Self-adaptive energy storage countercurrent control method integrating load edge identification and electricity price scheduling

Abstract

According to the adaptive energy storage countercurrent control method integrating load edge recognition and electricity price scheduling, the energy storage system is guided to actively release power in the load fluctuation active interval, energy storage resources are fully mobilized, and the energy release efficiency is improved; the scheduling strategy has edge identification and dynamic amplitude limiting capabilities, can actively adjust energy storage response in a high-load fluctuation period, and plays a role in dynamic voltage regulation of a power grid; since the PCS outputs electric energy to replace part of high-price power grid electricity purchasing, the overall operation electric charge expenditure is reduced, and the economic benefit of the system is improved; compared with a traditional scheme, the method has the advantages that higher dynamic adjusting capacity and response stability are shown under the complex load condition, and wider application adaptability is achieved.

Description

Technical Field

[0001] This application relates to the field of energy storage system scheduling and control technology, specifically to an adaptive energy storage reverse flow control method that integrates load edge identification and electricity price scheduling. Background Technology

[0002] With the large-scale integration of distributed photovoltaic and energy storage systems into the power grid, the problem of reverse power (current backflow) in the power system is becoming increasingly serious. Especially when the load fluctuates rapidly or the control strategy has a response lag, it can easily lead to grid protection malfunctions, power metering errors, and power quality degradation. Traditional energy storage reverse current control methods cannot ensure flexible and stable control of grid output power in such complex dynamic environments, and they cannot find the optimal balance between ensuring grid output power stability and minimizing reverse current under the constraint of maximizing energy storage benefits. Summary of the Invention

[0003] This application provides an adaptive energy storage reverse current control method that integrates load edge identification and electricity price scheduling, aiming to ensure stable grid output power, minimize reverse current, and optimize energy storage revenue.

[0004] To achieve this objective, the following technical solution is adopted in this application:

[0005] An adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling is provided, including:

[0006] Based on the current time User load sequences collected or constructed by the power grid Energy storage power station Grid output power Power output executed by PCS and daily electricity price function And by predicting the sudden change trend of user load, calculate in Timing correction control quantity ;

[0007] Dynamic adjustment Limiting threshold for constant energy storage reverse flow control And / or adaptively adjust the limiting factor for energy storage reverse flow control the following day. ;

[0008] In Constantly control the amplitude of energy storage reverse current to limit Within, and / or the next day, the control energy storage reverse current amplitude is limited to With the goal, in The time-scheduling control quantity is Energy storage capacity.

[0009] Preferably, the method for predicting sudden changes in user load includes the following steps:

[0010] A1, Calculate rising edge identification and falling edge identification ;

[0011] A2, Parallel Judgment Does it exceed the first threshold? Does it exceed the second threshold?

[0012] when When the first threshold is exceeded, it is determined that the user load has a sudden upward trend; when When the second threshold is exceeded, it is determined that the user load has a sudden downward trend.

[0013] Preferably, , It is calculated using the following formula (3):

[0014]

[0015] In formula (3), This represents the convolution operator;

[0016] These represent the matched filter functions used to detect rising and falling edges, respectively.

[0017] .

[0018] Preferably, when a sudden upward trend in user load is predicted, The calculation method is expressed by the following formula (1):

[0019] (1)

[0020] In formula (1), Indicates the first dynamic correction amount;

[0021] This represents the second dynamic correction amount, which is and The difference,

[0022] Indicates the current time The value of the dead zone buffer.

[0023] Preferably, The calculation method is expressed by the following formula (2).

[0024] (2)

[0025] In formula (2), , These represent the prediction errors respectively. , Integral error and The scaling factor of the differential error;

[0026] , These represent the prediction errors respectively. Integration error and prediction error The differential error;

[0027] The calculation method is as follows:

[0028] .

[0029] Preferably, when a sudden downward trend in user load is predicted, The calculation method is expressed by the following formula (5):

[0030]

[0031] In formula (5), Indicates that it occurs at the current time. The previous Historical secondary energy storage release;

[0032] This indicates the distance from the current time. Recent historical energy storage release counts;

[0033] Indicates the first The lowest point in the historical secondary energy storage release levels;

[0034] Indicates the current time The value of the dead zone buffer.

[0035] Preferably, (6);

[0036] In formula (6), Indicates the current time The next The value of the dead zone buffer at any given moment;

[0037] Indicates coefficient;

[0038] (7)

[0039] in, This represents the second threshold. This represents the first threshold;

[0040] numerical values , different.

[0041] Preferably, the limiting factor is adaptively adjusted. The method includes the following steps:

[0042] B1, Construct a comprehensive profit function to calculate the adjusted control amount. Perform power correction to the theoretical benefit at the end of the cycle. and actual benefits ;

[0043] B2, Calculate the electricity surplus ratio as a revenue indicator. Ratio to revenue forecast ;

[0044] B3, with and The limiting factor for the day After making adjustments, the amplitude limiting factor for the next day is obtained. .

[0045] Preferably,

[0046]

[0047]

[0048] Indicates the energy storage power station at the current moment The discharge power, Indicates the energy storage power station at the current moment The charging power;

[0049] The daily cumulative ideal reverse power energy, This indicates the daily cumulative actual reverse power consumption;

[0050] As a countercurrent penalty factor;

[0051] This indicates the electricity price in high-price zones; This indicates the electricity price in the low-price zone;

[0052] This indicates the actual electricity price.

[0053] Preferably,

[0054]

[0055]

[0056] , These represent the operating periods during high electricity price periods and low electricity price periods, respectively.

[0057] , These represent the changes in SOC during operation in high-electricity-price periods and low-electricity-price periods, respectively.

[0058] Preferably,

[0059]

[0060] in This is a flag indicating abnormal program operation.

[0061] This indicates the threshold value at 00:00 on the next day following the current day.

[0062] This indicates the threshold value for the current day at 0:00.

[0063] Preferably, the limiting threshold The dynamic adjustment method includes the following steps:

[0064] C1, calculate the user load sequence mean Standard deviation Simultaneously calculate the current power grid. mean as well as User load fluctuation level ;

[0065] C2, Dynamically calculates the limit increment The calculation method is expressed as follows:

[0066]

[0067] for The learning rate is automatically adjusted.

[0068] Indicates the fluctuation threshold; Indicates the mean threshold;

[0069] C3, Calculation With Sum as .

[0070] Preferably, in The time-scheduling control quantity is The method for storing energy power includes the following steps:

[0071] D1, Construct the offline perturbation optimization function ,

[0072] Indicates the current time Control correction instructions;

[0073] The first one to be optimized A composite gain factor, This indicates the number of composite gain factors to be optimized;

[0074] Indicates origin from user load sequence and grid output power Feedback function;

[0075] D2, Finding the optimal control parameter vector The search method includes the following steps:

[0076] D21, Define the initial gain vector and the corresponding perturbation step size vector and define the performance objective function. ,

[0077] This represents the grid output power at the k-th time step under parameter vector perturbation in the test case.

[0078] This represents the grid output power under the initial gain vector;

[0079] Indicates the execution of system test cases Ideal data for the output power of the secondary power grid;

[0080] D22, for the first Add perturbation to each composite gain factor ;

[0081] D23, after adding the perturbation, perform one control cycle and recalculate the objective function value. ;

[0082] D24, Determine Is it less than , This represents the minimum objective function value calculated by adding perturbations in the past. ,

[0083] If so, then the added perturbation is deemed valid, and it is set to... ;

[0084] If not, proceed to step D25;

[0085] D25 determines whether the same composite gain factor has undergone two reverse perturbations.

[0086] If so, restore the second added reverse perturbation to the state of the first added reverse perturbation, and then proceed to step D26;

[0087] If not, add a reverse perturbation, making Then return to step D22;

[0088] D26, Determine whether the search for the optimal control parameter vector has been achieved. The iteration termination condition,

[0089] If so, then based on the found optimal control parameter vector Calculate control correction instructions ;

[0090] If not, return to step D22.

[0091] Preferably, the optimal control parameter vector is found. The iteration termination condition is:

[0092] When the sum of all disturbance amplitudes in the optimization process Less than the set tolerance and / or performance objective function value Less than the error threshold And / or the difference in performance objective function values ​​after two adjacent optimizations. Less than the difference threshold If so, then the iteration termination condition is determined.

[0093] This application has the following beneficial effects:

[0094] 1. By predicting the sudden change trend of user load and performing dead-zone correction, the system can better handle... Timing correction control quantity The calculations are more accurate, which can effectively avoid backflow caused by sudden load changes and improve the response stability and safety of the energy storage system.

[0095] 2. By introducing a limiting factor During the scheduling process, a dynamic balance between economy and safety was achieved.

[0096] 3. By introducing a composite gain factor disturbance optimization mechanism and iterating the disturbance feedback parameter vector, the grid power deviation from the target value is minimized while taking into account the reverse flow risk that may occur during system operation. At the same time, the response sensitivity and error convergence speed of the feedback adjustment module to changes in system state are improved.

[0097] 4. By finding the optimal control parameter vector This makes the correction control quantity The calculations are more precise, and the control is more flexible. Attached Figure Description

[0098] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments of this application will be briefly described below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0099] Figure 1 This is a diagram illustrating the implementation steps of the adaptive energy storage reverse current control method that integrates load edge identification and electricity price scheduling, as provided in the embodiments of this application.

[0100] Figure 2 This is a flowchart of the PID energy storage reverse flow control method provided in this embodiment;

[0101] Figure 3 This is a schematic diagram of the matched filtering effect;

[0102] Figure 4 This is a schematic diagram for assessing user load fluctuations;

[0103] Figure 5 This is a comparison chart of the control effects of the adaptive energy storage countercurrent control method provided in this application and the traditional method. Detailed Implementation

[0104] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The accompanying drawings are for illustrative purposes only and represent schematic diagrams rather than actual pictures. They should not be construed as limiting this patent. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0105] The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling provided in this application embodiment, such as Figure 2 As shown, the overall process is divided into a three-layer structure, including:

[0106] The first layer is the data sensing layer, used to collect data operation parameters of the power grid and power plants, including User load sequence at time Energy storage power station Grid output power Power output executed by PCS and daily electricity price function ;

[0107] The second layer is the intelligent decision-making layer, which integrates functions such as load edge identification, electricity price analysis, feedback control, and strategy self-tuning. The intelligent decision-making layer includes a feedback adjustment module, a load edge identification and limiting intervention module, an electricity price-driven dispatching module, a fluctuation assessment and limiting self-adjustment module, and a parameter adaptive optimization module.

[0108] The third layer is the execution response layer, which outputs the regulated energy storage power command to the energy storage converter PCS to complete the actual power scheduling execution.

[0109] The following combination Figure 2 The workflow of the three-layer structure and the technical content of the five modules in the intelligent decision-making layer are explained in detail below:

[0110] First, the data perception layer performs data preprocessing, and at the current moment... User load sequences within a fixed-width sliding window are collected synchronously from the power grid. Energy storage power station Grid output power Power output executed by PCS Daily electricity price function The collected data is then input into the feedback regulation module, which is set to maintain the grid output power. Within safety limits Within, and by calculating the current prediction error as well as The derivative and integral of the product are used to output the first dynamic correction value through the composite gain factor. .

[0111] The calculation method is expressed by the following formula (2).

[0112] (2)

[0113] In formula (2), , These represent the prediction errors respectively. , Integral error and The scaling factor of the differential error;

[0114] , These represent the prediction errors respectively. Integration error and prediction error The differential error;

[0115] The calculation method is as follows:

[0116] The load edge identification and limiting intervention module uses a sliding array to identify dead-zone limiting and load edges. By identifying sudden changes in user load, it intervenes in energy storage output in advance, overcoming the problem of lag in feedback regulation response. Specifically, it performs the following actions:

[0117] First, based on the current moment Constructed user load sequence Within a fixed-width sliding time window, for The signal undergoes matched filtering, and the processing method is expressed by the following formula (3):

[0118]

[0119] In formula (3), The convolution operator is physically continuous in time, but it uses discrete sampling during computation. , These represent rising edge recognition and falling edge recognition, respectively.

[0120] , These represent the matched filter functions used to detect rising and falling edges, respectively.

[0121] The expression is:

[0122]

[0123] The expression is:

[0124]

[0125] This represents an intermediate variable used in the convolution operation. The value range depends on the time width of the load sliding window and is set according to the actual control cycle. For example, a value of 10.

[0126] The effect of matched filtering on the rising and falling edges of the user load is as follows: Figure 3 As shown. Figure 3In the diagram, the blue line represents the actual user load curve, while the black and red lines are the rising edge response curve and falling edge response curve extracted by the matched filtering algorithm, respectively. The matched filter is constructed based on a typical edge template and can sensitively capture and directionally respond to abrupt changes in the user load.

[0127] from Figure 3 As can be seen, when the user load changes rapidly, the matched filter output produces a significant peak (rising edge) or trough (falling edge) response at the corresponding time point, which is highlighted by the elliptical marking area. This characteristic verifies that the matched filter has good directional recognition capability and robustness for transient change signals, and can effectively remove background fluctuation noise and extract typical edge features.

[0128] User load sequence The construction method is briefly described as follows:

[0129] At the current moment User load data collected at several previous time points is arranged in chronological order to form a user load sequence. .

[0130] When detected When the first threshold is exceeded, it is determined that the user load is about to rise rapidly, and energy storage is maintained at this time. Release level at any time and continue The duration of the operation reduces the lag in energy storage feedback control when the user load decreases, thus preventing the user load from dropping or falling below the maximum value when the PCS reaches its peak, thereby avoiding backflow.

[0131] When a sudden upward trend in user load is predicted (a rapid increase is imminent), the feedback adjustment module calculates the corrected control quantity using the following formula (1). :

[0132] (1)

[0133] In formula (1), Indicates the first dynamic correction amount;

[0134] This represents the second dynamic correction amount, which is and The difference,

[0135] Indicates the current time The value of the dead zone buffer.

[0136] The method of obtaining is It uses initially given data and then updates it dynamically.

[0137] When detected If the second threshold is exceeded, it is determined that the user load is about to drop rapidly. At this point, the energy storage release level is forcibly lowered to the trough of the previous 10 cycles. This reduces the load level when the user load decreases, preventing backflow. Then, the estimated output is maintained continuously. Duration, to offset the reverse flow trend caused by a sudden drop in load.

[0138] When a sudden downward trend in user load is predicted, the feedback regulation module calculates the corrective control quantity using the following formula (5). :

[0139]

[0140] In formula (5), Indicates that it occurs at the current time. The previous Historical secondary energy storage release;

[0141] This indicates the distance from the current time. Recent historical energy storage release counts;

[0142] Indicates the first The lowest point in the historical secondary energy storage release levels; for example... At that time, the trough value was 1531.

[0143] Preferably, when a rising or falling edge change in user load is detected, the system will dynamically adjust the corresponding dead zone limiting threshold to twice the original set value to expand the buffer range of energy storage output control, effectively threshold the reverse flow risk caused by short-term load changes, and simultaneously reduce the instantaneous power level released by the energy storage device to ensure the safe and stable operation of the power system.

[0144] In addition, this application controls redundancy by expanding the algorithm for user load fluctuations and makes the dead zone limit value (dead zone buffer value) correspond to the rise time in the corresponding recognition mode. or descent time By maintaining consistency, more precise and adaptive amplitude limiting control is achieved. The method for increasing the redundancy of algorithmic control to mitigate user load fluctuations is expressed as follows:

[0145] (6)

[0146] In formula (6), Indicates the current time The next The value of the dead zone buffer at any given moment;

[0147] Indicates coefficient;

[0148] (7)

[0149] in, This represents the second threshold. This represents the first threshold.

[0150] After the above dead zone correction process The signal is the actual execution command issued to the PCS, which effectively avoids backflow caused by sudden load changes and improves the response stability and safety of the energy storage system.

[0151] Using the above method, it was calculated that... Timing correction control quantity Subsequently, the electricity price-driven dispatch module integrates day-ahead electricity price data to optimize energy storage charging and discharging plans, maximizing revenue. Specifically, the electricity price-driven dispatch module performs the following actions:

[0152] First, by calling the day-ahead electricity price function ( This refers to the number of hours in a 24-hour period, from 0:00 to 23:00 (when the electricity price is equal), dividing the daily electricity price range into high-price and low-price segments. The system control logic is as follows: in the low-price segment, the energy storage charging strategy is prioritized, while in the high-price segment, the discharging strategy is prioritized, in order to maximize the peak-valley arbitrage benefits of the energy storage system.

[0153] Then, a comprehensive profit function is constructed that takes into account both economic efficiency and security. Let... Indicates the energy storage power station at the current moment The discharge power, Indicates the energy storage power station at the current moment The charging power, The cumulative reverse power consumption per day. This is the reverse flow penalty factor (unit: yuan / kWh). According to... , Dynamic adjustment of status To achieve a balance between charging and discharging and maximize the overall benefit, the system's comprehensive benefit calculation model is as follows:

[0154]

[0155]

[0156] in: This represents the theoretical return based on the prediction model; This represents the actual revenue calculated after the operating cycle ends. The daily cumulative ideal reverse power energy, This indicates the daily cumulative actual reverse power consumption; This indicates the electricity price in high-price zones; This indicates the electricity price in the low-price zone;

[0157] Indicates the actual electricity price; , These represent the discharge power and charging power of the energy storage, respectively.

[0158] Countercurrent penalty factor The penalty factor is determined based on the penalty requirements for reverse power conditions at different power plants, and is generally a constant.

[0159] Furthermore, the revenue indicator of the remaining electricity ratio is defined. Ratio to revenue forecast Used to measure economic optimization problems:

[0160]

[0161]

[0162] , These represent the operating periods during high electricity price periods and low electricity price periods, respectively.

[0163] , These represent the changes in SOC during operation in high-electricity-price periods and low-electricity-price periods, respectively.

[0164] At the end of each day, the profit is calculated based on the difference between the estimated and actual revenue, and the remaining electricity. The correction factor, for the next day Make corrections to improve profitability.

[0165]

[0166] in This is a flag indicating abnormal program operation, representing abnormal working conditions and abnormal setting logic under manual intervention: special cases such as forced stop, forced start, and insufficient battery charge. These cases do not participate in the profit optimization steps. This indicates that the system is operating normally. This indicates that the system is not operating normally. This indicates the threshold value at 00:00 on the next day following the current day.

[0167] This indicates the threshold value for the current day at 0:00.

[0168] Table 1 below is a list of... Example of an adaptive tuning process:

[0169] Table 1

[0170] Serial Number (Yuan) (Yuan) 1 31.0 1400.0 1450.0 96.6 93.3 1.04 0.97 0 31.0 2 31.0 900.0 800.0 82.0 80.0 0.89 0.98 0 27.1 3 31.0 1450.0 1390.0 96.0 95.0 0.96 0.99 0 29.4 4 31.0 1450.0 500.0 96.0 40.0 0.34 0.42 1 31.0

[0171] The electricity price-driven dispatch module in this application introduces a limiting factor. A dynamic trade-off between economy and safety was achieved during the scheduling process. Analysis of actual operational data from multiple scheduling days shows that: based on... Through adaptive adjustments, the system can improve the robustness of scheduling response while ensuring benefits, thereby achieving the optimized scheduling goal of maximizing long-term economic benefits and ensuring system security.

[0172] In this application, the fluctuation assessment and amplitude limiting self-adjustment module implements reverse flow limit thresholds based on the statistical characteristics of user load data. ( The dynamic adjustment of the limit threshold of the energy storage reverse current amplitude at any time is used to improve the economy and robustness while ensuring the stability of system operation.

[0173] Specifically, the system first defines key statistical indicators in the user load curve, including peak power. and volatility level Within a sliding time window of fixed width, the user load sequence is... Perform statistical processing and calculate its mean. Standard deviation Simultaneously calculate the current power grid. mean The expression is as follows:

[0174]

[0175] in, The time width of the sliding window is set according to the actual control cycle.

[0176] Furthermore, define User load The fluctuation level is calculated using the mean. The mean value is used as an indicator to measure the intensity of user load fluctuations.

[0177]

[0178] A scaling factor that represents the level of fluctuation in user load.

[0179] Set fluctuation threshold and mean threshold Based on the actual monitored standard deviation and mean changes, different amplitude limiting adjustment strategies are adopted, and the adjustment model is shown below:

[0180]

[0181]

[0182] in for The learning rate is self-adjusted, and when t=0, in accordance with The daily update, in other words, the dynamic changes on the daily scale determine the baseline of the Limit, while the changes on the hourly scale are used to make subtle dynamic adjustments to it.

[0183] When standard deviation with the mean When all values ​​are within the set stable range, it indicates that the load fluctuation is relatively stable, and the system performs a limiting tightening operation to improve the economy and stability of operation;

[0184] When standard deviation with the mean When both values ​​increase sharply or fluctuate significantly, it indicates that the load fluctuation is intensifying, and the system should relax the limit to ensure the system's flexible response capability.

[0185] In other cases, the current limit value will remain unchanged to ensure stable system operation.

[0186] like Figure 4 As shown, the blue curve represents the actual load curve of users, showing obvious high-frequency disturbances and non-periodic fluctuations; the orange curve represents the mean change trend of the actual load curve of users within the sliding window, reflecting the medium- and long-term changes in the load center level; the black curve represents the standard deviation (i.e., load fluctuation intensity) change trend within the same window, used to quantify the disturbance level of user load; the red curve is the mean of the standard deviation curve (i.e., fluctuation benchmark), providing benchmark support for dynamic threshold adjustment.

[0187] Through the above assessment of the fluctuation characteristics of user load, the limiting threshold was set. The adaptive adjustment ensures the stable operation of the power system under multi-load conditions. The parameter adaptive optimization module provided in this application performs offline perturbation optimization on the parameter vector of feedback control based on existing user load values ​​and error correction parameters.

[0188] Specifically, to improve the response sensitivity and error convergence speed of the feedback control module to changes in system state, this application introduces a composite gain factor disturbance optimization mechanism. By iterating the disturbance feedback parameter vector, it minimizes the deviation of the power grid from the target value while taking into account the reverse flow risk that may occur during system operation. This method does not depend on the specific form of the controller and can be used for adaptive parameter correction under various feedback strategies. Moreover, it does not rely on a fixed proportional, integral, or derivative structure, but rather treats the control correction amount as being jointly determined by a set of feedback composite gain factors.

[0189] In this application, a composite gain factor perturbation optimization mechanism is introduced. The time-scheduling control quantity is The specific method for increasing energy storage capacity is as follows:

[0190] First, construct the offline perturbation optimization function. :

[0191] (15)

[0192] Indicates the current time Control correction instructions;

[0193] The first one to be optimized A composite gain factor (control parameter vector). This indicates the number of composite gain factors to be optimized;

[0194] Indicates origin from user load sequence and grid output power The feedback function, such as the pre-prediction error of the error value (e.g. ), error integral (e.g.) ), rate of change (e.g.) )wait.

[0195] For automatic optimization of control parameter vectors The self-tuning optimization process includes the following steps:

[0196] D21, Define the initial gain vector and the corresponding perturbation step size vector and define the performance objective function. The following is based on For example.

[0197]

[0198] This represents the grid output power at the k-th time step under parameter vector perturbation in the test case.

[0199] This represents the grid output power under the initial gain vector;

[0200] Indicates the execution of system test cases Ideal data for the output power of the secondary power grid, i.e., the average value without negative values, and .

[0201] Used to characterize the deviation between the perturbed gain vector and the desired grid output power. Reflects the overall performance index of the control system under the current parameter settings. When The smaller the value, the closer the gain vector after disturbance correction is to the target reference value, the lower the grid deviation, and the better the control accuracy and response stability. Through analysis of... Continuous calculation and comparison can achieve adaptive evaluation of the effect of parameter perturbation, thereby guiding the dynamic optimization and final convergence of the gain vector.

[0202] D22, for the first Add perturbation to each composite gain factor ;

[0203] D23, after adding the perturbation, perform one control cycle and recalculate the objective function value. ;

[0204] D24, Determine Is it less than , This represents the minimum objective function value calculated by adding perturbations in the past. ,

[0205] If so, then the added perturbation is deemed valid, and it is set to... ;

[0206] If not, proceed to step D25;

[0207] D25 determines whether the same composite gain factor has undergone two reverse perturbations.

[0208] If so, then restore the second added reverse perturbation to the state of the first added reverse perturbation (let...). Then proceed to step D26;

[0209] If not, add a reverse perturbation, making Then return to step D22;

[0210] D26, Determine whether the search for the optimal control parameter vector has been achieved. The iteration termination condition,

[0211] If so, then based on the found optimal control parameter vector Calculate control correction instructions ;

[0212] If not, return to step D22.

[0213] Preferably, ; ; .

[0214] Preferably, the optimal control parameter vector is found. The iteration termination condition is:

[0215] When the sum of all disturbance amplitudes in the optimization process Less than the set tolerance Or the performance objective function value Less than the error threshold Or the difference in the performance objective function value after two adjacent optimizations. Less than the difference threshold If so, then the iteration termination condition is determined.

[0216] Preferably, in this application, the tolerance Error threshold Difference threshold When any one of the termination conditions is met, the optimization process can be terminated, and the current control parameter vector... This is considered a convergent solution. This tolerance mechanism ensures that the algorithm maintains control accuracy while avoiding resource waste caused by excessive iteration, further improving the practicality and efficiency of the parameter adaptation module in multi-scenario scheduling.

[0217] This application seeks the optimal control parameter vector. The process is illustrated in Table 2 below:

[0218] Table 2

[0219] Iteration rounds Disturbed parameters <![CDATA[G = [G1, G2, G3]]]> Latest Error Optimal error Total step size ∑ΔG state 0 initial value [0.50,0.01,0.10] <![CDATA[1.2*10 4 ]]> <![CDATA[1.2*10 4 ]]> / / 1 <![CDATA[G 1 ↑]]> [0.60,0.01,0.10] <![CDATA[1.05*10 4 ]]> <![CDATA[Updated to 1.05*10 4 > 0.1 Successful perturbation 2 <![CDATA[G 2 ↑]]> [0.60,0.02,0.10] <![CDATA[1.08*10 4 ]]> <![CDATA[1.05*10 4 ]]> 0.2 fail 2* <![CDATA[G2 ↓]]> [0.60,0.011,0.10] <![CDATA[1.02*10 4 ]]> <![CDATA[Updated to 1.02*10 4 > 0.11 Successful perturbation 3 <![CDATA[G 3 ↑]]> [0.60,0.011,0.15] <![CDATA[9.8*10 3 ]]> <![CDATA[Updated to 9.8*10 3 > 0.16 Successful perturbation … … … … … … … N - [0.63,0.011,0.15] 0.009 termination

[0220] Figure 5 Figure (a) in the figure is a curve showing the changes in user load, PCS output power and grid output power under a certain operating cycle using the adaptive energy storage reverse current control method of integrated load edge identification and electricity price dispatch provided in this application. Figure 5 Figure (b) shows the curves of user load, PCS output power and grid output power changes under the same operating cycle when the traditional energy storage countercurrent control method is applied.

[0221] Figure 5The blue curve represents user load, the orange curve represents PCS power output, and the red curve represents power supplied by the grid. Comparing Figures (a) and (b), it can be seen that this application guides the energy storage system to actively release power during periods of active load fluctuations, fully mobilizing energy storage resources and improving energy release efficiency. The scheduling strategy possesses edge recognition and dynamic limiting capabilities, enabling it to actively adjust the energy storage response during periods of high load fluctuations, thus playing a role in dynamic grid voltage regulation. Since the PCS output power replaces some of the high-priced grid electricity purchases, the overall operating electricity cost is reduced, thereby improving the system's economic benefits. Compared with traditional solutions, this application exhibits stronger dynamic adjustment capabilities and response stability under complex load conditions, and has wider application adaptability.

[0222] In summary, as Figure 1 As shown, the adaptive energy storage reverse current control method for integrating load edge identification and electricity price dispatch provided in this application includes the following steps:

[0223] S1, based on the current time User load sequences collected or constructed by the power grid Energy storage power station Grid output power Power output executed by PCS Electricity price function And by predicting the sudden change trend of user load, calculate in Timing correction control quantity ;

[0224] S2, Dynamic Adjustment Limiting threshold for constant energy storage reverse flow control And / or adaptively adjust the limiting factor for energy storage reverse flow control the following day. ;

[0225] S3, in Constantly control the amplitude of energy storage reverse current to limit Within, and / or the next day, the control energy storage reverse current amplitude is limited to With the goal, in The time-scheduling control quantity is Energy storage capacity.

[0226] It should be stated that the above-described specific embodiments are merely preferred embodiments and technical principles applied in this application. Those skilled in the art should understand that various modifications, equivalent substitutions, and variations can be made to this application. However, such variations, as long as they do not depart from the spirit of this application, should be within the scope of protection of this application. Furthermore, some terminology used in this application's specification and claims is not limiting but merely for ease of description.

Claims

1. An adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling, characterized in that, include: Based on the current time User load sequences collected or constructed by the power grid Energy storage power station Grid output power Power output executed by PCS and daily electricity price function And by predicting the sudden change trend of user load, calculate in Timing correction control quantity ; Dynamic adjustment Limiting threshold for constant energy storage reverse flow control And / or adaptively adjust the limiting factor for energy storage reverse flow control the following day. ; In Constantly control the amplitude of energy storage reverse current to limit Within, and / or the next day, the control energy storage reverse current amplitude is limited to With the goal of, in The time-scheduling control quantity is Energy storage capacity.

2. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 1, characterized in that, Methods for predicting sudden changes in user load include the following steps: A1, Calculate rising edge identification and falling edge identification ; A2, Parallel Judgment Does it exceed the first threshold? Does it exceed the second threshold? when When the first threshold is exceeded, it is determined that the user load has a sudden upward trend; when When the second threshold is exceeded, it is determined that the user load has a sudden downward trend.

3. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 2, characterized in that, , It is calculated using the following formula (3): In formula (3), This represents the convolution operator; These represent the matched filter functions used to detect rising and falling edges, respectively. 。 4. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 2, characterized in that, When a sudden upward trend in user load is predicted The calculation method is expressed by the following formula (1): （1） In formula (1), Indicates the first dynamic correction amount; This represents the second dynamic correction amount, which is and The difference, Indicates the current time The value of the dead zone buffer.

5. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 4, characterized in that, The calculation method is expressed by the following formula (2). （2） In formula (2), , These represent the prediction errors respectively. , Integral error and The scaling factor of the differential error; , These represent the prediction errors respectively. Integration error and prediction error The differential error; The calculation method is as follows: 。 6. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 2, characterized in that, When a sudden downward trend in user load is predicted The calculation method is expressed by the following formula (5): In formula (5), Indicates that it occurs at the current time. The previous Historical secondary energy storage release; This indicates the distance from the current time. Recent historical energy storage release counts; Indicates the first The lowest point in the historical secondary energy storage release levels; Indicates the current time The value of the dead zone buffer.

7. The adaptive energy storage reverse current control method for integrating load edge identification and electricity price scheduling according to claim 4 or 6, characterized in that, （6）； In formula (6), Indicates the current time The next The value of the dead zone buffer at any given moment; Indicates coefficient; （7） in, This represents the second threshold. This represents the first threshold; numerical values , different.

8. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 1, characterized in that, Adaptive adjustment of limiting factor The method includes the following steps: B1, Construct a comprehensive profit function to calculate the adjusted control amount. Perform power correction to the theoretical benefit at the end of the cycle. and actual benefits ; B2, Calculate the electricity surplus ratio as a revenue indicator. Ratio to revenue forecast ; B3, with and The limiting factor for the day After making adjustments, the amplitude limiting factor for the next day is obtained. .

9. The adaptive energy storage reverse current control method for integrating load edge identification and electricity price scheduling according to claim 8, characterized in that, Indicates the energy storage power station at the current moment The discharge power, Indicates the energy storage power station at the current moment The charging power; The daily cumulative ideal reverse power energy, This indicates the daily cumulative actual reverse power consumption; As a countercurrent penalty factor; This indicates the electricity price in high-price zones; This indicates the electricity price in the low-price zone; This indicates the actual electricity price.

10. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 9, characterized in that, , These represent the operating periods during high electricity price periods and low electricity price periods, respectively. , These represent the changes in SOC during operation in high-electricity-price periods and low-electricity-price periods, respectively.

11. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 8, characterized in that, in This is a flag indicating abnormal program operation. This indicates the threshold value for the time limit at 00:00 on the next day following the current day. This indicates the threshold value for the current day at 0:

00.

12. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 1, characterized in that, The limiting threshold The dynamic adjustment method includes the following steps: C1, calculate the user load sequence mean Standard deviation Simultaneously calculate the current power grid. mean as well as User load fluctuation level ; C2, Dynamically calculates the limit increment The calculation method is expressed as follows: for The learning rate is automatically adjusted. Indicates the fluctuation threshold; Indicates the mean threshold; C3, Calculation With Sum as .

13. The adaptive energy storage reverse current control method integrating load edge identification and electricity price scheduling according to claim 1, characterized in that, exist The time-scheduling control quantity is The method for storing energy power includes the following steps: D1, Construct the offline perturbation optimization function , Indicates the current time Control correction instructions; The first one to be optimized A composite gain factor, This indicates the number of composite gain factors to be optimized; Indicates origin from user load sequence and grid output power Feedback function; D2, Finding the optimal control parameter vector The search method includes the following steps: D21, Define the initial gain vector and the corresponding perturbation step size vector and define the performance objective function. , This represents the grid output power at the k-th time step under parameter vector perturbation in the test case. This represents the grid output power under the initial gain vector; Indicates the execution of system test cases Ideal data for the output power of the secondary power grid; D22, for the first Add perturbation to each composite gain factor ; D23, after adding the perturbation, perform one control cycle and recalculate the objective function value. ; D24, Determine Is it less than , This represents the minimum objective function value calculated by adding perturbations in the past. , If so, then the added perturbation is deemed valid, and it is set to... ; If not, proceed to step D25; D25 determines whether the same composite gain factor has undergone two reverse perturbations. If so, restore the second added reverse perturbation to the state of the first added reverse perturbation, and then proceed to step D26; If not, add a reverse perturbation, making Then return to step D22; D26, Determine whether the search for the optimal control parameter vector has been achieved. The iteration termination condition, If so, then based on the found optimal control parameter vector Calculate control correction instructions ; If not, return to step D22.

14. The adaptive energy storage reverse current control method for integrating load edge identification and electricity price scheduling according to claim 13, characterized in that, Finding the optimal control parameter vector The iteration termination condition is: When the sum of all disturbance amplitudes in the optimization process Less than the set tolerance and / or performance objective function value Less than the error threshold And / or the difference in performance objective function values ​​after two adjacent optimizations. Less than the difference threshold If so, then the iteration termination condition is determined.

Images