In this paper, we review the various methods for robust autonomous mobile robot navigation and scene modeling in structured environments. The techniques vary with availibility of a priori knowledge of the environment ...
详细信息
A novel paradigm for the synthesis of convergent-axis stereo geometries is presented. This paradigm incorporates constraints that represent task-oriented properties that can be easily derived in many applications of s...
详细信息
A novel paradigm for the synthesis of convergent-axis stereo geometries is presented. This paradigm incorporates constraints that represent task-oriented properties that can be easily derived in many applications of stereo-ranging, particularly those involving manipulation. Given these properties, a solution that selects the set of optimal design parameters, in the sense of accuracy, under the given constraints is presented. Results of this algorithm are discussed and analyzed in terms of the optimality of the minimum-area technique used in the derivation.
In this paper, a robust robot force tracking impedance control scheme that uses a neural network as a compensator is proposed. The proposed neural compensator has the capability of making the robot track a specified d...
详细信息
In this paper, a robust robot force tracking impedance control scheme that uses a neural network as a compensator is proposed. The proposed neural compensator has the capability of making the robot track a specified desired force as well as of compensating for uncertainties in environment location and stiffness, and the uncertainties in robot dynamics. The neural compensator is trained separately for free space motion and contact space motion control using two different training signals. The proposed training signal for force control can be used regardless of the environment profile in order to achieve desired force tracking. Simulation studies with three link rotary robot manipulator are carried out to demonstrate the robustness of the proposed scheme under uncertainties in robot dynamics, environment position and environment stiffness. The results show that excellent force tracking is achieved by the neural network.
A neural network technique for robot manipulator control is proposed. This technique called reference compensation technique(RCT), compensates for uncertainties in robot dynamics at input trajectory level rather than ...
详细信息
A neural network technique for robot manipulator control is proposed. This technique called reference compensation technique(RCT), compensates for uncertainties in robot dynamics at input trajectory level rather than at the joint torque level. The ultimate goal of the proposed technique is to achieve an ideal computed-torque controlled system. Compensating at trajectory level carries several advantages over other neural network control schemes that compensate at robot joint torques. First, the position tracking performance is better. Second, the neural controller is more robust to feedback controller gain variations. Finally, practical implementation can be done with ease without changing the internal control algorithm. Simulation studies have been conducted for various neural network structures and different training signals. The results showed the superior performances of the RCT over other NN control schemes.
Programmable vector fields can be used to control a variety of flexible planar parts feeders. When a part is placed on our devices, the programmed vector field induces a force and moment upon it. Over time, the part m...
详细信息
Programmable vector fields can be used to control a variety of flexible planar parts feeders. When a part is placed on our devices, the programmed vector field induces a force and moment upon it. Over time, the part may come to rest in a dynamic equilibrium state. We demonstrate lower bounds on what the devices cannot do, and results on a classification of control strategies. We suggest sufficient conditions for programmable fields to induce well-behaved equilibria on every part placed on our devices. We define composition operators to build complex strategies from simple ones, and show the resulting fields are also well-behaved. We discuss whether fields outside this class can be useful and free of pathology. Using these tools, we describe new manipulation algorithms, and improve existing planning algorithms by a quadratic factor, and the plan-length by a linear factor. We relax earlier dynamic and mechanical assumptions to obtain more robust and flexible strategies. Finally, we consider parts feeders that can only implement a very limited "vocabulary" of vector fields. We discuss the trade-off between mechanical complexity and planning complexity.
Two neural network models were developed for the prediction of postural sway response due to exposure to risk factors including environmental lighting, job-tasks, standing surface firmness, surface oiliness, work load...
详细信息
Two neural network models were developed for the prediction of postural sway response due to exposure to risk factors including environmental lighting, job-tasks, standing surface firmness, surface oiliness, work load, peripheral vision conditions, age and gender. Variables used to measure the loss of balance were index of proximity to stability boundary and sway length. Tests showed that job-task is the main risk factor that changes the output while there is some impact by age or gender on the outcome of the model. The results from these models can be used to find risk factors that have great impact on loss of balance and therefore can help in designing intervention programs.
An automatic target recognition (ATR) classifier is proposed that uses modularly cascaded vector quantizers (VQs) and multilayer perceptrons (MLPs). A dedicated VQ codebook is constructed for each target class at a sp...
详细信息
An automatic target recognition (ATR) classifier is proposed that uses modularly cascaded vector quantizers (VQs) and multilayer perceptrons (MLPs). A dedicated VQ codebook is constructed for each target class at a specific range of aspects, which is trained with the K-means algorithm and a modified learning vector quantization (LVQ) algorithm. Each final codebook is expected to give the lowest mean squared error (MSE) for its correct target class at a given range of aspects. These MSEs are then processed by an array of window MLPs and a target MLP consecutively. In the spatial domain, target recognition rates of 90.3 and 65.3 percent are achieved for moderately and highly cluttered test sets, respectively. Using the wavelet decomposition with an adaptive and independent codebook per sub-band, the VQs alone have produced recognition rates of 98.7 and 69.0 percent on more challenging training and test sets, respectively.
The purpose of this article is to examine the stability of a robust and decentralized PD control law for robot manipulator control. The control law, which was developed by the authors, has been shown experimentally to...
Visibility analysis algorithms use digital elevation models (DEMs), which represent terrain topography, to determine visibility at each point on the terrain from a given location in space. This analysis can be computa...
Visibility analysis algorithms use digital elevation models (DEMs), which represent terrain topography, to determine visibility at each point on the terrain from a given location in space. This analysis can be computationally very demanding, particularly when manipulating high resolution DEMs accurately at interactive response rates. Massively data-parallel computers offer high computing capabilities and are very well-suited to handling and processing large regular spatial data structures. In the paper, the authors present a new scanline-based data-parallel algorithm for visibility analysis. Results from an implementation onto a MasPar massively data-parallel SIMD computer are also presented.
In the emerging paradigm of animate vision, the visual processes are not thought of as being independent of cognitive or motor processing, but as an integrated system within the context of visual behavior. Intimate co...
详细信息
In the emerging paradigm of animate vision, the visual processes are not thought of as being independent of cognitive or motor processing, but as an integrated system within the context of visual behavior. Intimate coupling of sensory and motor systems have found to improve significantly the performance of behavior based vision systems. In order to study active vision systems one requires sensory-motor systems. Designing, building, and operating such a test bed is a challenging task. In this paper we describe the status of on-going work in developing a sensory-motor robotic system, R2H, with ten degrees of freedoms (DOF) for research in active vision. To complement the R2H system a Graphical Simulation and Animation (GSA) environment is also developed. The objective of building the GSA system is to create a comprehensive design tool to design and study the behavior of active systems and their interactions with the environment. GSA system aids the researchers to develop high performance and reliable software and hardware in a most effective manner. The GSA environment integrates sensing and motor actions and features complete kinematic simulation of the R2H system, it's sensors and it's workspace. With the aid of the GSA environment a Depth from Focus (DFF), Depth from Vergence, and Depth from Stereo modules are implemented and tested. The power and usefulness of the GSA system as a research tool is demonstrated by acquiring and analyzing images in the real and virtual worlds using the same software implemented and tested in the virtual world.
暂无评论