Wind-solar-storage collaborative rolling frequency modulation method based on model predictive control

Abstract

The application provides a wind-solar-storage collaborative rolling frequency modulation method based on model predictive control, and relates to the field of new energy grid-connected frequency control.The method comprises the following steps: constructing a frequency response mathematical model of a wind-solar-storage grid-connected system, establishing a multi-objective optimization objective function with the minimization of grid frequency deviation and the minimization of total frequency modulation cost as the target; setting system operation constraints, obtaining current state data of the wind-solar-storage grid-connected system in each control cycle, taking the multi-objective optimization objective function as a rolling optimization objective based on the frequency response mathematical model, rolling solving the optimization problem in a prediction time domain, and outputting wind turbine active power instructions, photovoltaic system active power instructions and energy storage system active power instructions and delivering them to an execution end. The technical problems of insufficient frequency stability and poor frequency modulation economy of new energy participating in grid frequency modulation in the prior art are solved. The technical effects of improving the frequency stability of the new energy grid-connected system and optimizing the frequency modulation economy are achieved.

Description

Technical Field

[0001] This invention relates to the field of new energy grid-connected frequency control technology, specifically to a wind-solar-storage coordinated rolling frequency regulation method based on model predictive control. Background Technology

[0002] With the large-scale grid connection of renewable energy sources such as wind and solar power, the power system faces a severe challenge of declining frequency stability. After traditional synchronous generator units are replaced by new energy power generation units, the system inertia decreases and the frequency regulation capability weakens. When large power disturbances occur, grid frequency fluctuations intensify, which can trigger low-frequency load shedding or even cause large-scale power outages in severe cases.

[0003] Existing technical solutions for renewable energy to participate in grid frequency regulation have the following shortcomings: wind and solar power output is intermittent and fluctuates, and their response capability is limited when participating in frequency regulation alone. They lack an effective coordination mechanism. Although energy storage systems have the advantage of rapid response, their state of charge is prone to exceeding the safety boundary during frequent frequency regulation, leading to accelerated aging of energy storage and a decline in the economic efficiency of system operation. Furthermore, existing methods often take minimizing frequency deviation as the sole objective and fail to take into account the economic efficiency of frequency regulation, resulting in a high economic cost for renewable energy to participate in primary frequency regulation.

[0004] In summary, existing technologies suffer from insufficient frequency stability and poor economic efficiency when new energy sources participate in grid frequency regulation. Summary of the Invention

[0005] The purpose of this application is to provide a wind-solar-storage coordinated rolling frequency regulation method based on model predictive control, which can solve the technical problems of insufficient frequency stability and poor frequency regulation economy when new energy participates in grid frequency regulation in the existing technology.

[0006] In view of the above problems, this application provides a wind-solar-storage coordinated rolling frequency regulation method based on model predictive control. The method includes: constructing a frequency response mathematical model of the wind-solar-storage grid-connected system, which includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response sub-model; establishing a multi-objective optimization objective function with the goals of minimizing grid frequency deviation and minimizing total frequency regulation cost; setting system operation constraints; acquiring the current state data of the wind-solar-storage grid-connected system in each control cycle; solving the optimization problem in the prediction time domain based on the frequency response mathematical model and the multi-objective optimization objective function as the rolling optimization objective; outputting active power commands for the wind turbine, photovoltaic system, and energy storage system; and distributing these commands to the execution end through a hierarchical control architecture.

[0007] Optionally, the system operation constraints include power balance constraints of the wind-solar-storage combined system, upper and lower limit constraints of wind turbine output, upper and lower limit constraints of photovoltaic system output, charging and discharging power constraints of energy storage system, safety boundary constraints of energy storage system state of charge, and power ramping constraints.

[0008] Optionally, the mathematical expression for the wind turbine inertia response sub-model is: ;in, This refers to the actual active power output of the wind turbine. This refers to the output power in maximum power point tracking mode. This represents the increment of virtual inertia response power. The increment of the droop control power is calculated as follows: ;in, This is the virtual inertia coefficient; The power grid frequency change rate is given; the formula for calculating the droop control power increment is: ;in, This is the droop coefficient; This refers to the power grid frequency deviation.

[0009] Optionally, the mathematical expression of the photovoltaic system load shedding control sub-model is: ;in, The actual active power output of the photovoltaic system; For load reduction rate; This is the power at the maximum power point.

[0010] Optionally, the mathematical expression for the power response sub-model of the energy storage system is: ; ;in, The actual active power output of the energy storage system; This refers to the power command value for the energy storage system. Let be the state of charge at time t; This is the initial state of charge; To improve the charging and discharging efficiency of energy storage systems; This refers to the rated capacity of the energy storage system.

[0011] Optionally, the expression for the multi-objective optimization objective function is: ;in, To optimize the time domain; Let be the power grid frequency deviation at time t; and These are the weighting coefficients; Let be the total cost function of frequency modulation at time t.

[0012] Optionally, obtain ultra-short-term power prediction data and prediction error ranges for wind turbines and photovoltaic systems; construct a wind and solar power uncertainty set based on the prediction error range; and add robust constraints to the constraints of the rolling optimization problem based on the wind and solar power uncertainty set.

[0013] Optionally, the active power commands of the wind turbine, photovoltaic system, and energy storage system are verified to meet the power balance constraints. If the verification passes, the commands are sent to the execution end through the hierarchical control architecture. If the verification fails, the power commands are corrected step by step according to the priority order of the wind turbine, photovoltaic system, and energy storage system until the power balance constraints are met.

[0014] One or more technical solutions provided in this application have at least the following technical effects or advantages: The method provided in this application constructs a frequency response mathematical model for a wind-solar-storage grid-connected system. This model includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response sub-model. A multi-objective optimization objective function is established with the goals of minimizing grid frequency deviation and minimizing total frequency regulation cost. System operating constraints are set, and the current state data of the wind-solar-storage grid-connected system is acquired in each control cycle. Based on the frequency response mathematical model and using the multi-objective optimization objective function as the rolling optimization objective, the optimization problem is solved continuously in the prediction time domain. The method outputs active power commands for the wind turbine, photovoltaic system, and energy storage system, and distributes them to the execution end through a hierarchical control architecture. This achieves the technical effect of improving the frequency stability of the renewable energy grid-connected system while optimizing the economic efficiency of frequency regulation.

[0015] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1A flowchart illustrating the wind-solar-storage coordinated rolling frequency regulation method based on model predictive control provided in this application.

[0018] Figure 2 A schematic diagram illustrating the constraint terms of the system operation constraints in the wind-solar-storage coordinated rolling frequency regulation method based on model predictive control provided in this application. Detailed Implementation

[0019] This application provides a model predictive control-based rolling frequency regulation method for wind, solar, and energy storage coordination, designed to address the technical problems of insufficient frequency stability and poor economic efficiency when new energy sources participate in grid frequency regulation. It achieves the technical effect of improving the frequency stability of new energy grid-connected systems while optimizing frequency regulation economics.

[0020] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. It should also be noted that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0021] like Figure 1 As shown, this application provides a wind-solar-storage coordinated rolling frequency regulation method based on model predictive control, which includes: A frequency response mathematical model for a wind-solar-storage grid-connected system is constructed. The frequency response mathematical model includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response sub-model.

[0022] Specifically, a frequency response mathematical model for a wind-solar-storage grid-connected system is constructed, unifying the physical characteristics of the three heterogeneous power sources—wind, solar, and energy storage—into a mathematical expression solvable by an optimization algorithm. This frequency response mathematical model includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response sub-model. Specifically, the wind turbine inertia response sub-model simulates the inertia and primary frequency regulation characteristics of a synchronous generator through a virtual synchronous generator (VSG) control strategy. The virtual inertia response power increment and the droop control power increment together constitute the frequency regulation power support. Specifically, feedback quantities for the frequency change rate and frequency deviation are introduced through the virtual inertia coefficient and droop coefficient, respectively. An additional power increment is superimposed on the maximum power point tracking (MPPT) basis, enabling the wind turbine to actively simulate the inertia support and primary frequency regulation characteristics of a synchronous generator, thus compensating for the low inertia deficiency of new energy sources.

[0023] Photovoltaic system load shedding control sub-model: The photovoltaic system reserves backup power for frequency support through load shedding control. The load shedding rate is the adjustment variable, ranging from 5% to 22%. This causes the photovoltaic operating point to deviate from the maximum power point (MPP), thus reserving adjustable active power reserves. When the grid frequency drops, the backup power is quickly released by reducing the load shedding rate, achieving bidirectional regulation capability. Energy storage system power response sub-model: The energy storage system responds to power commands and achieves charge and discharge control through a bidirectional converter. The state of charge of the energy storage system... As the charging and discharging power changes cumulatively, it ensures that the energy storage will not shut down due to overcharging or over-discharging during the entire frequency regulation process.

[0024] By constructing a frequency response mathematical model that includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response sub-model, the inertia support capability and primary frequency regulation capability of wind turbines can be fully utilized. At the same time, the load shedding reserve power of the photovoltaic system can be used to participate in frequency regulation. Combined with the rapid response characteristics of the energy storage system, the advantages of wind, solar and energy storage resources can be complemented and coordinated for frequency regulation.

[0025] Furthermore, the mathematical expression for the wind turbine inertia response sub-model is as follows: ;in, This refers to the actual active power output of the wind turbine. This refers to the output power in maximum power point tracking mode. This represents the virtual inertia response power increment. The increment of the droop control power is calculated as follows: ;in, This is the virtual inertia coefficient; The power grid frequency change rate is given; the formula for calculating the droop control power increment is: ;in, This is the droop coefficient; This refers to the power grid frequency deviation.

[0026] Specifically, the actual output active power of the wind turbine Output from maximum power point tracking Superimposed virtual inertial response power increment and droop control power increment The mathematical expression for the wind turbine inertia response sub-model is as follows: ;in, This refers to the actual active power output of the wind turbine. This refers to the output power in maximum power point tracking mode. This represents the virtual inertia response power increment. This is for droop control power increment.

[0027] Virtual inertia response power increment With the rate of change of grid frequency Proportional, the proportionality coefficient is the virtual inertia coefficient The intensity of the inertial response of the wind turbine generator set simulating that of a synchronous generator set is characterized by the following formula for calculating the power increment of the virtual inertial response: ;in, This is the virtual inertia coefficient; This refers to the rate of change of the power grid frequency; when the power grid frequency drops rapidly, , If the frequency is positive, the wind turbines increase their power output to prevent the frequency from dropping further. However, when the grid frequency rises rapidly... , If the value is negative, the wind turbine reduces its power output to suppress frequency rise.

[0028] Droop control power increment Frequency deviation from the power grid Proportional, the proportionality coefficient is the droop coefficient The formula for calculating the droop control power increment, which characterizes the response intensity of the wind turbine unit participating in primary frequency regulation, is as follows: ;in, The droop coefficient is... This represents the power grid frequency deviation. Among them, the virtual inertia coefficient... and droop coefficient It adapts to changes in system frequency.

[0029] Virtual inertia acts on the transient phase where frequency changes drastically but for a short period of time, while droop control acts on the quasi-steady phase where frequency deviation persists. Together, they transform the wind turbine from a passive power source that moves with the wind into a flexible unit that can actively participate in grid frequency regulation, thereby providing a quantifiable boundary for the wind turbine's frequency regulation capability.

[0030] Furthermore, the mathematical expression for the photovoltaic system load shedding control sub-model is: ;in, The actual active power output of the photovoltaic system; For load reduction rate; The maximum power point power is; the reserve power is: ; wherein, the backup power Used for frequency modulation power support when the system frequency drops.

[0031] Specifically, the photovoltaic system reserves backup power through load shedding control to support frequency regulation power during system frequency drops. The mathematical expression of the photovoltaic system load shedding control sub-model is as follows: ;in, The actual active power output of the photovoltaic system. The power at the maximum power point. The load shedding rate (d) is dynamically set based on system frequency regulation requirements, ranging from 5% to 22%. When the system frequency is within the normal range, the photovoltaic system operates in load shedding mode, reserving backup power. When the system frequency drops below a preset frequency drop threshold, the photovoltaic system releases backup power and generates additional power to support frequency recovery. The normal range of the grid frequency is directly defined by the grid operation regulations. The preset frequency drop threshold refers to the frequency boundary value that triggers the photovoltaic system to switch from load shedding to power release. Its determination follows these principles: the preset frequency drop threshold is higher than the frequency emergency limit but lower than the upper limit of the primary frequency regulation dead zone, ensuring sufficient release of backup capacity before the frequency deteriorates to a dangerous level.

[0032] By operating under reduced load, photovoltaic systems transform themselves from uncontrollable power sources into flexible frequency-regulating resources with active up-regulation capabilities. When the frequency drops, they dynamically release reserve power to support the grid, while providing continuously adjustable control variables for model predictive control (MPC).

[0033] Furthermore, the mathematical expression for the power response sub-model of the energy storage system is: ; ;in, The actual active power output of the energy storage system; This refers to the power command value for the energy storage system. Let be the state of charge at time t; This is the initial state of charge; To improve the charging and discharging efficiency of energy storage systems; This refers to the rated capacity of the energy storage system.

[0034] Specifically, the mathematical expression for the power response sub-model of the energy storage system is as follows: ; ;in, This represents the actual active power output of the energy storage system; a positive value indicates discharging, and a negative value indicates charging. This refers to the power command value for the energy storage system. Let be the state of charge at time t; This is the initial state of charge; To improve the charging and discharging efficiency of energy storage systems; This refers to the rated capacity of the energy storage system.

[0035] State of charge of energy storage system As the charging and discharging power cumulatively changes, it needs to be maintained within the safety boundary. To prevent the SOC of the energy storage system from exceeding the limit during frequent frequency regulation, a system is set... and As a safety boundary, preferably, take , .when near When the energy storage system reduces its discharge power or actively charges to increase its state of charge (SOC); when near At this time, the energy storage system reduces charging power or actively discharges to lower the SOC.

[0036] A multi-objective optimization objective function is established with the goals of minimizing power grid frequency deviation and minimizing total frequency regulation cost.

[0037] Furthermore, the expression for the multi-objective optimization objective function is: ;in, To optimize the time domain; Let be the power grid frequency deviation at time t; and These are the weighting coefficients. Let be the total cost function of frequency modulation at time t.

[0038] Specifically, the multi-objective optimization objective function unifies the control objectives of minimizing grid frequency deviation and minimizing total frequency regulation cost into a single scalar optimization index through weighted summation. The expression of the multi-objective optimization objective function is: m ;in, To optimize the time domain; Let be the power grid frequency deviation at time t; Let be the total cost function of frequency modulation at time t; and These are the weighting coefficients. / The larger the ratio, the more the system tends to minimize frequency deviation, resulting in better frequency modulation performance, but the economic cost may increase. The larger the ratio, the more the system is biased towards frequency regulation economy, resulting in lower operating costs, but frequency fluctuations may increase. In practical applications, the weighting coefficient is adjusted online according to the operating conditions of the system: when the frequency safety margin is sufficient, the economic weighting is appropriately increased, and when the frequency fluctuations are severe, frequency stability is prioritized.

[0039] The total frequency regulation cost comprises three parts: wind turbine frequency regulation cost, photovoltaic system frequency regulation cost, and energy storage system frequency regulation cost. These respectively reflect the economic losses from wind turbines operating off their maximum power point, the economic losses from photovoltaic systems operating at reduced load, and the combined costs of energy storage systems deviating from their optimal power output and state of charge (SOC). The total frequency regulation cost function... The expression is: ;in, For the frequency regulation cost of wind turbine units; Frequency regulation cost of photovoltaic systems; This refers to the frequency regulation cost of the energy storage system.

[0040] The frequency regulation cost of the wind turbine The expression is: ;in, This refers to the frequency regulation cost coefficient for wind turbine generators. This represents the maximum output power of the wind turbine. Let t be the actual output power of the wind turbine at time t.

[0041] The frequency regulation cost of the photovoltaic system The expression is: ;in, This refers to the frequency regulation cost coefficient of the photovoltaic system. This represents the maximum output power of the photovoltaic system. Let t be the actual output power of the photovoltaic system at time t.

[0042] The frequency regulation cost of the energy storage system The expression is: ;in, The power cost factor for energy storage systems. The penalty factor for deviation of the SOC of the energy storage system. For the optimal state of charge, preferably, Take a value of 0.5 to 0.6. Penalty item. The system guides the energy storage system to actively maintain the State of Charge (SOC) near the optimal level during frequency regulation, thus avoiding frequent overshooting.

[0043] By establishing a multi-objective optimization objective function with the goals of minimizing grid frequency deviation and minimizing total frequency regulation cost, the system can effectively reduce the economic cost of new energy sources participating in primary grid frequency regulation while ensuring grid frequency stability and taking into account the economic efficiency of frequency regulation.

[0044] Set system operation constraints, acquire current state data of the wind-solar-storage grid-connected system in each control cycle, solve the optimization problem in the prediction time domain based on the frequency response mathematical model and the multi-objective optimization objective function as the rolling optimization objective, output active power commands for wind turbines, photovoltaic systems and energy storage systems, and send them to the execution end through a hierarchical control architecture.

[0045] Furthermore, such as Figure 2 As shown, the system operation constraints include power balance constraints of the wind-solar-storage combined system, upper and lower limit constraints of wind turbine output, upper and lower limit constraints of photovoltaic system output, charging and discharging power constraints of energy storage system, safety boundary constraints of energy storage system state of charge, and power ramping constraints.

[0046] Specifically, the system operation constraints include six categories: power balance constraints for the wind-solar-storage combined system, upper and lower limit constraints for wind turbine output, upper and lower limit constraints for photovoltaic system output, charging and discharging power constraints for the energy storage system, safety boundary constraints for the state of charge of the energy storage system, and power ramping constraints. Among these, the power balance constraints are: ;in, Let t be the actual active power output of the wind turbine. Let t be the actual active power output of the photovoltaic system. At time t, the actual active power output of the energy storage system is, Let be the load power at time t. The power balance constraint ensures a real-time balance between the active power output of the wind-solar-storage combined system and the load demand and network loss at time t.

[0047] Wind turbine output constraints: Wherein, the actual active power output of the wind turbine at time t is... This represents the maximum active power that the wind turbine is allowed to output at the current wind speed.

[0048] Photovoltaic system output constraints: ;in, This represents the maximum active power that the photovoltaic system is allowed to output under current light and temperature conditions.

[0049] Energy storage system power constraints: ;in, This represents the maximum charging power of the energy storage system (take a positive value). This represents the maximum discharge power of the energy storage system (take a positive value).

[0050] Energy storage system SOC constraints: ;in, This is the upper limit of the state of charge, and its value ranges from 0.1 to 0.2. This represents the upper limit of the state of charge, with a value ranging from 0.9 to 0.95.

[0051] Power ramping constraints: ; ; ;in, , , These represent the maximum ramp rates for wind turbines, photovoltaic systems, and energy storage systems, respectively. To control the cycle.

[0052] Then, in each control cycle, the current status data of the wind-solar-storage grid-connected system is acquired, including real-time acquisition of the grid frequency deviation at the current moment through frequency measurement devices at the grid connection point, such as PMU synchronous phasor measurement units or frequency transmitters. The actual output active power of the wind turbine unit at the current moment is collected in real time by the power transmitter at the wind turbine terminal or the SCADA system. The actual output active power of the photovoltaic system at the current moment The actual output active power of the energy storage system at the current moment is collected in real time by the power detection module of the energy storage converter (PCS). The current state of charge of the energy storage system is collected and reported in real time by the battery management system (BMS) of the energy storage system. Based on the frequency response mathematical model and with the multi-objective optimization objective function as the rolling optimization objective, the system state-space equation is established: ; ;in, The system state vector contains , , , and ; To control the input vector, it includes , and ; The output vector contains ; , , For each control cycle, the coefficient matrix is ​​used. Solve the following rolling optimization problem: ;in, To predict the time domain, The state weight matrix is... To control the weight matrix, under the premise of satisfying all system operation constraints, a set of future control sequences from the current time to the end of the prediction time domain is solved. Only the first control variable in the future control sequence is used as the output. The output includes the active power command of the wind turbine, the active power command of the photovoltaic system, and the active power command of the energy storage system in the current period. The active power commands are issued to each unit for execution through a hierarchical control architecture. The hierarchical control architecture includes a scheduling control layer, a coordination control layer, and a local control layer. The scheduling control layer is used to formulate the day-ahead scheduling plan of the wind, solar, and energy storage systems based on frequency prediction data. The coordination control layer is used to calculate the power commands of each unit on a rolling basis based on model predictive control methods. The local control layer is used to receive the power commands and control the converters of each unit to execute the commands.

[0053] Then, in the next control cycle (k+1), the actual state is re-acquired and the above process is repeated, thus forming a closed-loop rolling mechanism of prediction, optimization, execution and re-prediction, which enables the control strategy to continuously resist the random fluctuations of wind and solar power output and the interference caused by load changes, and always maintain optimal tracking of the grid frequency.

[0054] By setting system operating constraints, the optimized power command satisfies both the rigid requirements of instantaneous power balance in the power grid and the physical limits and dynamic response capabilities of each device, fundamentally ensuring the engineering feasibility and effectiveness of the control strategy. By employing model predictive control (MMC) to solve the optimization problem in a rolling manner, the control strategy can be adjusted in real time according to the system state, adapting to the dynamic changes in renewable energy output and load demand. Compared to traditional dual-loop control, it exhibits better dynamic regulation performance, effectively mitigating the impact of renewable energy output fluctuations on the power grid frequency, improving the frequency stability of the renewable energy grid-connected system, and optimizing frequency regulation economy.

[0055] Furthermore, before solving the optimization problem in the prediction time domain, the process includes: acquiring ultra-short-term power prediction data and prediction error intervals for wind turbines and photovoltaic systems; constructing a wind and solar power uncertainty set based on the prediction error intervals; and adding robust constraints to the constraints of the rolling optimization problem based on the wind and solar power uncertainty set.

[0056] Specifically, before rolling the optimization problem within the prediction time domain, in each control period k, ultra-short-term power prediction data for each moment in the future prediction time domain is obtained from the new energy power prediction system, including wind turbine power prediction values ​​and photovoltaic system power prediction values. At the same time, the prediction error interval for each moment is determined by the statistical distribution of historical prediction errors, such as a normal distribution. The prediction error interval is defined by the error quantile with a confidence level of 95% to 99% as the upper and lower bounds.

[0057] Based on the prediction error range, an uncertain set of wind and solar power is constructed. This uncertain set reflects all possible fluctuations in the actual output of wind and solar power in the future. Then, the uncertain set of wind and solar power is substituted into the constraints of the rolling optimization problem as a robust constraint. This robust constraint is added as an additional hard condition to the optimization solution framework of MPC, and together with the original constraints such as power balance, equipment limits, SOC boundary, and ramp rate, it forms the constraint system of the closed-loop optimization problem.

[0058] By quantifying the wind and solar power output prediction error into an uncertainty set and transforming it into a robust constraint, the MPC optimization command is feasible for any prediction deviation, fundamentally avoiding the risk of power imbalance caused by inaccurate prediction, improving the safety and reliability of the control strategy under strong volatile weather conditions, and reducing the economic waste of excessive reserve to cope with uncertainty.

[0059] Furthermore, after solving the optimization problem in the prediction time domain, the process includes: verifying whether the active power commands of the wind turbine, photovoltaic system, and energy storage system meet the power balance constraints; if the verification passes, the commands are sent to the execution end through a hierarchical control architecture; if the verification fails, the power commands are corrected step by step according to the priority order of the wind turbine, photovoltaic system, and energy storage system until the power balance constraints are met.

[0060] Specifically, after the Model Predictive Control (MPC) is solved, three sets of active power commands for the current period are obtained, including active power commands for wind turbines, photovoltaic systems, and energy storage systems. These commands are then input into the power balance constraints for verification. When the absolute deviation is less than the allowable deviation threshold, such as 0.1MW, the verification is considered successful, and the commands are sent to the execution end through the hierarchical control architecture.

[0061] If the verification fails, the power commands are corrected level by level according to the priority order of wind turbine, photovoltaic system, and energy storage system. The first level prioritizes adjusting the wind turbine commands, increasing or decreasing the power between its maximum available output and zero output at the current wind speed to eliminate the imbalance. If the wind turbine's adjustment capacity is insufficient to fully compensate for the deviation, the remaining imbalance is passed to the second level, the photovoltaic system, which continues to adjust between its maximum available output and zero output. If a deviation still exists after the first two levels of correction, the energy storage system will adjust, absorbing or releasing power to completely absorb the remaining imbalance under the dual constraints of the energy storage charging and discharging power upper limit and the SOC safety boundary. During the step-by-step correction process, the adjustment amount of each level of equipment must strictly meet its own upper and lower output limits and ramp rate constraints. After the correction is completed, the power balance is re-verified. If the deviation falls within the allowable threshold, the command is issued for execution. If the adjustment capacity of a certain level is exhausted and the deviation is still not eliminated, the command continues to be passed to the next level, and so on, until the power balance constraints are met, ensuring that the three sets of commands issued in the end satisfy both instantaneous power balance and remain within the physical safety boundaries of each device.

[0062] By verifying and correcting step by step, the system ensures that the issued instructions strictly meet the instantaneous power balance and avoids execution risks. At the same time, the system prioritizes the use of wind turbines and photovoltaics to reduce the deep charging and discharging of energy storage, thus balancing system safety and equipment economy.

[0063] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0064] Obviously, those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A wind-solar-storage coordinated rolling frequency regulation method based on model predictive control, characterized in that, include: A frequency response mathematical model for a wind-solar-storage grid-connected system is constructed. The frequency response mathematical model includes a wind turbine inertia response sub-model, a photovoltaic system load shedding control sub-model, and an energy storage system power response model. A multi-objective optimization objective function is established with the goals of minimizing power grid frequency deviation and minimizing total frequency regulation cost. Set system operation constraints, acquire current state data of the wind-solar-storage grid-connected system in each control cycle, solve the optimization problem in the prediction time domain based on the frequency response mathematical model and the multi-objective optimization objective function as the rolling optimization objective, output active power commands for wind turbines, photovoltaic systems and energy storage systems, and send them to the execution end through a hierarchical control architecture.

2. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, The system operation constraints include power balance constraints of the wind-solar-storage combined system, upper and lower limit constraints of wind turbine output, upper and lower limit constraints of photovoltaic system output, charging and discharging power constraints of energy storage system, safety boundary constraints of energy storage system state of charge, and power ramping constraints.

3. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, The mathematical expression for the wind turbine inertial response sub-model is as follows: ； in, This refers to the actual active power output of the wind turbine. This refers to the output power in maximum power point tracking mode. This represents the increment of virtual inertia response power. For droop control power increment; The formula for calculating the virtual inertia response power increment is: ； in, This is the virtual inertia coefficient; The rate of change of the power grid frequency; The formula for calculating the droop control power increment is: ; in, This is the droop coefficient; This refers to the power grid frequency deviation.

4. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, The mathematical expression for the photovoltaic system load shedding control sub-model is: ； in, The actual active power output of the photovoltaic system; For load reduction rate; This is the power at the maximum power point.

5. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, The mathematical expression for the power response sub-model of the energy storage system is: ； ； in, The actual active power output of the energy storage system; This refers to the power command value for the energy storage system. Let be the state of charge at time t; This is the initial state of charge; To improve the charging and discharging efficiency of energy storage systems; This refers to the rated capacity of the energy storage system.

6. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, The expression for the multi-objective optimization objective function is: ； in, To optimize the time domain; Let be the power grid frequency deviation at time t; and These are the weighting coefficients; Let be the total cost function of frequency modulation at time t.

7. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, Before rolling the solution of the optimization problem in the prediction time domain, the following are included: Obtain ultra-short-term power forecast data and forecast error range for wind turbines and photovoltaic systems; Construct a set of uncertain wind and solar power based on the prediction error interval; Robust constraints are added to the constraints of the rolling optimization problem based on the aforementioned uncertain set of wind and solar power.

8. The wind-solar-storage coordinated rolling frequency regulation method based on model predictive control as described in claim 1, characterized in that, After solving the optimization problem in the prediction time domain, the following steps are included: Verify whether the active power commands of wind turbines, photovoltaic systems, and energy storage systems meet the power balance constraints. If the verification passes, the instruction will be sent to the execution end through the hierarchical control architecture; If the verification fails, the power command will be corrected step by step according to the priority order of wind turbine, photovoltaic system and energy storage system until the power balance constraint is met.

Images