• The Journal of Korea Robotics Society
• /
• v.18 no.2
• /
• pp.143-154
• /
• 2023

In this study, a fast motion planning method for the swing motion of a 6x6 wheel-legged robot to traverse large obstacles and gaps is proposed. The motion planning method presented in the previous paper, which was based on trajectory optimization, took up to tens of seconds and was limited to two-dimensional, structured vertical obstacles and trenches. A deep neural network based on one-dimensional Convolutional Neural Network (CNN) is introduced to generate keyframes, which are then used to represent smooth reference commands for the six leg angles along the robot's path. The network is initially trained using the behavioral cloning method with a dataset gathered from previous simulation results of the trajectory optimization. Its performance is then improved through reinforcement learning, using a one-step REINFORCE algorithm. The trained model has increased the speed of motion planning by up to 820 times and improved the success rates of obstacle crossing under harsh conditions, such as low friction and high roughness.

• 제어로봇시스템학회:학술대회논문집
• /
• 1988.10a
• /
• pp.131-136
• /
• 1988

In this paper we describe a motion control algorithm for a 4-legged robot over slopped terrain and steps. The new concept of the mechanically constrained angle has been introduced and the algorithm has been developed based on the relationship between the gait stability margin and the slope angle. The result then has been extended to the case where the robot walks over steps.

• The Journal of Korea Robotics Society
• /
• v.2 no.1
• /
• pp.55-63
• /
• 2007

Mobile robots traveling on rough terrain need several algorithms to overcome obstacles. In this paper, we propose the step-type obstacle traversal algorithm to adapt the mobile robot with six arms and wheels to travel on rough terrain. Obstacle traversal is composed of two different stages: planning and control. In planning stage, the required joint torque of each arm as well as the interference between the wheels and the arms are analyzed to guarantee traversing obstacles. Control stage includes such steps as checking distance to obstacle, determining the height and length of obstacle, performing arm motion according to sensed torque data, and evaluating safety at every instance. The proposed algorithm is designed and implemented for CALEB 1 six legged robot developed in the laboratory and verified by simulation and experiment in outdoor environment.

• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.2
• /
• pp.141-148
• /
• 2012

This paper proposes fault tolerant gaits of a quadruped robot with feet of flat shape. Fault tolerant gaits make it possible for a legged robot to continue static walking against a leg failure. In the previous researches, it was assumed that a legged robot had feet that have point contact with the surface. When the robot is endowed with feet having flat shape, fault tolerant gaits can show better performance compared with the former gaits, especially in terms of the stride length and gait stability. In this paper, fault tolerant gaits of a quadruped robot against a locked joint failure are addressed in straight-line motion and crab walking, respectively.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.83.3-83
• /
• 2001

A method to adaptively control leg-ends trajectories of a five-legged walking robot, Cepheus-2, has been developed in terms of a kind of a table look-up method. Cepheus-2 is a five-legged robot with a pentagonal body with two joints of each leg. The robot control system has a hierarchical autonomic-integrated architecture with a main computer (PC), a manager and servo modules. Being given the goals of walking by the main computer, the manager module assigns a type of leg-end trajectories of which data are described with the work space coordinates for the legs. Every servo module generates the joint angle data. In steady walking of the robot on flat floor without obstacle, two joints have to generate the assigned trajectory and five legs ...

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.5
• /
• pp.622-628
• /
• 2009

A proper walking pattern is to be assigned for a walk of multi-legged robots. For the purpose of identifying a good walking pattern for multi-legged robots, this paper consider a simple model of quadruped robotic walking and analyze its walking balance based on the centroid of foot polygons formed in every step. A performance index to estimate the walking balance is also proposed. Simulation studies show that the centroid trajectory of foot polygons and the walking balance in a common quadruped walking are different according to the walking pattern employed. Based on the walking balance index and a bio-mimetic aspect, a useful walking pattern for quadruped robots is finally addressed.

• The Journal of Korea Robotics Society
• /
• v.17 no.4
• /
• pp.438-446
• /
• 2022

In this paper, we propose an intelligent three-legged landing system that can maintain stability and level even on rough terrain than conventional four-legged landing systems. Conventional landing gear has the limitation that it requires flat terrain for landing. The 3-leg landing system proposed in this paper extends the usable range of the legs and reduces the weight, allowing the quadcopter to operate in various environments. To do this, kinematics determine the joint angles and coordinates of the legs of the two-link structure. Based on the angle value of the quadcopter detected via the IMU sensor, the leg control method that corrects the posture is determined. A force sensor attached to the end of the leg is used to detect contact with the ground. At the moment of contact with the ground, landing control starts according to the value of the IMU sensor. The proposed system verifies its reliability in various environments through an indoor landing test stand. Finally, in an outdoor environment, the quadcopter lands on a 20 degree incline and 20 cm rough terrain after flight. This demonstrates the stability and effectiveness of the 3-leg landing system even on rough terrain compared to the 4-leg landing system.

• 제어로봇시스템학회:학술대회논문집
• /
• 1987.10b
• /
• pp.69-73
• /
• 1987

In this paper the continuous static stable gait, kinematics and the basic control algorithm of the quadruped walking robot have been discussed. The control method described in this paper will be extended for the walking robot to walk on an uneven terrain.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.9
• /
• pp.936-941
• /
• 2014

Walking mechanisms are very important for legged robots to ensure their stable locomotion. In this research, Klann-linkage is suggested as a walking mechanism for a water-running robot and is optimized using level average analysis. The structure of the Klann-linkage is introduced first and design variables for the Klann-linkage are identified considering the kinematic task of the walking mechanism. Next, the design problem is formulated as a path generation optimization problem. Specifically, the desired path for the foot-pad is defined and the objective function is defined as the structural error between the desired and the generated paths. A process for solving the optimization problem is suggested utilizing the sensitivity analysis of the design variables. As a result, optimized lengths of Klann-linkage are obtained and the optimum trajectory is obtained. It is found that the optimized trajectory improves the cost function by about 62% from the initial one. It is expected that the results from this research can be used as a good example for designing legged robots.

• The Journal of Korea Robotics Society
• /
• v.9 no.1
• /
• pp.57-66
• /
• 2014

This paper describes the design concept of a bio-inspired legged underwater and estimating its performance by implementing simulations. Especially the leg structure of an underwater organism, diving beetles, is fully adopted to our designing to employ its efficiency for swimming. To make it possible for the robot to both walk and swim, the transformable kinematic model according to applications of the leg is proposed. To aid in the robot development and estimate swimming performance of the robot in advance, an underwater simulator has been constructed and an approximated model based on the developing robot was set up in the simulation. Furthermore, previous work that we have done, the swimming locomotion produced by a swimming patten generator based on the control parameters, is briefly mentioned in the paper and adopted to the simulation for extensive studies such as path planning and control techniques. Through the results, we established the strategy of leg joints which make the robot swim in the three dimensional space to reach effective controls.