Accurate positioning method and system for wheelchair following

Abstract

The invention relates to the technical field of intelligent auxiliary equipment, and discloses an accurate positioning method and system for wheelchair following, and the method comprises the steps: constructing a global environment map of a target following object, and determining a target position of the target following object in the global environment map; setting a dynamic obstacle avoidance path from the wheelchair to the target position, and identifying an optimal following path from the wheelchair to the target following object; an automatic control instruction of the wheelchair is set, the motion change state of the wheelchair is monitored in real time, and a following path control unit of the wheelchair to the target following object is created; identifying an error positioning factor of the wheelchair, and setting a positioning error compensation mechanism of the wheelchair; and in combination with a global environment map, the optimal following path, a following path control unit and a positioning error compensation mechanism, executing accurate positioning processing of the wheelchair on the target following object. Accurate positioning of wheelchair following can be achieved, the applicability of the wheelchair following function is improved, a user can travel more safely and independently, and the life autonomy of the user is improved.

Description

Technical Field

[0001] The present invention relates to a precise positioning method and system for wheelchair following, belonging to the technical field of intelligent auxiliary equipment. Background Art

[0002] Wheelchair following refers to a function in which a wheelchair automatically identifies and tracks the motion trajectory of a target object with the help of onboard sensors and control systems, in order to maintain its relative position and distance with the target object. As an important tool to help people with physical disabilities achieve self-care, return to "normal" society and improve their sense of well-being in life, the wheelchair's following function brings great convenience to users, enabling them to interact with the surrounding environment more independently when moving. However, without precise positioning, the wheelchair may not be able to accurately perceive the positional relationship between itself and the target object in space, resulting in deviations in the following process. For example, a collision is likely to occur when the distance is too close, while the target may be lost if the distance is too far. Therefore, precise positioning is the core element to ensure the stable and efficient operation of the wheelchair's following function.

[0003] Traditional wheelchair tracking and positioning methods mainly use visual positioning methods, which use cameras to capture visual markers on the user, such as clothing of specific colors or skeletal joints, and calculate the relative position of the user and the wheelchair. However, this method is easily affected by changes in lighting, resulting in a decrease in positioning accuracy, which in turn affects the applicability of the wheelchair tracking function.

[0004] Therefore, there is an urgent need for a solution that can achieve accurate positioning of wheelchair following, improve the applicability of the wheelchair following function, allow users to travel more safely and independently, and enhance users' life autonomy. Summary of the Invention

[0005] The present invention provides a precise positioning method and system for wheelchair following, the main purpose of which is to achieve precise positioning of wheelchair following, improve the applicability of the wheelchair following function, enable users to travel more safely and independently, and enhance users' life autonomy.

[0006] To achieve the above objectives, the present invention provides a precise positioning method for wheelchair following, comprising:

[0007] Acquire a target following object of the wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data;

[0008] Tracking real-time position information of the target following object, identifying a motion trajectory of the target following object based on the real-time position information, and determining a target position of the target following object in the global environment map, and identifying dynamic obstacles between the wheelchair and the target position based on the global environment map;

[0009] setting a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle, analyzing a movement change trend of the target following object based on the motion trajectory, and identifying an optimal following path for the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path;

[0010] Based on the optimal following path, setting an automated control instruction for the wheelchair, monitoring the motion change state of the wheelchair in real time, and creating a following path control unit for the wheelchair to follow the target object by combining the motion change state and the automated control instruction;

[0011] identifying an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and setting a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state;

[0012] In combination with the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism, the wheelchair performs precise positioning processing on the target following object to obtain a precise positioning result.

[0013] Optionally, determining the target position of the target following object in the global environment map based on the real-time position information includes:

[0014] generating sensor coordinates of the target following object based on the real-time position information;

[0015] Mapping the real-time position information into a map coordinate system of the global environment map according to the sensor coordinates to obtain an initial target position of the target following object;

[0016] Extracting map environment features of the target following object based on the initial target position;

[0017] Identifying actual surrounding environment features of the target following object according to the sensor coordinates;

[0018] Calculating a feature matching degree between the actual surrounding environment feature and the map environment feature;

[0019] According to the feature matching degree, performing position calibration processing of the target following object to obtain a position calibration result;

[0020] Based on the position calibration result, a target position of the target following object in the global environment map is determined.

[0021] Optionally, identifying dynamic obstacles between the wheelchair and the target location based on the global environment map includes:

[0022] setting an initial path for the wheelchair to the target location according to the global environment map;

[0023] Collecting historical point cloud data of the initial path and extracting current frame data corresponding to the historical point cloud data;

[0024] Simulating a background model of the initial path based on the historical point cloud data;

[0025] identifying a motion region of the initial path according to the current frame data and the background model;

[0026] Extracting moving objects in the moving area and analyzing the movement change trend of the moving objects;

[0027] Performing spatial position matching processing on the motion change trend and the initial path to obtain a spatial position matching result;

[0028] Analyzing the degree of influence of the moving object on the initial path based on the spatial position matching result;

[0029] According to the impact degree, dynamic obstacles between the wheelchair and the target position are identified.

[0030] Optionally, setting a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle includes:

[0031] measuring a velocity parameter of the dynamic obstacle and identifying obstacle characteristics of the dynamic obstacle;

[0032] According to the characteristics of the obstacle, a basic safety distance between the dynamic obstacle and the wheelchair is set;

[0033] Based on the speed parameter and the basic safety distance, the dynamic safety distance between the wheelchair and the dynamic obstacle is calculated using the following formula:

[0034]

[0035] Among them, D represents the dynamic safety distance of the dynamic obstacle, D0 represents the basic safety distance between the dynamic obstacle and the wheelchair, T average It represents the average reaction time of the wheelchair user, which is usually 0.2-0.5 seconds. V represents the speed parameter of the dynamic obstacle. A represents the maximum braking acceleration of the wheelchair, which can be 3m / s. 2 ;

[0036] Constructing an obstacle avoidance zone for the wheelchair according to the dynamic safety distance and the dynamic obstacle;

[0037] measuring kinematic constraint parameters of the wheelchair, and setting a movable area of ​​the wheelchair based on the kinematic constraint parameters;

[0038] Defining an obstacle avoidance path constraint space from the wheelchair to the target position by combining the obstacle avoidance zone and the movable area;

[0039] determining a feasible path from the wheelchair to the target location according to the obstacle avoidance path constraint space;

[0040] A dynamic obstacle avoidance path from the wheelchair to the target position is set based on the obstacle avoidance path constraint space and the feasible path.

[0041] Optionally, the identifying an optimal following path from the wheelchair to the target following object according to the movement change trend and the dynamic obstacle avoidance path includes:

[0042] Analyzing usage scenarios and user needs of the wheelchair based on the movement change trend and the dynamic obstacle avoidance path;

[0043] Setting the following task priority of the wheelchair based on the usage scenario and user needs;

[0044] Constructing a task priority matrix of the wheelchair according to the following task priority;

[0045] Based on the task priority matrix, the path score from the wheelchair to the target following object is calculated using the following formula:

[0046]

[0047] Among them, S represents the path score from the wheelchair to the target following object, w j represents the weight of the jth follow-up task in the task priority matrix, which can be 0.2, s represents the path position, and f j (s) represents the execution function of the jth follower task with respect to the path position s, with a value range of [0,1], ds represents the integral variable, j represents the follower task index in the task priority matrix, and n represents the total number of follower tasks in the task priority matrix;

[0048] An optimal following path from the wheelchair to the target following object is identified according to the path score.

[0049] Optionally, setting the automatic control instructions of the wheelchair based on the optimal following path includes:

[0050] determining a driving direction and a driving speed of the wheelchair based on the optimal following path;

[0051] Identifying the user of the wheelchair and the driving surface, and analyzing the ground material of the driving surface;

[0052] collecting arm muscle signals of the user, and identifying the user's movement intention based on the arm muscle signals;

[0053] Analyzing the difficulty of the wheelchair in moving based on the ground material;

[0054] Identifying the distance between the wheelchair and obstacles in the optimal following path, and calculating the moving speed of the object in the optimal following path;

[0055] determining an environmental safety level of the optimal following path according to the obstacle distance and the moving speed of the object;

[0056] generating a decision control vector for the wheelchair based on the movement intention, the driving difficulty level, and the environmental safety level;

[0057] Setting a parameter adaptive adjustment mechanism of the driving direction and the driving speed according to the decision control vector;

[0058] In combination with the decision control vector and the parameter adaptive adjustment mechanism, an automated control instruction for the wheelchair is set.

[0059] Optionally, the combining the motion change state and the automation control instruction to create a following path control unit for the wheelchair to follow the target following object includes:

[0060] Analyzing the environmental adaptability of the wheelchair according to the motion change state;

[0061] extracting safety protection instructions from the automation control instructions based on the environmental adaptability;

[0062] Setting a boundary detection mechanism for the wheelchair according to the safety protection instruction;

[0063] Based on the boundary detection mechanism, an abnormal feedback system of the motion change state is established;

[0064] Analyzing the abnormal cause of the motion change state according to the abnormal feedback system;

[0065] Based on the abnormal cause, identifying the object attribute of the target following object;

[0066] Constructing a multi-target tracking network for the wheelchair according to the object attributes;

[0067] Combining the boundary detection mechanism, the abnormal feedback system and the multi-target tracking network, a control unit for the wheelchair to follow the target object is created.

[0068] Optionally, identifying an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle includes:

[0069] extracting, according to the motion change state, motion change characteristics of the wheelchair under the dynamic obstacle;

[0070] identifying an error pattern of the wheelchair under the dynamic obstacle based on the motion change characteristics;

[0071] Performing cumulative effect analysis and processing of the error pattern to obtain a cumulative law;

[0072] Analyzing the positioning deviation of the wheelchair according to the accumulation rule;

[0073] The error positioning factor of the wheelchair is identified by combining the error pattern and the positioning deviation.

[0074] Optionally, setting a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state includes:

[0075] extracting sensor nodes of the wheelchair based on the error location factor;

[0076] identifying a movement scene of the wheelchair according to the movement change state;

[0077] Analyzing the error positioning performance of the sensor node in the motion scenario;

[0078] determining an optimal sensor combination for the wheelchair based on the error positioning performance;

[0079] monitoring the target loss condition of the wheelchair in real time according to the optimal sensor combination;

[0080] Setting a target repositioning mechanism for the wheelchair based on the target loss condition and the optimal sensor combination;

[0081] In combination with the error positioning performance, the optimal sensor combination and the target repositioning mechanism, a positioning error compensation mechanism for the wheelchair is set.

[0082] In order to solve the above problems, the present invention further provides a precise positioning system for wheelchair following, the system comprising:

[0083] A target environment recognition module is used to obtain a target following object of the wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data;

[0084] The obstacle recognition module is used to track the real-time position information of the target following object, identify the motion trajectory of the target following object based on the real-time position information, determine the target position of the target following object in the global environment map, and identify dynamic obstacles from the wheelchair to the target position based on the global environment map.

[0085] a following path setting module, configured to set a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle, analyze the movement change trend of the target following object based on the motion trajectory, and identify an optimal following path for the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path;

[0086] a following control module, configured to set an automated control instruction for the wheelchair based on the optimal following path, monitor the motion change state of the wheelchair in real time, and create a following path control unit for the wheelchair to follow the target following object by combining the motion change state and the automated control instruction;

[0087] an error compensation module, configured to identify an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and set a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state;

[0088] The positioning execution module is used to combine the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism to perform precise positioning processing of the wheelchair on the target following object to obtain a precise positioning result.

[0089] Compared with the problems described in the background technology, the embodiment of the present invention constructs a global environment map of the target following object based on the environmental data, which can help the wheelchair clearly know the relative position relationship between itself and the target following object and other objects, and plan the following path; further, the embodiment of the present invention identifies the dynamic obstacles from the wheelchair to the target position according to the global environment map, which can optimize the wheelchair following path, avoid collision accidents, and ensure the personal safety of the user; the embodiment of the present invention sets a dynamic obstacle avoidance path from the wheelchair to the target position according to the dynamic obstacles, which can ensure that the wheelchair can adapt to various complex and changing environments, reduce the occurrence of wheelchairs and obstacles. The risk of collision is reduced, thereby ensuring the personal safety of the user during travel; further, the embodiment of the present invention identifies the optimal following path from the wheelchair to the target following object according to the action change trend and the dynamic obstacle avoidance path, so that the wheelchair can follow the target in the fastest and most efficient way, reduce unnecessary detours and waiting, and save time and energy consumption; the embodiment of the present invention sets the automatic control instructions of the wheelchair based on the optimal following path, so that the wheelchair can better cope with complex environmental changes, improve the safety and stability of wheelchair travel, and enhance the user's mobility autonomy and psychological sense of security; further, the embodiment of the present invention combines the motion change state and the automatic control instructions, Creating a following path control unit for the wheelchair to the target following object can not only make an immediate response according to the current real-time situation, but also predict future movement trends, make corresponding plans in advance, and improve the overall following effect and user experience; the embodiment of the present invention identifies the error positioning factor of the wheelchair according to the movement change state and the dynamic obstacle, and can optimize the movement control of the wheelchair according to these factors, so that the wheelchair can follow the movement trajectory of the target object more smoothly and accurately, thereby improving the positioning accuracy of the wheelchair; further, the embodiment of the present invention sets the positioning error compensation mechanism of the wheelchair based on the error positioning factor and the movement change state, which can be used for the wheelchair Targeted compensation is performed based on the error patterns and positioning deviations in different scenarios, thereby more effectively improving positioning accuracy, improving the robustness and reliability of the entire wheelchair following system, and reducing the risks of positioning errors and loss of control; finally, the embodiment of the present invention combines the global environment map, the optimal following path, the following path control unit, and the positioning error compensation mechanism to perform precise positioning processing of the wheelchair on the target following object, obtain a precise positioning result, accurately identify the position of the target following object, effectively avoid the influence of lighting changes, greatly improve positioning accuracy, and comprehensively enhance the applicability and reliability of the wheelchair following function in complex and diverse scenarios, bringing users a more stable and accurate following experience.Therefore, the precise positioning method and system for wheelchair following provided by the embodiments of the present invention can achieve precise positioning of wheelchair following, improve the applicability of the wheelchair following function, allow users to travel more safely and independently, and enhance users' life autonomy. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 A flowchart of a precise positioning method for wheelchair following provided by one embodiment of the present invention;

[0091] Figure 2 A schematic diagram of modules for implementing a precise positioning method for wheelchair following provided in one embodiment of the present invention.

[0092] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0093] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0094] The embodiments of the present application provide a precise positioning method for wheelchair following. The execution subject of the precise positioning method for wheelchair following includes but is not limited to at least one of electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiments of the present application. In other words, the precise positioning method for wheelchair following can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc.

[0095] Example 1:

[0096] Reference Figure 1 FIG. 1 is a flow chart of a precise positioning method for wheelchair following provided by an embodiment of the present invention. In this embodiment, the precise positioning method for wheelchair following includes:

[0097] S1. Acquire a target following object of a wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data.

[0098] By acquiring the target following object of the wheelchair, the embodiment of the present invention can provide core pointing for subsequent following positioning. The wheelchair refers to a wheelchair device with an intelligent following function, and the target following object refers to a specific object that the wheelchair needs to track and follow, such as a caregiver or a QR code on an object.

[0099] Furthermore, the embodiments of the present invention can construct a comprehensive and accurate global environmental map by collecting environmental data of the target following object, so that the wheelchair can have a holistic understanding of the environment in which it is located. The environmental data refers to the environmental information of the target following object of the wheelchair, including but not limited to spatial layout information, environmental feature information, and dynamic information, such as the open and close status of doors, the movement of objects, or the appearance of new obstacles.

[0100] Optionally, the environmental data of the target following object can be collected by a laser radar.

[0101] The embodiment of the present invention constructs a global environment map of the target following object based on the environmental data, which can help the wheelchair clearly understand the relative position relationship between itself and the target following object and other objects, and plan the following path. The global environment map refers to a map that comprehensively and accurately describes and models the entire environment in which the target following object is located, including spatial information, obstacle information and feature information, such as the position and range of rooms, corridors and stairs in indoor environments, and pedestrians and moving vehicles in outdoor environments.

[0102] Optionally, the global environment map of the target following object based on the environment data can be constructed using a GMapping algorithm, such as using a Rao-Blackwellized particle filter (RBPF) algorithm to fuse laser data and odometry information to generate a two-dimensional grid map.

[0103] S2. Track the real-time position information of the target following object, identify the motion trajectory of the target following object based on the real-time position information, and determine the target position of the target following object in the global environment map; and identify dynamic obstacles from the wheelchair to the target position based on the global environment map.

[0104] The embodiment of the present invention can help the wheelchair know the current specific location of the target in the global map by tracking the real-time position information of the target following object, and achieve precise following of the target following object. The real-time position information refers to the precise position data of the target following object relative to the reference coordinate system (such as the global map coordinate system) at a specific moment.

[0105] Optionally, the real-time position information of the target following object may be tracked by a visual sensor, such as a depth camera.

[0106] Furthermore, the embodiment of the present invention helps the wheelchair better understand the target's intention by identifying the motion trajectory of the target following object based on the real-time position information, thereby adjusting the following strategy and providing more intelligent services. The motion trajectory refers to the specific presentation in space of the path that the target following object moves through within a certain period of time.

[0107] In an optional embodiment of the present invention, based on the real-time position information, the motion trajectory of the target following object is identified using the following formula:

[0108]

[0109] Among them, Q(t) represents the motion trajectory of the target following the object, Q(t i ) indicates that the target follows the object at t i The target position at time t i represents the i-th time point, t represents any moment in the process of target following the object, and t can be taken i+1 and t i Any value between, such as 2, t i+1 Represents the time point t i Adjacent time points, t i In front, t i+1 Afterwards, Q(t i+1 ) indicates that the target follows the object at t i+1 Target location at the moment.

[0110] By determining the target position of the target following object in the global environment map based on the real-time position information, the embodiment of the present invention can help the wheelchair plan the optimal path from its current position to the target position based on information such as the roads and obstacle distribution in the global environment map. The target position refers to the specific spatial coordinate position of the target following object within the reference frame of the global environment map.

[0111] As an embodiment of the present invention, determining the target position of the target following object in the global environment map based on the real-time position information includes: generating sensor coordinates of the target following object based on the real-time position information; mapping the real-time position information to the map coordinate system of the global environment map according to the sensor coordinates to obtain the initial target position of the target following object; extracting the map environment features of the target following object based on the initial target position; identifying the actual surrounding environment features of the target following object according to the sensor coordinates; calculating the feature matching degree between the actual surrounding environment features and the map environment features; performing position calibration processing of the target following object according to the feature matching degree to obtain a position calibration result; and determining the target position of the target following object in the global environment map based on the position calibration result.

[0112] The sensor coordinates refer to the coordinate representation of the position of the target following object in the sensor's own coordinate system. For example, a lidar typically establishes a three-dimensional coordinate system with its own installation location as the origin. The target position data measured by the sensor is the coordinate value in this specific coordinate system. The map coordinate system refers to the coordinate system used by the global environment map. The initial target position refers to the position obtained by mapping the real-time position information of the target following object to the map coordinate system of the global environment map through operations such as coordinate conversion based on the sensor coordinates. The map environment features refer to relevant feature information extracted from the global environment map based on the initial target position, such as the topographic features of the area where the target following object is located on the map, the distribution and shape features of buildings, and the layout features of the road network. The actual surrounding environment features refer to the environmental features of the actual surroundings of the target following object identified based on the sensor coordinates. The feature matching degree refers to an indicator that measures the degree of similarity between the actual surrounding environment features and the map environment features. The position calibration process refers to the process of adjusting and correcting the initial target position of the target following object based on the feature matching degree, such as calculating an offset based on feature differences and performing transformations such as translation and rotation on the initial position.

[0113] Optionally, based on the real-time position information, the sensor coordinates of the target following object can be generated by trigonometric calculation, such as installing a laser radar on the target following object, establishing a coordinate system with the position of the laser radar as the origin, and calculating the coordinates of the target point in the coordinate system according to the emission angle of the laser beam and the measured distance by mathematical methods such as trigonometric functions. According to the sensor coordinates, mapping the real-time position information to the map coordinate system of the global environment map can be achieved by affine transformation. According to the feature matching degree, the position calibration processing of the target following object can be performed by a particle filtering algorithm, and the feature matching degree between the actual surrounding environment features and the map environment features can be calculated by Euclidean distance.

[0114] Furthermore, the embodiment of the present invention can optimize the path followed by the wheelchair, avoid collision accidents, and ensure the personal safety of the user by identifying dynamic obstacles from the wheelchair to the target position based on the global environment map. The dynamic obstacles refer to objects that can change their position and posture autonomously or due to external forces, and obstruct or interfere with the wheelchair's travel path from the current position to the target position, such as vehicles and pedestrians.

[0115] As an embodiment of the present invention, identifying dynamic obstacles from the wheelchair to the target position based on the global environment map includes: setting an initial path from the wheelchair to the target position based on the global environment map; collecting historical point cloud data of the initial path and extracting current frame data corresponding to the historical point cloud data; simulating a background model of the initial path based on the historical point cloud data; identifying a motion area of ​​the initial path based on the current frame data and the background model; extracting moving objects in the motion area and analyzing the motion change trend of the moving objects; performing spatial position matching processing on the motion change trend and the initial path to obtain a spatial position matching result; analyzing the degree of influence of the moving object on the initial path based on the spatial position matching result; and identifying dynamic obstacles from the wheelchair to the target position based on the degree of influence.

[0116] Among them, the initial path refers to a preset route from the current position to the target position planned for the wheelchair based on the global environment map; the historical point cloud data refers to the point cloud data of the scene around the initial path continuously collected over a period of time in the past; the current frame data refers to the point cloud data corresponding to the current moment in the continuously collected historical point cloud data; the background model refers to a mathematical model of the static environment around the initial path constructed based on the historical point cloud data; the motion area refers to an area that is different from the background model after comparing and analyzing the current frame data with the background model; the moving object refers to an object with motion characteristics detected in the motion area. These objects can be pedestrians, other moving vehicles or movable items, etc. The motion change trend refers to the law of change of the motion state of the moving object over time, such as the direction and speed of movement. The spatial position matching processing refers to the process of spatially comparing and matching the motion change trend of the moving object with the initial path. The degree of influence refers to a quantitative description of the degree of interference or obstruction of the moving object to the initial path based on the result of the spatial position matching processing. For example, if a moving object moves toward the initial path at a higher speed and is closer to the initial path, then its degree of influence on the initial path is greater.

[0117] Optionally, based on the global environment map, the initial path of the wheelchair to the target position can be set by a path planning algorithm, the current frame data corresponding to the historical point cloud data can be extracted using the Open3D library, the background model of the initial path based on the historical point cloud data can be simulated by a triangulated meshing method, the moving objects in the moving area can be extracted using a target detection algorithm, the motion change trend and the spatial position matching of the initial path can be performed by a Monte Carlo simulation algorithm, and based on the spatial position matching result, the degree of influence of the moving object on the initial path can be analyzed using the influence formula of distance and speed, such as Where I represents the degree of influence, v represents the speed of the moving object, d represents the distance between the moving object and the initial path, and k represents a constant used to avoid the situation where the denominator is zero. This formula shows that the greater the speed and the closer the distance to the path, the greater the degree of influence of the moving object on the initial path.

[0118] S3. According to the dynamic obstacle, set a dynamic obstacle avoidance path from the wheelchair to the target position; based on the motion trajectory, analyze the movement change trend of the target following object; and identify the optimal following path from the wheelchair to the target following object according to the movement change trend and the dynamic obstacle avoidance path.

[0119] The embodiment of the present invention ensures that the wheelchair can adapt to various complex and changing environments by setting a dynamic obstacle avoidance path from the wheelchair to the target position based on the dynamic obstacles, reduces the risk of collision between the wheelchair and obstacles, and protects the personal safety of the user during travel. The dynamic obstacle avoidance path refers to a safe and feasible driving path that can be planned and adjusted in real time according to the real-time position and movement status of dynamic obstacles in the environment during the process of the wheelchair driving from the current position to the target position.

[0120] As an embodiment of the present invention, setting a dynamic obstacle avoidance path for the wheelchair to the target position based on the dynamic obstacle includes: measuring a speed parameter of the dynamic obstacle and identifying obstacle characteristics of the dynamic obstacle; setting a basic safety distance between the dynamic obstacle and the wheelchair based on the obstacle characteristics; calculating a dynamic safety distance between the wheelchair and the dynamic obstacle based on the speed parameter and the basic safety distance; constructing an obstacle avoidance zone for the wheelchair based on the dynamic safety distance and the dynamic obstacle; measuring kinematic constraint parameters of the wheelchair and setting a movable area of ​​the wheelchair based on the kinematic constraint parameters; defining an obstacle avoidance path constraint space for the wheelchair to the target position in combination with the obstacle avoidance zone and the movable area; determining a feasible path for the wheelchair to the target position based on the obstacle avoidance path constraint space; and setting a dynamic obstacle avoidance path for the wheelchair to the target position based on the obstacle avoidance path constraint space and the feasible path.

[0121] Among them, the speed parameter refers to the relevant physical quantity that describes the movement speed of a dynamic obstacle, including linear velocity and angular velocity. The obstacle characteristics refer to the physical properties and behavioral characteristics of the dynamic obstacle, such as size, shape, material, surface characteristics and movement state (such as whether it is dynamic, movement direction, etc.). The basic safety distance refers to the minimum safety distance that should be maintained between the wheelchair and the dynamic obstacle, determined according to the obstacle characteristics. For example, for a stationary, large obstacle, the basic safety distance will be set larger. The dynamic safety distance refers to the safety distance that should be maintained between the wheelchair and the obstacle, which is obtained by adjusting the basic safety distance in combination with the speed parameter of the dynamic obstacle. The obstacle avoidance zone refers to the area with the dynamic obstacle as the center. The center is an area defined by a dynamic safety distance as a radius. The kinematic constraint parameters refer to the restriction parameters related to the wheelchair movement, including the maximum speed, maximum acceleration, maximum deceleration, minimum turning radius, steering angle range, etc. of the wheelchair. The movable area refers to the spatial range in which the wheelchair can move freely in the environment, determined according to the kinematic constraint parameters of the wheelchair. For example, due to the minimum turning radius limitation of the wheelchair, the movable area will be restricted in some narrow bends. The obstacle avoidance path constraint space refers to the safe space for the wheelchair to follow, which is jointly defined by the obstacle avoidance zone and the movable area. The feasible path refers to a path from the current position of the wheelchair to the target position within the obstacle avoidance path constraint space that meets all constraint conditions.

[0122] Optionally, the speed parameters of the dynamic obstacle can be measured using a millimeter-wave radar, and the obstacle characteristics of the dynamic obstacle can be identified by a machine learning model, such as a decision tree model. Based on the dynamic safety distance and the dynamic obstacle, the obstacle avoidance area of ​​the wheelchair can be constructed using a rapid search random tree (RRT) algorithm. For example, with the dynamic obstacle as the boundary condition, during the search tree expansion process, the branches that intersect with the obstacle are discarded, and only the parts that do not intersect with the obstacle are retained. As the search tree continues to grow, an area containing a feasible path for the wheelchair is formed. The kinematic constraint parameters of the wheelchair can be measured by a sensor, such as an accelerometer. Based on the obstacle avoidance path constraint space, the feasible path from the wheelchair to the target position is determined using the Dijkstra algorithm.

[0123] In an optional embodiment of the present invention, the dynamic safety distance between the wheelchair and the dynamic obstacle is calculated based on the speed parameter and the basic safety distance using the following formula:

[0124]

[0125] Among them, D represents the dynamic safety distance of the dynamic obstacle, D0 represents the basic safety distance between the dynamic obstacle and the wheelchair, T average It represents the average reaction time of the wheelchair user, which is usually 0.2-0.5 seconds. V represents the speed parameter of the dynamic obstacle. A represents the maximum braking acceleration of the wheelchair, which can be 3m / s. 2 .

[0126] Furthermore, the embodiments of the present invention can not only improve the accuracy and reliability of wheelchair following but also enhance the user's sense of control over the wheelchair by analyzing the movement change trend of the target following object based on the motion trajectory. The movement change trend refers to the behavior pattern and change direction exhibited by the target following object during the movement.

[0127] Optionally, based on the motion trajectory, the movement change trend of the target following object can be analyzed using a Kalman filter algorithm. For example, in an indoor environment, a wheelchair receives position information sent by a positioning device installed on the target. The Kalman filter algorithm can be used to remove noise interference in the signal and accurately determine whether the target is walking at a constant speed, accelerating or decelerating, as well as the angle and speed of the turn, etc.

[0128] By identifying the optimal following path from the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path, the embodiment of the present invention can enable the wheelchair to follow the target in the fastest and most efficient manner, reduce unnecessary detours and waiting, and save time and energy consumption. The optimal following path refers to a path that enables the wheelchair to reach the target position in the best manner.

[0129] As an embodiment of the present invention, identifying the optimal following path from the wheelchair to the target following object based on the action change trend and the dynamic obstacle avoidance path includes: analyzing the usage scenario and user needs of the wheelchair based on the action change trend and the dynamic obstacle avoidance path; setting the following task priority of the wheelchair based on the usage scenario and user needs; constructing a task priority matrix of the wheelchair based on the following task priority; calculating the path score from the wheelchair to the target following object based on the task priority matrix; and identifying the optimal following path from the wheelchair to the target following object based on the path score.

[0130] The usage scenario refers to the actual operating environment of the wheelchair, including but not limited to factors such as space size, obstacle distribution, and personnel density. For example, when driving in a hospital corridor, the user needs refer to the various expectations and requirements of the wheelchair user for the wheelchair during use, including safety, comfort, convenience, and following accuracy. For example, elderly users place greater emphasis on safety and comfort, and hope that the wheelchair has smooth driving and reliable obstacle avoidance functions. The following task priority refers to the ranking of the various tasks involved in the wheelchair following the target based on the usage scenario and user needs. For example, in a shopping mall with high traffic, the obstacle avoidance task that ensures user safety may be given the highest priority, followed by maintaining an appropriate distance from the target to facilitate user observation, and finally considering the following speed to avoid safety issues caused by excessive speed. The task priority matrix refers to a mathematical matrix that quantifies and systematically represents the priority of following tasks. The path score refers to a numerical value obtained by evaluating each path from the current position of the wheelchair to the target following object based on the task priority matrix.

[0131] Optionally, based on the action change trend and the dynamic obstacle avoidance path, the usage scenario of the wheelchair can be analyzed by a clustering algorithm, and the user needs of the wheelchair can be analyzed using a naive Bayes algorithm. Based on the follow-up task priority, the task priority matrix of the wheelchair is constructed using the hierarchical analysis method. For example, for task i and task j, if task i is much more important than task j, it can be assigned a value of 9; if they are equally important, they can be assigned a value of 1; if they are slightly more important, they can be assigned a value of 3, and so on. The value follows the 1-9 scaling method.

[0132] In an optional embodiment of the present invention, based on the task priority matrix, the path score from the wheelchair to the target following object is calculated using the following formula:

[0133]

[0134] Among them, S represents the path score from the wheelchair to the target following object, wj represents the weight of the jth follow-up task in the task priority matrix, which can be 0.2, s represents the path position, and f j (s) represents the execution function of the j-th follower task with respect to the path position s, and its value range is [0,1]. ds represents the integral variable, j represents the follower task index in the task priority matrix, and n represents the total number of follower tasks in the task priority matrix.

[0135] S4. Based on the optimal following path, set an automated control instruction for the wheelchair, monitor the motion change state of the wheelchair in real time, and create a following path control unit for the wheelchair to the target following object by combining the motion change state and the automated control instruction.

[0136] By setting automated control instructions for the wheelchair based on the optimal following path, the embodiment of the present invention can enable the wheelchair to better cope with complex environmental changes, improve the safety and stability of wheelchair travel, and enhance the user's mobility autonomy and psychological security. The automated control instructions refer to a series of commands or signals, such as speed control signals, used to control the movement and behavior of the wheelchair.

[0137] As an embodiment of the present invention, the setting of the automatic control instructions of the wheelchair based on the optimal following path includes: determining the driving direction and driving speed of the wheelchair based on the optimal following path; identifying the user of the wheelchair and the driving surface, and analyzing the ground material of the driving surface; collecting arm muscle signals of the user, and identifying the user's movement intention based on the arm muscle signals; analyzing the driving difficulty of the wheelchair based on the ground material; identifying the distance between the wheelchair and obstacles in the optimal following path, and calculating the object moving speed of the optimal following path; determining the environmental safety level of the optimal following path based on the obstacle distance and the object moving speed; generating a decision control vector for the wheelchair in combination with the movement intention, the driving difficulty and the environmental safety level; setting a parameter adaptive adjustment mechanism for the driving direction and the driving speed based on the decision control vector; and setting the automatic control instructions of the wheelchair in combination with the decision control vector and the parameter adaptive adjustment mechanism.

[0138] Among them, the driving direction refers to the direction the wheelchair is facing during the movement, the driving speed refers to the distance the wheelchair moves per unit time, the user refers to the person who operates and uses the wheelchair, the driving surface refers to the surface that the wheelchair contacts during the movement, which can be indoor floors, such as tile floors, wooden floors, etc., or various outdoor roads, such as cement roads and asphalt roads. The ground material refers to the type of material and its characteristics that constitute the driving surface. For example, tile floors are smooth and have low friction; carpets are soft and have high friction and will increase the resistance of the wheelchair to movement. The arm muscle signal refers to the bioelectric signal generated by the user's arm muscles during contraction and relaxation. The movement intention refers to the user's desire for the wheelchair to perform a certain movement expressed by the arm muscle movement, etc. For example, the user may want the wheelchair to accelerate, decelerate, turn or Stop. The driving difficulty level refers to the degree of difficulty encountered by the wheelchair when driving on a specific driving surface. For example, the driving difficulty level is relatively low on a flat hard surface, while the driving difficulty level is relatively high on soft sand or uneven roads. The obstacle distance refers to the real-time Euclidean distance between the wheelchair and the nearest dynamic / static obstacle on the optimal following path, which is calculated by the obstacle distance detected by the sensor and the wheelchair's own size radius. The object movement speed refers to the radial speed of the obstacle relative to the wheelchair, which is calculated by tracking multiple frames of sensor data. The environmental safety level refers to the normalized risk index of the comprehensive obstacle distance and movement speed, ranging from [0,1], where 0 = absolutely safe and 1 = extremely dangerous. The parameter adaptive adjustment mechanism refers to an algorithm or strategy that automatically adjusts parameters such as the wheelchair's driving direction and speed based on factors such as movement intention and driving difficulty level.

[0139] Optionally, based on the optimal following path, the driving direction and speed of the wheelchair can be determined by a kinematic model of the wheelchair. For example, according to the requirements of the optimal following path, the kinematic model can be used to calculate the rotational speed and steering angle of each wheel, thereby achieving precise control of the driving direction and speed of the wheelchair. The ground material of the driving surface can be analyzed using image processing and computer vision technology. The arm muscle signals of the user can be collected by a surface electromyography sensor. Based on the arm muscle signals, the movement intention of the user can be identified using a template matching method. For example, muscle signals of multiple standard movement actions are first collected as templates, and features are extracted and stored for these templates. The same feature extraction is then performed on the muscle signals to be identified, and then similarity is calculated with templates in a template library. The movement intention is judged based on the template with the highest similarity. Based on the ground material, the driving difficulty of the wheelchair can be analyzed by a fuzzy logic algorithm. According to the movement intention and the driving difficulty, the adaptive adjustment mechanism of the parameters of the driving direction and the driving speed can be set using an adaptive control algorithm, such as a model predictive control algorithm.

[0140] Furthermore, the embodiments of the present invention can help the wheelchair continuously adjust its route according to the actual motion state during travel by real-time monitoring of the motion change state of the wheelchair, guiding the user to reach the destination accurately and improving travel efficiency. The motion change state refers to the situation in which the motion state of the wheelchair changes over time during operation, such as the position gradually deviating from the safe distance.

[0141] Optionally, the movement change state of the wheelchair can be monitored in real time by a computer vision algorithm, for example, by identifying changes in lane lines or tile patterns on the road surface to determine whether the wheelchair deviates from a straight line, and estimating the wheelchair's travel speed based on the movement speed of objects in the image.

[0142] The embodiment of the present invention combines the motion change state and the automatic control instruction to create a following path control unit for the wheelchair to follow the target following object. It can not only respond immediately according to the current real-time situation, but also predict future motion trends and make corresponding plans in advance to improve the overall following effect and user experience. The following path control unit refers to a system or module for controlling the wheelchair to follow the target following object.

[0143] As an embodiment of the present invention, the combination of the motion change state and the automatic control instruction to create a following path control unit for the wheelchair to the target following object includes: analyzing the environmental adaptability of the wheelchair according to the motion change state; extracting the safety protection instructions in the automatic control instruction based on the environmental adaptability; setting the boundary detection mechanism of the wheelchair according to the safety protection instructions; creating an abnormal feedback system for the motion change state based on the boundary detection mechanism; analyzing the abnormal cause of the motion change state according to the abnormal feedback system; identifying the object attributes of the target following object based on the abnormal cause; constructing a multi-target tracking network for the wheelchair according to the object attributes; and creating a following path control unit for the wheelchair to the target following object in combination with the boundary detection mechanism, the abnormal feedback system and the multi-target tracking network.

[0144] Among them, the environmental adaptability refers to the ability of the wheelchair to operate normally, adapt to environmental changes and complete tasks under different environmental conditions (such as different terrain, lighting, spatial layout, etc.); the safety protection instructions refer to the instructions in the automated control instructions to ensure the safe operation of the wheelchair, such as speed limit, maintaining a safe distance, emergency braking, turning on the warning device, etc.; the boundary detection mechanism refers to a mechanism for detecting whether the wheelchair has exceeded the safety boundary or entered a dangerous area; the abnormal feedback system refers to a system for collecting, analyzing and processing abnormal information in the wheelchair movement change state; the abnormal cause refers to various factors that cause the wheelchair movement change state to be abnormal, such as sudden Obstacles, slippery ground, etc. The object attributes refer to various characteristics and attribute information of the target following object, such as the identity of the object (such as a specific user, medical staff, etc.), physical characteristics (such as height, body shape, walking posture, etc.). The multi-target tracking network refers to a system for simultaneously tracking multiple target objects. It integrates multiple sensor data and uses computer vision, machine learning and other technologies to identify, locate and track the target objects, establish a motion model of the target objects, and predict their future position and motion state, thereby providing accurate target information for the wheelchair's following path control, enabling the wheelchair to accurately follow the target object in a complex environment while avoiding other obstacles and interference.

[0145] Optionally, based on the motion change state, the environmental adaptability of the wheelchair can be analyzed by a fuzzy logic algorithm. Based on the safety protection instructions, the boundary detection mechanism of the wheelchair can be set using boundary identification technology. For example, a Bluetooth beacon is set in a specific activity area, and the wheelchair judges its distance from the boundary based on the strength of the received Bluetooth signal, and triggers the safety mechanism when it approaches the boundary. Based on the boundary detection mechanism, the abnormal feedback system of the motion change state can be created by a threshold based on boundary detection. For example, when the wheelchair approaches the boundary, an upper limit threshold of the speed is set; when an obstacle is detected within a specific distance, an allowable range of the steering angle is set. Once the motion state data exceeds these thresholds, it is determined to be abnormal. Based on the abnormal feedback system, the abnormal cause of the motion change state can be analyzed using cluster analysis. Based on the object attributes, the multi-target tracking network of the wheelchair can be constructed by a YOLO target detection algorithm.

[0146] S5. Identify an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and set a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state.

[0147] The embodiment of the present invention identifies the error positioning factors of the wheelchair based on the motion change state and the dynamic obstacle, and can optimize the motion control of the wheelchair based on these factors, so that the wheelchair can follow the motion trajectory of the target object more smoothly and accurately, thereby improving the positioning accuracy of the wheelchair. The error positioning factor refers to the relevant factors used to quantify and describe the deviation between the actual position and the theoretical position of the wheelchair, such as environmental interference factors and sensor measurement errors.

[0148] As an embodiment of the present invention, the identifying the error positioning factor of the wheelchair based on the motion change state and the dynamic obstacle includes: extracting the motion change characteristics of the wheelchair under the dynamic obstacle based on the motion change state; identifying the error pattern of the wheelchair under the dynamic obstacle based on the motion change characteristics; performing cumulative effect analysis processing on the error pattern to obtain a cumulative law; analyzing the positioning deviation of the wheelchair based on the cumulative law; and identifying the error positioning factor of the wheelchair in combination with the error pattern and the positioning deviation.

[0149] The motion change characteristics refer to the characteristics of a wheelchair's motion state when it moves in an environment with dynamic obstacles, such as wheelchair acceleration. The error pattern refers to the typical manifestation or pattern of deviation in wheelchair motion identified based on the motion change characteristics. For example, each time a wheelchair encounters a fast-moving obstacle, its positioning will shift in a fixed direction. The cumulative effect analysis and processing refers to the process of studying and analyzing the cumulative impact of the error pattern over time or the development of the motion process. The cumulative pattern refers to the specific pattern or trend of error accumulation obtained through the cumulative effect analysis and processing. For example, the rate of error growth increases over time. The positioning deviation refers to the deviation between the actual and ideal wheelchair positioning analyzed based on the cumulative pattern, including information such as the size, direction, and change trend of the positioning deviation. For example, the laser radar may misjudge or miss a dynamic obstacle, resulting in inaccurate wheelchair positioning information, thereby generating positioning deviation. This deviation may continue to increase over time, or in some cases, the deviation may fluctuate within a certain range but will not continue to increase.

[0150] Optionally, according to the motion change state, the motion change characteristics of the wheelchair under the dynamic obstacle can be extracted by a supervised learning algorithm, and based on the motion change characteristics, the error pattern of the wheelchair under the dynamic obstacle can be identified using a hidden Markov model. For example, the error pattern includes several states such as "normal driving", "obstacle avoidance error", and "sensor failure error". The transition probability between these states and the observation probability of generating different motion change characteristics in each state can be estimated based on historical data to determine the error pattern. The cumulative effect analysis and processing of the error pattern can be performed by a time series analysis method. According to the cumulative law, the positioning deviation of the wheelchair can be analyzed using a positioning deviation model. For example, if it is found that the positioning deviation is linearly related to time, a linear regression model can be established; if there is a nonlinear relationship, polynomial regression, exponential model or other suitable nonlinear models can be tried to describe the changing law of positioning deviation.

[0151] Furthermore, the embodiment of the present invention sets a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state, and can perform targeted compensation for the error pattern and positioning deviation of the wheelchair in different scenarios, thereby more effectively improving the positioning accuracy, improving the robustness and reliability of the entire wheelchair following system, and reducing the risk of positioning errors and loss of control. The positioning error compensation mechanism refers to a system that detects and corrects the positioning deviation of the wheelchair in real time due to various factors (such as sensor errors, environmental interference, mechanical wear, etc.) during the movement of the wheelchair.

[0152] As an embodiment of the present invention, the positioning error compensation mechanism of the wheelchair is set based on the error positioning factor and the motion change state, including: extracting the sensor nodes of the wheelchair based on the error positioning factor; identifying the motion scene of the wheelchair according to the motion change state; analyzing the error positioning performance of the sensor nodes in the motion scene; determining the optimal sensor combination of the wheelchair based on the error positioning performance; monitoring the target loss situation of the wheelchair in real time according to the optimal sensor combination; setting the target repositioning mechanism of the wheelchair based on the target loss situation and the optimal sensor combination; and setting the positioning error compensation mechanism of the wheelchair in combination with the error positioning performance, the optimal sensor combination and the target repositioning mechanism.

[0153] Among them, the sensor node refers to various equipment units with perception and data acquisition functions installed on the wheelchair, such as acceleration sensors and gyroscopes. The motion scene refers to a collection of various different environments and motion states in which the wheelchair is located during operation. For example, the wheelchair may be in different scenarios such as indoor flat ground, outdoor rugged road, turning, accelerating, decelerating, approaching obstacles, and moving away from obstacles. The error positioning performance refers to the deviation between the measured data and the true value of each sensor node in different motion scenes and the performance characteristics of the wheelchair positioning error caused by this. For example, in a turning scene, the gyroscope may produce a certain measurement error due to changes in angular velocity, resulting in deviations in the wheelchair posture estimation. The optimal sensor combination refers to a set of sensors that can comprehensively achieve the most accurate wheelchair positioning in various motion scenes. This combination is not fixed and may be dynamically adjusted according to different motion scenes and positioning requirements. For example, in an indoor environment, the combination of lidar and camera may provide more accurate positioning information; while in an open outdoor environment A combination of a global positioning system (GPS) with an accelerometer and a gyroscope may be more suitable. The target loss situation refers to a situation in which the wheelchair loses track of the intended target or cannot accurately determine the relationship between itself and the target due to various reasons while the wheelchair is moving in accordance with the target movement state. For example, the target may not be accurately sensed by the wheelchair due to reasons such as being beyond the sensor's detection range, being blocked by obstacles, sensor failure, or environmental interference, thereby preventing the wheelchair from continuing to move toward the target as originally planned. The target relocalization mechanism refers to a set of methods and strategies for re-determining the positional relationship between the wheelchair and the target and enabling the wheelchair to continue moving toward the target after the wheelchair loses target. For example, when the wheelchair is following a pedestrian and the pedestrian suddenly enters an area not covered by the map, resulting in target loss, the target relocalization mechanism may first use a camera to search for the pedestrian's features in the surrounding environment, and at the same time combine the lidar to detect the distribution of surrounding obstacles to infer the pedestrian's possible location, and then control the wheelchair to move in the inferred direction to re-establish tracking of the target.

[0154] Optionally, the error positioning performance of the sensor node in the motion scenario can be analyzed through experimental testing, such as allowing the wheelchair to operate in different motion scenarios, such as linear acceleration, deceleration, turning, driving on different roads, etc., while recording the sensor data and the actual motion state of the wheelchair, comparing the sensor measurement value with the actual value, and calculating the error size and change pattern. Based on the error positioning performance, the optimal sensor combination of the wheelchair can be determined using the information entropy method, such as calculating the information entropy of the measurement data of each sensor in different motion scenarios based on the error positioning performance. The lower the information entropy, the more certain and reliable the information provided by the sensor. Based on the target loss situation and the optimal sensor combination, the target relocation mechanism of the wheelchair can be set by a relocation algorithm based on visual feature matching, such as using a visual sensor to obtain image information of the environment around the wheelchair, extracting feature points or feature areas in the image, and then matching them with the features in a pre-stored environmental map or database, and determining the position of the wheelchair in the environment through the matched feature points.

[0155] S6. In combination with the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism, the wheelchair performs precise positioning processing on the target following object to obtain a precise positioning result.

[0156] The embodiments of the present invention combine the global environment map, the optimal following path, the following path control unit, and the positioning error compensation mechanism to perform precise positioning processing of the wheelchair relative to the target following object, obtaining a precise positioning result. This can accurately identify the position of the target following object, effectively avoid the influence of lighting changes, greatly improve positioning accuracy, and comprehensively enhance the applicability and reliability of the wheelchair following function in complex and diverse scenarios, providing users with a more stable and accurate following experience. The precise positioning processing refers to a series of operations that comprehensively utilize multiple elements such as the global environment map, the optimal following path, the following path control unit, and the positioning error compensation mechanism to accurately measure and adjust the positional relationship between the wheelchair and the target following object. The precise positioning result refers to information about the accurate positional relationship between the wheelchair and the target following object obtained after a series of precise positioning processing operations. For example, the precise positioning result may show that the target following object is 3 meters in front of the wheelchair and 5 degrees to the left, and the distance error between the wheelchair and the target object is within ±0.1 meter.

[0157] Compared with the problems described in the background technology, the embodiment of the present invention constructs a global environment map of the target following object based on the environmental data, which can help the wheelchair clearly know the relative position relationship between itself and the target following object and other objects, and plan the following path; further, the embodiment of the present invention identifies the dynamic obstacles from the wheelchair to the target position according to the global environment map, which can optimize the wheelchair following path, avoid collision accidents, and ensure the personal safety of the user; the embodiment of the present invention sets a dynamic obstacle avoidance path from the wheelchair to the target position according to the dynamic obstacles, which can ensure that the wheelchair can adapt to various complex and changing environments, reduce the occurrence of wheelchairs and obstacles. The risk of collision is reduced, thereby ensuring the personal safety of the user during travel; further, the embodiment of the present invention identifies the optimal following path from the wheelchair to the target following object according to the action change trend and the dynamic obstacle avoidance path, so that the wheelchair can follow the target in the fastest and most efficient way, reduce unnecessary detours and waiting, and save time and energy consumption; the embodiment of the present invention sets the automatic control instructions of the wheelchair based on the optimal following path, so that the wheelchair can better cope with complex environmental changes, improve the safety and stability of wheelchair travel, and enhance the user's mobility autonomy and psychological sense of security; further, the embodiment of the present invention combines the motion change state and the automatic control instructions, Creating a following path control unit for the wheelchair to the target following object can not only make an immediate response according to the current real-time situation, but also predict future movement trends, make corresponding plans in advance, and improve the overall following effect and user experience; the embodiment of the present invention identifies the error positioning factor of the wheelchair according to the movement change state and the dynamic obstacle, and can optimize the movement control of the wheelchair according to these factors, so that the wheelchair can follow the movement trajectory of the target object more smoothly and accurately, thereby improving the positioning accuracy of the wheelchair; further, the embodiment of the present invention sets the positioning error compensation mechanism of the wheelchair based on the error positioning factor and the movement change state, which can be used for the wheelchair Targeted compensation is performed based on the error patterns and positioning deviations in different scenarios, thereby more effectively improving positioning accuracy, improving the robustness and reliability of the entire wheelchair following system, and reducing the risks of positioning errors and loss of control; finally, the embodiment of the present invention combines the global environment map, the optimal following path, the following path control unit, and the positioning error compensation mechanism to perform precise positioning processing of the wheelchair on the target following object, obtain a precise positioning result, accurately identify the position of the target following object, effectively avoid the influence of lighting changes, greatly improve positioning accuracy, and comprehensively enhance the applicability and reliability of the wheelchair following function in complex and diverse scenarios, bringing users a more stable and accurate following experience.Therefore, the precise positioning method and system for wheelchair following provided by the embodiments of the present invention can achieve precise positioning of wheelchair following, improve the applicability of the wheelchair following function, allow users to travel more safely and independently, and enhance users' life autonomy.

[0158] Example 2:

[0159] like Figure 2 FIG. 1 is a functional module diagram of a precise positioning system for wheelchair following according to the present invention.

[0160] The precise positioning system 200 for wheelchair following described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the precise positioning system for wheelchair following can include a target environment recognition module 201, an obstacle recognition module 202, a following path setting module 203, a following control module 204, an error compensation module 205, and a positioning execution module 206. A module, also referred to as a unit, is a series of computer program segments that can be executed by an electronic device processor and perform a fixed function, and is stored in the electronic device's memory.

[0161] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0162] The target environment recognition module 201 is used to obtain a target following object of the wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data;

[0163] The obstacle identification module 202 is used to track the real-time position information of the target following object, identify the motion trajectory of the target following object based on the real-time position information, determine the target position of the target following object in the global environment map, and identify dynamic obstacles from the wheelchair to the target position based on the global environment map.

[0164] The following path setting module 203 is used to set a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle, analyze the movement change trend of the target following object based on the motion trajectory, and identify the optimal following path for the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path;

[0165] The following control module 204 is configured to set an automated control instruction for the wheelchair based on the optimal following path, monitor the motion change state of the wheelchair in real time, and create a following path control unit for the wheelchair to follow the target object by combining the motion change state and the automated control instruction;

[0166] The error compensation module 205 is configured to identify an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and set a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state;

[0167] The positioning execution module 206 is used to perform precise positioning processing of the wheelchair on the target following object in combination with the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism to obtain a precise positioning result.

[0168] In detail, the modules in the precise positioning system 200 for wheelchair following according to the embodiment of the present invention are used in the same manner as above. Figure 1 The same technical means as the precise positioning method for wheelchair following described in and can produce the same technical effects are not described here.

[0169] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0170] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A precise positioning method for wheelchair following, characterized in that: The method comprises: Acquire a target following object of the wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data; Tracking real-time position information of the target following object, identifying a motion trajectory of the target following object based on the real-time position information, and determining a target position of the target following object in the global environment map, and identifying dynamic obstacles between the wheelchair and the target position based on the global environment map; setting a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle, analyzing a movement change trend of the target following object based on the motion trajectory, and identifying an optimal following path for the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path; Based on the optimal following path, setting an automated control instruction for the wheelchair, monitoring the motion change state of the wheelchair in real time, and creating a following path control unit for the wheelchair to follow the target object by combining the motion change state and the automated control instruction; identifying an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and setting a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state; In combination with the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism, the wheelchair performs precise positioning processing on the target following object to obtain a precise positioning result.

2. A precise positioning method for wheelchair following according to claim 1, characterized in that: The determining, based on the real-time position information, a target position of the target following object in the global environment map includes: generating sensor coordinates of the target following object based on the real-time position information; Mapping the real-time position information into a map coordinate system of the global environment map according to the sensor coordinates to obtain an initial target position of the target following object; Extracting map environment features of the target following object based on the initial target position; Identifying actual surrounding environment features of the target following object according to the sensor coordinates; Calculating a feature matching degree between the actual surrounding environment feature and the map environment feature; According to the feature matching degree, performing position calibration processing of the target following object to obtain a position calibration result; Based on the position calibration result, a target position of the target following object in the global environment map is determined.

3. The precise positioning method for wheelchair following according to claim 1, characterized in that: The step of identifying dynamic obstacles between the wheelchair and the target location based on the global environment map includes: setting an initial path for the wheelchair to the target location according to the global environment map; Collecting historical point cloud data of the initial path and extracting current frame data corresponding to the historical point cloud data; Simulating a background model of the initial path based on the historical point cloud data; identifying a motion region of the initial path according to the current frame data and the background model; Extracting moving objects in the moving area and analyzing the movement change trend of the moving objects; Performing spatial position matching processing on the motion change trend and the initial path to obtain a spatial position matching result; Analyzing the degree of influence of the moving object on the initial path based on the spatial position matching result; According to the impact degree, dynamic obstacles between the wheelchair and the target position are identified.

4. The precise positioning method for wheelchair following according to claim 1, characterized in that: The step of setting a dynamic obstacle avoidance path from the wheelchair to the target position according to the dynamic obstacle comprises: measuring a velocity parameter of the dynamic obstacle and identifying obstacle characteristics of the dynamic obstacle; According to the characteristics of the obstacle, a basic safety distance between the dynamic obstacle and the wheelchair is set; Based on the speed parameter and the basic safety distance, the dynamic safety distance between the wheelchair and the dynamic obstacle is calculated using the following formula: Among them, D represents the dynamic safety distance of the dynamic obstacle, D0 represents the basic safety distance between the dynamic obstacle and the wheelchair, T average It represents the average reaction time of the wheelchair user, which is usually 0.2-0.5 seconds. V represents the speed parameter of the dynamic obstacle. A represents the maximum braking acceleration of the wheelchair, which can be 3m / s. 2 ; Constructing an obstacle avoidance zone for the wheelchair according to the dynamic safety distance and the dynamic obstacle; measuring kinematic constraint parameters of the wheelchair, and setting a movable area of ​​the wheelchair based on the kinematic constraint parameters; Defining an obstacle avoidance path constraint space from the wheelchair to the target position by combining the obstacle avoidance zone and the movable area; determining a feasible path from the wheelchair to the target location according to the obstacle avoidance path constraint space; A dynamic obstacle avoidance path from the wheelchair to the target position is set based on the obstacle avoidance path constraint space and the feasible path.

5. The precise positioning method for wheelchair following according to claim 1, characterized in that: The identifying, based on the movement change trend and the dynamic obstacle avoidance path, an optimal following path from the wheelchair to the target following object includes: Analyzing usage scenarios and user needs of the wheelchair based on the movement change trend and the dynamic obstacle avoidance path; Setting the following task priority of the wheelchair based on the usage scenario and user needs; Constructing a task priority matrix of the wheelchair according to the following task priority; Based on the task priority matrix, the path score from the wheelchair to the target following object is calculated using the following formula: Among them, S represents the path score from the wheelchair to the target following object, w j represents the weight of the jth follow-up task in the task priority matrix, which can be 0.2, s represents the path position, and f j (s) represents the execution function of the jth follower task with respect to the path position s, with a value range of [0,1], ds represents the integral variable, j represents the follower task index in the task priority matrix, and n represents the total number of follower tasks in the task priority matrix; An optimal following path from the wheelchair to the target following object is identified according to the path score.

6. The precise positioning method for wheelchair following according to claim 1, characterized in that: The step of setting an automatic control instruction for the wheelchair based on the optimal following path includes: determining a driving direction and a driving speed of the wheelchair based on the optimal following path; Identifying the user of the wheelchair and the driving surface, and analyzing the ground material of the driving surface; collecting arm muscle signals of the user, and identifying the user's movement intention based on the arm muscle signals; Analyzing the difficulty of the wheelchair in moving based on the ground material; Identifying the distance between the wheelchair and obstacles in the optimal following path, and calculating the moving speed of the object in the optimal following path; determining an environmental safety level of the optimal following path according to the obstacle distance and the moving speed of the object; generating a decision control vector for the wheelchair based on the movement intention, the driving difficulty level, and the environmental safety level; Setting a parameter adaptive adjustment mechanism of the driving direction and the driving speed according to the decision control vector; In combination with the decision control vector and the parameter adaptive adjustment mechanism, an automated control instruction for the wheelchair is set.

7. The precise positioning method for wheelchair following according to claim 1, characterized in that: The control unit for creating a following path for the wheelchair to follow the target object by combining the motion change state and the automation control instruction includes: Analyzing the environmental adaptability of the wheelchair according to the motion change state; extracting safety protection instructions from the automation control instructions based on the environmental adaptability; Setting a boundary detection mechanism for the wheelchair according to the safety protection instruction; Based on the boundary detection mechanism, an abnormal feedback system of the motion change state is established; Analyzing the abnormal cause of the motion change state according to the abnormal feedback system; Based on the abnormal cause, identifying the object attribute of the target following object; Constructing a multi-target tracking network for the wheelchair according to the object attributes; Combining the boundary detection mechanism, the abnormal feedback system and the multi-target tracking network, a control unit for the wheelchair to follow the target object is created.

8. The precise positioning method for wheelchair following according to claim 1, characterized in that: The identifying the error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle includes: extracting, according to the motion change state, motion change characteristics of the wheelchair under the dynamic obstacle; identifying an error pattern of the wheelchair under the dynamic obstacle based on the motion change characteristics; Performing cumulative effect analysis and processing of the error pattern to obtain a cumulative law; Analyzing the positioning deviation of the wheelchair according to the accumulation rule; The error positioning factor of the wheelchair is identified by combining the error pattern and the positioning deviation.

9. The precise positioning method for wheelchair following according to claim 1, characterized in that: The setting of the positioning error compensation mechanism of the wheelchair based on the error positioning factor and the motion change state includes: extracting sensor nodes of the wheelchair based on the error location factor; identifying a movement scene of the wheelchair according to the movement change state; Analyzing the error positioning performance of the sensor node in the motion scenario; determining an optimal sensor combination for the wheelchair based on the error positioning performance; monitoring the target loss condition of the wheelchair in real time according to the optimal sensor combination; Setting a target repositioning mechanism for the wheelchair based on the target loss condition and the optimal sensor combination; In combination with the error positioning performance, the optimal sensor combination and the target repositioning mechanism, a positioning error compensation mechanism for the wheelchair is set.

10. A precise positioning system for wheelchair following, characterized in that: The system comprises: A target environment recognition module is used to obtain a target following object of the wheelchair, collect environmental data of the target following object, and construct a global environmental map of the target following object based on the environmental data; The obstacle recognition module is used to track the real-time position information of the target following object, identify the motion trajectory of the target following object based on the real-time position information, determine the target position of the target following object in the global environment map, and identify dynamic obstacles from the wheelchair to the target position based on the global environment map. a following path setting module, configured to set a dynamic obstacle avoidance path for the wheelchair to the target position according to the dynamic obstacle, analyze the movement change trend of the target following object based on the motion trajectory, and identify an optimal following path for the wheelchair to the target following object based on the movement change trend and the dynamic obstacle avoidance path; a following control module, configured to set an automated control instruction for the wheelchair based on the optimal following path, monitor the motion change state of the wheelchair in real time, and create a following path control unit for the wheelchair to follow the target following object by combining the motion change state and the automated control instruction; an error compensation module, configured to identify an error positioning factor of the wheelchair according to the motion change state and the dynamic obstacle, and set a positioning error compensation mechanism for the wheelchair based on the error positioning factor and the motion change state; The positioning execution module is used to combine the global environment map, the optimal following path, the following path control unit and the positioning error compensation mechanism to perform precise positioning processing of the wheelchair on the target following object to obtain a precise positioning result.

Images