Battery energy storage adjustment performance evaluation method based on power instruction feature period extraction

Abstract

The application discloses a battery energy storage regulation performance evaluation method based on power instruction feature period extraction. It comprises the following steps: (1) designing an adaptive step formula to accelerate the convergence speed of the improved eagle algorithm; (2) using the improved eagle algorithm to find the global optimal compression offset of the rotating door algorithm; (3) based on the optimization result, using the improved rotating door algorithm to extract the feature trend of the power regulation instruction of the energy storage system, and then dividing it into multiple feature periods; (4) designing an evaluation method for the regulation performance of the energy storage system, and applying it to the regulation performance evaluation of the BESS. The application designs the rotating door algorithm optimized by the improved eagle algorithm, can extract the feature trend of the power regulation instruction of the energy storage system, and then can divide the power regulation instruction into multiple feature periods. The application designs the evaluation method for the regulation performance of the energy storage system, and can effectively evaluate the regulation characteristics of energy storage power stations of different types and different scales.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power systems, in particular to a battery energy storage system regulation performance evaluation method. TECHNICAL BACKGROUND

[0002] Electric energy has the characteristics of being difficult to store, and the load is changing at any time. When the power generation capacity cannot meet the load demand, the power generation and load will be out of balance, which will lead to a large power difference and affect the normal operation of the power system. Therefore, when the load changes, the unit output needs to be changed in time to meet the requirement of load change. Whether the power generation capacity is greater than or less than the load demand, it will lead to system fluctuation and cause safety hazards and economic losses.

[0003] However, due to the delay of the feedback from the load to the system, and the restriction of factors such as climbing speed and regulation capacity, battery energy storage systems are introduced in many regions to improve the regulation performance. For battery energy storage systems, due to the different climbing characteristics and scales of units, and the different types and scales of battery energy storage systems equipped, the regulation performance of different battery energy storage systems is quite different, so an effective method is needed to evaluate whether the regulation performance of the battery energy storage system meets the requirements and to calculate the regulation benefit. At present, there are few evaluation systems for energy storage systems in many regions, and only the direct calculation of indicators is performed without a relatively systematic evaluation process. Therefore, the present application designs a battery energy storage regulation performance evaluation method based on power instruction feature period extraction. SUMMARY

[0004] The present application aims to design a reasonable energy storage system regulation performance evaluation method to effectively evaluate the regulation performance of different energy storage systems. The present application provides an energy storage system regulation performance evaluation method based on an improved swing door trending (SDT) algorithm. The present application uses the swing door algorithm optimized by the improved eagle algorithm to extract the feature time of the power regulation instruction of the energy storage system, and then divides the power regulation instruction into multiple feature periods. Then, an evaluation method for the regulation performance of the energy storage system is designed, and the method is used to evaluate the regulation performance of the BESS.

[0005] The technical scheme of the present application is as follows: a battery energy storage regulation performance evaluation method based on power instruction feature period extraction, which includes the following steps:

[0006] (1) Design an adaptive step formula to improve the optimization process of the eagle algorithm to obtain an improved eagle optimization algorithm;

[0007] (2) Based on the power regulation instruction of the battery energy storage system, use the improved eagle algorithm to find the global optimal compression offset E of the swing door algorithm to obtain an improved swing door algorithm;

[0008] (3) Based on the optimal compression offset E, the feature trend of the battery energy storage system power regulation instruction is extracted by using the improved rotating door algorithm, and the power regulation instruction is divided into multiple feature periods;

[0009] (4) An evaluation method of the regulation performance of the battery energy storage system is designed, and the regulation performance of the BESS is evaluated by using the method.

[0010] In step (1), in order to solve the problems of fixed step size attenuation, insufficient optimization flexibility, and difficulty in balancing wide-range exploration and precise convergence of the traditional eagle algorithm, and to improve the optimization efficiency and accuracy of the algorithm for the compression offset E, the adaptive step size formula of the eagle algorithm optimization process is designed as follows:

[0011]

[0012] Wherein, n is the iteration number, a1, a2, a3, a4, a5 are step size parameters, and the values are 1.1, 0.2, 30, 1.5 and 1 respectively.

[0013] In step (2), the process of the improved eagle algorithm for finding the global optimal compression offset E of the rotating door algorithm is as follows: eagle algorithm parameter initialization, step size factor calculation, determination of fitness function, eagle selection of search space, eagle local search, eagle global exploration, and eagle range reduction.

[0014] In step (3), based on the optimal compression offset E found by the improved eagle algorithm, the feature time of the energy storage system power regulation instruction is extracted by using the SDT and the improved rotating door algorithm, and then the power regulation instruction is divided into multiple feature periods:

[0015] SDT algorithm calculation steps:

[0016] 1) Initialization

[0017]

[0018] In the formula, t0 and x0 are the initial time and the corresponding data value respectively; t1 and x1 are the first time and the corresponding data value respectively; k 1d and k 2d are the initial values of the upper and lower fulcrum door slopes respectively; E is the compression offset;

[0019] 2) Calculate the slope

[0020]

[0021] In the formula: t j and x j are the jth time and the corresponding data value respectively; t k and xk Respectively, the Rth moment and the corresponding data value;

[0022] 3) Slope update

[0023]

[0024] 4) Data extraction

[0025] k 1d ≥k 2d (5)

[0026] If formula (5) is satisfied, the data value x j-1 of the previous moment t j-1 is recorded as the characteristic data, and step 2) is returned, otherwise step 3) is returned.

[0027] In the step (4), an evaluation method of the adjustment performance of the energy storage system is designed, and the adjustment performance of the BESS is evaluated using the method as follows:

[0028] For each characteristic period, the comprehensive adjustment performance index R p i is calculated respectively.

[0029] First, the adjustment accuracy R1 is calculated, which ranges from 0 to 2, and the closer to 2, the higher the adjustment accuracy. The calculation formula of R1 is as follows:

[0030]

[0031] In the formula, ΔP is the deviation between the characteristic instruction at the last moment of each characteristic period and the output of the BESS, and ΔP N is the allowed deviation of adjustment;

[0032] When the adjustment accuracy R1 is the maximum value 2, the adjustment rate R2 and the response time R3 of the characteristic period are both taken as the maximum value 2; otherwise, the adjustment rate R2 is first calculated according to the following formula:

[0033]

[0034] In the formula, P S is the output of the battery energy storage system at the beginning of each characteristic period, P E is the output of the battery energy storage system after 4s at the beginning of each characteristic period, and v N is the standard adjustment rate of the battery energy storage system.

[0035] The calculation formula of the response time R3 is as follows:

[0036]

[0037] Wherein, At is the time required for each characteristic period to reach the deviation amount allowed by adjustment, At N is the standard response time.

[0038] The comprehensive adjustment performance index R of the i-th characteristic period p i The value is calculated as follows:

[0039] R P i = R1x R2x R3 (9)

[0040] The adjustment effect R of the battery energy storage system in a scheduling period p The calculation formula is as follows:

[0041]

[0042] Wherein, N is the number of characteristic periods;

[0043] According to the above analysis, the minimum value of R p is 0, the maximum value is 8, and the average comprehensive performance index R p The closer the value is to 8, the better the adjustment effect of the control strategy.

[0044] The technical scheme provided by the present application has the beneficial effects that:

[0045] The adaptive step formula of the optimization process of the eagle algorithm is designed to speed up its convergence speed, thereby improving the eagle algorithm; based on the power adjustment instruction of the energy storage system, the improved eagle algorithm is used to find the global optimal compression offset of the rotating door algorithm; based on the optimization result, the improved rotating door algorithm is used to extract the characteristic time of the BESS power adjustment instruction, and then it is divided into multiple characteristic periods; an evaluation method of the BESS adjustment performance is designed, and it is used to evaluate the response result of the BESS. The step formula of the improved eagle algorithm is used to speed up its convergence speed, the improved eagle algorithm is used to find the global optimal compression offset of the SDT algorithm, and the rotating door algorithm optimized by the improved eagle algorithm is used to extract the characteristic time of the BESS power adjustment instruction, and then it is divided into multiple characteristic periods. The evaluation method of the adjustment performance of the energy storage system is designed, and it is used to evaluate the adjustment result of the BESS. The adjustment characteristics of different types and different sizes of battery energy storage systems can be effectively evaluated, the calculation speed is significantly improved, and further, it can assist the trading center to settle accounts. BRIEF DESCRIPTION OF DRAWINGS

[0046] The present application will be further described below in conjunction with the drawings:

[0047] Figure 1 is the flowchart of the present application;

[0048] Figure 2 To improve the fitness function change in the optimization process of the algorithm;

[0049] Figure 3 To improve the process of searching for the global optimal compression offset by the algorithm;

[0050] Figure 4 To improve the feature extraction period of the SDT algorithm;

[0051] Figure 5 To improve the BESS regulation performance evaluation process;

[0052] Figure 6 To improve the result of the BESS responding to the power regulation instruction; Specific embodiments

[0053] In order to better understand the purpose, technical solution and technical effect of the present application, the present application will be further explained in combination with the drawings.

[0054] The present application proposes an improved rotating door algorithm for evaluating the regulation performance of energy storage systems, and the accompanying Figure 1 The flowchart of the present application, the implementation process includes the following detailed steps.

[0055] Step 1: The adaptive step formula of the optimization process of the algorithm is as follows:

[0056]

[0057] Wherein, n is the iteration number, a1, a2, a3, a4, a5 are step parameters, and the values are 1.1, 0.2, 30, 1.5 and 1 respectively.

[0058] Step 2: The process of finding the global optimal compression offset E by the improved algorithm is as follows:

[0059] 1) Initialization of algorithm parameters

[0060] Set the initial step (maximum step), the maximum number of iterations;

[0061] 2) Step factor calculation

[0062]

[0063] Wherein, n is the iteration number, a1, a2, a3, a4, a5 are step parameters, and the values are 1.1, 0.2, 30, 1.5 and 1 respectively.

[0064] 3) Determine the fitness function, as shown in the following formula:

[0065]

[0066] where f r and f cr represent the error and compression ratio of the feature period, respectively, x i is the power regulation instruction of the BESS at time t, y i is the extracted feature trend value at time t, N1 is the number of BESS power regulation instructions, N2 is the number of extracted feature times, and a1 and a2 are weights, whose values are 0.68 and 5, respectively;

[0067] 4) The search space of the vulture selection:

[0068]

[0069] where X1(n+1) is the solution of the next iteration of n, which is generated by the first search method X1(n), X best (n) is the optimal solution of the nth iteration, (1-n / N max ) is used to control the search range by controlling the number of iterations, X M (n) represents the average value of the current solution at the nth iteration. rand is a random value between 0 and 1. n and N max represent the current iteration and the maximum number of iterations, respectively.

[0070]

[0071] where dim represents the number of variables, and N represents the number of vultures.

[0072] 5) Local search of the vulture:

[0073] X2(n+1) = X best (n) * Levy(D) + X R (n) + (y-x) * rand (16)

[0074] where X2(n+1) is the solution of the next iteration of X2, which is generated by the second search method X2. D is the dimension space, and Levy(D) is the Levy flight distribution function. X R (n) is a random solution in the range of [1, N] at the nth iteration.

[0075]

[0076] where the parameter s takes 0.01, u and v are random numbers between 0 and 1, and σ is calculated using (18)

[0077]

[0078] where the parameter β takes 1.5.

[0079] 6) Global exploitation by Gull:

[0080] X3(n+1) = (X best (n) - X M (n)) x a - rand + ((UB - LB) x rand + LB) x d (19)

[0081] where X3(n+1) is the solution of the next iteration of n, generated by the third search method X3. X best (n) refers to the approximate location of the prey before the nth iteration (optimal solution), X M (n) represents the average value of the current solution at the nth iteration. rand is a random value between 0 and 1. a and d are both 0.1, LB represents the lower limit of the given problem, and UB represents the upper limit of the given problem.

[0082] 7) Gull narrowing range:

[0083] X4(n+1) = QF x X best (n) - (G1 x X(n) x rand) - m x Levy(D) + rand x G1 (20)

[0084]

[0085] where X4(n+1) is the solution of the next iteration of n, generated by the fourth search method X4. QF represents the quality function for balancing the search strategy, G1 represents the various movements of the gull in tracking the prey during hunting, G2 is a decreasing function from 2 to 0, representing the flight slope of the gull used to follow the prey from the first position 1 to the last position n, X(n) is the current solution of the nth iteration, represented by (22) and (23) respectively.

[0086] G1 = 2 x rand - 1 (22)

[0087]

[0088] where n is the number of iterations, a1, a2, a3, a4, a5 are step parameters, taking values of 1.1, 0.2, 30, 1.5 and 1 respectively. 8) Determine whether the iteration termination condition is met, if it is met, output the current E value as the global optimal compression offset of the SDT algorithm, if it is not met, return to step 4). Taking the power regulation command and response result of a certain battery energy storage system within one hour as the research object, the time resolution is 4s, the installed capacity of BESS is 100MW / 50MWh, and the optimal compression offset of the improved SDT by adaptive Gull search is used. The search process and results are shown in Figures 1 and 2. Figure 2 and Figure 2 Figure 3As shown, Step 3 uses the improved Skyhawk algorithm to find the optimal compression offset E, and then uses the improved rotating door algorithm to extract feature moments, thereby dividing the power adjustment command into multiple feature time periods: SDT algorithm calculation steps: 1) Initialization In the formula, t0 and x0 represent the initial time and the corresponding data value, respectively; t1 and x1 represent the first time and the corresponding data value, respectively; k 1d and k 2d These are the initial values ​​of the slopes of the upper and lower fulcrum gates, respectively; E is the compression offset. 2) Calculate the slope In the formula: t j and x j These represent the j-th time point and its corresponding data value; t k and x k These are the data values ​​at the Rth time point, respectively. 3) Slope Update 4) Data Extraction k 1d ≥k 2d (27)

[0089] If equation (23) is satisfied, then the previous time t will be... j-1 Data value x j-1 Record it as feature data and return to step 2); otherwise return to step 3.

[0090] The feature time periods extracted using the improved SDT are shown in the attached figure. Figure 4 As shown.

[0091] Step 4: Flowchart of the evaluation method for the regulation performance of the energy storage system is attached. Figure 5 As shown, the process of evaluating the regulation performance of BESS using this method is as follows:

[0092] For each characteristic time period, the comprehensive regulation performance index R is calculated separately. p i :

[0093] First, calculate the adjustment accuracy R1, which ranges from 0 to 2. The closer it is to 2, the higher the adjustment accuracy. The formula for calculating R1 is as follows:

[0094]

[0095] In the formula, ΔP represents the deviation between the characteristic command at the last moment of each characteristic period and the output of the battery energy storage system. Nto adjust the allowed deviation amount;

[0096] When the adjustment precision R1 is the maximum value 2, the adjustment rate R2 and the response time R3 of the feature period are both taken as the maximum value 2; otherwise, the adjustment rate R2 is first calculated according to the following formula:

[0097]

[0098] P S is the output of the battery energy storage system at the beginning of each feature period, P E is the output of the battery energy storage system after 4s at the beginning of each feature period, v N is the standard adjustment rate of the battery energy storage system, and since the standard adjustment rate of the battery energy storage system is 1.5% / min, v N is taken as 0.1MW / 4s.

[0099] The calculation formula of the response time R3 is as follows:

[0100]

[0101] In the formula, Δt is the time required for each feature period to reach the allowed deviation amount of adjustment, Δt N is the standard response time, which is taken as 1min.

[0102] The comprehensive adjustment performance index R p i of the i-th feature period is calculated according to the following formula:

[0103] R P i = R1×R2×R3 (31)

[0104] The adjustment effect R p of the battery energy storage system in a scheduling period is calculated according to the following formula:

[0105]

[0106] In the formula, N is the number of feature periods;

[0107] According to the above analysis, the minimum value of R p is 0, the maximum value is 8, and the average comprehensive performance index R p value closer to 8 indicates that the adjustment effect of the regulation strategy is better.

[0108] According to the adjustment performance evaluation model, the power adjustment command and response result of a certain battery energy storage system within one hour are taken as the research object, the time resolution is 4s, and the installed capacity of the BESS is 100MW / 50MWh. The response power adjustment command of the energy storage system is shown in the following table:Figure 5 The calculation results of the regulation accuracy R1, the regulation rate R2 and the response time R3 in the evaluation system are shown in Table 1 as follows.

[0109] The calculation result of the regulation accuracy R1 at the characteristic period of 75% is 2, indicating that the energy storage system can accurately respond to the power regulation instruction at most times; the calculation result of the regulation accuracy R1 at the characteristic period of 7% is 0, which is caused by the fact that the power regulation instruction exceeds the maximum charge and discharge power limit of the energy storage system; and the average value of the regulation accuracy R1 in the entire scheduling period is 1.84, indicating that the regulation accuracy of the energy storage system in the entire scheduling period is high.

[0110] The calculation result of the regulation rate R2 at the characteristic period of 83% is 2, indicating that the regulation speed of the energy storage system is fast at most times; the calculation result of the regulation rate R2 at the characteristic period of 11% is 0, which is caused by the fact that the power regulation instruction exceeds the maximum charge and discharge power limit of the energy storage system, resulting in that the energy storage system cannot quickly respond to the power regulation instruction; and the average value of the regulation rate R2 in the entire scheduling period is 1.70, indicating that the regulation speed of the energy storage system in the entire scheduling period is relatively fast.

[0111] The calculation result of the response time R3 at the characteristic period of 96% is 2, indicating that the energy storage system can quickly respond to the power regulation instruction at most times; the calculation result of R3 at the characteristic period of 0% is 0; and the average value of the response time R3 in the entire scheduling period is 1.95, indicating that the response time of the energy storage system in the entire scheduling period is short, and the energy storage system can quickly respond to the power regulation instruction.

[0112] Table 1 Calculation results of the battery energy storage system evaluation index system

[0113]

[0114]

[0115] The calculation results of the comprehensive evaluation index R of each power regulation characteristic period are shown in Table 2 as follows. p i

[0116] As can be seen from Table 2, the calculation results of the comprehensive evaluation index R of each power regulation characteristic period are shown in Table 2 as follows. p i The calculation result of the comprehensive evaluation index R of each power regulation characteristic period is the maximum value 8 at the characteristic period of 67%, indicating that the battery energy storage system can quickly and accurately respond to the power regulation instruction at most times; the calculation result of the comprehensive evaluation index R of each power regulation characteristic period is the minimum value 0 at the characteristic period of 12.5%, which is caused by the fact that the power regulation instruction suddenly increases or decreases issued by the power grid, exceeding the maximum charge and discharge power limit of the energy storage system; and the average value of the comprehensive evaluation index R of each power regulation characteristic period is 1.95, indicating that the battery energy storage system can quickly and accurately respond to the power regulation instruction in the entire scheduling period. p i The calculation result of the comprehensive evaluation index R of each power regulation characteristic period is the maximum value 8 at the characteristic period of 67%, indicating that the battery energy storage system can quickly and accurately respond to the power regulation instruction at most times; the calculation result of the comprehensive evaluation index R of each power regulation characteristic period is the minimum value 0 at the characteristic period of 12.5%, which is caused by the fact that the power regulation instruction suddenly increases or decreases issued by the power grid, exceeding the maximum charge and discharge power limit of the energy storage system; and the average value of the comprehensive evaluation index R of each power regulation characteristic period is 1.95, indicating that the battery energy storage system can quickly and accurately respond to the power regulation instruction in the entire scheduling period. p i ​The average of the calculation result is 6.51, which is close to the maximum value 8, indicating that the battery energy storage system has good tracking effect on the power regulation instruction in the whole scheduling period.

[0117] Table 2 comprehensive evaluation index R of each power regulation characteristic period p i The calculation result of

[0118]

Claims

1. A battery energy storage regulation performance evaluation method based on power instruction characteristic period extraction, characterized in that, The method comprises the following steps: (1) designing an adaptive step formula to improve the optimization process of the algorithm, and obtaining an improved algorithm; (2) based on the power regulation instruction of the battery energy storage system, using the improved algorithm to find the global optimal compression offset E of the rotating door algorithm, and obtaining the improved rotating door algorithm; (3) based on the optimal compression offset E, using the improved rotating door algorithm to extract the characteristic trend of the power regulation instruction of the battery energy storage system, and dividing the power regulation instruction into multiple characteristic periods; (4) designing an evaluation method for the regulation performance of the battery energy storage system, and using the method to evaluate the regulation performance of the battery energy storage system; The adaptive step of the improved algorithm in step (1) is obtained by the following formula: Wherein, n is the number of iterations, a1, a2, a3, a4 and a5 are respectively 1.1, 0.2, 30, 1.5 and 1; The BESS regulation performance evaluation method designed in the step (4) is as follows: for each characteristic period i, the regulation accuracy R1, the regulation speed R2 and the response time R3 indexes in the period are respectively calculated, and then the comprehensive regulation performance index is obtained The regulation accuracy R1 is in the range of 0-2, the closer to 2, the higher the regulation accuracy, and the calculation formula of R1 is as follows: where ΔP is the deviation between the characteristic command at the last moment of each characteristic period and the output of the battery energy storage system, ΔP N is the adjustment allowed deviation amount; When the regulation accuracy R1 is the maximum value 2, the regulation rate R2 and the response time R3 of the characteristic period are both taken as the maximum value 2; otherwise, the regulation rate R2 needs to be calculated according to the following formula: where P S is the battery energy storage system output at the beginning of each characteristic time period, P E is the battery energy storage system output 4s after the beginning of each characteristic time period, v N is the standard regulation rate of the battery energy storage system, which corresponds to v N has a value of 0.1 MW / 4s; The calculation formula of the response time R3 is as follows: where Δt is the time required to reach the allowed deviation from the regulation for each characteristic period, Δt N is the standard response time, which is taken as 1 min; The comprehensive adjustment performance index R of the ith feature period p i The value calculation formula is as follows: R P i = R1 x R2 x R3 (5) The regulation effect R of the battery energy storage system in a dispatching cycle p The calculation formula is as follows: In the formula, N is the number of characteristic periods; According to the above analysis, the minimum value of R p is 0, the maximum value is 8, and the average comprehensive performance index R p value closer to 8, the better the adjustment effect of the regulation strategy.

2. The battery energy storage regulation performance evaluation method based on power instruction characteristic period extraction according to claim 1, characterized in that, In step (2), the improved algorithm is used to find the global optimal compression offset of the SDT algorithm, wherein the fitness function of the algorithm is: wherein f r and f cr respectively represent the error and compression ratio of the feature period, x i is the power adjustment instruction of the BESS at time t, y i is the extracted feature trend value at time t, N1 is the number of BESS power adjustment instructions, N2 is the number of extracted feature times, and α1 and α2 are weights, the values of which are 0.68 and 5 respectively. 3.The battery energy storage regulation performance evaluation method based on power instruction characteristic period extraction according to claim 1, characterized in that, In step (3), the characteristic trend of the power regulation instruction of the battery energy storage system is extracted by using the improved rotating door algorithm, and the characteristic trend is segmented.

4. The battery energy storage regulation performance evaluation method based on power instruction characteristic period extraction according to claim 1, characterized in that, In step (4), the designed evaluation method of the regulation performance of the battery energy storage system is used to evaluate the result of the response power regulation instruction of the battery energy storage system, so as to analyze the advantages and disadvantages of the regulation performance.

Images