A hybrid energy storage capacity configuration method for a wind-storage combined power generation system

Abstract

The present invention is applicable to the technical field of wind-storage combined power generation system, and provides a hybrid energy storage capacity configuration method for wind-storage combined power generation system, which uses improved Kalman filtering to find the optimal Q , R parameters to smooth wind power fluctuations; the Kepler algorithm is used to optimize VMD decomposition parameters to achieve power allocation between batteries and supercapacitors; and the energy storage rated power and capacity are calculated based on charge and discharge efficiency and SOC constraints. This method effectively reduces energy storage configuration costs by over 50% and is suitable for energy storage optimization for large-scale wind power grid integration.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind-storage combined power generation systems, and in particular relates to a hybrid energy storage capacity configuration method for a wind-storage combined power generation system. Background Art

[0002] Wind power generation is a clean, green energy source, and the continued development of wind resources is a key component of the global energy transition toward sustainable development. However, the randomness and volatility of large-scale wind power grid integration pose multiple challenges to the stable operation of power systems. In recent years, hybrid energy storage systems have been widely used to smooth wind power fluctuations due to their flexible operating characteristics, and extensive research has been conducted on their control methods. However, in practical engineering applications, the economics of energy storage are often the primary consideration, making the study of energy storage costs in this context crucial. Summary of the Invention

[0003] The purpose of the embodiments of the present invention is to provide a hybrid energy storage capacity configuration method for a wind-storage combined power generation system, aiming to solve the problems raised in the above background technology.

[0004] The embodiment of the present invention is implemented as follows: a hybrid energy storage capacity configuration method for a wind-storage combined power generation system includes the following steps:

[0005] Step 1: Use the improved Kalman filter algorithm to smooth out the fluctuation and obtain the target grid-connected power;

[0006] Step 2: Use the KOA-VMD algorithm to implement power allocation that takes into account the characteristics of the battery and supercapacitor energy storage devices;

[0007] Step 3: Calculate the rated power and capacity of the hybrid energy storage based on the charge and discharge power and energy storage efficiency to complete the configuration.

[0008] Further technical solution, said step 1 comprises the following specific steps:

[0009] Step 1.1: Establish the time update equation and state update equation;

[0010] Step 1.2: Find the optimal Q and R-value;

[0011] Step 1.3: Use the KOA-VMD algorithm to smooth wind power fluctuations and find the optimal Q The wind power fluctuation is initially smoothed by R and R. After the wind power grid-connected power is obtained, the target power of the hybrid energy storage system is the difference between the grid-connected power and the original output power of the wind power.

[0012] Further technical solution, said step 1.1 includes the following specific steps:

[0013] The Kalman filter algorithm is used to smooth wind power fluctuations, which includes two parts: prediction and update. First, the time update equation is based on t The posterior state estimate at time -1 is calculated t The prior state estimate at the time instant is used to implement the Kalman filter prediction; then, the measurement update equation combines the prior estimate obtained from the time update equation with the actual measurement value to obtain a more accurate posterior estimate;

[0014] Based on the research background of Kalman filtering to smooth wind power fluctuations, the time update equation and state update equation are established:

[0015] The time update equation is as follows:

[0016] ;

[0017] ;

[0018] The state update equation is as follows:

[0019] ;

[0020] ;

[0021] ;

[0022] ;

[0023] Where: is the measurement residual; for t -1 moment t A priori estimate of the state at a given moment; for t -1 time after the grid-connected power is stabilized, unit: MW; is the a priori estimated covariance; for t -1 moment estimated covariance; for t Grid-connected power at the moment, unit: MW; for t The original wind power at the moment, unit: MW; is the Kalman filter gain; R is the measurement noise covariance matrix; Q is the process noise covariance matrix; P ( t | t )for t The estimated covariance of the moments.

[0024] Further technical solution, said step 1.2 includes the following specific steps:

[0025] When using energy storage devices to smooth wind power fluctuations, the current focus is on the fluctuation amount and energy storage capacity. Considering the mutual constraints between the two objectives, an objective function J that includes the fluctuation amount and energy storage capacity is constructed based on weighted coefficients to better balance the two conflicting fluctuation smoothing objectives:

[0026] ;

[0027] Where: b is the weighting coefficient, and its value is between 0 and 1. The sum of the weighting coefficients of energy storage capacity and fluctuation amount is 1. is the maximum throughput energy of energy storage; To smooth out the fluctuation of grid-connected power, it consists of 1-minute fluctuation and 10-minute fluctuation.

[0028] ;

[0029] Where: is the original wind power at time t; is the grid-connected power obtained after smoothing at time t.

[0030] The grid-connected power obtained after stabilization is allowed to fluctuate within a reasonable range. Therefore, the fluctuation amount in the objective function has a certain benchmark, and the fluctuation amount exceeding the benchmark is included in the calculation of the objective function as a penalty item.

[0031] The calculation formula of 1min time scale fluctuation is:

[0032] ;

[0033] The calculation formula of 10-minute time scale fluctuation is:

[0034] ;

[0035] ;

[0036] Where: It is the benchmark value of 1-minute fluctuation; It is the 10-minute fluctuation benchmark value; is the maximum fluctuation within the 10-min time scale at time t;

[0037] Weighting coefficient b The value of will lead to different optimization tendencies. According to the principle of Pareto frontier, different b The values ​​form multiple sets of optimal solutions, and the optimal solution is selected based on the minimum objective function. bThe simulated annealing algorithm is used to solve the optimal Q and R values, which can better balance the two conflicting goals and meet the demand for smoothing wind power fluctuations.

[0038] Further technical solution, said step 2 includes the following specific steps:

[0039] Step 2.1: Kepler algorithm optimizes VMD parameters;

[0040] VMD is performed by presetting the modal decomposition number K and penalty factor , adaptively decompose the non-stationary signal and transform the input signal f Decompose into K The Kepler optimization algorithm is introduced to optimize the parameters of VMD and find the best parameter combination. ;

[0041] Step 2.2: Use KOA-VMD algorithm to realize power allocation;

[0042] The KOA algorithm optimizes the VMD parameters, sets the maximum number of iterations, and searches for the optimal solution of the parameters; based on the decomposition parameters obtained by the optimization , decompose Phess, decompose the hybrid energy storage target power into multiple modal components, the low-frequency components are allocated to the battery, and the high-frequency components are allocated to the supercapacitor.

[0043] Further technical solution, said step 3 includes the following specific steps:

[0044] When configuring capacity, only the rated power and rated capacity of the hybrid energy storage device are considered. Since the energy conversion efficiency of the hybrid energy storage system is not 100% during the charging and discharging process, the following adjustments need to be made when configuring the rated power of the energy storage system:

[0045] ;

[0046] Where: is the rated power, unit: MW; t 0 is the initial sampling time; T is the sampling period; Energy storage device t Charging power at any moment, unit: MW; Energy storage device t Discharge power at any moment, unit: MW; charging efficiency of energy storage devices; is the discharge efficiency of the energy storage device.

[0047] After obtaining the charging and discharging power of the energy storage system at each moment, the cumulative energy storage capacity at each moment is calculated; considering the variation range of the energy storage state of charge, the rated capacity of the hybrid energy storage system is calculated within the sampling period:

[0048] ;

[0049] ;

[0050] Where: E BN is the rated capacity of the battery; E SN is the rated capacity of the supercapacitor; E ba ( t )for t The cumulative energy storage capacity of the battery at all times; E sc ( t )for t The cumulative energy storage capacity of the supercapacitor at all times; SOC max The maximum state of charge allowed for the energy storage device; SOC min The minimum state of charge allowed for the energy storage device.

[0051] The embodiment of the present invention provides a hybrid energy storage capacity configuration method for a wind-storage combined power generation system, which uses an improved Kalman filter to find the optimal Q , R parameters to smooth wind power fluctuations; the Kepler algorithm is used to optimize VMD decomposition parameters to achieve power allocation between batteries and supercapacitors; and the energy storage rated power and capacity are calculated based on charge and discharge efficiency and SOC constraints. This method effectively reduces energy storage configuration costs by over 50% and is suitable for energy storage optimization for large-scale wind power grid integration. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 The 24-hour wind farm output curves for two typical days are shown below;

[0053] Figure 2 This is the comparison of wind power before and after leveling on a typical day 1;

[0054] Figure 3 This is the comparison of wind power before and after leveling on a typical day 2;

[0055] Figure 4 To smooth out the fluctuations before and after 1 minute;

[0056] Figure 5 To smooth out the fluctuations in the 10 minutes before and after;

[0057] Figure 6 This is the KOA-VMD power decomposition flow chart;

[0058] Figure 7 is the IMF component obtained by KOA-VMD decomposition;

[0059] Figure 8 Decompose the Hilbert marginal spectrum for KOA-VMD;

[0060] Figure 9 Power distribution for hybrid energy storage;

[0061] Figure 10 Flowchart for step 1.2. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0063] The specific implementation of the present invention is described in detail below with reference to specific embodiments.

[0064] An embodiment of the present invention provides a hybrid energy storage capacity configuration method for a wind-storage combined power generation system, comprising the following steps:

[0065] Step 1: Use the improved Kalman filter algorithm to smooth out the fluctuation and obtain the target grid-connected power;

[0066] Step 1.1: Establish the time update equation and state update equation;

[0067] The Kalman filter algorithm is used to smooth wind power fluctuations, which mainly includes two parts: prediction and update. First, the time update equation is based on the previous moment ( t -1 moment), and calculate the posterior state estimate of the current moment ( t The Kalman filter uses the prior state estimate at the time instant to implement the prediction. The measurement update equation then combines the prior estimate obtained from the time update equation with the actual measurement value to obtain a more accurate posterior estimate.

[0068] Based on the research background of Kalman filtering to smooth wind power fluctuations, the time update equation and state update equation are established:

[0069] The time update equation is as follows:

[0070] ;

[0071] ;

[0072] The state update equation is as follows:

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] Where: is the measurement residual; for t -1 moment t A priori estimate of the state at a given moment; for t -1 time after the grid-connected power is stabilized, unit: MW; is the a priori estimated covariance; for t -1 moment estimated covariance; for t Grid-connected power at the moment, unit: MW; for t The original wind power at the moment, unit: MW; is the Kalman filter gain; R is the measurement noise covariance matrix; Q is the process noise covariance matrix; P ( t | t )for t The estimated covariance of the moments.

[0078] Step 1.2: Find the optimal Q and R-value;

[0079] Traditional Kalman filter algorithm R and Q The value is set artificially, however, the Kalman filter R and Q The selection of has a significant impact on the filtering results. When the wind power suddenly increases or decreases, the parameters cannot be effectively adjusted with the sudden change in wind power, resulting in the failure of the filtering results to converge.

[0080] When using energy storage devices to smooth wind power fluctuations, the current focus is on the fluctuation amount and energy storage capacity. Considering the mutual constraints between the two objectives, an objective function J that includes the fluctuation amount and energy storage capacity is constructed based on weighted coefficients to better balance the two conflicting fluctuation smoothing objectives:

[0081] ;

[0082] Where: bis the weighting coefficient, and its value is between 0 and 1. The sum of the weighting coefficients of energy storage capacity and fluctuation amount is 1. is the maximum throughput energy of energy storage; To smooth out the fluctuation of grid-connected power, it consists of 1-minute fluctuation and 10-minute fluctuation.

[0083] ;

[0084] Where: is the original wind power at time t; is the grid-connected power obtained after smoothing at time t.

[0085] The grid-connected power obtained after stabilization is allowed to fluctuate within a reasonable range. Therefore, the fluctuation amount in the objective function has a certain benchmark, and the fluctuation amount exceeding the benchmark is included in the calculation of the objective function as a penalty item.

[0086] The calculation formula of 1min time scale fluctuation is:

[0087] ;

[0088] The calculation formula of 10-minute time scale fluctuation is:

[0089] ;

[0090] ;

[0091] Where: It is the benchmark value of 1-minute fluctuation; It is the 10-minute fluctuation benchmark value; is the maximum fluctuation within the 10-min time scale at time t;

[0092] Weighting coefficient b The value will lead to different optimization tendencies. According to the principle of Pareto frontier, different b The values ​​form multiple sets of optimal solutions, and the optimal solution is selected based on the minimum objective function. b value, and on this basis, the simulated annealing algorithm is used to solve the optimal Q and R values, which can better balance the two conflicting goals and meet the demand for smoothing wind power fluctuations. The specific optimization process is as follows: Figure 10 shown.

[0093] Step 1.3: Preliminary smoothing of wind power fluctuations;

[0094] First, the K-means algorithm is used to cluster the power data of a wind farm in northwest my country with a sampling step of 1 minute and an installed capacity of 45 MW in January 2021. The algorithm cluster number K value is set to 6, and the wind power fluctuations on six typical days are obtained. Since the fluctuations on typical days 1, 3, 4, and 6 are relatively small, and the fluctuations rarely exceed the grid connection standard, in order to demonstrate the effectiveness of the proposed fluctuation smoothing method, the wind power output curves of typical days 2 and 5 with relatively strong fluctuations are selected as the smoothing targets. The wind power curves of typical days 2 and 5 are shown in Figure 2. Figure 1 .

[0095] In order to verify the effectiveness of the proposed method, simulation analysis of different schemes is carried out. Method 1: Using traditional Kalman filtering to smooth wind power fluctuations, Q and R The value is set randomly; Method 2: Use improved Kalman filtering to smooth wind power fluctuations, and find the optimal Q and R , so that it can adapt to sudden changes in wind power.

[0096] The wind power curves before and after the leveling of typical days 1 and 2 are partially enlarged, as shown in the figure below. Figure 2 and Figure 3 It can be seen that both Method 1 and Method 2 can effectively stabilize wind power, but the grid-connected power obtained by Method 2 after stabilization is more consistent with the original power operation trend, and can better cope with sudden changes in wind power, effectively reduce energy storage output, and alleviate energy storage burden.

[0097] The 1-minute and 10-minute fluctuation constraints specified in the national standard are used as the foothold and basis for the formulation of the control method. Taking typical day 1 as an example, the original wind power is smoothed by method 1 and method 2 to calculate the 1-minute and 10-minute fluctuations. Figure 4 and Figure 5 . Figure 4 and Figure 5 In the data, Fluctuation Sequence 1 represents the original wind power curve fluctuations, Fluctuation Sequence 2 represents the grid-connected power curve fluctuations obtained after smoothing using Method 2, and Fluctuation Sequence 3 represents the grid-connected power curve fluctuations obtained after smoothing using Method 1. The maximum power fluctuations on Typical Days 1 and 2 reached 8.93 MW and 7.726 MW, respectively. The maximum power fluctuations over 10 minutes reached 21.25 MW and 18.33 MW, respectively. The maximum fluctuations after smoothing wind power fluctuations using different methods are shown in Table 1.

[0098] Table 1 Fluctuation Smoothing Indicators

[0099]

[0100] Comparative Analysis Figure 4 and Figure 5As shown in Table 1, the wind power curves obtained by filtering with the two methods both meet the grid connection constraint conditions, which are shown in Table 2. However, the fluctuation of method 2 is significantly greater than that of method 1, which indicates that the improved Kalman filter is more efficient. Q and R These two parameters can effectively cope with sudden changes in wind power, avoid excessive smoothing, and reduce energy storage operating costs.

[0101] Table 2 Wind farm power fluctuation limits

[0102]

[0103] After obtaining the wind power grid-connected power, the target power of the hybrid energy storage system is the difference between the grid-connected power and the original output power of the wind power.

[0104] Step 2: Use the KOA-VMD algorithm to implement power allocation that takes into account the characteristics of the battery and supercapacitor energy storage devices;

[0105] Step 2.1: Kepler algorithm optimizes VMD parameters;

[0106] VMD is a signal processing technique whose core is to construct and solve a variational problem, with the goal of decomposing complex non-stationary signals into a series of intrinsic mode functions. K and penalty factor , adaptively decompose the non-stationary signal and transform the input signal f Decompose into K To avoid manually setting parameter combinations The subjectivity of the decomposition reduces the accuracy of the decomposition. The Kepler optimization algorithm is introduced to optimize the parameters of VMD and find the best parameter combination. , optimize the process such as Figure 6 shown.

[0107] Step 2.2: Implement power allocation using the KOA-VMD algorithm

[0108] Taking typical day 1 as an example, the KOA algorithm optimizes the VMD parameters and sets the maximum number of iterations to 50 to find the optimal solution for the parameters. , decompose Phess to obtain 6 modal components with frequency range from high to low, such as Figure 7 The frequency characteristics of the signal after the hybrid energy storage target power is processed by VMD and Hilbert transform are shown as follows. Figure 8 .

[0109] From the marginal spectrum, we can see that IMF1 and IMF2 have lower frequencies and higher amplitudes, while IMF3-IMF6 have higher frequencies and lower amplitudes. Reconstructing IMF1 and IMF2 into the reference power of the battery can reduce the number of times the battery is charged and discharged, slowing down its lifespan loss. Supercapacitors have high power density, so reconstructing IMF3-IMF6 into the reference power of the supercapacitor can prevent overcharging or over-discharging. The power that the battery and supercapacitor need to bear after component reconstruction is as follows: Figure 9 shown.

[0110] Depend on Figure 9 It can be seen that in the hybrid energy storage system, there is a significant difference in the power allocated to the battery and the supercapacitor. The battery mainly bears the gentle low-frequency power component, while the remaining high-frequency power component is borne by the supercapacitor. Figure 9 A zoomed-in view of the power distribution between 8:00 AM and 12:00 PM shows that the supercapacitor has a higher number of charge and discharge cycles, while the battery has a relatively lower number. This power distribution method reduces the number of charge and discharge transitions and reduces battery life loss. By adopting a reasonable power allocation strategy, the advantages of different energy storage devices in the hybrid energy storage system are fully utilized.

[0111] Step 3: Capacity configuration of the hybrid energy storage system;

[0112] When configuring capacity, only the rated power and rated capacity of the hybrid energy storage device are considered. Since the energy conversion efficiency of the hybrid energy storage system is not 100% during the charging and discharging process, the following adjustments need to be made when configuring the rated power of the energy storage system:

[0113] ;

[0114] Where: is the rated power, unit: MW; t 0 is the initial sampling time; T is the sampling period; Energy storage device t Charging power at any moment, unit: MW; Energy storage device t Discharge power at any moment, unit: MW; charging efficiency of energy storage devices; is the discharge efficiency of the energy storage device.

[0115] After obtaining the charge and discharge power of the energy storage system at each moment, the cumulative energy storage capacity at each moment can be calculated. Taking into account the range of variation in the energy storage state of charge, the rated capacity of the hybrid energy storage system is calculated within the sampling period.

[0116] ;

[0117] ;

[0118] Where: E BN is the rated capacity of the battery; E SN is the rated capacity of the supercapacitor; E ba ( t )for t The cumulative energy storage capacity of the battery at all times; E sc ( t )for t The cumulative energy storage capacity of the supercapacitor at all times; SOC max The maximum state of charge allowed for the energy storage device; SOC min SOC is the minimum state of charge allowed for the energy storage device; max , SOC min .

[0119] To compare the advantages and disadvantages of the two methods, different energy storage devices are configured for capacity. The energy storage system capacity configuration parameters are shown in Table 3. The hybrid energy storage target power obtained by different methods on typical days 1 and 2 is configured for capacity. The configuration results are shown in Table 4.

[0120] Table 3 Parameters related to energy storage system capacity configuration

[0121]

[0122] Table 4 Capacity configuration results

[0123]

[0124] As shown in Table 4, the energy storage cost for smoothing wind power fluctuations using Method 1 on Typical Day 1 was RMB 86.263 million, while the cost using Method 2 was reduced to RMB 32.484 million, a cost reduction of RMB 53.779 million. The cost of Method 2 on Typical Day 2 was RMB 50.590 million lower than that of Method 1, proving that Method 2 is superior to Method 1.

[0125] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hybrid energy storage capacity configuration method for a wind-storage combined power generation system, characterized in that: The following steps are involved: Step 1: Use the improved Kalman filter algorithm to smooth out the fluctuation and obtain the target grid-connected power; Step 2: Use the KOA-VMD algorithm to implement power allocation that takes into account the characteristics of the battery and supercapacitor energy storage devices; Step 3: Calculate the rated power and capacity of the hybrid energy storage based on the charge and discharge power and energy storage efficiency to complete the configuration; The step 1 includes the following specific steps: Step 1.1: Establish the time update equation and state update equation; Step 1.2: Find the optimal Q and R-value; Step 1.3: Use the KOA-VMD algorithm to smooth wind power fluctuations and find the optimal Q The wind power fluctuation is initially smoothed by R and R. After obtaining the wind power grid-connected power, the target power of the hybrid energy storage system is the difference between the grid-connected power and the original wind power output power. The step 2 includes the following specific steps: Step 2.1: Kepler algorithm optimizes VMD parameters; VMD is performed by presetting the modal decomposition number K and penalty factor , adaptively decompose the non-stationary signal and transform the input signal f Decompose into K The Kepler optimization algorithm is introduced to optimize the parameters of VMD and find the best parameter combination. ; Step 2.2: Use KOA-VMD algorithm to realize power allocation; The KOA algorithm optimizes the VMD parameters, sets the maximum number of iterations, and searches for the optimal solution of the parameters; based on the decomposition parameters obtained by the optimization , decompose Phess, decompose the hybrid energy storage target power into multiple modal components, allocate low-frequency components to batteries, and allocate high-frequency components to supercapacitors; The step 3 includes the following specific steps: When configuring capacity, only the rated power and rated capacity of the hybrid energy storage device are considered. Since the energy conversion efficiency of the hybrid energy storage system is not 100% during the charging and discharging process, the following adjustments need to be made when configuring the rated power of the energy storage system: ； Where: is the rated power, unit: MW; t 0 is the initial sampling time; T is the sampling period; Energy storage device t Charging power at any moment, unit: MW; Energy storage device t Discharge power at any moment, unit: MW; charging efficiency of energy storage devices; is the discharge efficiency of the energy storage device; After obtaining the charging and discharging power of the energy storage system at each moment, the cumulative energy storage capacity at each moment is calculated; considering the variation range of the energy storage state of charge, the rated capacity of the hybrid energy storage system is calculated within the sampling period: ； ； Where: E BN is the rated capacity of the battery; E SN is the rated capacity of the supercapacitor; E ba ( t )for t The cumulative energy storage capacity of the battery at all times; E sc ( t )for t The cumulative energy storage capacity of the supercapacitor at all times; SOC max The maximum state of charge allowed for the energy storage device; SOC min The minimum state of charge allowed for the energy storage device.

2. The hybrid energy storage capacity configuration method for a wind-storage combined power generation system according to claim 1, characterized in that: The step 1.1 includes the following specific steps: The Kalman filter algorithm is used to smooth wind power fluctuations, which includes two parts: prediction and update. First, the time update equation is based on t The posterior state estimate at time -1 is calculated t The prior state estimate at the moment realizes the prediction of Kalman filter; Subsequently, the measurement update equation combines the prior estimate obtained by the time update equation with the actual measurement value to obtain a more accurate posterior estimate; Based on the research background of Kalman filtering to smooth wind power fluctuations, the time update equation and state update equation are established: The time update equation is as follows: ； ； The state update equation is as follows: ； ； ； ； Where: is the measurement residual; for t -1 moment t A priori estimate of the state at a given moment; for t -1 time after the grid-connected power is stabilized, unit: MW; is the a priori estimated covariance; for t -1 moment estimated covariance; for t Grid-connected power at the moment, unit: MW; for t The original wind power at the moment, unit: MW; is the Kalman filter gain; R is the measurement noise covariance matrix; Q is the process noise covariance matrix; P ( t | t )for t The estimated covariance of the moments.

3. The hybrid energy storage capacity configuration method for a wind-storage combined power generation system according to claim 2, characterized in that: The step 1.2 includes the following specific steps: Wind power fluctuations are smoothed based on energy storage equipment. Considering the mutual constraints between fluctuations and energy storage capacity, an objective function including fluctuations and energy storage capacity is constructed based on weighted coefficients. J : ； Where: is the weighting coefficient, and its value is between 0 and 1; the sum of the weighting coefficients of energy storage capacity and fluctuation is 1; is the maximum throughput energy of energy storage; To smooth out the fluctuation of grid-connected power, it consists of two parts: the 1-minute time scale fluctuation and the 10-minute time scale fluctuation. ； Where: for t The original wind power at the moment; for t The grid-connected power obtained after smoothing at the moment; The grid-connected power obtained after stabilization is allowed to fluctuate, so the fluctuation amount in the objective function has a benchmark, and the fluctuation amount exceeding the benchmark is included in the calculation of the objective function as a penalty term; 1-minute time scale fluctuation The calculation formula is: ； 10-minute time scale fluctuation The calculation formula is: ； ； Where: It is the benchmark value of 1-minute fluctuation; It is the 10-minute fluctuation benchmark value; is the maximum fluctuation within the 10-min time scale at time t; According to the principle of Pareto frontier, different The values ​​form multiple sets of optimal solutions, and the optimal solution is selected based on the minimum objective function. value, and on this basis, the simulated annealing algorithm is used to solve the optimal Q , R value.

Images