Algorithm Research On Parking Lots Layout Problem Based On Segmentation

Thanks to the rapid economic development and urbanization of China,the amount of automobiles in urban areas has risen dramatically in recent years.Meanwhile,the lack of parking space is becoming increasingly acute.Underground parking lots are effective solutions to this dilemma.In addition,optimizing their layout would bring more parking spaces,contributing to an even better cure.At present,the design process of underground parking lots layout is closely associated with the experience of designers,leading to fluctuating results.Furthermore,algorithms of parking lots layout cannot tackle the difficulty caused by complex contours and obstacles that constitute underground parking lots.According to what has been mentioned above,the characteristics of this problem are analyzed,then rules are extracted.Finally,a segmentation-based three-stage algorithm,which can automatically generate layout designs for underground parking lots,is proposed.Genetic Algorithm is applied in stage one,to fulfill parking lots layout in marginal areas,and strategies are developed to overcome algorithm prematurity.As for stage two,zone segmentation is further divided into two phases,segmentation and clustering,solving by mesh segmentation and greedy algorithm separately.The last stage is completed via traversal.Parameter experiments are conducted first to explore full potential of Genetic Algorithm and greedy algorithm.Then case studies are done based on the chosen optimal parameters.The results demonstrate that the segmentation-based three-stage algorithm for underground parking lots layout problem is able to increase the parking space capacity,and can be finished in reasonable time.