Static walk gait and diagonal trot gait switching algorithm applied to quadruped bionic robot

Abstract

The invention relates to a static walk gait and diagonal trot gait switching algorithm applied to a quadruped bionic robot, and provides six conditions in which an intermittent static walk gait is switched to a diagonal trot gait, and an optimal switching method. According to the algorithm, a defining method for describing the positions of foot ends in a gait switching process is given, and the intermittent static gait and points, which can be switched, thereof as well as the diagonal trot gait and points, which can be accessed, thereof are provided; after that, rules of rhythm control and a strategy of mode control are given in the gait switching process, so that the smoothness of speed is guaranteed in the switching process; finally, an improved generic stability margin stability discrimination method is provided, so that the stability of a robot body which is in a state of acceleration can be accurately monitored during gait transition. The algorithm considers phase difference and switching speed at the same time, thus shortening the time difference from receiving a switching signal to beginning the switching while improving the stability and smoothness of gait switching, and enabling the gait switching of the robot to be stable and efficient.

Description

technical field [0001] The invention relates to an algorithm for switching between a static gait and a diagonal trotting gait of a quadruped bionic robot, and belongs to the technical field of artificial intelligence. Background technique [0002] At this stage, quadruped robots are gradually moving from indoors to outdoors, and their high dynamic performance, environmental adaptability and large load capacity have been significantly improved. In order to complete a given task, robots generally adopt a single movement mode, which significantly reduces the adaptability of robots to various complex environments, while quadruped mammals can switch between different movement modes according to different environments. Therefore, in order to improve the environmental adaptability of the robot, it is necessary to study the unified generation method of various motion patterns of the robot and the smooth switching strategy between them. [0003] In addition, energy utilization effic...

AI Technical Summary

Problems solved by technology

[0007] In recent years, there have been more studies on gait switching, but there is no relatively complete method that comprehensively considers the stability of the robot in the process of gait switching, smooth switching of speed and minimum switching time.

"Gait transition of quadruped robot using rhythm control and stability analysis" published in "IEEE International Conference on Robotics and Biomimetics" ("IEEE Robotics and Biomimics International Conference Proceedings") pp. 2535-2539 in 2013 Robot Gait Switching and Stability Analysis") and the paper "Automated Transitions Between Walking and Running in Legged Robots" ("Footed Robot Automatic Switching Between Walking and Running") Inverted pendulum model and rigid springs are used to complete gait switching. In 2014, a master thesis "Gait Control and Switching Strategy Analysis of Quadruped Bionic Robots" pointed out walk-trot gait switching and the influence of speed changes, but did not give the specific execution process of the switch. The article "Hysteresis in gait transition induced by changing waist joint stiffness of a quadruped robot driven by nonlinear oscillators with phase reset" uses the method of changing the duty cycle to complete the gait Switching, the article "A Gait-Transition Method for a Quadruped Walking Robot" proposes a switching method to stop the walking robot at a certain point that meets the conditions to start the next gait, but neither gait switching is considered Stability and smoothness problems caused by changes in step size and cycle; "A gait transition algorithm based on hybrid walking gait for a quadruped walking robot" first improves the intermittent static gait to a non-intermittent static gait, and then gives a non-intermittent static gait The switching process is completed in one state intermittently, and the switching process is divided into three parts with the same time consumption, without considering the problem of speed transition; a relatively complete walk-to-trot has been proposed in "Biologically inspired gaittransition control for a quadruped walking robot" Switch plan, but received switching message There is a long time difference to the execution switching point. If the static gait state does not meet the switching conditions, it needs to continue to run until it meets the initial state of the two switching situations given, which greatly reduces the efficiency of gait switching.

Images