Robot-based ship charging method and device

Abstract

The invention relates to a ship charging method and device based on a robot, and the method comprises the steps: obtaining sensor data and image data of a ship charging interface in a driving process when the robot drives to a preset range of a ship; then according to the image data and the sensor data, the operation state of the robot and external environment information are determined, and when the operation state of the robot changes and / or an obstacle is detected in the external environment information, adjustment data of operation parameters of the robot are obtained according to the operation state of the robot and the external environment information; the problem of positioning deviation caused by ship shaking and environmental interference is solved; the operation parameters of the robot are adjusted according to the adjustment data to obtain an adjustment result, and the robot is controlled to charge the ship through the charging interface according to the adjustment result, so that accurate docking of the robot to the ship charging interface in a complex environment is realized; the method has the advantages of improving the automation degree, adapting to complex environment changes and improving safety and efficiency.

Description

Technical Field

[0001] The present invention relates to the field of robotics technology, and in particular to a robot-based ship charging method and device. Background Art

[0002] With the booming global trade and shipping industries, the number of ships continues to increase, and the trend toward electrification and intelligent ships is becoming increasingly pronounced. Electric ships are increasingly widely used, from small electric vessels operating in inland waterways to large electric cargo and passenger ships operating along the coast. Efficient and safe charging solutions are essential. Traditional shore-based charging methods often require manual operation, requiring crew members to drag and connect charging cables to the ship's charging ports. This method is not only time-consuming and labor-intensive, but can also pose safety risks to operators in inclement weather or when the ship is poorly positioned. Robotic technology has made significant progress in both the industrial and service sectors. Mobile robots are capable of autonomous navigation and obstacle avoidance in complex environments. Navigation technologies include lidar positioning and visual SLAM (simultaneous localization and mapping). These technologies enable robots to accurately determine their position in unknown environments and plan appropriate paths.

[0003] In existing technologies, the ship charging environment is complex and changeable, including factors such as ship shaking, marine environmental corrosion, different lighting and weather conditions, which poses a huge challenge to the robot's precise positioning.

[0004] Therefore, there is an urgent need to propose a robot-based ship charging method and device to solve the technical problems in the existing technology that the ship charging environment is complex and changeable, including ship shaking, marine environment corrosion, different lighting and weather conditions, etc., which pose a huge challenge to the precise positioning of the robot. Summary of the Invention

[0005] In view of this, it is necessary to provide a robot-based ship charging method and device to solve the technical problems existing in the existing technology that the ship charging environment is complex and changeable, including ship shaking, marine environment corrosion, different lighting and weather conditions, etc., which pose great challenges to the precise positioning of the robot.

[0006] In order to solve the above problems, in a first aspect, the present invention provides a robot-based ship charging method, comprising: When the robot travels within the preset range of the ship, it obtains sensor data during the travel process and image data of the ship's charging interface; Determining the robot's operating state and external environment information based on the image data and the sensor data; When the operating state of the robot changes and / or an obstacle is detected in the external environment information, adjustment data of the operating parameters of the robot are obtained according to the operating state of the robot and the external environment information; The operating parameters of the robot are adjusted according to the adjustment data to obtain an adjustment result, and the robot is controlled to charge the ship through the charging interface according to the adjustment result.

[0007] In one possible implementation, a ship simulation model is constructed; the ship simulation model includes three-dimensional model feature points of the charging interface; Extracting features from the image data and matching them with the feature points of the three-dimensional model to obtain charging port posture information; Path planning is performed according to the charging port posture information to obtain an optimal path, and the robot is controlled to travel toward the ship according to the optimal path.

[0008] In one possible implementation, extracting features from the image data and then matching them with feature points of the three-dimensional model to obtain charging port posture information includes: Build a dynamic model based on the robot arm and mobile platform; After the dynamic model is trained, the image data is input into the dynamic model for feature extraction to obtain two-dimensional image feature points; The two-dimensional image feature points are matched with the three-dimensional model feature points to obtain charging port posture information.

[0009] In one possible implementation, matching the two-dimensional image feature points with the three-dimensional model feature points to obtain charging port posture information includes: Calculating the coordinates of the two-dimensional image feature points and the three-dimensional model feature points according to direct linear transformation to obtain linear equations of rotation parameters and translation parameters; Solving the linear equation according to singular value decomposition to obtain a least squares solution; Optimizing the least squares solution according to a reprojection error minimization optimization algorithm to obtain a reprojection error sum; The reprojection error and are iteratively updated to obtain charging port posture information.

[0010] In one possible implementation, performing path planning based on the charging port posture information to obtain an optimal path includes: Constructing a grid map according to the environment of the robot, wherein the grid map includes an obstacle status of each grid; Determine a starting point and an end point in the grid map according to the current position of the robot and the posture information of the charging port; The starting point and the end point in the grid map are searched according to a preset search algorithm and the obstacle status to obtain an optimal path.

[0011] In a possible implementation, searching the starting point and the end point in the grid map according to a preset search algorithm and the obstacle status to obtain an optimal path includes: Putting the starting point into an open list, and calculating the starting point using the heuristic function of the preset search algorithm to obtain a heuristic function value; Determine the node with the smallest heuristic function value from the open list as the current node and put it into the closed list; Traversing the adjacent nodes of the current node, and skipping the adjacent node when there is an obstacle in the adjacent node or the adjacent node is in the closed list; When the obstacle does not exist or the adjacent node is not in the closed list, calculating the heuristic function value of the adjacent node by using the heuristic function, and placing the adjacent node into the open list; The node with the smallest heuristic function value is confirmed again from the open list and searched until the open list is empty or the end point is searched, and the optimal path is generated by backtracking the parent node of each node in the closed list.

[0012] In a possible implementation, after controlling the robot to charge the ship through the charging interface according to the adjustment result, the method further includes: The operating state of the robot and the external environment information are continuously monitored. When the operating state of the robot changes and / or an obstacle is detected in the external environment information, the movement of the robot is adjusted according to the operating state of the robot and the external environment information.

[0013] In a possible implementation, after controlling the robot to charge the ship through the charging interface according to the adjustment result, the method further includes: The electrical parameters of the ship are continuously monitored, and when the electrical parameters meet the preset charging completion conditions, the robot is controlled to separate from the charging interface according to the external environment information.

[0014] In a possible implementation, before the robot travels within a preset range of the ship, the method further includes: Obtaining a current position of the ship and navigating the robot according to the current position; Determine whether the robot has navigated to within a preset range of the ship.

[0015] In a second aspect, the present invention further provides a robot-based ship charging device, comprising: A data acquisition module is used to acquire sensor data and image data of the ship's charging interface during the driving process when the robot travels within a preset range of the ship; An information confirmation module, configured to determine the robot's operating status and external environment information based on the image data and the sensor data; a data adjustment module, configured to obtain adjustment data for the operating parameters of the robot according to the operating state of the robot and the external environment information when the operating state of the robot changes and / or an obstacle is detected in the external environment information; The ship charging module is used to adjust the operating parameters of the robot according to the adjustment data to obtain an adjustment result, and control the robot to charge the ship through the charging interface according to the adjustment result.

[0016] The beneficial effects of the present invention are as follows: when the robot travels within a preset range of the ship, sensor data and image data of the ship's charging interface during the travel process are obtained; then the robot's operating status and external environment information can be determined based on the image data and sensor data; when the robot's operating status changes and / or an obstacle is detected in the external environment information, adjustment data of the robot's operating parameters can be obtained based on the robot's operating status and external environment information, thereby solving the positioning deviation problem caused by ship shaking and environmental interference; further, the robot's operating parameters can be adjusted according to the adjustment data to obtain adjustment results, and the robot can be controlled to charge the ship through the charging interface according to the adjustment results, thereby realizing precise docking of the robot with the ship's charging interface in complex environments, which has the advantages of improving the degree of automation, adapting to complex environmental changes, and improving safety and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of an embodiment of the robot-based ship charging method provided by the present invention; Figure 2 A schematic diagram of a flow chart of an embodiment of confirming the position information of a charging port provided by the present invention; Figure 3 For the present invention Figure 2 A schematic flow chart of an embodiment of step S203; Figure 4 A schematic diagram of a flow chart of an embodiment of determining an optimal path provided by the present invention; Figure 5 For the present invention Figure 4 A schematic flow chart of an embodiment of step S403; Figure 6 This is a schematic structural diagram of an embodiment of the robot-based ship charging device provided by the present invention. DETAILED DESCRIPTION

[0018] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0019] like Figure 1 As shown, a specific embodiment of the present invention discloses a robot-based ship charging method, comprising: S101. When the robot travels within a preset range of the ship, sensor data during the travel and image data of the ship's charging interface are acquired.

[0020] The robot's main body is constructed of a high-strength, lightweight alloy material to ensure stability and flexibility during movement. The robot is equipped with an adjustable robotic arm, whose joints are driven by high-precision motors, enabling multi-degree-of-freedom movement and accurately delivering the charging plug to the ship's charging port. The end of the robotic arm is designed with an adaptive clamping device that automatically adjusts the clamping force and angle based on the shape and size of the charging port, ensuring a tight connection between the charging plug and the port. The robot's base is equipped with a special mobility mechanism, such as Mecanum wheels or a tracked structure. Mecanum wheels enable full-range movement, facilitating flexible steering and positioning within the confined space of a ship's deck; tracked structures are more suitable for handling uneven or sloped ship surfaces, ensuring the robot maintains stable movement despite the ship's swaying.

[0021] Multiple high-resolution cameras are installed at key locations on the robot, forming a multi-view visual system. These cameras are located on the robot's head, arms, and sides, ensuring that they can capture image data of the ship's charging port from various angles. Sensor data can include data from the robot's visual sensors, attitude sensors, and distance sensors. When the robot is within a preset range of the ship, these cameras capture image data of the ship's charging port from various angles and receive sensor data in real time.

[0022] S102: Determine the robot's operating status and external environment information based on the image data and sensor data.

[0023] The robot's operating status and external environment information are integrated with sensor data and visual positioning results. For example, multi-source information is integrated through Kalman filtering or particle filtering algorithms to improve the reliability of state estimation. As the robot moves along the path, it continuously integrates real-time data from various sensors, including lidar and IMU, to obtain the robot's operating status and external environment information. Based on this external environment information and the robot's operating status, the robot updates the environmental map and detects dynamic obstacles.

[0024] S103. When the robot's operating state changes and / or an obstacle is detected in the external environment information, adjustment data for the robot's operating parameters is obtained according to the robot's operating state and the external environment information.

[0025] The robot's operating state and external environment information can be estimated using the least squares method, a mathematical optimization technique implemented using matrix operations in linear algebra. The method seeks the optimal function match for the data by minimizing the sum of squared errors, eliminating the effects of sensor noise and environmental interference on parameter estimation. The robot's operating state and external environment information refer to a data set containing the robot's own motion parameters and the dynamic changes of the external environment. This can be achieved using fused data from multi-axis accelerometers, gyroscopes, and lidar, and serves as the input conditions for constructing the robot's kinematic model. Inertial parameters are physical quantities that reflect the robot's mass distribution and motion inertia. These parameters can be implemented using the moment of inertia matrix and center of mass coordinate parameters, and are used to describe the robot's dynamic response characteristics. Motion adjustment refers to the correction of control instructions to the robot's actuators. This can be achieved using a PID controller combined with a dynamic compensation algorithm to eliminate posture deviations during path tracking.

[0026] Specifically, in a dynamic ocean environment, the swaying of ships and the impact of waves can cause sudden changes in the robot's motion state. By collecting information on the robot's joint torque, terminal acceleration, and distance to environmental obstacles, an observation equation containing a time series is established. The adaptive control algorithm first estimates these changing parameters. For example, the recursive least squares method is used to estimate the inertia parameters of the robot arm, and the parameter estimates are continuously updated based on the input and output data collected in real time. The updated inertia parameters are input into the dynamic feedforward controller to generate a compensation torque that is superimposed on the motor drive signal to offset the trajectory deviation caused by external disturbances. For example, when a lateral water flow impact is detected that causes the robot to displace laterally, the system reconstructs the motion trajectory by adjusting the center of gravity position parameters of the robot arm.

[0027] S104: Adjust the operating parameters of the robot according to the adjustment data to obtain an adjustment result, and control the robot to charge the ship through the charging interface according to the adjustment result.

[0028] Motion adjustment refers to correcting the robot's speed, direction, or arm's posture based on real-time feedback, for example, through dynamic deviation correction using a PID controller or model predictive control. If the robot detects a deviation or unexpected obstacle while moving along its path, it replans the local path or adjusts the arm's motion parameters to ensure accurate and secure docking with the charging port. Finally, the charging process is initiated by physically connecting the arm's end effector to the charging port.

[0029] Compared with existing technologies, traditional solutions rely on a single sensor for positioning, which is prone to matching errors when the ship shakes or the lighting changes. This solution significantly improves positioning robustness by fusing three-dimensional model feature points with multi-sensor data. In addition, existing path planning methods do not take into account dynamic obstacles and real-time status feedback. This solution achieves dynamic path optimization by continuously integrating external environmental information, effectively responding to ship berthing position offsets or temporary obstacle interference, solving positioning deviation problems caused by ship shaking and environmental interference, and achieving precise docking of robots to ship charging interfaces in complex environments. It has the advantages of increasing the degree of automation, adapting to complex environmental changes, and improving safety and efficiency.

[0030] Through the above technical solution, this application achieves high-precision positioning of the charging port in complex environments, solving the problem of docking failure caused by ship sway and inclement weather. Through dynamic path planning and real-time motion adjustment, the robot can still safely and reliably complete the charging operation even when obstacles change or the environment interferes, reducing the need for manual intervention and improving charging efficiency and safety.

[0031] In some embodiments of the present invention, step S101 includes: Construct a ship simulation model; the ship simulation model includes three-dimensional model feature points of the charging interface.

[0032] A 3D simulation model of the ship's charging port can also be created, generated through laser scanning or CAD modeling. This model accurately includes the shape and dimensions of the charging port, as well as the spatial coordinates of the 3D model's feature points. Before the robot is operational, this 3D model is stored in the robot's control system.

[0033] After feature extraction of the image data, it is matched with the feature points of the three-dimensional model to obtain the charging interface posture information.

[0034] Path planning is performed based on the charging port posture information to obtain the optimal path, and the robot is controlled to travel toward the ship along the optimal path.

[0035] The charging port pose information, including the position and orientation parameters of the port, is calculated through feature point matching and provides target coordinates for path planning. The optimal path is a global or local collision-free trajectory generated based on the environmental map and obstacle status, for example, using a rasterized map combined with a search algorithm. A camera captures images of the ship, extracts edge, corner, or texture features, and matches them with a pre-established 3D model of the charging port to calculate the precise position and pose of the port. The optimal path is then generated based on the current position and target pose, taking into account the distribution of environmental obstacles. This optimal path allows the robot to navigate toward the ship.

[0036] In some embodiments of the present invention, the process before step S101 includes: Get the current position of the ship and navigate the robot based on the current position; Determine whether the robot has navigated within the preset range of the ship.

[0037] Among them, the current position refers to the real-time position of the robot or ship in the environment. Specifically, it can be achieved by using the global positioning system or visual SLAM technology. Real-time positioning ensures that the robot can accurately approach the target ship. Navigation refers to planning the movement path based on the current position. Specifically, it can be achieved by using a path planning algorithm based on lidar or vision. By dynamically adjusting the path, obstacles are avoided and the robot is guided to the target area. The preset range refers to the safe distance range within which the robot can effectively collect images. Specifically, it can be set to a circular area with a specific radius around the ship. By limiting the distance, the robot avoids the risk of collision due to being too close to the ship. Multiple cameras refer to image acquisition devices installed at different positions on the robot. Specifically, they can be achieved by using a combination of wide-angle lenses, zoom lenses or infrared cameras. The integrity and clarity of the image data of the ship's charging port are ensured through multi-angle coverage.

[0038] Specifically, after startup, the robot first performs a comprehensive self-check of its hardware systems, including checking the proper functioning of its motors, sensors, and communication modules. Detected fault information is recorded and displayed, and if a critical fault is detected, the startup process is halted. Pre-set parameters, such as the robot's speed range, arm joint angle limits, and communication protocol templates for different ship charging systems, are loaded from a storage unit. These parameters provide the foundation for subsequent operations. The robot's multiple cameras are calibrated to ensure the accuracy of their intrinsic parameters (such as focal length and principal point coordinates) and extrinsic parameters (relative position and posture between cameras). This calibration process utilizes a specialized calibration algorithm, using a known calibration plate to capture images at different positions and angles. The robot first uses a positioning system to determine the ship's current position. Using a pre-set navigation algorithm, the robot generates a movement path and autonomously navigates using its own motion mechanisms (such as Mecanum wheels or tracks). Once within a pre-set range of the ship, the robot's cameras capture image data of the ship. During movement, the robot uses range sensors to monitor the surrounding environment in real time to avoid collisions with other objects. Simultaneously, the robot's attitude sensors (gyroscopes and accelerometers) continuously provide feedback on its posture to ensure stable navigation. Once the robot enters a preset range, multiple cameras activate simultaneously, capturing images of the ship's surface from different angles. For example, a wide-angle camera captures the ship's overall outline, a zoom camera focuses on details of the charging port, and an infrared camera supplements visible light data in low-light conditions. This simultaneous acquisition of multi-source image data effectively mitigates interference with a single camera caused by ship motion, changing lighting conditions, or inclement weather, ensuring the accuracy of subsequent feature matching and path planning.

[0039] In some embodiments of the present invention, Figure 2 As shown, after feature extraction of the image data, it is matched with the feature points of the 3D model to obtain the charging port posture information, including: S201. Construct a dynamic model based on the robot arm and mobile platform.

[0040] Among them, the dynamic model refers to a mathematical model that describes the motion laws of the robot's manipulator and mobile platform. It can be implemented using multi-body dynamics equations or Newton-Euler equations. By establishing the relationship between joint torque and motion state, the dynamic response of the manipulator in complex environments can be accurately predicted.

[0041] S202: After the dynamic model is trained, the image data is input into the dynamic model for feature extraction to obtain two-dimensional image feature points.

[0042] Training refers to optimizing the parameters of the dynamics model through a data-driven approach. Specifically, supervised learning methods can be used. For example, backpropagation is performed using the error between the measured data of the robot's motion trajectory and the model's predictions, thereby improving the model's ability to fit real-world physical behavior. Two-dimensional image feature points are pixels extracted from ship images that contain significant geometric or texture information. These include easily recognizable visual elements such as the geometric shape of the charging port edge and specific markings. These can be achieved using a scale-invariant feature transformation algorithm or an accelerated robust feature detector. By detecting key areas such as corners, edges, or spots in the image, a matching local feature descriptor is formed.

[0043] S203: Match the two-dimensional image feature points with the three-dimensional model feature points to obtain charging port posture information.

[0044] In a specific embodiment of the present invention, when the robot performs a charging task, the joint motion characteristics of the robotic arm and the mobile platform are modeled as a dynamic equation that takes into account factors such as joint inertia, friction, and external loads. During the training process, the robot is controlled to perform a series of preset actions while recording the actual motion trajectory and sensor data. By minimizing the difference between the model's predicted values ​​and the measured values, the dynamic model can accurately reflect the physical characteristics of the real system. When the ship image is input into the trained model, the computing unit inside the model processes the image layer by layer and extracts two-dimensional feature points related to the charging interface, such as the contour points of the interface edge or the center point of the positioning mark. These two-dimensional feature points are then spatially corresponded to the pre-constructed three-dimensional model feature points, and the position and posture information of the charging interface relative to the robot are calculated through coordinate transformation and projection relationship.

[0045] Compared to existing technologies, traditional methods typically treat the robotic arm as a static structure for feature matching, ignoring the effects of inertia and joint flexibility during motion. This results in decreased matching accuracy when the ship sways or the robot moves. However, this solution, by introducing a dynamic model, compensates for the dynamic deformation of the robotic arm in real time during the feature extraction phase. This makes the positioning of 2D image feature points more stable, thereby improving the accuracy of subsequent pose estimation.

[0046] Through the above technical solution, the present application can effectively solve the problem of charging interface posture estimation deviation caused by waves or mechanical vibrations of ships. By dynamically compensating for the motion error of the robotic arm, the matching accuracy between the two-dimensional feature points and the three-dimensional model is ensured, providing reliable spatial positioning information for subsequent path planning and charging operations.

[0047] In some embodiments of the present invention, Figure 3 As shown, step S203 includes: S301 , calculating the coordinates of the two-dimensional image feature points and the three-dimensional model feature points according to direct linear transformation to obtain linear equations of rotation parameters and translation parameters.

[0048] Among them, direct linear transformation refers to establishing a linear equation through the mapping relationship between the two-dimensional image coordinate system and the three-dimensional model coordinate system. Specifically, it can be implemented using a matrix transformation model to eliminate the influence of image distortion on coordinate matching.

[0049] S302. Solve the linear equation according to singular value decomposition to obtain a least squares solution.

[0050] Among them, singular value decomposition refers to decomposing a linear equation into the product of three matrices to obtain the optimal solution. Specifically, it can be implemented using numerical calculation methods to reduce the error in parameter solution caused by noise interference.

[0051] S303 , optimizing the least squares solution according to the reprojection error minimization optimization algorithm to obtain the reprojection error sum.

[0052] Among them, the minimization of reprojection error optimization algorithm refers to minimizing the total error of projecting the three-dimensional model onto the two-dimensional image by adjusting the pose parameters. Specifically, it can be implemented using a nonlinear optimization algorithm to improve the accuracy of pose estimation. S304: Iteratively update the reprojection error and to obtain charging port posture information.

[0053] Among them, iterative update refers to adjusting parameters and calculating errors multiple times until the error converges to a preset threshold. Specifically, it can be achieved by using the gradient descent method to enhance the stability of posture information.

[0054] In a specific embodiment of the present invention, in a ship charging scenario, the matching of the two-dimensional image feature points and the three-dimensional model feature points may deviate due to the shaking of the ship or changes in illumination. By establishing linear equations for the rotation and translation parameters through direct linear transformation, the geometric differences of the coordinate system can be preliminarily eliminated. Subsequently, the least squares solution of the linear equation is solved using singular value decomposition, which can effectively suppress parameter fluctuations caused by sensor noise or image blur. Furthermore, the initial solution is nonlinearly optimized by minimizing the reprojection error optimization algorithm, which can correct the projection error caused by environmental interference. Finally, the reprojection error is gradually converged through iterative updates, and accurate and stable charging interface posture information is ultimately output.

[0055] Specifically, the PnP algorithm is used to match the extracted 2D image feature points with the pre-stored 3D model feature points. By solving the corresponding mathematical equations. Assuming that there is n points P =( X i ,Y i , Z i ) T , its projection point in the image coordinate system is p i =( u i , v i ) T The intrinsic parameter matrix of the camera is known, which projects the points in the camera coordinate system to the image coordinate system. Its form is generally shown in formula (1): (1) Where, 、 The focal length is x 、 y The direction of the component, , are the coordinates of the center of the image.

[0056] Direct linear transformation (DLT) method is used to solve the problem. Let the external parameters of the camera (including rotation and translation) be [ R | t ],in R is a 3×3 rotation matrix, t is a 3×1 translation vector. According to the projection relationship, s i p i = K [ R | t ] P i ,in s i is a proportional factor. Expanding this formula yields the following formula (2): (2) In the formula, [ R | t ]for .

[0057] Further expansion yields and .

[0058] By eliminating , we can get the linear equations about the rotation and translation parameters, which can generally be written as Ax =0 form, where A is a coefficient matrix consisting of the coordinates of the known 3D points and the 2D projection points, x is a vector containing the rotation and translation parameters.

[0059] Singular value decomposition (SVD) is used to solve the least squares solution: Due to the existence of noise and other errors in actual measurement, the above linear equations are generally overdetermined, and the least squares solution is solved by singular value decomposition (SVD). A Perform SVD decomposition, , the least squares solution x yes V The vector corresponding to the last column of (except that the last element is 0). The obtained solution needs to be normalized and processed, because the directly obtained "does not fully comply with the physical constraints of rotation and translation.

[0060] The result obtained by direct linear transformation is only a preliminary estimate. In order to solve it more accurately, nonlinear optimization methods can be used. The commonly used method is the optimization algorithm based on minimizing the reprojection error. The reprojection error is defined as the observed image point and the points reprojected according to the estimated pose The distance between Use an optimization algorithm (such as Gauss-Newton or Levenberg-Marquardt) to minimize the sum of the reprojection errors of all points By iteratively updating the pose estimation until convergence, a more accurate position and pose information of the charging port in the robot coordinate system is obtained. Finally, based on the calculated charging port pose information, the robot control system plans the motion path of the manipulator, guiding the manipulator to accurately align the charging plug with the charging port.

[0061] In some embodiments of the present invention, Figure 4 As shown, path planning is performed based on the charging port posture information to obtain the optimal path, including: S401: Construct a grid map according to the robot's environment, where the grid map includes the obstacle status of each grid.

[0062] A grid map is a two- or three-dimensional map that discretizes the robot's environment into multiple grid cells. This can be achieved by using a rasterization method to divide the environment into square grid cells with a side length of, for example, 0.5 meters. Obstacle status refers to the information indicating whether each grid cell is occupied by an obstacle. This can be achieved by using lidar or visual sensors to detect environmental obstacles in real time and update the grid status. Map building: The robot's working environment is constructed as a grid map, where each grid cell has a corresponding status, such as whether it is an obstacle.

[0063] S402: Determine the starting point and the end point in the grid map based on the current position of the robot and the posture information of the charging port.

[0064] Among them, according to the current position of the robot and the posture information of the charging interface, the starting point is confirmed as the current position of the robot, and the end point is confirmed as the position of the charging interface.

[0065] S403: Search the starting point and the end point in the grid map according to a preset search algorithm and the obstacle status to obtain the optimal path.

[0066] The pre-set search algorithm refers to a graph-based path planning algorithm. Specifically, it can be implemented using the A algorithm, Dijkstra's algorithm, or a modified version thereof. It is used to find the shortest or lowest-cost path from a starting point to a destination on a grid map. The heuristic function in the A algorithm is generally the sum of the actual cost from the starting point to the current point and the estimated cost from the current point to the destination.

[0067] Specifically, in the ship charging scenario, the robot needs to plan its movement path based on the position information of the charging interface. First, the robot's surrounding environment is modeled as a grid map, and each grid mark indicates whether there are obstacles, such as fixed structures at the dock, temporary stacks, or dynamic obstacles caused by the shaking of the ship. The robot's current position is used as the starting point, and the charging interface coordinates are mapped to the grid map as the end point. Subsequently, a preset search algorithm is used to traverse the candidate nodes in the open list, gradually expanding the path and avoiding the grids marked with obstacles, and finally generating a collision-free optimal path connecting the starting point and the end point. For example, when the ship's position is offset due to wind and waves, the obstacle status in the grid map can be updated in real time, and the search algorithm dynamically adjusts the path to avoid collisions with newly appeared obstacles.

[0068] Compared with existing technologies, existing ship charging robots typically rely on fixed paths or simple obstacle avoidance strategies, making them difficult to adapt to the dynamic dock environment and the displacement of charging ports caused by ship sway. This method uses a grid map to represent the distribution of environmental obstacles in real time, and combines it with a heuristic search algorithm to achieve dynamic path planning. This method can effectively adapt to complex and changing charging scenarios, improving path planning efficiency and safety.

[0069] Through the above technical solution, this application can solve the problem of path failure caused by dynamic changes in obstacles in the ship charging environment. The discretization of the grid map reduces the complexity of environmental modeling. The preset search algorithm ensures path optimality while reducing computing resource consumption. This enables the robot to quickly generate a reliable charging path under complex conditions such as ship swaying and temporary obstacles, avoiding the risk of charging interruption or equipment damage caused by path planning failure.

[0070] In some embodiments of the present invention, Figure 5 As shown, step S403 includes: S501: Put the starting point into an open list, and calculate the starting point using a heuristic function of a preset search algorithm to obtain a heuristic function value.

[0071] The open list is a data structure that stores the nodes to be evaluated. It can be implemented as a priority queue and is used to dynamically maintain the current set of optimal candidate nodes. The heuristic function is an evaluation function used to estimate the distance from a node to the destination. It can be implemented as Manhattan distance or Euclidean distance and is used to guide the search direction toward the target location. The starting point is placed in the open list, and the heuristic function of the pre-set search algorithm is used to calculate the starting point's heuristic value (usually the sum of the estimated distance to the destination and the distance traveled).

[0072] S502: Identify the node with the smallest heuristic function value from the open list as the current node and put it into the closed list.

[0073] The current node refers to the node with the highest priority evaluated in each iteration. It can be determined by comparing the heuristic function values ​​of each node in the open list and is used as the reference point for path extension.

[0074] S503: Traverse the adjacent nodes of the current node. When there is an obstacle in the adjacent node or the adjacent node is in the closed list, skip the adjacent node.

[0075] Among them, the adjacent nodes refer to the grid units directly connected to the current node in the grid map, which can specifically include eight neighborhood grids in the up, down, left, right, and diagonal directions, which are used to construct the topological relationship of the path search.

[0076] S504: When there is no obstacle or the adjacent node is not in the closed list, the heuristic function value of the adjacent node is calculated using the heuristic function, and the adjacent node is placed in the open list.

[0077] The node with the smallest heuristic function value is selected from the open list as the current node and put into the closed list; the adjacent nodes of the current node are traversed, and if the adjacent node is an obstacle or the adjacent node is in the closed list, the adjacent node is skipped; otherwise, its heuristic function value is calculated and put into the open list or its value is updated.

[0078] S505. The node with the smallest heuristic function value is confirmed again from the open list and searched until the open list is empty or the search reaches the end, and the optimal path is generated by backtracking the parent node of each node in the closed list.

[0079] Steps S502-S504 are repeated until the open list is empty or the end point is found. If the end point is found, a path is formed by backtracking to the parent node. This effectively finds the shortest or best path from the starting point to the end point on the map, avoiding obstacles and obtaining the optimal path.

[0080] Compared to existing technologies, this method expands nodes from the starting point in ascending order of heuristic function values ​​until the end point is found or the entire map is searched. Its purpose is efficient planning: Algorithm A can quickly find an effective path from the robot's current position to the charging port, improving the robot's operating efficiency. Obstacle avoidance: During path planning, obstacles in the environment are taken into account, allowing the robot to safely move to the charging port for charging. Path planning must consider the robot's current position and posture, information about obstacles in the surrounding environment, and the kinematic constraints of the robotic arm. Traditional global path planning algorithms, such as the Dijkstra algorithm, require traversing all possible nodes, resulting in low computational efficiency. However, this method uses a heuristic function to guide the search direction, effectively narrowing the node evaluation range. It also uses a grid map and closed list mechanism to avoid repeated calculations, resulting in higher path reliability than random sampling algorithms. In ship charging scenarios, this method can quickly generate the shortest path that avoids obstacles and adapts to the complex layout of fixed obstacles and temporary storage in dock environments.

[0081] In some embodiments of the present invention, after step S104, the method further includes: The robot's operating status and external environment information are continuously monitored. When the robot's operating status changes and / or obstacles are detected in the external environment information, the robot's movement is adjusted based on the robot's operating status and external environment information.

[0082] Among them, electrical parameters refer to physical quantities that characterize the charging status during the charging process. Specifically, voltage sensors, current sensors or temperature sensors can be used to collect voltage, current or temperature data of the charging interface in real time to determine whether the charging has reached the preset threshold. The preset charging completion condition refers to the pre-set charging termination standard. Specifically, the voltage reaching the rated value, the current dropping to the maintenance value or the charging time exceeding the set threshold can be used as judgment conditions to trigger the robot separation action. External environment information refers to the status data of the space in which the robot is located. Specifically, lidar, visual sensors or inertial measurement units can be used to obtain obstacle distribution, sea surface fluctuation amplitude or robot posture data, which can be used to avoid dynamic obstacles when planning the separation path.

[0083] Specifically, during robot motion, the system's dynamic parameters change due to factors such as load variations (such as the weight of the charging plug carried at the end of the manipulator arm) and environmental disturbances (i.e., external environmental information, such as the impact of wind and waves on the ship). The adaptive control algorithm first estimates these changing parameters. For example, it uses recursive least squares to estimate the manipulator's inertial parameters, continuously updating the estimated parameters based on real-time input and output data. For example, if the robot's mobile platform detects an increase in ground friction (perhaps due to a wet deck, for example), the adaptive control algorithm will increase the motor's output torque accordingly to ensure the robot approaches the charging port along the predetermined trajectory. Throughout the docking process, the adaptive control algorithm automatically adapts to various uncertainties, improving the accuracy and stability of the robot's motion. The algorithm can adjust the robot's speed, direction, and the manipulator's trajectory in real time. For example, if the ship shakes, the robot can quickly adjust the manipulator's position using feedback from its attitude and vision sensors to maintain the relative position of the charging plug and port, ensuring a stable charging process.

[0084] In some embodiments of the present invention, after step S104, the method further includes: The electrical parameters of the ship are continuously monitored. When the electrical parameters meet the preset charging completion conditions, the robot is controlled to separate from the charging interface based on external environmental information.

[0085] In a specific embodiment of the present invention, the robot obtains electrical parameters from the ship's charging system (such as charging voltage, current, and charging mode). Based on these parameters, the robot sets its own charging output parameters and feeds back information indicating it is ready to charge to the ship's charging system. During the charging process, the robot continuously monitors parameters such as charging current and voltage, and transmits these parameters in real time to the ship's charging system via a communication module. Simultaneously, the robot's sensors monitor its surroundings and its own status. If any abnormalities occur (such as a loose charging port, severe ship shaking, or electrical failure), the robot promptly takes appropriate measures (such as pausing charging or repositioning the robotic arm). The charging system determines whether charging is complete based on preset charging completion conditions (such as the charging capacity reaching a set value or the charging current falling below a threshold). Once charging is confirmed to be complete, the robot prepares to detach from the charging port. The robot controls the robotic arm to slowly remove the charging plug from the charging port and moves it to its initial position or a safe parking position. During the evacuation process, care is taken to avoid damage to the charging port and the surrounding environment. Using the navigation system and movement mechanism, the robot leaves the ship according to a preset return path or instructions and returns to its designated parking location. During the return process, care is taken to avoid collisions with surrounding objects and maintain a stable driving posture.

[0086] Furthermore, the robot's housing and key electronic components are constructed from special corrosion-resistant materials or coatings, such as corrosion-resistant stainless steel, marine-grade aluminum alloy, and anti-corrosion paint. High-performance sealing structures, including rubber seals and sealants, are designed for water-prone areas such as joints and interfaces to prevent the intrusion of seawater and moisture.

[0087] Wind and wave-resistant design: The robot's mechanical structure is designed to lower its center of gravity and increase its stability. Simultaneously, the design of the mobile mechanism and robotic arm is optimized to automatically adjust its posture and force distribution when impacted by wind and waves. For example, when encountering strong winds, the robot can maintain a stable position on the ship's deck by adjusting the friction and driving force of its tracks or Mecanum wheels. The robotic arm can also employ flexible control strategies based on the ship's swaying caused by wind and waves to minimize the impact of this sway on the charging port.

[0088] The vision system based on the PnP algorithm in this embodiment of the present invention enables the robot to locate the ship's charging port with extremely high accuracy. Compared to existing robots that rely solely on traditional navigation technology, this robot can accurately identify the position and posture of the charging port even in the presence of complex interference factors such as ship motion, surface dirt, and seawater reflection. This precise positioning capability effectively avoids charging port connection failure or damage caused by inaccurate positioning. The adaptive clamping device at the end of the robot arm, combined with a precise motion control algorithm, ensures a tight and accurate connection between the charging plug and the port. Compared to the loose connection that can occur in existing technologies, this invention ensures a stable charging process, improves charging quality and efficiency, and reduces damage to the charging port and equipment caused by poor contact. The multifunctional communication module in the automated control system is compatible with the communication protocols of various ship charging systems. This enables the robot to seamlessly integrate with different types of ship charging systems, overcoming the poor compatibility between the automated control and ship charging systems in existing technologies. Regardless of the protocol used by the ship's charging system, the robot can accurately obtain charging parameters and achieve stable charging control. It also demonstrates excellent adaptability to marine environments. From material selection to structural design, the design fully considers harsh conditions such as seawater corrosion, high humidity, and wind and waves. Compared to conventional robots, this invention can operate stably and reliably in long-term marine environments, reducing the frequency of maintenance due to environmental factors, lowering operating costs, and extending the robot's service life. By integrating data from multiple sensors (visual, attitude, and distance sensors), and utilizing advanced Kalman filtering and motion control algorithms, the robot can perceive its own state and changes in the surrounding environment in real time. In dynamic environments such as ship sway, it can quickly and accurately adjust its movements to ensure a stable charging process. This intelligent control capability offers significant advantages over existing technologies in coping with complex environmental changes. During the charging process, it continuously monitors charging parameters and environmental conditions, enabling timely response to various abnormalities. Whether it's a loose charging port, severe ship swaying, or an electrical fault, the robot can quickly respond by pausing charging and repositioning its robotic arm, ensuring a safe and stable charging process and preventing damage to the ship's charging equipment due to unexpected situations.

[0089] In order to better implement the robot-based ship charging method in the embodiment of the present invention, based on the robot-based ship charging method, the embodiment of the present invention also provides a robot-based ship charging device, such as Figure 6 As shown, the robot-based ship charging device 600 includes: The data acquisition module 601 is used to acquire sensor data and image data of the ship's charging interface during the driving process when the robot travels within a preset range of the ship; An information confirmation module 602 is used to determine the robot's operating status and external environment information based on the image data and sensor data; The data adjustment module 603 is configured to obtain adjustment data for the robot's operating parameters based on the robot's operating state and the external environment information when the robot's operating state changes and / or an obstacle is detected in the external environment information; The ship charging module 604 is used to adjust the operating parameters of the robot according to the adjustment data, obtain the adjustment result, and control the robot to charge the ship through the charging interface according to the adjustment result.

[0090] The robot-based ship charging device 600 provided in the above embodiment can implement the technical solution described in the above embodiment of the robot-based ship charging method. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above embodiment of the robot-based ship charging method, which will not be repeated here.

[0091] The above is a detailed introduction to the robot-based ship charging method and device provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A robot-based ship charging method, characterized in that: include: When the robot travels within the preset range of the ship, it obtains sensor data during the travel process and image data of the ship's charging interface; Determining the robot's operating state and external environment information based on the image data and the sensor data; When the operating state of the robot changes and / or an obstacle is detected in the external environment information, adjustment data of the operating parameters of the robot are obtained according to the operating state of the robot and the external environment information; The operating parameters of the robot are adjusted according to the adjustment data to obtain an adjustment result, and the robot is controlled to charge the ship through the charging interface according to the adjustment result.

2. The robot-based ship charging method according to claim 1, characterized in that: After the robot travels to a preset range of the ship, the method further includes: Constructing a ship simulation model; the ship simulation model includes three-dimensional model feature points of the charging interface; Extracting features from the image data and matching them with the feature points of the three-dimensional model to obtain charging port posture information; Path planning is performed according to the charging port posture information to obtain an optimal path, and the robot is controlled to travel toward the ship according to the optimal path.

3. The robot-based ship charging method according to claim 2, characterized in that: The extracting features from the image data and matching them with the feature points of the three-dimensional model to obtain the charging port posture information includes: Build a dynamic model based on the robot arm and mobile platform; After the dynamic model is trained, the image data is input into the dynamic model for feature extraction to obtain two-dimensional image feature points; The two-dimensional image feature points are matched with the three-dimensional model feature points to obtain charging port posture information.

4. The robot-based ship charging method according to claim 3, characterized in that: The matching of the two-dimensional image feature points with the three-dimensional model feature points to obtain charging port posture information includes: Calculating the coordinates of the two-dimensional image feature points and the three-dimensional model feature points according to direct linear transformation to obtain linear equations of rotation parameters and translation parameters; Solving the linear equation according to singular value decomposition to obtain a least squares solution; Optimizing the least squares solution according to a reprojection error minimization optimization algorithm to obtain a reprojection error sum; The reprojection error and are iteratively updated to obtain charging port posture information.

5. The robot-based ship charging method according to claim 2, characterized in that: The performing path planning according to the charging port posture information to obtain the optimal path includes: Constructing a grid map according to the environment of the robot, wherein the grid map includes an obstacle status of each grid; Determine a starting point and an end point in the grid map according to the current position of the robot and the posture information of the charging port; The starting point and the end point in the grid map are searched according to a preset search algorithm and the obstacle status to obtain an optimal path.

6. The robot-based ship charging method according to claim 5, characterized in that: The searching the starting point and the end point in the grid map according to a preset search algorithm and the obstacle status to obtain an optimal path includes: Putting the starting point into an open list, and calculating the starting point using the heuristic function of the preset search algorithm to obtain a heuristic function value; Determine the node with the smallest heuristic function value from the open list as the current node and put it into the closed list; Traversing the adjacent nodes of the current node, and skipping the adjacent node when there is an obstacle in the adjacent node or the adjacent node is in the closed list; When the obstacle does not exist or the adjacent node is not in the closed list, calculating the heuristic function value of the adjacent node by using the heuristic function, and placing the adjacent node into the open list; The node with the smallest heuristic function value is confirmed again from the open list and searched until the open list is empty or the end point is searched, and the optimal path is generated by backtracking the parent node of each node in the closed list.

7. The robot-based ship charging method according to claim 1, characterized in that: After controlling the robot to charge the ship through the charging interface according to the adjustment result, the method further includes: The operating state of the robot and the external environment information are continuously monitored. When the operating state of the robot changes and / or an obstacle is detected in the external environment information, the movement of the robot is adjusted according to the operating state of the robot and the external environment information.

8. The robot-based ship charging method according to claim 1, characterized in that: After controlling the robot to charge the ship through the charging interface according to the adjustment result, the method further includes: The electrical parameters of the ship are continuously monitored, and when the electrical parameters meet the preset charging completion conditions, the robot is controlled to separate from the charging interface according to the external environment information.

9. The robot-based ship charging method according to claim 1, characterized in that: Before the robot travels to within the preset range of the ship, the method further includes: Obtaining a current position of the ship and navigating the robot according to the current position; Determine whether the robot has navigated to within a preset range of the ship.

10. A robot-based ship charging device, characterized in that: include: A data acquisition module is used to acquire sensor data and image data of the ship's charging interface during the driving process when the robot travels within a preset range of the ship; An information confirmation module, configured to determine the robot's operating status and external environment information based on the image data and the sensor data; a data adjustment module, configured to obtain adjustment data for the operating parameters of the robot according to the operating state of the robot and the external environment information when the operating state of the robot changes and / or an obstacle is detected in the external environment information; The ship charging module is used to adjust the operating parameters of the robot according to the adjustment data to obtain an adjustment result, and control the robot to charge the ship through the charging interface according to the adjustment result.

Images