A power coordination control method of a wind storage combined power generation system
By constructing a multi-branch scenario tree and using frequency domain decomposition technology, the problems of insufficient energy storage regulation and energy imbalance caused by the volatility and uncertainty of wind power in the wind-storage combined power generation system were solved, thereby achieving system stability and extending equipment life.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XINGNENG POWER CONSTR CO LTD
- Filing Date
- 2026-03-02
- Publication Date
- 2026-05-29
Smart Images
Figure CN121749280B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of wind power generation and energy storage system coordinated control technology, and relates to a power coordination control method for a wind-storage combined power generation system. Background Technology
[0002] Wind-storage combined power generation systems, by integrating wind power generation with energy storage devices, can effectively mitigate fluctuations in wind power output and are an important technological approach to improving grid dispatchability and operational stability. Through the charging and discharging regulation of energy storage devices, the intermittent and random nature of wind power is compensated for, enabling precise tracking of grid load demand and ensuring the safe and stable operation of the grid.
[0003] Currently, the existing power coordination control technology for wind and energy storage combined power generation systems has the following shortcomings: First, wind power has significant volatility and uncertainty in actual operation. Existing technologies usually rely on deterministic wind power forecasts to formulate energy storage charging and discharging plans. Rigid power plans based on a single forecast cannot adapt to the frequent deviations between actual output and forecast values, which can easily lead to insufficient regulation capacity of the energy storage system and may accelerate the aging of energy storage equipment.
[0004] Secondly, existing technologies often employ simple proportional allocation or fixed strategies when allocating total power demand to wind turbines and energy storage. However, wind-storage integrated systems need to cope with power fluctuations at different time scales in the power grid. Simple allocation strategies cannot fully leverage the rapid adjustment advantages of energy storage and may force wind turbines to undertake high-frequency adjustment tasks beyond their mechanical inertia, leading to unit fatigue damage.
[0005] Finally, the continuous accumulation of power deviation during the power coordination and control operation of existing wind-storage combined power generation systems will lead to energy imbalance on a long-term scale, causing the wind-storage combined system to gradually deviate from the optimal operating state and reduce the system's ability to track grid dispatch plans. Summary of the Invention
[0006] In view of this, in order to solve the problems mentioned in the background art, the present invention provides a power coordination control method for a wind-storage combined power generation system.
[0007] The objective of this invention can be achieved through the following technical solution: a power coordination control method for a wind-storage combined power generation system, comprising: S1, acquiring wind speed prediction data of a wind farm, calculating the corresponding wind power prediction power, and organizing a multi-branch scenario tree for uncertain paths based on the wind power prediction power.
[0008] S2. Collect the grid's electricity demand and the real-time energy storage power, and combine it with the wind power prediction in the multi-branch scenario tree to calculate the energy storage charging and discharging power plan required to achieve power balance.
[0009] S3. Select the optimal plan from the energy storage charging and discharging power plan, and extract the recommended operating range of energy storage for each time period from the optimal plan.
[0010] Starting from the current moment, S4 calculates the optimal power command sequence for future scheduling periods within the rolling time domain based on the predicted wind power and grid electricity demand, with reference to the suggested operating interval.
[0011] S5. By frequency domain decomposition, the total power demand in the optimal power command sequence is allocated to the wind turbine and energy storage for execution according to different frequency components.
[0012] S6. Calculate the deviation between the actual total power of the wind-storage system and the predicted wind power, and make corrections to subsequent regulation based on the deviation.
[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention obtains wind speed prediction data of wind farms and calculates probabilistic wind power prediction power, constructs a multi-branch scenario tree with uncertain output, and formulates a storage charging and discharging power plan that considers the probability distribution of multiple scenarios based on this. This overcomes the problem of relying on a single deterministic prediction to formulate a plan, enhances the stability of the plan to the actual fluctuation of wind power, thereby avoiding insufficient energy storage regulation capacity or frequent operation due to prediction deviation, and helps to extend the life of energy storage equipment.
[0014] (2) This invention, after solving the optimal power command sequence through rolling optimization, uses frequency domain decomposition technology to scientifically allocate the total power demand to wind turbines and energy storage according to different frequency components. This overcomes the limitations of simple proportional or fixed allocation strategies, allowing the fast-responding energy storage system to take priority on high-frequency fluctuation components, while the wind turbine with greater inertia mainly tracks low-frequency and base load components. This gives full play to the dynamic response advantages of each device, ensuring the stability of the power grid frequency while reducing mechanical fatigue damage caused by wind turbines undertaking inappropriate adjustment tasks.
[0015] (3) This invention calculates the deviation between the actual total power and the predicted power of the wind-storage system in real time, and dynamically corrects the benchmark value of the total power demand of the subsequent control links based on the deviation. This can promptly correct the energy deviation in the system operation and ensure that the wind-storage joint system tracks the grid dispatch plan in the long term. Attached Figure Description
[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a diagram illustrating the implementation steps of the method of the present invention.
[0018] Figure 2 This is a flowchart of the multi-branch scenario tree construction process for wind power prediction in this invention.
[0019] Figure 3 This is a flowchart illustrating the extraction process for the recommended operating range of the wind-storage combined power generation system of the present invention. Detailed Implementation
[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0021] Please see Figure 1 As shown, the present invention provides a power coordination control method for a wind-storage combined power generation system, including: S100, acquiring wind speed prediction data of a wind farm and calculating the corresponding wind power prediction power.
[0022] The calculation of the corresponding wind power prediction based on wind speed prediction data of wind farms includes: statistical analysis of wind farms by wind speed interval based on historical measured wind speed and power data.
[0023] Specifically, firstly, the wind speed range from the cut-in wind speed to the cut-out wind speed is evenly divided into several intervals, for example, with an interval of 0.5 m / s, forming multiple wind speed boxes. Then, each historical wind speed value is assigned to the corresponding wind speed box according to its value, and the power value corresponding to that moment is recorded in the box.
[0024] By using box-based statistics, a probability distribution of power can be constructed within each wind speed range, thereby supporting subsequent probabilistic wind power prediction based on quantiles.
[0025] Calculate the critical conditional quantiles of power data within each wind speed interval to form a set of power quantiles for each interval.
[0026] Since wind power output remains random at the same wind speed, relying solely on deterministic power curves cannot provide risk perception information for energy storage scheduling. By using multiple key quantiles, a probabilistic basis can be provided for subsequent scenario tree modeling.
[0027] In this invention, the key condition quantile refers to the percentile value that represents the probability distribution characteristic of power within a given wind speed range, based on all historical power data within that range; for example, but not limited to, quantiles such as 10%, 25%, 50%, 75%, and 90%. Each quantile represents the percentage probability that the wind power output is lower than that value under that wind speed condition.
[0028] Obtain the predicted wind speed sequence for the scheduling period and match each value in the predicted wind speed sequence to the corresponding wind speed interval.
[0029] Output the set of power quantiles corresponding to this interval, as the probabilistic wind power prediction power at each time point.
[0030] S101, Multi-branch scenario tree based on wind power prediction power organization of uncertain paths.
[0031] Because wind power forecasting has inherent uncertainties, a single forecast trajectory cannot characterize all possible future power output evolutions. However, a multi-branch scenario tree can intuitively depict the temporal evolution of stochastic processes and is applicable to the temporal uncertainties of wind power.
[0032] See Figure 2 As shown, the multi-branch scenario tree based on wind power prediction and uncertain path organization includes: defining each quantile value in the power quantile set at each scheduling time as the power value of the scenario node at that time, and assigning conditional probabilities to each quantile according to the cumulative probability corresponding to each quantile.
[0033] Each node is assigned a conditional probability, representing the probability of the node's power value occurring at the current moment, given that the state of a node occurred at the previous moment. This probability is calculated based on the probabilistic meaning represented by the quantiles themselves. Generally, if the quantiles are uniformly spaced, the probability interval between adjacent quantiles can be used as the conditional probability of the node's representative value. If the quantiles are not uniformly spaced, precise allocation is required based on the cumulative probability difference. For example, if the quantile node at the current moment represents a cumulative probability interval from 0 to 0.25, then the conditional transition probability of moving from any parent node at the previous moment to this node can be set to 0.25.
[0034] Using each node at the first scheduling time as the root node, connect each node to each node at the next time. The weight of each connection branch is equal to the conditional probability of the connected node at the next time.
[0035] Each path from the root node to the end node constitutes a wind power prediction scenario. The global probability of this scenario is the product of the weights of all branches of its path, generating a multi-branch scenario tree consisting of all scenarios and their probabilities.
[0036] The energy storage charging and discharging decisions at the current moment will change the energy storage power status, thereby affecting its ability to cope with future fluctuations. Therefore, it must be based on the judgment of the wind power evolution trend throughout the entire dispatch cycle, rather than the response to isolated points in time.
[0037] Therefore, one embodiment of the above steps is as follows: If it is only known that the wind power probability at 2 PM is 30%, it may be marked as low. It is impossible to determine whether this is a short-term fluctuation or the beginning of a sustained low power. If it is a short-term fluctuation, a small amount of discharge is sufficient; if it is a sustained low power, electricity should be used sparingly. The generated scene tree can identify a complete low power evolution path, such as wind power being moderate at 1 PM, then low at 2 PM, and then continuously low at 3 PM, and calculate its overall probability. It can assess whether the energy storage capacity is sufficient throughout this entire path.
[0038] S200 collects the grid's electricity demand and the real-time energy storage power, and combines it with the wind power prediction in the multi-branch scenario tree to calculate the energy storage charging and discharging power plan required to achieve power balance.
[0039] Since the operation of the power system must meet the basic physical constraint of real-time power balance, wind power output is intermittent and fluctuates, and its predicted power has multiple possible scenarios. In order to reliably match the combined output of wind power and energy storage with load demand, the energy storage charging and discharging power plan includes: calculating the expected difference between the grid electricity demand and the wind power predicted power for each time period for all wind power prediction scenarios in the multi-branch scenario tree, and using the standard deviation of the difference for each time period as the uncertainty measure of the difference for that time period.
[0040] The power demand of the grid is the total load of the entire grid. The expected difference represents the most likely power balance demand during this period. A positive value indicates that there is an average power shortage, requiring energy storage for discharge; a negative value indicates that there is an average power surplus, suitable for energy storage for charging. The uncertainty measure represents the fluctuation risk of power balance demand during this period. A small standard deviation means that the predicted differences of each scenario are very close, indicating low uncertainty and reliable decision-making; a large standard deviation means that various situations, from severe power shortage to severe surplus, may occur, indicating high decision-making risk.
[0041] If the expected difference is positive and the uncertainty measure is lower than the first set value, the scheduling period is marked as the planned discharge window; if the expected difference is negative and the uncertainty measure is lower than the second set value, the scheduling period is marked as the planned charging window.
[0042] It should be noted that if the expected difference is positive, it means that the load is greater than that of wind power, resulting in a power deficit. In this case, energy storage discharge should be used to make up the deficit. Since allocating this deficit to energy storage only has a predictable effect and value when the uncertainty of the difference is low, it is necessary to ensure that the energy storage can effectively release this part of the energy in the future and avoid ineffective charging and discharging.
[0043] The first and second set values are set based on statistical analysis of historical wind farm operation data. The specific determination process is as follows: First, the historical wind power prediction error sequence is statistically analyzed, and the distribution of its absolute value is calculated. Then, the 75th percentile of this distribution is selected as the set value for uncertainty measurement to ensure that the prediction of the planned charging and discharging window is relatively reliable in about 75% of cases.
[0044] All planned discharge windows are sorted in descending order of the absolute value of the expected difference, and the discharge power is allocated sequentially without exceeding the absolute value of the expected difference, provided that the energy storage capacity is not lower than the lower limit constraint.
[0045] The larger the absolute value of the expected difference, the more severe the power deficit during that period, and the greater the potential threat to the safe operation of the power grid if it is not mitigated. The higher the regulatory value that energy storage discharge can generate, the better. Adopting a descending order of priority and sequential allocation ensures that limited energy storage capacity is used first to address the most severe power imbalance problems, thereby maximizing the overall benefits of energy storage discharge at the global level.
[0046] The lower limit constraint uses the state of charge value corresponding to the maximum permissible depth of discharge specified by the battery manufacturer as an insurmountable constraint.
[0047] For all planned charging windows, sort them in descending order by the absolute value of the expected difference, and allocate charging power in sequence that does not exceed the absolute value of the expected difference, provided that the energy storage capacity does not exceed the upper limit constraint.
[0048] The larger the absolute value of the expected difference across all planned charging windows, the more severe the predicted power surplus during that period, the greater the potential wind curtailment, and the higher the renewable energy utilization value and grid support value created by energy storage charging. Adopting a descending order of priority and sequential allocation ensures that limited energy storage charging capacity is prioritized for absorbing the most severe power surplus, thereby maximizing wind power absorption and system operational safety at the global level.
[0049] The upper limit constraint is directly taken as the maximum safe state of charge corresponding to the rated capacity specified by the energy storage system manufacturer, in order to prevent overcharging.
[0050] S300: Select the optimal plan from the energy storage charging and discharging power plans.
[0051] Since there is always a most probable scenario path in a multi-branch scenario tree, the optimal plan should be the one that best matches this most probable wind power evolution path. Therefore, the candidate plan that achieves the most perfect power balance under this most probable scenario should be selected.
[0052] Based on this, the selection of the optimal plan from the energy storage charging and discharging power plan includes: selecting the wind power prediction scenario with the highest global probability from the multi-branch scenario tree as the most likely scenario.
[0053] For the most likely scenario, the difference between the grid electricity demand and the wind power value of that scenario is calculated for each time period to obtain the power demand sequence of the energy storage system for each time period.
[0054] Calculate the absolute value of the difference between the energy storage power in each time period of each energy storage charging and discharging power plan and the power demand in the corresponding time period of the power demand sequence, and sum the absolute values of all time periods to obtain the matching deviation value between the plan and the most likely scenario.
[0055] Compare the matching deviation values corresponding to all candidate plans, and select the candidate plan with the smallest matching deviation value as the optimal energy storage charging and discharging power plan.
[0056] S301. Extract the recommended operating range for energy storage for each time period from the optimal plan.
[0057] Considering that the operational safety and lifespan of energy storage devices are subject to rigid constraints on their power and energy capacity, if only a single power point is executed without a buffer zone, it is easy to cause power exceeding the limit or energy exceeding the limit under random disturbances. Therefore, it is necessary to set a power operation range centered on the optimal planned value for each time period to provide the necessary safety margin for real-time adjustment.
[0058] See Figure 3 As shown, the extraction of the recommended operating range for energy storage for each time period from the optimal plan includes: calculating the maximum absolute value of the power for all time periods in the optimal energy storage charging and discharging power plan, and using this maximum value as the power benchmark value.
[0059] Taking the maximum absolute power value across all time periods as the power benchmark value effectively identifies the maximum instantaneous power support capability required by the energy storage system throughout the entire scheduling cycle, ensuring the system's ability to cover the most stringent regulation demands and establishing a unified regulation benchmark for all time periods.
[0060] Multiply the power reference value by a preset proportional coefficient to obtain the power regulation margin shared by all time periods.
[0061] The preset proportional coefficient is determined based on the distribution of the absolute value of the deviation between the day-ahead predicted power and the ultra-short-term predicted power under similar wind speed conditions. For example, the 90th percentile of this absolute value distribution of deviation can be selected as the preset proportional coefficient to ensure that the adjustment margin can cover most error scenarios.
[0062] For each scheduling period in the optimal plan, with its planned power value as the center, an adjustment margin is extended in the direction of power decrease and power increase respectively to obtain the lower limit and upper limit of power for that period, forming the recommended operating range of energy storage for that period.
[0063] The proposed operating range for energy storage serves as a bridge connecting long-term probabilistic optimization and short-term rolling optimization, ensuring that the rolling optimization instructions follow the latest forecast information while remaining within the safety and economic boundaries assessed across multiple scenarios.
[0064] S400. Starting from the current moment, within the rolling time domain, based on the predicted wind power and grid electricity demand, and with the suggested operating interval as a reference, solve for the optimal power command sequence for future scheduling periods.
[0065] In this invention, the energy storage discharge power is defined as a positive value, and the charging power is defined as a negative value.
[0066] The optimal power command sequence solution process is as follows: at each rolling optimization start time, the difference between the grid electricity demand and the wind power prediction value is calculated for each time period, which is used as the initial power command of the energy storage system for that time period.
[0067] If the initial power command falls within the recommended operating range for energy storage, it will be retained; if it is below the lower limit, the initial power command will be raised to the lower limit; if it is above the upper limit, it will be lowered to the upper limit.
[0068] It should be noted that if the initial power command falls within the recommended operating range of energy storage during rolling optimization, it indicates that the real-time demand based on the latest forecast is highly coordinated with the long-term optimization plan, and therefore it is retained.
[0069] If the initial instruction is below the lower limit or above the upper limit of the interval, a limiting process is applied. The principle behind this is that the recommended operating interval is generated through prior random optimization across multiple scenarios, inherently considering energy storage capacity safety constraints and overall economic efficiency. The limiting operation ensures that the rolling optimization instructions do not exceed this optimized and validated safety boundary, thus inheriting the stability of the long-term plan.
[0070] Based on the energy change corresponding to the feasible instructions in the real-time energy storage power simulation range, if the predicted energy exceeds the upper and lower limits of the energy storage power safety, the first time period exceeding the limit is located, and the charging power of the previous time period is gradually reduced according to the preset step size until the energy of all time periods is within the safe range.
[0071] It should be added that the preset step size is set as a small percentage of the rated power of the energy storage system, such as 1% to 5%. For example, for an energy storage system with a rated power of 10MW, the step size can be set to 0.2MW, or 2%.
[0072] The safety range refers to the absolute upper and lower limits of the state of charge, which are determined by the electrochemical characteristics of the battery itself and are usually specified by the manufacturer. For example, for lithium-ion batteries, the safety range is usually [10%, 90%] of the rated capacity. Operating outside this range will seriously endanger the battery's safety and lifespan, and it is strictly forbidden to violate this range under any circumstances.
[0073] The corrected instruction sequence is output as the optimal power instruction sequence, and the first time-period instruction is executed. Then, the process is rolled over to the next time-period and the above steps are repeated.
[0074] S500 allocates the total power demand in the optimal power command sequence to wind turbines and energy storage for execution according to different frequency components through frequency domain decomposition.
[0075] Due to mechanical inertia and pitch system delays, wind turbines have a relatively slow active power regulation capability, making them more suitable for tracking power components with gradual changes. In contrast, energy storage systems, controlled by power electronic converters, have rapid charging and discharging capabilities and are adept at handling high-frequency fluctuations.
[0076] Therefore, by decomposing the optimal power command sequence obtained through rolling optimization into components of different frequencies in the frequency domain and matching and allocating them, the technical advantages of wind turbines and energy storage can be fully utilized to achieve synergy and complementarity. The specific steps for allocating the wind turbines and energy storage according to different frequency components are as follows: For the optimal power command sequence, a moving average window of fixed time length is used to calculate its moving average value point by point to form the target power command sequence of the wind turbine.
[0077] The optimal power command sequence is subtracted point by point from the target power command sequence of the wind turbine to obtain the power regulation command sequence of the energy storage.
[0078] The target power command sequence of the wind turbine is sent to the wind farm power control system, and the power regulation command sequence of the energy storage is sent to the energy storage converter, so that the wind turbine and the energy storage system can work together to execute the total power demand.
[0079] The total power command sequence is calculated by performing a moving average with a fixed window length. This is done by applying a low-pass filter with a cutoff frequency. The output sequence retains the low-frequency trend components in the original command while filtering out the high-frequency fluctuation components.
[0080] The smoothed low-frequency sequence is sent to the wind farm power control system, which has a slower response, in line with its ability to track slow power changes through torque control. The difference between the total command and the wind turbine command is sent to the energy storage converter, which has a very fast response, matching its ability to quickly compensate for power fluctuations through power electronic switching.
[0081] The length of the moving average window is adjusted based on the response time of the wind turbine power control closed loop to ensure that the extracted wind turbine commands are physically feasible.
[0082] S600, calculate the deviation between the actual total power of the wind-storage system and the predicted wind power.
[0083] The ultra-short-term wind power forecast here refers to the direct use of the latest measured data in S100 and S101 for the probabilistic wind power forecast corresponding to the current and future periods during the rolling execution phase, ensuring the timeliness and consistency of the deviation calculation benchmark.
[0084] Given the intermittency and volatility of wind power, it is difficult for a single feedforward open-loop control to maintain optimal performance over a long period of time. Any prediction-based plan will inevitably have an error between it and the actual operation. If this error is not corrected, it will continue to accumulate in subsequent scheduling and may lead to serious consequences. Therefore, it is necessary to establish real-time monitoring and feedback of this error.
[0085] Therefore, the deviation between the actual total power of the wind-storage system and the predicted wind power includes: real-time acquisition of the actual output power of the wind turbine generator and the actual charging and discharging power of the energy storage system.
[0086] The actual total output power of the wind turbine generator set is obtained by algebraically summing the actual charging and discharging power of the energy storage system.
[0087] Obtain the wind power ultra-short-term forecast power value for the same scheduling period, and calculate the instantaneous power deviation between the actual total output power and the wind power ultra-short-term forecast power value.
[0088] The specific steps for the ultra-short-term wind power prediction value are as follows: based on the target time period identifier, locate the power quantile set corresponding to that time period; select a specified quantile value from the set, such as the 50th quantile, i.e. the median, as the ultra-short-term wind power prediction value for output, ensuring that the predicted value used for deviation calculation is strictly synchronized with the actual power value on the time scale.
[0089] The instantaneous power deviation value within multiple consecutive scheduling periods is multiplied by the corresponding scheduling period length to obtain the energy deviation value for each period. The energy deviation values of all periods are summed to obtain the cumulative energy deviation value for that period.
[0090] S601. Correct subsequent adjustments based on the deviation.
[0091] The step of correcting subsequent regulation based on deviation includes: if the cumulative energy deviation value is positive, then a downward adjustment of the total output power is generated; otherwise, an upward adjustment of the total output power is generated.
[0092] The cumulative energy deviation is the algebraic sum of the actual total output energy of the wind-storage combined system and the reference wind power prediction energy over multiple consecutive scheduling periods. The reference wind power prediction energy is the cumulative energy of the 50th percentile value sequence in the same time period, such as the power quantile set of wind power prediction power sequence, which represents the degree of deviation between the actual charging and discharging energy of the energy storage system and the planned trajectory.
[0093] If the cumulative energy deviation is greater than 0, it indicates that the actual discharge energy of the energy storage system is greater than the planned discharge energy, meaning that the stored energy is reduced beyond the plan; conversely, it indicates that the actual discharge energy of the energy storage system is less than the planned discharge energy, meaning that the stored energy is increased beyond the plan.
[0094] To maintain the long-term balance of energy storage power during the dispatch cycle and ensure its sustainable regulation capability, the internal power benchmark for subsequent dispatch cycles is adjusted according to the sign of the cumulative energy deviation value: if the cumulative energy deviation is greater than 0, an instruction is generated to reduce the target benchmark value of the total output power that the wind-storage combined system needs to track in the subsequent dispatch cycle; otherwise, an instruction is generated to increase the target benchmark value of the total output power that the wind-storage combined system needs to track in the subsequent dispatch cycle.
[0095] The total output power target benchmark value refers to the joint output target reference value of the wind-storage combined system for internal rolling optimization and matching with the power grid demand in subsequent dispatch cycles. Its adjustment does not affect the external load demand instructions issued by the power grid dispatching department.
[0096] The adjusted total output power target benchmark value is fed back to the system total power demand benchmark value setting stage in subsequent scheduling cycles, serving as the directional basis for unidirectional dynamic correction of the benchmark value, gradually offsetting the accumulated energy deviation, and bringing the energy storage capacity back to the long-term planned trajectory.
[0097] The downward and upward adjustment instructions are fed back to the system total power demand baseline setting stage in subsequent scheduling cycles, serving as the direction for correction.
[0098] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.
[0099] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0100] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0102] Finally, 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, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A power coordination control method for a wind-storage combined power generation system, characterized in that: include: Obtain wind speed prediction data from wind farms, calculate the corresponding wind power prediction power, and organize a multi-branch scenario tree for uncertain paths based on the wind power prediction power. Collect the grid's electricity demand and the real-time energy storage power, and combine the wind power prediction in the multi-branch scenario tree to calculate the energy storage charging and discharging power plan required to achieve power balance; Select the optimal plan from the energy storage charging and discharging power plan, and extract the recommended operating range of energy storage for each time period from the optimal plan; Starting from the current moment, within the rolling time domain, based on the predicted wind power and grid electricity demand, and with the suggested operating interval as a reference, the optimal power command sequence for future dispatch periods is solved. The total power demand in the optimal power command sequence is allocated to wind turbines and energy storage according to different frequency components by frequency domain decomposition. Calculate the deviation between the actual total power of the wind-storage system and the predicted wind power, and make corrections to subsequent regulation based on the deviation; The energy storage charging and discharging power plan includes: For all wind power prediction scenarios in the multi-branch scenario tree, calculate the expected difference between grid electricity demand and wind power prediction power for each time period, and use the standard deviation of the difference for each time period as the uncertainty measure of the difference for that time period. If the expected difference is positive and the uncertainty measure is lower than the first set value, the scheduling period is marked as the planned discharge window; if the expected difference is negative and the uncertainty measure is lower than the second set value, the scheduling period is marked as the planned charging window. All planned discharge windows are sorted in descending order of the absolute value of the expected difference, and the discharge power is allocated sequentially without exceeding the absolute value of the expected difference, provided that the energy storage capacity is not lower than the lower limit constraint. For all planned charging windows, sort them in descending order by the absolute value of the expected difference, and allocate charging power in sequence that does not exceed the absolute value of the expected difference, provided that the energy storage capacity does not exceed the upper limit constraint.
2. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The step of obtaining wind speed prediction data from wind farms and calculating the corresponding predicted wind power includes: Based on historical measured wind speed and power data of wind farms, statistics are compiled by wind speed intervals. Calculate the key conditional quantiles of power data within each wind speed interval to form a set of power quantiles for each interval; Obtain the predicted wind speed sequence for the scheduling period and match each value in the predicted wind speed sequence to the corresponding wind speed interval; Output the set of power quantiles corresponding to this interval, as the probabilistic wind power prediction power at each time point.
3. The power coordination control method for a wind-storage combined power generation system according to claim 2, characterized in that: The multi-branch scenario tree based on wind power prediction power organization of uncertain paths includes: Each quantile value in the power quantile set at each scheduling time is defined as the power value of the scene node at that time, and a conditional probability is assigned to it based on the cumulative probability corresponding to each quantile. Using each node at the first scheduling time as the root node, connect each node to each node at the next time. The weight of each connection branch is equal to the conditional probability of the connected node at the next time. Each path from the root node to the end node constitutes a wind power prediction scenario. The global probability of this scenario is the product of the weights of all branches of its path, generating a multi-branch scenario tree consisting of all scenarios and their probabilities.
4. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The selection of the optimal plan from the energy storage charging and discharging power plan includes: The wind power prediction scenario with the highest global probability is selected from the multi-branch scenario tree as the most likely scenario; For the most likely scenario, the difference between the grid electricity demand and the wind power value of that scenario is calculated for each time period to obtain the power demand sequence of the energy storage system for each time period; Calculate the absolute value of the difference between the energy storage power in each time period of each energy storage charging and discharging power plan and the power demand in the corresponding time period of the power demand sequence, and sum the absolute values of all time periods to obtain the matching deviation value between the plan and the most likely scenario. Compare the matching deviation values corresponding to all candidate plans, and select the candidate plan with the smallest matching deviation value as the optimal energy storage charging and discharging power plan.
5. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The recommended operating ranges for energy storage extracted from the optimal plan for each time period include: Calculate the maximum absolute value of the power in all time periods of the optimal energy storage charging and discharging power plan, and use this maximum value as the power benchmark value; Multiply the power reference value by a preset proportional coefficient to obtain the power regulation margin shared by all time periods. For each scheduling period in the optimal plan, with its planned power value as the center, an adjustment margin is extended in the direction of power decrease and power increase respectively to obtain the lower limit and upper limit of power for that period, forming the recommended operating range of energy storage for that period.
6. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The process of solving the optimal power command sequence includes: At each rolling optimization start time, the difference between the grid electricity demand and the wind power forecast is calculated for each time period, and used as the initial power command of the energy storage system for that time period; If the initial power command falls within the recommended operating range for energy storage, it will be retained; if it is below the lower limit, the initial power command will be raised to the lower limit; if it is above the upper limit, it will be lowered to the upper limit. Based on the energy change corresponding to the feasible instructions in the real-time energy storage power simulation range, if the predicted energy exceeds the upper and lower limits of the energy storage power safety, the first time period exceeding the limit is located, and the charging power of the previous time period is gradually reduced according to the preset step size until the energy of all time periods is within the safe range. The corrected instruction sequence is output as the optimal power instruction sequence, and the first time-period instruction is executed. Then, the process is rolled over to the next time-period and the above steps are repeated.
7. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The allocation of different frequency components to the wind turbine and energy storage includes: For the optimal power command sequence, a moving average window of fixed time length is used to calculate its moving average point by point to form the target power command sequence of the wind turbine; The optimal power command sequence is subtracted point by point from the target power command sequence of the wind turbine to obtain the power regulation command sequence of the energy storage. The target power command sequence of the wind turbine is sent to the wind farm power control system, and the power regulation command sequence of the energy storage is sent to the energy storage converter, so that the wind turbine and the energy storage system can work together to execute the total power demand.
8. The power coordination control method for a wind-storage combined power generation system according to claim 1, characterized in that: The deviation between the actual total power of the calculated wind-storage system and the predicted wind power includes: Real-time acquisition of the actual output power of wind turbine generators and the actual charging and discharging power of energy storage systems; The actual total output power of the wind turbine generator set is obtained by algebraically summing the actual charging and discharging power of the energy storage system. Obtain the wind power ultra-short-term forecast power value for the same scheduling period, and calculate the instantaneous power deviation between the actual total output power and the wind power ultra-short-term forecast power value; The instantaneous power deviation value within multiple consecutive scheduling periods is multiplied by the corresponding scheduling period length to obtain the energy deviation value for each period. The energy deviation values of all periods are summed to obtain the cumulative energy deviation value for that period.
9. The power coordination control method for a wind-storage combined power generation system according to claim 8, characterized in that: The correction of subsequent regulation based on the deviation includes: If the cumulative energy deviation value is positive, the total output power will be reduced; otherwise, the total output power will be increased. The downward and upward adjustment instructions are fed back to the system total power demand baseline setting stage in subsequent scheduling cycles, serving as the directional basis for unidirectional correction.