Pull-type robot blind guiding system and method

Abstract

The invention discloses a pull-type robot blind guiding system and method, and relates to the technical field of mobile service robots, the pull-type robot blind guiding system comprises a robot body, the robot body comprises an environment sensing and recognition module and a path planning module, the environment sensing and recognition module is used for obtaining three-dimensional point cloud information of the surrounding environment, and the path planning module is used for planning the three-dimensional point cloud information of the surrounding environment; generating an environment composite map according to the three-dimensional point cloud information; the path planning module analyzes the environment composite map by introducing artificial potential field factors, and generates a safe and collision-free initial path for the blind; the robot body is physically connected with the blind person through the rigid guide rod. Multi-sensor coupling is introduced, and artificial potential field factors are added, so that the obstacle avoidance capability of the blind guiding robot in a complex environment is improved, safe walking of the blind in indoor and outdoor scenes is ensured, and the life autonomy of the blind is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile service robots, and in particular to a traction-type robot blind-guiding system and method. Background Art

[0002] Vision is one of the most important tools for people to perceive the surrounding environment. Visual impairment has brought many obstacles to people's lives. Nowadays, most visually impaired people use simple canes to assist in travel, but this simple cane can no longer meet the various needs of the blind. Biological guide methods represented by guide dogs are difficult to popularize due to high training costs, long training cycles and lifespans. With the accumulation and breakthroughs of technologies such as big data and deep learning, guide robots with excellent performance have become the optimal solution to solve the daily difficulties of visually impaired people.

[0003] The existing guide aids can generally be divided into three categories: cane guide aids, wearable guide aids and mobile walking aids. Cane guide aids often have the weakness of single function. Most smart guide canes only have the function of detecting obstacles and are less safe. In addition, guide canes do not have traction and route planning functions, and blind people can only use them in scenes or roads they are familiar with, which limits the range of activities of the blind. Wearable guide aids install sensors on the user's glasses or clothes, allowing the blind to walk like normal people with the entire system on their backs. However, due to the need to carry a large number of sensors and computers and other equipment, it will cause the blind to be burdened with too much equipment and too much information. In addition to interfering with the actions of the blind, the biggest drawback of wearable guide devices is that it is difficult to provide enough psychological security to the blind who walk "empty-handed". The functions that can be achieved by mobile guide robots completely cover other guide robots, and they are highly scalable. At the same time, the mobile guide robot adopts a navigation method in which the robot goes first and the blind person follows, which is similar to the guidance method of a real guide dog. It not only greatly reduces the probability of injury to the blind, significantly improves the efficiency and safety of the blind walking indoors and outdoors, but also provides sufficient sense of security on the psychological level.

[0004] Chinese patent document CN212187114U discloses "a mechanical guide dog based on machine vision". The mechanical guide dog detects the surrounding environment through a visual component installed on the main body, but has no traction device, path planning, voice interaction and other functions, and cannot be used for guiding the blind in complex environments.

[0005] Chinese patent document 113520812B discloses "a quadruped robot guide system and method". The robot guide system can plan the best path for the robot based on the perceived environmental information, but it cannot identify and avoid high-altitude obstacles, and only plans the best path for the robot during path planning, which makes it difficult to ensure that the route of the guided blind person is safe and obstacle-free.

[0006] Environmental perception and exploration of robots are the prerequisites for them to adapt to complex and unknown environments. Currently, robots' perception of the external environment is often concentrated in front of and on both sides of the robot, and they do not have the ability to identify and avoid obstacles behind and at high altitudes. In addition, current obstacle avoidance walking methods often focus on the robot's own obstacle avoidance, and do not consider unexpected situations that the towed blind person may encounter behind the robot. For example, when facing a narrow curve, the robot may successfully turn the corner but the towed blind person may hit the wall.

[0007] Therefore, technicians in this field are committed to developing a traction robot guide system and method, which can improve the obstacle avoidance ability of the guide robot in complex environments and ensure the safe walking of the blind in indoor and outdoor scenes. Summary of the invention

[0008] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is how to deal with dynamic obstacles in a complex environment and how to achieve coordinated movement between the blind and the robot.

[0009] To achieve the above-mentioned purpose, the present invention provides a traction robot blind guiding system, comprising a robot body, wherein the robot body comprises an environment perception and recognition module and a path planning module, wherein the environment perception and recognition module is used to obtain three-dimensional point cloud information of the surrounding environment, and generate an environment composite map according to the three-dimensional point cloud information; the path planning module analyzes the environment composite map by introducing an artificial potential field factor to generate a safe and collision-free initial path for the blind; the robot body is physically connected to the blind through a rigid guide rod.

[0010] Furthermore, the environment perception and recognition module obtains three-dimensional point cloud information of the surrounding environment through multi-sensor coupling; the sensors include three-dimensional laser radar and depth camera.

[0011] Furthermore, the three-dimensional laser radar is installed on the head of the robot body, and the depth camera is located at the front end of the three-dimensional laser radar.

[0012] Furthermore, the environment perception and recognition module performs semantic modeling according to the three-dimensional point cloud information to generate a global map of the environment; and marks dynamic obstacles in the global map of the environment to obtain the composite map of the environment.

[0013] Furthermore, the path planning module uses a greedy algorithm to optimize and smooth the turning points of the initial path, and adjusts the path in real time in combination with obstacle information to obtain a real-time path.

[0014] Furthermore, the path planning module can also generate a motion trajectory of the robot body by calculating tangents according to the real-time path, so as to ensure the coordinated movement of the blind person and the robot body.

[0015] Furthermore, one end of the rigid guide rod is hinged to the robot body via a turntable, and a six-dimensional force sensor is installed on the turntable.

[0016] Furthermore, the six-dimensional force sensor is used to monitor the tension and pressure signals on the rigid guide rod to adjust the travel speed of the robot body in real time.

[0017] Furthermore, it includes a voice interaction module, which is used to provide voice feedback, broadcast upcoming terrain features or obstacle information to the blind, and adjust the actions of the machine body according to the blind's voice instructions.

[0018] A traction robot blind-guiding method, the method comprising the following steps:

[0019] Acquire three-dimensional point cloud information of the surrounding environment through coupling of multiple sensors;

[0020] Generate a global map of the environment based on 3D point cloud data, mark dynamic obstacles, and form a composite map of the environment;

[0021] Analyzing the composite map of the environment through artificial potential field factors to generate an initial walking path for the blind person;

[0022] A greedy algorithm is used to optimize and smooth the turning points of the initial path, and the path is adjusted in real time based on obstacle information to obtain a real-time path.

[0023] The motion trajectory of the robot body is generated by calculating the tangent according to the real-time path, so as to ensure the coordinated motion of the blind person and the robot body.

[0024] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0025] 1. The present invention integrates multi-sensor information, performs environmental perception and modeling through 3D point cloud technology, and can identify dynamic and static obstacles in complex environments to ensure the safety of the blind's walking path. Compared with traditional guide devices, the present invention breaks through the detection limitations of high-altitude and rear obstacles, solves the problem of guiding the blind in complex scenes such as narrow curves, and significantly improves the efficiency and safety of guiding the blind;

[0026] 2. The present invention uses a six-dimensional force sensor to sense the pulling force or pressure applied by the blind, ensuring that the robot can adjust its movement according to the blind's stride and rhythm at any time, achieving a high-precision blind following and guiding function, and improving the comfort and safety of the blind guide;

[0027] 3. The present invention optimizes path planning by introducing artificial potential field factors, and combines the greedy algorithm to optimize the turning points and smooth the path. It can generate a safe and barrier-free walking path for the blind in real time and calculate the robot's motion trajectory; it greatly improves the robot's responsiveness in a dynamic environment, reduces the risk of collisions caused by sudden obstacles for the blind, and improves walking safety.

[0028] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a system structure diagram of a preferred embodiment of the present invention;

[0030] Figure 2 is a schematic diagram of path optimization of a preferred embodiment of the present invention;

[0031] Figure 3 It is a method framework diagram of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.

[0033] In the drawings, components with the same structure are indicated by the same numerical reference numerals, and components with similar structures or functions are indicated by similar numerical reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the components is appropriately exaggerated in some places in the drawings.

[0034] The present embodiment provides a traction-type robot blind guiding system, including a robot body 1, wherein the robot body includes an environment perception and recognition module, a path planning module, and a voice interaction module.

[0035] The environment perception and recognition module obtains the 3D point cloud information of the surrounding environment through multi-sensor coupling, and then performs semantic modeling based on the 3D point cloud information to generate a global map of the environment, mark dynamic obstacles, and finally obtain a composite map of the environment. The sensor includes a 3D laser radar 2 and a depth camera. The 3D laser radar 2 is installed on the head of the robot body 1, and the depth camera is located at the front end of the 3D laser radar 2.

[0036] The path planning module analyzes the composite map of the environment by introducing artificial potential field factors to generate a safe and collision-free initial path for the blind. Among them, the attraction field guides the blind person or robot to move towards the target. As the distance to the target approaches, the attraction gradually decreases to ensure a smooth path, which is calculated through quadratic optimization. The repulsive field is used to avoid obstacles. When approaching obstacles, the repulsive force increases sharply. Through gradient calculation, a collision-free path is generated. Finally, the combined force is calculated so that the path can quickly reach the target while safely avoiding obstacles. To adapt to complex environments, the module can ensure that path planning is safe and efficient through dynamic obstacle processing, multi-objective planning, path smoothness improvement, and weight adaptive adjustment optimization.

[0037] like Figure 2 As shown in the figure, the greedy algorithm is used to optimize and smooth the turning points of the initial path to obtain the optimized path. In order to improve safety and stability, the optimized path is adjusted in real time based on obstacle information to obtain a real-time path to ensure the safe passage of the blind. Figure 2 It can be seen that the path undergoes a significant change in direction at the key turning point. Through algorithm optimization, the turning point is smoothed to ensure that the path is safe and smooth when turning.

[0038] The path planning module calculates the robot's motion trajectory based on the blind person's real-time path by calculating the tangent, ensuring the coordinated movement of the blind person and the robot.

[0039] The voice interaction module is used to provide voice feedback, inform the blind of upcoming terrain features or obstacle information, and adjust the robot's actions according to the blind's voice commands, further improving the blind's user experience.

[0040] The robot body 1 is physically connected to the blind person through a rigid guide rod 4. One end of the rigid guide rod 4 is hinged to the robot body 1 through a turntable 3. The six-dimensional force sensor is installed at the turntable 3. The six-dimensional force sensor is used to monitor the tension and pressure signals on the rigid guide rod 4. The robot body 1 can automatically adjust the speed in real time according to the blind person's walking rhythm to match the blind person's stride and improve the blind person's following experience.

[0041] This embodiment also provides a traction robot blind guiding method, such as Figure 3 As shown, including:

[0042] Acquire three-dimensional point cloud information of the surrounding environment through coupling of multiple sensors;

[0043] Perform semantic modeling based on 3D point cloud data, generate a global map of the environment, mark dynamic obstacles, and form a composite map of the environment;

[0044] The composite map of the environment is analyzed through artificial potential field factors to generate the initial path for the blind to walk;

[0045] A greedy algorithm is used to optimize and smooth the turning points of the initial path, and the path is adjusted in real time based on obstacle information to obtain a real-time path.

[0046] The motion trajectory of the robot body 1 is generated by calculating the tangent according to the real-time path to ensure the coordinated movement of the blind person and the robot body 1, and finally a command is issued to perform motion control.

[0047] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A traction robot blind guide system, characterized in that: The robot body includes an environment perception and recognition module and a path planning module, wherein the environment perception and recognition module is used to obtain three-dimensional point cloud information of the surrounding environment and generate an environment composite map according to the three-dimensional point cloud information; the path planning module analyzes the environment composite map by introducing an artificial potential field factor to generate a safe and collision-free initial path for the blind; the robot body is physically connected to the blind through a rigid guide rod.

2. The traction robot guide system for the blind as claimed in claim 1, characterized in that: The environment perception and recognition module obtains three-dimensional point cloud information of the surrounding environment through multi-sensor coupling; the sensors include three-dimensional laser radar and depth camera.

3. The traction robot guide system for the blind as claimed in claim 2, characterized in that: The three-dimensional laser radar is installed on the head of the robot body, and the depth camera is located at the front end of the three-dimensional laser radar.

4. The traction robot blind guiding system according to claim 1, characterized in that: The environment perception and recognition module performs semantic modeling according to the three-dimensional point cloud information to generate a global map of the environment; and marks dynamic obstacles in the global map of the environment to obtain the composite map of the environment.

5. The traction robot guiding system for the blind as claimed in claim 1, characterized in that: The path planning module uses a greedy algorithm to optimize and smooth the turning points of the initial path, and adjusts the path in real time in combination with obstacle information to obtain a real-time path.

6. The traction robot blind guiding system according to claim 5, characterized in that: The path planning module can also generate a motion trajectory of the robot body by calculating tangents according to the real-time path, thereby ensuring coordinated movement between the blind person and the robot body.

7. The traction robot blind guiding system according to claim 1, characterized in that: One end of the rigid guide rod is hinged to the robot body through a turntable, and a six-dimensional force sensor is installed on the turntable.

8. The traction robot guiding system for the blind as claimed in claim 7, characterized in that: The six-dimensional force sensor is used to monitor the tension and pressure signals on the rigid guide rod to adjust the travel speed of the robot body in real time.

9. The traction robot blind guiding system according to claim 1, characterized in that: It includes a voice interaction module, which is used to provide voice feedback, broadcast upcoming terrain features or obstacle information to the blind, and adjust the actions of the machine body according to the blind's voice instructions.

10. A traction robot blind-guiding method, characterized in that: The method comprises the following steps: Acquire three-dimensional point cloud information of the surrounding environment through coupling of multiple sensors; Generate a global map of the environment based on 3D point cloud data, mark dynamic obstacles, and form a composite map of the environment; Analyzing the composite map of the environment through artificial potential field factors to generate an initial walking path for the blind person; A greedy algorithm is used to optimize and smooth the turning points of the initial path, and the path is adjusted in real time based on obstacle information to obtain a real-time path. The motion trajectory of the robot body is generated by calculating the tangent according to the real-time path, so as to ensure the coordinated motion of the blind person and the robot body.

Images