Krogh, B. and Thorpe, C. "Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles." In Proceedings of the IEEE Conference on Robotics and Automation, pages 1664-1669, 1986.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....all possible action combinations [20] 23] but it becomes computationally intractable with higher degrees of freedom. Our proposed architecture trains a self organizing neural network to continuously sample the low level configuration space. Other integrated architectures [2] 6] 9] 12] [14] utilize potential fields [11] in their reactive controllers to encode continuous responses, which are subject to local minima problems [13] In contrast, integrated architectures [24] 25] that employ discrete response encoding (i.e. finite, enumerated set of responses) encode high level ....

B. H. Krogh and C. E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proceedings of IEEE International Conference on Robotics and Automation, pages 1664--1669, 1986.

....do not attempt to model complicated maneuvers such as full turns or body reconfigurations. 2. Related Work Many researchers have developed methods that find suitable or efficient paths for agent navigation. In the field of robotics, path planning and collision avoidance are longstanding topics [4, 14, 15, 16] aimed at solving the challenge of navigating a robot through an arbitrary environment to a goal. Research stemming from robot navigation systems has been applied to virtual humans walking through environments in the contexts of computer animation [1, 2, 5, 9, 17] city planning [10] and fire ....

B. Krogh and C. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proceedings of the IEEE International Conference on Robotics and Automation, pages 1664--1669, Piscataway, N.J., 1986. IEEE Press.

....capability, and complexity of computation. Their strengths are complementary and their weaknesses can be mitigated by combining the two approaches in a hybrid architecture. Among the existing hybrid frameworks, 2, 3, 12, 14, 16, 28, 31, 32] emphasize high level task planning. On the other hand, [5, 6, 10, 23, 29] focus on integrating low level reactive motor control with motion path planning. This paper describes a method for goal directed, collisionfree navigation in complex, unpredictable environments that employs a behavior based hybrid architecture with asynchronously operating behavioral modules. It ....

....velocity or torque control of motors to perform fine, smooth motion control. 2. In our framework, the planning module produces a sequence of checkpoints instead of a complete motion path. The constraint of adhering strictly to a generated path no longer exists. This approach is adopted in [23] as well. 3. The real time performance of existing hybrid architectures is still not optimal because the capability of the reactive components has not been fully exploited. In extreme cases, the workload of the high level planning module far exceeds that of the low level reactive module (e.g. ....

B. H. Krogh and C. E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proc. ICRA, pages 1664--1669, 1986.

.... feasible paths for the robot [1, 20, 25, 56] This is related to the question concerning how to generate optimal as well as numerically feasible paths, since in autonomous robot applications it is crucial that the computations are performed on line due to the changing, unmodeled environment [29,52,72]. Modeling of the platform dynamics is typically not an easy task. In fact, some practical problems arose when trying to implementmodel dependentcontrollers for the car like robots, such as finding adhesion coefficients or cornering stiffness parameters [1,73] Difficultyinmodeling and a desire ....

B. Krogh and C. Thorpe. Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles, Proceedings of the 1986.

....rates and priority levels. The work presented in this paper is a step towards the full integration of task planning and motion control, motivated by Khatib s seminal work [17] that formulates a general framework for robot planning and control. This work and a few other integrated architectures [2, 5, 6, 8, 13, 19] have utilized methods based on potential fields [14] in their reactive control algorithms, while their planning and interface techniques differ. Potential field methods are implementations of continuous response encoding [1] infinite space of responses) which makes low level control possible. ....

B. H. Krogh and C. E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proc. ICRA, pages 1664--1669, 1986.

....capability, and complexity of computation. Their strengths are complementary and their weaknesses can be mitigated by combining the two approaches in a hybrid architecture. Among the existing hybrid frameworks, 2, 3, 12, 14, 16, 29, 32, 33] emphasize high level task planning. On the other hand, [5, 6, 10, 23, 30] focus on integrating low level reactive motor control with motion path planning. This paper describes a method for goal directed, collisionfree navigation in complex, unpredictable environments that employs a behavior based hybrid architecture with asynchronously operating behavioral modules. It ....

....velocity or torque control of motors to perform ne, smooth motion control. 2. In our framework, the planning module produces a sequence of checkpoints instead of a complete motion path. The constraint of adhering strictly to a generated path no longer exists. This approach is adopted in [23] as well. 3. The real time performance of existing hybrid architectures is still not optimal because the capability of the reactive components has not been fully exploited. In extreme cases, the workload of the high level planning module far exceeds that of the low level reactive module (e.g. ....

B. H. Krogh and C. E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proc. ICRA, pages 1664-1669, 1986.

....driver intentions) The perceptual input is a obstacle vehicle tracker which provides the relative position and velocity of the target. Internally, these reasoning objects construct a time to impact potential field around the relevant obstacle (similar to the generalized potential fields used in [59]) The vote against a particular action is inversely proportional to the estimated time to impact. Thus, such a reasoning object will prefer braking to accelerating when it sees a stationary obstacle ahead in its lane since decelerating will increase the time to impact. Naturally, when an action ....

B. Krogh and C. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proceedings of IEEE Conference on Robotics and Automation, 1986.

....for a given time interval is calculated, and the algorithm is repeated. Krogh [19] has enhanced this concept further by taking into consideration the robot s velocity in the vicinity of obstacles. Thorpe [27] has applied the potential field method to off line path planning and Krogh and Thorpe [20] suggest a combined method for global and local path planning, which uses a Generalized Potential Field approach. Newman and Hogan [15] introduce the construction of potential functions through combining individual obstacle functions with logical operations. Common to these methods is the ....

Krogh, B. H. and Thorpe, C. E., "Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles." Proceedings of the

....as if it were subject to this force. Equivalently, we may say that the robot selects, at each point, the action that brings it to a point of locally minimal potential. Following Khatib s suggestion, many methods for robot navigation have been proposed based on artificial potential fields (e.g. [1, 2, 5, 6]) All of these methods perform action selection in a very specific case: when the objective is to place the robot at a certain location, and the possible actions consist in robot displacements. It is not obvious if, and how, these methods can be used to solve a more general action selection ....

B.H. Krogh and C.E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proc. of the Intl. Conf. on Robotics and Automation, 1986.

....collision free robots, the planner can generate forward and reverse paths. path. The initial phase is often called off line path planning. The literature is replete of off li ne path planning algorithms, the 1989, Pin and Vasseur, 1990) search for paths for car like robo t vehicles. A few method s (Krogh and Thorpe, 1986, Warren, 1989) use artificial potential fields for off line path planning. 1 omroc cc vwidth r s s g lane expanded obstacle original obstacle 2 Figure 1 . Corner clearance on worst case turn. Figure 2 . Minimum lane width. The major contribut ions of this paper are the technique used to ....

Krogh , B. H. and Thorpe, C. E., 1986, "Integrated Path Plannin g and Dynamic Steering Control for Autonomous Vehicles," Proceedings, IEEE Int. Conf. Robotics Automat. , San Francisco , CA, pp. 1664-1669.

....of the second derivative, the resulting curve will be a cubic spline. This is a fortunate fact since it means that we will not be forced to relay on extensive world information or to do any heavy computations on line which tends to be the case when more sophisticated planning algorithms are used [12], 13] 18] It is thus an almost trivial task to generate the splines that connect the robot and the target in the approach target behavior, and the robot and the obstacle in the approach obstacle behavior, as seen in Figure 8. 3.2 Tracking We now have an on line method for producing low ....

B. Krogh and C. Thorpe. Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles, Proceedings of the 1986 IEEE International Conference on Robotics and Automation, San Francisco, CA, pp. 1664-1669, 1986.

.... it senses its environment during motion execution and should be attracted toward its goal (attractive potential field) while being repulsed by ob5 stacles (repulsive potential fields) This approach has been used in the past to solve path planning as well as on line navigation problems ( 6] 7] [8], 9] At every time interval Deltat, we build a potential function U which depends on the current world situation and the state of the vehicle. U is a combination of three potential functions: 1) a potential U s corresponding to a repulsive potential field representing the static obstacles in ....

B. Krogh and C. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In International Conference on Robotics and Automation, San Francisco, USA, Apr. 1986. IEEE.

....in a reactive system to avoid collision. This represents effectively a common solution to the problem. In our approach the robot was always obliged to minimize the drift from the original trajectory even while avoiding obstacles. Different approaches exist in the literature. Krogh and Thorpe [Krogh 86] proposed to replace the planned trajectory by critical points. The artificial potential field method [O. Khatib 86] is then applied to move the robot between these points. The drawback of this method is that the robot does not move on the original trajectory, which makes robot control between ....

B.H. Krogh & C.E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proc. IEEE International Conference on Robotics and Automation, San Francisco (USA)., 1986. 154 References 155

....as two separate phases. Many motion planning techniques have been proposed (see [7] for a survey) to produce a collision free and feasible path avoiding the obstacles of the workspace. Motion control then consists in making the robot follow this path, possibly using a potential field based method [12, 6] for reacting to unexpected obstacles sensed during the execution. Such an approach suffers from a lack of robustness when experimented in real world settings since the execution of the planned trajectory is conditioned by several sources of uncertainties. Control and sensing errors are ....

B.H. Krogh and C.E. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In IEEE International Conference on Robotics and Automation, San Francisco (USA)., 1986.

....2 norm of the second derivative, the resulting curve will be a spline. This is a fortunate fact since it means that we will not be forced to relay on extensive world information or to do any heavy computations on line, which tends to be the case when using more sophisticated planning algorithms. (Krogh and Thorpe, 1986; Latombe, 1991; Stenz, 1994) It should be pointed out that these trajectories actually do not stay away from the safety region around the obstacle at all times since we only interpolate through one distinct point. The reason for choosing such a solution will become clear further on when we fuse ....

Krogh B. and C. Thorpe (1986). Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles, Proceedings of the 1986 IEEE International Conference on Robotics and Automation, San Francisco, CA, pp. 1664-1669.

....programs that have the capability to maneuver in complex environments. The robot, when moving autonomously in an unknown environment, must have the capability to sense the environment to avoid obstacles and in addition should always know its relative position with respect to the environment[1] 6][8][14] This goal has a number of inherent difficulties such as obstacle detection and avoidance and movement in complex environments. In this paper, we present experiments in programming a mobile robot that allows for incremental learning. The experiments detail the objective, implementation and ....

Krogh, Bruce H. and Charles E. Thrope (1986). Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles. Proceedings of the IEEE International Conference on Robotics and Automation, Washington D.C., 1664-1669.

No context found.

B. Krogh and C. Thorpe, "Integrated path planning and dynamic steering control for autonomous vehicles," in Proceedings of IEEE Conference on Robotics and Automation, 1986.

No context found.

Krogh, B. and Thorpe, C. "Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles." In Proceedings of the IEEE Conference on Robotics and Automation, pages 1664-1669, 1986.

No context found.

B. H. Krogh and C. E. Thorpe. Integrated Path Planning and Dynamic Steering control for Autonomous Vehicles. In IEEE Int. Conf. on Robotics and Automation, pages 1664--1669, San Francisco, USA, 1986.

No context found.

B. H. Krogh and C. E. Thorpe, "Integrated path planning and dynamic steering control for autonomous vehicles," in Proc. IEEE Int. Conf. Robotics and Automation, San Francisco, CA, 1986, pp. 1664--1669.

No context found.

B. Krogh and C. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proceedings of the IEEE Conference on Robotics and Automation, pages 1664--1669, 1986.

No context found.

B. Krogh and C. Thorpe. Integrated path planning and dynamic steering control for autonomous vehicles. In Proceedings of IEEE Conference on Robotics and Automation, 1986.

No context found.

Krogh, B. H. and Thorpe, C. E., "Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles." Proceedings of the

No context found.

Krogh, B. H. and Thorpe, C. E., "Integrated Path Planning and Dynamic Steering Control for Autonomous Vehicles." Proceedings of the

No context found.

B.H. Krogh. Integrated path planning and dynamic steering control from autonomous vehicles. Proc. IEEE Int'l Conf. on Robotics and Automation, pages 1664-- 1669, April 1986.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC