In this paper, we introduce an approach for visual tracking in videos that predicts the bounding box location of a target object at every frame. This tracking problem is formulated as a sequential decision-making process where both historical and current information are taken into account to decide the correct object location. We develop a deep reinforcement learning based strategy, via which the target object position is predicted and decided in a unified framework. Specifically, a RNN based prediction network is developed where local features and global features are fused together to predict object movement. Together with the predicted movement, some predefined possible offsets and detection results form into an action space. A decision network is trained in a reinforcement manner to learn to select the most reasonable tracking box from the action space, through which the target object is tracked at each frame. Experiments in an existing tracking benchmark demonstrate the effectiveness and robustness of our proposed strategy.

This paper presents a parameter coordination approach based on constraints network to support multidisciplinary collaborative design. Firstly, from the point of view of collaborative design, the model of constraints network for parameter coordination is studied. In this model, interval boxes are adopted to describe the uncertainty of design parameters quantitatively to support multidisciplinary collaborative and enhance the design robustness. Secondly, a general consistency algorithm is designed to predict conflict and refine the intervals for parameter coordination, which verify the design process early in the process and assist the designers in determining design variables to reduce the multidisciplinary iterations in collaborative design. Finally, the method is demonstrated by a bogie parameter design problem

Knowledge-based engineering (KBE) and simulation analysis have been used widely in multidisciplinary concurrent and collaborative design process. However, the acquisition of knowledge keeps bottleneck yet in building knowledge base in KBE. In this paper, a framework of knowledge discovery from multidisciplinary simulation data is proposed. Correspondingly, a data mining algorithm named fuzzy-rough algorithm is developed to deal with the simulation data by combining the fuzzy set theory and rough set theory. The proposed knowledge discovery process is applied respectively to obtain some useful, implicit production rules with efficient measure. Finally, the method is demonstrated by a metal forming simulation problem. The results prove that knowledge discovery from simulation data is feasible, and the proposed method can be applied in other disciplinary simulation

This paper describes an uncertainty management method using generalized dynamic constraint networks (GDCN). Uncertainty of parameter in the concurrent and collaborative design is analyzed, and GDCN is presented to manage the uncertainty. Firstly, the model of GDCN including domain level constraints and knowledge level constraints is established. Secondly, the modeling of domain level constraints and knowledge level constraints are introduced; thirdly, the clearing up strategy of conflicts among constraints and consistency algorithm are put forward to gain consistency parameter intervals. Finally, a design example is analyzed to show effectiveness of this proposed method