CN-122009357-A - Terrain self-adaptive hexapod robot based on wheel foot composite structure and K230 visual recognition

CN122009357ACN 122009357 ACN122009357 ACN 122009357ACN-122009357-A

Abstract

The invention belongs to the field of mobile robots, and provides a terrain self-adaptive hexapod robot based on a wheel-foot composite structure and K230 visual identification, which comprises a wheel-foot composite mechanical body, a multi-mode sensing system and a self-adaptive control system, wherein the wheel-foot composite mechanical body is provided with a wheel-foot sensing system; the wheel-foot composite machine body comprises a leg wheel-foot integrated unit, a leg joint structure and a main body frame structure which are sequentially connected, wherein the multi-mode sensing system comprises a hardware layer, a sensing layer and a fusion layer which are sequentially connected, the wheel-type motion control unit, the foot-type motion control unit and the mode switching control unit are embedded in the self-adaptive control system, the road surface flatness can be judged in real time through the multi-mode sensing system, the motor output is dynamically optimized, the gait and the torque can be intelligently adjusted through accurate identification of the sensing layer on the type of an obstacle and the ground material, and the passing success rate of complex terrain is improved.

Inventors

YANG JINTAO
TAN ZHONGXIAN
HE YUN
FANG CHUNHUI

Assignees

云南交通职业技术学院

Dates

Publication Date: 20260512
Application Date: 20260309

Claims (7)

1. The terrain self-adaptive hexapod robot based on the wheel-foot composite structure and the K230 visual identification is characterized by comprising a wheel-foot composite mechanical body, a K230-based multi-mode sensing system and a self-adaptive control system, wherein the wheel-foot composite mechanical body comprises a leg wheel-foot integrated unit, a leg joint structure and a main body frame structure which are sequentially connected, the multi-mode sensing system comprises a hardware layer, a sensing layer and a fusion layer which are sequentially connected and are based on a YOLO-V5 framework, the self-adaptive control system is embedded with a wheel-type motion control unit, a foot-type motion control unit and a mode switching control unit, and the leg wheel-foot integrated unit comprises 6 groups of wheel-foot composite structures; The multi-mode sensing system is arranged on the wheel foot composite mechanical body, and the self-adaptive control system is respectively connected with the wheel foot composite mechanical body and the multi-mode sensing system; The fusion layer is used for synchronously aligning and filtering the recognition result output by the sensing layer and the gesture data and the ranging data acquired by the sensor to obtain an environment sensing result, wherein the environment sensing result comprises terrain type, accurate position of an obstacle, movement speed and ground inclination angle information, and the mode switching control unit is used for controlling the working states of the wheel type movement control unit and the foot type movement control unit by utilizing preset switching triggering conditions according to the environment sensing result.
2. The terrain self-adaptive hexapod robot based on the visual recognition of the wheel foot composite structure and the K230 is characterized in that the wheel foot composite structure is manufactured by an outer rotor type brushless direct current hub motor and a shank structure through an integrated injection molding process, a stator of the brushless direct current hub motor is fixed inside a shell at the tail end of the shank structure, a rubber tire is covered outside a rotor of the brushless direct current hub motor, and a rotation axis of the brushless direct current hub motor is in a space perpendicular relation with a swing axis of the shank structure.
3. The terrain self-adaptive hexapod robot based on the wheel-foot composite structure and the K230 visual recognition according to claim 2, wherein the leg joint structure comprises a big arm structure and a steering engine; the large arm structure is connected with the shank structure through one steering engine, and the large arm structure is connected with the side plate of the main body frame structure through two steering engines.
4. The terrain self-adaptive hexapod robot based on the wheel foot composite structure and the K230 visual recognition is characterized in that the main body frame structure is made of PLA material through 3D printing technology in an integrated mode, the supporting structure of the main body frame structure is honeycomb-shaped, and the K230 visual terminal of the multi-mode sensing system is fixed in an installation cabin at the front end of the top of the main body frame structure.
5. The terrain-adaptive hexapod robot based on the wheel-pod composite structure and the K230 visual recognition of claim 4, wherein the hardware layer comprises a microcontroller, the K230 visual terminal, a six-axis inertial measurement unit and an ultrasonic sensor; the six-axis inertial measurement unit is fixed at the gravity center position of the main body frame structure, the ultrasonic sensors are respectively arranged at the front end, the left side and the right side of the main body frame structure, the ultrasonic sensors at the front end of the main body frame structure are 150mm away from the ground, and the ultrasonic sensors at the left side and the right side of the main body frame structure are distributed at an included angle of 120 degrees.
6. The terrain adaptive hexapod robot based on the wheel-foot composite structure and the K230 visual recognition of claim 5, wherein the working process of the wheel-type motion control unit comprises: Processing the acquired image of the K230 visual terminal by using a preset path planning algorithm to obtain an optimal path; Extracting the error between the optimal path and the current direction to obtain the actual physical distance deviation; and regulating and controlling the brushless direct current hub motor by utilizing a PID controller according to the actual physical distance deviation.
7. The terrain adaptive hexapod robot based on the wheel-foot composite structure and the K230 visual recognition of claim 5, wherein the working process of the foot-type motion control unit comprises: Dividing the target ground into a passable area and an unviewable area according to the terrain category and a preset passable category range; and regulating and controlling the leg wheel-foot integrated unit in the passable area by utilizing a triangular gait algorithm.

Description

Terrain self-adaptive hexapod robot based on wheel foot composite structure and K230 visual recognition Technical Field The invention relates to the field of mobile robots, in particular to a terrain self-adaptive hexapod robot based on a wheel-foot composite structure and K230 visual identification. Background The existing mobile robots are mainly divided into wheel type, foot type and wheel-foot composite type. Wheeled robots, such as traditional AGV dollies, although moving fast on flat road surfaces and having low energy consumption, have poor capability of crossing obstacles, are easy to sink on soft road surfaces such as sand, mud and the like, and have serious defects in terrain adaptability. Foot robots, particularly hexapod robots, can effectively span complex terrain based on their discrete foot points. However, when the existing six-legged robot moves on a flat road surface, a plurality of joint steering engines need to be driven to simulate gait, and the problems of low moving speed, high energy consumption and high mechanical structure abrasion exist. In order to solve the above problems, a wheel-foot hybrid robot has appeared. For example, CN105923067a discloses a small wheel-foot composite hexapod robot, which is provided with a switchable wheel structure at the foot end. However, the prior art and similar wheel-foot compound schemes have the common defects that the environment sensing capability is weak, low-precision sensors such as infrared sensors and ultrasonic waves are depended on, the terrain types and dynamic obstacles cannot be accurately identified, the wheel-foot modes are switched by the aid of preset programs or simple distance triggering due to the insufficient sensing capability, intelligent decisions cannot be made according to real-time and complex environment information, mode switching is delayed or misjudgment is caused, and the overall efficiency in mixed terrains is limited. Disclosure of Invention In order to overcome the defects of the prior art, the invention aims to provide a terrain self-adaptive hexapod robot based on a wheel foot composite structure and K230 visual identification, and solve the problems of low moving efficiency and poor terrain adaptability of the traditional device. In order to achieve the above object, the present invention provides the following solutions: A terrain self-adaptive hexapod robot based on a wheel-foot composite structure and K230 visual recognition comprises a wheel-foot composite mechanical body, a multi-mode sensing system based on K230 and a self-adaptive control system, wherein the wheel-foot composite mechanical body comprises a leg wheel-foot integrated unit, a leg joint structure and a main body frame structure which are sequentially connected, the multi-mode sensing system comprises a hardware layer, a sensing layer based on a YOLO-V5 framework and a fusion layer which are sequentially connected, the self-adaptive control system is embedded with a wheel-type motion control unit, a foot-type motion control unit and a mode switching control unit, and the leg wheel-foot integrated unit comprises 6 groups of wheel-foot composite structures; The multi-mode sensing system is arranged on the wheel foot composite mechanical body, and the self-adaptive control system is respectively connected with the wheel foot composite mechanical body and the multi-mode sensing system; The fusion layer is used for synchronously aligning and filtering the recognition result output by the sensing layer and the gesture data and the ranging data acquired by the sensor to obtain an environment sensing result, wherein the environment sensing result comprises terrain type, accurate position of an obstacle, movement speed and ground inclination angle information, and the mode switching control unit is used for controlling the working states of the wheel type movement control unit and the foot type movement control unit by utilizing preset switching triggering conditions according to the environment sensing result. The wheel foot composite structure is preferably manufactured by an outer rotor type brushless direct current hub motor and a shank structure through an integrated injection molding process, a stator of the brushless direct current hub motor is fixed inside a shell at the tail end of the shank structure, a rubber tire is covered outside a rotor of the brushless direct current hub motor, and a rotation axis of the brushless direct current hub motor is in a space perpendicular relation with a swing axis of the shank structure. Preferably, the leg joint structure comprises a big arm structure and a steering engine; the large arm structure is connected with the shank structure through one steering engine, and the large arm structure is connected with the side plate of the main body frame structure through two steering engines. Preferably, the main body frame structure is made of PLA material through 3D printing technology in an integrated mode, t