A new energy vehicle wireless charging multi-modal dynamic positioning detection system and method

Abstract

The present application relates to the technical field of new energy vehicle wireless charging, and particularly relates to a new energy vehicle wireless charging multi-modal dynamic positioning detection system and method. The new energy vehicle wireless charging multi-modal dynamic positioning detection system comprises a millimeter wave radar module, a visual acquisition module and an inertial measurement module. The millimeter wave radar module is used to identify metal positioning points in a charging plate marking unit to obtain three-dimensional point cloud coordinates. The visual acquisition module is used to identify image acquisition marks in the charging plate marking unit to obtain two-dimensional image coordinates. The inertial measurement module is used to acquire vehicle body attitude data. A central processing module is used to fuse the three-dimensional point cloud coordinates and the two-dimensional image coordinates to obtain an initial relative position, and to perform dynamic error compensation on the initial relative position according to the vehicle body attitude data to obtain a corrected relative position. The system combines millimeter wave radar, visual semantic information and attitude sensing, solves the problem of low positioning accuracy of single electromagnetic induction, and realizes sub-centimeter level positioning.

Description

Technical Field

[0001] This invention relates to the field of wireless charging technology for new energy vehicles, and in particular to a multimodal dynamic positioning and detection system and method for wireless charging of new energy vehicles. Background Technology

[0002] In wireless charging technology for new energy vehicles, the positioning of the vehicle and the ground charging pad can be mainly divided into three categories: Electromagnetic induction positioning technology: By setting induction coils on the ground charging pad and the vehicle-mounted receiver, the relative position is determined by the strength of the electromagnetic coupling signal; such as Chinese patent application: CN116272345A; Visual positioning technology: The vehicle-mounted camera captures visual markings (such as QR codes or geometric patterns) on the ground charging pad, and combines them with image recognition algorithms to calculate the relative position, such as Chinese patent: CN217847896U; Satellite positioning technology: The absolute position of the vehicle is obtained by combining GPS / BeiDou positioning modules and compared with the preset coordinates of the charging board to achieve positioning, as described in the technical solution in the paper "Design of Wireless Charging Parking Guidance System Based on BeiDou Positioning" (Electronic Technology Application, Vol. 48, No. 3, 2022).

[0003] However, the above-mentioned solutions have technical problems such as low accuracy, susceptibility to interference, lack of dynamic compensation, and lack of fault tolerance. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a multimodal dynamic positioning and detection system and method for wireless charging of new energy vehicles.

[0005] This invention includes the following technical solutions: The first aspect of this invention provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, comprising: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0006] Furthermore, the charging pad marking unit includes metal positioning points and image acquisition marks, the image acquisition marks including a central QR code and a ring-shaped infrared dot matrix; The central QR code is located at the center of the ground wireless charging pad. When there is sufficient sunlight during the day, the visual acquisition module can identify the central QR code. The ring-shaped infrared dot array is arranged around the central QR code. In the dark or nighttime environment, the visual acquisition module can identify the ring-shaped infrared dot array. The central QR code, the ring-shaped infrared dot matrix, and the outer ring metal positioning points are all coaxially arranged with the coil center of the ground wireless charging pad.

[0007] Furthermore, the annular infrared dot array includes multiple uniformly distributed infrared LEDs.

[0008] Furthermore, the outer ring metal positioning points are stainless steel cylinders, and the outer ring metal positioning points are set at the four corners of the ground wireless charging plate.

[0009] Furthermore, the vehicle-mounted display module is used to display the corrected relative position and guidance instructions; A charging control module, connected to the central processing module, is used to control charging according to the charging instructions from the central processing module.

[0010] Furthermore, it also includes an automatic parking module. If the corrected relative position does not meet the charging start threshold, the central processing module sends a guidance command to the automatic parking module to automatically adjust the vehicle position until the corrected relative position meets the charging start threshold.

[0011] A second aspect of this invention provides a multimodal dynamic positioning and detection method for wireless charging of new energy vehicles, comprising the following steps: The three-dimensional point cloud coordinates of the ground wireless charging pad are obtained through the millimeter-wave radar module, the two-dimensional image coordinates of the ground wireless charging pad are obtained through the vision acquisition module, and the vehicle body attitude data are obtained through the inertial measurement module. The initial relative position is obtained by fusing the three-dimensional point cloud coordinates with the two-dimensional image coordinates; Based on the vehicle posture data, dynamic error compensation is performed on the initial relative position to obtain the corrected relative position; Determine whether the corrected relative position meets the charging start threshold. If it does, output a charging command; otherwise, output a guidance command.

[0012] Furthermore, the three-dimensional point cloud coordinates are fused with the two-dimensional image coordinates using an improved ICP registration algorithm.

[0013] Furthermore, the improved ICP registration algorithm fuses the three-dimensional point cloud coordinates with the two-dimensional image coordinates, including: Using two-dimensional image coordinates as the initial reference point, the matching range of three-dimensional point cloud coordinates is narrowed; The three-dimensional point cloud coordinates are assigned a first weight, and the two-dimensional image coordinates are assigned a second weight. The initial relative positions are obtained by weighted fusion.

[0014] Furthermore, the correction of relative position includes: X-axis coordinate, Y-axis coordinate, and angular deviation.

[0015] Furthermore, the charging start threshold is: X-axis deviation ≤ 0.1cm~0.5cm, Y-axis deviation ≤ 0.1cm~0.5cm, and angle deviation ≤ 0.1°~0.3°.

[0016] Furthermore, the visual acquisition module switches to normal recognition mode during the day when there is sufficient light to recognize the central QR code of the charging pad marking unit, and switches to infrared mode at night or in a dark environment to recognize the ring infrared dot matrix of the charging pad marking unit.

[0017] By adopting the above technical solution, the present invention has the following advantages: 1. The system of the present invention solves the problem of low positioning accuracy of single electromagnetic induction by combining millimeter-wave radar, visual semantic information and attitude perception, and achieves sub-centimeter-level positioning; it also realizes an automated closed loop from perception to execution by judging the charging start threshold and outputting the guidance command, thereby improving the intelligence level of the system.

[0018] 2. The system of the present invention completely solves the problem of low recognition rate of visual positioning in low light conditions by combining the "central QR code" (daytime) and the "ring infrared dot matrix" at night / in darkness.

[0019] 3. The central QR code, the ring infrared dot matrix, and the outer metal positioning points of this invention are all coaxial with the coil center of the ground wireless charging board, ensuring that the coordinates identified by vision / radar are the center of charging energy transmission, thus eliminating the cumulative error caused by intermediate conversion.

[0020] 3. This invention uses multiple evenly distributed infrared LEDs to ensure that the visual acquisition module can obtain stable light spot features at night or in dark environments, avoiding image distortion or recognition failure caused by uneven light sources.

[0021] 4. The outer ring metal positioning points of this invention are defined as stainless steel cylinders and placed at the four corners, which not only ensures that the millimeter-wave radar can obtain strong reflected signals (improving the signal-to-noise ratio), but also utilizes the crush resistance of stainless steel to adapt to the complex environment of parking lots.

[0022] 5. This invention introduces an automatic parking module, upgrading the positioning detection system into part of the autonomous driving system; when the deviation is too large, the vehicle can make minor adjustments on its own without the need for manual operation by the driver, thus achieving truly fully automatic wireless charging.

[0023] 6. This invention solves the problems of long iterations and slow convergence of traditional ICP algorithms by defining an "improved ICP registration algorithm", thereby reducing data processing latency to less than 10ms and meeting the requirements of low-speed dynamic positioning of vehicles.

[0024] 7. The present invention assigns radar weights of 0.6 (emphasizing stability) and vision weights of 0.4 (emphasizing two-dimensional accuracy). This specific weight allocation enables the fusion accuracy to reach within 0.3cm.

[0025] 8. The system of the present invention has extremely strong adaptability to complex environments and a very high positioning success rate.

[0026] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures pointed out in the description and the drawings. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the ground wireless charging pad in an embodiment of the present invention; Figure 3 This is a flowchart of a multimodal dynamic positioning and detection method for wireless charging of new energy vehicles in an embodiment of the present invention. Detailed Implementation

[0029] The following description provides many different embodiments or examples for implementing various features of the invention. The elements and arrangements described in the specific examples below are only for concise expression of the invention and are merely examples, not intended to limit the invention.

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of 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, 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.

[0031] Example 1 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, combined with... Figure 1 As shown, it includes: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0032] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0033] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0034] For example, the inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β) and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0035] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0036] Example 2 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, combined with... Figure 1 As shown, it includes: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0037] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0038] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0039] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0040] Combination Figure 2 As shown, the charging pad marking unit includes metal positioning points and image acquisition marks, and the image acquisition marks include a central QR code and a ring-shaped infrared dot matrix; The central QR code is located at the center of the ground wireless charging pad. When there is sufficient sunlight during the day, the visual acquisition module can identify the central QR code. The ring-shaped infrared dot array is arranged around the central QR code. In the dark or nighttime environment, the visual acquisition module can identify the ring-shaped infrared dot array. The central QR code, the ring-shaped infrared dot matrix, and the outer ring metal positioning points are all coaxially arranged with the coil center of the ground wireless charging pad.

[0041] Example 3 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, including: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0042] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0043] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0044] For example, the inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β) and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0045] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0046] The charging pad marking unit includes metal positioning points and image acquisition marks, and the image acquisition marks include a central QR code and a ring-shaped infrared dot matrix. The central QR code is located at the center of the ground wireless charging pad. When there is sufficient sunlight during the day, the visual acquisition module can identify the central QR code. The ring-shaped infrared dot array is arranged around the central QR code. In the dark or nighttime environment, the visual acquisition module can identify the ring-shaped infrared dot array. The central QR code, the ring-shaped infrared dot matrix, and the outer ring metal positioning points are all coaxially arranged with the coil center of the ground wireless charging pad.

[0047] The annular infrared dot array comprises multiple uniformly distributed infrared LEDs.

[0048] The outer ring metal positioning points are stainless steel cylinders, and the outer ring metal positioning points are set at the four corners of the ground wireless charging pad.

[0049] For example, combined Figure 2 As shown, the ground wireless charging pad adopts a circular structure with a diameter of 60cm, and the coil is located in the central area of ​​the charging pad (diameter 40cm).

[0050] For example, combined Figure 2 As shown, the central QR code uses the QR code format to store information such as the latitude and longitude coordinates of the charging pad coil center and the coil radius. The QR code is 10cm×10cm in size and is printed on wear-resistant and waterproof black printing material.

[0051] For example, combined Figure 2 As shown, the ring-shaped infrared dot array consists of six 940nm infrared LEDs arranged in a ring (20cm in diameter). The LEDs have a power of 50mW, a beam angle of 120°, and are powered by an independent power supply. They emit light continuously and can be recognized by the visual acquisition module even in complete darkness.

[0052] For example, combined Figure 2 As shown, the metal positioning points consist of four cylindrical positioning points made of stainless steel (1cm in diameter and 0.5cm in height), which are respectively set at the four corners of the ground wireless charging plate (30cm from the center). The material is 304 stainless steel to ensure that the millimeter-wave radar module can stably reflect radar waves.

[0053] Example 4 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, including: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0054] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0055] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0056] For example, the inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β) and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0057] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0058] The vehicle-mounted display module is used to display the relative position correction and guidance instructions; A charging control module, connected to the central processing module, is used to control charging according to the charging instructions from the central processing module.

[0059] For example, the vehicle display module uses a vehicle central control touch screen (10.25 inches, 1920×720 resolution) to display corrected relative position and guidance instructions to guide the user to adjust the vehicle position.

[0060] For example, the charging control module uses the Infineon TLE9879 chip as the main controller, which is connected to the central processing module. It receives the charging command from the central processing module, controls the vehicle wireless charging receiver (voltage level 220V, current 50A) to start wireless charging, and provides real-time feedback on the charging status.

[0061] Example 5 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, including: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0062] The millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST model radar, operating at a frequency of 60-64GHz, with a detection range of 0.1-10m, ranging accuracy of ±0.1cm, and angle accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0063] The visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0064] The inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β), and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0065] The central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0066] It also includes an automatic parking module. If the corrected relative position does not meet the charging start threshold, the central processing module sends a guidance command to the automatic parking module to automatically adjust the vehicle position until the corrected relative position meets the charging start threshold.

[0067] Example 6 This embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, including: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0068] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0069] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0070] For example, the inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β) and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0071] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0072] It also includes an automatic parking module. If the corrected relative position does not meet the charging start threshold, the central processing module sends a guidance command to the automatic parking module to automatically adjust the vehicle position until the corrected relative position meets the charging start threshold.

[0073] Example 7 The first aspect of this embodiment provides a multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, including: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

[0074] For example, the millimeter-wave radar module uses the Texas Instruments (TI) IWR6843BOOST radar, which operates at a frequency of 60-64 GHz, has a detection range of 0.1-10 m, a ranging accuracy of ±0.1 cm, and an angle measurement accuracy of ±0.1°. Its advantages include strong resistance to electromagnetic interference, the ability to penetrate dust and slight water accumulation, and immunity to light exposure. It scans metal positioning points and outputs the three-dimensional point cloud coordinates (Xr, Yr, Zr) of those points.

[0075] For example, the millimeter-wave radar module can also be a lidar (such as the Velodyne VLP-16).

[0076] For example, the visual acquisition module uses a Sony IMX477 high-definition camera (8 megapixels, 30fps) and an infrared filter switcher to output an image resolution of 3840×2160. It captures images of the image acquisition markers and outputs two-dimensional image coordinates (Xv, Yv).

[0077] For example, the inertial measurement module uses the MPU6050 module, which integrates a three-axis gyroscope and a three-axis accelerometer to collect the vehicle's tilt angle (α), pitch angle (β) and horizontal acceleration (a_x, a_y) in real time, and output the vehicle's attitude data (α, β).

[0078] For example, the central processing module uses the NVIDIA Jetson Xavier NX embedded AI chip, equipped with an ARM Cortex-A57 CPU and a Volta GPU, supports parallel computing, has a built-in Linux system, and comes pre-installed with drivers for the improved ICP registration algorithm and Kalman filter algorithm. It is responsible for receiving and processing multi-source data and outputting control commands. The chip has a computing power of 21 TOPS and ensures that the data processing latency is ≤10ms.

[0079] Furthermore, the charging pad marking unit includes metal positioning points and image acquisition marks, the image acquisition marks including a central QR code and a ring-shaped infrared dot matrix; The central QR code is located at the center of the ground wireless charging pad. When there is sufficient sunlight during the day, the visual acquisition module can identify the central QR code. The ring-shaped infrared dot array is arranged around the central QR code. In the dark or nighttime environment, the visual acquisition module can identify the ring-shaped infrared dot array. The central QR code, the ring-shaped infrared dot matrix, and the outer ring metal positioning points are all coaxially arranged with the coil center of the ground wireless charging pad.

[0080] Furthermore, the annular infrared dot array includes multiple uniformly distributed infrared LEDs.

[0081] Furthermore, the outer ring metal positioning points are stainless steel cylinders, and the outer ring metal positioning points are set at the four corners of the ground wireless charging plate.

[0082] Furthermore, the vehicle-mounted display module is used to display the corrected relative position and guidance instructions; A charging control module, connected to the central processing module, is used to control charging according to the charging instructions from the central processing module.

[0083] Furthermore, it also includes an automatic parking module. If the corrected relative position does not meet the charging start threshold, the central processing module sends a guidance command to the automatic parking module to automatically adjust the vehicle position until the corrected relative position meets the charging start threshold.

[0084] The second aspect of this embodiment provides a multimodal dynamic positioning and detection method for wireless charging of new energy vehicles, implemented using the system described above, combined with... Figure 3 As shown, it includes the following steps: The three-dimensional point cloud coordinates of the ground wireless charging pad are obtained through the millimeter-wave radar module, the two-dimensional image coordinates of the ground wireless charging pad are obtained through the vision acquisition module, and the vehicle body attitude data are obtained through the inertial measurement module. The initial relative position is obtained by fusing the three-dimensional point cloud coordinates with the two-dimensional image coordinates; Based on the vehicle posture data, dynamic error compensation is performed on the initial relative position to obtain the corrected relative position; Determine whether the corrected relative position meets the charging start threshold. If it does, output a charging command; otherwise, output a guidance command.

[0085] Furthermore, the three-dimensional point cloud coordinates are fused with the two-dimensional image coordinates using an improved ICP registration algorithm.

[0086] The improved ICP registration algorithm minimizes the matching error function using the least squares method and converges in ≤5 iterations, which is 60% faster than the traditional ICP algorithm and has a matching accuracy of ≤0.3cm.

[0087] By using an improved ICP registration algorithm for data matching, the large positioning error of single electromagnetic induction is overcome. This sub-centimeter positioning accuracy ensures optimal coupling between the coil of the ground wireless charging pad and the coil of the vehicle wireless charging pad, significantly improving the transmission efficiency of wireless charging.

[0088] Furthermore, the improved ICP registration algorithm fuses the three-dimensional point cloud coordinates with the two-dimensional image coordinates, including: Using two-dimensional image coordinates as the initial reference point, the matching range of three-dimensional point cloud coordinates is narrowed; The three-dimensional point cloud coordinates are assigned a first weight, and the two-dimensional image coordinates are assigned a second weight. The initial relative positions are obtained by weighted fusion.

[0089] Furthermore, the correction of relative position includes: X-axis coordinate, Y-axis coordinate, and angular deviation.

[0090] Further, the charging start-up thresholds are: X-axis deviation ≤ 0.1cm~0.5cm, Y-axis deviation ≤ 0.1cm~0.5cm, and angle deviation ≤ 0.1°~0.3°. The two values ​​of 0.1cm~0.5cm can be any value between 0.1cm, 0.5cm, or 0.1cm~0.5cm, and the value of 0.1°~0.3° can be any value between 0.1°, 0.3°, or 0.1°~0.3°. Preferably, the X-axis deviation is ≤ 0.5cm, the Y-axis deviation is ≤ 0.5cm, and the angle deviation is ≤ 0.3°.

[0091] Furthermore, the visual acquisition module switches to normal recognition mode during the day when there is sufficient light to identify the central QR code of the charging pad marking unit, and switches to infrared mode at night or in dim environments to identify the ring-shaped infrared dot matrix of the charging pad marking unit. Automatically switching between normal and infrared modes based on lighting conditions ensures that effective image acquisition marks can be identified in any lighting environment, further solidifying the system's high success rate of over 99.5% in complex environments.

[0092] Furthermore, a Kalman filter attitude compensation algorithm is used for dynamic error compensation.

[0093] The Kalman filter attitude compensation algorithm process is as follows to address the positioning deviation caused by vehicle body attitude shift: State equation: X(k)=A×X(k-1)+B×u(k)+w(k), where X(k) is the state vector at time k (positioning deviation ΔX, ΔY, angle deviation Δθ), A is the state transition matrix, B is the control matrix, u(k) is the attitude data (α,β) of the inertial measurement module, and w(k) is the process noise; Observation equation: Z(k) = H × X(k) + v(k), where Z(k) is the localization result after multimodal fusion, H is the observation matrix, and v(k) is the observation noise; Through prediction-update iteration, positioning deviations are corrected in real time, and the compensated positioning error is ≤0.5cm. Even if the vehicle body tilt angle reaches 5°, positioning accuracy can still be guaranteed.

[0094] For example: When the vehicle approaches the charging area (within 10m of the ground wireless charging pad), the user can activate the wireless charging reservation function through the vehicle's central control screen. After receiving the instruction, the central processing module will simultaneously activate the millimeter-wave radar module, vision acquisition module, and inertial measurement module, and each module will begin data acquisition.

[0095] The vehicle travels slowly at a speed of ≤5km / h. The millimeter-wave radar module continuously scans the ground. When it detects four outer ring metal positioning points, it outputs three-dimensional point cloud coordinates (Xr1-Xr4, Yr1-Yr4, Zr1-Zr4). At the same time, the vision acquisition module captures ground images and identifies the central QR code and the ring infrared dot array through image preprocessing (grayscale conversion, filtering, edge detection). It parses the absolute coordinate information in the QR code and outputs the two-dimensional image coordinates (Xv, Yv) of the infrared dot array center. The inertial measurement module collects vehicle attitude data (α, β, a_x, a_y) in real time, with a data output frequency of 100Hz.

[0096] After receiving three sets of data, the central processing module first performs data fusion using an improved ICP registration algorithm: based on the absolute coordinates of the center QR code parsing, the three-dimensional coordinates of the radar point cloud are projected onto a two-dimensional plane and matched with the visual image coordinates to calculate the initial relative position (ΔX0, ΔY0, Δθ0) between the coil of the vehicle wireless charging disconnection and the coil center of the ground charging plate. Based on the attitude data (α, β) from the inertial measurement module, dynamic error compensation is performed using the Kalman filter algorithm: the projection deviation ΔX1=h×tanα and ΔY1=h×tanβ caused by the vehicle body tilt are calculated (h is the installation height of the vehicle positioning terminal (which includes a central processing module, a millimeter-wave radar module, a vision acquisition module, and an inertial measurement module), which is preset to 1.2m), and the corrected relative position is obtained (ΔX=ΔX0+ΔX1, ΔY=ΔY0+ΔY1, Δθ=Δθ0). The central processing module determines whether the corrected positioning deviation meets the charging start conditions: X-axis deviation ≤ 0.5cm, Y-axis deviation ≤ 0.5cm, and angle deviation ≤ 0.3°. If the conditions are met (i.e., ΔX ≤ 0.5cm, ΔY ≤ 0.5cm, Δθ ≤ 0.3°), the central processing module sends a charging command to the charging control module. After the charging control module detects the coupling efficiency (≥ 90%) between the vehicle-mounted wireless charging terminal and the ground wireless charging pad, it starts wireless charging and displays that the positioning is successful and charging is in progress through the vehicle-mounted display module. If the conditions are not met: the central processing module outputs visual guidance instructions (such as "adjust 5cm to the left", "adjust 3cm forward", "adjust clockwise by 0.2°") on the vehicle display module according to the direction and magnitude of the deviation. The user adjusts the vehicle position according to the guidance, and at the same time, each module continues to collect data and repeats the above until the positioning deviation meets the conditions. During charging, each module maintains low power consumption and performs positioning calibration every 5 seconds. If the deviation exceeds the threshold (1cm) due to slight vehicle movement, charging is paused and a guidance command is output. Charging resumes after calibration is completed.

[0097] The technical effectiveness of this application was verified through actual testing. The test conditions and results are as follows: Test environment: normal lighting, no lighting at night, rainy weather, underground parking garage (no satellite signal), metal obstruction (toolbox placed around the charging pad); Test vehicle: The onboard wireless charging receiver is installed at a height of 1.2m; Test metrics: positioning accuracy, positioning success rate, dynamic response time, and attitude compensation effect; Test results: Positioning accuracy: The average positioning error is 0.32cm in all environments, the maximum error is 0.48cm, and both are ≤0.5cm; Location success rate: 100% in normal lighting, 99.8% at night without light, 99.6% in cloudy or rainy weather, 99.7% in underground parking garages, 99.5% with metal obstruction, with an average success rate of 99.7%. Dynamic response time: The average time from data acquisition to output of positioning results is 8.7ms, ≤10ms; Attitude compensation effect: When the vehicle tilt angle is 3°, the positioning error before compensation is 1.8cm and after compensation is 0.41cm; when the tilt angle is 5°, the positioning error before compensation is 3.2cm and after compensation is 0.47cm, showing a significant compensation effect.

[0098] In summary, the specific implementation method of this application, through clear hardware selection, optimized algorithm design and complete workflow, achieves sub-centimeter level, high reliability and strong environmental adaptability of wireless charging dynamic positioning detection for new energy vehicles, which fully meets the implementation requirements of those skilled in the art.

[0099] Example 8 This embodiment provides a multimodal dynamic positioning and detection method for wireless charging of new energy vehicles, combined with... Figure 3 As shown, it includes the following steps: The three-dimensional point cloud coordinates of the ground wireless charging pad are obtained through the millimeter-wave radar module, the two-dimensional image coordinates of the ground wireless charging pad are obtained through the vision acquisition module, and the vehicle body attitude data are obtained through the inertial measurement module. The initial relative position is obtained by fusing the three-dimensional point cloud coordinates with the two-dimensional image coordinates; Based on the vehicle posture data, dynamic error compensation is performed on the initial relative position to obtain the corrected relative position; Determine whether the corrected relative position meets the charging start threshold. If it does, output a charging command; otherwise, output a guidance command.

[0100] Furthermore, the three-dimensional point cloud coordinates are fused with the two-dimensional image coordinates using an improved ICP registration algorithm.

[0101] The improved ICP registration algorithm minimizes the matching error function using the least squares method and converges in ≤5 iterations, which is 60% faster than the traditional ICP algorithm and has a matching accuracy of ≤0.3cm.

[0102] By using an improved ICP registration algorithm for data matching, the large positioning error of single electromagnetic induction is overcome. This sub-centimeter positioning accuracy ensures optimal coupling between the coil of the ground wireless charging pad and the coil of the vehicle wireless charging pad, significantly improving the transmission efficiency of wireless charging.

[0103] Furthermore, the improved ICP registration algorithm fuses the three-dimensional point cloud coordinates with the two-dimensional image coordinates, including: Using two-dimensional image coordinates as the initial reference point, the matching range of three-dimensional point cloud coordinates is narrowed; The three-dimensional point cloud coordinates are assigned a first weight, and the two-dimensional image coordinates are assigned a second weight. The initial relative positions are obtained by weighted fusion.

[0104] Furthermore, the correction of relative position includes: X-axis coordinate, Y-axis coordinate, and angular deviation.

[0105] Furthermore, the charging start threshold is: X-axis deviation ≤ 0.5cm, Y-axis deviation ≤ 0.5cm, and angle deviation ≤ 0.3°.

[0106] Furthermore, the visual acquisition module switches to normal recognition mode during the day when there is sufficient light to identify the central QR code of the charging pad marking unit, and switches to infrared mode at night or in dim environments to identify the ring-shaped infrared dot matrix of the charging pad marking unit. Automatically switching between normal and infrared modes based on lighting conditions ensures that effective image acquisition marks can be identified in any lighting environment, further solidifying the system's high success rate of over 99.5% in complex environments.

[0107] Furthermore, a Kalman filter attitude compensation algorithm is used for dynamic error compensation.

[0108] The Kalman filter attitude compensation algorithm process is as follows to address the positioning deviation caused by vehicle body attitude shift: State equation: X(k)=A×X(k-1)+B×u(k)+w(k), where X(k) is the state vector at time k (positioning deviation ΔX, ΔY, angle deviation Δθ), A is the state transition matrix, B is the control matrix, u(k) is the attitude data (α,β) of the inertial measurement module, and w(k) is the process noise; Observation equation: Z(k) = H × X(k) + v(k), where Z(k) is the localization result after multimodal fusion, H is the observation matrix, and v(k) is the observation noise; Through prediction-update iteration, positioning deviations are corrected in real time, and the compensated positioning error is ≤0.5cm. Even if the vehicle body tilt angle reaches 5°, positioning accuracy can still be guaranteed.

[0109] For example: When the vehicle approaches the charging area (within 10m of the ground wireless charging pad), the user can activate the wireless charging reservation function through the vehicle's central control screen. After receiving the instruction, the central processing module will simultaneously activate the millimeter-wave radar module, vision acquisition module, and inertial measurement module, and each module will begin data acquisition.

[0110] The vehicle travels slowly at a speed of ≤5km / h. The millimeter-wave radar module continuously scans the ground. When it detects four outer ring metal positioning points, it outputs three-dimensional point cloud coordinates (Xr1-Xr4, Yr1-Yr4, Zr1-Zr4). At the same time, the vision acquisition module captures ground images and identifies the central QR code and the ring infrared dot array through image preprocessing (grayscale conversion, filtering, edge detection). It parses the absolute coordinate information in the QR code and outputs the two-dimensional image coordinates (Xv, Yv) of the infrared dot array center. The inertial measurement module collects vehicle attitude data (α, β, a_x, a_y) in real time, with a data output frequency of 100Hz.

[0111] After receiving three sets of data, the central processing module first performs data fusion using an improved ICP registration algorithm: based on the absolute coordinates of the center QR code parsing, the three-dimensional coordinates of the radar point cloud are projected onto a two-dimensional plane and matched with the visual image coordinates to calculate the initial relative position (ΔX0, ΔY0, Δθ0) between the coil of the vehicle wireless charging disconnection and the coil center of the ground charging plate. Based on the attitude data (α, β) from the inertial measurement module, dynamic error compensation is performed using the Kalman filter algorithm: the projection deviation ΔX1=h×tanα and ΔY1=h×tanβ caused by the vehicle body tilt are calculated (h is the installation height of the vehicle positioning terminal (which includes a central processing module, a millimeter-wave radar module, a vision acquisition module, and an inertial measurement module), which is preset to 1.2m), and the corrected relative position is obtained (ΔX=ΔX0+ΔX1, ΔY=ΔY0+ΔY1, Δθ=Δθ0). The central processing module determines whether the corrected positioning deviation meets the charging start conditions: X-axis deviation ≤ 0.5cm, Y-axis deviation ≤ 0.5cm, and angle deviation ≤ 0.3°. If the conditions are met (i.e., ΔX ≤ 0.5cm, ΔY ≤ 0.5cm, Δθ ≤ 0.3°), the central processing module sends a charging command to the charging control module. After the charging control module detects the coupling efficiency (≥ 90%) between the vehicle-mounted wireless charging terminal and the ground wireless charging pad, it starts wireless charging and displays that the positioning is successful and charging is in progress through the vehicle-mounted display module. If the conditions are not met: the central processing module outputs visual guidance instructions (such as "adjust 5cm to the left", "adjust 3cm forward", "adjust clockwise by 0.2°") on the vehicle display module according to the direction and magnitude of the deviation. The user adjusts the vehicle position according to the guidance, and at the same time, each module continues to collect data and repeats the above until the positioning deviation meets the conditions. During charging, each module maintains low power consumption and performs positioning calibration every 5 seconds. If the deviation exceeds the threshold (1cm) due to slight vehicle movement, charging is paused and a guidance command is output. Charging resumes after calibration is completed.

[0112] The technical effectiveness of this application was verified through actual testing. The test conditions and results are as follows: Test environment: normal lighting, no lighting at night, rainy weather, underground parking garage (no satellite signal), metal obstruction (toolbox placed around the charging pad); Test vehicle: The onboard wireless charging receiver is installed at a height of 1.2m; Test metrics: positioning accuracy, positioning success rate, dynamic response time, and attitude compensation effect; Test results: Positioning accuracy: The average positioning error is 0.32cm in all environments, the maximum error is 0.48cm, and both are ≤0.5cm; Location success rate: 100% in normal lighting, 99.8% at night without light, 99.6% in cloudy or rainy weather, 99.7% in underground parking garages, 99.5% with metal obstruction, with an average success rate of 99.7%. Dynamic response time: The average time from data acquisition to output of positioning results is 8.7ms, ≤10ms; Attitude compensation effect: When the vehicle tilt angle is 3°, the positioning error before compensation is 1.8cm and after compensation is 0.41cm; when the tilt angle is 5°, the positioning error before compensation is 3.2cm and after compensation is 0.47cm, showing a significant compensation effect.

[0113] In summary, the specific implementation method of this application, through clear hardware selection, optimized algorithm design and complete workflow, achieves sub-centimeter level, high reliability and strong environmental adaptability of wireless charging dynamic positioning detection for new energy vehicles, which fully meets the implementation requirements of those skilled in the art.

[0114] All terms used in this invention (including technical or scientific terms) have the same meaning as understood by one of ordinary skill in the art to which this invention pertains, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant art, and not as idealized or highly formalized, unless expressly defined herein.

[0115] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multimodal dynamic positioning and detection system for wireless charging of new energy vehicles, characterized in that, include: A charging pad marking unit is installed on a ground-based wireless charging pad; A millimeter-wave radar module is used to identify metal positioning points in the charging panel marking unit to obtain three-dimensional point cloud coordinates; The visual acquisition module is used to identify image acquisition marks in the charging pad marking unit to obtain two-dimensional image coordinates; Inertial measurement module, used to collect vehicle body attitude data; The central processing module is used to fuse three-dimensional point cloud coordinates and two-dimensional image coordinates to obtain an initial relative position, perform dynamic error compensation on the initial relative position based on the vehicle posture data to obtain a corrected relative position, and determine whether the corrected relative position meets the charging start threshold. If it does, a charging command is output; if it does not, a guidance command is output. Among them, the corrected relative position refers to the relative position between the coil of the vehicle-mounted wireless charging terminal and the center of the coil of the ground wireless charging pad.

2. The multimodal dynamic positioning and detection system for wireless charging of new energy vehicles according to claim 1, characterized in that, The charging pad marking unit includes metal positioning points and image acquisition marks, and the image acquisition marks include a central QR code and a ring-shaped infrared dot matrix. The central QR code is located at the center of the ground wireless charging pad. When there is sufficient sunlight during the day, the visual acquisition module can identify the central QR code. The ring-shaped infrared dot array is arranged around the central QR code. In the dark or nighttime environment, the visual acquisition module can identify the ring-shaped infrared dot array. The central QR code, the ring-shaped infrared dot matrix, and the outer ring metal positioning points are all coaxially arranged with the coil center of the ground wireless charging pad.

3. The multimodal dynamic positioning and detection system for wireless charging of new energy vehicles according to claim 2, characterized in that, The annular infrared dot array comprises multiple uniformly distributed infrared LEDs.

4. A multimodal dynamic positioning and detection system for wireless charging of new energy vehicles according to claim 2, characterized in that, The outer ring metal positioning points are stainless steel cylinders, and the outer ring metal positioning points are set at the four corners of the ground wireless charging pad.

5. The multimodal dynamic positioning and detection system for wireless charging of new energy vehicles according to claim 1, characterized in that, The vehicle-mounted display module is used to display the relative position correction and guidance instructions; A charging control module, connected to the central processing module, is used to control charging according to the charging instructions from the central processing module.

6. The multimodal dynamic positioning and detection system for wireless charging of new energy vehicles according to claim 1, characterized in that, It also includes an automatic parking module. If the corrected relative position does not meet the charging start threshold, the central processing module sends a guidance command to the automatic parking module to automatically adjust the vehicle position until the corrected relative position meets the charging start threshold.

7. A multimodal dynamic positioning and detection method for wireless charging of new energy vehicles, characterized in that, Includes the following steps: The three-dimensional point cloud coordinates of the ground wireless charging pad are obtained through the millimeter-wave radar module, the two-dimensional image coordinates of the ground wireless charging pad are obtained through the vision acquisition module, and the vehicle body attitude data are obtained through the inertial measurement module. The initial relative position is obtained by fusing the three-dimensional point cloud coordinates with the two-dimensional image coordinates; Based on the vehicle posture data, dynamic error compensation is performed on the initial relative position to obtain the corrected relative position; Determine whether the corrected relative position meets the charging start threshold. If it does, output a charging command; otherwise, output a guidance command.

8. The multimodal dynamic positioning and detection method for wireless charging of new energy vehicles according to claim 7, characterized in that, The three-dimensional point cloud coordinates are fused with the two-dimensional image coordinates using an improved ICP registration algorithm.

9. A multimodal dynamic positioning and detection method for wireless charging of new energy vehicles according to claim 8, characterized in that, The improved ICP registration algorithm fuses the three-dimensional point cloud coordinates with the two-dimensional image coordinates, including: Using two-dimensional image coordinates as the initial reference point, the matching range of three-dimensional point cloud coordinates is narrowed; The three-dimensional point cloud coordinates are assigned a first weight, and the two-dimensional image coordinates are assigned a second weight. The initial relative positions are obtained by weighted fusion.

10. A multimodal dynamic positioning and detection method for wireless charging of new energy vehicles according to claim 7, characterized in that, The corrected relative position includes: X-axis coordinate, Y-axis coordinate, and angular deviation.

11. A multimodal dynamic positioning and detection method for wireless charging of new energy vehicles according to claim 10, characterized in that, The charging start thresholds are: X-axis deviation ≤ 0.1cm~0.5cm, Y-axis deviation ≤ 0.1cm~0.5cm, and angle deviation ≤ 0.1°~0.3°.

12. The multimodal dynamic positioning and detection method for wireless charging of new energy vehicles according to claim 7, characterized in that, The visual acquisition module switches to normal recognition mode during the day when there is sufficient light to recognize the central QR code of the charging pad marking unit, and switches to infrared mode at night or in dark environments to recognize the ring infrared dot matrix of the charging pad marking unit.

Images