A Source-Storage Collaborative Planning and Configuration Method with Temporal Similarity Constraints of Source and Load

Abstract

The present invention discloses a source-storage collaborative planning and configuration method with source-load time-series similarity constraints, which relates to the technical field of power system unit planning. Based on the CE theory, the present invention proposes a source-load time-series similarity index and establishes a constraint set model for source-storage collaborative planning, and optimally configures the wind-solar-storage system with the objectives of minimizing the daily sharing cost, emission reduction cost and the operating cost of each unit. The source-storage collaborative planning method based on source-load similarity proposed by the present invention can effectively suppress the fluctuations of conventional units and effectively improve the new energy consumption capacity.

Description

Technical Field

[0001] The present invention relates to the technical field of power system unit planning, and in particular to a source-storage collaborative planning and configuration method with source-load time series similarity constraints, which is used for industrial parks with wind-solar-thermal-storage combined power generation systems. Background Art

[0002] In recent years, due to the increasing penetration rate of new energy in the power system year by year, the installed capacity of new energy has continued to rise. However, there are also serious problems of wind and light abandonment, resulting in the phenomenon of "installed capacity without output" of new energy power. Due to the randomness of power load and new energy output, the peak regulation capacity of conventional units is insufficient, and the operation mode of wind and light abandonment or load shedding is usually adopted. Simply relying on the frequent adjustment of the output of conventional units is difficult to meet the requirements of the safe operation of the system, and it is also difficult for the units to operate economically. How to maximize the regulation potential of sources, loads and storage under the premise of meeting the safety and reliability of the power system is an urgent problem. How to tap the regulation potential of the source-grid-load to achieve multi-energy complementary regulation is one of the key links in source-storage collaborative planning. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a source-storage collaborative planning and configuration method with source-load time series similarity constraints, which can effectively improve the new energy consumption capacity and facilitate meeting the economic requirements of investment and construction while achieving the regulation objectives of the power system.

[0004] To solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A source-storage collaborative planning and configuration method with source-load time series similarity constraints includes the following steps:

[0006] Step 1: Discount the initial construction costs C RE,n of wind and photovoltaic power plants and the initial construction cost C ESS,n of the energy storage system into daily amortized costs, and construct a multi-objective function model by combining the operation costs of conventional units, environmental emission reduction costs, and wind and light abandonment penalty costs;

[0007] Step 2: According to the source-load time series curve, establish a source-load time series similarity index based on the CE theory, and construct a constraint set model, including CE source-load time series similarity constraints, conventional unit ramp constraints, wind and light output constraints, etc.;

[0008] Step 3: Set up an energy storage compensation system to suppress wind and light ramp events, and construct an energy storage control strategy plan;

[0009] Step 4: Establish a two-layer planning model composed of an investment decision-making layer and a simulation operation layer. The investment decision-making layer determines the installed capacity of wind, light and storage, and the simulation operation layer determines the output of each unit, and uses the ideal point algorithm to achieve optimal planning and configuration.

[0010] A further improvement of the technical solution of the present invention lies in: in the multi-objective function model constructed in the step 1, the initial construction cost of the wind, light and their energy storage compensation systems is allocated to the daily allocated cost through the discount rate:

[0011] The energy storage compensation system mainly includes: an energy storage power module and an energy storage capacity module. The daily allocated cost of the energy storage compensation system is as follows:

[0012]

[0013] Where

[0014]

[0015] In the formula, C ESS,n is the investment cost of the energy storage system, S ESS,η and E ESS,η are the power capacity and energy capacity of the energy storage in the ηth year, α PCS,η and α ESU,η are the unit capacity construction costs thereof respectively; x ESS,η is whether to allocate the initial investment cost, and N ESS is the service life of the energy storage unit.

[0016] A further improvement of the technical solution of the present invention lies in: in the step A2, a source-load time-series similarity index is established based on the CE theory, and a constraint set model is constructed, including: installed capacity constraint, power quantity constraint, energy policy constraint, power balance constraint, wind and light output constraints, conventional unit output constraints, ramp rate constraint, start-stop time constraint, demand-side response constraint, energy storage output constraint, reverse peak shaving safety constraint, CE similarity constraint:

[0017] Among them, the energy policy constraint:

[0018]

[0019] In the formula, the numerator represents the total power generation of various renewable power sources in the nth year, and the denominator represents the total power generation of various power sources in the nth year; β re,n represents the lower limit value of the proportion of clean power generation required by the policy in the nth year.

[0020] A further improvement of the technical solution of the present invention lies in: in the step 2, the parameter estimation of the bivariate Copula function is adopted, and the parameter estimation of the bivariate Copula function adopts the Kendall τ similarity coefficient estimation method and the Monte Carlo method is used to simulate the entropy of the Copula function; the steps are as follows:

[0021] 2.1. Kendall τ similarity coefficient estimation method

[0022] The relationship between the Copula parameter and the Kendall τ similarity coefficient is as follows

[0023]

[0024] Estimate the parameters of the corresponding type of Copula function according to the Kendall τ similarity coefficient, and obtain the Copula probability density function c(u, v);

[0025] 2.2. Monte Carlo method for simulating Copula entropy

[0026] Let U ∈ [0, 1], i = 1, …, d, and the Copula entropy is expressed as:

[0027]

[0028] Therefore, estimating CE is converted to estimating the expectation of ln[c(U)] using the Monte Carlo method;

[0029] The reverse peak shaving safety constraint is mainly based on the reverse peak shaving characteristics of wind and light. During peak load periods, the reverse peak shaving capacity of the system is large and there is already a sufficient safety margin. However, during low load periods, the reverse peak shaving capacity is small and there is a risk of safety margin. Therefore, during low load periods, a safety margin constraint is defined to ensure that conventional units have a certain peak shaving space at any time to cope with the volatility of new energy output;

[0030]

[0031] In the formula: P Gi,min (t) is the minimum output of conventional unit i at time t; β aq is the upper limit value of the safety margin constraint; T l is the low load period;

[0032] The CE constraint reflects the similarity between the source-load output curves, and the power system can refer to this constraint for reasonable wind and light resource scheduling;

[0033]

[0034] In the formula: H w (t) is the time-series Copula entropy of wind-load at time t; H pv (t) is the time-series Copula entropy of light-load; P ess (t) is the power of the energy storage system at time t; P w (t), P pv (t) are the wind and light powers at time t; is the predicted wind and light power at time t.

[0035] A further improvement of the technical solution of the present invention lies in: the energy storage control strategy proposed in step 3:

[0036] Define the ramp rate constraints for wind and solar power

[0037]

[0038] Control the operation state of the energy storage system according to the magnitude of the output volatility of wind and solar power; specifically as follows:

[0039] 3.1. Energy storage does not operate

[0040] When At this time

[0041] P ess (t) = 0

[0042] In the formula, P′ m (t - Δt) is the output power of the m-th type of new energy after being smoothed at the previous moment; δ m Is the upper limit of the ramp rate of the m-th type of new energy; this formula indicates that the ramp rate of the system's new energy cannot exceed δ m Times its installed capacity;

[0043] 3.2. Energy storage discharges. At this time, the output of new energy suddenly decreases, and the ramp rate δ < -δ m Exceeds the limit, and the energy storage system discharges to smooth the output volatility;

[0044] When At this time

[0045] P ess (t) = η ess [(δ + δ m )S m + P′ m (t - Δt)]

[0046] In the formula: η ess Is the charge-discharge efficiency of the energy storage system;

[0047] C3. Energy storage charges. At this time, the output of new energy suddenly increases, and the ramp rate δ > δ m Exceeds the limit, and the energy storage system discharges to smooth the output volatility;

[0048] When At this time

[0049] P ess (t) = η ess [(δ m - δ)S m - P′ m (t - Δt)].

[0050] A further improvement of the technical solution of the present invention lies in: the ideal point algorithm proposed in step 4 is used to solve the bi-level programming model:

[0051] First, calculate the optimal solutions of each objective function, and then use this value as the expected value to transform the multi-objective optimization model into seeking the solution closest to the expected value among all solution sets, that is, transform it into a single-objective optimization model with the minimum sum of the deviation amounts between each objective function and a i The specific solution process is as follows:

[0052] In the first step, for the multi-objective problem of the bilevel programming, calculate the optimal values of each single objective function first

[0053]

[0054] In the second step, since it is difficult to have an x 0 that minimizes all objective functions, a distance norm ||·|| is introduced in the feasible region X;

[0055]

[0056] where represents the importance degree of different objective functions for the global solution and satisfies Based on this, the optimal solution of the multi-objective optimization is determined.

[0057] Due to the adoption of the above technical solutions, the technical progress achieved by the present invention is as follows:

[0058] The present invention meets the economic requirements of investment and construction while achieving the power system regulation objectives, maximally absorbs new energy, and optimizes CO with low importance 2 emissions and curtailment rates of wind and light; the present invention ensures the constraint on the coordination degree of the source-load curve and simultaneously maintains the evaluation index for the similarity relationship between each processing curve. The present invention ensures that the regulation potential of the source-network-load is maximally exploited and the frequent start-stop and ramping of conventional units are reduced. The load curve is expected to follow the new energy curve, that is, the source-load achieves a high similarity. The present invention realizes the maximum absorption of new energy, reduces the curtailment of wind and light, and makes the output curve of conventional units stable. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 is the collaborative planning diagram applicable to the present invention;

[0060] Figure 2 is the operation flow chart of the present invention;

[0061] Figure 3 is the CE similarity source-load time series diagram of the present invention;

[0062] Figure 4 is the conventional optimization operation diagram of the present invention;

[0063] Figure 5 is the coordinated optimization operation diagram of the source-load-storage of the present invention;

[0064] Figure 6 It is the source-load-storage CE combined operation diagram of the present invention. Specific embodiments

[0065] The present invention will be further described in detail below in conjunction with embodiments:

[0066] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present application.

[0067] The present invention is a source-storage collaborative planning and configuration method with source-load time-series similarity constraints, and is actually a source-storage collaborative planning and configuration method based on the source-load time-series similarity of Copula Entropy (CE). The present invention is applied to an industrial park including a wind-solar-thermal-storage combined power generation system, as Figure 1 shown. The conventional unit is 800 MW, as well as a to-be-built wind farm, a photovoltaic power station, and a energy storage compensation system. With a 30-minute scheduling period, the source-load time series is simulated for 8,760 hours throughout the year, and the optimal planning scheme is solved according to the ideal point method.

[0068] Figure 2 is the operation flow chart of the present invention, which is the operation flow chart of an embodiment of the temperature control load demand response control method based on cluster status statistical information. The method includes the following steps:

[0069] Step 1. In the investment decision-making layer, the initial construction costs C RE,n of the wind and photovoltaic power plants and the initial construction cost C ESS,n of the energy storage system are discounted to the daily sharing cost;

[0070] Step 2. In the simulation operation layer, based on the source-load similarity index, a CE similarity constraint is established according to the source-load time series curve, and constraints such as the ramp constraint of the conventional unit and the wind-solar output constraint are established by combining the operating characteristics of each unit to establish a constraint set model; to suppress the wind-solar ramp event, an energy storage compensation system is set up to construct an energy storage control strategy scheme;

[0071] Step 3. Multiple objective functions such as the daily sharing cost, the operating cost of the conventional unit, the environmental emission reduction cost, and the penalty cost for wind and light abandonment are established, and the optimal planning and configuration scheme is obtained by using the ideal point algorithm;

[0072] Step 4: Establish a two - layer programming model consisting of an investment decision - making layer and a simulation operation layer. The investment decision - making layer determines the installed capacity of wind, light, and energy storage, and the simulation operation layer determines the output of each unit. The ideal point algorithm is used to achieve the optimal planning configuration.

[0073] For Step 1, in the source - storage collaborative planning model based on the similarity of source - load - storage output curves considering the time - series characteristics of source and load, first is the constraint set of the collaborative planning module:

[0074] 1.1 Installed capacity constraint:

[0075]

[0076] In the formula, G N is the number of thermal power units; S G,n represents the installed capacity of the conventional unit in the nth year, S m,n represents the installed capacity of the m - th type of renewable energy source in the nth year, P lmax,n represents the maximum load in the nth year; R L,n is the reserve coefficient, which includes the additional reserve capacity added to the system to cope with the volatility of new - energy output.

[0077] 1.2 Electricity quantity constraint:

[0078]

[0079] In the formula, n is the number of years; G N,n is the number of thermal power units; S Gi,n is the capacity of the i - th thermal power unit; H Gi,n and H m,n are the utilization hours of the i - th thermal power unit and the m - th type of renewable energy source respectively; H l,n is the load utilization hours; R D,n is the reserve coefficient.

[0080] 1.3 Energy policy constraint:

[0081]

[0082] In the formula, the numerator represents the total power generation of various renewable energy sources in the nth year, and the denominator represents the total power generation of various power sources in the nth year; β re,n represents the lower limit value of the proportion of clean - power generation required by the policy in the nth year.

[0083] The multi - objective function model constructed in Step 1 apportions the initial construction cost of the wind, light, and their energy - storage compensation systems into a daily apportionment function through the discount rate:

[0084]

[0085] Among them

[0086]

[0087] In the formula, C RE,n is the investment cost of the renewable energy system, M represents the type of renewable power source (wind, solar, etc.), and S m,η is the installed capacity of the newly built m-type renewable energy in the ηth year, and α m,η and x m,η are respectively the construction cost per unit capacity of the m-type renewable power source in the ηth year and whether the initial investment cost needs to be shared, and N m is the service life of each type of power source; represents the discount rate.

[0088]

[0089] Among them

[0090]

[0091] In the formula, C ESS,n is the investment cost of the energy storage system, S ESS,η and E ESS,η are the power capacity and energy capacity of the energy storage in the ηth year, and α PCS,η and α ESU,η are respectively their construction costs per unit capacity; x ESS,η is whether the initial investment cost needs to be shared, and N ESS is the service life of the energy storage unit.

[0092] In step 2, in the source-load time series simulation operation module, the constraints include CE constraints, the economic constraint sets of each unit, and the energy storage control strategy scheme.

[0093] Let x ∈ R d be a d-dimensional random variable, where U i is a random variable subject to a uniform distribution, and u i is the specific value of the random variable U i . The entropy calculation expression of the Copula function is:

[0094]

[0095] In the formula, c(u 1 , u 2 , …, u d ) is the probability density function of the Copula function.

[0096] In view of the problem of difficult calculation of multiple integrals due to the large number of variables and high-dimensional data such as the source-load curve, the parameter estimation of the bivariate Copula function adopts the Kendall τ similarity coefficient estimation method and uses the Monte Carlo method to simulate the entropy of the Copula function. The steps are as follows:

[0097] 2.1. Kendall τ Similarity Coefficient Estimation Method

[0098] The relationship between Copula parameters and Kendall τ similarity coefficient is as follows

[0099]

[0100] Estimate the parameters of the corresponding type of Copula function and obtain the Copula probability density function c(u, v);

[0101] 2.2. Monte Carlo Method for Simulating Copula Entropy

[0102] Let U ∈ [0, 1], i = 1, …, d, and the Copula entropy is expressed as:

[0103]

[0104] Therefore, estimating CE is converted to estimating the expectation of ln[c(U)] using the Monte Carlo method. The CE source-load time series similarity diagram in the example is shown in Figure 3 as follows

[0105] The economic constraint sets of each unit include:

[0106] 1. Power Balance Constraint

[0107]

[0108] In the formula: P Gi (t), P m (t), P ess (t), P l (t) are the power of the conventional unit, the m-th type of new energy, energy storage charge and discharge, and load at time t, respectively; x i (t) is the start-stop state variable of the conventional unit

[0109] 2. Reverse Peak-Regulation Safety Constraint

[0110] During large load periods, the reverse peak-regulation capacity of the system is large and there is already an adequate safety margin. However, during small load periods, the reverse peak-regulation capacity is small and there is a risk in the safety margin. Therefore, during small load periods, safety margin constraints are defined to ensure that conventional units have a certain peak-regulation space at any time to cope with the volatility of wind power output

[0111]

[0112] In the formula: P Gi,min (t) is the minimum output of the conventional unit i at time t; β aq is the upper limit value of the safety margin constraint; T l is the low load period

[0113] 3. The CE constraint reflects the similarity between the output curves of the power source and load. The power system can refer to this constraint for reasonable scheduling of wind and solar resources.

[0114]

[0115] In the formula: H w (t) is the time-series Copula entropy of wind-load at time t; H pv (t) is the time-series Copula entropy of light-load; P ess (t) is the power of the energy storage system at time t; P w (t), P pv (t) are the wind and solar powers at time t; is the predicted wind and solar power at time t.

[0116] 4. Wind and solar output constraints

[0117]

[0118] In the formula is the predicted output value of the m-th type of new energy.

[0119] 5. Output constraints, ramping constraints, start-stop time constraints of conventional units

[0120] x h,i P G,i,h,min ≤P G,i,h ≤x h,i P G,i,h,max (14)

[0121] P i- ≤P G,i,h -P G,i,h-1 ≤P i+ (15)

[0122] (x h-1,i -x h,i )(T G,i,h-1 -T G,i,on )≥0 (16)

[0123]

[0124] In the formula, P G,i,h,min and P G,i,h,max are the minimum and maximum outputs of thermal power unit i within time period h, respectively; P i- and P i+ are the downward and upward ramping rates of thermal power unit i, respectively; T G,i,on and T G,i,off are the minimum continuous operation and shutdown times of conventional unit i, respectively; T G,i,h-1is the duration for the thermal power unit i to maintain the same state (running or shutdown) before the (h-1)th period.

[0125] 6. Demand-side response constraint

[0126] 1) Single-scheduling-cycle load response constraint

[0127] P Y min (t) ≤ ΔP(t) ≤ P Y max (t) (18)

[0128] In the formula, P Y min (t) and P Y max (t) represent the upper and lower limits of the demand-side response load at time t.

[0129] 2) Load response capacity constraint

[0130]

[0131] In the formula, S P min and S P max represent the upper and lower limits of the demand-side response capacity within the entire scheduling cycle T. This constraint mainly reflects that the electricity load should be stable within the entire scheduling cycle, and the increase or reduction of the demand-side response should be within a certain capacity.

[0132] The energy storage control strategy plan includes:

[0133] Define the wind and light ramp constraints

[0134]

[0135] Control the operation state of the energy storage system according to the magnitude of the wind and light output volatility. Specifically as follows:

[0136] 1. Energy storage does not operate

[0137] When At this time

[0138] P ess (t) = 0 (21)

[0139] In the formula, P′ m (t - Δt) is the output power of the mth type of new energy after being leveled at the previous moment; δ m is the upper limit value of the ramp of the mth type of new energy. This formula indicates that the ramp rate of the new energy in the system cannot exceed δ m times its installed capacity.

[0140] 2. Energy storage discharges. At this time, the output of the new energy suddenly decreases, and the ramp rate δ < -δ m exceeds the limit, and the energy storage system discharges to level the output volatility.

[0141] When When

[0142] P ess (t) = η ess [(δ + δ m )S m + P′ m (t - Δt)] (22)

[0143] Where: η ess is the charge - discharge efficiency of the energy storage system.

[0144] 3. Energy storage charging. At this time, the output of new energy suddenly increases, and the ramp rate δ > δ m exceeds the limit. The energy storage system discharges to suppress the volatility of the output.

[0145] When When

[0146] P ess (t) = η ess [(δ m - δ)S m - P′ m (t - Δt)] (23)

[0147] Energy storage output constraint

[0148] - S ESS ≤ P ess (t) ≤ S ESS (24)

[0149] δ min E ESS ≤ soc(t) ≤ δ max E ESS (25)

[0150] Where soc(t) is the state of charge of the energy storage system at time t. δ min and δ max represent the upper and lower limits of the energy capacity of the shallow charge - shallow discharge strategy adopted to protect the service life of the energy storage system.

[0151] The overall objective function is:

[0152] 1. Conventional unit operation cost

[0153]

[0154] Where C 1 is the conventional unit operation cost, a i , b i , c i are the consumption coefficients of conventional unit i; α i , β i , τi is the start - stop cost coefficient of the conventional unit i; τ is the unit shutdown time.

[0155] 2. Carbon emission reduction effect

[0156]

[0157] In the formula, C 2 mainly considers the effect brought by the emission of CO 2 emissions; is the amount of CO 2 emitted by the conventional unit and the m - th type of new energy; F G , F m are the carbon footprint emission factors of the conventional unit and the m - th type of new energy; R G is the start - stop coal consumption of the conventional unit; Q m (t) is the power generation of the conventional unit and the m - th type of new energy at time t

[0158] 3. Penalty for wind and light curtailment

[0159]

[0160] In the formula, C 3 is the penalty for wind and light curtailment; is the predicted power generation of the m - th type of new energy at time t.

[0161] Due to the source - load time - series correlation constraint and the energy storage control strategy scheme which includes maximizing the consumption of wind and light, the wind and light curtailment rate and the CO 2 emissions can be reduced. Therefore, on the premise of ensuring the economic operation of the system, maximizing the consumption of new energy, the ideal point method optimizes the system operation cost with high importance and optimizes the CO 2 emissions and the wind and light curtailment rate with low importance.

[0162] To sum up, the example of the present invention selects three operation modes. Operation mode A: The operation mode where the dispatching resources only consider the conventional units; Operation mode B: The source - load - storage coordinated operation mode where the dispatching resources consider the conventional units, demand - side response and energy storage system; Operation mode C: The CE joint operation mode where the dispatching resources consider the conventional units, demand - side response and energy storage system;

[0163] The planning scheme implemented by applying the present invention is as follows:

[0164] Table 1 Optimal planning scheme

[0165]

[0166] Under typical days, the diagrams of the three operation modes A, B, and C are as shown in Figures 4 - 6 , and the operation costs are shown in Table 2.

[0167] Table 2 Cost of three operation modes

[0168]

[0169] From the comparison between Table 1 and Table 2, it can be seen that although the proposed CE source-load time series similarity constraint increases the demand-side response cost, it effectively reduces the installed capacity of the energy storage system, with the lowest daily apportionment cost, obvious carbon emission reduction effect, and the lowest daily cost. It effectively solves the problem of system absorption of new energy.

[0170] The difference from the existing source load planning method is:

[0171] Existing research is mostly focused on the operation side, and has established overall or local coordinated optimization models of sources, grids, loads and storage. Often, the objective function only considers the overall economy or the safety and environmental protection of the power system, but lacks constraints on the degree of coordination between source and load curves. At the coordination optimization level, existing evaluation methods are mostly focused on economic benefits, and lack evaluation indicators for the similarity relationship between each processing curve.

[0172] The present invention focuses on the planning layer. In order to ensure the maximum potential of source-grid-load regulation and reduce the frequent start-stop and ramp-up of conventional units, the load curve is expected to achieve high similarity with the new energy curve, that is, the source-load curve. The present invention proposes a CE similarity measurement index to achieve the maximum absorption of new energy, reduce wind and solar power abandonment, and make the output curve of conventional units stable.

Claims

1. A source-storage collaborative planning and configuration method with source-load time-series similarity constraints, characterized in that: It includes the following steps: Step 1: Discount the initial construction costs C RE,n of the wind and photovoltaic power fields and the initial construction cost C ESS,n of the energy storage system into daily amortized costs, and construct a multi-objective function model by combining the operating costs of conventional units, environmental emission reduction costs, and wind and photovoltaic curtailment penalty costs; Step 2: According to the source-load time series curve, establish a source-load time series similarity index based on the CE theory, and construct a constraint set model, including CE source-load time series similarity constraints, conventional unit ramp constraints, wind and solar power output constraints, etc.; in the source-load time series simulation operation module, the constraints include CE constraints, the economic constraint set of each unit, and the energy storage control strategy scheme; let \(x\in R\) d be a d-dimensional random variable, where \(U\) i is a random variable subject to a uniform distribution, and \(u\) i is the specific value of the random variable \(U\) i . The entropy calculation expression of the Copula function is as follows: where c(u 1 , u 2 , …, u d ) is the probability density function of the Copula function; Adopt the parameter estimation of the bivariate Copula function. The parameter estimation of the bivariate Copula function uses the Kendallτ similarity coefficient estimation method and uses the Monte Carlo method to simulate the Copula function entropy. The steps are as follows: 2.

1. Kendallτ similarity coefficient estimation method The relationship between the Copula parameter and the Kendallτ similarity coefficient is as follows Estimate the parameters of the corresponding type of Copula function according to the Kendallτ similarity coefficient, and obtain the Copula probability density function c(u, v); 2.

2. Monte Carlo method to simulate Copula entropy Let U ∈ [0, 1], i = 1, …, d, and the Copula entropy is expressed as: Therefore, estimating CE is converted to estimating the expectation of ln[c(U)] using the Monte Carlo method; The CE constraint reflects the similarity between the source-load output curves. The power system can refer to this constraint for reasonable wind-solar resource scheduling: Where, H w (t) is the time series Copula entropy of wind-load at time t; H pv (t) is the time series Copula entropy of light-load; P ess (t) is the power of the energy storage system at time t; P w (t), P pv (t) are the wind and light powers at time t; is the predicted wind and light power at time t; Step 3: To suppress wind-solar ramping events, establish a energy storage compensation system and construct an energy storage control strategy plan; Step 4: Establish a two-layer planning model consisting of an investment decision-making layer and a simulation operation layer. The investment decision-making layer determines the installed capacity of wind-solar-storage, and the simulation operation layer determines the output of each unit, and uses the ideal point algorithm to achieve the optimal planning and configuration.

2. A source-storage collaborative planning and configuration method with source-load time-series similarity constraints according to claim 1, characterized in that: In the multi-objective function model constructed in the step 1, the initial construction costs of wind, solar and their energy storage compensation systems are allocated as daily allocated costs through the discount rate: The energy storage compensation system mainly includes: an energy storage power module and an energy storage capacity module. The daily allocated cost of the energy storage compensation system is as follows: Where Where C ESS,n is the investment cost of the energy storage system, S ESS,η and E ESS,η are the power capacity and energy capacity of the energy storage in the η-th year, α PCS,η and α ESU,η are the unit capacity construction costs thereof respectively; x ESS,η is whether the initial investment cost needs to be amortized, and N ESS is the service life of the energy storage unit.

3. A source-storage collaborative planning and configuration method with source-load time-series similarity constraints according to claim 1, characterized in that: In the step 2, a source-load time-series similarity index is established based on the CE theory, and a constraint set model is constructed, including: installed capacity constraint, power quantity constraint, energy policy constraint, power balance constraint, wind and solar output constraints, conventional unit output constraints, ramping constraint, start-stop time constraint, demand-side response constraint, energy storage output constraint, reverse peak shaving safety constraint, CE similarity constraint: Among them, the energy policy constraint: In the formula, the numerator represents the total power generation of various renewable power sources in the nth year, and the denominator represents the total power generation of various power sources in the nth year; β re,n represents the lower limit value of the proportion of clean power generation required by the policy in the nth year.

4. A source-storage collaborative planning and configuration method with source-load time-series similarity constraints according to claim 3, characterized in that: In the step 2, the reverse peak shaving safety constraint is mainly based on the reverse peak shaving characteristics of wind and solar. During the period of large load, the reverse peak shaving capacity of the system is large and there is already a sufficient safety domain. However, during the period of small load, the reverse peak shaving capacity is small and there is a safety domain risk. Therefore, during the period of small load, a safety domain constraint is defined to ensure that conventional units have a certain peak shaving space at any time to cope with the volatility of new energy output; Where: P Gi,min (t) is the minimum output of the conventional unit i at time t; β aq is the upper limit value of the security domain constraint; T l is the low load period.

5. A source-storage collaborative planning and configuration method with source-load time-series similarity constraints according to claim 1, characterized in that: The energy storage control strategy plan proposed in the step 3: Define wind and solar ramping constraints Control the operation state of the energy storage system according to the magnitude of the volatility of the wind and light output; specifically as follows: 3.

1. The energy storage does not operate When when P ess (t) = 0 where P′ m (t - Δt) is the output power of the m-th type of new energy after smoothing at the previous moment; δ m is the upper limit of the ramp of the m-th type of new energy; this equation means that the ramp rate of the new energy in the system cannot exceed δ m times its installed capacity; 3.

2. Energy storage discharge. At this time, the output of new energy suddenly decreases, and the ramp rate δ < -δ m exceeds the limit, and the energy storage system discharges to suppress the volatility of the output; When time P ess η(t) = ess [(δ + δ m )S m + P'(t - Δt)] m ​ where: η ess is the charge-discharge efficiency of the energy storage system; C3. Energy storage charging. At this time, the output of new energy suddenly increases, and the ramp rate δ > δ m exceeds the limit, and the energy storage system discharges to suppress the volatility of the output; When When P ess η(t) = ess [(δ m - δ)S m - P′ m (t - Δt)].

6. A source-storage collaborative planning and configuration method with source-load time series similarity constraints according to claim 1, characterized in that: The ideal point algorithm proposed in step 4 solves the bi-level programming model: First, calculate the optimal solutions of each objective function. Then, take this value as the expected value and transform the multi-objective optimization model into seeking the solution closest to the expected value among all solution sets, that is, transform it into a single-objective optimization model with the minimum sum of the deviation amounts between each objective function and a i The specific solution process is as follows: In the first step, for the multi-objective problem of bilevel programming, calculate the optimal values of each single objective function first In the second step, since it is difficult to have x 0 to minimize all objective functions, the distance norm ||·|| is introduced in the feasible region X; In the formula represents the importance degree of different objective functions for the global solution and there is Based on this, the optimal solution of multi-objective optimization is determined.

Images