A retractable solar roof power supply system for a vehicle and a control method thereof

Abstract

The present application relates to the technical field of vehicle power supply, and discloses a telescopic solar roof power supply system for vehicle and a control method thereof, which comprises a parking monitoring module that acquires gear state and battery state of charge, generates a hardware interrupt signal when the theoretical power generation is greater than the system power consumption, a pose compensation module that converts static point cloud matrix and correction coordinate data into real point cloud matrix based on parking pose parameters, a target solving module that calculates and outputs feedforward target stroke in combination with a shadow boundary function and an electrical characteristic attenuation model, a closed-loop control module that controls a stepping motor, calculates pressure gradient and power gradient to determine safe stroke and correction coordinate data, and a dynamic adjustment module that performs hysteresis maintenance and generates forced stretching instructions based on actual ambient light power. The present application avoids invalid energy consumption and spatial geometric deviation caused by blind stretching, effectively avoids local shadow and physical interference, and utilizes reflected light field to improve net energy efficiency integral and safety.

Description

Technical Field

[0001] This invention relates to the field of vehicle power supply technology, specifically to a retractable solar roof power supply system for vehicles and its control method. Background Technology

[0002] With the development of new energy vehicles, using rooftop solar power generation devices to supplement vehicle power has become an effective way to extend driving range. To increase the area exposed to sunlight, retractable solar roofs are increasingly being used. However, existing retractable solar power systems for vehicles still have shortcomings in real-world parking scenarios. In terms of control logic, most existing systems directly trigger the retractable actuator to extend after the vehicle is parked, lacking a pre-emptive energy efficiency assessment mechanism. In low-light environments such as underground parking garages or severely shaded areas, the power consumption of the control chip and drive motor often exceeds the actual power generation of the photovoltaic modules. Unconditional extension can lead to ineffective power loss or even battery depletion.

[0003] Regarding the accuracy of lighting positioning, conventional solar roof control strategies typically assume the vehicle is parked on an absolutely level surface, directly calculating light and shadow based on data from onboard sensors. In actual parking scenarios, road slopes and unevenness cause the vehicle to pitch and tilt, resulting in geometric projection deviations in the environmental perception data. Because the spatial offset caused by the parking posture is not corrected, and there is a lack of optimization mechanisms to avoid local shadows, the deployed photovoltaic modules can easily fall partially into the shaded area, leading to a reduction in the overall system energy efficiency.

[0004] In terms of safety and utilization of ambient light fields, existing systems typically rely on a single parameter for anti-pinch judgment during mechanical extension. This makes them unable to effectively identify small obstacles or hidden shadows that are missed by visual sensors, easily leading to physical interference with the mechanical structure or a sudden drop in electrical efficiency. Furthermore, existing control models usually calculate daylighting benefits based solely on direct sunlight, ignoring secondary reflected light fields generated by surfaces such as glass curtain walls of urban buildings. This makes it difficult for the system to utilize additional lighting resources in complex environments. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a retractable solar roof power supply system for vehicles and its control method. It solves the problems of existing technologies, such as lack of parking energy efficiency assessment, failure to correct geometric deviations caused by parking posture, inability to effectively deal with hidden interference and complex reflected light fields, easy power battery depletion, low accuracy of light-gathering optimization, and poor operational safety.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a retractable solar roof power supply system for vehicles, including a parking monitoring module, a pose compensation module, a target solving module, a closed-loop control module, and a dynamic adjustment module; The parking monitoring module is used to obtain the gear status and battery charge status to generate a system activation signal, obtain light intensity data, time data and latitude and longitude data to estimate the theoretical power generation, and generate a hardware interrupt signal when the theoretical power generation exceeds the sum of the chip startup power consumption and the motor operation power consumption. The pose compensation module is used to obtain parking posture parameters and use the spatial rotation matrix calculation formula to convert the obtained static point cloud matrix and corrected coordinate data into a real point cloud matrix. The target solution module generates a shadow boundary function based on the real point cloud matrix, and uses the feedforward target travel calculation formula to obtain the target travel sequence in combination with the electrical characteristic attenuation model. When the travel variance of the target travel sequence meets the condition, the feedforward target travel is output. The closed-loop control module controls the stepper motor based on the feedforward target stroke, calculates the pressure gradient and power gradient, performs electromechanical parallel judgment to generate a safe stroke, extracts coordinate data to generate corrected coordinate data and feeds it back to the pose compensation module. The dynamic adjustment module compares the latest feedforward target travel with the safe travel and intervenes based on the theoretical power generation and the actual ambient light power to generate a forced extension command.

[0007] Furthermore, the parking monitoring module includes a state activation unit, a revenue estimation unit, and a wake-up decision unit. The state activation unit generates a system activation signal when the vehicle is shifted into parking mode and the battery is not fully charged, in order to control the microcontroller to acquire light intensity data, time data, and latitude and longitude data. The revenue forecasting unit uses a low-order geometric model combined with light intensity data, time data, and latitude and longitude data to forecast the theoretical power generation. The wake-up decision unit generates a hardware interrupt signal when the theoretical power generation is greater than the sum of the chip startup power consumption and the motor operation power consumption, in order to wake up the SoC chip and sense the hardware to enter the working state; otherwise, it generates a sleep command to control the telescopic actuator to fully retract.

[0008] Furthermore, the pose compensation module includes an attitude extraction unit and a spatial mapping unit. The attitude extraction unit acquires parking attitude parameters, including yaw angle, pitch angle, and roll angle, from the chassis inertial sensors, and derives the spatial rotation matrix using the spatial rotation matrix calculation formula; The spatial mapping unit calls the sensing hardware to obtain the static point cloud matrix with the vehicle body as the origin. After combining the static point cloud matrix with the corrected coordinate data, it multiplies it by the spatial rotation matrix on the left and outputs the real point cloud matrix using the real point cloud matrix calculation formula. In the initial state, the corrected coordinate data is a zero matrix.

[0009] Furthermore, the objective solving module includes a shadow projection unit, a transient optimization unit, and a steady-state output unit. The shadow projection unit analyzes the sunlight vector based on time and latitude / longitude data, and generates a shadow boundary function by projecting the real point cloud matrix along the direction of the sunlight vector using a ray tracing algorithm. Within a set time window, the transient optimization unit combines the shadow boundary function at the corresponding moment with the topological mapping relationship of the photovoltaic module and the electrical characteristic attenuation model, and uses the feedforward target travel calculation formula to obtain multiple transient target travels to generate a target travel sequence. The steady-state output unit calculates the travel variance of the target travel sequence. If it is greater than the preset tolerance threshold, an action rejection command is generated. If it is less than or equal to the preset tolerance threshold, the average value is taken to generate the feedforward target travel. The preset tolerance threshold is preset based on the mechanical gear backlash of the telescopic actuator and the allowable motion jitter amplitude of the system.

[0010] Furthermore, the steady-state output unit is also used to generate a termination extension command when the net energy efficiency integral corresponding to the feedforward target stroke is negative.

[0011] Furthermore, the closed-loop control module includes a deceleration buffer unit, a gradient monitoring unit, and a redundant interrupt unit. The deceleration buffer unit generates a motor drive signal based on the feedforward target stroke to control the stepper motor. When the real-time physical stroke obtained by the motor encoder is greater than or equal to the difference between the feedforward target stroke and the safety buffer distance, a micro-stepping trigger signal is generated to control the stepper motor to decelerate. After acquiring the microstepping trigger signal, the gradient monitoring unit obtains the instantaneous pressure value and instantaneous output power. Based on the real-time physical stroke, it calculates the pressure gradient and power gradient using the transient pressure gradient calculation formula and the transient power gradient calculation formula. The redundant interrupt unit performs electromechanical parallel determination based on the pressure gradient and power gradient, locks the stepper motor to generate a safe stroke, and extracts the corresponding coordinate data to generate corrected coordinate data. The safety buffer distance is preset based on the rotor inertia of the stepper motor and the friction braking distance of the telescopic actuator.

[0012] Furthermore, the logic for the redundant interrupt unit to perform electromechanical parallel determination specifically includes: if the pressure gradient is greater than the set collision physical threshold, a forced retraction command is generated to lock the stepper motor. If the pressure gradient is less than or equal to the set collision physical threshold and the power gradient is less than the set tolerance negative step, an electrical interrupt command is generated to lock the stepper motor. If neither of the above two conditions applies, control the stepper motor to extend smoothly; The collision physical threshold is preset based on the yield strength of the photovoltaic module frame material and the compressive strength limit of the mechanical structure, and the negative step of the tolerance is preset based on the power ripple fluctuation range of the MPPT controller during the optimization process.

[0013] Furthermore, the dynamic adjustment module includes a hysteresis control unit and a gain intervention unit. The hysteresis control unit calculates the absolute difference between the latest feedforward target stroke and the safety stroke. After the absolute difference is greater than the set deadband width and is maintained for a set time period, it generates a fine-tuning permission command to control the stepper motor to perform mechanical adjustment actions. The gain intervention unit calculates the expected shadow power based on the theoretical power generation through a theoretical calculation model. When the actual ambient light power, represented by the actual light intensity data, is greater than the sum of the expected shadow power and the set reflective gain threshold, it generates a forced extension command and a projection shielding command. The dead zone width is preset based on the physical width of a single cell in the photovoltaic module and the mechanical transmission clearance. The time period is preset based on the solar offset angular velocity and the average cloud movement period. The reflectivity gain threshold is preset based on the reflectivity of the building glass and the minimum effective starting power of the photovoltaic module.

[0014] Furthermore, the dynamic adjustment module also includes a system reset unit. The system reset unit generates a system reset command upon receiving a gear disengagement signal or a vehicle unlock signal.

[0015] A second aspect of the present invention provides a power supply control method for a retractable solar roof for vehicles, applied to the aforementioned power supply system for a retractable solar roof for vehicles, comprising the following steps: The system uses a parking monitoring module to obtain gear status and battery charge status to generate a system activation signal, obtains light intensity data, time data and latitude and longitude data to estimate theoretical power generation, and generates a hardware interrupt signal when the theoretical power generation exceeds the sum of chip startup power consumption and motor operation power consumption. The parking attitude parameters are obtained using the pose compensation module and the static point cloud matrix and corrected coordinate data are converted into a real point cloud matrix using the spatial rotation matrix calculation formula. The target solution module generates a shadow boundary function based on the real point cloud matrix. Combined with the electrical characteristic attenuation model, the target travel sequence is obtained by using the feedforward target travel calculation formula. The feedforward target travel is output when the travel variance of the target travel sequence meets the condition. The closed-loop control module controls the stepper motor based on the feedforward target stroke, calculates the pressure gradient and power gradient to perform electromechanical parallel judgment to generate a safe stroke, and extracts coordinate data to generate corrected coordinate data to feed back to the pose compensation module. The dynamic adjustment module compares the latest feedforward target travel with the safe travel, and intervenes based on the theoretical power generation and the actual ambient light power to generate a forced extension command.

[0016] This invention provides a retractable solar roof power supply system for vehicles and its control method. It possesses at least one of the following beneficial effects: 1. This invention uses the acquired light intensity data, time data and latitude and longitude data through the parking monitoring module to obtain the theoretical power generation. Only when the theoretical power generation exceeds the sum of the chip startup power consumption and the motor operation power consumption will a hardware interrupt signal be generated to wake up the sensing hardware, so as to avoid the ineffective energy consumption caused by blindly driving the telescopic actuator when the vehicle is parked in an obstructed environment. 2. This invention utilizes a pose compensation module combined with parking attitude parameters output by the chassis inertial sensor to perform spatial mapping transformation on the static point cloud matrix, eliminating spatial geometric deviations caused by road tilt. At the same time, it utilizes a target solving module combined with the real point cloud matrix and the photovoltaic module electrical characteristic attenuation model to perform optimization calculations, enabling the system to avoid power reduction areas caused by local shadows and improve the net energy efficiency integral in complex three-dimensional light and shadow environments. 3. This invention extracts the instantaneous pressure value of the tactile sensor and the instantaneous output power of the MPPT controller through a closed-loop control module to calculate the pressure gradient and power gradient. In the event of physical collision or entering a hidden shadow, it can lock the safe travel distance through parallel judgment, ensuring the safe operation of the telescopic actuator. Furthermore, by comparing the actual ambient light power with the expected shadow power through a dynamic adjustment module, it can identify and utilize the reflected light field of building glass curtain walls, etc., thus expanding the actual lighting range of the system. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a retractable solar roof power supply system for vehicles according to the present invention; Figure 2 This is a flowchart of a retractable solar roof power supply control method for vehicles according to the present invention; Figure 3 This is an architecture diagram of the parking monitoring module of the present invention; Figure 4 This is an architectural diagram of the pose compensation module of the present invention; Figure 5 This is an architecture diagram of the target solving module of the present invention; Figure 6 This is an architecture diagram of the closed-loop control module of the present invention; Figure 7 This is an architectural diagram of the dynamic adjustment module of the present invention; Figure 8 This is a diagram showing the coupling relationship between the system's net benefit and the extension stroke in an application embodiment of the present invention; Figure 9 This is a schematic diagram of the retractable solar roof structure of the present invention.

[0018] Among them, 1. Photovoltaic modules; 2. Stepper motors. Detailed Implementation

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see the appendix Figure 1 and attached Figure 9 This invention provides a retractable solar roof power supply system for vehicles, including a parking monitoring module, a pose compensation module, a target solving module, a closed-loop control module, and a dynamic adjustment module.

[0021] The retractable solar roof includes a photovoltaic module 1 and a telescopic actuator for driving the photovoltaic module 1 to move. The telescopic actuator consists of a linear guide rail, a stepper motor 2 as a power source, and a transmission assembly; the transmission assembly adopts a mechanical configuration such as a lead screw, rack and pinion, or synchronous belt. The stepper motor 2 drives the photovoltaic module 1 to move linearly along the linear guide rail through the transmission assembly, thereby realizing the adjustment of the sun-receiving area of ​​the roof.

[0022] The system operates based on the vehicle's underlying hardware architecture, which includes a hierarchical computing unit comprising a microcontroller and a SoC chip; a perception side comprising onboard sensors, light intensity sensors, chassis inertial sensors, and perception hardware including onboard cameras and LiDAR; and an execution and energy side comprising a motor encoder, tactile sensors, and an MPPT controller. The microcontroller and SoC chip work together to acquire front-end perception data, thereby controlling the stepper motor 2 to drive the photovoltaic module 1 to physically extend and retract.

[0023] The parking monitoring module is used to collect basic data and predict energy efficiency while the vehicle is parked. This module acquires the vehicle's underlying gear position and battery state of charge, and generates a system activation signal when the vehicle is in parking gear and the battery is not fully charged.

[0024] Subsequently, the microcontroller maintains a resident monitoring state, continuously acquiring light intensity data, time data, and latitude and longitude data. The parking monitoring module further estimates the theoretical power generation using a low-order geometric model and calculates the difference between the theoretical power generation and the sum of the chip's startup power consumption and the motor's operating power consumption. When the difference is greater than zero, the parking monitoring module generates a hardware interrupt signal to wake up the SoC chip and sensing hardware; conversely, when the difference is less than or equal to zero, a sleep command is generated to instruct the retractable actuator to maintain its retraction.

[0025] The pose compensation module is used to correct spatial coordinate offsets caused by the vehicle's physical parking posture after the SoC chip is woken up. This module first acquires parking posture parameters from the chassis inertial sensors and calculates the spatial rotation matrix using the formula. Next, it calls the sensing hardware to acquire the static point cloud matrix and receives corrected coordinate data from the preceding process. Finally, the pose compensation module combines the static point cloud matrix with the corrected coordinate data and uses the formula for calculating the true point cloud matrix to obtain the true point cloud matrix, reflecting the relative positions of obstacles around the vehicle in the real physical world coordinate system.

[0026] The objective solving module performs 3D light and shadow mapping and solves for the optimal net energy efficiency feedforward travel. Based on time and latitude / longitude data, this module derives the solar vector and projects the real point cloud matrix along the solar vector to generate a shadow boundary function. Simultaneously, it extracts the topological mapping relationship and electrical characteristic degradation model of photovoltaic module 1, calculates the transient target travel within a time window using the feedforward target travel calculation formula, and generates a target travel sequence. Subsequently, it calculates the travel variance of the target travel sequence, generating an action rejection command if the travel variance exceeds a tolerance threshold. If the travel variance is less than or equal to the tolerance threshold, it generates a feedforward target travel and generates a termination extension command when the net energy efficiency integral corresponding to the feedforward target travel is negative.

[0027] The closed-loop control module is used to drive the telescopic actuator and combines electromechanical dual-path feedback for redundant control. This closed-loop control module first generates a motor drive signal based on the feedforward target stroke, controlling the operation of the stepper motor 2 of the telescopic actuator.

[0028] During operation, the real-time physical travel is acquired through the motor encoder. When the real-time physical travel is greater than or equal to the difference between the feedforward target travel and the safety buffer distance, a micro-step trigger signal is generated. Upon receiving this signal, the closed-loop control module synchronously acquires the instantaneous pressure value of the tactile sensor and the instantaneous output power of the MPPT controller. The pressure gradient is calculated using the transient pressure gradient calculation formula, and the power gradient is calculated using the transient power gradient calculation formula. Then, the electromechanical parallel system determines the negative step of the pressure gradient relative to the collision physical threshold and the power gradient relative to the tolerance, thereby generating a forced retraction command or an electrical interrupt command to lock the stepper motor 2. Finally, after locking the stepper motor 2, a safety travel is generated, and the corresponding coordinate data is extracted to generate corrected coordinate data and fed back to the pose compensation module.

[0029] The dynamic adjustment module is used to perform long-cycle system status updates and interventions. This module acquires the safe travel distance and calculates the absolute difference between the feedforward target travel distance and the safe travel distance. When the absolute difference exceeds the dead zone width and remains unchanged for a set time period, a fine-tuning permission command is generated.

[0030] The dynamic adjustment module acquires actual light intensity data from a light intensity sensor and calculates the expected shadow power based on the theoretical power generation using a theoretical calculation model. When it is determined that the actual ambient light power, as represented by the actual light intensity data, is greater than the sum of the expected shadow power and the reflective gain threshold, a forced extension command and a projection shielding command are generated. Additionally, upon receiving a gear disengagement signal or a vehicle unlock signal, the dynamic adjustment module generates a system reset command, controlling the telescopic actuator to fully retract.

[0031] See appendix Figure 2 In this embodiment, the present invention also provides a power supply control method for a retractable solar roof for vehicles, comprising the following steps: S1 utilizes the parking monitoring module to receive the gear position and battery state of charge from the vehicle's underlying network. When it is determined that the gear is in parking position and the battery is not fully charged, the parking monitoring module generates a system activation signal. Subsequently, the parking monitoring module puts the SoC chip into a deep sleep state and starts the microcontroller's resident monitoring mode, whereby the microcontroller continuously acquires light intensity data, time data, and latitude and longitude data from the vehicle's sensors.

[0032] After acquiring data, the parking monitoring module estimates the theoretical power generation using a pre-set low-order geometric model. The module then calculates the difference between this theoretical power generation and the sum of the chip's startup power consumption and the motor's operating power consumption. If this difference is greater than zero, a hardware interrupt signal is generated to wake up the SoC chip and sensing hardware (including the vehicle-mounted camera and LiDAR) to enter working mode; if the difference is less than or equal to zero, a sleep command is generated to control the telescopic actuator to remain fully retracted.

[0033] S2 utilizes the pose compensation module to perform pose compensation correction for spatial coordinate offsets after the SoC chip is woken up. The pose compensation module obtains parking attitude parameters from the chassis inertial sensors, including yaw angle, pitch angle, and roll angle. Based on the parking attitude parameters, the pose compensation module calculates the spatial rotation matrix using the spatial rotation matrix calculation formula.

[0034] The pose compensation module calls the sensing hardware to obtain a static point cloud matrix with the vehicle body as the origin, and obtains the corrected coordinate data fed back from subsequent processes. At this stage, the corrected coordinate data in the initial state is set to a zero matrix. After combining the static point cloud matrix with the corrected coordinate data, the pose compensation module multiplies it by a spatial rotation matrix on the left, and uses the real point cloud matrix calculation formula to output a real point cloud matrix mapped to the real physical world coordinate system.

[0035] S3 utilizes the objective solver module to perform optimization calculations for light and shadow occlusion. The objective solver module acquires the real point cloud matrix, time data, and latitude and longitude data, and derives the solar radiation vector. Using a ray tracing algorithm, the objective solver module projects the real point cloud matrix along the direction of the solar radiation vector onto a two-dimensional physical plane, generating a shadow boundary function. The shadow boundary function is then obtained, and the topological mapping relationship and electrical characteristic degradation model of photovoltaic module 1 are extracted.

[0036] Within a set time window, the target solving module combines the shadow boundary function and electrical characteristic attenuation model at the corresponding time point, and uses the feedforward target travel calculation formula to derive the transient target travel. Multiple transient target travels are then combined to form a target travel sequence. The target solving module acquires the target travel sequence and calculates its travel variance. If the travel variance is greater than a preset tolerance threshold, an action rejection command is generated; if the travel variance is within the tolerance threshold range, the average is taken to generate the final feedforward target travel. If the net energy efficiency integral corresponding to the feedforward target travel is negative, a termination extension command is generated.

[0037] S4 utilizes a closed-loop control module to perform controlled extension and real-time feedback monitoring of the mechanical structure. The closed-loop control module acquires the feedforward target stroke and generates a motor drive signal to control the stepper motor 2 of the telescopic actuator to run at full speed. The closed-loop control module acquires the real-time physical stroke through a motor encoder. When the real-time physical stroke is greater than or equal to the difference between the feedforward target stroke and the safety buffer distance, a micro-step trigger signal is generated to control the stepper motor 2 to decelerate.

[0038] After receiving the micro-stepping trigger signal, the closed-loop control module obtains the instantaneous pressure value from the tactile sensor and the instantaneous output power from the MPPT controller. Based on the real-time physical travel, the pressure gradient is calculated using the transient pressure gradient calculation formula, and the power gradient is calculated using the transient power gradient calculation formula.

[0039] The closed-loop control module acquires the pressure gradient and power gradient and performs parallel judgments. If the pressure gradient is greater than the collision physical threshold, a forced retraction command is generated to lock stepper motor 2; if the pressure gradient is less than or equal to the collision physical threshold and the power gradient is less than the tolerance negative step, an electrical interruption command is generated to lock stepper motor 2. When neither of these two conditions is met, stepper motor 2 extends smoothly. Subsequently, the closed-loop control module generates a safe travel distance from the final stopping position and extracts coordinate data to generate corrected coordinate data, which is then fed back to the pose compensation module.

[0040] S5 utilizes the dynamic adjustment module to perform long-cycle state maintenance and intervention. The dynamic adjustment module acquires the safe travel distance and the latest feedforward target travel distance, calculates the absolute difference between the latest feedforward target travel distance and the safe travel distance, and generates a fine-tuning permission command after the absolute difference is greater than the dead zone width and maintained for a set time period.

[0041] The dynamic adjustment module continuously acquires actual light intensity data from the light intensity sensor and calculates the expected shadow power based on the theoretical power generation using a theoretical calculation model. If the actual ambient light power exceeds the sum of the expected shadow power and the reflective gain threshold, the dynamic adjustment module generates a forced extension command and a projection shielding command. The system detects and acquires the vehicle's underlying gear engagement signal or vehicle unlock signal. Upon receiving these signals, a system reset command is directly generated to control the telescopic actuator to fully retract.

[0042] The technical solutions in the embodiments of the present invention will be described in detail below: See appendix Figure 3 In this embodiment, the parking monitoring module includes a status activation unit, a revenue estimation unit, and a wake-up decision unit.

[0043] As the initial triggering element of the system, the state activation unit is used to perform basic state identification and power consumption management in the early stages of the system.

[0044] Specifically, the state activation unit receives the gear position and battery charge status from the vehicle's underlying system via the vehicle local area network bus. When it is determined that the gear position is engaged in parking and the battery charge status is not fully charged, a system activation signal is generated.

[0045] As a preferred approach, the determination of incomplete charge here depends on a set full charge threshold, which is set according to the chemical characteristics of the power battery and the overcharge protection strategy, for example, it can be set to SOC reaching 95% or 98%.

[0046] After generating the system activation signal, the state activation unit defaults to placing the SoC chip into a deep sleep state, with the microcontroller maintaining constant monitoring. This mechanism allows for the disconnection of power supply circuits for high-performance computing components before positive energy gains are confirmed, effectively reducing initial static power consumption. During constant monitoring, the microcontroller continuously acquires light intensity, time, and latitude / longitude data from the vehicle's sensors.

[0047] For the process of a microcontroller acquiring low-level sensor parameters via a bus, those skilled in the art can use standard CAN bus or LIN bus communication protocols to parse data messages. The specific data transmission and decoding mechanisms are well-known technologies in the field and will not be elaborated here.

[0048] The revenue forecasting unit establishes a data connection with the state activation unit to assess the photovoltaic power generation potential of the current environment with low computing power overhead.

[0049] This revenue estimation unit acquires light intensity data, time data, and latitude and longitude data transmitted from the microcontroller, and estimates the theoretical power generation capacity using a pre-set low-order geometric model. It should be noted that this low-order geometric model differs from the high-order spatial model that relies on 3D point clouds; its internal calculation process mainly includes: Based on time and latitude / longitude data, the solar altitude angle and azimuth angle of the vehicle's current location are calculated using existing astronomical calendar algorithms. Subsequently, combined with the current ambient irradiance characterized by light intensity data, the external irradiance is projected along the solar altitude angle onto the horizontal physical area of ​​the photovoltaic module 1 in its fully deployed state (for example, by multiplying the irradiance value by the sine of the solar altitude angle to obtain the equivalent irradiance of the photovoltaic panel's horizontal surface), and then multiplied with the photoelectric conversion efficiency coefficient to obtain the theoretical power generation.

[0050] This operation involves only scalar multiplication and addition and basic trigonometric function operations, making it suitable for operation under the computing power limitations of microcontrollers. It also provides data support for subsequent wake-up actions.

[0051] Based on the above theoretical power generation, the wake-up decision unit is used to assess the net energy efficiency boundary and control the start-up and shutdown of the system-level hardware.

[0052] Specifically, the wake-up decision unit calculates the difference between the theoretical power generation and the sum of the chip startup power consumption and the motor operation power consumption. In this embodiment, the chip startup power consumption is calculated and set based on the steady-state operating current and nominal voltage of the SoC chip, for example, a value of 50W-80W;

[0053] The motor operating power consumption is set based on the calibrated mechanical energy consumption of the stepper motor 2 driving the telescopic actuator to complete a single full stroke of expansion and contraction, and the time is evenly distributed over the equivalent operating power of the entire estimated parking cycle. For example, a value of 15W-20W can be used.

[0054] If the calculated difference is greater than zero, it indicates that the expected power generation revenue from the current lighting environment can cover the additional energy consumption caused by system spatial calculation and mechanical action. The wake-up decision unit then generates a hardware interrupt signal to wake up the SoC chip and sensing hardware to enter the working state, where the sensing hardware includes vehicle cameras and LiDAR.

[0055] Conversely, if the calculated difference is less than or equal to zero (for example, when the vehicle is parked in an underground garage or in a cloudy environment with severe obstruction), it indicates that the deployment of the telescopic actuator may not be able to replenish effective electrical energy, and may instead cause a static charge loss in the power battery system. In this case, the wake-up decision unit will generate a hibernation command.

[0056] The hibernation command is sent to the underlying control network to maintain the SoC chip's hibernation state and instruct the retractable actuator to remain fully retracted. By introducing this reward-based game mechanism, ineffective actions of the retractable actuator and unnecessary consumption of vehicle battery power can be effectively avoided.

[0057] See appendix Figure 4 In this embodiment, the pose compensation module includes a pose extraction unit and a spatial mapping unit.

[0058] In real-world applications, vehicles are often affected by factors such as road slope and uneven suspension load when parked, causing the vehicle's surface to deviate from the absolute horizontal plane. Directly using environmental perception data relative to the vehicle's coordinate system for light projection typically results in geometric distortion.

[0059] To minimize the impact of the aforementioned spatial coordinate offset, in this embodiment, the attitude extraction unit acquires parking attitude parameters from the chassis inertial sensor after the SoC chip is woken up. Specifically, the parking attitude parameters include the yaw angle, which characterizes the vehicle's rotation about its vertical axis. Pitch angle of rotation about the lateral axis The roll angle relative to the rotation about the longitudinal axis Based on the aforementioned parking attitude parameters, the attitude extraction unit calculates the spatial rotation matrix using the spatial rotation matrix calculation formula. Its formula is: ; In the formula: It is a spatial rotation matrix; This is the basic rotation matrix corresponding to the Z-axis; This is the basic rotation matrix corresponding to the Y-axis; This is the basic rotation matrix corresponding to the X-axis.

[0060] As a preferred approach, for the trigonometric function expansion operations inside the basic rotation matrix, those skilled in the art can use the Euler angle rotation theorem of the conventional three-dimensional coordinate system to construct the matrix (for example, by applying the sine and cosine functions corresponding to each rotation axis to construct the sub-transformation matrix). The specific derivation of the matrix elements is a well-known technique in the art and will not be elaborated here.

[0061] Through the above calculations, the system establishes a transformation relationship that can map the vehicle's local coordinate system to the absolute horizontal coordinate system.

[0062] The spatial mapping unit calls the sensing hardware to obtain a static point cloud matrix with the vehicle body as the origin. The sensing hardware outputs a set of 3D coordinates of surrounding obstacles through underlying ranging and vision algorithms, forming the static point cloud matrix. Simultaneously, the spatial mapping unit acquires the spatial rotation matrix generated by the pose extraction unit. And obtain the corrected coordinate data of the subsequent loop feedback of the control process.

[0063] The corrected coordinate data represents the amount of positional offset compensation between the physical interference point detected by the telescopic actuator during historical operation and the visual boundary of the sensing system. In the initial state of the system, since the physical contact verification of the telescopic actuator has not yet been performed, the corrected coordinate data is set to a zero matrix (i.e., it is assumed that the point cloud data output by the current sensing system does not have any static offset caused by mechanical installation or sensor blind spots).

[0064] After obtaining the aforementioned basic data, the spatial mapping unit will generate a static point cloud matrix. After combining the corrected coordinate data, multiply by the aforementioned spatial rotation matrix on the left. The formula for calculating the real point cloud matrix is ​​used to output the real point cloud matrix mapped to the real physical world coordinate system. Specifically, after the static point cloud matrix undergoes coordinate translation or data fusion with the corrected coordinate data, attitude compensation is achieved through rotation transformation. The formula is as follows: ; Through a spatial matrix transformation mechanism, the system removes the tilting effect caused by non-horizontal parking posture on environmental perception data, thereby ensuring that the three-dimensional environmental model on which subsequent light and shadow projection depends is as consistent as possible with the horizontal reference of real natural sunlight.

[0065] See appendix Figure 5 In this embodiment, the target solving module includes a shadow projection unit, a transient optimization unit, and a steady-state output unit.

[0066] After establishing the spatial environment model, the next crucial step is to analyze the coupling relationship between light and shadow and the physical topology of the photovoltaic panel. To this end, the shadow projection unit is used to acquire the actual point cloud matrix, time data, and latitude and longitude data.

[0067] This shadow projection unit uses time and latitude / longitude data, combined with conventional astronomical algorithms, to deduce the solar vector at the vehicle's current location. Based on this, the unit uses a ray tracing algorithm to project obstacle feature points from the real point cloud matrix along the direction of the solar vector onto the two-dimensional physical plane where the vehicle roof is located.

[0068] The ray tracing algorithm is used to calculate the intersection process between rays and the bounding box. In this embodiment, the system divides the roof plane into a two-dimensional discrete grid. By determining whether the incident rays emitted in the reverse direction from each grid point are occluded by points in the real point cloud matrix, a shadow boundary function composed of Boolean values ​​is generated on the two-dimensional physical plane. This function can clearly characterize the lighting and occlusion status of each coordinate region on the roof.

[0069] To accurately quantify the expected power generation benefits under different extension positions, the transient optimization unit is used to obtain the aforementioned shadow boundary function and extract the topological mapping relationship and electrical characteristic degradation model of the photovoltaic module 1 loaded inside the system. Here, the topological mapping relationship characterizes the physical spatial arrangement coordinates of each cell within the photovoltaic module 1 and the series-parallel electrical connection logic between them. The electrical characteristic degradation model describes the nonlinear drop in output power of the entire series branch when some cells are covered by shadow, triggering the conduction of the bypass diode. Considering the instantaneous fluctuations in outdoor sunlight conditions, the transient optimization unit performs high-frequency sampling calculations within a set time window.

[0070] As a preferred approach, the time window is set based on the typical frequency characteristics of cloud movement or leaves swaying in the wind, for example, it can be set to 30-60 seconds.

[0071] Within this time window, the transient optimization unit combines the shadow boundary function and electrical characteristic attenuation model at the corresponding time point, and uses the feedforward target travel calculation formula to derive the transient target travel that maximizes net energy efficiency. .

[0072] Specifically, since partial shading can cause multiple local extrema in the power curve, traditional gradient descent algorithms are prone to getting trapped in local optima. Therefore, the system prefers to use a discrete traversal sweep method with a fixed step size to calculate the difference between the expected total power generation and the mechanical execution energy consumption under different strokes, thereby finding the physical location that maximizes this difference. The formula is as follows: ; In the formula: For the first The transient target travel time of the next sample; The operators corresponding to the independent variables when solving for the maximum value of the objective function; For the extension stroke variable; This is the integration operator; For the first The initial timestamp of the next sample; This refers to the duration of parking. For extended journey With time The expected power generation function is as follows; The derivative of the time variable; For extended journey The mechanical power consumption function.

[0073] After multiple sampling and solving within the time window, the transient optimization unit combines the multiple transient target travels to generate a target travel sequence.

[0074] In real-world physical environments, such transient optimization results are often prone to high-frequency fluctuations due to environmental noise. If these results are directly used as commands, they can cause frequent start-stop cycles of mechanical components. To address this issue, a steady-state output unit is used to acquire the target travel sequence and calculate its travel variance. This travel variance reflects the degree of fluctuation in the expected optimal parking position within the time window.

[0075] Subsequently, the steady-state output unit determines whether the stroke variance exceeds a preset tolerance threshold. As a preferred method, the tolerance threshold is set based on the mechanical gear backlash of the telescopic actuator and the allowable motion jitter amplitude of the system; for example, it can be set to correspond to a stroke variance value of 0.02m. 2 .

[0076] If the travel variance is greater than the preset tolerance threshold, the steady-state output unit determines that the current ambient light and shadow are extremely disordered, and then generates an action rejection command to block the current extension action, thereby preventing unnecessary power consumption.

[0077] If the travel variance is less than or equal to the preset tolerance threshold, it indicates that the shadow boundary is relatively stable. The steady-state output unit takes the mean of the target travel sequence for smoothing and calculates the final feedforward target travel.

[0078] In addition, the steady-state output unit also verifies the energy gain baseline. When the net energy efficiency integral corresponding to the feedforward target stroke (i.e. the difference between the integral term and the power consumption term in the feedforward target stroke calculation formula) is negative, it indicates that the system is still in a state of power depletion even when the optimal position is reached. At this time, a termination extension command is directly generated to stop the corresponding mechanical drive action.

[0079] See appendix Figure 6In this embodiment, the closed-loop control module includes a deceleration buffer unit, a gradient monitoring unit, and a redundant interrupt unit.

[0080] After obtaining the optimal stroke parameters issued by the front-end module, targeted driving and control of the telescopic actuator can be carried out.

[0081] Specifically, the deceleration buffer unit is used to obtain the feedforward target stroke and generate a motor drive signal accordingly to control the stepper motor 2 of the telescopic actuator to run at full speed in the non-boundary area.

[0082] During the full-speed operation of stepper motor 2, the reduction buffer unit acquires the real-time physical travel characterizing the current position of the component through the motor encoder. Considering the structural inertia of mechanical components during high-speed movement, emergency braking at the target position can easily lead to mechanical impact on the transmission gears or panel vibration. Therefore, the deceleration and buffer unit continuously compares the difference between the real-time position and the target position, and adjusts the deceleration and buffer unit based on the real-time physical travel. When the difference between the feedforward target stroke and the safety buffer distance is greater than or equal to the value of the difference between the target stroke and the safety buffer distance, the deceleration buffer unit generates a micro-stepping trigger signal to control the stepper motor 2 to decelerate and switch it into a low-speed micro-stepping approximation mode.

[0083] As a preferred approach, the safety buffer distance is set based on the rotor inertia of the stepper motor 2 and the friction braking distance of the telescopic actuator, for example, it can be set to 20ms-50ms. The specific electrical implementation for controlling the stepper motor 2 to switch into microstepping mode can be accomplished by adjusting the drive pulse frequency or using the microstepping control pins of the drive chip, which are well-known technologies in the field and will not be elaborated upon here.

[0084] During the operation of stepper motor 2 in micro-stepping mode, in order to compensate for the possible blind spots of the environmental sensing hardware, the gradient monitoring unit is used to obtain the instantaneous pressure value from the tactile sensor at the front end of photovoltaic module 1 after acquiring the micro-stepping trigger signal. Obtain instantaneous output power from MPPT controller And simultaneously obtain real-time physical travel. The tactile sensor here can take the form of a pressure-sensitive resistor strip or force-sensitive conductive rubber, physically attached to the leading edge of the photovoltaic module 1, to sense physical resistance in the direction of travel. Based on real-time physical travel... The gradient monitoring unit derives the pressure gradient, which characterizes the rate of change of physical contact force with displacement, using the transient pressure gradient calculation formula. Its formula is: ; In the formula: This is the partial derivative operator.

[0085] Simultaneously, the gradient monitoring unit derives the power gradient, which characterizes the rate of change of photoelectric conversion efficiency with displacement, using the transient power gradient calculation formula. Its formula is: ; In actual microcontroller engineering operations, the aforementioned partial derivatives are usually approximated by the difference ratio within the discrete sampling period (i.e., the ratio of a small displacement increment to the corresponding pressure or power change).

[0086] Based on the gradient data calculated above, the redundant interrupt unit is used to obtain the pressure gradient. With power gradient And perform parallel determination. If the pressure gradient If the collision physical threshold is greater than the set threshold, it indicates that the front end of photovoltaic module 1 has physically come into contact with a hard obstacle (such as a hidden twig or cable) that was missed by visual perception.

[0087] At this time, the redundant interrupt unit triggers the physical anti-pinch protection mechanism, generates a forced retraction command to lock the stepper motor 2, and commands the telescopic actuator to retract by a set step length to disengage from physical contact. Then, the retracted position is generated and locked as the safe travel.

[0088] The collision physical threshold is set based on the yield strength of the photovoltaic module 1 frame material and the compressive strength limit of the mechanical structure. For example, it can be set to a corresponding gradient value of 5N / mm-10N / mm.

[0089] If pressure gradient Less than or equal to the set collision physics threshold and power gradient A negative step below the set tolerance indicates that no physical collision occurred at the front end, but the photovoltaic panel entered a hidden shadow area that was not predicted by the spatial model, causing the internal bypass diode to conduct and triggering a step drop in power.

[0090] At this point, the redundant interrupt unit generates an electrical interrupt command to lock stepper motor 2 and directly generates the current stop position as the safe travel distance. The negative step tolerance here is set based on the power ripple fluctuation range of the MPPT controller during normal optimization, for example, it can be set to -5W / mm to distinguish between normal illumination fluctuations and efficiency drops caused by shadow occlusion.

[0091] If the current state is not under either of the above two intervention conditions, it indicates that there is no physical interference or abnormal light and shadow occlusion in the current micro-stepping range. The redundant interrupt unit controls the stepper motor 2 to extend smoothly until it accurately reaches the feedforward target stroke, and generates the feedforward target stroke as the safe stroke.

[0092] Subsequently, the redundant interruption unit extracts the position coordinate data corresponding to the safe travel distance to generate corrected coordinate data, which is then fed back to the spatial mapping unit for coordinate fusion compensation of the underlying model.

[0093] See appendix Figure 7 In this embodiment, the dynamic adjustment module includes a hysteresis control unit, a gain intervention unit, and a system reset unit.

[0094] Once the photovoltaic panel has completed its optimal extension and achieved physical locking, considering that the solar altitude angle will slowly change over time during the parking period, the shadow boundary will also shift accordingly. To avoid frequent minor adjustments to the mechanical structure due to this gradually changing environment, a hysteresis control unit is used to receive the safety travel generated by the forward feedback and perform long-term position maintenance throughout the overall parking period.

[0095] Specifically, the microcontroller is equipped with a hardware timer that periodically wakes up the steady-state output unit at a fixed frequency, enabling the system to rerun the optimization algorithm based on the updated time data, thereby obtaining the latest recalculated feedforward target distance.

[0096] After acquiring the latest parameters, the hysteresis control unit calculates the absolute difference in travel between the latest feedforward target travel and the currently locked safe travel. Subsequently, the hysteresis control unit determines whether the absolute difference in travel is greater than the set dead zone width and continuously monitors whether the deviation state is maintained for a set set time period.

[0097] As a preferred approach, the dead zone width here is set based on the physical width of a single cell in the photovoltaic module 1 and the mechanical transmission clearance, for example, it can be set to 80ms-120ms; The time period is set based on the angular velocity of the sun's offset caused by the Earth's rotation and the average period of cloud movement, for example, it can be set to 10-15 minutes.

[0098] The hysteresis control unit generates a fine-tuning permission command only when the absolute difference in the stroke exceeds the set dead zone width and remains so for a set time period. Upon receiving this fine-tuning permission command, the underlying control logic is reactivated and the motor is controlled to perform subsequent mechanical adjustment actions. By introducing a spatial dead zone and a time-delay confirmation mechanism, the system can filter out mechanical vibrations caused by gradual environmental changes and occasional disturbances.

[0099] For conventional geometric optical projection models, the system mainly relies on direct sunlight for shadow calculation, making it difficult to directly handle complex indirect light sources in urban environments (such as reflections from glass curtain walls). To compensate for this limitation of the algorithm, a gain intervention unit is used to continuously acquire actual light intensity data from an externally configured light intensity sensor and receive the theoretical power generation generated by the system front end.

[0100] The gain intervention unit derives the expected shadow power based on the theoretical power generation using an internal theoretical calculation model. Specifically, this theoretical calculation model multiplies the theoretical power generation by a scaling factor characterizing the ambient diffuse background intensity, based on the current solar altitude angle and atmospheric scattering coefficient, thereby calculating the expected shadow power that should be present within the ideal shadow shading area.

[0101] The atmospheric scattering coefficient and proportionality coefficient mentioned above can be obtained through pre-calibrated empirical values. In specific implementation, those skilled in the art can configure them as a local one-dimensional look-up table based on the season and basic weather conditions, so that the microcontroller can quickly call and calculate them.

[0102] In real-world parking scenarios, if a vehicle is parked next to a building with a large glass curtain wall, the strong specular reflection from the building's surface will typically project a large amount of light energy into the shadow area predicted by the algorithm. At this time, the gain intervention unit compares the actual environmental parameters with the theoretical model parameters in real time. If it finds that the actual ambient light power, represented by the actual light intensity data, is greater than the sum of the expected shadow power and the set reflective gain threshold, it determines that the vehicle is currently in a high-value reflected light field.

[0103] In this embodiment, the reflectivity gain threshold is set based on the reflectivity of common building glass and the minimum effective start-up power of photovoltaic module 1. For example, it can be set to a power difference of 100W-150W.

[0104] When the above-mentioned conditions are met, the gain intervention unit generates a forced extension command and sends it to the telescopic actuator to force it to fully extend. At the same time, it generates a projection shielding command to shield the geometric ray tracing algorithm at the front end, thereby making the most of the complex reflected light sources in the external environment.

[0105] When the vehicle faces a state transition during the parking period, the system needs to ensure the safe retraction of the physical structure. To this end, the system reset unit maintains communication with the vehicle's underlying network, monitoring for engagement / disengagement signals or vehicle unlocking signals from the chassis. At any point during parking, upon receiving either an engagement / disengagement signal or a vehicle unlocking signal, indicating that the vehicle is about to disengage, the system reset unit triggers the highest-level hardware interrupt response. Ignoring any currently executing optimization state or mechanical position, the system reset unit directly generates a system reset command, controlling the telescopic actuator to quickly retract to its initial reference position with maximum available safe torque, ensuring that the vehicle's driving physical boundaries remain unaffected.

[0106] To further illustrate the working principle and effect of the present invention in practical application scenarios, the following description is based on a specific parking power supply scenario.

[0107] In this application example, a vehicle equipped with this system is parked on an uneven road surface with urban building shadows. The parking time is 2:00 PM in summer, and the road surface has a 5° side tilt angle and a 3° pitch angle. The vehicle is adjacent to a 15m high glass curtain wall building on its right side. This building casts a significant dynamic shadow on the vehicle's roof, and due to the reflection effect of the glass curtain wall, there is a strong secondary reflected light field at the edge of the shadow.

[0108] During system startup, the parking monitoring module detected that the vehicle was in parking position and the power battery's state of charge was 60%. The microcontroller obtained an ambient irradiance of 800W / m². 2 The theoretical power generation is estimated to be approximately 115W using a low-order geometric model. At this point, the system queries the chip's startup power consumption, which is 60W, and the motor's operating power consumption, amortized over the parking cycle, is approximately 15W, totaling 75W. Calculating the difference between the theoretical power generation benefit and the system operating cost yields 115W - 75W = 40W. Since the difference is greater than zero, the wake-up decision unit immediately generates a hardware interrupt signal, waking the SoC chip to enter a high-precision calculation state.

[0109] After the pose compensation module intervenes, pose data is read through the chassis inertial sensors. The roll angle of 5°, pitch angle of 3°, and yaw angle of 0° are substituted into the spatial rotation matrix calculation formula. In the specific calculation, the basic rotation matrix constructed from Euler angles is multiplied to obtain the result. : ; Subsequently, the spatial mapping unit obtains a static point cloud matrix containing the building outline. By left-multiplying the matrix above, it corrects the point cloud matrix from the tilted vehicle coordinate system to the absolute horizontal coordinate system, thus obtaining the true point cloud matrix.

[0110] During the optimization phase, the transient optimization unit performs a net energy efficiency assessment based on the calculated shadow boundary function. Assuming the maximum stroke of the telescopic actuator is 1000mm, the estimated parking time is... The time is 14400 seconds (4 hours). When inputting the feedforward target travel calculation formula, the system calculates the revenue for different travel distances L using a discrete traversal sweep method: ; In the specific substitution calculation process, if the journey Taking 0mm (i.e., maintaining the initial state), the corresponding expected power generation function The average power consumption is 115W, mechanical power consumption. The initial energy gain is 0J, and the total energy gain is 115 × 14400 = 1656000J. (The remaining text appears to be incomplete and requires further context.) Increasing the thickness to 650mm, the photovoltaic panels effectively avoid the building's master shadow and utilize lateral reflection gain, resulting in an expected increase in power generation function. The average power consumption increases to 185W, at which point the mechanical power required to move to this stroke is reduced. The calculated value is approximately 18,000 J. Substituting this into the formula, the total net energy gain is 185 × 14,400 - 18,000 = 2,646,000 J. Through algorithmic comparison, it can be seen that... The difference reaches its maximum value around 650mm, thus determining the transient target travel at that moment to be 650mm.

[0111] During execution, the deceleration buffer unit controls the stepper motor 2 to drive the telescopic actuator. When it reaches 635mm, the gradient monitoring unit detects that the instantaneous pressure value remains stable, but the power gradient exhibits a negative step and exceeds the set threshold. The redundant interruption unit determines that there is a hidden shadow at this position, and immediately stops the extension and locks it at 635mm as a safe stroke. The system net energy efficiency power calculation process at this time is as follows: the expected power generation of photovoltaic module 1 at this stroke is approximately 185W. After deducting the system operating cost of 75W, the final output system net energy efficiency power is 110W.

[0112] See appendix Figure 8 This demonstrates the coupling relationship between system benefits and travel in the above scenario.

[0113] Figure 8 The thick solid line with a circle mark represents the dynamic benefit curve of the present invention. This curve reflects the trend of power increasing with the increase of stroke during the optimization process. It is marked with a vertical dotted line at the safe stroke of 635mm, at which the corresponding vertical axis power value is about 110W.

[0114] The thin dashed line represents the conventional control curve without pose compensation and optimization. Because it does not take into account the projection deviation caused by vehicle tilt, the maximum power point is shifted and the value is only about 90W.

[0115] The dotted line represents the baseline for the revenue of a fixed solar roof, where the travel distance is constant at 0mm and the power output is constant at 40W (i.e., 115W-75W). A comparison shows that the net revenue of this invention, after optimization and locking, is approximately 175% higher than that of a fixed roof, and effectively avoids the power loss caused by blindly extending the roof.

[0116] To further verify the rationality of the present invention, this embodiment constructs a special experimental environment and uses measured data to quantitatively evaluate the beneficial effects of the technical solution.

[0117] The experiment was conducted on a test vehicle equipped with this system. The edge of an urban road network with complex light and shadow characteristics was selected as the test site to ensure that the environment included the shadows of tall buildings, the obstruction of roadside trees, and the reflection of glass curtain walls.

[0118] First, under ideal conditions where the vehicle is parked horizontally and without obstruction, the photovoltaic module 1 is calibrated to obtain standard output curves under different irradiance levels. These curves serve as the basis for subsequent calculations of theoretical power generation. Simultaneously, the sampling accuracy of the tactile sensor and the MPPT controller is calibrated.

[0119] After entering a complex lighting and shadow scene, the vehicle was artificially parked on a sloping road, and the roll and pitch angle data output by the chassis inertial sensors were recorded. The actual point cloud matrix generated by the pose compensation module was then observed. By comparing the physical projection deviation of the shadow boundary on the photovoltaic panel before and after compensation, the effectiveness of the spatial rotation matrix calculation formula in eliminating geometric distortion was verified.

[0120] In the optimization process verification, the instantaneous output power at different physical positions was recorded by changing the extension stroke of the telescopic actuator, and then compared with the predicted value obtained by the transient optimization unit using the feedforward target stroke calculation formula. This step aims to verify the accuracy of the coupled modeling of the expected power generation function and the mechanical power consumption function in the formula by the degree of coincidence between the measured power curve and the theoretical integral curve.

[0121] To verify the reliability of the closed-loop control circuit, a scenario was simulated where the sensing hardware failed to detect a small obstacle, based on the triggering logic of the safety mechanism. The pressure gradient data output by the gradient monitoring unit was recorded. The response time of the redundant interrupt unit in generating a forced retraction command after detecting an over-limit gradient, as well as the physical distance between the locked safe travel distance and the obstacle boundary, were observed.

[0122] Finally, throughout the entire parking cycle, the frequency of action of the hysteresis control unit and the forced extension behavior of the gain intervention unit when detecting the reflected light field are recorded by the dynamic adjustment module. The total energy gain during the entire parking process is calculated, and the power consumption generated by the operation of each unit in the system is deducted to obtain the final net energy efficiency power of the system.

[0123] Table 1. Performance Comparison of Power Supply Schemes in Different Parking Scenarios

[0124] in conclusion: Experimental data show that Scenario 1, under ideal conditions, demonstrates the fundamental contribution of increasing the light-receiving area to power improvement. Scenarios 2 and 3 fully demonstrate the synergistic advantages of the pose compensation module and the target solution module. When road slope exists, conventional telescopic schemes, lacking pose correction, often deviate from the optimal light-receiving point in their feedforward target travel, causing some solar cells to be in a partially obscured state, triggering bypass diodes. The proposed solution corrects the projection deviation through a spatial rotation matrix calculation formula, resulting in a system net energy efficiency power superior to conventional schemes.

[0125] Data from Scenario 3 further demonstrates the unique value of the gain intervention unit in complex urban environments. By comparing actual light intensity data with expected shadow power, the system successfully identified and utilized the additional reflected light field from the glass curtain wall, achieving power replenishment exceeding the theoretical direct sunlight limit. Furthermore, throughout the verification process in all scenarios, the deviation between the predicted and measured values ​​obtained through the feedforward target travel calculation formula remained at a low level, proving that the mathematical model's description of the game-theoretic relationship between photovoltaic power degradation and mechanical energy consumption is reasonable and accurate.

Claims

1. A retractable solar roof power supply system for vehicles, characterized in that, include: The parking monitoring module is used to obtain the gear position and battery charge state to generate a system activation signal, obtain light intensity data, time data and latitude and longitude data to estimate the theoretical power generation, and generate a hardware interrupt signal when the theoretical power generation is greater than the sum of the chip start-up power consumption and the motor operation power consumption. The pose compensation module is used to obtain parking posture parameters and use the spatial rotation matrix calculation formula to convert the obtained static point cloud matrix and corrected coordinate data into a real point cloud matrix. The target solution module generates a shadow boundary function based on the real point cloud matrix, combines the electrical characteristic attenuation model with the feedforward target travel calculation formula to obtain the target travel sequence, and outputs the feedforward target travel when the travel variance of the target travel sequence meets the condition. The closed-loop control module controls the stepper motor (2) based on the feedforward target stroke, calculates the pressure gradient and power gradient to perform electromechanical parallel judgment to generate a safe stroke, extracts coordinate data to generate the corrected coordinate data and feeds it back to the pose compensation module. The dynamic adjustment module is used to compare the latest feedforward target travel with the safe travel, and intervene based on the theoretical power generation and the actual ambient light power to generate a forced extension command.

2. The retractable solar roof power supply system for vehicles according to claim 1, characterized in that, The parking monitoring module includes: A state activation unit is used to generate the system activation signal when the gear position is shifted into parking gear and the battery is not fully charged, so as to control the microcontroller to acquire the light intensity data, the time data and the latitude and longitude data; The revenue estimation unit is used to estimate the theoretical power generation by combining the light intensity data, the time data and the latitude and longitude data through a low-order geometric model. The wake-up decision unit is used to generate the hardware interrupt signal when the theoretical power generation is greater than the sum of the chip startup power consumption and the motor operation power consumption, so as to wake up the SoC chip and the sensing hardware to enter the working state; otherwise, it generates a sleep command to control the telescopic actuator to fully retract.

3. A retractable solar roof power supply system for vehicles according to claim 2, characterized in that, The pose compensation module includes: An attitude extraction unit is used to obtain the parking attitude parameters, including yaw angle, pitch angle and roll angle, from the chassis inertial sensor, and to obtain the spatial rotation matrix using the spatial rotation matrix calculation formula. The spatial mapping unit is used to call the sensing hardware to obtain the static point cloud matrix with the vehicle body as the origin, combine the static point cloud matrix with the corrected coordinate data, multiply it by the spatial rotation matrix, and output the real point cloud matrix using the real point cloud matrix calculation formula. In the initial state, the corrected coordinate data is a zero matrix.

4. A retractable solar roof power supply system for vehicles according to claim 3, characterized in that, The objective solving module includes: The shadow projection unit analyzes the sunlight vector based on the time data and the latitude and longitude data, and generates the shadow boundary function by projecting the real point cloud matrix along the direction of the sunlight vector using a ray tracing algorithm. The transient optimization unit is used to combine the shadow boundary function and the topological mapping relationship of the photovoltaic module (1) at the corresponding time and the electrical characteristic attenuation model within a set time window, and use the feedforward target travel calculation formula to obtain multiple transient target travels to combine and generate the target travel sequence. A steady-state output unit is used to calculate the travel variance of the target travel sequence. If it is greater than a preset tolerance threshold, an action rejection instruction is generated. If it is less than or equal to the tolerance threshold, the average value is taken to generate the feedforward target travel. The tolerance threshold is preset based on the mechanical gear clearance of the telescopic actuator and the allowable motion jitter amplitude of the system.

5. A retractable solar roof power supply system for vehicles according to claim 4, characterized in that, The steady-state output unit is also used to generate a termination extension command when the net energy efficiency integral corresponding to the feedforward target stroke is negative.

6. A retractable solar roof power supply system for vehicles according to claim 4, characterized in that, The closed-loop control module includes: The deceleration buffer unit generates a motor drive signal based on the feedforward target stroke to control the operation of the stepper motor (2). When the real-time physical stroke obtained by the motor encoder is greater than or equal to the difference between the feedforward target stroke and the safety buffer distance, it generates a micro-step trigger signal to control the deceleration of the stepper motor (2). The gradient monitoring unit is used to acquire the instantaneous pressure value and instantaneous output power after acquiring the micro-step trigger signal, and to obtain the pressure gradient and the power gradient based on the real-time physical stroke through the transient pressure gradient calculation formula and the transient power gradient calculation formula. A redundant interrupt unit is used to make the electromechanical parallel determination based on the pressure gradient and the power gradient, lock the stepper motor (2) to generate the safe stroke and extract the corresponding coordinate data to generate the corrected coordinate data; The safety buffer distance is preset based on the rotor inertia of the stepper motor (2) and the friction braking distance of the telescopic actuator.

7. A retractable solar roof power supply system for vehicles according to claim 6, characterized in that, The logic for the redundant interrupt unit to perform the electromechanical parallel determination specifically includes: If the pressure gradient is greater than the set collision physical threshold, a forced retraction command is generated to lock the stepper motor (2). If the pressure gradient is less than or equal to the set collision physical threshold and the power gradient is less than the set tolerance negative step, an electrical interrupt command is generated to lock the stepper motor (2). If neither of the above two situations applies, control the stepper motor (2) to extend smoothly; The collision physical threshold is preset based on the yield strength of the frame material of the photovoltaic module (1) and the compressive strength limit of the mechanical structure, and the negative step of the tolerance is preset based on the power ripple fluctuation range of the MPPT controller during the optimization process.

8. A retractable solar roof power supply system for vehicles according to claim 6, characterized in that, The dynamic adjustment module includes: The hysteresis control unit is used to calculate the absolute difference between the latest feedforward target stroke and the safety stroke, and generate a fine-tuning allow command after the absolute difference is greater than the set dead zone width and is maintained for a set time period, so as to control the stepper motor (2) to perform mechanical adjustment action. The gain intervention unit calculates the expected shadow power based on the theoretical power generation through a theoretical calculation model. When the actual ambient light power, as represented by the acquired actual light intensity data, is greater than the sum of the expected shadow power and the set reflective gain threshold, it generates the forced extension command and the projection shielding command. The dead zone width is preset based on the physical width of a single cell of the photovoltaic module (1) and the mechanical transmission gap. The time period is preset based on the solar offset angular velocity and the average cloud movement period. The reflective gain threshold is preset based on the reflectivity of the building glass and the minimum effective starting power of the photovoltaic module (1).

9. A retractable solar roof power supply system for vehicles according to claim 8, characterized in that, The dynamic adjustment module also includes a system reset unit; The system reset unit is used to generate a system reset command when it receives a disengaged gear signal or a vehicle unlock signal.

10. A method for controlling power supply to a retractable solar roof for vehicles, characterized in that, The retractable solar roof power supply system for vehicles according to any one of claims 1-9 includes the following steps: The system uses a parking monitoring module to obtain gear status and battery charge status to generate a system activation signal, obtains light intensity data, time data and latitude and longitude data to estimate theoretical power generation, and generates a hardware interrupt signal when the theoretical power generation is greater than the sum of chip startup power consumption and motor operation power consumption. The parking posture parameters are obtained using the pose compensation module and the static point cloud matrix and corrected coordinate data are converted into a real point cloud matrix using the spatial rotation matrix calculation formula. The target solution module generates a shadow boundary function based on the real point cloud matrix, and the target travel sequence is obtained by combining the electrical characteristic attenuation model with the feedforward target travel calculation formula. The feedforward target travel is output when the travel variance of the target travel sequence meets the condition. The closed-loop control module controls the stepper motor (2) based on the feedforward target stroke, calculates the pressure gradient and power gradient and performs electromechanical parallel determination to generate a safe stroke, extracts coordinate data to generate the corrected coordinate data and feeds it back to the pose compensation module. The dynamic adjustment module compares the latest feedforward target travel with the safe travel, and intervenes based on the theoretical power generation and the actual ambient light power to generate a forced extension command.

Images