constraint model of the park, a three-stage stochastic optimization model for the day before, a two-stage stochastic rolling optimization model for the day, and a real-time rolling optimization model are constructed respectively. By setting the solver to solve the three-stage day-ahead random optimization model taking into account CVaR, the two-stage intraday random rolling optimization model and the real-time rolling optimization model respectively, the day-ahead stage scheduling strategy, the intraday stage scheduling strategy and the real-time stage scheduling strategy of the park's integrated energy system are obtained.

2. According to claim 1, a risk-averse multi-time scale scheduling method for a park integrated energy system is characterized in that: In the process of modeling the renewable energy output equipment in the park comprehensive energy system by setting the modeling strategy, the constraint model expression of the renewable energy output equipment is: In the formula, The actual output of PV; is the maximum output of PV; In the process of modeling the energy coupling link in the park comprehensive energy system by setting the modeling strategy, the energy coupling link includes: cogeneration unit, ground source heat pump, electric boiler and gas boiler; the constraint model expression of the energy coupling link is: In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; Indicates the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices; In the process of modeling the energy storage equipment in the park comprehensive energy system by setting the modeling strategy, the energy storage equipment includes: power storage equipment and heat storage equipment; the constraint model expression of the energy storage equipment is: In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment; In the process of modeling the busbar structure in the park integrated energy system by setting the modeling strategy, the constraint model expression of the busbar structure is: In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

3. A risk-averse multi-time scale scheduling method for a park integrated energy system according to claim 2, characterized in that: In the process of generating the day-ahead scenario through the scenario generation strategy and the scenario reduction strategy, a sample of correlation is generated through the scenario generation strategy according to the distribution information of photovoltaic output, electric load and electricity price; a typical scenario is clustered through the scenario reduction strategy according to the sample of correlation; the scenario generation strategy is a scenario generation strategy based on CD-LHS; The scene reduction strategy is a scene reduction strategy based on GMM clustering.

4. A risk-averse multi-time scale scheduling method for a park integrated energy system according to claim 3, characterized in that: The expression of the three-stage stochastic optimization model taking into account CVaR is: In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient; The expression of the intraday two-stage stochastic rolling optimization model is: In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set; The expression of the real-time rolling optimization model is: In the formula, A collection of time periods for the intraday scheduling model.

5. A risk-averse multi-time scale scheduling method for a park integrated energy system according to claim 4, characterized in that: The expression of the day-ahead scheduling strategy is: In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP; The expression of the intraday stage scheduling strategy is: In the formula, Scheduling strategies for the intraday phase; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power; The expression of the real-time stage scheduling strategy is: In the formula, Scheduling strategy for the real-time phase; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

6. A risk-averse multi-time-scale scheduling device for a park integrated energy system, using a risk-averse multi-time-scale scheduling method for a park integrated energy system according to any one of claims 1 to 5, characterized in that: include: The park comprehensive energy system modeling module is used to model the renewable energy output equipment, energy coupling links, energy storage equipment and busbar structure in the park comprehensive energy system by setting modeling strategies to obtain the park comprehensive energy system constraint model; A scenario generation module, used to generate a day-ahead scenario through a scenario generation strategy and a scenario reduction strategy; Generate intraday scenarios based on the forecast data of the day-ahead scenarios through a scenario tree strategy; Generate a real-time scenario based on the intraday scenario through a scenario tree strategy; An optimization model construction module, for constructing a three-stage stochastic optimization model for the day before, a two-stage stochastic rolling optimization model for the day, and a real-time rolling optimization model taking into account CVaR according to the constraint model of the park comprehensive energy system; The scheduling strategy acquisition module is used to solve the three-stage random optimization model taking into account CVaR, the two-stage random rolling optimization model within the day, and the real-time rolling optimization model by setting a solver to obtain the day-ahead stage scheduling strategy, the intraday stage scheduling strategy, and the real-time stage scheduling strategy of the park's comprehensive energy system.

7. The multi-time scale scheduling device for a risk-averse park integrated energy system according to claim 6 is characterized in that: In the park comprehensive energy system modeling module, in the process of modeling the renewable energy output equipment in the park comprehensive energy system by setting the modeling strategy, the constraint model expression of the renewable energy output equipment is: In the formula, The actual output of PV; is the maximum output of PV; In the process of modeling the energy coupling link in the park comprehensive energy system by setting the modeling strategy, the energy coupling link includes: cogeneration unit, ground source heat pump, electric boiler and gas boiler; the constraint model expression of the energy coupling link is: In the formula, and is the electrical, gas and thermal power of unit j; and is the real-time adjustment of the electric, gas and thermal power of unit j; Indicates the start and stop status of the CHP unit; and H j It is the upper and lower limits of the unit output; and Δ H j It is the upper and lower limits of the unit climbing rate; is the CHP gas-to-electricity conversion efficiency; is the CHP gas-to-heat conversion efficiency; is the HP electrothermal conversion coefficient; is the EB electrothermal conversion efficiency; is GB gas-to-heat conversion efficiency; A collection of energy coupling devices; In the process of modeling the energy storage equipment in the park comprehensive energy system by setting the modeling strategy, the energy storage equipment includes: power storage equipment and heat storage equipment; the constraint model expression of the energy storage equipment is: In the formula, and It is the energy storage charging and discharging power; and The upper limit of energy storage charging and discharging power; μ j is the energy storage self-discharge rate; and W is the energy storage charging and discharging efficiency; j is the capacity of the energy storage device; and SOC j The upper and lower limits of energy storage SOC; Assemble for energy storage equipment; In the process of modeling the busbar structure in the park integrated energy system by setting the modeling strategy, the constraint model expression of the busbar structure is: In the formula, Purchase electricity for the PIES day-ahead market; Purchase electricity for the PIES intraday market; Purchase electricity for the PIES real-time market; Purchase heat for PIES; Gas purchase volume for PIES.

8. The multi-time scale scheduling device for a risk-averse park integrated energy system according to claim 7, characterized in that: In the scenario generation module, in the process of generating the day-ahead scenario through the scenario generation strategy and the scenario reduction strategy, a sample of correlation is generated through the scenario generation strategy according to the distribution information of photovoltaic output, electric load and electricity price; based on the sample of correlation, a typical scenario is clustered through the scenario reduction strategy; the scenario generation strategy is a scenario generation strategy based on CD-LHS; The scene reduction strategy is a scene reduction strategy based on GMM clustering.

9. The multi-time scale scheduling device for a risk-averse park integrated energy system according to claim 8, characterized in that: In the optimization model construction module, the expression of the three-stage stochastic optimization model taking into account CVaR is: In the formula, is the day-ahead scheduling cost; is the intraday scheduling cost; For real-time scheduling costs; The scene is in the recent scene set; It is the scene set of the day before; is the time period set of the day-ahead scheduling model; is the expected value under the day-ahead scenario set; is the CVaR under the day-ahead scenario set; Γ is the risk preference coefficient; The expression of the intraday two-stage stochastic rolling optimization model is: In the formula, It is the time period set of the intraday scheduling model; For the intraday scene set; is the expected value under the intraday scenario set; The expression of the real-time rolling optimization model is: In the formula, A collection of time periods for the intraday scheduling model.

10. A risk-averse multi-time scale scheduling device for a park integrated energy system according to claim 9, characterized in that: In the scheduling strategy acquisition module, the expression of the day-ahead scheduling strategy is: In the formula, is the day-ahead scheduling strategy; Purchase electricity for the PIES day-ahead market; Charging power for ES; is the ES discharge power; Charge power for HS; is the HS energy release power; It is the start and stop status of CHP; The expression of the intraday stage scheduling strategy is: In the formula, It is a two-stage scheduling strategy within the day; Purchase electricity for the PIES intraday market; is the CHP electric power; is the HP electrical power; is the EB electric power; is GB gas power; The expression of the real-time stage scheduling strategy is: In the formula, It is a real-time rolling scheduling strategy; Purchase electricity for the PIES real-time market; Gas purchase volume for PIES; Purchase heat for PIES; is the PV power consumption; is the CHP power adjustment value; is the HP power adjustment value; is the EB power adjustment value; It is the GB power adjustment value.

Images