Distributed photovoltaic system and control method

Abstract

The invention relates to the technical field of photovoltaic power generation, and discloses a distributed photovoltaic system and a control method.The distributed photovoltaic system comprises a photovoltaic cascade adaptive impedance matching module, an intelligent control module, a layered dynamic reconstruction convergence module and an energy management and interface module, the intelligent control module diagnoses the health state of the string by comparing the waveform with a pre-stored fingerprint and generates a reconstruction instruction by combining the adjustment margin with external data and by feedforward prediction and feedback adjustment dual logic, and the hierarchical dynamic reconstruction convergence module adjusts convergence topology according to the instruction and performs convergence control on the string according to the reconstruction instruction. The control method comprises the corresponding steps of impedance matching, decision control, topology reconstruction and energy management. According to the invention, the power generation efficiency, the operation stability and the environmental adaptability of the system are improved through group cascade power optimization and a pre-judgment type active reconstruction mechanism.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation technology, and in particular to distributed photovoltaic systems and control methods. Background Technology

[0002] As an important form of renewable energy utilization, distributed photovoltaic (PV) power generation efficiency is directly affected by the external environment and the condition of its own components. In actual operation, factors such as cloud cover, building shadows, surface stains, or uneven aging of the PV modules themselves often lead to inconsistencies in the output characteristics of different PV strings in the system, i.e., mismatch.

[0003] Traditional centralized or string photovoltaic (PV) systems typically employ uniform maximum power point tracking (MPPT) control. When addressing mismatch issues between PV strings, this control method can only find a balanced global maximum power point, failing to optimize for each individual string. This results in some strings deviating from their optimal operating state, leading to a loss of overall system power generation. Furthermore, this structure lacks rapid adaptive adjustment capabilities at the individual string level, making it difficult to cope with complex and changing local operating conditions.

[0004] Furthermore, the performance degradation of photovoltaic strings is a gradual process. Existing systems have limited means of monitoring string status, usually relying on monitoring macroscopic parameters such as output power or voltage. This approach can only respond after performance has significantly declined or a fault has occurred, lacking insight into the evolution trend of the string's internal health status. Therefore, the system cannot predict and intervene in advance to avoid potential power generation losses caused by performance degradation, affecting the long-term stable operation of the system.

[0005] To address the mismatch problem, some technical solutions have proposed methods for dynamically reconstructing the photovoltaic array topology. However, these methods often lack a collaborative mechanism with underlying fine-grained power optimization. The system either relies on frequent topology switching or cannot smoothly transition to global topology optimization when local adjustment capabilities are insufficient. This indicates that existing technologies lack a hierarchical and collaborative control strategy in combining rapid local optimization with global topology reconstruction, thus limiting the overall adaptability and robustness of the system under different levels of disturbance. Summary of the Invention

[0006] The purpose of this invention is to provide a distributed photovoltaic system and control method, which solves the problems in the prior art that it cannot perform independent and refined power optimization for individual photovoltaic strings, lacks predictive health status diagnosis and active reconfiguration capabilities, and lacks a collaborative control mechanism between local optimization and global topology adjustment.

[0007] To address the aforementioned technical problems, the first aspect of this invention provides a distributed photovoltaic system, comprising: at least one photovoltaic cascade adaptive impedance matching module, an intelligent control module, a hierarchical dynamic reconfiguration combiner module, and an energy management and interface module.

[0008] The photovoltaic string adaptive impedance matching module is electrically connected to the photovoltaic strings. Its function is to receive the raw DC power output from the photovoltaic strings and perform real-time impedance matching processing on it. This processing aims to ensure that each photovoltaic string operates near its instantaneous maximum power point, and the processed output is DC power optimized by instantaneous impedance matching. While performing impedance matching, the module also generates two key output signals: the complete response waveform of the high-frequency probe signal and its own regulation margin.

[0009] In one specific embodiment, the photovoltaic cascade adaptive impedance matching module consists of a high-frequency detection and feedback unit and an adaptive matching network unit. The high-frequency detection and feedback unit injects a high-frequency detection signal into the original DC current and calculates the reflection coefficient reflecting the current impedance matching state according to the reflection coefficient formula. This reflection coefficient is used as a feedback signal and sent to the adaptive matching network unit. Based on this feedback signal, the adaptive matching network unit adjusts its own network impedance, thereby completing the impedance matching process for the original DC current.

[0010] The process by which the high-frequency detection and feedback unit generates the complete response waveform and regulation margin of the high-frequency detection signal is as follows: the response of the monitored high-frequency detection signal is digitized by its built-in analog-to-digital converter to generate the complete response waveform of the high-frequency detection signal; at the same time, the regulation margin is calculated according to the current control value of the variable reactance element inside the adaptive matching network unit and the upper and lower limits of its adjustment range, based on the regulation margin formula.

[0011] The intelligent control module, serving as the system's decision-making core, communicates with the photovoltaic cascade adaptive impedance matching module. It receives the complete response waveform and adjustment margin of the high-frequency detection signal uploaded by the former, and, combined with external environmental data obtained from the energy management and interface module, ultimately generates reconfiguration instructions for adjusting the system topology.

[0012] In one specific embodiment, the intelligent control module includes a string health status diagnosis unit and a topology reconfiguration decision unit. The string health status diagnosis unit uses the complete response waveform of the received high-frequency probe signal, compares it with a pre-stored high-frequency response feature fingerprint, and calculates the health status deviation characterizing the current performance state of the photovoltaic string according to the health status deviation formula. This high-frequency response feature fingerprint is pre-set based on the response characteristics of the photovoltaic string under ideal or initial health conditions. The topology reconfiguration decision unit then integrates the health status deviation output by the string health status diagnosis unit, the adjustment margin reported by the photovoltaic string-level adaptive impedance matching module, and external environmental data to generate the final reconfiguration command.

[0013] To achieve both active and passive adjustment of the system topology, the topology reconstruction decision unit runs feedforward prediction triggering logic and feedback adjustment triggering logic in parallel, and incorporates an arbitration mechanism.

[0014] The feedforward prediction triggering logic is triggered when the predicted power attenuation value calculated based on health status deviation and external environment data exceeds a preset power attenuation threshold. This is a proactive reconfiguration based on performance degradation prediction.

[0015] The triggering condition for the feedback adjustment triggering logic is as follows: when the topology reconfiguration decision unit detects that the number of photovoltaic cascade adaptive impedance matching modules with a regulation margin less than a preset regulation margin threshold in the system reaches or exceeds a preset module number threshold, this logic is triggered. This is a passive response reconfiguration based on the depletion of module regulation capability. When any triggering logic is triggered, a corresponding reconfiguration command is generated; if two triggering logics are triggered simultaneously, the final output reconfiguration command is determined by an arbitration mechanism. The power attenuation threshold, regulation margin threshold, and module number threshold are all preset based on the system design's safety redundancy and regulation sensitivity requirements.

[0016] In addition, the intelligent control module may also include an energy dispatch management unit. This unit generates energy dispatch commands based on local load, power grid, and energy storage status information contained in external environmental data, and sends these commands to the energy management and interface module. Upon receiving the commands, the energy management and interface module adjusts its own energy conversion and storage behavior accordingly.

[0017] The hierarchical dynamic reconfiguration combiner module has its input terminal electrically connected to the output terminal of the photovoltaic cascade adaptive impedance matching module, and is also communicatively connected to the intelligent control module. Its function is to receive and execute reconfiguration commands, perform topology reconfiguration on the input DC power that has undergone instantaneous impedance matching optimization, and output topology-optimized combined DC power after combining.

[0018] In one specific embodiment, the hierarchical dynamic reconfiguration bus module consists of a switch matrix unit and a bus and protection unit. The switch matrix unit contains a hierarchically arranged switch array, which is responsible for receiving and parsing reconfiguration commands and driving the switching elements in the array to adjust the bus topology of each DC power supply. The bus and protection unit then performs power collection and safety protection on the electrical paths adjusted by the switch matrix unit.

[0019] The energy management and interface module has its input terminal electrically connected to the output terminal of the hierarchical dynamic reconfiguration combiner module and is communicatively connected to the intelligent control module. It is responsible for energy conversion and storage of the topology-optimized combiner DC power output by the system, and serves as the system's environmental sensing entry point, collecting external environmental data and sending it to the intelligent control module.

[0020] A second aspect of the present invention provides a control method for a distributed photovoltaic system, the method being applied to the aforementioned distributed photovoltaic system, comprising the following steps: S1. Impedance matching step: The photovoltaic string adaptive impedance matching module performs impedance matching processing on the raw DC output of the photovoltaic string, outputs DC after instantaneous impedance matching optimization, and generates the complete response waveform of the high-frequency detection signal and its own adjustment margin.

[0021] S2. Decision-making and control steps: The intelligent control module receives the complete response waveform of the high-frequency detection signal and the adjustment margin, calculates the health status deviation, and generates a reconstruction instruction in combination with external environmental data.

[0022] S3. Topology Reconfiguration Step: The hierarchical dynamic reconfiguration bus module receives the reconfiguration command, performs topology reconfiguration on the input DC power that has undergone instantaneous impedance matching optimization, and outputs the topology-optimized bus DC power.

[0023] S4. Energy Management Steps: The energy management and interface module performs energy conversion and storage on the topology-optimized DC bus, and collects the external environmental data.

[0024] In summary, the present invention has at least one of the following beneficial technical effects: 1. This invention achieves independent and precise adjustment of the output power of each photovoltaic string by setting a photovoltaic string-level adaptive impedance matching module on each photovoltaic string. It can quickly respond to instantaneous changes such as local shading or performance inconsistency, and compensate for power loss at the string level in real time. This avoids the adverse effects of performance fluctuations of a single string on the entire system, thereby improving the overall energy output efficiency of the system under complex operating conditions.

[0025] 2. This invention introduces the complete response waveform of a high-frequency detection signal as a representation of the string health status, and combines it with feedforward prediction triggering logic to enable the system to predict the performance degradation trend of photovoltaic strings. Before the performance declines significantly, the system can actively perform topology reconstruction to isolate or reassemble potentially faulty or inefficient strings, thereby ensuring the efficient and stable operation of the photovoltaic system throughout its entire life cycle.

[0026] 3. This invention combines two triggering logics, feedforward prediction and feedback regulation, to construct a dual regulation mechanism. The photovoltaic cascade adaptive impedance matching module provides the first level of fast local optimization, while the feedback regulation triggering logic based on regulation margin provides the second level of global topology reconstruction guarantee when the system's local regulation capability reaches its limit. This hierarchical and collaborative regulation method enhances the system's adaptability and robustness in the face of disturbances and faults of different degrees. Attached Figure Description

[0027] Figure 1 This is a diagram of the distributed photovoltaic system architecture of the present invention; Figure 2 This is an architecture diagram of the intelligent control module of the present invention; Figure 3 This is a flowchart of the control method for the distributed photovoltaic system of the present invention.

[0028] Among them, 10 is the photovoltaic cascade adaptive impedance matching module; 20 is the hierarchical dynamic reconfiguration combiner module; 30 is the intelligent control module; and 40 is the energy management and interface module. Detailed Implementation

[0029] The following is in conjunction with the appendix Figure 1 -Appendix Figure 3 The present invention will be further described in detail below.

[0030] This invention provides a distributed photovoltaic system, including: multiple photovoltaic string adaptive impedance matching modules 10 connected in parallel, a hierarchical dynamic reconfiguration combiner module 20, an intelligent control module 30, and an energy management and interface module 40.

[0031] In this system, the output of each photovoltaic string in the photovoltaic module array is electrically connected to the input of a photovoltaic string-level adaptive impedance matching module 10. The number of photovoltaic string-level adaptive impedance matching modules 10 corresponds to the number of photovoltaic strings. The outputs of all photovoltaic string-level adaptive impedance matching modules 10 are electrically connected to the input of the hierarchical dynamic reconfiguration combiner module 20. The output of the hierarchical dynamic reconfiguration combiner module 20 is electrically connected to the input of the energy management and interface module 40.

[0032] The intelligent control module 30 serves as the control core of the system, and its signal interactions are distributed throughout the system. It establishes bidirectional data communication links with each photovoltaic cascade adaptive impedance matching module 10, hierarchical dynamic reconfiguration combiner module 20, and energy management and interface module 40.

[0033] During system operation, the photovoltaic string adaptive impedance matching module 10 receives the raw DC power output from the photovoltaic string. This module injects a high-frequency detection signal into the DC circuit through its internal high-frequency detection and feedback unit, and monitors the matching status between its input impedance and the output impedance of the photovoltaic string in real time according to the reflection coefficient formula. The reflection coefficient formula is as follows: In the formula: Γ is the reflection coefficient of the high-frequency detection signal; Z L Z is the input impedance of the photovoltaic cascade adaptive impedance matching module 10; S The instantaneous optimal output impedance of the photovoltaic string; For Z S .

[0034] Simultaneously, the photovoltaic cascade adaptive impedance matching module 10, through its internal adaptive matching network unit, performs high-speed adjustment based on the aforementioned reflection coefficient Γ, processing the input raw DC power into DC power optimized by instantaneous impedance matching before outputting it. During this process, the module also sends data including the complete response waveform containing the high-frequency detection signal and its own adjustment margin to the intelligent control module 30.

[0035] The string health status diagnosis unit of the intelligent control module 30 receives the complete response waveform of the high-frequency detection signal and calculates the health status deviation of each photovoltaic string according to the health status deviation formula. The health status deviation formula is as follows: ΔH i =||H i (t)-H i,ref ||2; Where: ΔH i H represents the health status deviation value of the i-th photovoltaic string; i (t) is the health status feature vector of the i-th string at the current time t; H i,ref Let be the baseline health feature vector of this string; ||·||2 represents the L2 norm operation of the vector.

[0036] The topology reconfiguration decision unit of the intelligent control module 30 calculates and makes decisions based on the health status deviation, the environmental data reported by the energy management and interface module 40, and the adjustment margin from the photovoltaic cascade adaptive impedance matching module 10. The adjustment margin is calculated based on the adjustment margin formula. The adjustment margin formula is as follows: Where: M j (t) represents the adjustment margin of the j-th photovoltaic cascade adaptive impedance matching module 10 at time t; X j (t) represents the current control value of the variable reactance element in the photovoltaic cascade adaptive impedance matching module 10; X max With X min These represent the upper and lower limits of the adjustment range for the control value, respectively; |·| represents the absolute value operation.

[0037] After the decision is made, the intelligent control module 30 generates a reconstruction instruction and sends it to the hierarchical dynamic reconstruction confluence module 20.

[0038] The hierarchical dynamic reconfiguration bus module 20, based on the received reconfiguration command, dynamically adjusts and aggregates the topology of all input DC power that has undergone instantaneous impedance matching optimization through its internal switch matrix unit, and finally outputs the topology-optimized bus DC power to the energy management and interface module 40.

[0039] The energy management and interface module 40 receives topology-optimized DC power and, according to the energy scheduling instructions from the intelligent control module 30, performs AC-DC conversion through its internal inverter unit or stores and releases energy through its energy storage unit. Simultaneously, its internal parameter monitoring unit continuously collects external environmental data and reports it to the intelligent control module 30, providing data support for the system's predictive decision-making.

[0040] The photovoltaic cascade adaptive impedance matching module 10 is the core device for realizing the instantaneous energy transfer optimization of the physical layer in this invention. The input terminal of the photovoltaic cascade adaptive impedance matching module 10 is electrically connected to the output terminal of a single photovoltaic string, and the output terminal is connected to the hierarchical dynamic reconfiguration combiner module 20. Unlike traditional devices that use DC-DC conversion for maximum power point tracking, the photovoltaic cascade adaptive impedance matching module 10 directly adjusts its input impedance through high-speed analog circuits, responding to changes in the output characteristics of the photovoltaic string in a near-instantaneous manner.

[0041] The photovoltaic cascade adaptive impedance matching module 10 mainly includes a high-frequency detection and feedback unit and an adaptive matching network unit.

[0042] The high-frequency detection and feedback unit is the sensing and data generation center of the photovoltaic string adaptive impedance matching module 10. Its main function is to acquire the output impedance information of the photovoltaic string in real time without loss and to provide the data required for decision-making by other modules in the system. In a specific embodiment, this unit integrates a signal generator, a coupling circuit, and a signal separation and processing circuit.

[0043] The signal generator is used to generate a signal with a characteristic frequency f. probe A high-frequency detection signal with an amplitude much smaller than the DC component is superimposed onto the original DC output of the photovoltaic string through a coupling circuit (as a signal injection device), forming a DC current superimposed with the high-frequency detection signal. Because this detection signal has a high frequency and low energy, it will not cause substantial interference to the transmission of the main DC power.

[0044] The signal separation and processing circuit is used to monitor and extract reflected signals caused by impedance mismatch. This circuit can use directional couplers or circulators as signal separation devices to accurately separate the reflected high-frequency signal components from the main circuit. This unit calculates the reflection coefficient (Γ), which characterizes the current matching state, by analyzing the amplitude and phase information of the reflected signal. This physical relationship is determined by the reflection coefficient formula. The reflection coefficient formula is as follows: In the formula: Γ is the reflection coefficient of the high-frequency detection signal; Z L Z is the input impedance of the photovoltaic cascade adaptive impedance matching module 10; S The instantaneous optimal output impedance of the photovoltaic string; For Z S .

[0045] One output of the calculated reflection coefficient (Γ) is directly transmitted to the adaptive matching network unit inside the module as an error signal for high-speed analog feedback regulation.

[0046] In addition, the high-frequency detection and feedback unit also undertakes the task of providing key data to the intelligent control module 30. This unit digitizes the complete response waveform of the monitored high-frequency detection signal using its built-in analog-to-digital converter (ADC) and then sends it to the string health status diagnosis unit of the intelligent control module 30. Simultaneously, the intelligent control module 30 calculates the regulation margin representing its own regulation capability based on the operating status of its internal regulating elements and reports this data to the topology reconfiguration decision unit of the intelligent control module 30.

[0047] The adaptive matching network unit is the physical mechanism in the photovoltaic cascade adaptive impedance matching module 10 that performs impedance matching. This adaptive matching network unit receives the reflection coefficient (Γ) provided by the high-frequency detection and feedback unit as a real-time error feedback signal, and performs impedance transformation processing on the main DC power flowing through it.

[0048] In one specific embodiment, the circuit topology of the adaptive matching network unit is an adjustable passive impedance matching network, such as an L-type, T-type, or π-type network. This network is connected in series or parallel in the main DC circuit path, and its core consists of one or more variable reactance elements and an analog feedback control circuit.

[0049] Variable reactance elements are key to achieving continuously adjustable network impedance. In different implementations, this variable reactance element can be: One or more varactor diodes. The size of their junction capacitance can be precisely changed by altering the DC reverse bias voltage applied to them.

[0050] A MEMS (Micro-Electro-Mechanical System) variable capacitor. The capacitance value is adjusted by changing the distance or overlap of the plates through electrostatic or thermal forces.

[0051] An electrically controlled variable inductor. For example, a saturated reactor that changes the core saturation by controlling the bias magnetic field, or a graded adjustable inductor consisting of multiple fixed inductors and a switch array.

[0052] The analog feedback control circuit receives the reflection coefficient (Γ). Since Γ is a complex number, the circuit demodulates it into two DC error voltages representing the degree of mismatch (mode length) and phase deviation, respectively. Based on these two error voltages, the control circuit generates and outputs one or more DC control signals. In embodiments employing varactor diodes, this control signal is its reverse bias voltage.

[0053] The analog feedback control circuit operates as a high-speed analog closed-loop negative feedback process. When the reflection coefficient (Γ) is not zero, the control circuit adjusts the output DC control signal, thereby changing the reactance value of the variable reactance element. This change in reactance value causes an overall change in the input impedance Z of the photovoltaic cascade adaptive impedance matching module 10. L Changes occur. The goal of this adjustment process is to drive Z. L Approaching the instantaneous optimal output impedance Z of the photovoltaic string S complex conjugate This causes the modulus of the reflection coefficient (Γ) to approach zero.

[0054] Since the feedback loop is composed entirely of analog circuits, its response speed is limited only by the physical characteristics of the components, reaching microseconds or even faster, which is far superior to the traditional maximum power point tracking method based on algorithm iteration using digital signal processors (DSPs).

[0055] After processing by this unit, the impedance characteristics of the DC power superimposed with the high-frequency detection signal are optimized, and the power transmission efficiency is maximized under the current conditions. This results in a DC power supply optimized by instantaneous impedance matching, which is then output to the hierarchical dynamic reconfiguration bus module 20. Filtering out the high-frequency detection signal can be achieved by setting a simple low-pass filter at the module output, a design known in the art and will not be elaborated upon here.

[0056] The intelligent control module 30 is the central data processing and control command generation unit of the distributed photovoltaic system of the present invention. In a preferred embodiment, the intelligent control module 30 can be implemented by a digital signal processor (DSP), a field-programmable gate array (FPGA), or an industrial computer with corresponding interfaces. It interacts with all photovoltaic cascade adaptive impedance matching modules 10, hierarchical dynamic reconfiguration combiner modules 20, and energy management and interface modules 40 in the system through a bidirectional data communication link.

[0057] The intelligent control module 30 mainly integrates a string health status diagnosis unit, a topology reconfiguration decision unit, and an energy scheduling management unit.

[0058] The string health status diagnosis unit is a functional unit within the intelligent control module 30. Its function is to receive and process data from the physical layer to evaluate the performance status of each photovoltaic string.

[0059] The string health status diagnostic unit receives the complete response waveform of the high-frequency detection signal uploaded by each photovoltaic string adaptive impedance matching module 10 after being digitized as input data.

[0060] Upon receiving the response waveform, the unit first processes it to extract key parameters characterizing the string's electrical properties, thereby generating a health state feature vector. In one specific embodiment, this processing includes performing a Fast Fourier Transform (FFT) on the time-domain response waveform data to obtain its frequency domain spectrum. Subsequently, preset feature quantities are extracted from this spectrum. These feature quantities may include, but are not limited to, the amplitude attenuation coefficient of the fundamental frequency component of the probe signal, the phase shift of the fundamental frequency component, and the normalized amplitude of one or more harmonic components. These extracted feature quantities collectively constitute the health state feature vector, the composition of which can be defined by the health state vector formula. The health state vector formula is as follows: H i (t)=[A i (t),φ i (t),D i,h (t),…]; In the formula: H i (t) is the health status feature vector of the i-th photovoltaic string at time t; A i(t) is the attenuation coefficient of the fundamental frequency component of the probe signal; φ i (t) represents the phase shift of the fundamental frequency component of the probe signal; D i,h (t) represents the distortion of the h-th harmonic component.

[0061] After generating the health state feature vector, the unit compares it with a pre-stored high-frequency response feature fingerprint to quantify the difference between the current state and the baseline state. The high-frequency response feature fingerprint is a reference vector H. i,ref This fingerprint data characterizes the response of the photovoltaic string under ideal or initial healthy conditions (e.g., during initial system installation and commissioning). This fingerprint data is stored in the non-volatile memory of the intelligent control module 30.

[0062] The comparison process is achieved by calculating the vector distance between the current health status feature vector and the high-frequency response feature fingerprint. The scalar result obtained from this calculation is the health status bias (ΔH), which follows the health status bias formula. The health status bias formula is as follows: ΔH i =||H i (t)-H i,ref ||2; Where: ΔH i H represents the health status deviation value of the i-th photovoltaic string; i (t) is the health status feature vector of the i-th string at the current time t; H i,ref This is the baseline health feature vector (i.e., high-frequency response feature fingerprint) of the string; ||·||2 represents the L2 norm operation of the vector.

[0063] Finally, the string health status diagnosis unit will calculate the health status deviation (ΔH) value for each photovoltaic string and output it to the topology reconfiguration decision unit inside the intelligent control module 30 as the key basis for its predictive decision-making.

[0064] The topology reconfiguration decision unit is a functional unit within the intelligent control module 30 responsible for formulating macroscopic array topology adjustment strategies. This topology reconfiguration decision unit receives data from the string health status diagnostic unit, the photovoltaic string adaptive impedance matching module 10, and the energy management and interface module 40, and generates reconfiguration commands based on this data to control the hierarchical dynamic reconfiguration combiner module 20.

[0065] In one specific embodiment, the unit integrates two parallel decision logics: one is a feedforward prediction triggering logic based on long-term performance degradation and environmental prediction, and the other is a feedback regulation triggering logic based on the real-time adjustment capability of the physical layer.

[0066] The feedforward prediction triggering logic aims to achieve predictive and proactive system optimization. This logic receives a vector consisting of the health status deviation (ΔH) of all strings output by the string health status diagnosis unit, and an external environment data vector W(t) (e.g., short-term light intensity prediction, temperature change trend, etc.) provided by the parameter monitoring unit of the energy management and interface module 40.

[0067] Internally, this logic can be implemented as a decision model. This decision model can be: a system based on an expert rule base; a multidimensional lookup table; or a machine learning model trained offline, such as a feedforward neural network or a regression tree. This decision model establishes a mapping relationship between input data (string health status, environmental changes) and the optimal array topology. Its decision-making process can be formally described by the feedforward prediction and reconstruction formula. The feedforward prediction and reconstruction formula is as follows: C R =f predict (W(t),ΔH); In the formula: C R For the generated refactoring instructions; f predict (·) represents the feedforward prediction decision model; W(t) is the external environment data vector at time t; ΔH is a vector containing the health status deviation values ​​of all strings.

[0068] The feedback adjustment trigger logic serves as a real-time adjustment mechanism to ensure stable system operation. This logic continuously receives and monitors the adjustment margin reported by each photovoltaic cascade adaptive impedance matching module 10. The calculation method for the adjustment margin is defined within the photovoltaic cascade adaptive impedance matching module 10, and its calculation follows the adjustment margin formula. The adjustment margin formula is as follows: Where: M j (t) represents the adjustment margin of the j-th photovoltaic cascade adaptive impedance matching module 10 at time t; X j (t) represents the current control value of the variable reactance element in the photovoltaic cascade adaptive impedance matching module 10; X max With X min These represent the upper and lower limits of the adjustment range for the control value, respectively; |·| represents the absolute value operation.

[0069] This logic determines whether to generate a reconfiguration command by checking if a preset feedback adjustment trigger condition is met. This condition is met when a certain number of photovoltaic cascade adaptive impedance matching modules 10 in the system have instantaneous impedance matching capabilities that tend to saturate (i.e., the adjustment margin is too low). The feedback adjustment trigger condition is as follows: In the formula: It is an existential quantifier; "S" is a universal quantifier; "S" is the set of indices for all photovoltaic cascade adaptive impedance matching modules 10 in the system; S sub A subset of S; |S sub | represents the number of elements in the subset; N th The preset threshold for the number of modules; st indicates that the condition is met; M th This is the preset adjustment margin threshold.

[0070] This condition means: when there are at least N in the system th Each module has its own adjustment margin M. i (t) are all less than the threshold M th When this happens, feedback adjustment is triggered.

[0071] Once any logic is triggered, the topology reconfiguration decision unit generates a corresponding reconfiguration instruction. This reconfiguration instruction is a data packet containing the appropriate states of all switches within the hierarchical dynamic reconfiguration bus module 20. The topology reconfiguration decision unit also includes an arbitration mechanism to handle situations where two logics are triggered simultaneously, for example, setting the priority of feedback regulation triggers higher than that of feedforward prediction triggers. Finally, the determined reconfiguration instruction is sent to the hierarchical dynamic reconfiguration bus module 20.

[0072] The energy dispatch management unit is a functional unit within the intelligent control module 30 responsible for the unified dispatch of the overall energy flow of the system. The decision-making objective of this unit is to ensure that the electrical energy generated by the photovoltaic array and optimized in the previous stage can be economically and efficiently distributed to local loads, the power grid, or energy storage systems.

[0073] The energy dispatch management unit receives and integrates data from various parts of the system. This data includes: the current total power generation of the system and the power forecast for the short term output by the topology reconfiguration decision unit; the real-time power consumption of the local load and the current electricity price information of the grid provided by the parameter monitoring unit inside the energy management and interface module 40; and information such as the state of charge (SoC) and state of health (SoH) reported by the energy storage unit's own management system (BMS).

[0074] Based on the above input information, the energy dispatch management unit executes a preset energy management strategy, generating and outputting control commands for the inverter unit and energy storage unit in the energy management and interface module 40. Commands for the inverter unit may include specific active and reactive power output settings; commands for the energy storage unit may include target charging or discharging power and corresponding cutoff conditions.

[0075] In one specific embodiment, the energy management strategy can be implemented based on a set of preset rules. For example, when the total power generation of the system exceeds the local load demand and the state of charge (SoC) of the energy storage unit is below a preset charging limit (e.g., 95%), the unit generates an instruction to prioritize using the excess power generation to charge the energy storage unit. Once the energy storage unit is fully charged, the inverter unit is then instructed to feed the remaining power back into the grid.

[0076] In another embodiment, the energy dispatch management unit formulates a dispatch strategy by solving an optimization problem. This optimization problem has the objective function of maximizing economic benefits or minimizing energy costs. The objective function can be expressed as minimizing the sum of the total cost of purchasing electricity from the grid and the cyclic loss cost of energy storage devices, or maximizing the total revenue from selling electricity to the grid, within a predetermined time period (e.g., 24 hours).

[0077] The optimization process is subject to a series of constraints that ensure the physical feasibility and safety of the system operation. These constraints may include: system power balance constraints (i.e., at any given time, the sum of power generation, energy storage discharge, and purchased power equals the load power consumption, and the sum of energy storage charging power and sold power), upper and lower limits of the energy storage unit's state of charge (SoC), and rated charging and discharging power constraints of the inverter unit and energy storage unit. For solving this optimization problem, those skilled in the art can use algorithms such as dynamic programming, linear programming, or model predictive control (MPC). The specific programming and implementation of these algorithms are well-known in the field and will not be elaborated upon here.

[0078] In addition, the energy dispatch management unit also undertakes the safety monitoring function of the system. It continuously monitors the status information reported by each unit, and once it detects a fault or unsafe condition such as battery overheating or abnormal grid voltage, it will immediately execute the preset safety protection strategy and generate instructions to disconnect the corresponding part or put the system into a safe operation mode.

[0079] The hierarchical dynamic reconfiguration combiner module 20 replaces the traditional photovoltaic combiner box in the system of this invention. Its function is to receive and execute instructions from the intelligent control module 30 to adjust the physical connection topology of the DC inputs from multiple photovoltaic cascade adaptive impedance matching modules 10. The input terminal of the hierarchical dynamic reconfiguration combiner module 20 is electrically connected to the output terminals of all photovoltaic cascade adaptive impedance matching modules 10, and its output terminal provides topology-optimized combined DC power to the energy management and interface module 40.

[0080] The hierarchical dynamic reconfiguration bus module 20 mainly includes a switch matrix unit and a bus and protection unit.

[0081] The switch matrix unit is the actuator for dynamically reconfiguring the photovoltaic array topology. The core hardware of this switch matrix unit is a switch array composed of a large number of controllable switching elements, equipped with a local microcontroller. The switch matrix unit receives reconfiguration commands from the intelligent control module 30 and precisely controls the on / off state of each switching element in the switch array according to these commands.

[0082] In one specific embodiment, the controllable switching element can be a solid-state switch, such as a metal-oxide-semiconductor field-effect transistor (MOSFET) or an insulated-gate bipolar transistor (IGBT). Using a solid-state switch enables fast switching action and reduces conduction losses. In applications where high switching frequency is not required but higher electrical isolation in the off-state is necessary, an electromechanical relay can also be used as the switching element.

[0083] The switch array can be arranged as a fully cross-connected switch matrix. For a matrix with N input ports (corresponding to N photovoltaic cascade adaptive impedance matching modules 10) and M internal busbars, it is possible to connect any input to any busbar.

[0084] In a preferred embodiment of the present invention, to reduce the number of switches and system complexity, the switch array adopts a hierarchical structure. For example, for a 16-input system, the first-layer switch network can group the 16 inputs into pairs, selecting whether each group is connected in series or in parallel via switches; the second-layer switch network is responsible for further paralleling the eight sub-groups of the first-layer output according to instructions, ultimately merging them onto one or more main output buses. This hierarchical structure clarifies the physical meaning of hierarchical dynamic reconfiguration in the present invention, namely, achieving a combinatorial topology through a multi-level switch network.

[0085] The local microcontroller of the switch matrix unit is responsible for receiving and decoding the reconstruction instruction. This instruction can be encoded as a binary string or an address list, the contents of which are directly mapped to the target state (on or off) of each switch in the switch array. After parsing the instruction, the microcontroller generates the corresponding gate drive signal or coil excitation current to drive the corresponding switching element.

[0086] To ensure electrical safety during topology switching and prevent short circuits between inputs or between inputs and ground due to improper switching timing, the local microcontroller strictly enforces a break-before-make switching logic. That is, when disconnecting an input from an old connection point and connecting it to a new connection point, the microcontroller ensures that the input is first disconnected from all connections before being closed to the new target point, with a preset safety dead time in between.

[0087] The power collection and protection unit is a functional unit within the hierarchical dynamic reconfiguration power collection module 20 responsible for power collection and safety assurance. The input of this unit is directly connected to the output node of the switch matrix unit, and its output constitutes the external output terminal of the entire hierarchical dynamic reconfiguration power collection module 20.

[0088] The bus and protection unit contains one or more low-impedance busbars. The various electrical paths formed by the switch matrix units according to reconfiguration commands are ultimately connected to these busbars. In this way, currents from different photovoltaic string adaptive impedance matching modules 10, after topology combination, are physically collected on these busbars.

[0089] In addition to its busbar function, the busbar and protection unit also integrates multiple electrical protection functions to ensure the safe and stable operation of the system under various working conditions.

[0090] In one specific embodiment, the bus and protection unit has an overcurrent protection device connected in series on the main output bus. This device can be a DC fuse or a DC circuit breaker. When the total current flowing through the main output bus exceeds a preset safety threshold, the device disconnects the circuit, thereby protecting the downstream energy management and interface module 40 and other equipment from overcurrent surges.

[0091] The bus and protection unit may also include a reverse current protection device. In different embodiments, a reverse protection diode may be connected in series before each input branch formed by the switch matrix unit merges into the main bus. The function of this diode is to prevent current from flowing back from the main bus into the branch under specific topology or fault conditions (e.g., an abnormal drop in the voltage of a string), thereby avoiding damage to the upstream photovoltaic string adaptive impedance matching module 10 or the photovoltaic string.

[0092] To cope with voltage surges caused by lightning strikes or grid disturbances, this busbar and protection unit is also equipped with a surge protector. This surge protector is connected in parallel between the positive and negative terminals of the DC output bus and the protective ground. When a transient overvoltage occurs on the line, the surge protector quickly switches to a low-impedance state, discharging the surge current to the ground, thereby protecting the entire module and connected equipment.

[0093] In addition, a manual DC disconnect switch can be installed at the output of the bus and protection unit. This switch is used to achieve complete physical isolation between the hierarchical dynamic reconfiguration bus module 20 and the downstream system during system installation, maintenance, or emergencies, to ensure the safety of operators.

[0094] After being combined and protected by the busbar and protection unit, the topology-optimized DC power is delivered to the energy management and interface module 40 through the module's output terminals.

[0095] The energy management and interface module 40 serves as the energy collection and processing hub of the distributed photovoltaic system of this invention, connecting it to the external environment. The DC input terminal of this module 40 is electrically connected to the output terminal of the hierarchical dynamic reconfiguration combiner module 20, receiving topology-optimized combiner DC power. This module 40 is responsible for the final power conversion, storage scheduling, and acquisition of key external parameters.

[0096] The energy management and interface module 40 integrates a parameter monitoring unit, an inverter unit, and an energy storage unit.

[0097] The parameter monitoring unit serves as the window through which the system interacts with the external environment. This unit integrates a series of sensors and corresponding data acquisition circuits to obtain external information that significantly impacts the system's power generation and energy consumption strategies.

[0098] In one specific embodiment, the parameter monitoring unit includes: one or more solar intensity meters for measuring light intensity; one or more temperature sensors for measuring ambient temperature; an energy meter or current transformer for monitoring local load power consumption; and a communication interface for acquiring grid status (such as voltage, frequency, and phase) and electricity price information. The parameter monitoring unit processes and digitizes all collected raw data to form an external environment data vector W(t), which is then uploaded to the intelligent control module 30 to provide a basis for decisions by the topology reconfiguration decision unit and the energy dispatch management unit.

[0099] The inverter unit is responsible for converting DC power into AC power that meets the requirements of the power grid or local load. The input of this inverter unit is connected to the output bus of the hierarchical dynamic reconfiguration combiner module 20, receiving topology-optimized DC power. The inverter unit performs DC / AC conversion according to control commands issued by the energy dispatch management unit in the intelligent control module 30.

[0100] In one specific embodiment, the inverter unit is a single-phase or three-phase bridge inverter, which internally employs sinusoidal pulse width modulation (SPWM) or space vector pulse width modulation (SVPWM) technology. The commands issued by the energy dispatch management unit can precisely set its output active and reactive power, thereby enabling the function of feeding power to the grid or supplying power to local loads. The specific circuit topology, switching device selection, and underlying control algorithm implementation of the inverter can be chosen by those skilled in the art according to the application scenario; these are well-known technologies in the field and will not be elaborated upon here.

[0101] Energy storage units are used to realize time shifting of electrical energy, that is, storing energy when there is a surplus of electricity generation and releasing energy when needed.

[0102] In a preferred embodiment, the energy storage unit includes a battery pack, a battery management system (BMS), and a bidirectional DC / DC converter. The battery pack is the physical carrier of energy and can be a lithium-ion battery, lead-acid battery, or supercapacitor. The BMS monitors the voltage, temperature, and other states of each cell within the battery pack, calculates and reports the overall state of charge (SoC) and state of health (SoH) of the battery pack, and performs necessary balancing and protection operations. The bidirectional DC / DC converter is connected between the battery pack and the DC bus of the inverter unit. Based on the charge / discharge power commands issued by the energy dispatch management unit, it controls the bidirectional flow of energy between the battery pack and the DC bus.

[0103] This invention also provides a control method for a distributed photovoltaic system. This method operates through a hierarchical, collaborative control flow, organically combining real-time, high-speed adaptive adjustment at the physical layer with macroscopic, predictive decision-making at the system layer.

[0104] Step S1: At the physical layer of the system, each photovoltaic cascade adaptive impedance matching module 10 operates independently and continuously. When the photovoltaic module array generates raw DC power under illumination, this DC power is input to its corresponding photovoltaic cascade adaptive impedance matching module 10. Its internal high-frequency detection and feedback unit injects a high-frequency detection signal into the main circuit, forming a DC power superimposed with the high-frequency detection signal. By analyzing the reflected waveform, the reflection coefficient is calculated in real time according to the reflection coefficient formula. This reflection coefficient is directly used as the error feedback signal for the adaptive matching network unit, driving the unit to perform near-instantaneous impedance adjustment, thereby outputting DC power optimized by instantaneous impedance matching. During this process, the high-frequency detection and feedback unit also simultaneously completes two data generation tasks: first, digitizing the complete response waveform of the detection signal to generate the complete response waveform of the high-frequency detection signal; second, calculating the adjustment margin according to the adjustment margin formula based on the operating point of its internal variable reactance element. These two data points are uploaded to the intelligent control module 30.

[0105] Step S2: In the system's control layer, the string health status diagnosis unit within the intelligent control module 30 receives the complete response waveforms of the high-frequency probe signals uploaded by all photovoltaic string adaptive impedance matching modules 10. This string health status diagnosis unit processes the received waveform data, for example, by extracting its frequency domain features through Fourier transform, and constructs a health status feature vector characterizing the current electrical characteristics of the string based on the health status vector formula. Subsequently, the string health status diagnosis unit compares this vector with the high-frequency response feature fingerprint representing the ideal state of the string, which is pre-stored in memory, and calculates a scalar value of the quantified difference, i.e., the health status deviation, based on the health status deviation formula.

[0106] The topology reconfiguration decision unit within the intelligent control module 30 receives a health status deviation vector containing information about all strings, calculated by the string health status diagnosis unit, and simultaneously receives the regulation margin reported by each photovoltaic string adaptive impedance matching module 10. This topology reconfiguration decision unit operates two decision logics in parallel. The first is a feedforward prediction trigger logic, which takes the health status deviation vector and external environmental data (such as sunlight prediction) provided by the parameter monitoring unit of the energy management and interface module 40 as input, and generates a predictive reconfiguration command based on an internal decision model and the feedforward prediction reconfiguration formula. The second is a feedback regulation trigger logic, which monitors all regulation margin values ​​in real time. When the feedback regulation trigger condition is met—that is, when the number of modules in the system whose regulation capacity is approaching saturation reaches a preset threshold—a reconfiguration command is generated to address the current state. The topology reconfiguration decision unit ultimately outputs an arbitrated reconfiguration command.

[0107] Step S3: At the system's execution layer, the hierarchical dynamic reconfiguration combiner module 20 receives a reconfiguration command from the intelligent control module 30. The switch matrix unit inside the hierarchical dynamic reconfiguration combiner module 20 drives its switch array to adjust the electrical connection relationship of the DC power from all photovoltaic cascade adaptive impedance matching modules 10, which has undergone instantaneous impedance matching optimization, thereby realizing the series-parallel topology reconfiguration of the photovoltaic array. The recombined current converges in the combiner and protection unit, ultimately forming a topology-optimized combiner DC power and outputting it from the hierarchical dynamic reconfiguration combiner module 20.

[0108] Step S4: The topology-optimized DC power is delivered to the energy management and interface module 40. The energy dispatch management unit in the intelligent control module 30 issues a final energy dispatch command to the energy management and interface module 40 based on information such as the total system power generation, local load demand, energy storage status, and grid electricity price. The energy management and interface module 40, according to the command, converts the DC power into AC power through its internal inverter unit to supply local loads or feed it into the grid, and manages the charging and discharging of the electrical energy through its energy storage unit.

Claims

1. A distributed photovoltaic system, characterized in that, include: At least one photovoltaic string adaptive impedance matching module is electrically connected to the photovoltaic string, which is used to receive the raw DC power output by the photovoltaic string, perform impedance matching processing, output DC power optimized by instantaneous impedance matching, and generate the complete response waveform and adjustment margin of the high-frequency detection signal. An intelligent control module, which is communicatively connected to the photovoltaic cascade adaptive impedance matching module, is used to receive the complete response waveform of the high-frequency detection signal and the adjustment margin, and generate a reconstruction command in combination with external environmental data; A hierarchical dynamic reconfiguration bus module is electrically connected to at least one of the photovoltaic cascade adaptive impedance matching modules and communicatively connected to the intelligent control module. It is used to receive the reconfiguration command and perform topology reconfiguration, and after the input DC power optimized by instantaneous impedance matching is combined, it outputs topology-optimized bus DC power. An energy management and interface module is electrically connected to the hierarchical dynamic reconfiguration combiner module and communicatively connected to the intelligent control module. It is used to perform energy conversion and storage on the topology-optimized combiner DC power and to collect external environmental data and send it to the intelligent control module.

2. The distributed photovoltaic system according to claim 1, characterized in that, The photovoltaic cascade adaptive impedance matching module includes: A high-frequency detection and feedback unit is used to inject a high-frequency detection signal into the original DC current, calculate the reflection coefficient characterizing the impedance matching state according to the reflection coefficient formula, and generate the complete response waveform of the high-frequency detection signal and the adjustment margin. An adaptive matching network unit is used to receive the reflection coefficient as a feedback signal and adjust the network impedance of the adaptive matching network unit itself to complete the impedance matching process of the original DC power.

3. The distributed photovoltaic system according to claim 2, characterized in that, The specific process by which the high-frequency detection and feedback unit generates the complete response waveform of the high-frequency detection signal and the adjustment margin is as follows: The high-frequency detection and feedback unit uses an analog-to-digital converter built into it to digitally process the response of the detected high-frequency detection signal, thereby generating a complete response waveform of the high-frequency detection signal. Based on the current control value and the upper and lower limits of the adjustment range of the variable reactance element inside the adaptive matching network unit, the adjustment margin is calculated according to the adjustment margin formula.

4. The distributed photovoltaic system according to claim 1, characterized in that, The intelligent control module includes: A string health status diagnostic unit is used to receive the complete response waveform of the high-frequency detection signal and, based on the health status deviation formula, calculate the health status deviation characterizing the current performance status of the photovoltaic string by comparing the complete response waveform of the high-frequency detection signal with a pre-stored high-frequency response feature fingerprint. A topology reconfiguration decision unit is configured to receive and generate the reconfiguration instruction based on the health status deviation, the adjustment margin, and the external environment data; The high-frequency response feature fingerprint is preset based on the response characteristics of the photovoltaic string under ideal or initial healthy conditions.

5. The distributed photovoltaic system according to claim 4, characterized in that, The topology reconfiguration decision unit is specifically used for: The feedforward prediction triggering logic and the feedback adjustment triggering logic run in parallel, and an arbitration mechanism is built in. When either triggering logic is triggered, the corresponding reconstruction instruction is generated. When both triggering logics are triggered simultaneously, the arbitration mechanism determines the final output of the reconstruction instruction.

6. The distributed photovoltaic system according to claim 5, characterized in that, The triggering condition for the feedforward prediction triggering logic is: When the predicted power attenuation value calculated based on the health status deviation and the external environment data exceeds a preset power attenuation threshold, the feedforward prediction triggering logic is triggered.

7. The distributed photovoltaic system according to claim 6, characterized in that, The triggering condition for the feedback adjustment triggering logic is: When the topology reconfiguration decision unit detects that the number of photovoltaic cascade adaptive impedance matching modules with an adjustment margin less than a preset adjustment margin threshold reaches or exceeds a preset module number threshold, the feedback adjustment triggering logic is triggered. The power attenuation threshold, the adjustment margin threshold, and the module number threshold are preset based on the safety redundancy and adjustment sensitivity requirements of the system design.

8. The distributed photovoltaic system according to claim 1, characterized in that, The intelligent control module also includes: An energy dispatch management unit is used to generate energy dispatch instructions based on the local load, power grid and energy storage status information contained in the external environment data and send them to the energy management and interface module. The energy management and interface module receives the energy dispatch instructions and regulates energy conversion and storage accordingly.

9. The distributed photovoltaic system according to claim 1, characterized in that, The hierarchical dynamic reconfiguration merge module includes: A switch matrix unit, which contains a switch array arranged in a hierarchical structure, is used to receive and parse the reconstruction command, drive the switching elements in the switch array to operate, and adjust the bus topology of the DC output of the photovoltaic cascade adaptive impedance matching module. A bus and protection unit is used to collect electrical energy and provide safety protection for the electrical path adjusted by the switch matrix unit.

10. A control method for a distributed photovoltaic system, characterized in that, Applied to any one of claims 1-9, the method comprises the following steps: S1. The photovoltaic string adaptive impedance matching module performs impedance matching processing on the raw DC power output by the photovoltaic string, outputs DC power optimized by instantaneous impedance matching, and generates the complete response waveform of the high-frequency detection signal and its own adjustment margin. S2. The intelligent control module receives the complete response waveform of the high-frequency detection signal and the adjustment margin, calculates the health status deviation, and generates a reconstruction command by combining it with external environmental data. S3. The hierarchical dynamic reconfiguration bus module receives the reconfiguration command, performs topology reconfiguration on the input DC power that has undergone instantaneous impedance matching optimization, and outputs the topology-optimized bus DC power. S4. The energy management and interface module performs energy conversion and storage on the topology-optimized DC bus, and collects the external environmental data.

Images