A three-dimensional photovoltaic array and a power generation control method thereof

Abstract

The present application relates to the field of photovoltaic power generation, and specifically provides a three-dimensional photovoltaic array and a power generation control method thereof, the array comprising eastward, westward, southeastward and southwestward photovoltaic panels, and adopting a three-dimensional arrangement form of opposite and front sides; each photovoltaic panel is configured with a component-level power optimizer and a backside light sensor, and both are electrically connected with a four-direction MPPT cooperative controller. The controller cooperatively adjusts the output power of each photovoltaic panel according to the current time period and the equivalent front-side irradiance converted from the backside irradiance of each photovoltaic panel. The present application improves the utilization rate of land and irradiance, realizes fine regulation and control of photovoltaic panel power, avoids irradiance acquisition error, achieves global MPPT cooperative optimization of the array, effectively reduces power mismatching problem, adapts the output power of each photovoltaic panel to its own light and the global demand of the array, and improves the overall power generation efficiency and power output stability of the photovoltaic array.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation, and specifically to a three-dimensional photovoltaic array and its power generation control method. Background Technology

[0002] In photovoltaic (PV) power generation applications such as industrial and commercial distributed generation and building-integrated photovoltaics (BIPV), the arrangement of PV arrays is constrained by factors such as site space and solar irradiance utilization efficiency. Traditional planar PV arrays suffer from problems such as a single angle of sunlight reception, low irradiance utilization at different times of day, and the output power of individual PV panels is easily affected by changes in sunlight, making it difficult to improve the overall power generation efficiency of the array. Current solutions to this problem often employ a single-oriented PV array paired with a centralized MPPT controller, or install simple power regulation modules on some PV panels, while simultaneously collecting irradiance data through front-mounted light sensors to assist in power regulation. For example, the traditional arrangement of a single-oriented PV array involves mounting PV modules on a support frame at a certain tilt angle, with spacing between modules to avoid shading. However, installing at a certain tilt angle results in low land utilization, limiting the power generation per unit area. Furthermore, the single angle of sunlight reception leads to low utilization of solar irradiance at different times, and the output power of a single photovoltaic panel is easily affected by changes in sunlight, making it difficult to improve the overall power generation efficiency of the array. The centralized MPPT controller cannot independently adjust the working status of a single photovoltaic panel, which easily leads to power mismatch problems. The simple power adjustment module lacks multi-panel collaborative control logic, making it difficult to achieve global power optimization of the array. Summary of the Invention

[0003] To address the aforementioned issues, this invention provides a three-dimensional photovoltaic array and its power generation control method. Through the three-dimensional arrangement of photovoltaic panels in the east, west, southeast, and southwest directions, combined with independent module-level power optimizers and back-side illumination sensors for each panel, and based on a four-directional MPPT collaborative controller, it achieves equivalent front-side irradiance conversion from back-side irradiance and global collaborative regulation across time periods. This not only expands the light reception angle and enables all-time solar irradiance capture, improving land and irradiance utilization, but also achieves refined control of photovoltaic panel power, avoids irradiance acquisition errors, and achieves global MPPT collaborative optimization of the array. This effectively reduces power mismatch issues, allowing the output power of each photovoltaic panel to adapt to its own illumination and the overall needs of the array, thereby improving the overall power generation efficiency and power output stability of the photovoltaic array.

[0004] In a first aspect, the technical solution of the present invention provides a three-dimensional photovoltaic array, including multiple photovoltaic module units. The photovoltaic module units include: a west-facing photovoltaic panel, an east-facing photovoltaic panel, a southeast-facing photovoltaic panel, and a southwest-facing photovoltaic panel. The east-facing photovoltaic panel and the west-facing photovoltaic panel are arranged opposite to each other. The southeast-facing photovoltaic panel is arranged in front of the east-facing photovoltaic panel, and the southwest-facing photovoltaic panel is arranged in front of the west-facing photovoltaic panel. Each photovoltaic panel is equipped with a module-level power optimizer and a light sensor; the module-level power optimizer is electrically connected to the corresponding photovoltaic panel and is used to adjust the operating point of the photovoltaic panel; the light sensor is installed on the back of the corresponding photovoltaic panel and is used to collect the back irradiance data of the corresponding photovoltaic panel in real time. The module-level power optimizer and the light sensor are electrically connected to a four-way MPPT co-controller. The four-way MPPT co-controller is configured to send control commands to each module-level power optimizer to coordinately adjust the output power of each photovoltaic panel based on the current time period and the equivalent front irradiance of each photovoltaic panel. The equivalent front irradiance is calculated from the corresponding back irradiance data.

[0005] Secondly, the technical solution of the present invention provides a power generation control method for a three-dimensional photovoltaic array, applied to the aforementioned three-dimensional photovoltaic array, including a four-way MPPT cooperative control step, specifically including: Get the current time period; Based on the pre-stored time period-orientation weight model, determine the weight coefficients of west-facing photovoltaic panels, east-facing photovoltaic panels, southeast-facing photovoltaic panels, and southwest-facing photovoltaic panels in the current time period; The back irradiance data of each photovoltaic panel is converted into equivalent front irradiance. Based on the equivalent front irradiance and the actual output power of the photovoltaic panel, the maximum usable power of each photovoltaic panel under the current irradiance is estimated. Calculate the target power of each photovoltaic panel based on the global target power and the weighting coefficient of each photovoltaic panel; Control commands are sent to the module-level power optimizer of each photovoltaic panel according to the target power, and the operating point of each photovoltaic panel is adjusted.

[0006] As can be seen from the above technical solutions, this application has the following advantages: By setting up photovoltaic panels facing east, west, southeast, and southwest in a three-dimensional arrangement relative to each other and on the front, the light reception angle of the photovoltaic array is expanded, realizing the capture of solar irradiance at different times. This solves the problems of single light reception and low utilization rate of irradiance during different time periods in traditional planar photovoltaic arrays, increases the total amount of light received by the photovoltaic array as a whole, improves land utilization, and thus improves power generation efficiency. Each photovoltaic panel is independently equipped with a module-level power optimizer and a light sensor. The module-level power optimizer can independently adjust the operating point of a single photovoltaic panel, achieving precise power control of a single photovoltaic panel and reducing power mismatch problems. The light sensor is installed on the back of the photovoltaic panel, effectively avoiding irradiance acquisition errors caused by front shading, and ensuring the authenticity and validity of irradiance data. A four-way MPPT collaborative controller is set up to establish an electrical connection with each component-level power optimizer and light sensor. The equivalent front irradiance is calculated by converting the back irradiance data and the output power of each photovoltaic panel is adjusted in conjunction with the current time period. This achieves global MPPT collaborative optimization of the array based on actual light conditions and time period characteristics, so that the output power of each photovoltaic panel is adapted to its own light conditions and the global needs of the array, thereby improving the overall power generation efficiency and power output stability of the photovoltaic array. By converting the back irradiance data into equivalent front irradiance, a correlation is established between the back irradiance and the actual effective front irradiance of the photovoltaic panel. This avoids the problem of data distortion caused by the front irradiance sensor being easily affected by shadows, dust, and obstructions, providing a more accurate irradiance data basis for adjusting the operating point and optimizing the power of the photovoltaic panel, and further improving the reliability and accuracy of power regulation. Attached Figure Description

[0007] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of a photovoltaic module unit structure.

[0009] Figure 2 This is a schematic diagram of the four-way MPPT collaborative control process.

[0010] Figure 3 This is a schematic diagram of the dynamic reconstruction process for shadows.

[0011] In the diagram, 1 represents an east-facing photovoltaic panel, 2 represents a west-facing photovoltaic panel, 3 represents a southeast-facing photovoltaic panel, and 4 represents a southwest-facing photovoltaic panel. Detailed Implementation

[0012] To make the purpose, features, and advantages of this application more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solution protected by this application. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0014] The following explains the relevant terms used in this invention.

[0015] MPPT: Maximum Power Point Tracking.

[0016] This invention provides a three-dimensional photovoltaic array comprising multiple photovoltaic module units. Figure 1 This is a schematic diagram of a photovoltaic module unit structure, such as... Figure 1 As shown, the photovoltaic module unit includes: a west-facing photovoltaic panel 2, an east-facing photovoltaic panel 1, a southeast-facing photovoltaic panel 3, and a southwest-facing photovoltaic panel 4. The east-facing photovoltaic panel 1 and the west-facing photovoltaic panel 2 are positioned opposite each other, facing due east and due west respectively. The southeast-facing photovoltaic panel 3 is positioned in front of the east-facing photovoltaic panel 1 (i.e., to the south), facing southeast; the southwest-facing photovoltaic panel 4 is positioned in front of the west-facing photovoltaic panel 2 (i.e., to the south), facing southwest. In practical applications, the photovoltaic module unit can serve as a basic module, with multiple units arranged along a north-south direction to form a complete photovoltaic array. The spacing between adjacent units is optimized and determined based on the local latitude and the solar altitude angle on the winter solstice.

[0017] In this embodiment, both the west-facing photovoltaic panel 2 and the east-facing photovoltaic panel 1 are rectangular structures, with an angle of inclination of 60°±3° to the ground. In one specific embodiment, the front side of the west-facing photovoltaic panel 2 and the front side of the east-facing photovoltaic panel 1 form an equilateral triangle with the horizontal line of the ground, meaning the spatial angle between the west-facing photovoltaic panel 2 and the east-facing photovoltaic panel 1 is 60°. The southeast-facing photovoltaic panel 3 and the southwest-facing photovoltaic panel 4 are both triangular structures, their shapes adapted to the spatial position of the rectangular photovoltaic panels in the three-dimensional structure, ensuring minimal gaps between them and adjacent photovoltaic panels.

[0018] In this embodiment, the west-facing photovoltaic panel 2 and the east-facing photovoltaic panel 1 are set opposite each other at a 60° tilt angle, with their front sides forming an equilateral triangle with the ground. This achieves a doubling of the number of photovoltaic panels within the same land projection area. Taking a single photovoltaic module unit as an example, four photovoltaic panels are distributed in both the vertical and horizontal directions. Compared with the traditional planar arrangement, 1-2 more photovoltaic panels can be arranged within the same projection area, increasing the installed capacity per unit land area.

[0019] In this embodiment, the four photovoltaic panels with different orientations correspond to the optimal light-receiving directions at different times of the day: the east-facing photovoltaic panel 1 faces due east, and when the sun is at a low angle to the east in the morning, the 60° tilt angle ensures that the light is nearly perpendicular, efficiently capturing morning sunlight; the southeast-facing photovoltaic panel 3 faces southeast, and when the sun gradually moves south in the morning, this orientation is in the best light-receiving state; the southwest-facing photovoltaic panel 4 faces southwest, and when the sun moves west in the afternoon, this orientation outputs power efficiently; the west-facing photovoltaic panel 2 faces due west, and when the sun is at a low angle to the west in the evening, the 60° tilt angle ensures that the light is nearly perpendicular, efficiently capturing evening sunlight. Through the relay coordination of the four orientations, the photovoltaic module unit can achieve efficient power generation throughout the day, effectively extending the effective power generation duration. Furthermore, the west-facing photovoltaic panel 2 and the east-facing photovoltaic panel 1 are arranged at a 60° tilt angle, while the southeast-facing photovoltaic panel 3 and the southwest-facing photovoltaic panel 4 are placed at the front, making full use of the three-dimensional space above and to the south of the photovoltaic array, improving space utilization.

[0020] In this embodiment, the west-facing photovoltaic panel 2 and the east-facing photovoltaic panel 1 adopt a rectangular structure, which ensures reliable connection with the support and is easy to install. The southeast-facing photovoltaic panel 3 and the southwest-facing photovoltaic panel 4 adopt a triangular structure, which matches the layout of the rectangular photovoltaic panels in three-dimensional space, ensuring both structural compactness and avoiding material waste. The triangular photovoltaic panels can be directly attached to the inclined surface of the three-dimensional support without the need for additional frame support, reducing the overall weight and the load on the support.

[0021] In this embodiment, the photovoltaic module unit is equipped with a control system, including a module-level power optimizer, a light sensor, and a four-way MPPT collaborative controller. Each photovoltaic panel is equipped with one module-level power optimizer and one light sensor; the module-level power optimizer is electrically connected to the corresponding photovoltaic panel and is used to adjust the operating point of the photovoltaic panel, while the light sensor is installed on the back of the corresponding photovoltaic panel to collect the back irradiance data of the corresponding photovoltaic panel in real time. The module-level power optimizer and the light sensor are each electrically connected to a four-way MPPT collaborative controller, which is configured to send control commands to each module-level power optimizer to collaboratively adjust the output power of each photovoltaic panel based on the current time period and the equivalent front irradiance of each photovoltaic panel; wherein, the equivalent front irradiance is calculated from the corresponding back irradiance data.

[0022] The four-way MPPT collaborative controller connects the power optimizers and light sensors of the photovoltaic panels in four directions simultaneously. It can acquire the operating status and light data of all photovoltaic panels in real time, and coordinately adjust the output power of each photovoltaic panel according to the current time and the actual light received by each photovoltaic panel, so that the overall output power of the array is more stable and more efficient.

[0023] Specifically, each photovoltaic panel is independently equipped with a module-level power optimizer. The input of this optimizer is electrically connected to the output of the corresponding photovoltaic panel, and it is used to independently adjust the operating point of that photovoltaic panel. The module-level power optimizer has maximum power point tracking capability, and can adjust the output voltage and current of the photovoltaic panel according to the received control commands, so that the photovoltaic panel operates at the target operating point.

[0024] Each photovoltaic panel has a light sensor installed on its back to collect real-time irradiance data. Installing the sensor on the back of the panel, rather than the front, avoids blocking the sunlight-receiving area on the front, ensuring the panel's power generation efficiency remains unaffected. It also reduces the sensor's environmental exposure, minimizing dust accumulation and UV aging, thus extending its lifespan. Furthermore, back-mounted wiring is convenient, allowing it to be routed along with the photovoltaic panel's junction box, simplifying the installation process.

[0025] During installation, the DC output terminal of each photovoltaic panel is connected to the input terminal of the corresponding module-level power optimizer. The output terminals of all module-level power optimizers are then combined and connected to a DC combiner box, and then converted to AC power by an inverter for grid connection or local load use. The signal output terminals of each light sensor are connected to the analog input interface of the four-way MPPT co-controller via shielded cables. The communication interfaces of each module-level power optimizer are connected to the communication interface of the four-way MPPT co-controller via RS485 bus or power line carrier communication. The four-way MPPT co-controller sends control commands to each module-level power optimizer via the communication bus and receives the operating status data returned by each optimizer. The four-way MPPT co-controller and light sensors are powered by an auxiliary power supply, which can be taken from the DC side of the photovoltaic array or an external AC power supply, ensuring that the control system can still operate normally when the main circuit is disconnected.

[0026] The photovoltaic module unit in this embodiment is also equipped with a shading dynamic reconfiguration system, which includes an electrical switch matrix and a shading dynamic reconfiguration controller.

[0027] An electrical switch matrix, with its input terminals electrically connected to the output terminals of each module-level power optimizer and its control terminal signal-connected to the output terminal of the shadow dynamic reconfiguration controller, is used to dynamically switch the series-parallel connection topology between photovoltaic panels in response to control commands sent by the shadow dynamic reconfiguration controller. The electrical switch matrix consists of multiple controllable switching devices.

[0028] Specifically, the electrical switch matrix is ​​configured with the following connection method: each photovoltaic panel can be arbitrarily combined into series strings, and multiple strings can be connected in parallel to the inverter; by changing the on and off state of the switch matrix, the photovoltaic panels that were originally connected in series can be recombined, or the photovoltaic panels that originally belonged to different strings can be exchanged, thereby realizing the "dynamic reorganization" of photovoltaic panels in the array.

[0029] In some embodiments, the switching devices in the electrical switch matrix are either DC contactors or solid-state relays. DC contactors are suitable for high-current scenarios, with low contact resistance and low conduction losses; solid-state relays have no mechanical contacts, fast response speed, and long lifespan, making them suitable for scenarios with frequent switching. The electrical switch matrix adopts a "fully connected" topology, meaning that the output of each photovoltaic panel can be connected to any string bus via a switching device. Taking M photovoltaic panels and N strings as an example, the switch matrix needs to contain M×N switching devices to achieve optional connection from any photovoltaic panel to any string.

[0030] The shadow dynamic reconstruction controller is connected to each light sensor and the electrical switch matrix signal respectively, and is configured to send control commands to the electrical switch matrix to control the electrical switch matrix to perform electrical reconstruction based on the predicted shadow distribution.

[0031] Specifically, the input terminal of the shading dynamic reconfiguration controller is connected to the output terminal of each light sensor to acquire real-time backside irradiance data of each photovoltaic panel; the output terminal of the shading dynamic reconfiguration controller is connected to the control terminal of the electrical switch matrix to send control commands to the electrical switch matrix. The shading dynamic reconfiguration controller is configured to generate an electrical reconfiguration strategy based on the predicted shading distribution and drive the electrical switch matrix to execute the reconfiguration through control commands. In this embodiment, at a preset time (e.g., 1-5 minutes) before the predicted shading occurs, the shading dynamic reconfiguration controller sends control commands to the electrical switch matrix, driving the electrical switch matrix to execute the pre-generated reconfiguration strategy and change the series-parallel connection relationship between the photovoltaic panels. By performing the reconfiguration in advance before the shading occurs, proactive prevention is achieved, avoiding instantaneous power oscillations caused by shading.

[0032] In this embodiment, the shading dynamic reconfiguration controller is electrically connected to the four-way MPPT co-controller. After performing electrical reconfiguration, the shading dynamic reconfiguration controller sends an update command to the four-way MPPT co-controller. Responding to the update command, the four-way MPPT co-controller re-acquires the maximum available power of each photovoltaic panel and recalculates the target power of each photovoltaic panel. Before the shading dynamic reconfiguration controller performs reconfiguration, the four-way MPPT co-controller calculates the target power of each photovoltaic panel based on the current array topology and real-time illumination data of each photovoltaic panel, according to a preset time-orientation weight model, and maintains each photovoltaic panel at its corresponding operating point through a component-level power optimizer. At this time, the system's electrical topology is in its initial configuration state. When the shading dynamic reconfiguration controller determines that reconfiguration is required based on the shading prediction results, it first generates a reconfiguration strategy and then sends a control command to the electrical switch matrix at a preset time before the predicted shading occurs, completing the dynamic switching of the series-parallel connection topology between photovoltaic panels. After the reconfiguration is completed, the shading dynamic reconfiguration controller immediately sends an update command to the four-way MPPT co-controller through the communication interface. This update command includes the following information: a marker indicating that reconfiguration is complete, a description of the reconfigured array topology, and the string number to which each photovoltaic panel currently belongs. After receiving the update command, the four-way MPPT collaborative controller triggers the parameter recalculation process: (1) The four-way MPPT collaborative controller reacquires the real-time operating data of each photovoltaic panel through the power optimizer of each component, including the actual output power, output voltage, output current, etc. (2) The four-way MPPT collaborative controller reacquires the back irradiance data of each photovoltaic panel through each light sensor and recalculates the equivalent front irradiance of each photovoltaic panel according to the preset conversion model. (3) The four-way MPPT co-controller re-estimates the maximum available power of each photovoltaic panel based on the re-acquired equivalent front irradiance and the actual output power of the photovoltaic panel; the maximum available power refers to the maximum power that the photovoltaic panel can theoretically output under the current illumination conditions. (4) The four-way MPPT collaborative controller recalculates the target power of each photovoltaic panel based on the current time period, the pre-stored time period-orientation weight model and the re-estimated maximum available power of each photovoltaic panel, and sends the updated control command to each component-level power optimizer through the communication bus.

[0033] After receiving the updated control command from the four-way MPPT co-controller, each component-level power optimizer adjusts its operating point according to the target power value in the command.

[0034] The foregoing has described in detail an embodiment of a three-dimensional photovoltaic array. Based on the three-dimensional photovoltaic array described in the above embodiment, this invention also provides a power generation control method applied to the three-dimensional photovoltaic array. The method includes a step of four-way MPPT cooperative control and a step of shadow dynamic reconstruction.

[0035] Figure 2 This is a schematic diagram of the four-way MPPT collaborative control process, which includes the following steps.

[0036] S1, Get the current time period.

[0037] The four-way MPPT collaborative controller reads the current time, including hours and minutes, through its built-in real-time clock chip and determines the current time period interval according to preset time period division rules.

[0038] In this embodiment, the day is divided into five time periods: morning (6:00-9:00), late morning (9:00-12:00), noon (12:00-14:00), afternoon (14:00-17:00), and evening (17:00-20:00). It should be noted that these time periods can be dynamically adjusted based on the latitude and season of the photovoltaic array's location. For example, in high-latitude regions where summer sunshine hours are longer, the morning and evening time periods can be appropriately extended.

[0039] The four-way MPPT collaborative controller determines the current time period number t by comparing the current time with the start and end times of each time period interval, and uses this number as the index for the query weight model.

[0040] S2, based on the pre-stored time period-orientation weight model, determine the weight coefficients of west-facing photovoltaic panel 2, east-facing photovoltaic panel 1, southeast-facing photovoltaic panel 3, and southwest-facing photovoltaic panel 4 in the current time period.

[0041] The four-way MPPT collaborative controller has a pre-stored time-orientation weight model, which stores the weight coefficients of the four-orientation photovoltaic panels in tabular form for different time periods. The weight coefficients reflect the expected contribution ratio of each orientation photovoltaic panel to the total output power of the system in different time periods. In this embodiment, the specific values ​​of the time-orientation weight model are shown in Table 1 below.

[0042] Table 1: Time Period-Orientation Weight Model

[0043] In the morning (6:00-9:00): the sun is at a low angle on the east side, and the east-facing photovoltaic panel 1 is in the best light-receiving state, so it is given the highest weight of 0.6; the southeast-facing panel also has a good light-receiving angle at this time, so it is given a weight of 0.3; the west-facing and southwest-facing panels have almost no direct sunlight and generate electricity only by scattered light and ground reflected light, so they are given a lower weight of 0.05.

[0044] During the morning period (9:00-12:00): As the sun gradually moves southward, the southeast-facing photovoltaic panel 3 enters its optimal light-receiving state and is assigned the highest weight of 0.6; the east-facing panel receives less light at this time, and its weight drops to 0.2; the southwest-facing panel begins to receive a small amount of direct sunlight, and its weight rises to 0.15; the west-facing panel still receives mainly diffused light, and its weight remains at 0.05.

[0045] During the midday period (12:00-14:00): The sun is in the south, and the southeast and southwest facing panels have a better angle of light reception, with a weight of 0.4 for both. The east and west facing panels have a poorer angle of light reception at this time, with a weight of 0.1 for both.

[0046] Afternoon period (14:00-17:00): As the sun moves westward, the southwest-facing photovoltaic panel 4 enters its optimal light-receiving state and is assigned the highest weight of 0.6; the light-receiving of the west-facing panel gradually improves, and the weight rises to 0.2; the light-receiving of the southeast-facing panel deteriorates, and the weight drops to 0.15; the weight of the east-facing panel remains at 0.05.

[0047] During the evening period (17:00-20:00): the sun is located at a low angle on the west side, and the west-facing photovoltaic panel 2 is in the best light-receiving state, and is given the highest weight of 0.6; the southwest-facing panel also has a good light-receiving angle at this time, and is given a weight of 0.3; the east-facing and southeast-facing panels have basically no direct sunlight, and are given a lower weight of 0.05.

[0048] It should be noted that the above weighting coefficients are preferred values ​​for regions with latitudes of 40°-50°. In practical applications, they can be dynamically adjusted according to the specific latitude and season of the photovoltaic array's location. For example, in higher latitude regions, where the solar altitude angle is lower, the weights of the east-facing and west-facing panels can be appropriately increased during the morning and evening hours; during the long hours of sunshine in summer, the time intervals for the morning and evening hours can be extended, and the weighting allocation can be adjusted accordingly.

[0049] The four-way MPPT collaborative controller retrieves the weight coefficients of the four photovoltaic panels from a pre-stored weight model based on the current time period t determined in step S1. These correspond to the east-facing, west-facing, southeast-facing, and southwest-facing photovoltaic panels, respectively. This embodiment utilizes a time-time-orientation weighted model, enabling the system to differentiate control of photovoltaic panels facing different directions based on time-time characteristics. Given a limited global target power, priority is given to ensuring that photovoltaic panels with high power generation potential in the current time period generate more electricity, thus improving overall power generation efficiency.

[0050] S3 converts the back irradiance data of each photovoltaic panel into equivalent front irradiance. Based on the equivalent front irradiance and the actual output power of the photovoltaic panel, it estimates the maximum usable power of each photovoltaic panel under the current irradiance.

[0051] The four-way MPPT co-controller collects real-time back irradiance data of each photovoltaic panel through various light sensors. Since the sensor is installed on the back of the photovoltaic panel, it collects the irradiance received on the back side, and converts the back side irradiance data into an equivalent front side irradiance.

[0052] In this embodiment, the back irradiance data is converted into equivalent front irradiance using the following formula:

[0053] In the formula, For the first A solar panel in Equivalent frontal irradiance at time t, in W / m 2 ; For the first The actual output power of the photovoltaic panel, in W, is collected in real time by the module-level power optimizer. For the first The bifaciality of a photovoltaic panel, which is the ratio of the power generation efficiency on the back to that on the front, is provided by the photovoltaic panel manufacturer and is typically between 0.7 and 0.95. For the first Irradiance measured by sensors on the back of the photovoltaic panel, in W / m². 2 ; For the first The area of ​​a photovoltaic panel, in m² 2 .

[0054] The total output power of a photovoltaic (PV) panel equals the power contributed by the front-side irradiance plus the power contributed by the back-side irradiance. The power contributed by the back-side irradiance is calculated by multiplying the back-side irradiance by the PV panel area and then by the bifaciality. Therefore, subtracting the back-side contribution from the total output power and then dividing by the PV panel area yields the equivalent front-side irradiance.

[0055] S4. Calculate the target power of each photovoltaic panel based on the global target power and the weighting coefficient of each photovoltaic panel.

[0056] After obtaining the equivalent frontal irradiance, the four-way MPPT co-controller estimates the maximum available power of each photovoltaic panel based on this irradiance. Maximum available power refers to the maximum power that the photovoltaic panel can theoretically output under the current illumination conditions, and is estimated using linear interpolation or a lookup table method.

[0057] 1) Linear interpolation method Photovoltaic panel manufacturers typically provide the maximum power, Pstc, under standard test conditions (STC, irradiance 1000 W / m², temperature 25°C). Under non-standard irradiance conditions, the maximum power of a photovoltaic panel is approximately proportional to the irradiance. Therefore, the usable maximum power can be estimated using the following formula:

[0058] In the formula, For the first The maximum power of a photovoltaic panel under standard test conditions, expressed in watts (W). The power temperature coefficient of a photovoltaic panel is typically 0.003-0.005 / ℃. For the first The current temperature of the photovoltaic panel, in °C, can be collected by a temperature sensor.

[0059] If no temperature sensor is configured, the temperature correction term can be ignored, and a simplified formula can be used:

[0060] 2) Table lookup method The maximum power curves of the photovoltaic panel under different irradiances are obtained in advance through experimental testing or simulation, and the curves are discretized into a table and stored in the controller. During estimation, the corresponding available maximum power is obtained by looking up the table based on the equivalent front irradiance, and then corrected according to temperature.

[0061] S5 sends control commands to the module-level power optimizer of each photovoltaic panel according to the target power, and adjusts the operating point of each photovoltaic panel.

[0062] After obtaining the maximum available power and weighting coefficient of each photovoltaic panel, the four-way MPPT collaborative controller calculates the target power of each photovoltaic panel based on the global target power.

[0063] Global target power This refers to the total expected output power of the entire photovoltaic array at the current moment, which can be given by grid dispatch instructions or determined according to the charging demand or discharging plan of the energy storage system. When there are no external dispatch instructions and no energy storage coordination requirements, the global target power is set by default to the sum of the maximum available power of each photovoltaic panel.

[0064] The global target power is compared with the sum of the maximum available power of each photovoltaic panel. At that time, the global target power is allocated according to the weighting coefficients. Let the orientation be... The weighting coefficient for the current time period is This orientation has a total of The first photovoltaic panel, then the second The target power of the photovoltaic panel is:

[0065] Where M represents the total number of photovoltaic panels. For the first The orientation of the photovoltaic panel.

[0066] when At that time, each photovoltaic panel outputs its maximum available power, that is:

[0067] The four-way MPPT collaborative controller sends control commands to each module-level power optimizer via a communication bus based on the calculated target power of each photovoltaic panel. Each module-level power optimizer receives the target power value of the corresponding photovoltaic panel and adjusts the operating point of that photovoltaic panel so that its output power approaches the target power.

[0068] Specifically, if the target power equals the current maximum available power of the photovoltaic panel, the optimizer operates directly in maximum power point tracking (MPPT) mode, tracking the maximum power point in real time using either the perturbation observation method or the incremental conductance method. If the target power is less than the current maximum available power of the photovoltaic panel, the optimizer switches to constant power control mode. The optimizer adjusts the output voltage of the photovoltaic panel to stabilize its output power near the target power value. When changes in illumination conditions cause the maximum available power to fall below the target power, the optimizer automatically switches back to MPPT mode. After adjusting the operating point, the module-level power optimizer feeds back the current output voltage, output current, and output power, among other operating data, to the four-way MPPT co-controller via the communication bus. Upon receiving the feedback data, the four-way MPPT co-controller compares it with the actual output power and the target power. If the deviation is less than a preset threshold, the adjustment is considered complete, and the current command is maintained. If the deviation is greater than the preset threshold, the adjusted control command is resent, forming a closed-loop control. This embodiment achieves global optimization of multi-panel collaboration through the unified allocation of the global target power, enabling the system output power to better match external demands.

[0069] Figure 3 This is a schematic diagram of the shadow dynamic reconstruction process, which includes the following steps.

[0070] SS1 uses a ray tracing algorithm to calculate the shadow distribution of each photovoltaic panel at each moment in a preset future time period, based on the three-dimensional geometric model of the three-dimensional photovoltaic array and the solar position parameters at future times.

[0071] The three-dimensional geometric model of a three-dimensional photovoltaic array includes the spatial coordinates, orientation angle, and size parameters of each photovoltaic panel. Specifically, each photovoltaic panel... The three-dimensional coordinates of its vertices Let k be the vertex identifier, representing its projection area on the ground. The orientation angle of the photovoltaic panel is determined by the polygon formed after these vertices are projected perpendicularly onto the z=0 plane. and azimuth This indicates that it is used for shadow effect assessment.

[0072] Calculate future times using astronomical algorithms. Solar position parameters, including solar altitude angle and solar azimuth :

[0073]

[0074] in, The latitude is the local latitude. The solar declination angle, It is the solar hour angle.

[0075] Based on a three-dimensional geometric model and solar position parameters, a ray tracing algorithm is used to calculate the position of each photovoltaic panel at time [time value missing]. The geometric shadow distribution. The basic principle of ray tracing is: to emit rays from each point in the projection area of ​​the photovoltaic panel in the opposite direction of sunlight, to determine whether the ray intersects with other photovoltaic panels or support structures, and if so, to record the height of the intersection point.

[0076] Specifically, for any photovoltaic panel Its shadow area Determined by the following system of inequalities:

[0077] In the formula, For photovoltaic panels In the projection area on the ground, For point The shadow height at a point represents the lowest point in the vertical direction that the sun can reach.

[0078] For any point In photovoltaic panels projection area Inside, its shadow height The height of the shadow is determined by the height of the obstruction and the position of the sun. By iterating through all possible obstructions (including other photovoltaic panels and support structures), the shadow height produced by each obstruction at that point is calculated, and the maximum value is taken as the shadow height at that point.

[0079] Among them, Obstacles are a collection of objects that may produce shadows, including other photovoltaic panels and support structures; For covering The height can be extracted from the three-dimensional geometric model; For point to the obstruction The horizontal distance is calculated based on the position coordinates of the obstruction in the three-dimensional geometric model; For covering Relative to point The azimuth angle can be provided by a three-dimensional geometric model.

[0080] obstruction At point The height of the shadow cast at a point is equal to the height of the obstruction minus the shadow offset caused by the sun's tilt. By comparing the shadow heights cast by all obstructions at that point, the maximum value is taken as the actual shadow height at that point. If a point is not obstructed by any object, then... This indicates that the point is completely exposed to sunlight.

[0081] SS2 calculates the basic shadow coverage of each photovoltaic panel at the corresponding time based on the predicted shadow distribution.

[0082] Through the above ray tracing calculations, the shadow height of each point within the projection area of ​​each photovoltaic panel can be obtained, thereby determining the time of each photovoltaic panel. Shadow distribution Shadow coverage It can be calculated from the ratio of the shaded area to the projected area of ​​the photovoltaic panel:

[0083] in, Indicates area.

[0084] SS3 obtains cloud data based on meteorological forecast data, corrects the basic shadow coverage rate based on the cloud data to obtain the final shadow coverage rate, estimates the expected mismatch loss of each photovoltaic panel based on the final shadow coverage rate, and sums up the expected mismatch losses to obtain the total expected mismatch loss.

[0085] Obtain meteorological forecast data, including cloud distribution maps and cloud heights. Cloud movement speed and direction of movement .

[0086] The cloud distribution map is represented in a grid format, with each grid point containing information on cloud thickness or cloud cover. Based on the cloud movement speed and direction, the current cloud distribution map is projected onto future timeframes. Get the time Cloud distribution. For photovoltaic panels. For any point within the projection area, determine whether it is covered by cloud shadows based on the sun's position and cloud height, and then calculate the shadow coverage contributed by the clouds. .

[0087] The shadow distribution is corrected based on cloud data to obtain the final shadow coverage. :

[0088] The expected mismatch loss of each photovoltaic panel is estimated based on the shading coverage rate. The mismatch loss exhibits a non-linear relationship with the shading coverage rate, which is fitted using a quadratic function model:

[0089] Among them, linear terms The quadratic term reflects the direct power loss caused by shadow coverage. This reflects the additional mismatch loss caused by uneven shadow distribution. , These are empirical coefficients, obtained by fitting historical data.

[0090] The linear term represents the direct power loss caused by shading. When the shading coverage is small, the loss is approximately proportional to the coverage. The quadratic term represents the additional loss caused by uneven shading distribution. When the shading coverage is large, the spatial unevenness of the shading distribution intensifies, leading to more severe mismatch losses. Because when some cells of a photovoltaic panel are shaded, the bypass diodes may conduct, causing a sharp drop in the output power of the entire string. This effect becomes more pronounced after the shading coverage reaches a certain threshold. Furthermore, in a string, even if only one panel is partially shaded, the current of the entire string will be limited by that panel, preventing the power of the unshaded panels from being fully output. This causes the mismatch loss to increase non-linearly with the shading coverage. Simultaneously, for the same shading coverage, if the shading is concentrated in a localized area, the loss is often greater than with uniformly distributed shading. Therefore, the mismatch loss of a photovoltaic panel and the shading coverage are not a simple linear relationship; a quadratic function is used, with the quadratic term reflecting the additional losses caused by unevenness.

[0091] Calculate the total expected mismatch loss of the system :

[0092] SS4: If the total expected mismatch loss is greater than the preset loss threshold, it is determined that electrical reconfiguration needs to be performed. The shadow type is identified, the electrical reconfiguration strategy is determined according to the shadow type, and a control command is sent to the electrical switch matrix at a preset time before the predicted shadow occurs, driving the electrical switch matrix to execute the pre-generated reconfiguration strategy.

[0093] like If so, it is determined that a refactoring is required, where The preset mismatch loss threshold ranges from 0.1 to 0.2. This threshold can be determined based on the balance between the benefits and costs of reconfiguration: when the mismatch loss exceeds this threshold, the power generation loss that reconfiguration can recover is greater than the switching life loss and transient disturbance costs brought about by the reconfiguration itself, thus having a positive benefit.

[0094] In this embodiment, the type is identified through the following steps.

[0095] SS4.1, obtain the final shadow coverage sequence and spatial coordinates of each photovoltaic panel.

[0096] Final shade coverage of each photovoltaic panel ,in Number the photovoltaic panels. To predict the time, Spatial coordinates refer to the coordinate positions of each photovoltaic panel. In this embodiment, the geometric center point of each photovoltaic panel is selected as its spatial coordinate position, which is calculated from the vertex coordinates of the photovoltaic panel.

[0097] SS4.2, construct a spatial adjacency matrix between photovoltaic panels, and identify the connected regions formed by photovoltaic panels covered by shadows based on the spatial adjacency matrix and shadow coverage rate.

[0098] Construct a spatial adjacency matrix D between photovoltaic panels, where:

[0099] Criteria for determining spatial adjacency: The center-to-center distance between two photovoltaic panels is less than a preset threshold. Alternatively, two photovoltaic panels can be directly connected in structure.

[0100] For time Define the shading status of each photovoltaic panel:

[0101] in, The threshold for determining cluster shadows is 0.3-0.5.

[0102] A depth-first search algorithm is used, based on the adjacency matrix. and shadow coverage status Identify the connected regions formed by the photovoltaic panels covered by shadows:

[0103] Each connected region contains a continuous set of shaded photovoltaic panels.

[0104] SS4.3, when at least one connected region contains a number of photovoltaic panels greater than or equal to a preset cluster size threshold, and the average shading coverage of the photovoltaic panels in that region is greater than or equal to a preset average cluster threshold, and this state continues for a preset period of time. If all predicted times are true, the shadow type is determined to be a clustered shadow.

[0105] To be classified as a cluster shadow, conditions for size, coverage, and duration must be met simultaneously.

[0106] Size condition: There exists at least one connected component. The number of photovoltaic panels it contains ,in This is the minimum cluster size threshold, ranging from 3 to 5 blocks, with 4 blocks being the preferred value.

[0107] Coverage condition: Average shading coverage of photovoltaic panels within the connected area ,in:

[0108] This is the preset average threshold for the cluster.

[0109] Duration condition: The above conditions apply continuously All predicted times were true, among which The value range is 2-3 time points.

[0110] SS4.4 If the cluster shadow determination condition is not met, calculate the proportion of photovoltaic panels covered by shadow and the coefficient of variation of shadow coverage rate; when the proportion of photovoltaic panels is greater than or equal to a preset proportion threshold and the coefficient of variation is less than or equal to a preset coefficient threshold, the shadow type is determined to be scattered shadow.

[0111] Calculate the percentage of photovoltaic panels covered by shade:

[0112] in, For indicator functions, This is the minimum shadow coverage threshold.

[0113] Calculate the spatial uniformity index of shadow coverage, i.e., the coefficient of variation:

[0114] in:

[0115]

[0116] The smaller the coefficient of variation, the more uniform the shadow distribution; the larger the coefficient of variation, the more concentrated the shadow distribution.

[0117] To be classified as a moderate shadow, the following conditions must be met simultaneously: non-clustered, coverage area, and dispersion.

[0118] Non-cluster condition: The criteria for determining cluster shadow are not met.

[0119] Coverage conditions: Shading panel ratio Among them, the preset ratio threshold The value range is 0.2-0.3.

[0120] Dispersion condition: coefficient of variation This indicates that the shadow distribution is relatively uniform, where the preset coefficient threshold... The value range is 0.5-0.7.

[0121] In this embodiment, when the shadow type is clustered shadow, a uniformly distributed reconstruction strategy is adopted to evenly distribute the shaded photovoltaic panels into each string of the entire array. The optimization objective of this strategy is to balance the weighted output power of each string, and the objective function is expressed as:

[0122] in, This represents the total number of photovoltaic panels. The number of strings. For the first A string is a set of photovoltaic panels. For the first The ideal maximum power of a photovoltaic panel For the first The availability factor of the photovoltaic panels. The objective function aims to make the ratio of the weighted output power of each string to the total weighted output power of the system as close as possible to 1 / N, thereby achieving power balance among the strings.

[0123] The constraints for solving the objective function include: each photovoltaic panel must be assigned to one string, the number of photovoltaic panels in each string is within a reasonable range, and the photovoltaic panels in the same string should have the same or similar orientation to avoid long-term mismatch due to different orientations.

[0124] A genetic algorithm is used to solve the objective function, obtain the optimal topology, and then perform reconstruction.

[0125] Integer encoding is used, with each individual being an array of length M. ,in Indicates the first The string number assigned to each photovoltaic panel.

[0126] The fitness function is:

[0127] in, It should be a small positive number to avoid division by zero.

[0128] Genetic algorithms include the following steps: Step U1: Initialize the population by randomly generating P individuals; Step U2: Calculate the fitness value for each individual; Step U3: Select individuals with high fitness values ​​to enter the next generation (tournament selection or roulette selection); Step U4: Perform a crossover operation on the selected individuals, using a two-point crossover; Step U5: Perform mutation operation on the crossover individuals to randomly change the string allocation of some photovoltaic panels; Step U6: Check if the constraints are met; if not, make corrections. Step U7: Repeat steps U2-U6 until the maximum number of iterations is reached or the fitness value converges; Step U8: Output the individual with the highest fitness value as the optimal topology.

[0129] In this embodiment, when the shadow type is moderate shadow, a dynamic series-parallel reconstruction strategy is adopted, based on the final shadow coverage of each photovoltaic panel. With preset threshold , Based on the comparison results, the photovoltaic panels were divided into three levels: light shading, moderate shading, and heavy shading, and were processed separately.

[0130] 1) For A lightly shaded panel, maintaining the original topology.

[0131] The impact of the shadow is relatively small, and the benefits of the adjustment are insufficient to offset the reconstruction costs. Therefore, the original topology is maintained and no adjustments are made.

[0132] 2) For The medium shaded panels are redistributed to different groups of strings to balance the number of medium shaded panels in each group.

[0133] Identify all medium shaded panels and denote them as the set. Count the number of moderately shaded panels in each current string group. If a certain string exists of Then, some of the moderately shaded panels in that string are swapped with the lightly shaded or normal panels in other strings. The swapping principle is to make the number of moderately shaded panels in each string as balanced as possible, that is:

[0134] in, This represents the number of strings.

[0135] 3) For For heavily shaded panels, bypass them or connect them to a separate optimizer.

[0136] Bypassing refers to using an electrical switch to bypass the heavily shaded photovoltaic panel, removing it from the series circuit and preventing it from affecting other photovoltaic panels in the same string. After bypassing, the other photovoltaic panels in the string can still operate normally.

[0137] Connecting to a standalone optimizer refers to separating heavily shaded panels from their original string and connecting them to a separate module-level power optimizer to achieve independent MPPT control. The standalone optimizer allows heavily shaded panels to operate at a lower power point without affecting other photovoltaic panels.

[0138] When a photovoltaic panel meets multiple conditions simultaneously, it is processed with priority given to heavy shading, followed by moderate shading, and finally light shading.

[0139] It should be noted that if it is neither a cluster shadow nor a scatter shadow, it is determined that no reconstruction is needed and the current topology remains unchanged.

[0140] A string is the basic electrical unit in a photovoltaic (PV) power generation system, referring to a DC power generation unit formed by connecting several PV modules (PV panels) in series. When the performance of PV panels within a string is inconsistent (due to uneven illumination, different aging levels, shading, etc.), mismatch losses occur. Specifically, poor-performing panels limit the current of the entire string, while high-performing panels cannot fully utilize their power generation capacity, and excess power is dissipated as heat on the poor-performing panels. To mitigate the mismatch problem, PV panels incorporate bypass diodes. When a panel is severely shaded, the bypass diode conducts, bypassing that panel and allowing current to flow around it, preventing the entire string from failing. In this embodiment, dynamic string reconfiguration is achieved through an electrical switch matrix. This means that the series-parallel relationship between PV panels can be adjusted in real time according to the shading distribution; the same PV panel may belong to different strings at different times; and through reconfiguration, shaded panels are distributed to different strings, avoiding multiple shaded panels connected in series. The composition of strings is not limited to the same orientation; before reconfiguration, they may be grouped by orientation, and after reconfiguration, panels with different orientations can be mixed into strings to achieve power balance.

[0141] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A three-dimensional photovoltaic array, comprising multiple photovoltaic module units, characterized in that, The photovoltaic module unit includes: a west-facing photovoltaic panel, an east-facing photovoltaic panel, a southeast-facing photovoltaic panel, and a southwest-facing photovoltaic panel. The east-facing photovoltaic panel is arranged opposite to the west-facing photovoltaic panel, the southeast-facing photovoltaic panel is arranged in front of the east-facing photovoltaic panel, and the southwest-facing photovoltaic panel is arranged in front of the west-facing photovoltaic panel. Each photovoltaic panel is equipped with a module-level power optimizer and a light sensor; the module-level power optimizer is electrically connected to the corresponding photovoltaic panel and is used to adjust the operating point of the photovoltaic panel; the light sensor is installed on the back of the corresponding photovoltaic panel and is used to collect the back irradiance data of the corresponding photovoltaic panel in real time. The module-level power optimizer and the light sensor are electrically connected to a four-way MPPT co-controller. The four-way MPPT co-controller is configured to send control commands to each module-level power optimizer to coordinately adjust the output power of each photovoltaic panel based on the current time period and the equivalent front irradiance of each photovoltaic panel. The equivalent front irradiance is calculated from the corresponding back irradiance data.

2. The three-dimensional photovoltaic array according to claim 1, characterized in that, It also includes an electrical switch matrix and a shadow dynamic reconfiguration controller; An electrical switch matrix, whose input terminal is electrically connected to the output terminal of each component-level power optimizer and whose control terminal is signal-connected to the output terminal of the shadow dynamic reconfiguration controller, is used to dynamically switch the series-parallel connection topology between photovoltaic panels in response to control commands sent by the shadow dynamic reconfiguration controller. The shadow dynamic reconstruction controller is connected to each light sensor and the electrical switch matrix signal respectively, and is configured to send control commands to the electrical switch matrix to control the electrical switch matrix to perform electrical reconstruction based on the predicted shadow distribution.

3. The three-dimensional photovoltaic array according to claim 2, characterized in that, The shadow dynamic reconfiguration controller is electrically connected to the four-way MPPT co-controller. After performing electrical reconfiguration, the shadow dynamic reconfiguration controller sends an update command to the four-way MPPT co-controller. In response to the update command, the four-way MPPT co-controller reacquires the maximum available power of each photovoltaic panel and recalculates the target power of each photovoltaic panel.

4. The three-dimensional photovoltaic array according to any one of claims 1 to 3, characterized in that, Both the west-facing and east-facing photovoltaic panels are rectangular structures, with an angle of 60°±3° between them and the ground; the southeast-facing and southwest-facing photovoltaic panels are triangular structures.

5. A power generation control method for a three-dimensional photovoltaic array, characterized in that, The method applied to the three-dimensional photovoltaic array according to any one of claims 1 to 4 includes a step of four-way MPPT cooperative control, specifically including: Get the current time period; Based on the pre-stored time period-orientation weight model, determine the weight coefficients of west-facing photovoltaic panels, east-facing photovoltaic panels, southeast-facing photovoltaic panels, and southwest-facing photovoltaic panels in the current time period; The back irradiance data of each photovoltaic panel is converted into equivalent front irradiance. Based on the equivalent front irradiance and the actual output power of the photovoltaic panel, the maximum usable power of each photovoltaic panel under the current irradiance is estimated. Calculate the target power of each photovoltaic panel based on the global target power and the weighting coefficient of each photovoltaic panel; Control commands are sent to the module-level power optimizer of each photovoltaic panel according to the target power, and the operating point of each photovoltaic panel is adjusted.

6. The power generation control method according to claim 5, characterized in that, The following formula is used to convert the back irradiance data into the equivalent front irradiance: In the formula, For the first A solar panel in Equivalent positive irradiance at any given moment. For the first The actual output power of the photovoltaic panel. For the first The bifaciality of a photovoltaic panel For the first Irradiance measured by sensors on the back of a photovoltaic panel. For the first The area of ​​a photovoltaic panel.

7. The power generation control method according to claim 5, characterized in that, The method also includes a step of dynamic shadow reconstruction, specifically including: Based on the three-dimensional geometric model of the three-dimensional photovoltaic array and the solar position parameters at future times, the ray tracing algorithm is used to calculate the shadow distribution of each photovoltaic panel at each time within a preset future time period. Based on the predicted shadow distribution, calculate the basic shadow coverage rate of each photovoltaic panel at the corresponding time. Cloud data is obtained from meteorological forecast data. The basic shadow coverage rate is corrected based on the cloud data to obtain the final shadow coverage rate. The expected mismatch loss of each photovoltaic panel is estimated based on the final shadow coverage rate. The total expected mismatch loss is obtained by summing the expected mismatch losses. If the total expected mismatch loss is greater than the preset loss threshold, it is determined that electrical reconfiguration needs to be performed. The shadow type is identified, the electrical reconfiguration strategy is determined based on the shadow type, and a control command is sent to the electrical switch matrix at a preset time before the predicted shadow occurs, driving the electrical switch matrix to execute the pre-generated reconfiguration strategy.

8. The power generation control method according to claim 7, characterized in that, Identifying shadow types specifically includes: Obtain the final shadow coverage sequence and spatial coordinates of each photovoltaic panel; Construct a spatial adjacency matrix between photovoltaic panels, and identify the connected regions formed by photovoltaic panels covered by shadows based on the spatial adjacency matrix and shadow coverage rate; When at least one connected region contains a number of photovoltaic panels greater than or equal to a preset cluster size threshold, and the average shading coverage of the photovoltaic panels in that region is greater than or equal to a preset average cluster threshold, and this state is maintained for a continuous preset period... If all predicted times are true, the shadow type is determined to be a clustered shadow. If the cluster shadow determination condition is not met, calculate the proportion of photovoltaic panels covered by shadow and the coefficient of variation of shadow coverage rate; when the proportion of photovoltaic panels is greater than or equal to a preset proportion threshold and the coefficient of variation is less than or equal to a preset coefficient threshold, the shadow type is determined to be scattered shadow.

9. The power generation control method according to claim 8, characterized in that, When the shading type is clustered shading, a uniformly distributed reconstruction strategy is adopted to evenly distribute the shaded photovoltaic panels into each string of the entire array. The optimization objective of this strategy is to balance the weighted output power of each string, and the objective function is expressed as: in, This represents the total number of photovoltaic panels. The number of strings. For the first Each string contains a set of photovoltaic panels. For the first The ideal maximum power of a photovoltaic panel For the first The availability factor of a photovoltaic panel is determined by the expected mismatch loss; A genetic algorithm is used to solve the objective function, obtain the optimal topology, and then perform reconstruction.

10. The power generation control method according to claim 8, characterized in that, When the shading type is moderate shading, a dynamic series-parallel reconfiguration strategy is adopted, based on the final shading coverage of each photovoltaic panel. With preset threshold , Based on the comparison results, photovoltaic panels were classified into three levels: light shading, moderate shading, and heavy shading, and then processed accordingly: for A lightly shaded panel, maintaining the original topology; for The moderately shaded boards are redistributed to different groups of strings to balance the number of moderately shaded boards in each group. The distribution and exchange principle is expressed as follows: in, This represents the number of moderately shaded panels within each current string group. For all identified sets of moderately shaded panels, This represents the number of strings. for For heavily shaded panels, bypass them or connect them to a separate optimizer.

Images