Research On Key Algorithms Of Residential Planning And Their Implementation

With the development of computer technology, Computer Aided Design(CAD) hasspread into all aspects of modern technology. At the same time, custom softwares for specialfield of industry are also developed by combining CAD and professional knowledge,including the field of urban plan and architecture design. However, due to the process ofurban plan and architecture design refers to lots of details in design criterion and professionalempirical evidence, most of CAD softwares in the field of architecture only serve the drawingof special graphic modules and the calculations of regular formula. Base on the residentialplan CAD system, the dissertation researches and implements the layout algorithm forbuildings in the residential area, the main road on the residential area and database storage ofthe residential data. The main work of the dissertation is as follows:Firstly, the dissertation abstracted the layout model according to the residential buildingslayout problem, and transformed the practical problem into2D layout problem. Referring tothe concept and algorithm of no-fit polygon in the terms of polygon placement problem, thealgorithm of the inner available area was proposed in order to avoid the overlap between theshape of building and the edge of the residential area. In the light of the difference betweenthe layout of mono-style buildings and multi-style buildings, the different simplified modelswere constructed in this dissertation. The perimetric and elongated automatic layout for themono-style buildings was implemented by applying the bricklaying rule and the intelligentlayout for the multi-style buildings was implemented by applying the genetic algorithm. Theresults of computer simulation experiment show the proposed methods is effective andpractical in the building layout problem.Second, the dissertation has introduced the basic content and characteristic of the mainroad on the residential area. The MAKLINK Graph theory was utilized to construct the freespace model of the residential area, and the Floyd algorithm was adopted to find asub-optimal collision-free path based on MAKLINK Graph. Then, the genetic algorithm wasapplied to optimize the location of the sub-optimal path so as to generate the globally optimalpath. The result of computer simulation experiment shows the proposed method is effectiveand practical in the main road planning problem on the residential area.At last, the dissertation has introduced the functional modules and the data storagestructure in the residential planning system in which the algorithms were applied. Also, thedissertation has showed the implements of the algorithms and other assisted methods.