Research On The Constraint Solving For Direct Manipulation Of Feature Model

Feature modeling is product-oriented and it is an important milestone of CAD development, which really makes integrated CAD/CAPP/CAM possible. It is also the theory and technological foundation of solving products problems. In recent years, the structure of CAD software is undertaking significant reforms, ch-anging from solid modeling centered systems which are called" geometric CAD" to the feature and constraint based, which are called" applied CAD".As an indispensable part in any CAD systems, technology concerning constraint model-ing is highly emphasized, and many useful solutions have been raised which encourage the reform and application of CAD systems.In current commercial feature modelling systems, support for direct manipulation of features is not commonly available. This is partly due to the lack of speed of current constraint solvers, but also to deficient interactive facilities.In this dissertation, an approach to the optimization of geometric constraint solving for direct manipulation of feature dimensions, orientation, and position is described. Firstly, we analyse the direct manipulation of features, direct manipulation of a real-valued feature parameter consists of two phases: selection phase and interaction phase. As to the characteristics of direct manipulation of features, we presents a new technique that the original constraint model is compiled into a manipulation constraint model and each geometric variable and geometric constraint of the high-level constraint graph must be mapped to the Kramer graph as one or more variables or constraints, during the mapping, determine which constraints are rigid and which are not. The presented constraint solver-driven model compilation approach not only considerably reduced the number of constraints that have to be solved during interaction, it also avoids the need for a separate constraint model analysis algorithm. At the same time, presenting degrees of freedom analysis to solve over-and underconstrained models, optimally solves constraints, so can obtain the real-time feedback on the consequences of the operation and enhance the intuitiveness of user interaction. Finally, this approach is elementarily applied to constraint modeling of HUST-CAID system.