A robot for charging pile guiding docking

Abstract

The application discloses a kind of robots for charging pile guiding butt joint, it is related to electric vehicle charging technical field, including charging pile main body, the surface of charging pile main body is fixedly connected with mechanical arm main body, the front end of mechanical arm main body is fixedly connected with charging connector, the top of charging connector is fixedly connected with camera assembly at the front end of mechanical arm main body, and laser radar is fixedly connected at the top of camera assembly at the front end of mechanical arm main body;The robot for charging pile guiding butt joint, by the setting that the front end annular conductive induction ring of charging connector is electrically connected with control component, the capacitance variation between it and metal terminal is detected in real time when being close to charging port to judge coaxiality deviation and feedback to mechanical arm main body to carry out position correction, to construct independent capacitive detection correction channel outside vision and force sense, effectively avoid inserting skew card stagnation, improve butt joint precision and success rate, and need not vehicle end additional auxiliary device.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle charging technology, and more specifically to a robot for guiding and docking charging piles. Background Technology

[0002] With the rapid popularization of electric vehicles, automatic charging technology has become a key direction for solving the problem of charging convenience. Currently, automatic charging solutions based on vision guidance and LiDAR perception using robotic arms are developing rapidly. By identifying the location of the charging port using a camera and acquiring depth information using LiDAR, the robotic arm is guided to complete the automatic docking. This technology has been gradually applied to scenarios such as fixed parking spaces and unattended charging stations. In existing technologies, a strategy of fusing vision and LiDAR perception is often adopted to compensate for the perception limitations of single sensors under conditions such as changes in lighting and differences in target surface characteristics. The camera is responsible for providing the texture features and contour information of the charging port, while the LiDAR is responsible for providing three-dimensional spatial coordinates and depth data. The two are used together through a fusion algorithm to calculate the six-degree-of-freedom pose of the charging port relative to the end effector of the robotic arm, thereby achieving guided docking.

[0003] However, the aforementioned fusion perception scheme still has a drawback in engineering practice: when the vehicle charging port area is exposed to strong direct light, surface reflection, or is obscured by dirt, the local texture of the image captured by the camera degrades, causing the charging port contour features extracted by the image processing unit to deviate. Furthermore, the fusion positioning unit lacks a self-evaluation mechanism for the effectiveness of the current frame image data quality and still fuses the deviated features with the LiDAR point cloud data according to a fixed weight, resulting in the final output spatial pose deviating from the true value. When the robotic arm guides the charging connector closer to the charging port based on this deviated pose, an angular deviation occurs between the front end face of the charging connector and the end face of the charging port, causing the charging connector to jam or the terminal to scratch when inserted. In severe cases, it may even be impossible to insert the connector. Therefore, it needs to be improved. Summary of the Invention

[0004] The purpose of this invention is to provide a robot for guiding and docking charging piles, in order to solve the problems in the prior art where the image texture is degraded due to direct strong light, surface reflection, or dirt covering the charging port area. In such cases, the fusion positioning unit lacks a self-evaluation mechanism for the quality of image data and still fuses it with the LiDAR data according to a fixed weight, causing the spatial pose calculation to deviate from the true value. This leads to angular deviations between the charging connector and the charging port, jamming during insertion, or terminal scratching.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a robot for guiding docking of charging piles, comprising a charging pile body, a robotic arm body fixedly connected to the surface of the charging pile body, a charging connector fixedly connected to the front end of the robotic arm body, a camera assembly fixedly connected to the top of the charging connector at the front end of the robotic arm body, and a lidar fixedly connected to the top of the camera assembly at the front end of the robotic arm body.

[0006] The bottom end of the charging connector is fixedly connected to a wire, and the inner wall of the charging pile body is rotatably connected to a winding rod through a damping shaft. The wire is wound and connected to the surface of the winding rod, and the bottom end of the wire is electrically connected to a connecting wire, which is inserted through and plugged into the bottom end of the charging pile body.

[0007] The front end of the charging connector is embedded with an annular conductive induction coil, which is electrically connected to the signal input terminal of the control component. When the charging connector is close to the charging port, the coil is used to detect the change in capacitance between the annular conductive induction coil and the metal terminal inside the charging port to determine the coaxiality deviation between the charging connector and the charging port, and to feed back the deviation signal to the main body of the robotic arm for position correction.

[0008] Furthermore, a control component is fixedly connected to the inner wall of the charging pile body, a transmission component is electrically connected to the top of the control component, a computing chip is electrically connected to the top of the transmission component, and an input component is electrically connected to the top of the computing chip.

[0009] Furthermore, a display panel is fixedly connected to the surface of the charging pile body.

[0010] Furthermore, the control components include: an image processing unit, a point cloud processing unit, a fusion positioning unit, and a trajectory generation unit;

[0011] The image processing unit is used to receive and process image data collected by the camera component, and to identify the position and orientation of the vehicle charging port.

[0012] The point cloud processing unit is used to receive and process the point cloud data collected by the lidar, and generate three-dimensional spatial information of the vehicle charging port and its surrounding environment.

[0013] The fusion positioning unit is used to fuse the recognition results of the image processing unit and the point cloud processing unit to calculate the spatial pose of the charging port relative to the charging connector.

[0014] The trajectory generation unit is used to generate a drive control signal for the robotic arm body to move from the current position to the docking position based on the spatial pose, and outputs it to the execution drive end of the robotic arm body through the transmission component.

[0015] The environmental semantic density value is calculated using the following formula:

[0016]

[0017] In the formula, For a point in space The environmental semantic density value at that location; The perception space domain is defined by the intersection of the lidar field of view and the camera's field of view. For camera image plane The magnitude of the gradient vector at that point; This represents the radial distance value measured by the lidar at this point; This is the arithmetic mean of the distances between all point clouds in the current frame; The standard deviation of the distance values; This is an approximation of the Gaussian curvature of the point cloud at that point. It is a preset, extremely small positive number used to prevent numerical overflow; For parameter-based The non-extensive entropy function, ;

[0018] The fusion coefficient is calculated using the following formula:

[0019]

[0020] In the formula, The fusion coefficient; This represents the total number of valid feature points projected from the current sampling point cloud onto the image plane. For the first Index of feature points; For the first The image neighborhood grayscale distribution of each point forms a normal distribution; This is a normal distribution fitted to the local curvature distribution of the corresponding 3D point cloud; The Kullback-Leibler divergence; This represents the local variance of the point on the image gradient. This represents the local variance of the point in the curvature of the point cloud. It is the natural logarithm function.

[0021] Furthermore, the control component also includes: a status monitoring unit and an error compensation unit;

[0022] The status monitoring unit is used to monitor the motion status of each joint of the robotic arm body and the docking process of the charging connector in real time.

[0023] The error compensation unit is used to correct the drive control signal in real time based on the feedback data from the status monitoring unit when the charging connector approaches the charging port.

[0024] The dynamic compensation vector is calculated using the following formula:

[0025]

[0026] In the formula, for The three-dimensional position compensation vector output at any given time; This is the proportional gain coefficient matrix; This represents the six-degree-of-freedom pose error vector at the current moment; This is the integral gain coefficient matrix; This is the initial time. From the initial moment up to the current moment of Fractional integral operator of order, ; This is the differential gain coefficient matrix; for Fractional differential operators of order, ; Compliance control factor; This is the vector difference between the current joint angle and the target joint angle of the robotic arm; It is the hyperbolic tangent function; The fusion coefficient;

[0027] The fractional integral operator and the fractional differential operator are uniformly defined as follows:

[0028]

[0029] In the formula, For order, when When the value is negative, it corresponds to a fractional integral, when When the value is positive, it corresponds to the fractional derivative; greater than The smallest integer; It is the Gamma function; Let be the integral variable, representing the time from the initial moment. up to the current moment At any time between.

[0030] Furthermore, the input component includes: a vehicle information input port and a docking command input port;

[0031] The vehicle information input port is used to receive the vehicle model information and charging port model information of the vehicle to be charged;

[0032] The docking command input port is used to receive docking start signals triggered externally or control commands that require manual intervention.

[0033] Furthermore, the computing chip incorporates: a charging port recognition algorithm, a spatial positioning algorithm, and a path interpolation algorithm;

[0034] The charging port recognition algorithm is used to identify the outline features of the charging port from images captured by the camera component;

[0035] The spatial positioning algorithm is used to calculate the three-dimensional coordinates of the charging port by combining the point cloud data of the lidar.

[0036] The path interpolation algorithm is used to calculate the motion interpolation trajectory of each joint of the robotic arm body from the starting point to the target point.

[0037] Furthermore, the transmission component includes: a drive signal output terminal and a sensor signal input terminal;

[0038] The drive signal output terminal is electrically connected to the drive motors of each joint of the robotic arm body and is used to transmit pulse width modulation drive signals.

[0039] The sensor signal input terminal is electrically connected to the signal output terminals of the camera assembly and the lidar, and is used to receive sensing data.

[0040] Furthermore, the display panel is electrically connected to the control component and is used to display at least one of the following information: the motion status of the robotic arm body, the docking progress of the charging connector, the charging port identification result, and system fault alarm information.

[0041] Compared with existing technologies, the present invention provides a robot for guiding docking of charging piles. Through the electrical connection of a ring-shaped conductive induction coil embedded at the front end of the charging connector to the signal input terminal of the control component, the ring-shaped conductive induction coil can detect the capacitance change between the charging connector and the metal terminals inside the charging port in real time when the charging connector approaches the charging port. Based on this capacitance change, it accurately determines the coaxiality deviation between the charging connector and the charging port, and simultaneously feeds the deviation signal back to the robotic arm for position correction. This establishes an independent capacitive coaxiality detection and correction channel, in addition to visual and force-based guidance. This allows the charging connector to smoothly advance along the axis of the charging port at an extremely low insertion speed at the docking end, effectively avoiding problems such as misalignment, jamming, or terminal scraping caused by visual positioning residuals, robotic arm joint gaps, or end-effector jitter. This significantly improves the positioning accuracy and insertion success rate at the docking end. Furthermore, by utilizing the existing metal terminals of the charging port as the detection target, no additional auxiliary devices are required on the vehicle side, demonstrating good versatility and engineering practicality.

[0042] By using the fusion positioning unit in the control component to quantify and evaluate the feature richness of the perception area based on the environmental semantic density value calculation formula, and adaptively adjust the fusion weight of image data and point cloud data based on the fusion coefficient calculation formula, and by using the error compensation unit to calculate the proportional error term, fractional integral error accumulation term, fractional differential prediction term, and nonlinear compliance correction term in sequence according to the fractional dynamic compensation vector algorithm and superimpose them to generate the final dynamic compensation vector, a deep coupling between the visual sensor and the LiDAR at the spatial consistency level is achieved. This enables the automatic judgment of sensor data credibility and adjustment of the fusion strategy in complex environments such as sudden changes in illumination, reflection, and texture loss, ensuring the robustness of pose calculation in all-weather scenarios. On the other hand, by utilizing the long-term memory effect of fractional calculus on historical errors and its ability to predict future error trends, the nonlinear friction and start-stop jitter of the robotic arm during low-speed movement are effectively suppressed. The two work together to form a complete error suppression link from the environmental perception layer to the motion execution layer. Thus, without relying on high-precision absolute encoders or expensive force sensors, the compliance and positioning reproduction accuracy of the robotic arm at the docking end are significantly improved, and the system hardware cost and calibration complexity are reduced. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0044] Figure 1 This is one of the overall structural schematic diagrams provided in the embodiments of the present invention;

[0045] Figure 2 This is the second overall structural schematic diagram provided for an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of the control component structure provided in an embodiment of the present invention.

[0047] Explanation of reference numerals in the attached figures:

[0048] 1. Charging pile main body; 2. Robotic arm main body; 3. Charging connector; 4. Camera assembly; 5. LiDAR; 6. Wire; 7. Wrapping rod; 8. Connecting wire; 9. Control assembly; 10. Transmission assembly; 11. Computing chip; 12. Input assembly; 13. Display panel. Detailed Implementation

[0049] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0050] As attached Figure 1 To be continued Figure 3 As shown:

[0051] Example 1:

[0052] The present invention provides a robot for guiding docking of charging piles, including a charging pile body 1, a robotic arm body 2 fixedly connected to the surface of the charging pile body 1, a charging connector 3 fixedly connected to the front end of the robotic arm body 2, a camera assembly 4 fixedly connected to the front end of the robotic arm body 2 at the top of the charging connector 3, and a lidar 5 fixedly connected to the front end of the robotic arm body 2 at the top of the camera assembly 4.

[0053] The bottom end of the charging connector 3 is fixedly connected to a wire 6. The inner wall of the charging pile body 1 is rotatably connected to a winding rod 7 through a damping shaft. The wire 6 is wound and connected to the surface of the winding rod 7. The bottom end of the wire 6 is electrically connected to a connecting wire 8, and the connecting wire 8 is inserted through and inserted into the bottom end of the charging pile body 1.

[0054] The front end of the charging connector 3 is embedded with an annular conductive induction coil, which is electrically connected to the signal input terminal of the control component 9. When the charging connector 3 is close to the charging port, the coil is used to detect the change in capacitance between the annular conductive induction coil and the metal terminal inside the charging port to determine the coaxiality deviation between the charging connector 3 and the charging port, and to feed back the deviation signal to the robotic arm body 2 for position correction.

[0055] In use, the charging pile body 1 serves as the fixed base and power supply core of the entire robot. It houses the control components, transmission components, and computing chip, and provides a stable mounting surface for the robotic arm body 2, ensuring the rigidity and stability of the overall structure during movement. Simultaneously, it connects to the external power grid via a connecting cable 8 inserted at its bottom, providing a continuous and stable power input to the entire system. The robotic arm body 2 is fixedly connected to the surface of the charging pile body 1, serving as the core motion mechanism for performing docking actions. It has multiple rotary joints and degrees of freedom, enabling it to drive the charging connector 3 mounted at the front end to perform precise positioning and attitude adjustment in three-dimensional space under the drive of the control components, achieving a precise positioning and attitude adjustment from the initial position. The trajectory movement between the position and the vehicle charging port; the charging connector 3 is fixedly connected to the front end of the robotic arm body 2, serving as the terminal execution component that directly contacts and transmits power to the vehicle charging port. Its bottom end is fixedly connected to a wire 6 to introduce power, and it is inserted into the vehicle charging port under the drive of the robotic arm body 2, completing the charging circuit connection; the camera assembly 4 is fixedly connected to the front end of the robotic arm body 2 and located on top of the charging connector 3, used to collect real-time image data of the vehicle charging port and surrounding area during the docking process. By identifying the outline, edge features, and preset marker points of the charging port, it provides visual positioning information to the control components, achieving preliminary perception and tracking of the charging port position; the lidar 5 is fixedly mounted. A laser beam is fixedly connected to the front end of the robotic arm body 2 and located on top of the camera assembly 4. During docking, it emits a laser beam and receives reflected signals to acquire 3D point cloud data of the vehicle charging port and its surrounding environment. By measuring the distance and orientation of the charging port relative to the lidar, it provides precise spatial depth information to the control components, compensating for the camera assembly's perception limitations under insufficient lighting or texture loss conditions. One end of the wire 6 is fixedly connected to the bottom end of the charging connector 3, and the other end is wrapped around the surface of the winding rod 7 and electrically connected to the connecting wire 8. Serving as a flexible channel for power transmission, it can freely bend and stretch with the movement of the charging connector 3 during the movement of the robotic arm body 2, ensuring power is transmitted from the charging pile. The charging connector 3 is stably delivered inside the body 1. The winding rod 7 is rotatably connected to the inner wall of the charging pile body 1 through a damping shaft, and the wire 6 is wound and connected to the surface of the winding rod 7. When the robotic arm body 2 is in a non-working state or the charging connector 3 is retracted, the constant damping torque provided by the damping shaft will automatically rotate and wind up the wire 6, so that the wire 6 is orderly wound on the surface of the winding rod 7, avoiding the wire 6 from being messy, tangled, or excessively bent inside the charging pile body 1. At the same time, when the robotic arm body 2 is extended, the winding rod 7 is passively rotated and released the wire 6 as the wire 6 is pulled, ensuring that the extension length of the wire 6 matches the range of motion of the robotic arm body 2, and realizing automatic winding and unwinding management of the wire 6.The connecting wire 8 is inserted through and electrically connected to the bottom end of the charging pile body 1 and the bottom end of the wire 6. It serves as the power input interface between the charging pile body 1 and the external power grid, introducing external AC or DC power into the charging pile body 1. This power is then transmitted via the wire 6 to the charging connector 3, ultimately providing charging current to the vehicle being charged. These structures work together, using the dual-sensor fusion perception of the camera assembly 4 and the lidar 5 to accurately identify the position and orientation of the charging port. The multi-degree-of-freedom movement of the robotic arm body 2 enables autonomous guidance and precise docking of the charging connector 3. The automatic retraction and extension mechanism of the wire 6 and the winding rod 7 ensures orderly management of the power transmission lines. Together, they achieve a fully automated, high-precision, safe, and reliable guidance and docking operation between the charging pile and the vehicle's charging port.

[0056] Example 2:

[0057] This embodiment is basically the same as the previous embodiment, except that a control component 9 is fixedly connected to the inner wall of the charging pile body 1, a transmission component 10 is electrically connected to the top of the control component 9, a computing chip 11 is electrically connected to the top of the transmission component 10, an input component 12 is electrically connected to the top of the computing chip 11, and a display panel 13 is fixedly connected to the surface of the charging pile body 1. The control component 9 includes: an image processing unit, a point cloud processing unit, a fusion positioning unit, and a trajectory generation unit.

[0058] The image processing unit is used to receive and process image data acquired by the camera component 4, and to identify the position and orientation of the vehicle charging port;

[0059] The point cloud processing unit is used to receive and process the point cloud data collected by the lidar 5 to generate three-dimensional spatial information of the vehicle charging port and its surrounding environment.

[0060] The fusion positioning unit is used to fuse the recognition results of the image processing unit and the point cloud processing unit to calculate the spatial pose of the charging port relative to the charging connector 3.

[0061] The trajectory generation unit is used to generate drive control signals for the robotic arm body 2 to move from the current position to the docking position according to the spatial pose, and outputs them to the execution drive end of the robotic arm body 2 through the transmission component 10;

[0062] The environmental semantic density value is calculated using the following formula:

[0063]

[0064] In the formula, For a point in space The environmental semantic density value at that location; The perception space domain is defined by the intersection of the lidar field of view and the camera's field of view. For camera image plane The magnitude of the gradient vector at that point; This represents the radial distance value measured by the lidar at this point; This is the arithmetic mean of the distances between all point clouds in the current frame; The standard deviation of the distance values; This is an approximation of the Gaussian curvature of the point cloud at that point. It is a preset, extremely small positive number used to prevent numerical overflow; For parameter-based The non-extensive entropy function, ;

[0065] The fusion coefficient is calculated using the following formula:

[0066]

[0067] In the formula, The fusion coefficient; This represents the total number of valid feature points projected from the current sampling point cloud onto the image plane. For the first Index of feature points; For the first The image neighborhood grayscale distribution of each point forms a normal distribution; This is a normal distribution fitted to the local curvature distribution of the corresponding 3D point cloud; The Kullback-Leibler divergence; This represents the local variance of the point on the image gradient. This represents the local variance of the point in the curvature of the point cloud. It is the natural logarithm function.

[0068] Control component 9 also includes: a status monitoring unit and an error compensation unit;

[0069] The status monitoring unit is used to monitor the motion status of each joint of the robotic arm body 2 and the docking process of the charging connector 3 in real time.

[0070] The error compensation unit is used to correct the drive control signal in real time based on the feedback data from the status monitoring unit when the charging connector 3 approaches the charging port.

[0071] The dynamic compensation vector is calculated using the following formula:

[0072]

[0073] In the formula, for The three-dimensional position compensation vector output at any given time; This is the proportional gain coefficient matrix; This represents the six-degree-of-freedom pose error vector at the current moment; This is the integral gain coefficient matrix; This is the initial time. From the initial moment up to the current moment of Fractional integral operator of order, ; This is the differential gain coefficient matrix; for Fractional differential operators of order, ; Compliance control factor; This is the vector difference between the current joint angle and the target joint angle of the robotic arm; It is the hyperbolic tangent function; The fusion coefficient;

[0074] Fractional integral operators and fractional differential operators are uniformly defined as follows:

[0075]

[0076] In the formula, For order, when When the value is negative, it corresponds to a fractional integral, when When the value is positive, it corresponds to the fractional derivative; greater than The smallest integer; It is the Gamma function; Let be the integral variable, representing the time from the initial moment. up to the current moment At any time between.

[0077] Input component 12 includes: a vehicle information input port and a docking command input port;

[0078] The vehicle information input port is used to receive the vehicle model information and charging port model information of the vehicle to be charged;

[0079] The docking command input port is used to receive docking start signals triggered externally or control commands that require manual intervention.

[0080] The computing chip 11 has built-in: a charging port recognition algorithm, a spatial positioning algorithm, and a path interpolation algorithm;

[0081] The charging port recognition algorithm is used to identify the outline features of the charging port from the images captured by the camera component 4;

[0082] The spatial positioning algorithm is used to calculate the three-dimensional coordinates of the charging port by combining the point cloud data of the LiDAR 5.

[0083] The path interpolation algorithm is used to calculate the motion interpolation trajectory of each joint of the robotic arm body 2 from the starting point to the target point.

[0084] The transmission component 10 includes: a drive signal output terminal and a sensor signal input terminal;

[0085] The drive signal output terminal is electrically connected to the drive motors of each joint of the robotic arm body 2, and is used to transmit pulse width modulation drive signals;

[0086] The sensor signal input terminal is electrically connected to the signal output terminals of the camera assembly 4 and the lidar 5 to receive sensing data.

[0087] The display panel 13 is electrically connected to the control component 9 and is used to display at least one of the following information: the motion status of the robotic arm body 2, the docking progress of the charging connector 3, the charging port identification result, and the system fault alarm information.

[0088] In use, the control component 9 is fixedly connected to the inner wall of the charging pile body 1, serving as the core of the robot's decision-making and control. It integrates an image processing unit, a point cloud processing unit, a fusion positioning unit, a trajectory generation unit, a state monitoring unit, and an error compensation unit. It receives multi-source perception data from the camera component 4 and the lidar 5, processes, fuses, and solves the data using built-in algorithms, generating drive control signals for each joint of the robotic arm body 2. Simultaneously, it monitors the motion state of the robotic arm body 2 and the docking progress of the charging connector 3 in real time, dynamically correcting the drive signals based on feedback data at the docking end to ensure that the charging connector 3 can be inserted into the vehicle charging port with the desired pose accuracy. The bottom end of the transmission component 10 is connected to... The top of the control component 9 is electrically connected to the computing chip 11, serving as a data and command transmission bridge between the control component 9 and the computing chip 11. Internally, it has a drive signal output terminal and a sensor signal input terminal. The drive signal output terminal is electrically connected to the drive motors of each joint of the robotic arm body 2, used to transmit the pulse width modulation drive signal generated by the control component 9 to each joint motor to drive the robotic arm body 2 to execute corresponding motion commands. The sensor signal input terminal is electrically connected to the signal output terminals of the camera component 4 and the lidar 5, used to transmit the image data collected by the camera component 4 and the point cloud data collected by the lidar 5 to the control component 9 in real time for processing, thereby ensuring that the perceived data and control commands are transmitted within the system. High-speed, reliable bidirectional communication; the bottom of the computing chip 11 is electrically connected to the top of the transmission component 10, and the top is electrically connected to the input component 12. As the core hardware carrier for algorithm operation and data processing, it has a built-in charging port recognition algorithm, spatial positioning algorithm, and path interpolation algorithm. The charging port recognition algorithm is used to extract the contour features, edge information, and preset marker points of the charging port from the image acquired by the camera component 4. The spatial positioning algorithm is used to calculate the three-dimensional coordinates and attitude angles of the charging port in the robot coordinate system by combining the point cloud data of the lidar 5. The path interpolation algorithm is used to calculate the motion interpolation trajectory of each joint of the robotic arm body 2 from the current joint angle to the target joint angle based on the target pose output by the spatial positioning algorithm, thereby controlling the... The decision instructions of component 9 are converted into specific kinematic parameters; the bottom of input component 12 is electrically connected to the top of computing chip 11, serving as an interface module for human-computer interaction and external information input. It is equipped with a vehicle information input port and a docking instruction input port. The vehicle information input port is used to receive the vehicle model information and charging port model information of the vehicle to be charged, so that the control component 9 can adaptively adjust the docking strategy and trajectory planning parameters according to different vehicle models and charging port specifications. The docking instruction input port is used to receive externally triggered docking start signals or manual intervention control commands, so that operators can remotely start the docking program or send stop, rollback and other intervention commands in emergency situations, thereby enhancing the system's operational flexibility and safety controllability.The display panel 13 is fixedly connected to the surface of the charging pile body 1 and electrically connected to the control component 9. Serving as a visual feedback terminal for the system's operating status, it displays in real-time at least one of the following: the motion status of the robotic arm body 2, the docking progress of the charging connector 3, the charging port identification result, and system fault alarm information. This allows operators to intuitively understand the status of each stage of the current docking operation, the processing results of the sensing data, and whether the system is operating normally. In case of an anomaly, it promptly displays the fault type and location for troubleshooting. The aforementioned structures work together to receive external commands and vehicle information through the input component 12, execute core algorithm calculations through the computing chip 11, achieve high-speed bidirectional data and command flow through the transmission component 10, complete the fusion processing of multi-source sensing data and robotic arm motion control through the control component 9, and provide real-time visual feedback of status information through the display panel 13. This collectively ensures the integrity and reliability of the entire robot system's intelligent decision-making, precise control, stable communication, and human-machine interaction during the charging docking operation.

[0089] Application example:

[0090] With the rapid popularization of electric vehicles, traditional manual plug-and-play charging methods are no longer sufficient to meet the ever-increasing charging demand. This is especially true in scenarios such as unattended public charging stations at night, underground parking lots, and residential parking spaces, where car owners often need to get out of their vehicles to manually plug and unplug the charging gun. This is not only time-consuming and physically demanding, but also poses inconvenience and safety hazards in rainy or snowy weather, or when the charging port is positioned too high or too low. Furthermore, with the gradual implementation of automatic parking technology and intelligent driving systems, manual intervention is still required for charging after the vehicle has autonomously parked, creating a "automatic parking, manual charging" experience gap that severely restricts the level of intelligence in charging infrastructure. To address this, this invention provides a robot for guiding and docking charging piles. After the vehicle has come to a complete stop, the robot automatically completes the entire process from charging port identification, robotic arm path planning, precise insertion of the charging connector, to automatic unplugging and retraction of the charging gun after charging is complete, without human intervention. It is suitable for various passenger cars, commercial vehicles, and logistics vehicles with standard charging interfaces, and is particularly suitable for fixed parking spaces, closed parks, centralized charging stations, and automated valet parking scenarios, effectively improving charging convenience, safety, and operational efficiency.

[0091] When a vehicle to be charged is parked in the designated charging space according to the parking guidance system or driver operation, the vehicle parking brake is locked, the engine or drive motor is turned off, and the standby program inside the charging pile body 1 obtains the vehicle model information and charging port model information sent by the vehicle communication module or cloud platform through the vehicle information input port of the input component 12, so that the control component 9 completes the basic parameter configuration before docking.

[0092] After the docking start signal is received through the docking command input port of the input component 12, the calculation chip 11 calls the path interpolation algorithm to generate the initial motion trajectory based on the current zero position state of each joint of the robotic arm body 2. The control component 9 sends pulse width modulation signals to the drive motors of each joint of the robotic arm body 2 through the drive signal output terminal of the transmission component 10, so that the robotic arm body 2 gradually unfolds from the folded storage posture and is raised to the preset search height. During this process, the camera component 4 and the lidar 5 are powered on and start up synchronously and complete self-calibration. The camera component 4 acquires the global image of the rear or side of the vehicle, and the lidar 5 emits a laser beam to scan the surrounding environment. The perception data acquired by both are transmitted back to the control component 9 through the sensor signal input terminal of the transmission component 10. The image processing unit in the control component 9 performs filtering, noise reduction and edge extraction on the image data, and the point cloud processing unit performs filtering, segmentation and clustering on the point cloud data to identify the position information of the charging port in the image plane and three-dimensional space respectively.

[0093] The fusion positioning unit receives the output results from the image processing unit and the point cloud processing unit. First, it quantifies and evaluates the feature richness of the current perceived environment using the environmental semantic density value calculation formula. If the environmental semantic density value is lower than a preset threshold, the control component 9 automatically adjusts the exposure parameters of the camera component 4 and triggers the power enhancement mode of the lidar 5 to re-acquire data until the environmental semantic density value meets the positioning requirements. Then, the fusion positioning unit quantifies the consistency confidence of the camera component 4 and the lidar 5 at the current spatial point using the fusion coefficient calculation formula. When the fusion coefficient is higher than the high confidence threshold, it is determined to be a high confidence scenario. The fusion positioning unit jointly solves the two-dimensional pixel coordinates output by the image processing unit and the three-dimensional spatial coordinates output by the point cloud processing unit to output the accurate three-dimensional coordinates and attitude angle of the charging port in the robot's base coordinate system. When the fusion coefficient is lower than the low confidence threshold, it is determined to be a sensor conflict scenario. The fusion positioning unit automatically reduces the fusion weight of the image data and enhances the ranging data weight of the lidar 5, and re-iterates the calculation until reliable six-degree-of-freedom pose information is output.

[0094] The trajectory generation unit, based on the six-DOF pose information of the charging port output by the fusion positioning unit and combined with the real-time angle encoding values ​​of each joint of the current robotic arm body 2, generates a set of joint spatial motion trajectories from the current position to the target position through inverse kinematics calculation. This motion trajectory is transmitted in real-time to the drive motors of each joint of the robotic arm body 2 via the drive signal output terminal of the transmission component 10. The robotic arm body 2 drives each rotary joint sequentially according to the trajectory planning order, causing the charging connector 3 installed at its front end to gradually approach the vehicle charging port from the search position. During this approach process, the state monitoring unit reads the angle feedback values ​​of each joint encoder and the force data at the end of the charging connector 3 in real-time with a millisecond-level sampling period, preventing errors. The error compensation unit corrects the drive control signal in real time based on the feedback data from the state monitoring unit. When the charging connector 3 moves to a position a certain distance from the surface of the charging port, the camera component 4 acquires a local high-resolution image of the charging port again, performs fine registration with the previously stored template image, and outputs a micron-level relative pose deviation. The error compensation unit then starts the fractional-order dynamic compensation vector algorithm, calculates the proportional error term, fractional-order integral error accumulation term, fractional-order differential prediction term, and nonlinear compliance correction term based on the fusion coefficient in sequence, and generates the final dynamic compensation vector by superimposing the four calculation results onto the current motion trajectory, so that the axis of the charging connector 3 gradually approaches the axis of the charging port.

[0095] When the annular conductive induction coil embedded at the front end of the charging connector 3 enters the charging port and the capacitance change between it and the metal terminal inside the charging port reaches a predetermined range, the annular conductive induction coil sends a coaxiality deviation signal to the control component 9. The error compensation unit generates a final set of fine-tuning drive signals based on the deviation signal, so that the charging connector 3 is smoothly pushed along the axis of the charging port at an extremely low insertion speed until it is fully inserted. During this process, the status monitoring unit continuously monitors the insertion depth and contact force of the charging connector 3. When the contact force reaches the preset locking threshold, the docking is determined to be completed. The control component 9 sends a locking command to the locking mechanism inside the charging connector 3 through the transmission component 10, and at the same time sends a power-on permission signal to the charging control circuit inside the charging pile body 1. The power from the external power grid is introduced into the charging pile body 1 through the connecting wire 8, and then transmitted to the charging connector 3 through the wire 6, ultimately providing charging current for the vehicle's power battery.

[0096] During charging, the status monitoring unit continuously monitors the static holding torque of the drive motors of each joint, as well as the temperature rise and contact resistance of the charging connector 3. When any parameter exceeds the preset safety range, the control component 9 immediately cuts off the power output and displays the corresponding fault alarm information on the display panel 13. At the same time, it sends a fault notification to the remote management platform through the input component 12. After charging is completed, the control component 9 first disconnects the power output and then sends an unlocking command to release the locking mechanism inside the charging connector 3. The trajectory generation unit generates a retraction path opposite to the insertion trajectory based on the current position of each joint of the robotic arm body 2. The robotic arm body 2 drives the charging connector 3 to be smoothly pulled out of the charging port along this path and retracted to the folded storage posture. Under the constant damping torque provided by the damping shaft, the winding rod 7 automatically rotates as the wire 6 is retracted, orderly winding the wire 6 onto the surface of the winding rod 7, restoring the wire 6 to its initial neat winding state. At the same time, the control component 9 controls the camera component 4 and the lidar 5 to enter standby sleep mode through the transmission component 10 and displays the status information "charging complete, equipment reset" on the display panel 13, waiting for the trigger signal of the next docking task.

[0097] Working principle: When a vehicle to be charged parks in the designated charging space, the control component 9 inside the charging pile body 1 receives the vehicle model information and charging port model parameters transmitted by the vehicle communication module through the input component 12. Simultaneously, an external trigger signal activates the entire system to enter the docking preparation state via the docking command input port. At this time, the camera component 4 and the lidar 5 start synchronously. The camera component 4 acquires images of the rear or side area of ​​the vehicle and transmits the image data back to the control component 9 through the transmission component 10. The lidar 5 emits a laser beam towards the target area and receives the reflected signal to obtain three-dimensional point cloud data, which is then transmitted back to the control component 9 through the transmission component 10. The image processing unit within the control component 9 filters and denoises the image data. Edge extraction and contour recognition are used to extract the two-dimensional pixel coordinates and pose information of the charging port. The point cloud processing unit performs filtering, segmentation, clustering analysis, and surface fitting on the point cloud data to extract the three-dimensional coordinates and depth information of the charging port in space. After receiving the output results from the image processing unit and the point cloud processing unit, the fusion positioning unit first quantifies and evaluates the feature richness of the current perception area using the environmental semantic density value calculation formula to determine whether the current frame data meets the positioning requirements. If the environmental semantic density value is lower than the preset threshold, the control component 9 automatically adjusts the exposure parameters of the camera component 4 and triggers the power enhancement mode of the lidar 5 to re-acquire data until the requirements are met. Subsequently, the fusion positioning unit quantifies the feature richness of the camera component 4 using the fusion coefficient calculation formula. Based on the consistency confidence of LiDAR 5 at the current spatial point, the fusion weights of image data and point cloud data are adaptively adjusted according to the fusion coefficient. The precise six-degree-of-freedom pose information of the charging port relative to the charging connector 3 is output through joint calculation. The trajectory generation unit generates a set of joint spatial motion trajectories from the current position to the target position by combining the pose information with the real-time angle encoding values ​​of each joint of the current robotic arm body 2 through inverse kinematics calculation. The motion trajectory is sent in real time to the drive motors of each joint of the robotic arm body 2 via the drive signal output terminal of the transmission component 10. Each joint of the robotic arm body 2 rotates and moves in sequence according to the trajectory planning, driving the charging connector 3 installed at its front end to gradually unfold from the folded storage posture and approach the vehicle. During the approach to the charging port, the state monitoring unit reads the angle feedback values ​​of each joint encoder and the force data at the end of the charging connector 3 in real time with a millisecond sampling period and transmits them to the error compensation unit. The error compensation unit corrects the current motion trajectory in real time based on the feedback data from the state monitoring unit. When the charging connector 3 moves to a certain distance from the surface of the charging port, the camera component 4 again acquires a local high-resolution image of the charging port and performs fine registration with the stored template image, and outputs the micron-level relative pose deviation to the error compensation unit. The error compensation unit then starts the fractional-order dynamic compensation vector algorithm to calculate the proportional error term, the fractional-order integral error accumulation term, the fractional-order differential prediction term, and the nonlinear compliance correction term based on the fusion coefficient in sequence.The four calculation results are superimposed to generate the final dynamic compensation vector, which is then superimposed onto the current motion trajectory. This causes the axis of the charging connector 3 to gradually approach the axis of the charging port. When the annular conductive induction coil embedded at the front end of the charging connector 3 enters the charging port and the capacitance change between it and the metal terminals inside the charging port reaches a predetermined range, the annular conductive induction coil feeds back a coaxiality deviation signal to the control component 9. The error compensation unit generates a final set of fine-tuning drive signals based on this deviation signal, causing the charging connector 3 to smoothly advance along the axis of the charging port until it is fully inserted. At this point, the charging connector 3 and the charging port are physically connected. The control component 9 then sends a signal to the charging connector... The internal locking mechanism sends a locking command and connects the wire 6, which is fixedly connected to the bottom of the charging connector 3, to the connecting wire 8. Power from the external power grid is sequentially introduced into the charging pile body 1 via the connecting wire 8, transmitted through the wire 6 to the charging connector 3, and finally provides charging current to the vehicle. Simultaneously, the winding rod 7 is rotatably connected to the inner wall of the charging pile body 1 via a damping shaft, and the wire 6 is wound around the surface of the winding rod 7. During the extension of the robotic arm body 2, the winding rod 7 passively rotates as the wire 6 is pulled to release the wire 6, thus ensuring that the extension length of the wire 6 matches the range of motion of the robotic arm body 2. During the charging process, a status monitoring system is activated. The control unit continuously monitors the static holding torque of the drive motors of each joint and the temperature rise and contact resistance of the charging connector 3. When any parameter exceeds the preset safety range, the control component 9 immediately cuts off the power output and displays the corresponding fault alarm information on the display panel 13. After charging is completed, the control component 9 first disconnects the power output and then sends an unlocking command to release the locking mechanism inside the charging connector 3. The trajectory generation unit generates a return path based on the current position of each joint of the robotic arm body 2 and drives each joint of the robotic arm body 2 to perform reverse movement through the transmission component 10, so that the charging connector 3 is smoothly pulled out of the charging port. The winding rod 7 rotates in a damped manner. Under the constant damping torque provided by the shaft, the wire 6 automatically rotates and rewinds as it is retracted, ensuring it is orderly coiled around the surface of the winding rod 7 to prevent it from becoming messy or tangled inside the charging pile body 1. Simultaneously, the control component 9, through the transmission component 10, controls the robotic arm body 2 to fold and reset, and puts the camera component 4 and lidar 5 into standby mode. Finally, the display panel 13 shows the charging completion and equipment reset status information, awaiting the next docking task. This completes the entire process from charging port identification, robotic arm-guided docking, power transmission to automatic gun removal and retraction after charging.

[0098] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A robot for guiding and docking charging piles, comprising a charging pile body (1), characterized in that, The surface of the charging pile body (1) is fixedly connected to the robotic arm body (2). The front end of the robotic arm body (2) is fixedly connected to the charging connector (3). The front end of the robotic arm body (2) is fixedly connected to the top of the charging connector (3). The front end of the robotic arm body (2) is fixedly connected to the top of the charging connector (3). The front end of the robotic arm body (2) is fixedly connected to the top of the camera assembly (4). The bottom end of the charging connector (3) is fixedly connected to a wire (6), and the inner wall of the charging pile body (1) is rotatably connected to a winding rod (7) through a damping shaft. The wire (6) is wound and connected to the surface of the winding rod (7). The bottom end of the wire (6) is electrically connected to a connecting wire (8), and the connecting wire (8) is inserted through and inserted into the bottom end of the charging pile body (1). The front end of the charging connector (3) is embedded with an annular conductive induction coil. The annular conductive induction coil is electrically connected to the signal input terminal of the control component (9). When the charging connector (3) is close to the charging port, the change in capacitance between the annular conductive induction coil and the metal terminal inside the charging port is detected to determine the coaxiality deviation between the charging connector (3) and the charging port, and the deviation signal is fed back to the main body of the robotic arm (2) for position correction.

2. The robot for guiding docking of charging piles according to claim 1, characterized in that, A control component (9) is fixedly connected to the inner wall of the charging pile body (1). A transmission component (10) is electrically connected to the top of the control component (9). A computing chip (11) is electrically connected to the top of the transmission component (10). An input component (12) is electrically connected to the top of the computing chip (11).

3. The robot for guiding docking of charging piles according to claim 1, characterized in that, A display panel (13) is fixedly connected to the surface of the charging pile body (1).

4. A robot for guiding docking of charging piles according to claim 2, characterized in that, The control component (9) includes: an image processing unit, a point cloud processing unit, a fusion positioning unit, and a trajectory generation unit; The image processing unit is used to receive and process the image data collected by the camera component (4) and identify the position and orientation of the vehicle charging port; The point cloud processing unit is used to receive and process the point cloud data collected by the lidar (5) to generate three-dimensional spatial information of the vehicle charging port and the surrounding environment. The fusion positioning unit is used to fuse the recognition results of the image processing unit and the point cloud processing unit to calculate the spatial pose of the charging port relative to the charging connector (3). The trajectory generation unit is used to generate a drive control signal for the robotic arm body (2) to move from the current position to the docking position according to the spatial pose, and outputs it to the execution drive end of the robotic arm body (2) through the transmission component (10); The environmental semantic density value is calculated using the following formula: In the formula, For a point in space The environmental semantic density value at that location; The perception space domain is defined by the intersection of the lidar field of view and the camera's field of view. For camera image plane The magnitude of the gradient vector at that point; This represents the radial distance value measured by the lidar at this point; This is the arithmetic mean of the distances between all point clouds in the current frame; The standard deviation of the distance values; This is an approximation of the Gaussian curvature of the point cloud at that point. It is a preset, extremely small positive number used to prevent numerical overflow; For parameter-based The non-extensive entropy function, ; The fusion coefficient is calculated using the following formula: In the formula, The fusion coefficient; This represents the total number of valid feature points projected from the current sampling point cloud onto the image plane. For the first Index of feature points; For the first The image neighborhood grayscale distribution of each point forms a normal distribution; This is a normal distribution fitted to the local curvature distribution of the corresponding 3D point cloud; The Kullback-Leibler divergence; This represents the local variance of the point on the image gradient. This represents the local variance of the point in the curvature of the point cloud. It is the natural logarithm function.

5. A robot for guiding docking of charging piles according to claim 4, characterized in that, The control component (9) further includes: a status monitoring unit and an error compensation unit; The status monitoring unit is used to monitor the motion status of each joint of the main body (2) of the robotic arm and the docking process of the charging connector (3) in real time. The error compensation unit is used to correct the drive control signal in real time based on the feedback data from the status monitoring unit when the charging connector (3) approaches the charging port; The dynamic compensation vector is calculated using the following formula: In the formula, for The three-dimensional position compensation vector output at any given time; This is the proportional gain coefficient matrix; This represents the six-degree-of-freedom pose error vector at the current moment; This is the integral gain coefficient matrix; This is the initial time. From the initial moment up to the current moment of Fractional integral operator of order order, ; This is the differential gain coefficient matrix; for Fractional differential operators of order, ; Compliance control factor; This is the vector difference between the current joint angle and the target joint angle of the robotic arm; It is the hyperbolic tangent function; The fusion coefficient; The fractional integral operator and the fractional differential operator are uniformly defined as follows: In the formula, For order, when When the value is negative, it corresponds to a fractional integral, when When the value is positive, it corresponds to the fractional derivative; greater than The smallest integer; It is the Gamma function; Let be the integral variable, representing the time from the initial moment. up to the current moment At any time between.

6. A robot for guiding docking of charging piles according to claim 2, characterized in that, The input component (12) includes: a vehicle information input port and a docking command input port; The vehicle information input port is used to receive the vehicle model information and charging port model information of the vehicle to be charged; The docking command input port is used to receive docking start signals triggered externally or control commands that require manual intervention.

7. A robot for guiding docking of charging piles according to claim 2, characterized in that, The computing chip (11) has built-in: a charging port recognition algorithm, a spatial positioning algorithm, and a path interpolation algorithm; The charging port recognition algorithm is used to identify the outline features of the charging port from the image acquired by the camera component (4); The spatial positioning algorithm is used to calculate the three-dimensional coordinates of the charging port by combining the point cloud data of the lidar (5); The path interpolation algorithm is used to calculate the motion interpolation trajectory of each joint of the robotic arm body (2) from the starting point to the target point.

8. A robot for guiding docking of charging piles according to claim 2, characterized in that, The transmission component (10) includes: a drive signal output terminal and a sensor signal input terminal; The drive signal output terminal is electrically connected to the drive motors of each joint of the robotic arm body (2) for transmitting pulse width modulation drive signals. The sensor signal input terminal is electrically connected to the signal output terminal of the camera assembly (4) and the lidar (5) to receive sensing data.

9. A robot for guiding docking of charging piles according to claim 3, characterized in that, The display panel (13) is electrically connected to the control component (9) and is used to display at least one of the following information: the motion status of the robotic arm body (2), the docking progress of the charging connector (3), the charging port identification result, and the system fault alarm information.

Images