A distributed new energy cluster collaborative network construction method based on fuzzy control

Abstract

The application belongs to the field of new energy power generation and grid-connected control technology, and relates to a distributed new energy cluster collaborative network construction method based on fuzzy control. n The application belongs to the field of new energy power generation and grid-connected control technology, and relates to a distributed new energy cluster collaborative network construction method based on fuzzy control. S1: constructing a photovoltaic cluster grid-connected equivalent model containing a distributed power supply, performing Taylor expansion and linearization, and obtaining a small signal dynamic equation of power output; S2: constructing a small signal model of virtual synchronization active-frequency and reactive-voltage control of the photovoltaic cluster considering power coupling; S3: determining the influence law of virtual inertia, damping coefficient and key parameters of device dynamic characteristics on the frequency and voltage response process; S4: introducing a fuzzy logic supervision layer, taking the absolute value of the frequency change rate of the grid-connected point, the absolute value of the frequency deviation and the absolute value of the voltage deviation as inputs, taking the normalized adjustment coefficient as the output of the fuzzy controller, and mapping the normalized adjustment coefficient to the actual virtual inertia parameter and the damping parameter, so as to enhance the dynamic adaptive ability of the system to random load disturbance and coupling working condition change.

Description

Technical Field

[0001] This invention belongs to the field of new energy power generation and grid connection control technology, specifically involving a distributed new energy cluster collaborative grid construction method based on fuzzy control. Background Technology

[0002] Distributed photovoltaic (PV) power generation has become an important force in energy transition. However, the large-scale integration of distributed PV also poses serious challenges to system operational security: on the one hand, it may cause power reverse flow in the distribution network and local voltage exceedances, thereby threatening power supply security; on the other hand, it may weaken the system's overall inertia and frequency regulation capabilities, resulting in insufficient system immunity. Therefore, exploring and utilizing the flexible grid support capabilities of distributed PV is of great significance for ensuring the safety of local voltage in the distribution network and the stability of the system frequency.

[0003] Regarding distributed power generation cluster control methods, in terms of system control, to address the difficulty of centralized scheduling of massive distributed photovoltaic systems, the literature... [1] This paper proposes a hierarchical cluster coordination control architecture for distributed photovoltaic (PV) power generation, which aggregates physically adjacent or electrically similar distributed power sources into controllable units and employs reasonable cluster partitioning. However, it faces challenges related to frequency support technology for distributed PV power grids under uncertainties. (References...) [2] A cluster partitioning strategy based on an improved community algorithm is proposed, which optimizes voltage control by comprehensively considering power balance and node coupling; [Reference] [3] This introduces a comprehensive electrical distance modularity index, enabling proactive zonal control of the distribution network while taking into account source-load uncertainties. (Reference) [4] A predictive control method for active power stratification of wind power clusters considering multi-temporal and spatial scale coordination is proposed, which realizes the rational allocation of wind power and improves the wind power absorption capacity. However, most of the above studies focus on economic dispatch and spatiotemporal coordination and complementary operation control of different equipment inside and outside the cluster, and pay insufficient attention to the dynamic adjustment and recovery of equipment within the cluster. They also do not fully discuss the safety issues such as instantaneous overvoltage and overcurrent that may be induced by frequent photovoltaic fluctuations.

[0004] In inverter control, a large body of literature has conducted in-depth research on the participation of distributed photovoltaic (PV) systems in distribution network voltage regulation. The main methods of PV participation include reactive power-voltage regulation, active power reduction control, and active power-reactive power coordinated control. Besides participating in distribution network voltage regulation, distributed PV can also provide frequency support to the system by rapidly adjusting its active power output. By reducing a portion of active power for reserve, distributed PV can respond to system frequency deviations in real time based on control schemes such as droop, virtual inertia, and virtual synchronous machines, providing instantaneous power support and thus reducing system active power imbalance after disturbances. Some studies also focus on the aggregated control of distributed PV, improving the overall frequency support effect by centrally optimizing parameters such as droop gain and virtual inertia.

[0005] Currently, with the continuous integration of distributed power sources into medium and low voltage distribution networks, the inertia and damping characteristics of the distribution network, especially at the low-voltage end, are reduced, resulting in a poorer ability to cope with power fluctuations and severe power coupling. To improve its operating characteristics, virtual synchronous generator control structures and power feedback decoupling designs have been proposed and applied. (Literature) [5] By analogy with the equations of a virtual synchronous generator, the controllable photovoltaic and energy storage within a photovoltaic cluster are modeled as a whole. Inertial equivalence and dynamic feedback are achieved through control design and power allocation. (Reference) [6] By analyzing the power coupling characteristics under low-voltage conditions, a decoupling control method corresponding to droop control is proposed. (Reference) [7] Based on the virtual synchronous generator, further considering the power coupling effect, the low-voltage operation stability of the virtual synchronous generator is improved through feedback decoupling design. The above is an application at the inverter control level. At the level of coordinated control among multiple converters, the main modes currently used are master-slave control, peer control, and hierarchical control. The master control converter unit typically uses V / F control as a stable control unit for the voltage and frequency of the parallel system, and the output current of the master control unit serves as a reference quantity for the slave control units. Peer control distributes and equalizes the control of the parallel system; each converter in the system uses the output power of other generating units as a reference signal, and after parsing the power signal, obtains its own unit's voltage reference signal. The hierarchical control mode has a central controller in the multi-parallel converter system to coordinate and control the entire system.

[0006] However, when the system operating conditions change with load switching, local voltage disturbances and changes in the dynamic state of equipment, a set of fixed inertia parameters obtained by offline tuning is often insufficient to cover all operating conditions, resulting in insufficient adaptive capability of the controller.

[0007] References: [1] Wan Can, Qin Hongpei, He Zihan, et al. Frequency-voltage coordinated support control for large-scale distributed photovoltaic hierarchical clusters [J]. Proceedings of the CSEE, 2025 [2] Ge Jinming, Liu Yingru, Pang Dan, et al. Cluster partitioning voltage control strategy for distribution networks with high photovoltaic penetration [J]. High Voltage Engineering, 2024, 50(1): 74-82. [3] Li Xiaobo, Xie Jianfeng, Zhang Shile. Robust two-stage control strategy for voltage and reactive power in active distribution network based on cluster partitioning [J]. Proceedings of the CSEE, 2025, 45(18): 7058-7071. [4] Ye Lin, Zhang Cihang, Tang Yong, et al. Active power hierarchical predictive control method for wind power clusters with multi-temporal and spatial scale coordination [J]. Proceedings of the CSEE, 2018, 38(13): 3767-3780. [5] Zhang Congyue, Dou Xiaobo, Sheng Wanxing, et al. Robust virtual synchronization control method for distributed photovoltaic clusters [J]. Proceedings of the CSEE, 2020, 40(02): 510-522. [6] Wang Changyun, Li Chunlan, Shi Zhai, et al. Research on droop control optimization of low voltage microgrid based on decoupling theory [J]. Renewable Energy, 2017, 35(2): 264-270. [7] Li Mingxuan, Wang Yue, Xu Ningyi, et al. Power decoupling strategy for virtual synchronous generators with relaxed small power angle constraints [J]. Automation of Electric Power Systems, 2018, 42(9): 59-68. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention proposes a distributed new energy cluster collaborative network construction method based on fuzzy control. This method introduces the concept of a virtual synchronous generator into the multi-converter collaborative control layer. By constructing a power feedback loop, an equivalent model of the system inertia is established, enabling the system to exhibit synchronous machine-like inertial and damping characteristics, thereby enhancing the dynamic support capability of the multi-converter system. Furthermore, to overcome the insufficient adaptability of fixed virtual inertia parameters under different disturbance conditions, a fuzzy logic supervision layer is designed. Using the frequency change rate, frequency deviation, and voltage deviation at the grid connection point as inputs, the virtual inertia and damping parameters are adaptively tuned online, thereby enhancing the controller's adaptability to random load fluctuations and coupled disturbances.

[0009] The technical solution adopted in this invention is a distributed new energy cluster collaborative network construction method based on fuzzy control, comprising the following steps: S1: Construct an equivalent model of a photovoltaic cluster connected to the grid with n distributed power sources, calculate the output power of each distributed power source, perform Taylor expansion and linearization on the power to obtain the small-signal equation of the power output; S2: Construct a small-signal model for virtual synchronization of active-frequency and reactive-voltage control of photovoltaic clusters, considering power coupling. S3: Based on the small-signal model, the influence of virtual inertia, damping coefficient and key parameters of equipment dynamic characteristics on the frequency and voltage response process was clarified; S4: Based on the aforementioned influence law, a fuzzy logic supervision layer is introduced, using the absolute value of the frequency change rate at the grid connection point, the absolute value of the frequency deviation, and the absolute value of the voltage deviation as inputs, and the normalized adjustment coefficient λ as the output of the fuzzy controller, and mapping the normalized adjustment coefficient λ to the actual virtual inertia parameter and damping parameter.

[0010] Furthermore, S1 specifically refers to: The output power of each distributed power source is: ; In the formula, , These are the active power and reactive power outputs of the distributed power source, respectively. and The equivalent resistance and inductive reactance of the line; This is the value of the work angle; This refers to the amplitude of the common bus voltage. Let be the output voltage amplitude of the i-th DG unit; The small-signal equation for output power is: ; ; In the formula, and These represent the changes in the active and reactive power outputs of the distributed power source, respectively. Let be the output voltage amplitude of the i-th DG unit; The change in phase angle of the output voltage of the i-th DG unit; This represents the change in the output voltage amplitude of the i-th DG unit; , , , This is the power coupling gain coefficient.

[0011] Furthermore, S2 specifically includes: The small-signal model for virtual synchronization of active power and frequency control in photovoltaic clusters is as follows: ; ; ; ; In the formula, This represents the change in the system's angular velocity. This is a reference value for angular velocity; and These represent the changes in active power output of m energy storage units and n photovoltaic units, respectively. This refers to uncontrollable disturbances within and outside the system, including internal and external load disturbances, line losses, and external angular frequency variations. This represents the change in active power. This is the angular velocity droop gain coefficient; and These are the cluster's virtual inertia and damping coefficient, respectively. The active power output of m energy storage units within the cluster; and These represent the changes in active power input for the j-th energy storage device and the i-th photovoltaic device, respectively. and For the dynamic transmission characteristics of input and output of energy storage and photovoltaic equipment; The small-signal model for virtual synchronization of reactive power-voltage control in photovoltaic clusters is as follows: ; ; ; ; In the formula, The change in output voltage amplitude; and These are the integral gain coefficient and the proportional gain coefficient, respectively. and These represent the changes in total reactive power output for controllable energy storage and photovoltaics, respectively. This represents the change in reactive power. This refers to the uncontrollable reactive power disturbance changes inside and outside the cluster. and These represent the changes in reactive power input for the i-th photovoltaic and j-th energy storage devices, respectively.

[0012] Furthermore, S3 specifically refers to: The phase response function of cluster virtual synchronization in grid-connected mode is: ; assumed H PV = 1 / ( T PV s +1) Simplifies the input-output transmission characteristics of photovoltaic devices. H Bess = 1 / ( T Bess s+1) To simplify the input-output transmission characteristics of energy storage devices, the phase response function of cluster virtual synchronization is simplified to: ; When power disturbances occur during grid connection, the disturbance will be shared by the external power grid and the photovoltaic cluster, and the power balance relationship is as follows: ; ; Based on the reactive-voltage control small-signal model, the system voltage response function is obtained as follows: ; In the formula, For the dynamic time of photovoltaic equipment; This refers to the dynamic time of energy storage devices.

[0013] Furthermore, S4 specifically includes: The absolute values ​​of the rate of change of frequency at the grid connection point, frequency deviation, and voltage deviation are used as input variables for the fuzzy controller and are defined as follows: ; ; ; In the formula, and These represent the real-time frequency and voltage amplitude on the PCC side, respectively. and These are the rated frequency and rated voltage, respectively. Characterizes the rate of frequency change in the initial stage of the disturbance; This characterizes the degree to which the current frequency of the system deviates from its rated value; It reflects the impact of low-voltage distribution network power coupling and local disturbances on voltage dynamics, and serves as an auxiliary correction input for virtual inertia and damping coefficient tuning; A normalized adjustment coefficient λ is introduced as the output variable of the fuzzy controller, and λ is mapped to the actual virtual inertia parameter and damping parameter: ; ; ; In the formula, This represents a fuzzy reasoning mapping relationship. This is the inertia normalization adjustment coefficient; and These are the minimum and maximum allowable values ​​for virtual inertia, respectively. and These are the minimum and maximum allowable values ​​for the damping coefficient, respectively.

[0014] Furthermore, the input and output variables are further divided into linguistic variables, specifically: Input variables , and Each is divided into four language variables: very small (VS), small (S), medium (M), and large (L); The output variables are divided into five language variables: small (S), medium small (MS), medium (M), medium large (ML), and large (L).

[0015] Furthermore, a triangular membership function is uniformly applied to both input and output variables, with the expression: ; In the formula, This is the membership function.

[0016] The beneficial effects of this invention are: (1) Based on the original photovoltaic cluster virtual synchronization control framework, this invention establishes a small-signal model of active-frequency and reactive-voltage control of photovoltaic cluster virtual synchronization, which takes into account the power coupling, for the collaborative operation scenario of new energy power generation cluster. The equivalent characterization of cluster layer inertia and damping is realized through the power feedback mechanism of secondary control, so that the system presents the characteristics of a synchronous network.

[0017] (2) This invention introduces a fuzzy logic supervision layer, using the grid connection point frequency change rate, frequency deviation, and voltage deviation as inputs, and simultaneously performs online adaptive tuning of the virtual inertia and damping coefficient. Without changing the original main control structure, it achieves adaptive adjustment of key control parameters to changes in operating conditions. This method not only enhances the initial frequency suppression capability of disturbances but also improves the system oscillation attenuation and dynamic recovery process.

[0018] (3) Establish a small-signal model of a multi-converter collaborative network cluster considering power coupling, systematically analyze the influence of virtual inertia, damping coefficient and equipment dynamic characteristics on frequency and voltage response, and conduct simulation verification on a six-machine hybrid system consisting of three photovoltaic converters and three energy storage converters. Verify that the proposed fuzzy adaptive collaborative network method is superior to the fixed parameter control scheme in terms of frequency suppression, RoCoF suppression and voltage recovery. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 It is an equivalent model for grid-connected photovoltaic clusters of distributed power sources; Figure 2 This is a block diagram of the active power frequency control for virtual synchronization of photovoltaic clusters; Figure 3 This is a block diagram of the virtual synchronization reactive voltage control for photovoltaic clusters. Figure 4 The comparison shows the phase response of virtual synchronous generators and cluster virtual synchronization, as well as the impact of equipment dynamics on the phase response characteristics of cluster virtual synchronization. Among them, (a) shows the impact of different inertia J on the phase response of virtual synchronous generators, (b) shows the impact of different inertia Jp on the phase response of cluster virtual synchronization, and (c) shows the impact of different equipment dynamics on the phase response of cluster virtual synchronization. Figure 5 The active power and frequency response trajectories of the photovoltaic cluster under different inertia parameters are shown; where (a) is the active power output response curve of the virtual synchronization system; and (b) is the frequency response curve of the virtual synchronization system. Figure 6 The voltage response characteristics of the virtual synchronization of the cluster are shown in (a), where (a) is the influence of different devices on the voltage response of the virtual synchronization of the cluster, and (b) is the influence of different voltage control coefficients kp on the voltage response of the virtual synchronization of the cluster. Figure 7 It is the membership function of the absolute value of the rate of change of frequency; Figure 8 It is the membership function of the absolute value of the frequency deviation; Figure 9 It is the membership function of the absolute value of the voltage deviation; Figure 10 This is a schematic diagram of the active power loop structure; Figure 11 The random fluctuation of the load power of the photovoltaic cluster system is shown in (a) and (b) is shown in (b) the random fluctuation of the active power of the photovoltaic cluster system load. Figure 12 This is a comparison of the frequency regulation of photovoltaic cluster systems; Figure 13 This refers to the voltage regulation status of the photovoltaic cluster system; Figure 14 It refers to the adjustment of photovoltaic active power within the cluster system; Figure 15 It is the adjustment of active power of energy storage within the cluster system; Figure 16 It is the adjustment of photovoltaic reactive power within the cluster system; Figure 17 It is the reactive power adjustment of the energy storage within the cluster system. Detailed Implementation

[0021] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0022] This invention provides a method for collaborative network construction of distributed new energy clusters based on fuzzy control, comprising the following steps: S1: Construct an equivalent model of a photovoltaic cluster connected to the grid with n distributed power sources, calculate the output power of each distributed power source, perform Taylor expansion and linearization on the power to obtain the small-signal equation of the power output; It should be further explained that the equivalent model is as follows: Figure 1 As shown, the output power of each distributed power source is calculated as follows: ; In the formula, , These are the active power and reactive power outputs of the distributed power source, respectively. and The equivalent resistance and inductive reactance of the line; This is the value of the work angle; This refers to the amplitude of the common bus voltage. Let be the output voltage amplitude of the i-th DG unit; Since both active and reactive power are affected by the power angle and voltage amplitude, a Taylor expansion and linearization are performed near the equilibrium point to obtain the small-signal dynamic equation for power output: ; ; In the formula, and These represent the changes in the active and reactive power outputs of the distributed power source, respectively. Let be the output voltage amplitude of the i-th DG unit; The change in phase angle of the output voltage of the i-th DG unit; This represents the change in the output voltage amplitude of the i-th DG unit; , , , This is the power coupling gain coefficient.

[0023] S2: Construct a small-signal model for virtual synchronization of active-frequency and reactive-voltage control of photovoltaic clusters, considering power coupling. It should be further explained that a power feedback loop is added to the cluster layer to construct an inertial equivalent model, and combined with the dynamic and power coupling characteristics of photovoltaic / energy storage, an active power-frequency control structure for virtual synchronization of the photovoltaic cluster is established, the control block diagram of which is shown below. Figure 2 As shown. Considering the active power output of the cluster-controlled photovoltaic and energy storage devices and various uncontrollable power disturbances, a coupled feedback and inertial process are designed to obtain the active power-frequency control small-signal model as follows: ; ; ; ; In the formula, This represents the change in the system's angular velocity. This is a reference value for angular velocity; and These represent the changes in active power output of m energy storage units and n photovoltaic units, respectively. This refers to uncontrollable disturbances within and outside the system, including internal and external load disturbances, line losses, and external angular frequency variations. This represents the change in active power. This is the angular velocity droop gain coefficient; and These are the cluster's virtual inertia and damping coefficient, respectively. The active power output of m energy storage units within the cluster; and These represent the changes in active power input for the j-th energy storage device and the i-th photovoltaic device, respectively. and For the dynamic transmission characteristics of input and output of energy storage and photovoltaic equipment; In voltage regulation control design, the regulation parameter k is introduced. p and k i And approximate the output voltage E and measured voltage U of the cluster system to obtain Figure 3 The control block diagram and the small-signal model of virtual synchronization reactive power-voltage control for photovoltaic clusters are shown below: ; ; ; ; In the formula, The change in output voltage amplitude; and These are the integral gain coefficient and the proportional gain coefficient, respectively. and These represent the changes in total reactive power output for controllable energy storage and photovoltaics, respectively. This represents the change in reactive power. This refers to the uncontrollable reactive power disturbance changes inside and outside the cluster. and These represent the changes in reactive power input for the i-th photovoltaic and j-th energy storage devices, respectively.

[0024] S3: Based on the small-signal model, the influence of virtual inertia, damping coefficient and key parameters of equipment dynamic characteristics on the frequency and voltage response process was clarified; Assuming no power disturbance, the phase response functions of the virtual synchronous generator and the cluster virtual synchronization under grid-connected conditions are as follows: ; assumed H PV = 1 / ( T PV s +1) Simplifies the input-output transmission characteristics of photovoltaic devices. H Bess = 1 / ( T Bess s +1) To simplify the input-output transmission characteristics of energy storage devices, the phase response function of cluster virtual synchronization is simplified to: ; According to the above formula, the photovoltaic cluster virtual synchronization architecture has a higher system order due to the coupling of the dynamic characteristics of photovoltaics and energy storage. The introduction of this higher-order dynamics makes its phase response characteristics not completely equivalent to those of traditional virtual synchronous generators. Figure 4 The comparison between the phase response of virtual synchronous generators and cluster virtual synchronization is shown, as well as the impact of equipment dynamics on the phase response characteristics of cluster virtual synchronization. Figure 4 (b) and (c) reveal that under constant damping conditions, the phase dynamic response is the result of the coupling between the virtual inertia of the control layer and the dynamics of the physical layer devices. The dynamic time delay characteristics of the underlying photovoltaic and energy storage units can effectively enhance the equivalent inertia of the system, that is, the faster the device dynamics, the more significant the system's ability to resist phase changes.

[0025] When power disturbances occur during grid connection, the disturbance will be shared by the external power grid and the photovoltaic cluster, and the power balance relationship is as follows: ; ; To simplify the analysis, the photovoltaic and energy storage units are assumed to have the same dynamic response time constant. Based on the simplified model above, the active power and frequency response trajectories of the photovoltaic cluster under different inertia parameters are as follows: Figure 5 As shown in the figure. The results indicate that the grid-connected photovoltaic cluster exhibits significant inertial support characteristics, and this dynamic behavior is mainly influenced by the virtual inertia parameter J. p Dominantly, specifically, the overshoot and rate of change of the system frequency response are related to J. p There is a significant negative correlation.

[0026] Based on the reactive-voltage control small-signal model, the system voltage response function is obtained as follows: ; In the formula, For the dynamic time of photovoltaic equipment; This refers to the dynamic time of energy storage devices.

[0027] This leads to the obtaining of the equipment dynamic process G. PV and G Bess The influence of damping characteristics on the voltage response of the cluster system, such as Figure 6 As shown, it is observed that, under the premise that the system phase does not change, the virtual synchronization voltage response characteristics are simultaneously affected by the transmission characteristics of the devices in the region and the voltage control parameter k. p The influence of a smaller voltage control parameter k p Slower dynamic processes can also lead to reduced system stability.

[0028] S4: Based on the aforementioned influence law, a fuzzy logic supervision layer is introduced. The absolute value of the frequency change rate at the grid connection point, the absolute value of the frequency deviation, and the absolute value of the voltage deviation are used as inputs, and the normalized adjustment coefficient λ is used as the output of the fuzzy controller. The normalized adjustment coefficient λ is mapped to the actual virtual inertia parameter and damping parameter, thereby enhancing the system's dynamic adaptability to random load disturbances and changes in coupled operating conditions.

[0029] It should be further explained that, in order for the fuzzy controller to reflect both the degree of frequency change in the initial stage of the disturbance and the current frequency deviation level and the active-voltage coupling characteristics in the low-voltage network, the absolute value of the rate of change of frequency at the grid connection point, the absolute value of frequency deviation, and the absolute value of voltage deviation are used as the input variables of the fuzzy controller, and are defined as follows: ; ; ; In the formula, and These represent the real-time frequency and voltage amplitude on the PCC side, respectively. and These are the rated frequency and rated voltage, respectively. The rate of frequency change in the early stage of a disturbance is the main criterion for the inertia support requirement in the early stage of a disturbance. It characterizes the degree to which the current frequency of the system deviates from the rated value and can be used to determine whether the frequency support is sufficient. It reflects the impact of low-voltage distribution network power coupling and local disturbances on voltage dynamics, and serves as an auxiliary correction input for virtual inertia and damping coefficient tuning; Considering that the fuzzy control output needs to be easily mapped continuously in simulation, a normalized adjustment coefficient λ is introduced as the output variable of the fuzzy controller, and λ is mapped to the actual virtual inertia parameter and damping parameter: ; ; ; In the formula, This represents a fuzzy reasoning mapping relationship. This is the inertia normalization adjustment coefficient; and These are the minimum and maximum allowable values ​​for virtual inertia, respectively. and These are the minimum and maximum allowable values ​​for the damping coefficient, respectively.

[0030] A Mamdani-type fuzzy inference structure is adopted, and the fuzzy control process is divided into three steps: fuzzification, rule inference, and defuzzification. Considering the computational complexity of online implementation and the clarity of rule expression, the input and output variables are classified as linguistic variables, specifically: Input variables , and Each is divided into four language variables: very small (VS), small (S), medium (M), and large (L); The output variables are divided into five language variables: small (S), medium small (MS), medium (M), medium large (ML), and large (L).

[0031] To facilitate implementation and parameter tuning, a triangular membership function is uniformly adopted for both input and output variables, with the expression: ; In the formula, This is the membership function.

[0032] Figure 7 , Figure 8 and Figure 9 Membership functions for the three input variables are given. All three use triangular membership functions and are divided into four fuzzy subsets: very small (VS), small (S), medium (M), and large (L).

[0033] Example verification: To verify the effectiveness of the proposed method, a 6-node photovoltaic cluster system topology is used as an example. The number and capacity of photovoltaic and energy storage units are described here. The controllable photovoltaic inverter adopts grid-following control, and the energy storage inverter adopts grid-connected control. The specific simulation parameters are shown in Table 1.

[0034] Table 1: Simulation parameters and values ;

[0035] The verification is carried out using a six-unit photovoltaic energy storage hybrid system as an example. The system topology is as follows: Figure 10 As shown, the DC input power of the inverter is set as a constant voltage source. The rated active power of the photovoltaic inverter is 20kW, and the rated active power of the energy storage inverter is 20kW, while the rated reactive power is 6kvar. The inter-node lines are set at 0.5km. In the power feedback loop control, the active power participation factor and reactive power participation factor of the three photovoltaic and three energy storage devices are both 1 / 6.

[0036] To evaluate the dynamic adjustment capability of the designed control method under random power fluctuations in a photovoltaic cluster system, random power fluctuations within the system are achieved through load switching. The time-domain variation of the cluster power is used as the external equivalent disturbance input, and the overall power fluctuation is as follows: Figure 11 As shown, the cluster system is in the initialization phase before 0.5s, with an initial load of 120kW + 36kvar. During the period from 0.5s to 1s, a load of 10kW + 2kvar is applied; during the period from 1.5s to 2s, a load of 5kW + 4kvar is applied; and during the period from 2.5s to 3s, a load of 6kW + 3kvar is applied. The inertial equivalent implementation process of the system is observed.

[0037] according to Figure 12 The frequency regulation waveform of the photovoltaic cluster system can be observed. When random power fluctuations occur in the system, the application of fuzzy control method can effectively suppress the rapid change of the photovoltaic cluster frequency, and further realize the rapid dynamic recovery of voltage, thus ensuring the safe and stable operation of the system. Figure 13 This demonstrates that during periods of reactive power fluctuation, the system's external voltage amplitude quickly stabilizes at 311V after the disturbance, suppressing the impact of power coupling on voltage and achieving rapid voltage recovery, thereby enhancing the stability of system operation under random fluctuation conditions.

[0038] Figure 14 and Figure 15The active power regulation process of photovoltaic (PV) and energy storage devices during random power fluctuations is demonstrated. During the period when load 1 is put into operation, the active power output of a single PV unit increased from 20.1kW to a stable 21.77kW, and the active power output of a single energy storage unit increased from 20.1kW to a stable 21.75kW. Simultaneously, after the secondary control is activated, the PV devices respond rapidly in the initial stage of the disturbance, but a short-term overshoot occurs. This allows for rapid absorption and compensation of power surges and deficits in a short period, delaying sudden changes in the overall power and frequency of the cluster system and reducing the probability of large overshoot.

[0039] Depend on Figure 16 and Figure 17 As can be seen, the control method of this invention effectively mobilizes the reactive power support capability of the cluster's internal resources. Upon detecting output voltage fluctuations, the feedback loop rapidly generates compensation commands and distributes them to each distributed node. At this time, the energy storage inverter and the photovoltaic inverter exhibit good collaborative characteristics, utilizing their respective reactive power capacities to quickly compensate the system. During the period when load 1 is engaged, the reactive power output of a single photovoltaic unit stabilizes at 6.36 kvar, and the reactive power output of a single energy storage unit stabilizes at 6.36 kvar, up from 6.03 kvar. This multi-machine collaborative adjustment significantly improves the voltage recovery speed, forcing the grid connection point voltage to quickly return to the rated level, verifying the photovoltaic cluster's ability to maintain stable voltage as a whole.

[0040] In summary, to address the issues of power surges, voltage fluctuations, and decreased frequency immunity under low inertia conditions after a high proportion of distributed photovoltaic (PV) systems are integrated into the distribution network, the method of this invention introduces the concept of a virtual synchronous generator and a power feedback mechanism into the cluster's secondary control structure. This enables the distributed PV cluster to exhibit inertial response characteristics similar to a synchronous machine, reducing the rate of frequency change during power disturbances and fundamentally improving the dynamic vulnerability caused by insufficient inertia under conditions of high-penetration power electronics grid connection. Based on the original virtual synchronization control framework, the introduction of a fuzzy logic supervision layer means that the virtual inertia and damping coefficients are no longer fixed but can be adaptively tuned online based on system state variables such as the rate of frequency change, frequency deviation, and voltage deviation. Therefore, the controller can enhance inertia support in the early stages of disturbances and improve damping and vibration suppression capabilities during the oscillation decay phase, thereby extending from fixed-parameter control to dual-parameter adaptive control.

[0041] The above embodiments are preferred implementations of the present invention. In addition, the present invention can be implemented in other ways. Any obvious substitutions without departing from the concept of the present technical solution are within the protection scope of the present invention.

[0042] To facilitate understanding by those skilled in the art of the improvements of this invention over the prior art, some of the accompanying drawings and descriptions have been simplified, and for clarity, some other elements have been omitted from this application. Those skilled in the art should realize that these omitted elements may also constitute the content of this invention.

Claims

1. A method for collaborative network construction of distributed new energy clusters based on fuzzy control, characterized in that, Includes the following steps: S1: Construct an equivalent model of a photovoltaic cluster connected to the grid with n distributed power sources, calculate the output power of each distributed power source, perform Taylor expansion and linearization on the power to obtain the small-signal equation of the power output; S2: Construct a small-signal model for virtual synchronization of active-frequency and reactive-voltage control of photovoltaic clusters, considering power coupling. S3: Based on the small-signal model, the influence of virtual inertia, damping coefficient and key parameters of equipment dynamic characteristics on the frequency and voltage response process was clarified; S4: Based on the aforementioned influence law, a fuzzy logic supervision layer is introduced, using the absolute value of the frequency change rate at the grid connection point, the absolute value of the frequency deviation, and the absolute value of the voltage deviation as inputs, and the normalized adjustment coefficient λ as the output of the fuzzy controller, and mapping the normalized adjustment coefficient λ to the actual virtual inertia parameter and damping parameter.

2. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 1, characterized in that, Specifically, S1 is: The output power of each distributed power source is: ； In the formula, , These are the active power and reactive power outputs of the distributed power source, respectively. and The equivalent resistance and inductive reactance of the line; This is the value of the work angle; This refers to the amplitude of the common bus voltage. Let be the output voltage amplitude of the i-th DG unit; The small-signal equation for output power is: ； ； In the formula, and These represent the changes in the active and reactive power outputs of the distributed power source, respectively. Let be the output voltage amplitude of the i-th DG unit; The change in phase angle of the output voltage of the i-th DG unit; This represents the change in the output voltage amplitude of the i-th DG unit; , , , This is the power coupling gain coefficient.

3. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 2, characterized in that, Specifically, S2 is: The small-signal model for virtual synchronization of active power and frequency control in photovoltaic clusters is as follows: ； ； ； ； In the formula, This represents the change in the system's angular velocity. This is a reference value for angular velocity; and These represent the changes in active power output of m energy storage units and n photovoltaic units, respectively. This refers to uncontrollable disturbances within and outside the system, including internal and external load disturbances, line losses, and external angular frequency variations. This represents the change in active power. This is the angular velocity droop gain coefficient; and These are the cluster's virtual inertia and damping coefficient, respectively. The active power output of m energy storage units within the cluster; and These represent the changes in active power input for the j-th energy storage device and the i-th photovoltaic device, respectively. and For the dynamic transmission characteristics of input and output of energy storage and photovoltaic equipment; The small-signal model for virtual synchronization of reactive power-voltage control in photovoltaic clusters is as follows: ； ； ； ； In the formula, The change in output voltage amplitude; and These are the integral gain coefficient and the proportional gain coefficient, respectively. and These represent the changes in total reactive power output for controllable energy storage and photovoltaics, respectively. This represents the change in reactive power. This refers to the uncontrollable reactive power disturbance changes inside and outside the cluster. and These represent the changes in reactive power input for the i-th photovoltaic and j-th energy storage devices, respectively.

4. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 3, characterized in that, Specifically, S3 is: The phase response function of cluster virtual synchronization in grid-connected mode is: ； assumed H PV = 1 / ( T PV s +1) Simplifies the input-output transmission characteristics of photovoltaic devices. H Bess = 1 / ( T Bess s +1) To simplify the input-output transmission characteristics of energy storage devices, the phase response function of cluster virtual synchronization is simplified to: ； When power disturbances occur during grid connection, the disturbance will be shared by the external power grid and the photovoltaic cluster, and the power balance relationship is as follows: ； ； Based on the reactive-voltage control small-signal model, the system voltage response function is obtained as follows: ； In the formula, For the dynamic time of photovoltaic equipment; This refers to the dynamic time of energy storage devices.

5. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 1, characterized in that, Specifically, S4 is: The absolute values ​​of the rate of change of frequency at the grid connection point, the absolute values ​​of frequency deviation, and the absolute values ​​of voltage deviation are selected as input variables for the fuzzy controller, and are defined as follows: ； ； ； In the formula, and These represent the real-time frequency and voltage amplitude on the PCC side, respectively. and These are the rated frequency and rated voltage, respectively. Characterizes the rate of frequency change in the initial stage of the disturbance; This characterizes the degree to which the current frequency of the system deviates from its rated value; It reflects the impact of low-voltage distribution network power coupling and local disturbances on voltage dynamics, and serves as an auxiliary correction input for inertia and damping tuning; A normalized adjustment coefficient λ is introduced as the output variable of the fuzzy controller, and λ is mapped to the actual virtual inertia parameter and damping parameter: ； ； ； In the formula, This represents a fuzzy reasoning mapping relationship. This is the inertia normalization adjustment coefficient; and These are the minimum and maximum allowable values ​​for virtual inertia, respectively. and These are the minimum and maximum allowable values ​​for the damping coefficient, respectively.

6. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 5, characterized in that, The input and output variables are classified as language variables, specifically: Input variables , and Each is divided into four language variables: very small (VS), small (S), medium (M), and large (L); The output variables are divided into five language variables: small (S), medium small (MS), medium (M), medium large (ML), and large (L).

7. The method for collaborative network construction of distributed new energy clusters based on fuzzy control as described in claim 6, characterized in that, The triangular membership function is used uniformly for both input and output variables, and its expression is: ； In the formula, This is the membership function.

Images