基于粒度分层模型的布局算法

Layout Algorithm Based on the Hierarchical Granularity Model

研究了问题求解的粒度理论在布局求解中的应用 .综合约束问题求解和回溯控制策略 ,给出了在粒度分层模型基础上的布局算法 .以往采用单一粒度模型的布局算法 ,由于粒度过粗或过细 ,不能充分利用各种约束来降低求解复杂度 .文中提出的算法采用了粒度分层模型 ,能够利用设计过程不同阶段的约束 ,来缩小合适解的范围 ,通过多层面的约束 ,降低求解的复杂度 ,克服了以往布局算法的缺陷 .

The application of granularity theory in the domain of layout design problem solving was discussed. A layout algorithm based on the multi granularity model was proposed by adopting strategies of both constraint problem solving and backtracking. The constraints were not sufficiently used in previous layout algorithms because they usually adopted singular granularity model that was either too raw to let design process go further or too fine to cut search branches earlier, thus the complexity was inevitable. In the proposed algorithm a multi granularity model was adopted and constraints of different design periods can be used in the algorithm. Hence the drawbacks of former approaches were overcome. A case study in a house layout problem was carefully analyzed using the algorithm.