Wireless charging transmitting and receiving coil alignment method and system based on vehicle-mounted vision

Abstract

The application provides a wireless charging transmitting and receiving coil alignment method and system based on vehicle-mounted vision, and relates to the technical field of electric vehicle wireless charging. The alignment method comprises the following steps: pre-calibrating the relative coordinates of the center point of the receiving coil in the vehicle body coordinate system; identifying the parking space origin by using the vehicle-mounted vision system to establish a parking space coordinate system, and analyzing the coordinates of the center point of the transmitting coil in the parking space coordinate system; acquiring the pose of the origin of the vehicle coordinate system in real time during the parking process of the vehicle; calculating the theoretical position of the center point of the receiving coil in the parking space coordinate system according to the vehicle pose and the relative coordinates of the receiving coil through coordinate transformation; calculating the deviation between the theoretical position and the coordinates of the center point of the transmitting coil, and judging whether the alignment is completed and outputting the guidance information according to the comparison result of the deviation and the preset threshold value. The application realizes low-cost and high-precision coil automatic alignment by using vehicle-mounted vision, and effectively overcomes the problem of visual blind area caused by the obstruction of the vehicle bottom.

Description

Technical Field

[0001] This invention relates to the field of wireless charging technology for electric vehicles, and more specifically, to a method and system for aligning wireless charging transmitting and receiving coils based on vehicle vision. Background Technology

[0002] In wireless charging systems for electric vehicles, the alignment accuracy between the ground transmitting coil and the on-board receiving coil directly determines the system's energy transfer efficiency. When there is a significant misalignment between the two coils, the coupling coefficient decreases significantly, leading to a substantial reduction in charging efficiency or even preventing normal charging.

[0003] Currently, to achieve high-precision automatic alignment of the primary and secondary coils, existing technologies mainly adopt the following solutions: One method is low-frequency positioning, which uses multiple low-frequency antennas on the ground to transmit signals and vehicle-mounted receivers to detect signal strength to calculate relative position. However, the guidance distance is short and the accuracy is limited, requiring additional hardware.

[0004] Secondly, there are differential induction positioning systems and ultra-wideband ranging methods. The former integrates auxiliary coils outside the main charging coil area to analyze signal strength and determine the offset direction, while the latter uses radio signal flight time for triangulation positioning. However, both are susceptible to environmental noise and multipath effects, and the consistency of radio frequency signal strength is difficult to guarantee.

[0005] Thirdly, passive beacon technology and magnetic field sensor arrays can achieve millimeter-level positioning accuracy by arranging passive beacon circuits or multiple magnetic induction coils at the vehicle or ground end, but this requires a large amount of additional sensor hardware.

[0006] Fourthly, vision-based solutions have been proposed. Some studies have suggested using panoramic camera image processing to obtain the positional relationship between the transmitting and receiving coils and plan the driving path. Other solutions use a binocular vision system to acquire parking space images and calculate alignment parameters. However, these vision solutions usually require the camera optical axis to be strictly aligned with the coil center or require complex on-site calibration, resulting in low system fault tolerance.

[0007] In addition, there are solutions that use vision for lane detection for coarse adjustment, and then combine it with Hall sensors or detection coils for fine adjustment, but these require a one-dimensional actuator to move the receiving coil, making the system topology complex. Summary of the Invention

[0008] The problem solved by this invention is one or more of the aforementioned related technical problems.

[0009] To address the aforementioned problems, this invention provides a method and system for aligning wireless charging transmitting and receiving coils based on vehicle vision.

[0010] In a first aspect, the present invention provides a method for aligning wireless charging transmitting and receiving coils based on vehicle vision, comprising: The horizontal and vertical relative coordinates of the center point of the receiving coil in the vehicle coordinate system are pre-calibrated. The vehicle coordinate system takes a preset point on the vehicle body as the origin, with the horizontal axis along the width of the vehicle and the vertical axis along the front of the vehicle. Based on the vehicle-mounted vision system, the parking space features are identified, the origin of the parking space is determined, and a parking space coordinate system is established. The parking space coordinate system has the origin of the parking space as its origin, the horizontal axis along the width of the parking space, and the vertical axis along the length of the parking space. Based on the image of the marker at the center point of the transmitting coil acquired by the vehicle-mounted vision system, the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system are analyzed by a visual ranging and positioning algorithm. During the process of parking a vehicle into a parking space, the vehicle vision system acquires in real time the lateral position, longitudinal position, and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system. Based on the lateral position, longitudinal position, and orientation angle of the origin of the vehicle's own coordinate system, as well as the pre-calibrated lateral and longitudinal relative coordinates of the center point of the receiving coil, the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system are calculated through coordinate transformation. Calculate the lateral deviation between the theoretical lateral position and the lateral coordinates of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinates of the center point of the transmitting coil; Based on the comparison results of the lateral and longitudinal deviations with the corresponding preset thresholds, it is determined whether the alignment is complete, and guidance information is output.

[0011] Optionally, the coordinate transformation includes: Vp_x=Vo_x+Vx×cosθ-Vy×sinθ; Vp_y=Vo_y+Vx×sinθ-Vy×cosθ; Wherein, Vp_x is the theoretical lateral position, Vp_y is the theoretical longitudinal position, Vo_x is the lateral position of the origin of the vehicle's own coordinate system, Vo_y is the longitudinal position of the origin of the vehicle's own coordinate system; Vx is the lateral relative coordinate of the center point of the receiving coil, Vy is the longitudinal relative coordinate of the center point of the receiving coil; θ is the orientation angle of the origin of the vehicle's own coordinate system.

[0012] Optionally, the criteria for determining whether alignment is complete include: the absolute values ​​of both the lateral deviation and the longitudinal deviation are less than or equal to the corresponding preset thresholds.

[0013] Optionally, the vehicle vision system includes at least one of a surround-view camera, a front-view camera, or a rear-view camera, for identifying parking lines, parking corners, or visual markings affixed to the ground.

[0014] Optionally, determining the origin of the parking space based on the parking space features identified by the vehicle vision system includes: Based on the top view output by the surround view camera in the vehicle vision system, the parking space lines are obtained by identifying the top view based on edge detection and Hough transform, and the corner point of the parking space is determined as the origin of the parking space.

[0015] Optionally, the step of identifying parking lines from the top view based on edge detection and Hough transform, and determining the corner point of the parking space as the origin of the parking space, includes: Based on the edge detection, the edge feature points of the parking lines are extracted from the top view; By using the Hough transform, the straight line equation of the parking space boundary line is fitted to the edge feature points, and the intersection point of the straight line equations of two adjacent parking space boundary lines is found. The intersection point is determined as the corner point of the parking space, and the corner point of the parking space is used as the origin of the parking space.

[0016] Optionally, the step of resolving the lateral and longitudinal coordinates of the center point of the transmitting coil in the parking space coordinate system using a visual ranging and positioning algorithm includes: Acquire an image containing the ArUco visual marker code pasted at the center point of the transmitting coil; The perspective transformation algorithm is used to analyze the two-dimensional projection of the center point of the ArUco visual marker code in the parking space coordinate system, and the horizontal and vertical coordinates are used as the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system.

[0017] Optionally, the method of outputting the guidance information includes: outputting the lateral deviation and the longitudinal deviation as correction values ​​to the automatic parking system on the vehicle, and the automatic parking system controlling the vehicle to make fine adjustments based on the correction values.

[0018] Optionally, controlling the vehicle to make fine-tuning movements based on the correction amount includes: Based on the vehicle's Ackerman steering geometry model, with the lateral and longitudinal deviations approaching preset target values ​​as control objectives, a smooth trajectory is dynamically planned, and real-time speed and front wheel steering angle commands are output to control the vehicle's movement.

[0019] Secondly, the present invention provides a wireless charging transmitting and receiving coil alignment system based on vehicle vision, used to implement the wireless charging transmitting and receiving coil alignment method based on vehicle vision as described in the first aspect, including: The acquisition unit is used to pre-calibrate the lateral and longitudinal relative coordinates of the center point of the receiving coil in the vehicle coordinate system, wherein the vehicle coordinate system takes a preset point on the vehicle body as the origin, the horizontal axis is along the width of the vehicle, and the vertical axis is along the front of the vehicle. The parsing unit is used to identify parking space features based on the vehicle vision system, determine the origin of the parking space, and establish a parking space coordinate system. The parking space coordinate system has the origin of the parking space as its origin, with the horizontal axis along the width direction of the parking space and the vertical axis along the length direction of the parking space. Based on the image of the marker at the center point of the transmitting coil acquired by the vehicle vision system, the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system are parsed using a visual ranging and positioning algorithm. The acquisition unit is also used to acquire, in real time, the lateral position, longitudinal position, and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system based on the vehicle vision system during the process of parking the vehicle into the parking space. The transformation unit is used to calculate the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system by means of coordinate transformation, based on the lateral position, longitudinal position and orientation angle of the origin of the vehicle's own coordinate system, and the lateral and longitudinal relative coordinates of the center point of the receiving coil that have been pre-calibrated. The calculation unit is used to calculate the lateral deviation between the theoretical lateral position and the lateral coordinates of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinates of the center point of the transmitting coil; The comparison unit is used to determine whether the alignment is complete based on the comparison results of the lateral deviation and longitudinal deviation with the preset threshold, and output guidance information.

[0020] Thirdly, the present invention provides a wireless charging transmitting and receiving coil alignment device based on vehicle vision, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the wireless charging transmitting and receiving coil alignment method based on vehicle vision as described in the first aspect.

[0021] The beneficial effects of the wireless charging transmitting and receiving coil alignment method and system based on vehicle vision of the present invention are: First, this method fully utilizes the vehicle's existing onboard vision system (such as surround-view cameras, front-view cameras, or rear-view cameras), eliminating the need for additional expensive hardware (such as low-frequency antennas, ultra-wideband modules, and magnetic induction coil arrays). This significantly reduces system hardware costs and facilitates deployment in mass-produced vehicles. Second, by establishing independent and completely decoupled vehicle coordinate systems and parking space coordinate systems, this method separates the calibration of the receiving coil from the identification of the transmitting coil. The vision system only needs to identify any parking space corner to establish the parking space coordinate system, fundamentally avoiding the stringent requirements of traditional solutions where the camera optical axis must be strictly aligned with the coil center or complex on-site calibration is required. This effectively reduces the interference of visual occlusion and blind spots on positioning accuracy, while simultaneously improving... The method achieves significantly higher alignment accuracy than traditional positioning schemes. Secondly, it fuses the pre-calibrated relative position of the receiving coil with the real-time vehicle pose through coordinate transformation, directly deriving the theoretical position of the receiving coil in the vehicle's coordinate system. This theoretical position is then compared with the visually recognized position of the transmitting coil, eliminating the need for cumulative extrapolation based on vehicle positioning. This fundamentally eliminates the accumulated errors caused by integral drift in odometer or inertial navigation systems, ensuring the consistency and reliability of each alignment result. Finally, the method completely decouples the coil alignment process from the charging function. Alignment determination and guidance output are both autonomously completed by the vehicle, eliminating the need for long-link signal interaction between the ground and vehicle ends. This simplifies the system control process and improves the robustness and engineering feasibility of the entire wireless charging system. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a method for aligning wireless charging transmitting and receiving coils based on vehicle vision, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the vehicle body coordinate system and the calibration of the receiving coil according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the parking space coordinate system and transmitting coil identification according to an embodiment of the present invention; Figure 4 This is a schematic diagram showing the coordinates and posture of the vehicle body during the parking process according to an embodiment of the present invention. Detailed Implementation

[0023] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0024] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0025] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the description below. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0026] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0027] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0028] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for aligning wireless charging transmitting and receiving coils based on vehicle vision, comprising: Step S100: Pre-calibrate the lateral and longitudinal relative coordinates of the center point of the receiving coil in the vehicle coordinate system. The vehicle coordinate system takes a preset point on the vehicle body as the origin, with the horizontal axis along the width of the vehicle and the vertical axis along the front of the vehicle.

[0029] Specifically, in an electric vehicle wireless charging system, the energy transmitting device (i.e., the primary coil, or ground-side transmitting coil GA device) is installed at the ground parking space to transmit electromagnetic energy to the vehicle; the energy receiving device (i.e., the secondary coil, or vehicle-side receiving coil VA device) is installed under the vehicle chassis to receive electromagnetic energy from the ground and convert it into electrical energy stored in the vehicle battery. For the wireless charging system to operate efficiently, the alignment accuracy between the primary and secondary coils is crucial. The closer the center points of the two coils are aligned, the higher the electromagnetic coupling efficiency, and the stronger the charging efficiency and power transmission capability. Therefore, the core task of this application's technical solution is to guide the vehicle's precise movement through an onboard vision system, ensuring precise spatial alignment between the center point of the secondary coil installed on the vehicle chassis and the center point of the primary coil at the ground parking space.

[0030] When the energy receiving device (VA device) is installed in the vehicle at the factory, or after the device is installed on the chassis, technicians precisely calibrate the center point V of the secondary coil according to the vehicle's structural dimensions and measurement methods. The prerequisite for calibration is first determining the origin of the vehicle's coordinate system. The vehicle's coordinate system is a two-dimensional planar coordinate system established with a preset point on the vehicle body as its origin, where the horizontal axis (X-axis) is along the vehicle's width and the vertical axis (Y-axis) is along the vehicle's front direction. The selection of the vehicle's origin is flexible; multiple different points on the vehicle body can be selected as the origin, such as the left rear corner, the vehicle's geometric center, or the rear axle center. Different origin selections only result in different coordinate values ​​for the receiving coil's center point, but do not affect the validity of the coordinate transformation model in this application's technical solution, nor do they affect the final alignment result. In the example of this application, the left rear corner of the vehicle body is used as the origin of the vehicle's coordinate system.

[0031] After determining the vehicle body origin, the distance from the center point V of the receiving coil to the vehicle body origin is measured from the vehicle body origin using measuring tools (such as a laser rangefinder, tape measure, or coordinate measuring machine). This gives the lateral relative coordinate Vx. The distance from the center point V of the receiving coil to the vehicle body origin is measured from the vehicle body origin along the vehicle width direction (lateral, i.e., the X-axis direction). This gives the longitudinal relative coordinate Vy.

[0032] For example: Figure 2 As shown, with the left rear of the vehicle body as the origin, the lateral relative coordinates Vx = 800mm and the longitudinal relative coordinates Vy = 500mm of the center point V of the receiving coil are measured. These two coordinate values, once calibrated, are stored in the vehicle's storage unit (such as the memory of the onboard controller) for use in subsequent coordinate transformation calculations during alignment. Since the relative position between the receiving coil and the vehicle body remains fixed once the receiving coil is installed on the chassis, this calibration process only needs to be completed once when the vehicle leaves the factory and does not need to be repeated during each charging. Figure 2This diagram illustrates the coordinate system of the vehicle body where VA is located. It shows a two-dimensional vehicle body coordinate system established with a preset point on the vehicle body (in this embodiment, the left rear corner of the vehicle body) as the origin O_v. The horizontal axis (X-axis) runs along the width of the vehicle, and the vertical axis (Y-axis) runs along the front of the vehicle. The diagram also shows the position of the center point V of the receiving coil (secondary coil) mounted on the chassis, as well as its lateral relative coordinates Vx and longitudinal relative coordinates Vy in the vehicle body coordinate system. This diagram visually illustrates the fixed spatial positional relationship of the receiving coil's center point relative to the vehicle body origin.

[0033] By pre-calibrating and storing the fixed relative coordinates of the receiving coil center point in the vehicle coordinate system, a deterministic geometric relationship is established between the position information of the receiving coil and the vehicle body. Subsequently, no matter how the vehicle moves or turns, as long as the real-time pose of the vehicle itself in the parking space coordinate system is known, the absolute position of the receiving coil center point can be calculated at any time through coordinate transformation, without having to re-identify or measure the position of the receiving coil during each parking process, thus greatly reducing the online recognition burden of the vision system. At the same time, since the vehicle coordinate system and the parking space coordinate system are completely decoupled, the calibration of the receiving coil can be completed independently of the parking space environment and the layout of the ground transmitting coil, without relying on any ground markers or signals, reducing the calibration complexity of the system during on-site deployment and improving the versatility and portability of the solution in different vehicle models and different parking space scenarios.

[0034] Step S200: Based on the vehicle vision system's identification of parking space features, determine the parking space origin and establish a parking space coordinate system. The parking space coordinate system has the parking space origin as its origin, with the horizontal axis along the width direction of the parking space and the vertical axis along the length direction of the parking space. Based on the vehicle vision system, acquire an image containing the marker at the center point of the transmitting coil, and use a visual ranging and positioning algorithm to analyze the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system.

[0035] Specifically, after the vehicle body coordinates of the receiving coil center point are calibrated in step S100, step S200 turns to the ground end, establishes a parking space coordinate system through the vehicle vision system, and identifies the position of the transmitting coil center point within it. It should be noted that there is no strict sequential execution order between steps S200 and S100. The establishment of the parking space coordinate system and the identification of the transmitting coil position in step S200 occur during the process of the vehicle parking into the target parking space.

[0036] When a vehicle begins to park in the target parking space, the onboard vision system (such as a surround-view camera, a front-view camera, or a rear-view camera) first identifies the parking space features, often including parking lines, corner points, and other visual information, to determine the parking space origin O. The selection of the parking space origin is also flexible; theoretically, any corner point of the parking space can be chosen as the origin. Different corner point choices only result in different coordinate values ​​of the transmitting coil's center point and do not affect the effectiveness of subsequent coordinate transformations. In the example of this application, such as... Figure 3 As shown, the left rear corner of the parking space is taken as the origin O of the parking space coordinate system. Figure 3 This diagram illustrates the visual recognition coordinates of the parking space and the center point G of the transmitting equipment GA, i.e., a schematic diagram of the parking space coordinate system. It shows the parking space coordinate system established with the origin O at a corner point of the parking space (the left rear corner point in this embodiment) after the parking space features are identified by the vehicle-mounted vision system. The horizontal axis (X-axis) runs along the width of the parking space, and the vertical axis (Y-axis) runs along the length of the parking space. The diagram also marks the position of the center point G of the ground transmitting coil (primary edge coil), and its horizontal coordinates Gx and vertical coordinates Gy in the parking space coordinate system after being analyzed by the visual ranging and positioning algorithm. This diagram visually illustrates the spatial relationship between the center point of the transmitting coil and the origin of the parking space.

[0037] After determining the origin O of the parking space, the vision system establishes a parking space coordinate system. The horizontal axis (X-axis) of the parking space coordinate system is along the width direction of the parking space, and the vertical axis (Y-axis) is along the length direction of the parking space, thus transforming the physical space of the target parking space into a two-dimensional plane that can be quantified by coordinates. Simultaneously, the vision system continues to perform image recognition on the parking space area, identifying the energy transmitting device (primary-side coil GA device) installed on the ground, especially the position of its center point G. To improve the accuracy and robustness of visual recognition, a visual marker code (such as an ArUco code) with a specific ID is usually affixed at the center point G of the ground transmitting coil. The center of this marker code coincides with the physical center point of the transmitting coil. After acquiring an image containing this marker code, the vision system uses visual ranging and positioning algorithms (such as perspective transformation / PnP algorithm) to parse the two-dimensional projection coordinates of the center point of the marker code, i.e., the center point G of the transmitting coil, in the established parking space coordinate system.

[0038] For example, after the vision system identifies the origin O of the parking space, it further identifies the marking code at the center point G of the transmitting coil. The algorithm analyzes and obtains the horizontal coordinates of point G in the parking space coordinate system as Gx=1200mm and the vertical coordinates as Gy=1200mm. That is, the position of the center point G of the transmitting coil relative to the origin of the left rear corner of the parking space is: 1200mm along the width direction of the parking space and 1200mm along the length direction of the parking space.

[0039] By establishing an independent parking space coordinate system on the parking space side and identifying and representing the coordinates of the center point of the transmitting coil in this coordinate system, the position of the transmitting coil is visualized and quantified. This coordinate value and the coordinate value of the receiving coil on the vehicle side are in two completely decoupled independent coordinate systems. The two do not need to be unified in advance or referenced to each other, which fundamentally eliminates the harsh conditions required by traditional vision solutions, such as strict alignment of the camera optical axis with the coil center or complex on-site calibration. At the same time, since the coordinates of the center point of the transmitting coil are visually identified and stored once during the vehicle parking process, even if the vehicle continues to move forward and the chassis of the vehicle gradually obscures the ground markers, the system can still continue to complete the alignment calculation based on the stored coordinate values ​​of the transmitting coil. This effectively overcomes the visual blind spot problem caused by the vehicle undercarriage obstruction, and significantly improves the robustness and environmental adaptability of the entire alignment process.

[0040] Step S300: During the process of parking the vehicle into the parking space, the vehicle vision system acquires in real time the lateral position, longitudinal position and orientation angle of the origin of the vehicle's own coordinate system in the parking space coordinate system. Step S400: Based on the lateral position, longitudinal position, and orientation angle of the origin of the vehicle's own coordinate system, and the pre-calibrated lateral and longitudinal relative coordinates of the center point of the receiving coil, calculate the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system through coordinate transformation.

[0041] Specifically, after establishing the parking space coordinate system and identifying the center point coordinates of the transmitting coil in step S200, steps S300 and S400 proceed to the stage of real-time vehicle pose acquisition and calculation of the theoretical position of the receiving coil. These two steps are continuously executed in a loop throughout the entire process of the vehicle parking in the parking space until alignment is completed.

[0042] As the vehicle moves toward the target parking space, the onboard vision system continuously acquires and recognizes images of the parking area. The vision system identifies the vehicle's position and orientation in real time within the established parking space coordinate system. Specifically, by recognizing fixed reference features such as parking lines and corner points, and combining this with the vehicle's geometric dimensions, the vision system calculates the lateral position Vo_x and longitudinal position Vo_y of the vehicle's own coordinate system origin Vo in the parking space coordinate system, as well as the vehicle's orientation angle θ. Here, Vo_x represents the offset of the vehicle's own coordinate system origin relative to the parking space origin O along the width of the parking space; Vo_y represents the offset of the origin relative to the parking space origin O along the length of the parking space; and θ represents the angle between the vehicle's longitudinal axis (i.e., the direction of the vehicle's front) and the longitudinal axis of the parking space coordinate system (i.e., the direction of the parking space's length). When θ = 0°, it indicates that the vehicle is perfectly aligned, with the front of the vehicle parallel to the length of the parking space; positive or negative values ​​of θ indicate clockwise or counterclockwise deflection of the vehicle.

[0043] For example, the vision system calculates in real time the lateral position Vo_x=600mm and the longitudinal position Vo_y=800mm of the vehicle's own coordinate system origin Vo in the parking space coordinate system at a certain moment, and the vehicle's orientation angle θ=10°.

[0044] After obtaining the real-time pose (Vo_x, Vo_y, θ) of the vehicle's own coordinate system origin, the lateral relative coordinates Vx and longitudinal relative coordinates Vy of the receiving coil center point V in the vehicle body coordinate system, which were pre-calibrated and stored in step S100, are called. Using a two-dimensional coordinate rotation transformation formula, the receiving coil center point V is transformed from the vehicle body coordinate system to the current parking space coordinate system, yielding the theoretical lateral position Vp_x and theoretical longitudinal position Vp_y of the receiving coil center point in the parking space coordinate system. The physical significance of this coordinate transformation is that, since the receiving coil is installed in a fixed position on the vehicle chassis, its relative coordinates (Vx, Vy) in the vehicle body coordinate system remain constant. However, when the vehicle moves or deflects within the parking space (i.e., Vo_x, Vo_y, θ change), the absolute position of the receiving coil center point in the parking space coordinate system also changes accordingly. The coordinate transformation formula is essentially a rotation and translation transformation on a two-dimensional plane. First, the fixed offset (Vx, Vy) of the center point of the receiving coil relative to the origin of the vehicle body is rotated around the origin of the vehicle body by the current orientation angle θ of the vehicle. Then, it is translated to the current position (Vo_x, Vo_y) of the origin of the vehicle's own coordinate system in the parking space coordinate system, thereby transforming the center point of the receiving coil into the parking space coordinate system for expression.

[0045] By acquiring the vehicle's pose information in the parking space coordinate system in real time through a vision system and performing coordinate transformations based on the pre-calibrated fixed coordinates of the receiving coil, the system can continuously calculate the real-time theoretical position of the receiving coil's center point in the parking space coordinate system as the vehicle moves. This achieves indirect, dynamic, and continuous acquisition of the receiving coil's position. Since this calculation process is entirely based on mathematical calculations of coordinate transformation, the vision system does not need to re-identify the receiving coil itself for each calculation. Even if the receiving coil is not visible due to being obscured by the vehicle chassis, the system can still accurately determine its current theoretical position, fundamentally solving the problem of blind spots caused by vehicle undercarriage obstruction during vehicle entry. At the same time, the continuous updating of the real-time pose makes the entire calculation process a dynamic feedback loop. The deviation calculation for each subsequent step is based on the latest vehicle pose rather than the accumulated historical values, effectively avoiding the amplification of positioning errors caused by integral drift accumulated over time in odometer or inertial navigation solutions, ensuring the accuracy and consistency of the alignment guidance process.

[0046] Step S500: Calculate the lateral deviation between the theoretical lateral position and the lateral coordinate of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinate of the center point of the transmitting coil.

[0047] Step S600: Based on the comparison results of the lateral deviation and longitudinal deviation with the corresponding preset threshold, determine whether the alignment is complete and output guidance information.

[0048] Specifically, after calculating the theoretical lateral position Vp_x and theoretical longitudinal position Vp_y of the center point of the receiving coil in the parking space coordinate system through coordinate transformation in step S400, the theoretical position is compared with the lateral coordinate Gx and longitudinal coordinate Gy of the center point of the transmitting coil in the parking space coordinate system that have been identified and stored in step S200, and the deviation between the two is calculated respectively.

[0049] Wherein, the lateral deviation Diff_x is equal to the lateral coordinate Gx of the center point of the transmitting coil minus the theoretical lateral position Vp_x of the center point of the receiving coil, and the longitudinal deviation Diff_y is equal to the longitudinal coordinate Gy of the center point of the transmitting coil minus the theoretical longitudinal position Vp_y of the center point of the receiving coil.

[0050] After calculating the lateral deviation Diff_x and longitudinal deviation Diff_y in step S500, these two deviation values ​​are compared with preset thresholds. The preset thresholds include a lateral preset threshold and a longitudinal preset threshold, which can be the same or different. That is, since the tolerance of the wireless charging system to deviation may vary in different directions, this application allows setting different threshold requirements for the lateral and longitudinal directions respectively.

[0051] For example, the preset lateral threshold is ±30mm, and the preset longitudinal threshold is ±50mm. Alignment is considered complete only when the absolute values ​​of the lateral and longitudinal deviations do not exceed their respective thresholds. First, Diff_x = -101mm is compared with the preset lateral threshold of ±30mm. Since 101mm > 30mm, the lateral deviation exceeds the limit. Simultaneously, Diff_y = -231mm is compared with the preset longitudinal threshold of ±50mm. Since 231mm > 50mm, the longitudinal deviation also exceeds the limit. Therefore, it is determined that alignment is not yet complete, and guidance information is output. For example, the deviation direction and distance information are displayed visually on the driver's interface, or directly sent to the automatic parking system as control commands to guide the vehicle to continue moving towards the target position. As the vehicle continuously adjusts its posture according to the guidance information, the vision system continuously updates the vehicle's posture data. Steps S300 to S600 are executed cyclically.

[0052] In this example, after adjustment, the vehicle's new pose is Vo_x = 450mm, Vo_y = 700mm, θ = 3°. After coordinate transformation, Vp_x ≈ 1222mm and Vp_y ≈ 1241mm. The recalculated deviations are Diff_x = 1200 - 1222 = -22mm and Diff_y = 1200 - 1241 = -41mm. At this point, |Diff_x| = 22mm ≤ 30mm and |Diff_y| = 41mm ≤ 50mm. Since the deviations in both directions fall within the preset threshold range, the system determines that alignment is complete, triggers locking, and sends a "alignment complete, charging allowed" signal to the wireless charging station to begin charging.

[0053] By calculating the difference between the theoretical position of the receiving coil and the actual position of the transmitting coil, the abstract goal of "alignment" is transformed into a quantifiable and comparable specific deviation value, providing a precise quantitative basis for subsequent guidance decisions. The system further achieves objective and automatic determination of whether alignment is complete by automatically comparing the deviation value with a preset threshold, without the need for manual intervention or external equipment confirmation, truly completing a fully automated closed loop from pose perception to alignment determination. At the same time, different thresholds can be set for the horizontal and vertical directions, allowing this method to flexibly adapt to the different tolerance requirements of different wireless charging systems in the X and Y directions, further improving the applicability and engineering practicality of the solution. The dynamic feedback loop formed by continuous deviation calculation and real-time comparison enables the vehicle to immediately obtain new deviation information and make a new round of determination after each pose adjustment, guiding the vehicle to gradually approach the target position and finally stop precisely in the alignment area, ensuring that the wireless charging system operates at the highest transmission efficiency.

[0054] In this embodiment, firstly, the method fully utilizes the vehicle's existing onboard vision system (such as surround-view cameras, front-view cameras, or rear-view cameras), eliminating the need for additional expensive hardware (such as low-frequency antennas, ultra-wideband modules, magnetic induction coil arrays, etc.). This significantly reduces system hardware costs and facilitates deployment in mass-produced vehicles. Secondly, by establishing independent and completely decoupled vehicle coordinate systems and parking space coordinate systems, the method separates the calibration of the receiving coil from the identification of the transmitting coil. The vision system only needs to identify any corner point of the parking space to establish the parking space coordinate system, fundamentally avoiding the stringent requirements of traditional solutions where the camera optical axis must be strictly aligned with the coil center or complex on-site calibration is required. This effectively reduces the interference of visual occlusion and blind spots on positioning accuracy. The method improves alignment accuracy significantly compared to traditional positioning schemes. Secondly, it fuses the pre-calibrated relative position of the receiving coil with the real-time vehicle pose through coordinate transformation, directly deriving the theoretical position of the receiving coil in the vehicle's coordinate system. This theoretical position is then compared with the visually recognized position of the transmitting coil, eliminating the need for cumulative extrapolation based on vehicle positioning. This fundamentally eliminates the accumulated errors caused by integral drift in odometer or inertial navigation systems, ensuring the consistency and reliability of each alignment result. Finally, the method completely decouples the coil alignment process from the charging function. Alignment determination and guidance output are both autonomously completed by the vehicle, eliminating the need for long-link signal interaction between the ground and vehicle ends. This simplifies the system control process and enhances the robustness and engineering feasibility of the entire wireless charging system.

[0055] Optionally, the coordinate transformation includes: Vp_x=Vo_x+Vx×cosθ-Vy×sinθ; Vp_y=Vo_y+Vx×sinθ-Vy×cosθ; Wherein, Vp_x is the theoretical lateral position, Vp_y is the theoretical longitudinal position, Vo_x is the lateral position of the origin of the vehicle's own coordinate system, Vo_y is the longitudinal position of the origin of the vehicle's own coordinate system; Vx is the lateral relative coordinate of the center point of the receiving coil, Vy is the longitudinal relative coordinate of the center point of the receiving coil; θ is the orientation angle of the origin of the vehicle's own coordinate system.

[0056] Specifically, in step S300, the lateral position Vo_x and longitudinal position Vo_y of the origin of the vehicle's own coordinate system in the parking space coordinate system are obtained in real time, as well as the vehicle's orientation angle θ. In step S100, the lateral relative coordinates Vx and longitudinal relative coordinates Vy of the center point of the receiving coil in the vehicle body coordinate system have been pre-calibrated and stored. In step S400, the theoretical lateral position Vp_x and theoretical longitudinal position Vp_y of the center point of the receiving coil in the parking space coordinate system are calculated using the following coordinate transformation formula.

[0057] The physical meaning of the above coordinate transformation formula is as follows: The fixed relative coordinates (Vx, Vy) of the receiving coil center point in the vehicle coordinate system are rotated around the origin of the vehicle's own coordinate system (Vo) by the current vehicle orientation angle θ, and then translated to the parking space coordinate system for expression. Specifically, Vx×cosθ-Vy×sinθ represents the rotational projection component of the receiving coil center point relative to the vehicle's origin along the width direction of the parking space, and Vx×sinθ+Vy×cosθ represents the rotational projection component of the receiving coil center point relative to the vehicle's origin along the length direction of the parking space. These two components are added to the lateral position (Vo_x) and longitudinal position (Vo_y) of the vehicle's own coordinate system origin in the parking space coordinate system, respectively, to obtain the absolute theoretical position of the receiving coil center point in the parking space coordinate system.

[0058] like Figure 4 As shown in the figure, the figure intuitively displays the real-time position (lateral position Vo_x, longitudinal position Vo_y) of the vehicle's own coordinate system origin Vo in the established parking space coordinate system and the vehicle's orientation angle θ during the process of parking the vehicle into the target parking space. Figure 4 The process of "real-time acquisition of the lateral position, longitudinal position and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system" in step S300 is presented graphically. At the same time, it provides visual assistance for "calculating the theoretical position of the receiving coil center point in the parking space coordinate system through coordinate transformation" in step S400. That is, Vo_x, Vo_y and θ shown in the figure are the three key input parameters of the coordinate transformation formula.

[0059] Calculations are performed using example data from a specific embodiment of this application: When the vehicle leaves the factory, the lateral relative coordinates of the center point of the receiving coil in the vehicle coordinate system are pre-calibrated as Vx = 800mm and the longitudinal relative coordinates as Vy = 500mm. At a certain moment during the vehicle parking process, the vision system acquires in real time the lateral position Vo_x = 600mm and the longitudinal position Vo_y = 800mm of the origin of the vehicle's own coordinate system in the parking space coordinate system, with a vehicle orientation angle θ = 10°. Substituting these values ​​into the above formula, the theoretical lateral position of the center point of the receiving coil in the parking space coordinate system is calculated as Vp_x = 600 + 800 × cos10° - 500 × sin10° ≈ 600 + 787.8 - 86.8 ≈ 1301mm, and the theoretical longitudinal position is Vp_y = 800 + 800 × sin10° + 500 × cos10° ≈ 800 + 138.9 + 492.4 ≈ 1431mm. The calculation results show that, under the current vehicle pose, the center point of the receiving coil installed on the chassis is located approximately 1301 mm in width and 1431 mm in length from the origin of the parking space in the parking space coordinate system.

[0060] By fusing the fixed relative coordinates of the receiving coil in the vehicle coordinate system with the real-time changing vehicle pose information, the system can quickly and accurately calculate the real-time theoretical position of the receiving coil center point in the parking space coordinate system in each control cycle without relying on any additional position sensors or ground auxiliary signals. This is achieved solely through mathematical transformations, resulting in low computational complexity and fast response, meeting the real-time control requirements during vehicle parking. Furthermore, since the input parameters (Vo_x, Vo_y, θ) in the formula are all results of independent real-time measurements by the vision system, rather than cumulative values ​​based on historical positions, the calculated Vp_x and Vp_y do not contain any drift errors accumulated over time. This fundamentally ensures the accuracy and consistency of the theoretical position calculation results, providing a reliable data foundation for subsequent deviation calculations and alignment determination.

[0061] Optionally, the criteria for determining whether alignment is complete include: the absolute values ​​of both the lateral deviation and the longitudinal deviation are less than or equal to the corresponding preset thresholds.

[0062] Specifically, after calculating the lateral deviation Diff_x and longitudinal deviation Diff_y in step S500, these two deviation values ​​are compared with preset thresholds. The preset thresholds include a lateral preset threshold and a longitudinal preset threshold, which can be the same or different. Since the tolerance of the wireless charging system to deviation may differ in different directions, this application allows setting different threshold requirements for the lateral and longitudinal directions respectively. The criterion for determining that alignment is complete is: the absolute value of the lateral deviation Diff_x is less than or equal to the lateral preset threshold, and the absolute value of the longitudinal deviation Diff_y is less than or equal to the longitudinal preset threshold.

[0063] For example, if the preset horizontal threshold is 30mm and the preset vertical threshold is 50mm, then when |Diff_x|≤30mm and |Diff_y|≤50mm, the alignment is considered complete.

[0064] By quantifying the abstract alignment target into deterministic rules that compare the lateral and longitudinal deviations with their respective thresholds, the objective and automatic determination of whether alignment is complete is achieved without manual intervention or reliance on external devices for confirmation. At the same time, different thresholds can be set for the lateral and longitudinal directions, enabling this method to flexibly adapt to the different tolerance requirements of different wireless charging systems in the X and Y directions, further improving the applicability and engineering practicality of the solution.

[0065] Optionally, the vehicle vision system includes at least one of a surround-view camera, a front-view camera, or a rear-view camera, for identifying parking lines, parking corners, or visual markings affixed to the ground.

[0066] Optionally, determining the origin of the parking space based on the parking space features identified by the vehicle vision system includes: Based on the top view output by the surround view camera in the vehicle vision system, the parking space lines are obtained by identifying the top view based on edge detection and Hough transform, and the corner point of the parking space is determined as the origin of the parking space.

[0067] Optionally, the step of identifying parking lines from the top view based on edge detection and Hough transform, and determining the corner point of the parking space as the origin of the parking space, includes: Based on the edge detection, the edge feature points of the parking lines are extracted from the top view; By using the Hough transform, the straight line equation of the parking space boundary line is fitted to the edge feature points, and the intersection point of the straight line equations of two adjacent parking space boundary lines is found. The intersection point is determined as the corner point of the parking space, and the corner point of the parking space is used as the origin of the parking space.

[0068] Specifically, the vehicle vision system includes at least one of surround-view cameras, front-view cameras, or rear-view cameras. These cameras are all existing visual sensor hardware in the vehicle and require no additional installation. Surround-view cameras are typically installed at the front, rear, left, and right of the vehicle, and can output a top-down view of the vehicle's surroundings through image stitching, suitable for recognizing ground features such as parking lines and parking space corners. Front-view cameras are installed at the front of the vehicle, and rear-view cameras are installed at the rear, respectively suitable for recognizing parking space areas and ground markings in forward or reverse parking scenarios. These cameras work together to recognize parking lines, parking space corners, or visual markings (such as ArUco codes) affixed to the ground.

[0069] The entire alignment process is completed by making full use of the vehicle's existing visual sensor resources, without the need to add any additional hardware such as low-frequency antennas, UWB modules, or magnetic induction coil arrays. This significantly reduces system hardware costs while simplifying the equipment layout on the vehicle side, making it easy to directly promote and deploy in mass-produced vehicles.

[0070] Based on the top-view output from the surround-view camera in the vehicle vision system, edge detection and Hough transform are used to identify the parking space lines, and then the corner points of the parking spaces are determined as the origin points. The specific implementation process is as follows: After the surround-view camera captures images of the area around the vehicle, the images are stitched together and processed using inverse perspective transformation to output a top-down view of the area surrounding the vehicle. In this top-down view, parking lines appear as straight edge regions with certain width and directional characteristics. First, the top-down view is preprocessed (e.g., image grayscale conversion, filtering, and denoising). Then, an edge detection algorithm (e.g., the Canny edge detection operator) is used to extract locations in the top-down view where grayscale changes drastically; these locations correspond to the edge feature points of the parking lines. Subsequently, the system uses Hough transform to process the extracted edge feature points. The core idea of ​​Hough transform is to transform discrete edge points in image space, which are difficult to fit directly, into easily detectable peak points through parameter space mapping, thereby fitting the straight line equations of the parking space boundary lines. After obtaining the straight line equations of multiple parking space boundary lines, the intersection point of the straight line equations of two adjacent parking space boundary lines is found. This intersection point is a corner point of the parking space, and this corner point is determined as the origin O of the parking space.

[0071] For example, during the process of a vehicle parking in a target parking space, a surround-view camera captures an image containing the target parking space and generates a top-down view. Canny edge detection is performed on the top-down view to extract the edge feature points of the four boundary lines of the parking space. Then, Hough transform is used to process these edge feature points and fit the linear equations of the four parking space boundary lines. The intersection point of the linear equations of two adjacent boundary lines is selected. For example, the intersection point of the left boundary line and the rear boundary line of the parking space is selected. This intersection point is the left rear corner point of the parking space, which is determined as the origin O of the parking space, and a parking space coordinate system is established accordingly.

[0072] By combining edge detection and Hough transform, the visual recognition problem of parking lines is transformed into a mathematical problem of image edge extraction and straight line parameter fitting, achieving accurate and stable recognition of parking space corner points. Edge detection effectively extracts the gray-level difference boundary between the parking line and the ground background, while Hough transform stably fits the straight line equation of the parking space boundary line even under interference from noise, partial occlusion, or uneven lighting. The two complement each other, significantly improving the robustness of parking space feature recognition. The use of a top-down view makes the parking lines appear as a more regular geometric shape in the image, eliminating the interference of perspective distortion on straight line detection, reducing the complexity of subsequent straight line fitting and intersection point calculation, and further improving the accuracy and reliability of parking space origin positioning. Finally, the parking space corner points obtained from the intersection point calculation are used as the parking space origin, giving the establishment of the parking space coordinate system a clear physical reference, providing a stable and unified coordinate benchmark for subsequent transmitting coil coordinate recognition and vehicle pose calculation.

[0073] Optionally, the step of resolving the lateral and longitudinal coordinates of the center point of the transmitting coil in the parking space coordinate system using a visual ranging and positioning algorithm includes: Acquire an image containing the ArUco visual marker code pasted at the center point of the transmitting coil; The perspective transformation algorithm is used to analyze the two-dimensional projection of the center point of the ArUco visual marker code in the parking space coordinate system, and the horizontal and vertical coordinates are used as the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system.

[0074] Specifically, to improve the accuracy and robustness of the visual sensor in recognizing the center point G of the transmitting coil, this application affixes an ArUco visual marker with a specific ID at the center point G of the transmitting coil. The physical center of this marker precisely coincides with the geometric center of the transmitting coil. The ArUco code is a reference marker widely used in camera pose estimation. It encodes unique ID information internally, which can be quickly detected and recognized by visual algorithms. At the same time, its outer black border provides stable corner features, facilitating high-precision image localization.

[0075] During actual parking, the vehicle-mounted vision system (such as surround-view cameras, front-view or rear-view cameras) acquires images of the ground area in real time. When the vehicle enters the parking space and the ArUco code at the center point of the transmitting coil enters the camera's field of view, an image containing the ArUco visual marker code is obtained. Subsequently, a visual ranging and positioning algorithm is used to process the image. Specifically, the Perspective-n-Point (PnP) algorithm is used to analyze the two-dimensional projection coordinates of the ArUco code's center point in the parking space coordinate system.

[0076] The basic principle of the PnP algorithm is as follows: given the coordinates of several three-dimensional points in the world coordinate system and their corresponding two-dimensional projection points in the image, the camera's pose relative to the world coordinate system is recovered by solving the camera's rotation and translation matrices, and then the coordinates of any target point in the world coordinate system are calculated.

[0077] In the application scenario of this application, the physical dimensions of the ArUco code are known (i.e., the three-dimensional coordinates of its four corner points in the world coordinate system are known). After the vision system detects the two-dimensional pixel coordinates of the four corner points of the ArUco code in the image, it combines the camera intrinsic parameters (focal length, principal point coordinates, etc.) and uses the PnP algorithm to solve for the three-dimensional position of the ArUco code center point relative to the origin of the parking space coordinate system. Then, it extracts its two-dimensional projection on the parking space plane, that is, the horizontal and vertical coordinates of the ArUco code center point in the parking space coordinate system. Since the ArUco code center point physically coincides with the center point of the transmitting coil, these two-dimensional projection coordinates are the horizontal coordinates Gx and vertical coordinates Gy of the transmitting coil center point in the parking space coordinate system.

[0078] For example, after the ArUco code is pasted at the center point G of the ground transmitting coil, the vision system successfully recognizes the ArUco code during the vehicle parking process. The PnP algorithm is used to analyze the code and obtain its lateral coordinates Gx=1200mm and longitudinal coordinates Gy=1200mm in the parking space coordinate system.

[0079] By affixing ArUco visual markers to the center point of the ground transmitting coil, the center point of the physical object, the transmitting coil, is transformed into a recognizable visual target with a unique ID and stable corner features. This enables the vision system to quickly and accurately detect the position of the transmitting coil in complex ground environments, significantly reducing the difficulty of target detection and localization in image processing. Simultaneously, combined with the PnP perspective transformation algorithm, the algorithm can accurately calculate the precise coordinates of the marker's center point in the parking space coordinate system, even if the camera's optical axis is not strictly perpendicular to the marker plane, even if the vehicle parks at a certain angle and the camera captures the ground marking from a tilted perspective. This effectively overcomes the limitations of viewing angle and the complexity of on-site calibration in traditional vision solutions. Furthermore, the ArUco code's own ID encoding mechanism allows multiple different parking spaces or different transmitting coils to use their own unique IDs, avoiding confusion between markers and further improving the system's reliability and anti-interference capabilities in complex parking lot environments.

[0080] Optionally, the method of outputting the guidance information includes: outputting the lateral deviation and the longitudinal deviation as correction values ​​to the automatic parking system on the vehicle, and the automatic parking system controlling the vehicle to make fine adjustments based on the correction values.

[0081] Optionally, controlling the vehicle to make fine-tuning movements based on the correction amount includes: Based on the vehicle's Ackerman steering geometry model, with the lateral and longitudinal deviations approaching preset target values ​​as control objectives, a smooth trajectory is dynamically planned, and real-time speed and front wheel steering angle commands are output to control the vehicle's movement.

[0082] Specifically, in step S600, when the system determines that alignment is not yet complete based on the comparison between the deviation and a preset threshold, it needs to output guidance information to instruct the vehicle to continue adjusting its posture. The guidance information is output by using the lateral deviation Diff_x and longitudinal deviation Diff_y calculated in step S500 as correction values ​​and outputting them to the automatic parking system (APA) on the vehicle. Upon receiving this correction value, the automatic parking system uses it as a control target, ensuring that the lateral and longitudinal deviations approach zero (or approach a preset target value, which is the allowable deviation range corresponding to complete alignment, such as a lateral deviation approaching 0mm, a longitudinal deviation approaching 0mm, or any acceptable value within a preset threshold range).

[0083] Subsequently, based on the vehicle's Ackerman steering geometry model and the control objective, the automatic parking system dynamically plans a continuous and smooth trajectory from the vehicle's current position to the target alignment position (with zero deviation or falling within a preset threshold range). During trajectory planning, real-time speed and front wheel steering angle commands are output to the vehicle's chassis actuators (such as the motor drive system and steering system) in each control cycle, controlling the vehicle to make fine adjustments along the planned trajectory. Each time the vehicle moves, the onboard vision system re-acquires the vehicle's pose (i.e., step S300), recalculates the theoretical position and deviation (i.e., steps S400 to S500), and feeds the updated deviation back to the automatic parking system. The model prediction controller then corrects the subsequent trajectory online based on the latest deviation information. This cycle of "real-time pose acquisition → deviation calculation → feedback correction → MPC trajectory planning → output control commands → vehicle movement" continues until both lateral and longitudinal deviations meet the alignment completion conditions.

[0084] For example, the initial calculation yielded a lateral deviation of Diff_x = -101mm and a longitudinal deviation of Diff_y = -231mm, both exceeding the preset thresholds (lateral ±30mm, longitudinal ±50mm). The system outputs Diff_x = -101mm and Diff_y = -231mm as correction values ​​to the automatic parking system. The model predictive controller in the automatic parking system is based on the vehicle's Ackerman steering geometry model. This model simplifies the vehicle into a kinematic system with the front wheel steering angle as input and the vehicle's position and orientation angle as states. It describes the vehicle's motion in low-speed parking scenarios, aiming for lateral and longitudinal deviations to approach zero. It plans a smooth trajectory from the current pose (Vo_x = 600mm, Vo_y = 800mm, θ = 10°) to the target alignment position (zero deviation), and outputs corresponding speed and front wheel steering angle commands to control the vehicle for fine-tuning. After the vehicle is adjusted, the vision system acquires a new pose (Vo_x=450mm, Vo_y=700mm, θ=3°), and recalculates to obtain Diff_x=-22mm and Diff_y=-41mm. Both fall within the preset threshold range, so the alignment is determined to be complete, and the fine-tuning movement stops.

[0085] By directly inputting the deviation value calculated by the vision system as a correction value into the automatic parking system, seamless integration between the vision perception layer and the control execution layer is achieved, completing the entire closed loop from deviation detection to vehicle adjustment without manual intervention. The model predictive controller performs trajectory planning based on the vehicle's Ackerman steering geometry model, fully considering the vehicle's own kinematic constraints (such as minimum turning radius, maximum steering angle, etc.), making the planned trajectory not only geometrically feasible but also physically executable. This avoids the problems of control commands failing to execute or having excessive execution deviations due to neglecting vehicle kinematic characteristics in traditional control methods. At the same time, the MPC algorithm has the inherent advantage of predictive control, that is, it predicts the system behavior for a period of time in the future based on the current state in each control cycle and solves the optimal control sequence. Therefore, it can continuously correct the trajectory online as the vehicle approaches the target position, effectively cope with environmental disturbances and model errors, and ensure that the vehicle finally stops accurately in the alignment position. The real-time closed-loop feedback mechanism composed of vision system and MPC controller makes each fine adjustment based on the latest measured vehicle pose rather than historical recursive values, fundamentally eliminating the impact of accumulated errors on the final alignment accuracy and ensuring high precision and high reliability of the entire alignment process.

[0086] This invention provides a wireless charging transmitter and receiver coil alignment system based on vehicle vision, comprising: The acquisition unit is used to pre-calibrate the lateral and longitudinal relative coordinates of the center point of the receiving coil in the vehicle coordinate system, wherein the vehicle coordinate system takes a preset point on the vehicle body as the origin, the horizontal axis is along the width of the vehicle, and the vertical axis is along the front of the vehicle. The parsing unit is used to identify parking space features based on the vehicle vision system, determine the origin of the parking space, and establish a parking space coordinate system. The parking space coordinate system has the origin of the parking space as its origin, with the horizontal axis along the width direction of the parking space and the vertical axis along the length direction of the parking space. Based on the image of the marker at the center point of the transmitting coil acquired by the vehicle vision system, the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system are parsed using a visual ranging and positioning algorithm. The acquisition unit is also used to acquire, in real time, the lateral position, longitudinal position, and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system based on the vehicle vision system during the process of parking the vehicle into the parking space. The transformation unit is used to calculate the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system by means of coordinate transformation, based on the lateral position, longitudinal position and orientation angle of the origin of the vehicle's own coordinate system, and the lateral and longitudinal relative coordinates of the center point of the receiving coil that have been pre-calibrated. The calculation unit is used to calculate the lateral deviation between the theoretical lateral position and the lateral coordinates of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinates of the center point of the transmitting coil; The comparison unit is used to determine whether the alignment is complete based on the comparison results of the lateral deviation and longitudinal deviation with the preset threshold, and output guidance information.

[0087] This invention provides a wireless charging transmitter and receiver coil alignment device based on vehicle vision, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the wireless charging transmitter and receiver coil alignment method based on vehicle vision as described above when the computer program is executed.

[0088] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the above-described method for aligning wireless charging transmitting and receiving coils based on vehicle vision.

[0089] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for aligning wireless charging transmitting and receiving coils based on vehicle vision, characterized in that, include: The horizontal and vertical relative coordinates of the center point of the receiving coil in the vehicle coordinate system are pre-calibrated. The vehicle coordinate system takes a preset point on the vehicle body as the origin, with the horizontal axis along the width of the vehicle and the vertical axis along the front of the vehicle. Based on the vehicle vision system, the parking space features are identified, the origin of the parking space is determined, and a parking space coordinate system is established. The parking space coordinate system takes the origin of the parking space as its origin, with the horizontal axis along the width direction of the parking space and the vertical axis along the length direction of the parking space. The vehicle vision system acquires an image containing a marker at the center point of the transmitting coil, and the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system are analyzed by a visual ranging and positioning algorithm. During the process of parking a vehicle into a parking space, the vehicle vision system acquires in real time the lateral position, longitudinal position, and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system. Based on the lateral position, longitudinal position, and orientation angle of the origin of the vehicle's own coordinate system, as well as the pre-calibrated lateral and longitudinal relative coordinates of the center point of the receiving coil, the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system are calculated through coordinate transformation. Calculate the lateral deviation between the theoretical lateral position and the lateral coordinates of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinates of the center point of the transmitting coil; Based on the comparison results of the lateral and longitudinal deviations with the corresponding preset thresholds, it is determined whether the alignment is complete, and guidance information is output.

2. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 1, characterized in that, The coordinate transformation includes: Vp_x=Vo_x+Vx×cosθ-Vy×sinθ; Vp_y=Vo_y+Vx×sinθ-Vy×cosθ; Wherein, Vp_x is the theoretical lateral position, Vp_y is the theoretical longitudinal position, Vo_x is the lateral position of the origin of the vehicle's own coordinate system, Vo_y is the longitudinal position of the origin of the vehicle's own coordinate system; Vx is the lateral relative coordinate of the center point of the receiving coil, Vy is the longitudinal relative coordinate of the center point of the receiving coil; θ is the orientation angle of the origin of the vehicle's own coordinate system.

3. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 1, characterized in that, The criteria for determining whether alignment is complete include: the absolute values ​​of both the lateral deviation and the longitudinal deviation are less than or equal to the corresponding preset thresholds.

4. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 1, characterized in that, The vehicle vision system includes at least one of a surround-view camera, a front-view camera, or a rear-view camera, used to identify parking lines, parking corner points, or visual markings affixed to the ground.

5. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 4, characterized in that, The step of determining the origin of a parking space based on the characteristics of the parking space identified by the vehicle-mounted vision system includes: Based on the top view output by the surround view camera in the vehicle vision system, the parking space lines are obtained by identifying the top view based on edge detection and Hough transform, and the corner point of the parking space is determined as the origin of the parking space.

6. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 5, characterized in that, The step of identifying parking lines from the top view based on edge detection and Hough transform, and determining the corner points of the parking spaces as the origin points of the parking spaces, includes: Based on the edge detection, the edge feature points of the parking lines are extracted from the top view; By using the Hough transform, the straight line equation of the parking space boundary line is fitted to the edge feature points, and the intersection point of the straight line equations of two adjacent parking space boundary lines is found. The intersection point is determined as the corner point of the parking space, and the corner point of the parking space is used as the origin of the parking space.

7. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 1, characterized in that, The step of resolving the lateral and longitudinal coordinates of the center point of the transmitting coil in the parking space coordinate system using a visual ranging and positioning algorithm includes: Acquire an image containing the ArUco visual marker code pasted at the center point of the transmitting coil; The perspective transformation algorithm is used to analyze the two-dimensional projection of the center point of the ArUco visual marker code in the parking space coordinate system, and the horizontal and vertical coordinates are used as the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system.

8. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 1, characterized in that, The method of outputting guidance information includes: outputting the lateral deviation and the longitudinal deviation as correction values ​​to the automatic parking system on the vehicle, and the automatic parking system controlling the vehicle to make fine adjustments based on the correction values.

9. The wireless charging transmitting and receiving coil alignment method based on vehicle vision according to claim 8, characterized in that, The step of controlling the vehicle to make fine adjustments based on the correction amount includes: Based on the vehicle's Ackerman steering geometry model, with the lateral and longitudinal deviations approaching preset target values ​​as control objectives, a smooth trajectory is dynamically planned, and real-time speed and front wheel steering angle commands are output to control the vehicle's movement.

10. A wireless charging transmitting and receiving coil alignment system based on vehicle vision, characterized in that, For implementing the wireless charging transmitting and receiving coil alignment method based on vehicle vision as described in any one of claims 1 to 9, the system comprises: The acquisition unit is used to pre-calibrate the lateral and longitudinal relative coordinates of the center point of the receiving coil in the vehicle coordinate system, wherein the vehicle coordinate system takes a preset point on the vehicle body as the origin, the horizontal axis is along the width of the vehicle, and the vertical axis is along the front of the vehicle. The parsing unit is used to identify parking space features based on the vehicle vision system, determine the origin of the parking space, and establish a parking space coordinate system. The parking space coordinate system has the origin of the parking space as its origin, with the horizontal axis along the width direction of the parking space and the vertical axis along the length direction of the parking space. Based on the image of the marker at the center point of the transmitting coil acquired by the vehicle vision system, the horizontal and vertical coordinates of the center point of the transmitting coil in the parking space coordinate system are parsed using a visual ranging and positioning algorithm. The acquisition unit is also used to acquire, in real time, the lateral position, longitudinal position, and orientation angle of the vehicle's own coordinate system origin in the parking space coordinate system based on the vehicle vision system during the process of parking the vehicle into the parking space. The transformation unit is used to calculate the theoretical lateral and longitudinal positions of the center point of the receiving coil in the current parking space coordinate system by means of coordinate transformation, based on the lateral position, longitudinal position and orientation angle of the origin of the vehicle's own coordinate system, and the lateral and longitudinal relative coordinates of the center point of the receiving coil that have been pre-calibrated. The calculation unit is used to calculate the lateral deviation between the theoretical lateral position and the lateral coordinates of the center point of the transmitting coil, and the longitudinal deviation between the theoretical longitudinal position and the longitudinal coordinates of the center point of the transmitting coil; The comparison unit is used to determine whether the alignment is complete based on the comparison results of the lateral deviation and longitudinal deviation with the preset threshold, and output guidance information.

Images