Path planning and formation structure forming are two of the most important problems for autonomous underwater vehicles(AUVs) to collaborate with each other.In this work,a dynamic formation model was proposed,in which...Path planning and formation structure forming are two of the most important problems for autonomous underwater vehicles(AUVs) to collaborate with each other.In this work,a dynamic formation model was proposed,in which several algorithms were developed for the complex underwater environment.Dimension changeable particle swarm algorithm was used to find an optimized path by dynamically adjusting the number and the distribution of the path nodes.Position relationship based obstacle avoidance algorithm was designed to detour along the edges of obstacles.Virtual potential point based formation-keeping algorithm was employed by incorporating dynamic strategies which were decided by the current states of the formation.The virtual potential point was used to keep the formation structure when the AUV or the formation was deviated.Simulation results show that an optimal path can be dynamically planned with fewer path nodes and smaller fitness,even with a concave obstacle.It has been also proven that different formation-keeping strategies can be adaptively selected and the formation can change its structure in a narrow area and restore back after passing the obstacle.展开更多
In this paper, we study the kinematic mechanism and path planning for a two-caster nonholonomic vehicle (the Essboard) which is a recent variant of skateboard. Different from the most studied Snakeboard, the Essboard ...In this paper, we study the kinematic mechanism and path planning for a two-caster nonholonomic vehicle (the Essboard) which is a recent variant of skateboard. Different from the most studied Snakeboard, the Essboard consists of a torsion bar and two platforms, each of which contains a pedal and a caster. We first investigate the relationship between the tilt angles of the pedals and the wheel directions of the casters. This relationship reveals how to control the wheel directions by adjusting the tilt angles. Next, the rotational radius of the Essboard is derived for a given pair of tilt angles of both pedals. The rotational radius of the Essboard is much different than that of the Snakeboard. Then we develop a path-planning algorithm for the Essboard to move from a start position to the goal, using a series of consecutively connected arcs, which are tangent to each other at the connected points. It is shown from a kinematic point of view that the path planning of the Essboard can be solved by a series of pairs of pedals' tilt angles. Three experiments are conducted to confirm the correctness of the main results. The results in this paper are a foundation for further study of the Essboard.展开更多
基金Project(NS2013091)supported by the Basis Research Fund of Nanjing University of Aeronautics and Astronautics,China
文摘Path planning and formation structure forming are two of the most important problems for autonomous underwater vehicles(AUVs) to collaborate with each other.In this work,a dynamic formation model was proposed,in which several algorithms were developed for the complex underwater environment.Dimension changeable particle swarm algorithm was used to find an optimized path by dynamically adjusting the number and the distribution of the path nodes.Position relationship based obstacle avoidance algorithm was designed to detour along the edges of obstacles.Virtual potential point based formation-keeping algorithm was employed by incorporating dynamic strategies which were decided by the current states of the formation.The virtual potential point was used to keep the formation structure when the AUV or the formation was deviated.Simulation results show that an optimal path can be dynamically planned with fewer path nodes and smaller fitness,even with a concave obstacle.It has been also proven that different formation-keeping strategies can be adaptively selected and the formation can change its structure in a narrow area and restore back after passing the obstacle.
基金supported by the National Natural Science Foundation of China (Grant Nos. 51105012 and 61175079)
文摘In this paper, we study the kinematic mechanism and path planning for a two-caster nonholonomic vehicle (the Essboard) which is a recent variant of skateboard. Different from the most studied Snakeboard, the Essboard consists of a torsion bar and two platforms, each of which contains a pedal and a caster. We first investigate the relationship between the tilt angles of the pedals and the wheel directions of the casters. This relationship reveals how to control the wheel directions by adjusting the tilt angles. Next, the rotational radius of the Essboard is derived for a given pair of tilt angles of both pedals. The rotational radius of the Essboard is much different than that of the Snakeboard. Then we develop a path-planning algorithm for the Essboard to move from a start position to the goal, using a series of consecutively connected arcs, which are tangent to each other at the connected points. It is shown from a kinematic point of view that the path planning of the Essboard can be solved by a series of pairs of pedals' tilt angles. Three experiments are conducted to confirm the correctness of the main results. The results in this paper are a foundation for further study of the Essboard.
