A zero-carbon industrial park peak regulation resource configuration method based on time sequence production simulation

Abstract

The application discloses a zero-carbon industrial park peak regulation resource configuration method based on time sequence production simulation. The application is carried out according to the following steps: A. extracting features to generate multiple scenes, and obtaining the expected value of the main index based on multiple random scene optimization; B. production simulation optimization model, including objective function and constraint condition; C. peak regulation resource / energy storage system configuration process based on income sensitivity calculation. The application adopts long-time-scale time sequence scenes, meets the calculation demand of long-time-scale energy storage regulation resources in the park, and proposes the peak regulation resource / energy storage system configuration process based on income sensitivity calculation, so that the optimal energy storage capacity configuration under certain index requirements is obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of peak regulation resource optimization configuration and power storage system planning, and more particularly to a zero-carbon industrial park peak regulation resource configuration method based on time sequence scenario generation and reduction, which is mainly applied to green power supply parks, zero-carbon industrial parks and newly added load parks with peak regulation resource configuration requirements. BACKGROUND

[0002] At present, the park-level distribution side peak regulation resource and energy storage system configuration mainly adopts a method based on theoretical analysis and a method based on typical day power generation and power consumption curve optimization. The method based on theoretical analysis usually takes meeting a certain technical index as the goal, such as renewable energy generation fluctuation rate index and renewable energy generation prediction deviation index, and obtains the power capacity and energy capacity configuration of the peak regulation resource and energy storage system through relatively strict theoretical calculation. However, this method often ignores the economic factor of the energy storage configuration result and is difficult to take into account multiple interrelated constraint indexes, so it is less used in actual projects. The method based on typical day power generation and power consumption curve optimization is widely used. This method calculates the extreme value of the objective function under certain constraint conditions through an optimization algorithm to obtain the optimal value of the to-be-optimized variable. The objective function may contain a single objective or multiple objectives to meet different optimization needs. However, this method has the following technical defects: ① The typical day data cannot represent the power generation and power consumption situation in a long time scale. ② The optimization process is limited to the time scale of the typical day and cannot reflect the correlation between long time scale data. The obtained peak regulation resource and energy storage system configuration result cannot meet the long time scale energy storage regulation resource demand in the park. ③ The method fails to identify the scene characteristics through power generation and power consumption data and then generate power generation and power consumption scenes containing random information, resulting in single scenes and loss of random information. ④ The existing technical solution does not generate multiple scenes, so the optimization calculation is performed on multiple random scenes, resulting in poor robustness of the obtained energy storage configuration result and the inability to adapt to the constant changes of scenes with certain characteristic indexes. ⑤ The method does not calculate the revenue sensitivity of the peak regulation resource and energy storage system configuration through the expected value of the index, resulting in the inability to obtain the optimal energy storage capacity configuration under certain index requirements. SUMMARY

[0003] The present application solves the technical problem of overcoming the deficiencies in the prior art and provides a zero-carbon industrial park peak regulation resource configuration method based on time sequence scenario generation and reduction.

[0004] The present application provides a zero-carbon industrial park peak regulation resource configuration method based on time sequence scenario generation and reduction, which is implemented through the following technical solutions and performed according to the following steps:

[0005] A. Extract features to generate multiple scenes and obtain the expected value of the main index based on multiple random scenes:

[0006] 1) Wind power data selection

[0007] ① Obtain the wind speed at the hub height of the wind turbine using the annual wind data from the wind tower;

[0008] ② Obtain the annual wind power output data according to the installed capacity of wind power in the park through the output model;

[0009] ③ Extract the characteristics of the wind power output scene, use the scene generation method to obtain multiple output scenes, and perform scene verification, discard scenes that do not meet the requirements of volatility and randomness indicators, and obtain N scenes that meet the requirements of the indicators.

[0010] 2) Photovoltaic data selection

[0011] ① Obtain the annual light data, and obtain the light data of the region combined with the characteristics of the region where the park is located;

[0012] ② Obtain the annual photovoltaic output data according to the installed capacity of photovoltaic power in the park through the output model;

[0013] ③ Extract the characteristics of the photovoltaic output scene, use the scene generation method to obtain multiple output scenes, and perform scene verification, discard scenes that do not meet the requirements of volatility and randomness indicators, and obtain N scenes that meet the requirements of the indicators.

[0014] 3) New load data

[0015] ① Divide the new load in the park into conventional load and controllable load;

[0016] ② Obtain the annual electricity consumption and maximum load utilization hours of the conventional load, and obtain the typical daily electricity consumption curve of the conventional load in the park by referring to the historical data of similar loads;

[0017] ③ Obtain the capacity and maximum load utilization hours of the controllable load, and obtain the typical daily electricity consumption curve of the controllable load in the park by combining the all-day output of wind power and photovoltaic power and taking the maximum renewable energy consumption as the target;

[0018] ④ Superimpose the typical daily electricity consumption curves of the conventional load and the controllable load to obtain the typical daily electricity consumption curve of the new load in the park.

[0019] B. Production simulation optimization model, including objective function and constraint conditions:

[0020] 1) Objective function of optimization model

[0021]

[0022] Where, Obj is the value of the objective function, T is the total number of optimization periods, P w (t) is the wind power output (MW) at time period t, Pv (t) is the photovoltaic output (MW) at time period t, P s (t) is the energy storage output (MW) at time period t, positive when discharging and negative when charging, Var{·} is the variance of the time series, and |·| is the absolute value;

[0023] ① Renewable energy consumption maximization sub-goal: For a zero-carbon industrial park, the goal is to minimize the curtailment rate (an important indicator) without sending renewable energy to the grid, to minimize the amount of electricity sent to the grid without using grid peak-shaving resources, and to consider that the price of green electricity purchased due to the amount of electricity sent to the grid is greater than the energy storage charging and discharging cost;

[0024] ② Peak-shaving resource / energy storage system charging and discharging power variance minimization sub-goal: The energy storage charging and discharging power is an optimization variable, and the variance minimization is a component of the objective function, which can minimize the total amount of electricity charged and discharged by the peak-shaving resource / energy storage system while maximizing renewable energy consumption, and achieve small current charging and discharging to maximize the operating life of the peak-shaving resource / energy storage system;

[0025] 2) Optimization model constraints

[0026] ① No renewable energy generation: Make the zero-carbon park a pure load park for the power system to ensure that it does not cause peak-shaving pressure on the system;

[0027] P w (t) + P v (t) + P s (t) - P l (t) ≤ 0, t ∈ (0, T)

[0028] where P l (t) is the park load (MW) at time period t;

[0029] ② Energy storage charging capacity equals energy storage discharging capacity in the optimization period: The actual constraint is that regardless of the charging and discharging power curve of the peak-shaving resource / energy storage system during the optimization process, the sum of the energy storage charging and discharging in the optimization period is 0;

[0030]

[0031] This constraint has little effect in a single optimization period, and ignoring this constraint can ensure the continuity of the peak-shaving resource / energy storage system operating parameters in the case of continuous multiple optimization periods;

[0032] ③ Wind power output constraint: Ensure that the actual output of wind power in each time period during the optimization process is less than the minimum output and the predicted output;

[0033] Pw,min ≤ P w (t) ≤ P w,max , t e (0, T)

[0034] where P w,min is the minimum technical output of wind power in period t, P w,max is the predicted maximum output of wind power in period t;

[0035] (iv) Photovoltaic output constraint: ensure that the actual output of photovoltaic in each period is taken as the value before the minimum output and the predicted output in the optimization process;

[0036] P v,min ≤ P v (t) ≤ P v,max , t e (0, T)

[0037] where P v,min is the minimum technical output of photovoltaic in period t, P v,max is the predicted maximum output of photovoltaic in period t;

[0038] (v) Peak shaving resource / energy storage system power constraint: ensure that the charging or discharging power value of energy storage in each period is taken as the value between the maximum charging power and the maximum discharging power in the optimization process;

[0039] P s,min ≤ |P s (t)| ≤ P s,max , t e (0, T)

[0040] where P s,min is the minimum technical output of energy storage in period t, P s,max is the maximum technical output of energy storage in period t;

[0041] (vi) Energy storage capacity constraint: always keep its SOC between 0-1;

[0042] 0 ≤ SOC s (t) ≤ 1, t e (0, T)

[0043] where SOC s (t) is the state of charge of peak shaving resource / energy storage system in period t;

[0044] (vii) Network power constraint: meet the network power constraint, and the branch flow does not exceed the limit;

[0045] B i,min ≤ B i (t) ≤ B i,max

[0046] where B i (t) is the branch transmission power (MW) of the i-th period t, B i,minMW is the lower limit of transmission power of the ith branch (MW) i,max MW is the upper limit of transmission power of the ith branch (MW)

[0047] 8. Network voltage constraint: meet the network voltage constraint, the node voltage does not exceed the limit;

[0048] U j,min ≤U j (t)≤U j,max

[0049] Wherein, U j (t) is the node voltage (kV) of the jth node at period t, U j,min is the lower limit of node voltage (kV) of the jth node, U j,max is the upper limit of node voltage (kV) of the jth node;

[0050] C. Peak shaving resource / energy storage system configuration process based on revenue sensitivity calculation, the steps are as follows:

[0051] ① Generate multiple time series scenarios: use the input and processing of data and the proposed method to generate N photovoltaic, wind power, and new load time series scenarios, N is 10;

[0052] ② Select a group of power capacity and energy capacity of peak shaving resource / energy storage system, and perform production simulation on N time series scenarios to obtain N groups of data, including wind power, photovoltaic actual output data, net power data, and energy storage charging and discharging power data, calculate the main indicators of N time series scenarios, and obtain the expected value of the main indicators. Calculate the revenue after installing the peak shaving resource / energy storage system in the whole life cycle;

[0053] ③ Increase the power capacity (△P) and energy capacity (△E) of the peak shaving resource / energy storage system by a certain numerical step, repeat step ②, and obtain the increase value (△IP, △IE) of the revenue, respectively. Calculate the sensitivity (SP) of the increase value (△IP) of the revenue to the power capacity (△P) and the sensitivity (SE) of the increase value (△IE) of the revenue to the energy capacity (△E). If the sensitivity (SP or SE) is greater than 1, it means that the corresponding power capacity or energy capacity can be increased;

[0054] ④ Compare the size of SP and SE, if SP>SE and SP>1, increase the power capacity by a certain numerical step, if SE>SP and SE>1, increase the energy capacity by a certain numerical step, repeat step ③, if SP and SE are less than 1, stop calculating the calculation power capacity and the calculation energy capacity;

[0055] V. According to the calculated power capacity and the calculated energy capacity obtained in step IV, in combination with the typical power and energy configuration of the peak regulation resource / energy storage system, the power capacity and energy capacity configuration of the peak regulation resource / energy storage system of the zero-carbon park are determined.

[0056] Compared with the prior art, the beneficial effects of the present application are:

[0057] a) long time scale time sequence scenarios are used to meet the calculation requirements of long time scale energy storage regulation resources in the park;

[0058] b) by identifying the characteristics of long time scale power generation and power consumption data, power generation and power consumption scenarios containing random information are generated, avoiding the problems of single scenario and loss of random information caused by using typical day data;

[0059] c) multiple scenarios are generated and the expected values of the main indicators are obtained based on the production simulation of multiple random scenarios, which improves the robustness of the energy storage configuration result and further improves the ability to adapt to diversified scenarios;

[0060] d) a peak regulation resource / energy storage system configuration process based on the yield sensitivity calculation is proposed to obtain the optimal energy storage capacity configuration under certain index requirements. BRIEF DESCRIPTION OF DRAWINGS

[0061] Figure 1 is the data input and processing flowchart of the present application;

[0062] Figure 2 is the production simulation optimization model flowchart of the present application;

[0063] Figure 3 is the peak regulation resource / energy storage system configuration flowchart of the present application. DETAILED DESCRIPTION

[0064] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the present application will be described in detail below with reference to the accompanying drawings.

[0065] I. Data input and processing, as shown in Figure 1

[0066] 1. Selection of wind power data

[0067] ① Use the annual wind data of the wind tower to obtain the wind speed at the hub height of the wind turbine;

[0068] ② Obtain the annual wind power output data according to the wind power installed capacity in the park through the output model;

[0069] ③ Extract the wind power output scenario characteristics, use the scenario generation method to obtain multiple output scenarios, and perform scenario verification, discard the scenarios that do not meet the fluctuation and randomness index requirements, and obtain N scenarios that meet the index requirements.​

[0070] 2. Photovoltaic data selection

[0071] ① Obtain annual light data, and obtain light data of the region according to the characteristics of the region where the park is located;

[0072] ② Obtain annual photovoltaic output data according to the installed capacity of photovoltaic in the park through the output model;

[0073] ③ Extract photovoltaic output scene characteristics, obtain multiple output scenes using scene generation method, and perform scene verification, discard scenes that do not meet the requirements of volatility and randomness indicators, and obtain N scenes that meet the requirements of the indicators.

[0074] 3. New load data

[0075] ① Divide the new load in the park into regular load and controllable load;

[0076] ② Obtain the annual electricity consumption and maximum load utilization hours of the regular load, and obtain the typical daily electricity consumption curve of the regular load in the park by referring to the historical data of similar loads;

[0077] ③ Obtain the capacity and maximum load utilization hours of the controllable load, combine the all-day output of wind power and photovoltaic, and obtain the typical daily electricity consumption curve of the controllable load in the park with the maximum renewable energy consumption as the target;

[0078] ④ Superimpose the typical daily electricity consumption curves of the regular load and the controllable load to obtain the typical daily electricity consumption curve of the new load in the park.

[0079] II. Production simulation optimization model, as shown in Figure 2

[0080] 1. Objective function of optimization model

[0081]

[0082] Where, Obj is the value of the objective function, T is the total number of optimization periods, P w (t) is the wind power output (MW) at period t, P v (t) is the photovoltaic output (MW) at period t, P s (t) is the energy storage output (MW) at period t, positive when discharging and negative when charging, Var{·} is the variance of the time series, and |·| is the absolute value.

[0083] ​① Renewable energy consumption maximum sub-target: For a zero-carbon industrial park, the abandoned power rate should be minimized under the premise of not sending renewable energy generation, which is an important indicator, to minimize the grid power as much as possible without occupying the grid peak shaving resources, and it is considered that the price of green electricity purchased due to grid power is greater than the storage charging and discharging cost.

[0084] ② Peak shaving resource / storage system charging and discharging power variance minimum sub-target: The storage charging and discharging power is an optimization variable, and the variance minimum is a component of the objective function, which can minimize the total power of the peak shaving resource / storage system charging and discharging while maximizing the renewable energy consumption, and achieve the goal of supplementing charging and discharging with smaller current to maximize the operating life of the peak shaving resource / storage system.

[0085] The objective function shown in formula (1) is a multi-objective optimization problem, but its essence is a single-objective optimization problem. The main reason is that on the one hand, the numerical order of magnitude of renewable energy consumption is much larger than that of the peak shaving resource / storage system charging and discharging power variance, and the multi-objective optimization problem first takes renewable energy consumption as the main optimization target, i.e., renewable energy consumption has a large weight, and on the other hand, the optimization of the peak shaving resource / storage system charging and discharging power variance is equivalent to finding the minimum value of the variance in the variance space formed when the renewable energy consumption is maximum, i.e., the two sub-targets are actually optimized separately, but the difference between the target optimization is that the optimization of the peak shaving resource / storage system charging and discharging power variance is optimized in the variance space formed when the renewable energy consumption is maximum.

[0086] 2. Optimization model constraint conditions

[0087] ① No renewable energy generation: Make the zero-carbon park still a pure load park for the power system to ensure that it does not cause peak shaving pressure on the system.

[0088] P w (t) is the park load (MW) at time t. v (t) is the park load (MW) at time t. s (t) is the park load (MW) at time t. l (t) is the park load (MW) at time t.

[0089] P l (t) is the park load (MW) at time t.

[0090] ② The storage charging capacity in the optimization period is equal to the storage discharging capacity: The actual constraint condition is that no matter what charging and discharging power curve the peak shaving resource / storage system runs in the optimization process, the sum of the storage charging and discharging in the optimization period is 0.

[0091]

[0092] This constraint has little effect in the case of a single optimization period, and can be ignored. In the case of multiple optimization periods, this constraint can guarantee the continuity of the operating parameters of the peak regulation resource / energy storage system.

[0093] ③Wind power output constraint: In the optimization process, the actual output of wind power in each period is ensured to be less than the minimum output and the predicted output.

[0094] P w,min ≤P w (t)≤P w,max ,t∈(0,T)

[0095] where P w,min is the minimum technical output of wind power in period t, and P w,max is the predicted (maximum) output of wind power in period t.

[0096] ④Photovoltaic output constraint: In the optimization process, the actual output of photovoltaic in each period is ensured to be less than the minimum output and the predicted output.

[0097] P v,min ≤P v (t)≤P v,max ,t∈(0,T)

[0098] where P v,min is the minimum technical output of photovoltaic in period t, and P v,max is the predicted (maximum) output of photovoltaic in period t.

[0099] ⑤Peak regulation resource / energy storage system power constraint: In the optimization process, the charging or discharging power value of the energy storage in each period is ensured to be between the maximum charging power and the maximum discharging power.

[0100] P s,min ≤|P s (t)|≤P s,max ,t∈(0,T)

[0101] where P s,min is the minimum technical output of energy storage in period t, and P s,max is the maximum technical output of energy storage in period t.

[0102] ⑥Energy storage capacity constraint: Always keep its SOC between 0 and 1.

[0103] 0≤SOC s (t)≤1,t∈(0,T)

[0104] where SOC s (t) is the state of charge of the peak regulation resource / energy storage system in period t.

[0105] Network power constraint: meet the network power constraint, branch flow does not exceed the limit.

[0106] B i,min ≤B i (t)≤B i,max

[0107] Wherein, B i (t) is the branch transmission power (MW) of the i-th time period t, B i,min is the lower limit of the transmission power of the i-th branch (MW), B i,max is the upper limit of the transmission power of the i-th branch (MW).

[0108] Network voltage constraint: meet the network voltage constraint, node voltage does not exceed the limit.

[0109] U j,min ≤U j (t)≤U j,max

[0110] Wherein, U j (t) is the node voltage (kV) of the j-th node at time period t, U j,min is the lower limit of the node voltage of the j-th node (kV), U j,max is the upper limit of the node voltage of the j-th node (kV).

[0111] Three, peak regulation resource configuration process, as Figure 3 shown:

[0112] 1. Main indicators for calculation

[0113] ① Renewable energy curtailment revenue: renewable energy curtailment rate is the ratio of total wind and light curtailment to total wind and photovoltaic forecast power in the production simulation period. The lower the curtailment rate, the higher the renewable energy utilization rate.

[0114] ② Grid off-grid power revenue: grid off-grid power ratio is the ratio of total grid power to total power load in the production simulation period. The lower the off-grid power ratio, the higher the renewable energy power ratio.

[0115] ③ Off-grid power carbon tax cost revenue: off-grid power carbon tax revenue is the carbon tax cost generated by purchasing non-clean energy power from the grid in the production simulation period.

[0116] The above two main indicators have a positive correlation, that is, the decrease or increase of renewable energy curtailment rate is accompanied by the decrease and then increase of the proportion of grid-connected power. If curtailment power and curtailment capacity are generated in period one, there is curtailment cost in period one, and if grid-connected power and grid-connected capacity are generated in period two, there is purchase cost in period two. Assuming that the curtailment capacity in period one is reduced by charging and the grid-connected capacity in period two is reduced by discharging due to the installation of the peak shaving resource / energy storage system, although intuitively, the peak shaving resource / energy storage system is optimized to reduce curtailment cost and purchase cost, but regardless of the subject form (including new energy subjects containing wind power, photovoltaic, and energy storage or park subjects containing wind power, photovoltaic, energy storage, and load), when calculating the benefits obtained after installing the peak shaving resource / energy storage system, only one of the above indicators can be calculated, and the grid-connected power proportion indicator is used in the technical solution. In addition, the bases of the two indicators are different, and the curtailment capacity and grid-connected capacity can be used for calculation in the technical solution.

[0117] In a zero-carbon park, the proportion of grid-connected power is reduced after installing the peak shaving resource / energy storage system, so the carbon tax cost is reduced, and the reduced part is the benefit obtained after installing the peak shaving resource / energy storage system. Therefore, the total benefit obtained includes the reduced grid-connected capacity benefit and the reduced carbon tax cost benefit.

[0118] It should be pointed out that the peak shaving resource / energy storage system can obtain benefits through leasing, participating in the spot market or ancillary services, obtaining policy subsidies, etc., but it is not clear whether these benefits can be obtained in a zero-carbon park, so they are not considered as part of the energy storage benefit.

[0119] 2. Peak shaving resource configuration calculation process based on benefit sensitivity

[0120] ① Generate multiple time series scenarios: use the input and processing method of data to generate N photovoltaic, wind power, and new load time series scenarios, and N can be 10;

[0121] ② Select a group of power capacity and energy capacity of the peak shaving resource / energy storage system, and perform production simulation on the N time series scenarios to obtain N groups of data, including wind power, photovoltaic actual output data, grid-connected power data, and energy storage charging and discharging power data, calculate the main indicators of the N time series scenarios, and obtain the expected value of the main indicators. The expected value of the indicators is used to calculate the benefit after installing the peak shaving resource / energy storage system in the whole life cycle;

[0122] ③Increase the power capacity (△P) and the energy capacity (△E) of the peak shaving resource / energy storage system by a certain numerical step, repeat step ②, respectively obtain the increase value (△IP, △IE) of the benefit, respectively calculate the sensitivity (SP) of the increase value (△IP) of the benefit to the power capacity (△P) and the sensitivity (SE) of the increase value (△IE) of the benefit to the energy capacity (△E), if the sensitivity (SP or SE) is greater than 1, it represents that the corresponding power capacity or energy capacity can be increased;

[0123] ④Compare the sizes of SP and SE, if SP>SE and SP>1, increase the power capacity by a certain numerical step, if SE>SP and SE>1, increase the energy capacity by a certain numerical step, repeat step ③, if SP and SE are both less than 1, stop calculating to obtain the calculation power capacity and the calculation energy capacity;

[0124] ⑤According to the calculation power capacity and the calculation energy capacity obtained in step ④, combined with the typical power and energy configuration of the peak shaving resource / energy storage system, the power capacity and the energy capacity configuration of the peak shaving resource / energy storage system of the zero-carbon park are determined.

[0125] The above only describes the preferred embodiments of the present application, it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method for peak-shaving resource allocation in zero-carbon industrial parks based on time-series production simulation, characterized in that, Follow these steps: A. Extract features to generate multiple scene-based datasets, and optimize based on multiple random scenes to obtain the expected values ​​of key metrics: 1) Selection of wind power data ① The wind speed at the height of the wind turbine hub is obtained using the annual wind measurement data from the wind measurement tower; ② Based on the installed wind power capacity within the park, annual wind power output data is obtained through an output model; ③ Extract the characteristics of wind power output scenarios, use the scenario generation method to obtain multiple output scenarios, and perform scenario verification. Discard scenarios that do not meet the requirements of volatility and randomness indicators, and obtain N scenarios that meet the requirements. 2) Selection of photovoltaic data ① Obtain annual sunshine data and combine it with the characteristics of the area where the park is located to obtain the sunshine data for that area; ② Based on the photovoltaic installed capacity within the park, the annual photovoltaic power output data is obtained through a power output model; ③ Extract the characteristics of photovoltaic power output scenarios, use the scenario generation method to obtain multiple power output scenarios, and perform scenario verification. Discard scenarios that do not meet the requirements of volatility and randomness indicators, and obtain N scenarios that meet the requirements. 3) New load data ① Divide the new load in the park into conventional load and controllable load; ② Obtain the annual electricity consumption and maximum load utilization hours of the regular load, and refer to historical data of similar loads to obtain the typical daily electricity consumption curve of the park's regular load; ③ Obtain the capacity and maximum load utilization hours of the controllable load, and combine the total daily output of wind and solar power to obtain the typical daily electricity consumption curve of the controllable load in the park with the goal of maximizing renewable energy consumption; ④ By superimposing the typical daily electricity consumption curves of conventional loads and controllable loads, the typical daily electricity consumption curve of the newly added load in the park is obtained; B. Production simulation optimization model, including objective function and constraints: 1) Optimize the objective function of the model Where Obj is the objective function value, T is the total number of optimization periods, and P is the total number of optimization periods. w (t) represents the wind power output during time period t, P v (t) represents the photovoltaic output during time period t, P s (t) represents the energy storage output during time period t, which is positive during discharge and negative during charging. Var{·} is the variance of the time series, and |·| is the absolute value. ① The sub-target of maximizing renewable energy consumption: For zero-carbon industrial parks, the curtailment rate should be minimized without sending renewable energy power generation to the grid, and the amount of electricity sent to the grid should be minimized without occupying the grid's peak-shaving resources. It is also assumed that the price of green electricity required to purchase electricity sent to the grid is greater than the cost of energy storage charging and discharging. ② Minimize the variance of charging and discharging power of peak-shaving resources / energy storage system: The charging and discharging power of energy storage is the optimization variable. By minimizing the variance as a component of the objective function, it is possible to reduce the total charging and discharging power of peak-shaving resources / energy storage system while maximizing the consumption of renewable energy, and to achieve supplementary charging and discharging with a smaller current, thereby maximizing the operating life of peak-shaving resources / energy storage system. 2) Optimize model constraints ① No renewable energy generation is transmitted: This ensures that the zero-carbon park remains a pure load park for the power system, thus preventing any peak-shaving pressure on the system. P w (t)+P v (t)+P s (t)-P l (t)≤0,t∈(0,T) Among them, P l (t) represents the park load during time period t; ② The energy storage charging amount equals the energy storage discharging amount during the optimization period: The actual constraint is that no matter what charging and discharging power curve the peak-shaving resources / energy storage system operate with during the optimization process, the sum of energy storage charging and discharging during the optimization period is 0. This constraint has a relatively small impact in the case of a single optimization period. Ignoring this constraint, in the case of multiple consecutive optimization periods, this constraint can ensure the consistency of the operating parameters of the peak-shaving resources / energy storage system. ③ Wind power output constraint: During the optimization process, the actual wind power output in each time period is taken before the minimum output and the predicted output; P w,min ≤P w (t)≤P w,max ,t∈(0,T) Among them, P w,min For the minimum technical output of wind power in time period t, P w,max The predicted maximum wind power output for time period t; ④ Photovoltaic output constraint: During the optimization process, ensure that the actual photovoltaic output in each time period is between the minimum output and the predicted output; P v,min ≤P v (t)≤P v,max ,t∈(0,T) Among them, P v,min For the minimum technical output of photovoltaic power during time period t, P v,max The maximum predicted photovoltaic output for time period t; ⑤ Peak-shaving resource / energy storage system power constraints: During the optimization process, ensure that the energy storage charging or discharging power value in each time period falls between the maximum charging power and the maximum discharging power; P s,min ≤|P s (t)|≤P s,max ,t∈(0,T) Among them, P s,min For the minimum technical output of energy storage in time period t, P s,max To maximize the technical output of energy storage during time period t; ⑥ Energy storage capacity constraint: Always keep its SOC between 0 and 1; 0 ≤ SOC s (t)≤1,t∈(0,T) Among them, SOC s (t) represents the state of charge of the peak-shaving resources / energy storage system during time period t; ⑦ Network power constraints: Network power constraints must be met; branch power flow must not exceed limits. i,min ≤B i (t)≤B i,max Among them, B i (t) represents the branch transmission power (MW) in the i-th time period t, B i,min B is the lower limit of the transmission power of the i-th branch. i,max This represents the upper limit of transmission power for the i-th branch; ⑧ Network voltage constraints: The network voltage constraints must be met, and the node voltage must not exceed the limit; U j,min ≤U j (t)≤U j,max Among them, U j (t) represents the node voltage (kV) at the j-th node time period t. j,min U is the lower limit of the node voltage of the j-th node. j,max This represents the upper limit of the node voltage of the j-th node; C. The configuration process for peak-shaving resources / energy storage systems based on revenue sensitivity calculation is as follows: ① Generate multiple time-series scenarios: Using the proposed method for data input and processing, generate N time-series scenarios for photovoltaic, wind power, and new load, where N is 10; ② Select a set of peak-shaving resources / energy storage systems with power capacity and energy capacity, perform production simulation on N time-series scenarios, obtain N sets of data, including actual output data of wind power and photovoltaic power, grid-connected power data, and energy storage charging and discharging power data, calculate the main indicators of N time-series scenarios, and obtain the expected values ​​of the main indicators. Use the expected values ​​of the indicators to calculate the benefits after installing peak-shaving resources / energy storage systems throughout the entire life cycle. ③ Increase the power capacity △P and energy capacity △E configured in the peak-shaving resource / energy storage system by a certain numerical step size, repeat step ②, and obtain the increase in revenue △IP and △IE respectively. Calculate the sensitivity SP of the increase in revenue △IP to the power capacity △P and the sensitivity SE of the increase in revenue △IE to the energy capacity △E respectively. If the sensitivity SP or SE is greater than 1, it means that the corresponding power capacity or energy capacity can be increased. ④ Compare the magnitudes of SP and SE. If SP > SE and SP > 1, increase the power capacity by a certain numerical step. If SE > SP and SE > 1, increase the energy capacity by a certain numerical step. Repeat step ③. If both SP and SE are less than 1, stop the calculation to obtain the calculated power capacity and calculated energy capacity. ⑤ Based on the calculated power capacity and calculated energy capacity obtained in step ④, and combined with the typical power and energy configuration of the peak-shaving resource / energy storage system, determine the power capacity and energy capacity configuration of the peak-shaving resource / energy storage system in the zero-carbon park.

Images