Frequency modulation and peak regulation cooperative control method for wind storage combined power station

Abstract

The invention discloses a frequency modulation and peak regulation cooperative control method for a wind storage combined power station, and belongs to the technical field of new energy power station operation control. The method aims to solve the problem that control targets are not matched due to the fact that frequency modulation and peak regulation demand characteristics cannot be effectively distinguished when an existing wind storage combined power station responds to a power grid dispatching instruction. According to the method, a received power station level total active power instruction is decomposed into a high-frequency component which changes rapidly and a low-frequency component which changes gently, and the power adjusting ranges of the high-frequency component and the low-frequency component are determined based on wind power plant output prediction data and the real-time charge state of an energy storage system. The high-frequency component is mainly distributed to the energy storage system to serve as a core part of a charging and discharging instruction, meanwhile, the low-frequency component is combined with a wind power plant basic power generation plan to form a target power generation instruction, and finally the instruction is issued to a corresponding subsystem to be executed. The method is used for improving the cooperative response capability of the wind storage combined power station to the power grid frequency modulation and peak regulation composite demand, and resource optimization configuration and system stable operation are achieved.

Description

Technical Field

[0001] This invention belongs to the field of new energy power plant operation and control technology, specifically relating to a frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant. Background Technology

[0002] The randomness and volatility of wind energy resources pose a continuous challenge to the stable operation of the power grid. As wind power penetration in the power system continues to increase, the grid demands that wind farms participate in ancillary services such as frequency regulation and peak shaving. Wind-storage combined power plants, by configuring energy storage systems, aim to smooth out wind power output fluctuations and improve controllability; however, their collaborative control strategies still face many technical challenges in practical applications.

[0003] Existing wind-storage joint control methods typically compare the total command with the predicted wind power output when responding to grid dispatch commands, with the difference borne by the energy storage system. This approach fails to effectively differentiate the control requirements with different time scales and physical characteristics inherent in the dispatch commands. Specifically, grid commands simultaneously include a fast frequency regulation component to address short-term load fluctuations or frequency deviations, and a slow peak-shaving component to track the daily load curve. Treating these two components together leads to a mismatch in response characteristics: the energy storage system may over-respond to the slow-changing component, accelerating its energy state deviation, while wind turbines, due to their greater inertia, struggle to track the fast-changing component. This control method cannot simultaneously meet the grid's dual requirements for rapid frequency regulation and long-term power balance.

[0004] Maintaining the state of charge (SOC) is a critical challenge in the operation and management of energy storage systems. During frequent charging and discharging, the SOC of energy storage units tends to gradually deviate from its normal operating range, approaching the limits of charging or discharging. When the SOC of an energy storage system deviates significantly due to continuously undertaking power tasks in one direction, its subsequent regulation capability will decrease significantly, or even be completely lost. Although additional control strategies can correct the SOC, simple correction methods are prone to conflict with the primary power regulation task, making it difficult to balance maintaining energy balance and providing power services.

[0005] Existing methods for coordinating wind and energy storage power allocation often rely on initial allocation strategies and lack adjustment mechanisms to address uncertainties in actual operation. Wind power output forecasting inevitably contains errors, and equipment operating conditions may change in real time. These factors lead to deviations between the initial power allocation scheme and actual demand. Failure to eliminate these deviations in a timely manner will affect the overall power plant's accuracy in tracking dispatch commands. However, designing an effective deviation compensation mechanism faces challenges: the dynamic response characteristics of wind turbines and energy storage systems differ significantly. An inappropriate compensation strategy may cause operational conflicts between the two subsystems, or even exacerbate power oscillations.

[0006] In terms of system operating condition identification and switching, wind-storage combined power plants need to cope with various operating scenarios, including normal frequency regulation and peak shaving, and emergency support during grid faults. Different operating scenarios require different control strategies, and smooth transitions between scenarios are a prerequisite for stable operation. Existing methods often rely on simple threshold judgments for operating condition identification, which are prone to misjudgment due to instantaneous fluctuations in grid signals, leading to frequent switching of control modes. At the same time, improper handling of transitions between different control modes can generate power surges, causing secondary disturbances to the grid.

[0007] Furthermore, in terms of improving control performance, relying solely on feedback control is insufficient to meet the requirements for rapid response, while introducing feedforward control presents new challenges. In particular, feedforward compensation based on wind power forecasting is highly dependent on forecast accuracy. Due to the inherent uncertainty in wind power forecasting, distorted feedforward signals may actually amplify power deviations. How to assess forecast reliability and dynamically adjust the feedforward compensation strength accordingly becomes a key challenge in improving system performance.

[0008] These problems limit the overall performance of wind-storage combined power plants, and there is an urgent need to study more effective collaborative control methods. Summary of the Invention

[0009] One object of the present invention is to solve at least the above-mentioned problems and / or defects, and to provide at least the advantages described below.

[0010] One objective of this invention is to solve the problem of mismatched control objectives when wind-storage combined power plants respond to grid dispatch commands.

[0011] One objective of this invention is to solve the problem of difficulty in maintaining the state of charge of energy storage systems during continuous service provision.

[0012] One objective of this invention is to solve the power tracking deviation problem in wind-storage combined power plants caused by prediction errors and equipment limitations.

[0013] One objective of this invention is to solve the problem of unstable identification and switching of wind-storage combined power plants under multiple operating conditions.

[0014] One objective of this invention is to solve the problems of insufficient instruction decomposition accuracy and response delay affecting control performance.

[0015] One objective of this invention is to solve the problem of invalid device operation caused by minute correction signals during SOC management.

[0016] One objective of this invention is to solve the oscillation problem caused by the mismatch between wind and energy storage response speeds during power compensation.

[0017] One objective of this invention is to address the problems of insufficient reliability in identifying transient operating conditions and the limited range of support strategies.

[0018] One objective of this invention is to solve the problem of amplified control deviation due to prediction error distortion in feedforward compensation.

[0019] One objective of this invention is to address the problem that feedforward reliability assessment models are complex and difficult to implement.

[0020] One object of the present invention is to provide a frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant, comprising the following steps: Receive the total active power command at the power station level issued by the power grid dispatch center; Acquire forecast data of wind farm output and real-time state of charge of energy storage systems; The total active power command is decomposed into rapidly changing high-frequency components and slowly changing low-frequency components. Based on the predicted wind power output data and the real-time state of charge of the energy storage, the upper and lower limits of the power regulation capability of the wind farm at the current moment, and the charging and discharging power limit of the energy storage system at the current moment are determined respectively. Perform coordinated allocation of power commands: Within the charging and discharging power limit range of the energy storage system, the high-frequency component is allocated to the energy storage system as the core part of its charging and discharging power command; The low-frequency component is combined with the basic power generation plan of the wind farm to form the target power generation command of the wind farm, and the command is ensured to be within its power regulation capability range. The target power generation command is sent to the wind farm, and the charging and discharging power command is sent to the energy storage system.

[0021] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station, after the step of allocating the high-frequency component to the energy storage system, the method further includes a step of dynamically managing the state of charge of the energy storage system: Real-time monitoring of the state of charge of the energy storage system; Set the target median value and allowable fluctuation range of the state of charge; When the real-time state of charge deviates from the target median, a power correction signal is generated. The direction of the power correction signal is configured to pull the real-time state of charge back to the target median, and the rate of change of the power correction signal is configured to be lower than the rate of change of the high-frequency component. The power correction signal is superimposed on the charge and discharge power command of the energy storage system, and the total command after superposition is ensured not to exceed its instantaneous charge and discharge power limit.

[0022] Preferably, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station further includes a power coordination and matching step before forming the final target power command for the wind farm and energy storage system: The deviation between the actual total active power at the power station's grid connection point and the power station-level total active power command is calculated in real time. Based on this power deviation, a power compensation signal is dynamically generated. The power compensation signal is dynamically allocated to the wind power system and / or the energy storage subsystem according to a preset rule based on the current state of charge and the charging / discharging power limit of the energy storage system, in order to eliminate the power deviation.

[0023] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station, before the step of receiving the total active power command at the power station level, an operating condition sensing and switching step is also included: Continuously monitor key electrical quantities characterizing grid fault status, the real-time state of charge of the energy storage system, and the adjustable power range of the wind farm; Based on preset criteria, the current operating condition of the combined power plant is determined to be one of the following: normal frequency regulation and peak shaving mode, transient emergency support mode, or energy recovery mode. When it is determined that a mode switch is required, a power command ramp function is used to smoothly transition the system's control output from the output value corresponding to the previous mode to the target output value corresponding to the current mode within a preset transition time.

[0024] Preferably, in the frequency regulation and peak shaving coordinated control method of the wind-storage combined power station, the decomposition of the total active power command into high-frequency and low-frequency components is specifically achieved through a first-order inertial filter or a Butterworth filter; the high-frequency component corresponds to the fast fluctuation part in the passband of the filter, and the low-frequency component corresponds to the slow change part in the stopband of the filter; and when issuing the target charge and discharge power command to the energy storage subsystem, the high-frequency component is superimposed with a feedforward compensation signal, which is generated based on the rate of change of the ultra-short-term prediction data of wind power output.

[0025] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power plant, the step of generating the power correction signal further includes: A static dead zone or a dynamic adaptive dead zone is set around the target median value of the state of charge. When the real-time state of charge is within the dead zone, the power correction signal is forcibly set to zero; when the real-time state of charge exceeds the dead zone, a non-zero power correction signal is generated based on the degree of deviation.

[0026] Preferably, in the frequency regulation and peak shaving coordinated control method of the wind-storage combined power station, the step of dynamically allocating power to the wind power system and / or the energy storage subsystem according to preset rules specifically means: prioritizing and fully allocating the power compensation signal to the energy storage subsystem for rapid response; Simultaneously, based on the magnitude and direction of the power compensation signal, a corresponding wind turbine compensation command with delay characteristics is generated and sent to the wind electronic system; When the actual output of the wind power system begins to respond to the wind turbine compensation command, the power compensation signal undertaken by the energy storage subsystem is reduced at the same rate, and finally the wind power system takes over all steady-state compensation tasks.

[0027] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power plant, the step of determining the current operating condition of the combined power plant as a transient emergency support mode based on a preset criterion specifically includes: The system continuously monitors the grid voltage and frequency. Only when the value exceeds the preset threshold and the duration reaches a set delay confirmation time will it finally confirm the entry into the transient emergency support mode. Furthermore, based on the combination of voltage drop depth and frequency deviation magnitude, the transient emergency support mode is further refined into reactive power priority support mode or active power priority recovery mode, and the corresponding preset support strategy is invoked.

[0028] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power plant, the superposition of the high-frequency component with a feedforward compensation signal specifically includes: Real-time calculation of the short-term prediction error between ultra-short-term wind power output forecast data and actual wind farm output; A feedforward reliability coefficient is dynamically calculated based on the magnitude of the short-term prediction error; wherein, the reliability coefficient decreases as the prediction error increases. The feedforward compensation signal is multiplied by the confidence coefficient to obtain a weighted feedforward compensation amount, which is then superimposed on the high-frequency component.

[0029] Preferably, in the frequency regulation and peak shaving coordinated control method for the wind-storage combined power plant, the step of dynamically calculating a feedforward reliability coefficient based on the magnitude of the short-term prediction error is specifically achieved through the following steps: S1: Set a baseline error threshold; S2: Calculate the average value of the absolute value of the short-term prediction error within a preset time window; S3: Compare the baseline error threshold with the average value, and calculate the feedforward confidence coefficient using a predefined mapping function; wherein the mapping function is configured such that: when the average value is less than or equal to the baseline error threshold, the output value is 1; when the average value is greater than the baseline error threshold, the output value monotonically decreases between 0 and 1 as the average value increases.

[0030] The present invention has at least the following beneficial effects: This invention achieves optimized matching of resource characteristics by decomposing grid dispatch commands according to frequency characteristics and allocating them to the most suitable response units. This method allows fast-response energy storage systems to focus on handling high-frequency fluctuation components, while wind turbines and energy storage work together to handle steady-state and slowly changing components, thereby improving the overall response quality of the combined power plant to complex dispatch commands. This characteristic-based allocation strategy satisfies the grid's control needs at different time scales while avoiding the operational pressure caused by a single device handling power across the entire frequency band, enhancing the system's adaptability and control rationality.

[0031] This invention achieves long-term stability of the state of charge (SOC) of an energy storage system by introducing a slow power correction signal for dynamic management. While ensuring the energy storage system can complete its main power regulation tasks, this method effectively prevents the SOC from continuously deviating from its normal operating range through directional correction, ensuring the continuous availability of the energy storage device. This management approach maintains the energy balance of the energy storage system without affecting immediate power service, thus improving the long-term operational reliability of the combined power plant.

[0032] This invention improves the tracking accuracy of dispatch commands by combined power plants through real-time power deviation detection and compensation. This method can promptly correct power deviations caused by prediction errors and equipment limitations, forming closed-loop control and enhancing system robustness. By dynamically allocating compensation based on system status, the effectiveness and stability of the correction process are ensured, maintaining a high degree of consistency between the power plant's output power and dispatch requirements.

[0033] This invention improves the adaptability of combined power plants under complex operating conditions by establishing a multi-condition identification and smooth switching mechanism. This method utilizes multi-parameter monitoring and delayed confirmation to ensure the reliability of condition judgment, and achieves smooth transitions between modes through a ramp transition function, avoiding power surges. This mechanism enables the system to safely transition between different modes such as normal frequency regulation and peak shaving, and transient emergency support, ensuring operational continuity.

[0034] This invention improves the dynamic response characteristics of a system by combining precise frequency domain decomposition and feedforward compensation. The method utilizes a specific filter to achieve accurate command separation and generates control commands in advance using a feedforward signal based on the predicted rate of change, effectively reducing response delay. This design enhances the ability to track rapid power changes, making the control process more timely and precise.

[0035] This invention optimizes the operating efficiency of energy storage devices by establishing a dead-zone mechanism for State of Charge (SOC) management. This method shields unnecessary correction signals when the SOC is within a reasonable range, avoiding frequent device actions caused by the superposition of SOC and main power commands. This mechanism maintains long-term SOC stability and reduces ineffective device operations, thus helping to extend the device's lifespan.

[0036] This invention ensures the coordination of the power compensation process by establishing a master-slave compensation and timing handover mechanism. This method utilizes the rapid response characteristics of energy storage to initiate the process, followed by gradual takeover by the wind turbine, forming an orderly compensation flow. This arrangement avoids system oscillations caused by differences in response speed, making the compensation process smooth and stable.

[0037] This invention improves the effectiveness of support during power grid faults by refining transient condition identification and support strategies. The method combines multi-dimensional judgments of voltage and frequency to achieve accurate identification and differentiated responses to different fault types. This precise support strategy can better adapt to the actual needs of the power grid and provide more targeted auxiliary services.

[0038] This invention enhances the system's adaptability to prediction uncertainty by introducing a feedforward reliability assessment mechanism. This method dynamically adjusts the feedforward compensation strength based on the prediction error, fully utilizing the feedforward advantage when the prediction is reliable and automatically reducing its impact when the prediction is unreliable. This adaptive mechanism improves the robustness of the control system and avoids performance degradation caused by prediction distortion.

[0039] This invention achieves reasonable control of feedforward compensation intensity through a concise reliability assessment method. Based on the comparison between the average prediction error and a threshold, this method establishes a clear reliability mapping mechanism, reducing implementation complexity while ensuring assessment effectiveness. This design makes adaptive adjustment of feedforward compensation simple and feasible, facilitating engineering applications.

[0040] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Detailed Implementation

[0041] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0042] This invention provides a frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant, comprising the following steps: receiving a total active power command at the power plant level issued by the power grid dispatch center; acquiring predicted power output data of the wind farm and the real-time state of charge of the energy storage system; decomposing the total active power command into rapidly changing high-frequency components and slowly changing low-frequency components; based on the predicted power output data of the wind farm and the real-time state of charge of the energy storage system, determining the upper and lower limits of the power regulation capability of the wind farm at the current moment, and the charging and discharging power limit of the energy storage system at the current moment; performing coordinated allocation of power commands: within the charging and discharging power limit range of the energy storage system, allocating the high-frequency components to the energy storage system as the core part of its charging and discharging power command; combining the low-frequency components with the basic power generation plan of the wind farm to form a target power generation command for the wind farm, and ensuring that the command is within its power regulation capability range; issuing the target power generation command to the wind farm and issuing the charging and discharging power command to the energy storage system.

[0043] This method first receives the total active power command at the power plant level issued by the power grid dispatch center in real time through the data acquisition system, and at the same time obtains the ultra-short-term power prediction data of the wind farm and the real-time charge status of the energy storage system.

[0044] In the command processing stage, a first-order inertial filter is used to perform frequency domain decomposition on the total active power command. The time constant of the filter is set according to the frequency regulation requirements of the power grid, and parameters that can separate fluctuation components at the minute level or below are typically selected. Through this processing, the total command is separated into high-frequency components reflecting rapid power changes and low-frequency components reflecting trend-based power changes.

[0045] In the power allocation phase, the system calculates the feasible power range of the wind farm under current wind speed conditions based on wind power prediction data, and simultaneously determines its safe charging and discharging power limits based on the real-time state of charge of the energy storage system. The decomposed high-frequency components are primarily allocated to the energy storage system, forming the core of its charging and discharging power command. The low-frequency components are combined with the wind farm's basic power generation plan and, after power limiting processing, form the wind farm's target power generation command. Within the energy storage system's charging and discharging power limits, the high-frequency components are preferentially and fully allocated to it. Only when the energy storage system reaches its power limit is the excess high-frequency component allocated to the wind farm. In actual calculations, the high-frequency component command P can be... high With the current maximum discharge power P of the energy storage ess_max and maximum charging power Pe ss_min After comparison and processing by the limiting function, it is directly used as the core part of the energy storage command, namely P. ess_ref =limit(P high ,P ess_min ,Pess_max This allocation logic ensures that the fastest-responding device undertakes the most critical tasks involving rapid fluctuations.

[0046] Finally, the system sends the generated target power generation command to the wind farm's central controller via the communication network, and the charging and discharging power command to the energy storage converter control system. The two subsystems coordinate their operation according to their respective commands to jointly ensure that the power at the power plant's grid connection point accurately tracks the dispatch command.

[0047] Existing technologies typically compare the total dispatch command with the predicted wind power and then compensate for the difference using energy storage. This method fails to distinguish the power demand characteristics at different time scales within the command. In contrast, this invention categorizes commands according to their characteristics through frequency domain decomposition. This allows the fast-responding energy storage system to focus on handling high-frequency fluctuation components, while the wind turbines primarily handle trend-based power components. This approach fully leverages the technological advantages of each component and avoids energy state imbalances caused by the energy storage system over-responding to slow-changing components, thereby improving the overall operating efficiency and stability of the system.

[0048] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station further includes, after the step of allocating the high-frequency component to the energy storage system, a step of dynamically managing the state of charge (SOC) of the energy storage system: real-time monitoring of the SOC of the energy storage system; setting a target median value and an allowable fluctuation range for the SOC; generating a power correction signal when the real-time SOC deviates from the target median value, wherein the direction of the power correction signal is configured to pull the real-time SOC back to the target median value, and the rate of change of the power correction signal is configured to be lower than the rate of change of the high-frequency component; superimposing the power correction signal onto the charging and discharging power command of the energy storage system, and ensuring that the total command after superposition does not exceed its instantaneous charging and discharging power limit.

[0049] This invention provides a specific implementation method for dynamic management of the state of charge (SOC) of an energy storage system. After completing high-frequency component allocation, the method continuously monitors the real-time SOC of the energy storage system and sets a target median and allowable fluctuation range. When the real-time SOC deviates from the target median, the system generates a slowly changing power correction signal. The generation rate of this signal is much slower than the rate of change of the power command responding to grid frequency regulation requirements, and its direction always points towards pulling the SOC back to the target median. Finally, this power correction signal is superimposed on the original charge / discharge power command, ensuring that the total command after superposition does not exceed the instantaneous power limit of the energy storage system.

[0050] The power correction signal is generated by setting a time constant much larger than that of the frequency regulation control loop. For example, the power correction signal can be generated by a proportional-integral (PI) controller with an integral time constant set in the tens of minutes to hours (e.g., 30 minutes to 2 hours), while the integral time constant of the controller responding to the grid frequency regulation command is typically in the seconds to minutes (e.g., 10 seconds to 5 minutes). This difference in magnitude ensures that the power correction process does not interfere with the main rapid power regulation task. The direction of the signal always points in the direction that pulls the state of charge back to the target median value. Finally, the power correction signal is superimposed on the original charge and discharge power command, ensuring that the total command after superposition does not exceed the instantaneous power limit of the energy storage system.

[0051] Existing technologies typically employ a fixed threshold method for state of charge (SOC) management, meaning correction is only performed when the SOC reaches set upper or lower limits. This approach leads to frequent switching of operating modes by the energy storage system near critical states. In contrast, this invention introduces a slowly varying power correction signal, achieving continuous and smooth adjustment of the SOC. This ensures the energy storage system's continued ability to participate in grid regulation while avoiding drastic switching of operating modes, enabling the system to maintain energy balance without service interruption. This gradual adjustment method also effectively reduces interference with primary power regulation tasks, improving the overall operational stability of the system.

[0052] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station further includes a power coordination and matching step before forming the final target power command for the wind farm and the energy storage system: calculating in real time the deviation between the actual total active power at the grid connection point of the power station and the total active power command at the power station level; dynamically generating a power compensation signal based on this power deviation; and dynamically allocating the power compensation signal to the wind power system and / or the energy storage subsystem according to a preset rule based on the current state of charge and the charging and discharging power limit of the energy storage system, so as to eliminate the power deviation.

[0053] This invention provides a specific implementation method for power coordination and matching. After generating initial power commands for the wind farm and energy storage system, the method continuously calculates the deviation between the actual total active power at the grid connection point and the dispatch command. Based on this deviation, a corresponding power compensation signal is dynamically generated, and according to the current state of charge and power limits of the energy storage system, the compensation signal is allocated to the wind power system and the energy storage subsystem according to preset rules. During the allocation process, the system prioritizes the rapid response characteristics of the energy storage system while also considering the power regulation potential of the wind farm. By dynamically adjusting the compensation task allocation ratio between the two subsystems, the power deviation is ensured to be eliminated in a timely manner.

[0054] The dynamic allocation according to preset rules is implemented as follows: The system calculates the power deviation ΔP=P in real time. order -P actual Initially, the compensation signal ΔP is fully allocated to the energy storage subsystem for rapid response. Simultaneously, a first-order inertial element G=1 / (T) is initiated. delay +1) ， (where T) delay A wind turbine compensation command with a time delay (5-30 seconds to accommodate the wind turbine's response delay) is generated. The system continuously monitors the rate of change of the wind farm's actual output. Once it confirms that the wind farm has begun to respond to its compensation command, a power transfer procedure is initiated: at the same rate, the compensation amount of energy storage is gradually reduced while the compensation amount of the wind farm is increased. Within 30-90 seconds, all steady-state compensation tasks are transferred to the wind farm. This rule effectively utilizes the speed of energy storage and the sustainability of the wind turbine.

[0055] Existing technologies typically employ fixed-ratio allocation or simple priority allocation methods to handle power deviations. These methods are ill-suited to the real-time changing operating states of wind farms and energy storage systems. This invention, however, establishes a dynamic allocation mechanism based on the real-time system status. This mechanism flexibly adjusts the compensation strategy according to parameters such as the energy storage system's state of charge and the wind farm's regulation capabilities. It fully utilizes the rapid regulation characteristics of energy storage while considering the actual operational limitations of the wind farm, avoiding the decline in compensation effectiveness due to changes in equipment operating conditions, and improving the reliability and adaptability of power deviation elimination.

[0056] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station further includes an operating condition perception and switching step before the step of receiving the total active power command at the power station level: continuously monitoring key electrical quantities characterizing the grid fault state, the real-time state of charge of the energy storage system, and the power adjustable range of the wind farm; based on preset criteria, determining the current operating condition of the combined power station as one of normal frequency regulation and peak shaving mode, transient emergency support mode, or energy recovery mode; when it is determined that a mode switch is required, a power command ramp function is used to smoothly transition the system's control output from the output value corresponding to the previous mode to the target output value corresponding to the current mode within a preset transition time.

[0057] This invention provides a specific implementation method for operational condition sensing and switching. Before receiving dispatch instructions, the method continuously monitors key electrical quantities of the power grid, including but not limited to the effective value of the grid connection point voltage, grid frequency, voltage / frequency change rate, and negative-sequence voltage and zero-sequence current used to determine asymmetrical faults and grounding faults. Simultaneously, it tracks the state of charge of the energy storage system and the adjustable power range of the wind farm. Based on multi-dimensional operating parameters, the system identifies the current operating condition as a normal frequency regulation and peak shaving mode, a transient emergency support mode, or an energy recovery mode using preset criteria. When a switch to a different operating mode is detected, a power command ramp function is used to smoothly transition the control system output from the original mode to the target value corresponding to the new mode within a preset transition time, ensuring continuous and stable power output during the mode transition process.

[0058] Existing technologies typically rely on a single parameter threshold for operating mode determination, which is prone to misjudgment due to instantaneous signal fluctuations. Furthermore, mode switching often employs a step-like approach, which can easily cause power surges. This invention, however, improves the accuracy of operating condition identification through multi-parameter comprehensive judgment and a delay confirmation mechanism. Simultaneously, it utilizes a ramp transition function to achieve smooth transitions between different modes, avoiding frequent mode switching due to misjudgments and preventing sudden power spikes during switching, thereby enhancing the system's adaptability and stability under various operating conditions.

[0059] In a preferred embodiment, in the frequency regulation and peak shaving coordinated control method of the wind-storage combined power station, the decomposition of the total active power command into high-frequency and low-frequency components is specifically achieved through a first-order inertial filter or a Butterworth filter; the high-frequency component corresponds to the fast fluctuation portion within the passband of the filter, and the low-frequency component corresponds to the slow change portion within the stopband of the filter; furthermore, when issuing the target charge / discharge power command to the energy storage subsystem, the high-frequency component is superimposed with a feedforward compensation signal, which is generated based on the rate of change of ultra-short-term wind power output prediction data.

[0060] This invention provides a specific implementation method for command decomposition and feedforward compensation. The method employs a first-order inertial filter to perform frequency domain decomposition on the total active power command. By appropriately setting the filter time constant, the total command is separated into high-frequency components reflecting rapid fluctuations and low-frequency components reflecting slow changes. Simultaneously, the system calculates the changing trend of wind power output based on ultra-short-term forecast data and generates a corresponding feedforward compensation signal. When issuing a power command to the energy storage system, this feedforward compensation signal is superimposed with the decomposed high-frequency components to form the final charging and discharging power command.

[0061] This method first receives the total active power command at the power plant level issued by the power grid dispatch center in real time through the data acquisition system, and at the same time obtains the ultra-short-term power prediction data of the wind farm (0-4 hours in the future, with a time resolution of 15 minutes) and the real-time state of charge of the energy storage system.

[0062] In a specific implementation, the transfer function of the first-order inertial filter is G = 1 / (T+1), where the time constant T is set according to the power grid frequency regulation requirements, typically ranging from 30s to 300s. For example, when T = 60s, the filter can effectively separate power commands with fluctuation periods of less than 1 minute as the high-frequency component, while power commands with more gradual fluctuations are identified as the low-frequency component. The Butterworth filter can be a low-pass filter, with its cutoff frequency typically set between 0.01Hz and 0.05Hz to ensure effective separation of frequency regulation and peak regulation components.

[0063] Existing technologies typically rely solely on feedback control based on the current state or employ simple differential methods for command decomposition, which suffer from response lag and insufficient decomposition accuracy. This invention, however, combines precise frequency domain decomposition with prediction-based feedforward compensation, ensuring both accurate command decomposition and effectively reducing system response delay through proactive action. This composite control approach better adapts to the power grid's demand for rapid frequency regulation, improving the tracking performance and response speed of combined power plants to power commands.

[0064] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power plant further includes the following steps in generating the power correction signal: setting a static dead zone or a dynamic adaptive dead zone around the target median value of the state of charge; forcing the power correction signal to zero when the real-time state of charge is within the dead zone range; and generating a non-zero power correction signal based on the degree of deviation when the real-time state of charge exceeds the dead zone range.

[0065] This invention provides a specific implementation method for state-of-charge (POC) dead-zone management. During the generation of the power correction signal, this method sets a reasonable dead-zone range around the target median POC value. When the real-time POC is detected to be within this dead-zone range, the system forces the power correction signal to zero, allowing the energy storage system to focus on performing its primary power regulation task. Only when the POC exceeds the dead-zone range is a corresponding power correction signal generated based on the degree of deviation, and the signal magnitude is positively correlated with the degree of deviation.

[0066] A specific embodiment of a static dead zone or a dynamic adaptive dead zone is as follows: A static dead zone is set around a target median (e.g., 50%), ranging from ±5% (i.e., 45% to 55%). When the real-time SOC is within this range, the power correction signal is forcibly set to zero. The dynamic adaptive dead zone can be finely adjusted according to the recent net power direction undertaken by the energy storage system. If the recent activity is mainly discharge, the dead zone can be shifted upwards (e.g., 48% to 58%) to reserve more space for charging, and vice versa.

[0067] Existing technologies typically employ continuous correction methods to manage the state of charge (SOC). This approach generates minute correction signals even when the SOC approaches the target value, potentially leading to unnecessary and frequent actions in the energy storage system due to the combined effect of the main power command and the correction signals. This invention, however, introduces a dead-zone management mechanism that pauses correction operations when the SOC is within a reasonable range. This maintains long-term SOC stability and avoids ineffective equipment actions caused by minute corrections, thereby improving the operating efficiency and lifespan of the energy storage system.

[0068] In a preferred embodiment, in the frequency regulation and peak shaving coordinated control method of the wind-storage combined power station, the step of dynamically allocating power to the wind power system and / or the energy storage subsystem according to preset rules specifically involves: prioritizing and fully allocating the power compensation signal to the energy storage subsystem for rapid response; simultaneously, generating a corresponding wind turbine compensation command with delay characteristics based on the magnitude and direction of the power compensation signal, and sending it to the wind power system; when it is detected that the actual output of the wind power system begins to respond to the wind turbine compensation command, the power compensation signal undertaken by the energy storage subsystem is reduced accordingly at the same rate, and finally the wind power system takes over all steady-state compensation tasks.

[0069] This invention provides a specific implementation method for power compensation master-slave handover. The method prioritizes allocating all power compensation signals to the energy storage system for rapid response, while simultaneously generating corresponding wind turbine compensation commands based on the magnitude and direction of the compensation signals. These wind turbine compensation commands are then sent to the wind farm after an appropriate delay, allowing sufficient time for power adjustment. The system continuously monitors the actual power output changes of the wind farm, and when it begins to respond to the compensation commands, it gradually reduces the compensation power borne by the energy storage system at a corresponding rate, ultimately achieving complete takeover of the steady-state compensation task by the wind farm.

[0070] Existing technologies typically distribute compensation signals to the wind and energy storage systems simultaneously at a fixed ratio. Due to the significant difference in response speed between the two systems, this can easily lead to regulation conflicts or power oscillations. This invention, however, establishes a master-slave compensation and timing handover mechanism, fully leveraging the rapid response of energy storage and the continuous operation of wind turbines. This ensures timely compensation response while avoiding mutual interference between the two subsystems during the compensation process, achieving a smooth transition from rapid response to steady-state maintenance and improving the coordination and stability of the power compensation process.

[0071] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method of the wind-storage combined power station, wherein the determination of the current operating condition of the combined power station as a transient emergency support mode based on preset criteria, specifically includes: continuously monitoring the grid voltage and frequency, and only confirming entry into the transient emergency support mode when the values ​​exceed preset thresholds and the duration reaches a set delay confirmation time; and further refining the transient emergency support mode into a reactive power priority support mode or an active power priority recovery mode according to the combination of voltage drop depth and frequency deviation magnitude, and calling the corresponding preset support strategy.

[0072] This invention provides a specific implementation method for transient condition identification and refined support strategies. The method continuously monitors grid voltage and frequency, and only determines whether to enter a transient emergency support mode when the values ​​exceed a preset threshold and remain there for a set confirmation time. Based on the combination of voltage drop depth and frequency deviation magnitude, the system further distinguishes between a reactive power priority support mode and an active power priority recovery mode. Corresponding preset support strategies are invoked for different modes, including differentiated control methods such as reactive current priority injection and rapid active power adjustment.

[0073] Existing technologies typically employ simple threshold judgments and provide only a uniform support strategy, which is prone to malfunctions due to transient disturbances and cannot adapt to diverse power grid fault types. This invention, however, improves the reliability of operating condition identification through a delayed confirmation mechanism and refines the support strategy based on the combination of fault characteristics. This avoids unnecessary mode switching and provides the most effective support for specific fault types, enhancing the accuracy and adaptability of combined power plants during power grid transient processes.

[0074] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station, wherein superimposing the high-frequency component with a feedforward compensation signal specifically includes: calculating in real time the short-term prediction error between the ultra-short-term prediction data of wind power output and the actual output of the wind farm; dynamically calculating a feedforward reliability coefficient based on the magnitude of the short-term prediction error; wherein, when the prediction error increases, the reliability coefficient decreases; multiplying the feedforward compensation signal with the reliability coefficient to obtain a weighted and corrected feedforward compensation amount, and then superimposing it with the high-frequency component.

[0075] This invention provides a specific implementation method for feedforward reliability weighting. This method calculates the short-term prediction error between ultra-short-term wind power forecast data and actual power output in real time, and dynamically calculates the feedforward reliability coefficient based on the magnitude of this error. When the prediction error is small, the reliability coefficient is close to its maximum value, and the feedforward compensation signal is fully utilized; when the prediction error increases, the reliability coefficient decreases accordingly, automatically weakening the strength of the feedforward compensation. Finally, the reliability-weighted feedforward compensation is superimposed with high-frequency components to form the final control command for the energy storage system.

[0076] Existing technologies typically use predicted data directly to generate feedforward signals without considering the impact of prediction uncertainty on control performance. When prediction errors are significant, distorted feedforward signals can interfere with normal system operation. This invention, however, introduces a reliability assessment mechanism based on prediction errors, enabling adaptive adjustment of the feedforward compensation strength. When predictions are reliable, it fully leverages the speed advantage of feedforward control; when predictions are unreliable, it automatically reduces their impact, thus improving the system's adaptability and control reliability under varying prediction accuracy conditions.

[0077] In a preferred embodiment, the frequency regulation and peak shaving coordinated control method for the wind-storage combined power station is implemented by dynamically calculating a feedforward reliability coefficient based on the magnitude of the short-term prediction error through the following steps: S1: Set a benchmark error threshold; S2: Calculate the average value of the absolute value of the short-term prediction error within a preset time window; S3: Compare the benchmark error threshold with the average value, and calculate the feedforward reliability coefficient through a predefined mapping function; wherein, the mapping function is configured such that: when the average value is less than or equal to the benchmark error threshold, the output value is 1; when the average value is greater than the benchmark error threshold, the output value monotonically decreases between 0 and 1 as the average value increases.

[0078] This invention provides a specific method for calculating the feedforward reliability coefficient. The method first sets a baseline error threshold, and then calculates the average value of the absolute value of wind power prediction errors within a preset time window. By comparing the average value with the baseline error threshold, a predefined mapping function is used to calculate the reliability coefficient. This mapping function is set to output the maximum value when the average value does not exceed the threshold, and the output value decreases as the average value increases when the average value exceeds the threshold. The feedforward reliability coefficient obtained through this calculation method can objectively reflect the reliability level of the prediction data.

[0079] The predefined mapping function can be implemented in the following specific form: Set the reference error threshold E base (For example, 5% of rated power). Calculation time window T w The average absolute value of the prediction error over a period of time (e.g., 10 minutes). .

[0080] The formula for calculating the feedforward reliability coefficient α is: Where k is the attenuation coefficient, used to control the rate at which the confidence level decreases as the error increases; for example, k = 0.1 can be chosen. This linear mapping function is simple to calculate and easy to implement in engineering.

[0081] Existing technologies typically employ complex intelligent algorithms or rely on large amounts of historical data for credibility assessment, which are computationally burdensome and difficult to implement. This invention, however, establishes a simple and parameter-defined credibility assessment method based on a concise mapping relationship between average error and threshold comparison. This method ensures the reasonableness of the assessment results while significantly reducing computational complexity and implementation difficulty, making the feedforward compensation credibility adjustment mechanism easier to apply and promote in practical engineering.

[0082] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be readily implemented by those skilled in the art. Therefore, the present invention is not limited to the specific details without departing from the general concept defined by the claims and their equivalents.

Claims

1. A frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant, characterized in that, Includes the following steps: Receive the total active power command at the power station level issued by the power grid dispatch center; Acquire forecast data of wind farm output and real-time state of charge of energy storage systems; The total active power command is decomposed into rapidly changing high-frequency components and slowly changing low-frequency components. Based on the predicted wind power output data and the real-time state of charge of the energy storage, the upper and lower limits of the power regulation capability of the wind farm at the current moment, and the charging and discharging power limit of the energy storage system at the current moment are determined respectively. Perform coordinated allocation of power commands: Within the charging and discharging power limit range of the energy storage system, the high-frequency component is allocated to the energy storage system as the core part of its charging and discharging power command; The low-frequency component is combined with the basic power generation plan of the wind farm to form the target power generation command of the wind farm, and the command is ensured to be within its power regulation capability range. The target power generation command is sent to the wind farm, and the charging and discharging power command is sent to the energy storage system.

2. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 1, characterized in that, Following the step of allocating the high-frequency components to the energy storage system, the method further includes a step of dynamically managing the state of charge of the energy storage system. Real-time monitoring of the state of charge of the energy storage system; Set the target median value and allowable fluctuation range of the state of charge; When the real-time state of charge deviates from the target median, a power correction signal is generated. The direction of the power correction signal is configured to pull the real-time state of charge back to the target median, and the rate of change of the power correction signal is configured to be lower than the rate of change of the high-frequency component. The power correction signal is superimposed on the charge and discharge power command of the energy storage system, and the total command after superposition is ensured not to exceed its instantaneous charge and discharge power limit.

3. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 1, characterized in that, Before the final target power command for wind farms and energy storage systems is formed, a power coordination and matching step is also included: The deviation between the actual total active power at the power station's grid connection point and the power station-level total active power command is calculated in real time. Based on this power deviation, a power compensation signal is dynamically generated. The power compensation signal is dynamically allocated to the wind power system and / or the energy storage subsystem according to a preset rule based on the current state of charge and the charging / discharging power limit of the energy storage system, in order to eliminate the power deviation.

4. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 1, characterized in that, Before the step of receiving the total active power command at the power plant level, the system also includes a step of operating condition sensing and switching: Continuously monitor key electrical quantities characterizing grid fault status, the real-time state of charge of the energy storage system, and the adjustable power range of the wind farm; Based on preset criteria, the current operating condition of the combined power plant is judged to be one of the following: normal frequency regulation and peak shaving mode, transient emergency support mode, or energy recovery mode. When it is determined that a mode switch is required, a power command ramp function is used to smoothly transition the system's control output from the output value corresponding to the previous mode to the target output value corresponding to the current mode within a preset transition time.

5. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 1, characterized in that, The decomposition of the total active power command into high-frequency and low-frequency components is specifically achieved through a first-order inertial filter or a Butterworth filter. The high-frequency component corresponds to the fast-fluctuating portion within the passband of the filter, and the low-frequency component corresponds to the slowly changing portion within the stopband of the filter. Furthermore, when issuing the target charge / discharge power command to the energy storage subsystem, the high-frequency component is superimposed with a feedforward compensation signal, which is generated based on the rate of change of ultra-short-term wind power output forecast data.

6. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 2, characterized in that, The step of generating the power correction signal further includes: A static dead zone or a dynamic adaptive dead zone is set around the target median value of the state of charge. When the real-time state of charge is within the dead zone, the power correction signal is forcibly set to zero; when the real-time state of charge exceeds the dead zone, a non-zero power correction signal is generated based on the degree of deviation.

7. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 3, characterized in that, The dynamic allocation of power compensation signals to the wind power system and / or the energy storage subsystem according to preset rules specifically means: prioritizing and fully allocating the power compensation signals to the energy storage subsystem for rapid response; Simultaneously, based on the magnitude and direction of the power compensation signal, a corresponding wind turbine compensation command with delay characteristics is generated and sent to the wind electronic system; When the actual output of the wind power system begins to respond to the wind turbine compensation command, the power compensation signal undertaken by the energy storage subsystem is reduced at the same rate, and finally the wind power system takes over all steady-state compensation tasks.

8. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 4, characterized in that, The method of determining the current operating condition of the combined power plant as a transient emergency support mode based on preset criteria specifically includes: The system continuously monitors the grid voltage and frequency. Only when the value exceeds the preset threshold and the duration reaches a set delay confirmation time will it finally confirm the entry into the transient emergency support mode. Furthermore, based on the combination of voltage drop depth and frequency deviation magnitude, the transient emergency support mode is further refined into reactive power priority support mode or active power priority recovery mode, and the corresponding preset support strategy is invoked.

9. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 5, characterized in that, The superposition of the high-frequency component with a feedforward compensation signal specifically includes: Real-time calculation of the short-term prediction error between ultra-short-term wind power output forecast data and actual wind farm output; A feedforward reliability coefficient is dynamically calculated based on the magnitude of the short-term prediction error; wherein, the reliability coefficient decreases as the prediction error increases. The feedforward compensation signal is multiplied by the confidence coefficient to obtain a weighted feedforward compensation amount, which is then superimposed on the high-frequency component.

10. The frequency regulation and peak shaving coordinated control method for a wind-storage combined power plant according to claim 9, characterized in that, The dynamic calculation of a feedforward reliability coefficient based on the magnitude of the short-term prediction error is achieved through the following steps: S1: Set a baseline error threshold; S2: Calculate the average value of the absolute value of the short-term prediction error within a preset time window; S3: Compare the baseline error threshold with the average value, and calculate the feedforward confidence coefficient using a predefined mapping function; wherein the mapping function is configured such that: when the average value is less than or equal to the baseline error threshold, the output value is 1; when the average value is greater than the baseline error threshold, the output value monotonically decreases between 0 and 1 as the average value increases.