Wind and light storage coordination control method and system for virtual synchronous machine

Abstract

The invention discloses a wind and light storage coordination control method and system for a virtual synchronous machine, and the method comprises the steps: building a wind power plant-energy storage power station double-layer response mechanism at a first stage, and employing a multi-time scale coordination control strategy for a second-stage energy storage system; and in combination with the virtual synchronous control characteristics of the optical storage power station, establishing a multi-scene optimization model of a wind and light storage multi-source collaborative double-layer coupling mechanism, and realizing dynamic optimization of output of each unit among inertia support, frequency modulation capacity and energy storage health state. According to the method, wind power plant rotor kinetic energy release combined overspeed load shedding control, a fuzzy control means based on an energy storage charge state and a VSG control means based on a virtual inertia and damping coefficient self-adaption photovoltaic system are considered, then wind and light energy storage frequency modulation coordination is optimized, and the system inertia safety is effectively improved on various working conditions and different time scales.

Description

Technical Field

[0001] The present invention relates to a wind-solar-storage coordinated frequency regulation control method for an electric power system, and in particular to a wind-solar-storage coordinated control method and system for a virtual synchronous machine. Background Art

[0002] With the rapid development of new energy technologies, the global application of wind and photovoltaic power generation continues to expand. However, integrating a high proportion of renewable energy into the grid also presents significant technical challenges. Traditional power systems primarily rely on synchronous generators for inertia support and frequency regulation. However, renewable energy generation equipment such as wind and photovoltaics, due to their use of power electronics interfaces, are weakly coupled to the grid and cannot directly provide traditional rotational inertia support. This significantly reduces the inertia level of the grid system. The volatility and randomness of renewable energy power plants further exacerbate system frequency instability. In scenarios with a high proportion of renewable energy integration, the grid faces the following challenges: insufficient system inertia and significantly weakened frequency response capabilities, resulting in an increased rate of change of frequency (RoCoF) when the grid encounters disturbances. Traditional frequency regulation resources are insufficient and slow to respond, making it difficult to meet the time-sensitive frequency regulation requirements imposed by renewable energy volatility. Although energy storage devices offer fast response characteristics, their control methods lack deep synergy with renewable energy power plants, failing to fully realize their frequency regulation potential.

[0003] To address these issues, virtual synchronous generator (VSG) technology has been widely researched and applied in recent years. VSG technology uses power electronics to simulate the inertia characteristics and frequency regulation behavior of traditional synchronous generators, enabling renewable energy power plants to provide inertia support and frequency regulation. However, existing VSG technology still has the following shortcomings: Insufficient research on the coordinated control of wind power, photovoltaic power plants, and energy storage makes it difficult to fully tap the regulation potential of various renewable energy power plants. Failure to optimize the state of charge (SOC) of the energy storage system can lead to excessive discharge or charging of the energy storage system, affecting its operating life and efficiency. The lack of dynamic adaptive control methods under different operating conditions makes it impossible to achieve efficient coordinated frequency control of wind and photovoltaic power plants. Therefore, how to use VSG technology to deeply coordinate wind farms, photovoltaic power plants, and energy storage systems, fully tap the regulation potential of regional wind, photovoltaic, and energy storage systems, and enhance frequency security while improving system inertia remains a key issue that needs to be addressed in the current field of power system technology. Summary of the Invention

[0004] Purpose of the invention: In response to the above problems, the present invention proposes a virtual synchronous machine wind, solar and storage coordinated control method and system, which effectively improves the inertia level and frequency stability of high-proportion new energy power grids, and provides important technical support for the safe operation of the "double-high" power system.

[0005] Technical Solution: The technical solution adopted by the present invention is a virtual synchronous machine wind, solar and energy storage coordinated control method, including: the wind farm uses rotor kinetic energy release combined with overspeed load reduction control to provide inertial support in the early stage of frequency response; the early stage is the 0-5 seconds period after the grid frequency disturbance occurs;

[0006] The energy storage system provides frequency response in a short time scale and power balance support in a long time scale at different time scales; the short time scale is milliseconds to seconds; the long time scale is minutes;

[0007] According to the real-time monitoring of the energy storage system's state of charge and the rate of change of the grid frequency, the system's virtual inertia is dynamically adjusted through two-dimensional fuzzy control;

[0008] The PV power station uses virtual inertia to coordinate the damping coefficient for spontaneous dynamic response, including: automatic adjustment of virtual inertia and damping coefficient according to the system frequency deficit and frequency change rate;

[0009] A wind, solar and storage multi-source collaborative double-layer coupling mechanism and a multi-scenario optimization model are established, including an upper-layer model and a lower-layer model. The upper-layer model is a wind, solar and storage capacity optimization configuration model, and the lower-layer model is a multi-scenario optimization model. The lower-layer model feeds back the multi-scenario weighted operating cost to the upper-layer model to update the objective function of the upper-layer model.

[0010] The rotor kinetic energy release combined with overspeed load reduction control includes: maintaining the rotor speed higher than the rated speed by adjusting the pitch angle of the wind turbine and the generator torque command; reserving rotor kinetic energy for release, releasing the rotor kinetic energy of the wind turbine when the frequency drops to increase the system inertia; automatically reducing the load when the frequency exceeds a set threshold, so that the wind turbine operating output curve remains above the original load reduction operating output curve, and the wind turbine output power is kept higher than the output power before the frequency drops. When the system frequency stabilizes within the allowable range, the overspeed load reduction stops.

[0011] An optimal expression for the wind turbine output power under overspeed load reduction control is:

[0012]

[0013] Where, P out is the output power of the wind turbine under overspeed load reduction control; P max is the maximum mechanical power of the wind turbine when it is running; ΔP dis the output power variation of the wind turbine under overspeed load reduction control; C P,out (λ,β) is the wind energy utilization coefficient under overspeed load reduction control; C P,max (λ,β) is the wind energy utilization coefficient when the wind turbine outputs the maximum mechanical power; ρ is the air density; r is the rotor radius; v is the wind speed; and d is the load reduction rate.

[0014] The energy storage system provides frequency response within a short time scale and power balance support within a long time scale at different time scales, including: supercapacitor energy storage undertakes the frequency regulation task within a short time scale by participating in the initial frequency regulation; battery pack energy storage is responsible for power balance support within a long time scale by adjusting power output.

[0015] The method of dynamically adjusting the virtual inertia of the system through two-dimensional fuzzy control based on the real-time monitored state of charge of the energy storage system and the rate of change of the grid frequency includes: using a two-dimensional fuzzy control method to output a virtual adjustment coefficient of the virtual synchronous machine interface based on the state of charge of the energy storage system and the rate of change of the grid frequency, and dynamically adjusting the virtual inertia of the system based on the virtual adjustment coefficient.

[0016] The virtual inertia of the system is dynamically adjusted according to the virtual adjustment coefficient, wherein a preferred calculation formula is:

[0017]

[0018] Where H is the virtual inertia, H0 is the inertia constant when the state of charge is normal; k1 is the virtual adjustment coefficient; k2 is the virtual inertia control parameter, SOC is the state of charge of the energy storage system, a is the discharge limit value of the energy storage battery, and b is the charging limit value of the energy storage battery.

[0019] The automatic adjustment of virtual inertia and damping coefficient according to the system frequency deficit and frequency change rate includes:

[0020] When the system is running in steady state, the virtual inertia and damping coefficient remain at the basic values of steady state operation;

[0021] When Δω>0 and dw / dt increases and is greater than 0, both the virtual inertia and the damping coefficient are increased to suppress the frequency change rate and offset. When Δω>0 and dw / dt decreases and is less than 0, the virtual inertia is reduced to shorten the settling time of the frequency step response, but the damping coefficient is kept increased to suppress overshoot. Δω is the system frequency deficit, and dw / dt is the frequency change rate.

[0022] When the output power stabilizes at the reference value, the virtual inertia and damping coefficient return to the steady-state value.

[0023] The virtual inertia and damping coefficient are automatically adjusted according to the system frequency deficit and the frequency change rate. One preferred calculation formula is:

[0024]

[0025] Where, J is the virtual inertia, D p is the damping coefficient, ω is the output angular frequency, and t is the time; J0 and D0 are the steady-state values of inertia and damping during steady-state operation, respectively, and k j ,k d They are virtual inertia J and damping coefficient D respectively p Dynamic adjustment coefficient; P ref is the reference active power, P e is the output active power; C j is the virtual inertia active power adjustment coefficient, C d is the damping coefficient active power regulation coefficient, P n is the initial active power threshold of the adaptive control strategy.

[0026] The establishment of a wind, solar and storage multi-source collaborative double-layer coupling mechanism and a multi-scenario optimization model also includes: an upper-layer wind, solar and storage capacity optimization configuration model to minimize the total investment cost C total With the expected operating cost as the optimization target, the decision variable is the capacity parameter, which is: photovoltaic capacity W pv 、Fan capacity W wt And energy storage capacity W bat , and is subject to the constraints of initial investment cost and operation and maintenance cost; the lower-level multi-scenario optimization model takes minimizing the weighted operation cost of multiple scenarios, maximizing the wind-solar coordination complementation rate (minimizing the wind-solar coordination complementation imbalance rate) and the system load power failure rate as the optimization objectives, and the decision variable is the real-time output P of the photovoltaic system pv , fan system real-time output P wt And energy storage charging and discharging power P ch / P dis , and is subject to power balance, equipment operating limits and dynamic frequency safety constraints; on this basis, the upper and lower layers are iteratively optimized, and the upper layer model passes the maximum capacity parameter to the lower layer model. The maximum capacity parameter multiplied by the system frequency serves as the boundary constraint of the real-time maximum power in the lower layer model.

[0027] The present invention proposes a virtual synchronous machine wind, solar and storage coordinated control system, including a memory, a processor and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements any one of the virtual synchronous machine wind, solar and storage coordinated control methods described.

[0028] Beneficial effects: Compared with the prior art, the present invention has the following six advantages, which are described in detail:

[0029] 1. Reconstruction and improvement of inertia support capabilities.

[0030] In traditional power systems, the rotational inertia of synchronous generator sets is the physical basis for maintaining the stability of the grid frequency. However, new energy power stations (such as wind farms and photovoltaic power stations) are connected to the grid through power electronic devices. Their output characteristics are essentially different from those of synchronous units, resulting in a significant reduction in the system's equivalent inertia. This problem is particularly prominent in "double-high" systems with a high proportion of renewable energy access. The core innovation of this invention lies in the construction of a hybrid inertia response mechanism for wind, solar and storage, which redefines the inertia support capacity of new energy power stations.

[0031] 1.1 Conversion of equivalent inertia of wind turbine rotor kinetic energy.

[0032] When a wind turbine is in operation, the mechanical kinetic energy stored in its rotating blades and generator rotor is not effectively utilized in traditional control strategies. The present invention uses overspeed load reduction control to actively adjust the wind turbine speed beyond the rated value at the moment the grid frequency disturbance occurs, converting the rotor kinetic energy into electromagnetic power output. This process is equivalent to the inertia response of a synchronous generator: when the system frequency drops, the stored kinetic energy is released to delay the frequency drop; when the frequency rises, the energy is restored through speed recovery. Theoretically, this mechanism breaks through the limitation that power electronic equipment can only provide "virtual inertia", realizes the coupling and superposition of physical inertia and virtual inertia, and enables new energy power stations to have real inertia support capabilities.

[0033] 1.2 Synergistic mechanism of multi-time-scale inertia response.

[0034] At different stages after the disturbance occurs, the system's demand for inertia support varies significantly:

[0035] The 0-5 second emergency response phase: During this period, the system frequency changes at its fastest rate, requiring rapid release of inertial energy to suppress sudden frequency fluctuations. The mechanical inertia response of wind turbines is inherently fast due to their physical properties, releasing kinetic energy within milliseconds, thus addressing the latency limitations of traditional virtual synchronous machines that rely on power electronics control.

[0036] After 5 seconds, the continuous adjustment phase begins: The short-lived nature of mechanical inertia makes it difficult to maintain long-term, necessitating the virtual synchronous control of the energy storage system to provide sustained inertia support. These two elements work together over time, forming a full-cycle inertia response chain for wind energy storage: "mechanical inertia buffering + continuous virtual inertia compensation." This theoretically covers the entire process from disturbance occurrence to frequency recovery.

[0037] 2. Collaborative optimization of frequency regulation at multiple time scales.

[0038] Power system frequency stability involves dynamic processes that occur on a timescale of seconds, minutes, and even hours. A single control strategy cannot meet the regulation requirements across the entire time domain. The proposed "hybrid energy storage hierarchical control" architecture achieves precise resolution of frequency issues through differentiated allocation of regulation resources.

[0039] 2.1 Supercapacitor energy storage has a fast response time of seconds.

[0040] Supercapacitors, with their high power density, are specifically designed to suppress high-frequency frequency fluctuations (e.g., df / dt ≥ 1 Hz / s). From a control theory perspective, they function like a "differential controller" in power systems: rapid charging and discharging offsets the acceleration of frequency fluctuations, reducing the peak ROCOF. This response, rather than relying on complex algorithms, is based on an instinctive reaction to physical properties. This theoretically avoids the vulnerability of the differential element in traditional PID control to noise.

[0041] 2.2 Minute-level power balance of lithium battery energy storage.

[0042] Lithium-ion batteries regulate low-frequency components (Δf ≥ 0.2Hz), and their control logic is closer to "integral control": by continuously adjusting the charge and discharge power, they gradually eliminate steady-state frequency deviations. Compared to the instantaneous operation of supercapacitors, lithium-ion batteries must comprehensively consider the sustainability of the state of charge. Therefore, their control strategy must embed state prediction and energy management algorithms to ensure that the SOC remains within a safe threshold during the regulation process, avoiding a sudden drop in regulation capability due to excessive discharge.

[0043] 3. Dynamic adaptive mechanism of virtual synchronous machine parameters.

[0044] In traditional virtual synchronous machine control, the virtual inertia coefficient (J) and the damping coefficient (D p ) is usually set to a fixed value, which makes it difficult to adapt to the dynamic changes in the power grid operation state. The three-dimensional fuzzy inference system proposed in this invention reconstructs the logical framework of parameter setting from a theoretical level.

[0045] 3.1 Multi-dimensional state perception and fuzzy processing.

[0046] The system monitors two key variables in real time:

[0047] Frequency change rate (df / dt): reflects the urgency of system inertia loss.

[0048] Energy storage SOC: characterizes the sustainable capacity of regulation resources.

[0049] Wind and solar power output volatility: indicates the uncertainty of primary energy supply.

[0050] These continuous variables are converted into fuzzy linguistic variables through membership functions, breaking the rigid boundaries of traditional threshold control and providing more practical mathematical tools for the description of nonlinear systems.

[0051] 3.2 Collaborative optimization logic of fuzzy rule base.

[0052] The established fuzzy rule base includes:

[0053] Rule 1: If the SOC is high and the frequency changes dramatically, increase the J value to enhance the inertia support and moderately reduce D p value to avoid over-damping.

[0054] Rule 2: If SOC is close to the lower limit and the fluctuation rate is low, reduce J value to maintain the energy storage capacity and increase D p The value suppresses low frequency oscillation.

[0055] These rules essentially encode expert experience into executable logical statements, and implement nonlinear mapping of "state-parameter" through a fuzzy inference engine, so that the virtual synchronous machine parameters can dynamically adapt to the current operating conditions.

[0056] 3.3 Physical meaning of the defuzzification process.

[0057] Use the centroid method to defuzzify and get accurate J and D p The physical meaning of value is:

[0058] Dynamic adjustment of the J value: This is equivalent to changing the mass of the virtual rotating body in real time according to the system strength. In a strong power grid, the inertia is reduced to reduce energy loss, and in a weak power grid, the inertia is increased to enhance stability.

[0059] D p Coordinated change of value: forming a complementary relationship with J value, when J value increases, D is appropriately reduced p value, to avoid the system from responding hysteresis due to excessive damping, otherwise increase D p value to suppress oscillation.

[0060] This mechanism can theoretically achieve the best matching between the virtual synchronous machine parameters and the dynamic characteristics of the power grid.

[0061] 4. Dynamic adaptive mechanism of virtual synchronous machine parameters.

[0062] By analyzing the step response of active power and angular frequency when setting different values of virtual inertia and damping coefficient, we can understand how the overshoot, stabilization time and peak time of active power and angular frequency change when the system frequency changes. Finally, a dynamic adaptive mechanism for virtual synchronous machine parameters is designed that takes into account the changes in system active power and frequency.

[0063] 5. Spontaneous dynamic response of the multi-source collaborative double-layer coupling optimization model.

[0064] The coordination of wind, solar and storage multi-source systems must meet both technical performance and economic requirements. The "double-layer coupling + multi-scenario optimization" model proposed in this invention theoretically constructs a path to achieve multi-dimensional target collaboration.

[0065] 5.1 Two-way feedback between upper-layer capacity configuration and lower-layer operation optimization.

[0066] The upper-level model determines the installed capacity of wind, solar, and storage with the goal of minimizing investment costs. Its theoretical innovation lies in incorporating the expected operating costs of the lower-level model into the optimization objective, rather than the static capacity planning of the traditional single-level model. This design considers the impact of capacity allocation on long-term operating economics, avoiding the subsequent surge in regulation costs caused by excessive pursuit of initial investment savings.

[0067] The lower-level model optimizes real-time output with the goal of minimizing regulation costs under given capacity constraints. Its breakthrough lies in the introduction of multiple scenarios, fully considering the output of wind, solar, and storage in different scenarios.

[0068] 5.2 Game equilibrium of wind, solar, and storage output coupling.

[0069] During the regulation process, the output distribution of wind turbines, photovoltaics and energy storage must meet the following requirements:

[0070] Technical constraints: The release of wind turbine rotor kinetic energy may affect maximum power point tracking, and photovoltaic power generation needs to reserve a regulation margin.

[0071] Economic constraints: Frequent adjustment of energy storage will accelerate capacity decay and increase the cost of the entire life cycle.

[0072] The model introduces a Nash bargaining game framework, using the differences in regulation costs among units as bargaining power weights to ultimately achieve a Pareto optimal solution. This theoretical approach avoids the subjectivity of traditional weighted summation methods and ensures that output allocation meets both technical requirements and economic rationality.

[0073] 6. Theoretical leap in overall system performance.

[0074] Based on the above technological innovations, the solution of the present invention theoretically achieves a paradigm shift from "passive grid connection" to "active support" for new energy power stations, which is specifically manifested in the following aspects:

[0075] 6.1 Structural Restructuring of Frequency Stability.

[0076] Through the coordination of mechanical inertia and virtual inertia, and the integration of multi-time-scale regulation resources, the system frequency response characteristics have undergone fundamental changes:

[0077] Inertial response stage: The physical inertia support of the wind turbine fills the initial delay of the power electronic equipment response and significantly reduces the frequency drop speed.

[0078] Primary frequency regulation stage: The rapid power injection of energy storage shortens the time to reach the lowest frequency point.

[0079] Secondary frequency regulation stage: Energy storage supplements the power lacking in primary frequency regulation through the energy storage of lithium battery packs over a long period of time.

[0080] 6.2 Economic optimization throughout the entire life cycle.

[0081] The two-layer optimization model breaks through the limitations of traditional "segmented optimization":

[0082] Capacity-operation coupling: Avoid local optimality caused by optimizing configuration or operation separately, and minimize global costs.

[0083] Extended equipment life: Dynamic parameter optimization reduces ineffective charge and discharge cycles of energy storage and reduces the capacity attenuation rate.

[0084] In summary, this invention constructs a comprehensive theoretical framework encompassing four dimensions: inertia reconstruction, time-scale coordination, parameter adaptation, and multi-source optimization. This not only addresses the frequency stability challenges associated with high renewable energy integration, but also provides a new methodological foundation for dynamic characteristic analysis and control strategy design for novel power systems. This invention effectively addresses the "low inertia, weak damping" technical bottleneck of traditional renewable energy power plants, significantly enhancing the frequency stability and dynamic regulation capabilities of power systems with high wind and photovoltaic power levels under grid fault conditions, providing crucial technical support for the development of novel power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 It is the block diagram of energy storage hierarchical control;

[0086] Figure 2 This is a diagram of the virtual inertia adjustment structure based on fuzzy control according to the present invention;

[0087] Figure 3 It is the photovoltaic VSG topology diagram;

[0088] Figure 4 is the step response of active power when J changes according to the present invention;

[0089] Figure 5 is the step response of the angular frequency when J changes according to the present invention;

[0090] Figure 6 It is the principle of spontaneous dynamic response of virtual inertia and damping coefficient described in the present invention;

[0091] Figure 7This is a block diagram of the wind, solar and storage dual-layer multi-scenario optimization model described in the present invention;

[0092] Figure 8 It is the double-layer model decoupling flow chart of the present invention;

[0093] Figure 9 It is the overall system framework structure diagram of the present invention. DETAILED DESCRIPTION

[0094] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0095] The virtual synchronous generator wind, solar, and storage coordinated control method described in this invention fully utilizes the potential for coordinated regulation of multiple resources through coordinated control of wind and solar power stations and storage power stations, effectively improving system inertia safety under different time scales and operating conditions. The specific control method includes the following steps:

[0096] 1) The wind farm provides inertial support in the early stages of frequency response by releasing rotor kinetic energy in combination with overspeed load reduction control;

[0097] 2) The energy storage system provides fast frequency response and long-term power balancing support at different time scales. Supercapacitor energy storage is responsible for fast frequency regulation within seconds, while battery energy storage is responsible for continuous power balancing support within minutes.

[0098] 3) A fuzzy logic-based dynamic optimization algorithm for virtual synchronous machine parameters monitors the state of charge of the energy storage system and the rate of change of the grid frequency in real time, and dynamically adjusts the virtual inertia coefficient through a two-dimensional fuzzy inference system;

[0099] 4) Combined with the spontaneous dynamic response of the photovoltaic and storage power station based on the virtual inertia coordinated damping coefficient, it is explained how the photovoltaic and storage power station coordinates with the wind and storage to provide frequency regulation support under different operating conditions, fully tapping the regulation potential of regional wind, photovoltaic and storage power stations;

[0100] 5) Establish a wind, solar and storage multi-source collaborative double-layer coupling mechanism and a multi-scenario CVaR optimization model, where the upper-layer model is the wind, solar and storage capacity optimization configuration model, and the lower-layer model multi-scenario optimization model is coupled with each other using the upper-layer objective function output as a constraint condition.

[0101] The wind farm rotor kinetic energy release combined with overspeed load reduction control is:

[0102] The rotor inertia controls the rotor speed by controlling the rotor-side converter, which absorbs / releases the kinetic energy stored in the wind turbine in a short time. During normal operation, the kinetic energy stored in the wind turbine rotor is expressed as:

[0103]

[0104] Where, Ew Kinetic energy stored in the wind turbine rotor; J w is the equivalent inertia of the wind turbine; ω r is the wind turbine speed.

[0105] When the disturbance frequency of the system decreases, the electromagnetic power released by the fan unit through the rotor kinetic energy is expressed as:

[0106]

[0107] However, simply releasing the rotor kinetic energy will cause the wind turbine unit to continuously drop in rotor speed due to the output power exceeding the input power for a long time when the power disturbance is large. When the rotor speed reaches the speed set by the protection device, the protection device will be activated and the wind turbine will exit the frequency regulation, causing the frequency to drop further. To avoid this situation, overspeed load reduction control is considered. Preferably, the output power expression of the wind turbine unit at this time is:

[0108]

[0109] Where, P out is the output power of the wind turbine under overspeed load reduction control; P max is the maximum mechanical power of the wind turbine during operation; ΔP d is the output power variation of the wind turbine under overspeed load reduction control; C P,out (λ,β) is the wind energy utilization coefficient under overspeed load reduction control; C P,max (λ,β) is the wind energy utilization coefficient when the wind turbine outputs the maximum mechanical power; ρ is the air density; r is the rotor radius; v is the wind speed; and d is the load reduction rate.

[0110] A hybrid energy storage hierarchical architecture is used to achieve power balance support on different time scales through modular energy storage and its control strategy. The specific control method includes the following steps:

[0111] A complete electrochemical energy storage system mainly consists of battery packs, a battery management system (BMS), an energy management system (EMS), a power conversion system (PCS), and other electrical equipment. First, the energy storage system is divided into layers and grades, and energy storage at different levels responds to demand through autonomous coordination. First, the bottom-level control uses high-frequency sampling and real-time control algorithms to quickly respond to the frequency in milliseconds to seconds. The second-level control uses model predictive control (MPC) to predict and optimize the short-term grid status and respond to the system frequency at a medium speed of seconds to minutes. The third-level control uses high-level control to predict long-term loads and plan the overall charging and discharging strategy of the energy storage system to cope with long-term load changes and fluctuations in renewable energy.

[0112] Based on the above, different types of energy storage can be divided into different modules according to different levels of control objectives. Figure 1 To achieve fast response of the bottom-level control, supercapacitors with high power density and fast charge and discharge are considered as energy storage devices for the fast response module; to achieve mid-level and high-level control, high energy density energy storage devices can be selected, and long-term energy storage modules can be used for long-term power balance.

[0113] Taking into account the state of charge of the energy storage and the rate of change of the grid frequency, the adaptive adjustment of the virtual inertia is achieved through fuzzy processing, so that the virtual inertia can be adjusted dynamically. The specific control method includes the following steps:

[0114] First, the working state of the energy storage system is divided into the following categories according to different states of charge:

[0115]

[0116] Where a is the discharge limit value of the energy storage battery, and b is the charging limit value of the energy storage battery.

[0117] The traditional VSG control strategy does not consider the state of charge of the energy storage. According to the rotor motion equation of the synchronous generator, its active power-frequency control can be expressed as:

[0118]

[0119] Where, P refis the reference active power; P is the actual active power output of the VSG; D is the damping coefficient; H is the virtual inertia; ω and ω0 are the angular frequencies of the VSG unit output and the grid, respectively. Equation (5) shows that its virtual inertia coefficient is a constant. When a large power disturbance occurs, the energy storage battery pack will over-discharge, which will seriously affect its service life and system stability. Because the inverse tangent function can effectively limit the output to a certain range, this paper considers the state of charge of the energy storage and combines the inverse tangent function to automatically adjust the virtual inertia H.

[0120]

[0121] Where H0 is the inertia constant when the charge state is normal; k1 is the virtual adjustment coefficient; and k2 is the virtual inertia control parameter.

[0122] This paper adopts fuzzy control, and obtains fuzzy sets by fuzzifying the input variables SOC and ROCOF through membership functions. Then, by establishing a fuzzy rule base and fuzzy reasoning, the precise control quantity k1 is finally defuzzified. Figure 3 shown.

[0123] Based on the different operating conditions of the state of charge and the different scenarios of the system frequency change rate, triangular, trapezoidal, and Z-shaped membership functions can be used to fuzzify them into fuzzy sets of {negative large (NB), negative small (NL), zero (ZO), positive small (PS), positive large (PL)}, with a value range of [0, 1]. The established fuzzy rule base is shown in Table 1. When the SOC is near depletion, the SOC protection principle takes precedence over the frequency regulation principle. At this time, if the system frequency drops suddenly, the SOC is restricted from discharging the VSG. If the system frequency rises, the SOC absorbs excess system power to restore system frequency stability and adjust the SOC back to the intermediate state. When the SOC is near full, the frequency regulation rule has the highest weight. At this time, if the system frequency drops suddenly, the SOC is quickly adjusted to discharge the system to replenish the system active power shortage and adjust the SOC back to the intermediate state. If the system frequency rises, the VSG is restricted from charging the SOC. The center of gravity method is used for defuzzification, and the output virtual adjustment coefficient k1 is also in the range of [0,1], using triangle and Z-type membership functions. The virtual inertia adjustment structure based on fuzzy control is shown in the figure below: Figure 2 shown.

[0124] Table 1 Fuzzy rule base

[0125]

[0126] The photovoltaic power station achieves power balance and optimized distribution by automatically responding to and coordinating the real-time output of wind power and photovoltaic power generation with the dynamic charging and discharging of the energy storage system through virtual inertia coordination damping coefficient. Figure 3As shown in Figure 1, the photovoltaic VSG samples the grid current and voltage in real time by simulating the characteristics of the synchronous generator, and determines the load power disturbance by calculating the active power and reactive power. The system frequency change can be further obtained to adjust the virtual inertia and damping coefficient in real time, thereby improving the frequency response characteristics of the system. Using the second-order synchronous machine model to model it, the power and power angle equations can be obtained as shown in Equation (7):

[0127]

[0128] Where: J is the virtual moment of inertia; ω and ω0 are the output angular frequency and the system rated angular frequency respectively; δ is the VSG power angle; D p is the damping coefficient.

[0129] Since the filter reactance is much larger than the filter resistance in actual engineering, the effect of the filter resistance on the Figure 4 The simplified active power output can be obtained by simplifying the topological diagram, as shown in formula (8):

[0130]

[0131] Where: X is the filter reactance, E is the electromotive force amplitude obtained by reactive voltage modulation; U1 is the voltage at the output of VSG.

[0132] Combining equations (7) and (8) to analyze, we can get ref to P e The transfer function is:

[0133]

[0134] From P ref The transfer function to ω is:

[0135]

[0136] Performing an inverse Laplace transform on equation (9) yields equation (11):

[0137]

[0138] According to formula (11), the step response of active power when J changes can be obtained as follows: Figure 4 shown.

[0139] The step response of angular frequency when J changes is as follows: Figure 5 As shown, observe Figure 5 It can be found that when the damping coefficient is kept constant and the virtual inertia is increased, the step response stabilization time t of the active power is s will increase, the peak time t p Will increase, overshoot M pwill also increase. Considering the error coefficient and performing quantitative calculation, we can get formula (12):

[0140]

[0141] It can be seen that under the condition of keeping the damping coefficient constant, t s It will increase as J increases, that is, the larger the moment of inertia value is set, the longer the active power stabilization time will be.

[0142] Similarly, performing the inverse Laplace transform on equation (10) yields equation (13):

[0143]

[0144] According to formula (13), the step response of angular frequency when J changes can be obtained as follows: Figure 5 shown.

[0145] like Figure 6 The following shows the principle of spontaneous dynamic response of virtual inertia and damping coefficient. Figure 6 It can be found that when the damping coefficient is kept constant and the virtual inertia is increased, the step response stabilization time t of the angular frequency is s will increase, the angular frequency peak time will also increase, and the overshoot M p Will decrease.

[0146] Based on the above analysis, we can draw the following conclusions: when the virtual inertia J increases, the stability of the active power output will be suppressed, but the overshoot of the frequency response can be reduced, making the frequency response more stable.

[0147] Similarly, the analysis of the damping coefficient can be concluded as follows: Damping coefficient D p When it increases, the active power and angular frequency response transition to a stable state is smoother, but the transition time will increase.

[0148] Therefore, based on the above analysis, this paper proposes a regulation strategy for the spontaneous dynamic response of photovoltaic power stations by coordinating the damping coefficient through virtual inertia:

[0149] (1) When the system is in steady-state operation, the virtual inertia and damping coefficient remain at the basic values of steady-state operation;

[0150] (2) When power is insufficient or the frequency changes rapidly, the virtual inertia and damping coefficient are adjusted according to the different frequency change states. When Δω>0 and dw / dt increases rapidly and is greater than 0, the virtual inertia and damping coefficient are increased simultaneously to suppress the frequency change rate and offset. When Δω>0 and dw / dt decreases rapidly and is less than 0, the virtual inertia is appropriately reduced to shorten the stabilization time of the frequency step response, but the damping coefficient is kept increasing to suppress overshoot and smooth the frequency transition. The subsequent oscillation state is similar to that described above and will not be repeated here.

[0151] (3) When the output power stabilizes at the reference value, J and D p also returns to the steady-state value.

[0152] According to the above analysis, there are many specific collaborative strategies. The preferred ones are shown in formulas (14) and (15):

[0153]

[0154]

[0155] Where J0 and D0 are the steady-state values of inertia and damping during steady-state operation, respectively, and k j ,k d J and D respectively p The dynamic adjustment coefficient of the control principle is as follows. Figure 6 shown.

[0156] Aiming at the problem of coordinated optimization of wind-solar-storage combined system in frequency regulation, in order to give full play to the regulation potential of wind-solar-storage power station, this paper proposes a multi-scenario dual-layer coupling optimization configuration of wind-solar-storage. The model configuration diagram is shown as follows: Figure 7 Preferably, the objective function of the upper model is:

[0157]

[0158] Where C total is the total investment cost of the system; C wt is the wind turbine installation cost per unit capacity; C pv is the photovoltaic installation cost per unit capacity; C bat is the energy storage installation cost per unit capacity; W wt is the installed capacity of the wind turbine; W pv is the photovoltaic installed capacity; W bat is the installed capacity of energy storage; ρ s is the scene probability; C s operation is one of the objective functions of the lower model.

[0159] Preferably, the objective function of the lower model is:

[0160]

[0161] Where C s operation is the weighted operating cost of multiple scenarios, α, β, γ are weight factors, P wt (t), P pv (t), P ch (t), P dis (t), P L(t), P bat (t) are the power generation of the wind turbine system at time t, the power generation of the photovoltaic system at time t, the charging capacity of the energy storage system at time t, the discharging capacity of the energy storage system at time t, the load at time t, and the power generation of the energy storage system at time t, respectively. is the average load. cp is the system wind-solar complementary imbalance rate, η LP is the system load power failure rate.

[0162] For the upper model, the capacity of each wind, solar and storage subsystem should be less than the maximum system cost, as shown in formula (20):

[0163] 0≤W wt ≤W wt,max

[0164] 0≤W pv ≤W pv,max

[0165] 0≤W bat ≤W bat,max (20)

[0166] For the lower model, the real-time maximum power it generates is approximately limited by the installed capacity of the upper model:

[0167] 0≤P wt (t)≤W wt,max f wt

[0168] 0≤P pv (t)≤W pv,max f pv

[0169] 0≤P bat (t)≤W bat,max f bat (twenty one)

[0170] At the same time, in order to protect the health of the energy storage battery, the battery status should be charged and discharged strictly in accordance with the method described above, and the upper and lower limits should be strictly set. wt , f pv , f bat They are wind turbine system frequency, photovoltaic system frequency, and energy storage system frequency respectively.

[0171] SOC min ≤SOC(t)≤SOC max (twenty two)

[0172] In addition, to ensure power supply reliability, the dynamic frequency security constraints include:

[0173] |Δf|≤0.5Hz (23)

[0174] The decoupling flow chart of the two-layer model is as follows Figure 8 As shown, first initialize the system capacity configuration, determine the corresponding scenario based on the capacity configuration, and then call the lower-level optimization to perform the optimization according to different scenarios. Finally, the optimization results are judged for convergence and divergence. If converged, the optimal configuration is output. If not, the lower-level optimization is called again until the convergence judgment is passed. In this example, the convergence condition is |C (k+1) total -C k total |≤ε.

[0175] Figure 9 This is the overall system framework structure diagram of this patent. The system first establishes a database of different operating states. In steady state, the power system adjusts the output of wind turbines and photovoltaics and reserves a certain amount of active power for standby. At the same time, it adjusts the charge and discharge of energy storage to cope with the insufficient inertia after the system is disturbed. When the system is disturbed and enters a new operating state, the dispatching center first calculates the active power imbalance in this state, and then takes emergency response measures based on the established operating state database: the wind turbine releases the rotor kinetic energy, and the photovoltaics spontaneously and dynamically respond to the active power shortage through virtual inertia and damping coefficient. After the emergency response measures, the power shortage is calculated again to see if it is within the allowable range. If it is still beyond the range, the load is appropriately cut to ensure the stability of the system frequency.

[0176] In one embodiment, a virtual synchronous machine wind, solar and storage coordinated control system is proposed, comprising a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, any one of the virtual synchronous machine wind, solar and storage coordinated control methods is implemented.

[0177] The present invention addresses the issues of insufficient system inertia and frequency safety caused by the access of a high proportion of renewable energy to the power grid ("double-high" power system). By coordinating the control of wind power, photovoltaic power, and energy storage, the present invention fully taps the potential for regulation and control of regional wind, photovoltaic, and energy storage power stations, effectively improving the system's inertia level and frequency stability.

Claims

1. A virtual synchronous machine wind, solar and energy storage coordinated control method, characterized in that: include: The wind farm provides inertial support in the early stages of frequency response by releasing rotor kinetic energy in conjunction with overspeed load reduction control. The early stages are the 0-5 seconds period after a grid frequency disturbance occurs. The energy storage system provides frequency response in a short time scale and power balance support in a long time scale at different time scales; the short time scale is milliseconds to seconds; the long time scale is minutes; According to the real-time monitoring of the energy storage system's state of charge and the rate of change of the grid frequency, the system's virtual inertia is dynamically adjusted through two-dimensional fuzzy control; The PV power station uses virtual inertia to coordinate the damping coefficient for spontaneous dynamic response, including: automatic adjustment of virtual inertia and damping coefficient according to the system frequency deficit and frequency change rate; A wind, solar and storage multi-source collaborative double-layer coupling mechanism and a multi-scenario optimization model are established, including an upper-layer model and a lower-layer model. The upper-layer model is a wind, solar and storage capacity optimization configuration model, and the lower-layer model is a multi-scenario optimization model. The lower-layer model feeds back the multi-scenario weighted operating cost to the upper-layer model to update the objective function of the upper-layer model.

2. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 1 is characterized by: The rotor kinetic energy release combined with overspeed load reduction control includes: maintaining the rotor speed higher than the rated speed by adjusting the pitch angle of the wind turbine and the generator torque command; reserving rotor kinetic energy for release, releasing the rotor kinetic energy of the wind turbine when the frequency drops to increase the system inertia; automatically reducing the load when the frequency exceeds a set threshold, so that the wind turbine operating output curve remains above the original load reduction operating output curve, and the wind turbine output power is kept higher than the output power before the frequency drops. When the system frequency stabilizes within the allowable range, the overspeed load reduction stops.

3. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 2 is characterized in that: The output power expression of the wind turbine generator set under the overspeed load reduction control is: Where, P out is the output power of the wind turbine under overspeed load reduction control; P max is the maximum mechanical power of the wind turbine when it is running; ΔP d is the output power variation of the wind turbine under overspeed load reduction control; C P,out (λ,β) is the wind energy utilization coefficient under overspeed load reduction control; C P,max (λ,β) is the wind energy utilization coefficient when the wind turbine outputs the maximum mechanical power; ρ is the air density; r is the rotor radius; v is the wind speed; and d is the load reduction rate.

4. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 1, characterized in that: The energy storage system provides frequency response within a short time scale and power balance support within a long time scale at different time scales, including: supercapacitor energy storage undertakes the frequency regulation task within a short time scale by participating in the initial frequency regulation; battery pack energy storage is responsible for power balance support within a long time scale by adjusting power output.

5. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 1 is characterized in that: The method of dynamically adjusting the virtual inertia of the system through two-dimensional fuzzy control based on the real-time monitored state of charge of the energy storage system and the rate of change of the grid frequency includes: using a two-dimensional fuzzy control method to output a virtual adjustment coefficient of the virtual synchronous machine interface based on the state of charge of the energy storage system and the rate of change of the grid frequency, and dynamically adjusting the virtual inertia of the system based on the virtual adjustment coefficient.

6. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 5, characterized in that: The virtual inertia of the system is dynamically adjusted according to the virtual adjustment coefficient, and the calculation formula is: Where H is the virtual inertia, H0 is the inertia constant when the state of charge is normal; k1 is the virtual adjustment coefficient; k2 is the virtual inertia control parameter, SOC is the state of charge of the energy storage system, a is the discharge limit value of the energy storage battery, and b is the charging limit value of the energy storage battery.

7. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 1, characterized in that: The automatic adjustment of virtual inertia and damping coefficient according to the system frequency deficit and frequency change rate includes: When the system is running in steady state, the virtual inertia and damping coefficient remain at the basic values of steady state operation; When Δω>0 and dw / dt increases and is greater than 0, both the virtual inertia and the damping coefficient are increased to suppress the frequency change rate and offset. When Δω>0 and dw / dt decreases and is less than 0, the virtual inertia is reduced to shorten the settling time of the frequency step response, but the damping coefficient is kept increased to suppress overshoot. Δω is the system frequency deficit, and dw / dt is the frequency change rate. When the output power stabilizes at the reference value, the virtual inertia and damping coefficient return to the steady-state value.

8. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 7 is characterized in that: The virtual inertia and damping coefficient are automatically adjusted according to the system frequency shortage and the frequency change rate. The calculation formula is: Where, J is the virtual inertia, D p is the damping coefficient, ω is the output angular frequency, and t is the time; J0 and D0 are the steady-state values of inertia and damping during steady-state operation, respectively, and k j ,k d They are virtual inertia J and damping coefficient D respectively p Dynamic adjustment coefficient; P ref is the reference active power, P e is the output active power; C j is the virtual inertia active power adjustment coefficient, C d is the damping coefficient active power regulation coefficient, P n is the initial threshold of active power of the adaptive control strategy.

9. The virtual synchronous machine wind-solar-storage coordinated control method according to claim 1, characterized in that: The establishment of a wind, solar and storage multi-source collaborative double-layer coupling mechanism and a multi-scenario optimization model also includes: an upper-layer wind, solar and storage capacity optimization configuration model to minimize the total investment cost C total With the expected operating cost as the optimization target, the decision variable is the capacity parameter, which is: photovoltaic capacity W pv , fan capacity W wt And energy storage capacity W bat , and is subject to the constraints of initial investment cost and operation and maintenance cost; the lower-level multi-scenario optimization model takes minimizing the weighted operation cost of multiple scenarios, minimizing the wind-solar coordinated complementary imbalance rate and the system load power failure rate as the optimization objectives, and the decision variable is the real-time output P of the photovoltaic system pv , fan system real-time output P wt And energy storage charging and discharging power P ch / P dis , and is subject to power balance, equipment operating limits and dynamic frequency safety constraints; on this basis, the upper and lower layers are iteratively optimized, and the upper layer model passes the maximum capacity parameter to the lower layer model. The maximum capacity parameter multiplied by the system frequency serves as the boundary constraint of the real-time maximum power in the lower layer model.

10. A virtual synchronous machine wind, solar and energy storage coordinated control system, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the virtual synchronous machine wind-solar-storage coordinated control method according to any one of claims 1 to 9 is implemented.

Images