Renewable energy system seasonal energy storage configuration method and equipment under ENSO event

Abstract

The invention relates to a renewable energy system seasonal energy storage configuration method and device under an ENSO event, and the method comprises the following steps: obtaining ENSO indexes and meteorological parameters under different SSP paths, inputting a renewable energy system seasonal energy storage configuration model under the ENSO event, and obtaining a renewable energy system seasonal energy storage configuration result under the ENSO event; the renewable energy system seasonal energy storage configuration model under the ENSO event comprises a prediction module, a calculation module and an optimization configuration module. The prediction module comprises a plurality of parallel sub-prediction models and is used for predicting renewable energy output and load demands in a planning year under different SSP paths according to a decomposition-coupling-synthesis mode; the calculation module is used for calculating a power gap index and a seasonal energy storage configuration requirement; and the optimal configuration module is used for substituting the calculation result into the target function and the constraint condition for solving. Compared with the prior art, the method can fully consider the influence of the ENSO event on renewable energy power generation and load demands, and effectively carries out energy storage configuration.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy storage system optimization configuration, and in particular, relates to a seasonal energy storage configuration method and device for a renewable energy system under ENSO events. Background Art

[0002] The El Nino-Southern Oscillation (ENSO) event is a phenomenon of sea-air interaction at low latitudes. It is one of the main reasons for extreme climate in recent years and may further cause the problem of seasonal supply and demand mismatch in the power system. With the shift of energy strategy to green and low-carbon concepts, a high proportion of renewable energy grid connection will become a basic feature of the future power system. However, both renewable energy generation and power system load demand have seasonal variation characteristics, which brings challenges to the effective consumption of large-scale renewable energy power.

[0003] The Chinese patent application with publication number CN117726093A discloses a method for coordinated planning of flexible supply and demand of distribution networks with high penetration of renewable energy. By considering the temporal fluctuation characteristics of net load in different seasons, introducing a flexibility deficiency rate evaluation index and establishing a multi-objective two-level planning model, it can rationalize the flexibility planning structure of the distribution network and effectively improve the anti-interference ability of the distribution network with high penetration RE. However, when renewable energy penetration is high, there is currently a lack of relevant research on how to fill the supply and demand relationship of the system affected by ENSO events. Therefore, it is urgent to design a seasonal energy storage configuration method for renewable energy systems, fully consider the impact of ENSO events on renewable energy generation and load demand, and achieve effective energy storage configuration. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a seasonal energy storage configuration method and device for a renewable energy system under ENSO events, fully consider the impact of ENSO events on renewable energy power generation and load demand, and achieve effective energy storage configuration.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] The present invention provides a seasonal energy storage configuration method for a renewable energy system under an ENSO event, comprising the following steps:

[0007] Obtain ENSO indicators and meteorological parameters under different SSP paths, input them into the seasonal energy storage configuration model of renewable energy systems under ENSO events, and obtain the seasonal energy storage configuration results of renewable energy systems under ENSO events;

[0008] Among them, the seasonal energy storage configuration model of the renewable energy system under the ENSO event includes a prediction module, a calculation module and an optimization configuration module connected in sequence; the prediction module includes a plurality of parallel sub-prediction models, which are used to predict the renewable energy output and load demand in the planning year under different SSP paths according to the decomposition-coupling-synthesis method; the calculation module is used to calculate the power gap index and the seasonal energy storage configuration demand according to the prediction results of the prediction module, the power gap index includes the electric power gap, the electric energy gap and the seasonal energy storage charging and discharging behavior index, and the seasonal energy storage configuration demand includes the power capacity demand and energy capacity demand of the seasonal energy storage configuration; the optimization configuration module is used to substitute the calculation results of the calculation module into the objective function and the constraint conditions for solving, and obtain the configuration result, the objective function is constructed based on minimizing the comprehensive cost of the planning year, the comprehensive cost of the planning year includes investment constraints and operation constraints, and the constraints are constructed based on investment constraints and operation constraints.

[0009] Furthermore, the prediction process of the prediction module is as follows:

[0010] Obtain ENSO indicators and meteorological parameters and perform multi-time scale decomposition respectively;

[0011] Each sub-forecast model uses deep learning methods to couple the decomposition results of ENSO indicators and meteorological parameters at the same time scale to generate the forecast results of renewable energy output and load demand at that time scale.

[0012] The outputs of all sub-prediction models are integrated through the inverse transformation of multi-time scale decomposition.

[0013] Furthermore, the intrinsic time scale decomposition method, the empirical mode decomposition method or the variational mode decomposition method is used to perform multi-time scale decomposition.

[0014] Furthermore, the deep learning method includes a long short-term memory network model, a generative adversarial network model and a diffusion model.

[0015] Furthermore, the calculation formula of the power shortage index is as follows:

[0016] Electric power gap:

[0017]

[0018] Among them, P gap,t is the power gap value, P L,t and P ε,t are the load and power output at time t, T τ,n is the nth time period with a time scale of τ, Ω G for renewable energy collection;

[0019] Electricity gap:

[0020]

[0021] Among them, E gap,τ,n is the energy gap value in the nth time period with a time scale of τ, t0 is the starting time of each calculation cycle, and the calculation cycle is determined according to the time scale τ of seasonal energy storage;

[0022] Seasonal energy storage charging and discharging behavior indicators:

[0023] u gap,t =sgn(P gap,t )

[0024]

[0025] Among them, u gap,t It is the charge and discharge index of the power gap, and its value is 1, -1 or 0, which means discharge, charge and no action respectively. and are the total number of times the energy storage needs to be discharged and charged in the nth time period of time scale τ, u τ,n It is an indicator of seasonal energy storage charging and discharging behavior, with values ​​of 1, -1 or 0, and the distribution represents discharging, charging and no action.

[0026] Furthermore, the calculation formula for the seasonal energy storage configuration requirement is as follows:

[0027] Power capacity requirements for seasonal energy storage configurations:

[0028]

[0029] in, The function is to find the second maximum value of x in the domain, P τ,inv,need P is the seasonal energy storage power capacity configuration demand value with a time scale of τ. τ-1,inv,need Configure the demand value for seasonal energy storage power capacity at the next smaller time scale;

[0030] Energy capacity requirements:

[0031]

[0032] Among them, E τ,inv,need The seasonal energy storage capacity configuration requirement value with a time scale of τ, E τ-1,inv,need Configure the required value for seasonal energy storage capacity at the next smaller time scale.

[0033] Furthermore, the expression of the objective function is specifically:

[0034]

[0035] Among them, C inv is the annualized investment cost, C op,j,i,s is the daily operating cost on day j of month i of year s, Y is the investment period, D i,s is the number of days in month i of year s.

[0036] Furthermore, the annualized investment cost C inv and daily operating cost C op,j,i,s The calculation formulas are as follows:

[0037]

[0038] Among them, P τ,inv and E τ,inv are the power capacity and energy capacity of the seasonal energy storage configuration with a time scale of τ, c τ,P and c τ,E are the corresponding unit power and unit energy costs respectively, r is the annual interest rate; Ω SES For seasonal energy storage collection;

[0039]

[0040] Among them, C op,t C op,j,i,s , t represents the subscript j, i, s, C τ,op,t is the operating cost of seasonal energy storage at time t with a time scale of τ, P τ,ch, t and P τ,di, t is the charge and discharge power respectively, c τ,P,ch and c τ,P,di are the unit charging and discharging operating costs, and They are respectively the charging and discharging actions of seasonal energy storage at time t with a time scale of τ 0-1 variables, and their values ​​are based on u τ,n OK, if u τ,n 1 is 1 and is 0, if u τ,n If -1 is 0 and is 1, if u τ,n If is 0, and Both are 0; C w,ab,t is the wind curtailment cost, P w,ab, t is the abandoned wind power, c w,ab is the unit wind curtailment cost; C PV,ab,t is the cost of abandoned light, P PV,ab, t is the abandoned light power, c PV,abis the unit abandoned light cost; C LS,t is the load shedding cost, P LS,t is the load shedding power, c LS is the unit load shedding cost; C ther,t is the thermal power operation cost, P ther,t is the thermal power operation power, a2, a1, a0 are the quadratic term, linear term, and constant term cost coefficients respectively, c coal is the coal consumption cost coefficient.

[0041] Furthermore, the constraint condition is specifically expressed as:

[0042] Investment constraints:

[0043]

[0044] Among them, P τ,inv,need and E τ,inv,need They are respectively…;

[0045] The operation constraints include power constraints, power constraints, wind and solar curtailment constraints, load shedding constraints, energy storage operation constraints, energy storage SOC constraints and thermal power operation constraints, and the expressions are as follows:

[0046] Power Constraints:

[0047]

[0048] Power Constraints:

[0049]

[0050] Among them, E ε,t is the energy provided by the power supply at time t, P ε,ab,t and E ε,ab,t are the power and amount of electricity abandoned by renewable energy at time t, Ω new and Ω G are new energy collection and renewable energy collection, E L,t is the load power size at time t;

[0051] Abandonment of wind and solar constraints:

[0052] 0≤P w,ab,t ≤P w,t

[0053] 0≤P PV,ab,t ≤P PV,t

[0054] Among them, P w,t and P PV,t They are respectively…;

[0055] Load shedding constraints:

[0056] 0≤P LS,t ≤P L,t

[0057] Energy storage operation constraints:

[0058]

[0059] Energy storage SOC constraints:

[0060]

[0061] Among them, SOC τ,t is the SOC state of the energy storage at time t, SOC τ,min、 SOC τ,max are the minimum and maximum values ​​of SOC respectively, are the charge and discharge efficiency respectively;

[0062] The upper and lower limits of thermal power unit output:

[0063] P ther,min ≤P ther,t ≤P ther,N

[0064] Where P ther,min is the minimum output of thermal power unit; P ther,N is the rated power of the thermal power unit.

[0065] The present invention also provides an electronic device, comprising a memory, a processor, and a program stored in the memory, wherein the processor implements the above method when executing the program.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] The present invention designs a seasonal energy storage configuration model for renewable energy systems under ENSO events, which can obtain seasonal energy storage configuration results of renewable energy systems under ENSO events according to ENSO indicators and meteorological parameters under different SSP paths. The model includes a prediction module, a calculation module and an optimization configuration module connected in sequence, wherein the prediction module includes a plurality of parallel sub-prediction models, and predicts the renewable energy output and load demand in the planned year under different SSP paths according to the decomposition-coupling-synthesis method. Specifically, the input data is first decomposed into multiple time scales, and then each sub-prediction model uses a deep learning method to couple the decomposition results of the ENSO indicators and the decomposition results of the meteorological parameters at the same time scale to generate the renewable energy output and load demand at the time scale. Demand forecast results. Finally, the outputs of all sub-forecast models are integrated through the inverse transformation of multi-time scale decomposition. This method helps to capture the changing characteristics on different time scales and improve the prediction accuracy and calculation efficiency. The calculation module is used to calculate the power gap index and seasonal energy storage configuration requirements according to the prediction results of the prediction module, and then evaluate the power gap caused by seasonal source-load imbalance from a long time scale. The optimization configuration module is used to substitute the calculation results of the calculation module into the objective function and constraints for solution to obtain the configuration results. The objective function is constructed based on minimizing the annual comprehensive cost of planning, and the constraints are constructed based on investment constraints and operation constraints. The above method fully considers the impact of ENSO events on renewable energy power generation and load demand, and can reasonably and effectively configure energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 This is a schematic diagram of the structure of the seasonal energy storage configuration model for renewable energy systems under ENSO events;

[0069] Figure 2 Schematic diagram of the weather-source / load D-CT-PC prediction method based on deep learning considering multi-time scale feature coupling;

[0070] Figure 3 Schematic diagram of the demand calculation process for seasonal energy storage configuration;

[0071] Figure 4 Schematic diagram of the model configuration and solution flow for seasonal energy storage. DETAILED DESCRIPTION

[0072] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0073] Example 1

[0074] Predicting future climate change plays an important role in enhancing human understanding of the climate system and determining social risks and response plans. At present, some scholars have predicted future climate through a scenario framework that comprehensively considers socioeconomic scenarios and climate scenarios, where socioeconomic scenarios are represented by shared socioeconomic pathways (SSPs); climate scenarios are represented by representative concentration pathways (RCPs). Phase 6 of the Coupled Model Intercomparison Project (CMIP6) provides multi-model climate forecasts based on alternative scenarios of future emissions and land use changes generated by comprehensive assessment models. The ENSO indicators mentioned in this embodiment are from the CMIP6 dataset, and the relevant meteorological parameters are from the CMIP6 and POWER datasets. In this embodiment, "SSP" refers to the "SSP-RCP coupled model" in the CMIP6 dataset, and different SSP-RCP coupled models are considered to be their corresponding climate change scenario models.

[0075] This embodiment provides a method for seasonal energy storage configuration of a renewable energy system under an ENSO event, comprising the following steps:

[0076] The ENSO indicators and related meteorological parameters under different SSP paths are obtained, and the seasonal energy storage configuration model of the renewable energy system under ENSO events is input to obtain the seasonal energy storage configuration results of the renewable energy system under ENSO events.

[0077] like Figure 1 As shown in Figure 1, the seasonal energy storage configuration model of renewable energy systems under ENSO events includes a prediction module, a calculation module, and an optimization configuration module connected in sequence. The functions and specific settings of each module are as follows:

[0078] A. Prediction Module

[0079] The prediction module includes multiple parallel sub-prediction models, which are used to predict the renewable energy output and load demand in the planning year under different SSP paths. In this regard, the present invention proposes a meteorological-source / load D-CT-PC prediction method based on deep learning considering multi-time scale feature coupling, and trains and obtains several sub-prediction models. D-CT-PC stands for "Decomposition-Coupling and Training-Prediction and Composition", such as Figure 2 As shown, specifically:

[0080] Step S1.1, decomposition of raw data: The raw data include ENSO indicators, relevant meteorological indicators and source / load data. All selected raw data are decomposed into multiple time scales to obtain component curves corresponding to multiple time scales.

[0081] Specifically, ENSO indicators include: Nino 3.4 indicator, Cumulative Sea Surface Temperature (CSST) indicator, and ENSO type indicator distinguished by occurrence location - I EP and I CP Indicators: Relevant meteorological indicators include: monthly average temperature, precipitation, surface air pressure, wind speed at 10m and downward shortwave radiation flux.

[0082] Multiple methods such as intrinsic time scale decomposition (ITD), empirical mode decomposition (EMD) or variational mode decomposition (VMD) can be used for multi-time scale decomposition. This embodiment takes the intrinsic time scale decomposition (ITD) method as an example, and then:

[0083] X(t)=L[X(t)]+H[X(t)]=LX(t)+HX(t)=L(t)+H(t)

[0084] Among them, X(t) is the original data (also called "original signal"), L(t) is the trend curve obtained after ITD decomposition (also called "baseline signal" or "basic signal"), H(t) is the component curve (also called "rotation signal component" or "component signal"), L(·) and H(·) only represent the decomposition trend function and decomposition component function of ITD decomposition on the original data. The calculation methods of the trend curve and component curve are as follows: if the set of abscissas of all poles of the original data X(t) is {t k , k=1,2,3,...} and define t0=0. If L(t) and H(t) are in the interval [0,t k ] is defined and X(t k ) in [0,t k+2 ] is defined, then:

[0085]

[0086] H(t)=X(t)-L(t)

[0087] In the above formula, α is the signal ratio, 0<α<1, usually α=0.5. At the same time, the initial value of the trend curve L(0) is set as:

[0088]

[0089] After the original data is decomposed into trend curve and component curve, the trend curve can be used as the given new original data to continue to decompose until the obtained trend curve is a monotonic signal or reaches the preset maximum decomposition level N ITD Taking L(·) and H(·) in the above formula as operators, the process can be expressed as:

[0090]

[0091] The last equation in the above formula indicates that the original data X(t) is finally processed N times. ITD Layer decomposition, HL n-1 X(t) is the ITD The nth layer component curve after layer decomposition, For N ITD Trend curve after layer decomposition.

[0092] Step S1.2, coupling-training: using a deep learning method, the component data of power output and load demand are coupled with the component data of ENSO indicators and related meteorological parameters at the corresponding time scales to obtain a prediction model composed of multiple sub-prediction models.

[0093] The deep learning methods that can be used include the Long-Short Term Memory Network (LSTMN) model, the Generative Adversarial Network (GAN) model, and the diffusion model. The specific process of this step is as follows:

[0094] Step S1.2.1, the trend data and component data of each original data are coupled and paired respectively, and the sub-data of ENSO index and related meteorological index are determined as independent variables, and the sub-data of output and load of each unit of the system are determined as dependent variables, which are input into the deep learning method, and multiple sub-prediction models are obtained through machine learning training according to the data;

[0095] Step S1.2.2, adjust the parameters of each sub-prediction model separately, select each sub-prediction model based on the effects of various evaluation indicators, and finally form an overall prediction model for this configuration method.

[0096] Step S1.3, prediction-synthesis: Input the ENSO index and related meteorological parameters under each future SSP into the prediction model to obtain the renewable energy output and load demand under a high proportion of renewable energy in the future. The specific process is as follows:

[0097] Step S1.3.1, obtain ENSO indicators and related meteorological parameters under each future SSP, perform multi-time scale decomposition respectively, and obtain component curves of each parameter under each future SSP;

[0098] Step S1.3.2, input the component curves into the prediction model, automatically couple the component curves under the corresponding decomposition levels and input them into the sub-prediction model, and use the sub-prediction model to predict the component curves of the power output and load demand of the renewable energy system in the future planning year;

[0099] Step S1.3.3, synthesize the component curves of power output and load demand into new "raw data", and the synthesis process is the reverse process of the decomposition process.

[0100] B. Computing Module

[0101] like Figure 3 As shown, the calculation module is used to calculate the power gap index and the seasonal energy storage configuration demand according to the prediction results of the prediction module. The power gap index includes the electric power gap, the electric energy gap and the seasonal energy storage charging and discharging behavior index. The seasonal energy storage configuration demand includes the power capacity demand and energy capacity demand of the seasonal energy storage configuration.

[0102] (1) Calculation of power shortage index

[0103] ①Electric power gap:

[0104]

[0105] Among them, P gap,t is the power gap value, P L,t and P ε,t are the load and power output at time t, T τ,n is the nth time period with a time scale of τ, Ω G Assemble for renewable energy.

[0106] ②Electricity gap:

[0107]

[0108] Among them, E gap,τ,n is the electric energy gap value in the nth time period with a time scale of τ, t0 is the starting time of each calculation cycle, and the calculation cycle is determined according to the time scale τ of the seasonal energy storage. If the seasonal energy storage is a monthly output, τ is 1 month (the number of days is also the specific number of days in the corresponding month of the planned year); if the seasonal energy storage is a quarterly output, τ is 1 quarter (3 months, according to the meteorological division method, March to May is spring, June to August is summer, September to November is autumn, December and January to February are winter, and the number of days is also the specific number of days in the corresponding season of the planned year).

[0109] ③ Seasonal energy storage charging and discharging behavior indicators:

[0110] u gap,t=sgn(P gap,t )

[0111]

[0112] Among them, u gap,t It is the charge and discharge index of the power gap. If it is 1, it means there is a positive power gap, the load demand is greater than the power supply capacity, and energy storage discharge is required. If it is -1, it is the opposite. 0 means there is no power gap. and are the total number of times the energy storage needs to be discharged and charged in the nth time period of time scale τ. τ,n It is an indicator of seasonal energy storage charging and discharging behavior. If the value is 1, it means that in the nth time period with a time scale of τ, the seasonal energy storage with a time scale of τ needs to be discharged; if it is -1, it means that charging is required; if it is 0, it means that the required charging and discharging times are the same. The specific need for charging or discharging or no action is not decided here, but is decided in the optimization configuration model.

[0113] (2) Calculation of seasonal energy storage requirements

[0114] This embodiment takes seasonal energy storage with monthly and quarterly output as an example to configure energy storage. Then:

[0115] ①Power capacity requirements for seasonal energy storage configuration:

[0116]

[0117] Among them, the definition The function is to find the second maximum value of x in the domain, P τ,inv,need P is the seasonal energy storage power capacity configuration demand value with a time scale of τ. τ-1,inv,need The seasonal energy storage power capacity configuration requirement value for the next smaller time scale is, in this embodiment, P 2,inv,need represents seasonal energy storage with quarterly output, P 1,inv,need Represents seasonal energy storage with monthly output.

[0118] ②Energy capacity requirements:

[0119]

[0120] Among them, E τ,inv,need The seasonal energy storage capacity configuration requirement value with a time scale of τ, E τ-1,inv,need Configure the required value for seasonal energy storage capacity at the next smaller time scale.

[0121] C. Optimize configuration module

[0122] The optimization configuration module is used to substitute the calculation results of the calculation module into the objective function and constraints for solving and obtaining the configuration results. The objective function is constructed based on minimizing the annual comprehensive cost of planning. The annual comprehensive cost of planning includes investment constraints and operation constraints. The constraints are constructed based on investment constraints and operation constraints.

[0123] (1) Objective function

[0124]

[0125] Among them, C inv is the annualized investment cost, C op,j,i,s is the daily operating cost on day j of month i of year s, Y is the investment period, D i,s is the number of days in month i of year s.

[0126] Annual investment cost C inv The calculation formula is as follows:

[0127]

[0128] Among them, P τ,inv and E τ,inv are the power capacity and energy capacity of the seasonal energy storage configuration with a time scale of τ, c τ,P and c τ,E are the corresponding unit power and unit energy costs respectively, r is the annual interest rate; Ω SES For seasonal energy storage collection.

[0129] Daily operating cost C op,j,i,s The calculation formula is as follows:

[0130]

[0131] Among them, C op,t C op,j,i,s , for the sake of simplicity, t is used to represent the subscripts j, i, and s. τ,op,t is the operating cost of seasonal energy storage at time t with a time scale of τ, P τ,ch, t and P τ,di, t is the charge and discharge power respectively, c τ,P,ch and c τ,P,di are the unit charging and discharging operating costs, and They are respectively the charging and discharging actions of seasonal energy storage at time t with a time scale of τ 0-1 variables, and their values ​​are based on u τ,n OK, if u τ,n 1 is 1 and is 0, if u τ,n If -1 is 0 and is 1, if u τ,n If is 0, and Both are 0; C w,ab,t is the wind curtailment cost, P w,ab, t is the abandoned wind power, c w,ab is the unit wind curtailment cost; C PV,ab,t is the cost of abandoned light, P PV,ab, t is the abandoned light power, c PV,ab is the unit abandoned light cost; C LS,t is the load shedding cost, P LS,t is the load shedding power, c LS is the unit load shedding cost; C ther,t is the thermal power operation cost, P ther,t is the thermal power operation power, a2, a1, a0 are the quadratic term, linear term, and constant term cost coefficients respectively, c coal is the coal consumption cost coefficient.

[0132] (2) Constraints

[0133] ① Investment constraints:

[0134]

[0135] Among them, P τ,inv,need and E τ,inv,need They are respectively…;

[0136] (2) Operation constraints

[0137] Operation constraints include power constraints, power constraints, wind and solar curtailment constraints, load shedding constraints, energy storage operation constraints, energy storage SOC constraints, and thermal power operation constraints. The expressions are as follows:

[0138] ① Power constraints:

[0139]

[0140] ②Power Constraint:

[0141]

[0142] Among them, E ε,t is the energy provided by the power output at time t, P ε,ab,t and E ε,ab,t are the power and amount of electricity abandoned by renewable energy at time t, Ω new They are new energy collection, E L,t is the load power at time t.

[0143] ③ Constraints on abandoning wind and solar power:

[0144] 0≤Pw,ab,t ≤P w,t

[0145] 0≤P PV,ab,t ≤P PV,t

[0146] Among them, P w,t and P PV,t They are respectively…

[0147] ④ Load shedding constraints:

[0148] 0≤P LS,t ≤P L,t

[0149] ⑤ Energy storage operation constraints:

[0150]

[0151] ⑥ Energy storage SOC constraints:

[0152]

[0153] Among them, SOC τ,t is the SOC state of the energy storage at time t, SOC τ,min、 SOC τ,max are the minimum and maximum values ​​of SOC respectively, are the charge and discharge efficiency respectively.

[0154] ⑦ Upper and lower limits of thermal power unit output:

[0155] P ther,min ≤P ther,t ≤P ther,N

[0156] Where P ther,min is the minimum output of thermal power unit; P ther,N is the rated power of the thermal power unit.

[0157] The operating cost of the thermal power unit in the objective function is a convex quadratic function and includes some integer variables. Therefore, the model is a mixed integer quadratic programming problem (MIQP). A suitable commercial solver is selected for calculation and solution. After the calculation is completed, it is determined whether all SSP scenarios have been solved: if so, the process ends; if not, the SSP path is changed and the solution continues. The schematic diagram of the solution process is shown in the figure. Figure 4 shown.

[0158] In the context of ENSO events, the above method takes into account the impact of ENSO event indicators and related meteorological indicators on renewable energy generation and load demand, and then evaluates the power gap caused by seasonal source-load imbalance from a long-term scale, providing support for subsequent reasonable configuration. The decomposition method and deep learning method in the meteorological-source / load D-CT-PC prediction method based on deep learning considering multi-time scale feature coupling are replaceable items, and the decomposition method and deep learning method can be flexibly replaced according to the existing computer vision industry technology development. Historical operating data can be used for simulation analysis and verification to prove the applicability of the proposed indicators, and the configuration requirements and corresponding configuration results under different SSP paths are calculated by simulation, providing a seasonal energy storage configuration scheme for a high proportion of renewable energy systems under the context of ENSO events for the optimal configuration of energy storage systems.

[0159] Example 2

[0160] This embodiment provides an electronic device, including a memory and a processor, wherein the processor is used to execute a program stored in the memory, wherein the program includes a number of instructions and can execute all or part of the steps of the method described in Example 1. The memory includes a computer-readable storage medium, which can be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0161] The above description of the embodiments is to facilitate the understanding and use of the invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments, and improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the present invention should be within the scope of protection of the present invention.

Claims

1. A seasonal energy storage configuration method for renewable energy systems under ENSO events, characterized in that: The following steps are involved: Obtain ENSO indicators and meteorological parameters under different SSP paths, input them into the seasonal energy storage configuration model of renewable energy systems under ENSO events, and obtain the seasonal energy storage configuration results of renewable energy systems under ENSO events; Among them, the seasonal energy storage configuration model of the renewable energy system under the ENSO event includes a prediction module, a calculation module and an optimization configuration module connected in sequence; the prediction module includes a plurality of parallel sub-prediction models, which are used to predict the renewable energy output and load demand in the planning year under different SSP paths according to the decomposition-coupling-synthesis method; the calculation module is used to calculate the power gap index and the seasonal energy storage configuration demand according to the prediction results of the prediction module, the power gap index includes the electric power gap, the electric energy gap and the seasonal energy storage charging and discharging behavior index, and the seasonal energy storage configuration demand includes the power capacity demand and the energy capacity demand of the seasonal energy storage configuration; the optimization configuration module is used to substitute the calculation results of the calculation module into the objective function and the constraint conditions for solving, and obtain the configuration result, the objective function is constructed based on minimizing the comprehensive cost of the planning year, the comprehensive cost of the planning year includes investment constraints and operation constraints, and the constraints are constructed based on investment constraints and operation constraints.

2. A seasonal energy storage configuration method for a renewable energy system under an ENSO event according to claim 1, characterized in that: The prediction process of the prediction module is as follows: Obtain ENSO indicators and meteorological parameters and perform multi-time scale decomposition respectively; Each sub-forecast model uses deep learning methods to couple the decomposition results of ENSO indicators and meteorological parameters at the same time scale to generate the forecast results of renewable energy output and load demand at that time scale. The outputs of all sub-prediction models are integrated through the inverse transformation of multi-time scale decomposition.

3. A seasonal energy storage configuration method for a renewable energy system under ENSO events according to claim 2, characterized in that: Multi-time scale decomposition is performed using the intrinsic time scale decomposition method, the empirical mode decomposition method or the variational mode decomposition method.

4. The seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 2, characterized in that: The deep learning method includes a long short-term memory network model, a generative adversarial network model and a diffusion model.

5. The seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 1, characterized in that: The calculation formula of the electricity gap index is as follows: Electric power gap: Among them, P gap,t is the power gap value, P L,t and P ε,t are the load and power output at time t, T τ,n is the nth time period with a time scale of τ, Ω G for renewable energy collection; Electricity gap: Among them, E gap,τ,n is the energy gap value in the nth time period with a time scale of τ, t0 is the starting time of each calculation cycle, and the calculation cycle is determined according to the time scale τ of seasonal energy storage; Seasonal energy storage charging and discharging behavior indicators: u gap,t =sgn(P gap,t ) Among them, u gap,t It is the charge and discharge index of the power gap, and its value is 1, -1 or 0, which means discharge, charge and no action respectively. and are the total number of times the energy storage needs to be discharged and charged in the nth time period of time scale τ, u τ,n It is an indicator of seasonal energy storage charging and discharging behavior, with values ​​of 1, -1 or 0, and the distribution represents discharging, charging and no action.

6. A seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 5, characterized in that: The calculation formula for the seasonal energy storage configuration requirement is as follows: Power capacity requirements for seasonal energy storage configurations: in, The function is to find the second maximum value of x in the domain, P τ,inv,need P is the seasonal energy storage power capacity configuration demand value with a time scale of τ. τ-1,inv,need Configure the demand value for seasonal energy storage power capacity at the next smaller time scale; Energy capacity requirements: Among them, E τ,inv,need The seasonal energy storage capacity configuration requirement value with a time scale of τ, E τ-1,inv,need Configure the required value for seasonal energy storage capacity at the next smaller time scale.

7. The seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 1, characterized in that: The expression of the objective function is specifically: Among them, C inv is the annualized investment cost, C op,j,i,s is the daily operating cost on day j of month i of year s, Y is the investment period, D i,s is the number of days in month i of year s.

8. A seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 7, characterized in that: Annual investment cost C inv and daily operating cost C op,j,i,s The calculation formulas are as follows: Among them, P τ,inv and E τ,inv are the power capacity and energy capacity of the seasonal energy storage configuration with a time scale of τ, c τ,P and c τ,E are the corresponding unit power and unit energy costs respectively, r is the annual interest rate; Ω SES For seasonal energy storage collection; Among them, C op,t C op,j,i,s , t represents the subscript j, i, s, C τ,op,t is the operating cost of seasonal energy storage at time t, P τ,ch, t and P τ,di, t is the charge and discharge power respectively, c τ,P,ch and c τ,P,di are the unit charging and discharging operating costs, and They are respectively the charging and discharging actions of seasonal energy storage at time t with a time scale of τ 0-1 variables, and their values ​​are based on u τ,n OK, if u τ,n 1 is 1 and is 0, if u τ,n If -1 is 0 and is 1, if u τ,n If is 0, and Both are 0; C w,ab,t is the wind curtailment cost, P w,ab, t is the abandoned wind power, c w,ab is the unit wind curtailment cost; C PV,ab,t is the cost of abandoned light, P PV,ab, t is the abandoned light power, c PV,ab is the unit abandoned light cost; C LS,t is the load shedding cost, P LS,t is the load shedding power, c LS is the unit load shedding cost; C ther,t is the thermal power operation cost, P ther,t is the thermal power operation power, a2, a1, a0 are the quadratic term, linear term, and constant term cost coefficients respectively, c coal is the coal consumption cost coefficient.

9. A seasonal energy storage configuration method for renewable energy systems under ENSO events according to claim 8, characterized in that: The expression of the constraint condition is specifically: Investment constraints: The operation constraints include power constraints, power constraints, wind and solar curtailment constraints, load shedding constraints, energy storage operation constraints, energy storage SOC constraints and thermal power operation constraints, and the expressions are as follows: Power Constraints: Power Constraints: Among them, E ε,t is the energy provided by the power supply at time t, P ε,ab,t and E ε,ab,t are the power and amount of electricity abandoned by renewable energy at time t, Ω new and Ω G are new energy collection and renewable energy collection, E L,t is the load power size at time t; Abandonment of wind and solar constraints: 0≤P w,ab,t ≤P w,t 0≤P PV,ab,t ≤P PV,t Among them, P w,t and P PV,t are the output of wind power and photovoltaic power at time t respectively; Load shedding constraints: 0≤P LS,t ≤P L,t Energy storage operation constraints: Energy storage SOC constraints: Among them, SOC τ,t is the SOC state of the energy storage at time t, SOC τ,min , SOC τ,max are the minimum and maximum values ​​of SOC respectively, are the charge and discharge efficiency respectively; Output upper and lower limits of thermal power units: P ther,min ≤P ther,t ≤P ther,N Where P ther,min is the minimum output of thermal power unit; P ther,N is the rated power of the thermal power unit.

10. An electronic device comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images