Research On Kernel Density Estimation Constrained By Linear Unit

Most of the various point-like things in geospatial space are closely related to linear elements,such as towns on both sides of the main line,villages on both sides of the river,and attractions on both sides of the tourist route.The analysis of events or activities constrained by linear unit is more advantageous in terms of accuracy,credibility,and application accuracy.The existing analysis methods for spatial points' elements mostly use a single point as the analysis subject,and there are few researches on the spatial analysis methods of point events constrained by linear unit,and the algorithm is much more complicated and difficult to implement.In this paper,the linear kernel density analysis method is taken as an example.Multi-source media data(mainly based on tourism data)and different scale road network data are used as data sources to study the kernel density analysis method under linear constraints.The aim is to provide new effective and practical quantitative methods for tourism planning management,personalized route recommendation for scenic spots,intelligent route identification,and even urban research fields,so that it can dig more detailed information.Through the study of the traditional kernel density estimation,we retain the core algorithm and linearly constrain the point datasets involved in the calculation by linear cell gridding.A kernel density algorithm under linear constraints is proposed.The algorithm implementation process mainly includes:(1)Projecting spatial discrete point data onto the line according to fixed rules to ensure that the point elements are online;(2)Grading the linear cells and ensuring the spatial adjacency relationship between each grid in the linear unit after gridding and the attribute table;(3)Using the gridded linear space unit as the calculation subject,calculating the number of event points falling into each grid,and calculating the kernel density value as the occurrence element;(4)Taking the kernel density value as the visual attribute field,and based on the linear space unit for visualization.The horizontal width of the line is used as an important indicator for identifying the hot zone.According to the above steps,the calculation and application of the kernel density value constrained by linear unit are basically realized.The algorithm is proved by the national road network of different scales and the Pingyao ancient city road network of the scenic spot.The results show that:(1).The linear kernel density algorithm can well indicate the hotspot area at the road segment level,and accurately locate the ?fuzzy hot zone? identified in the traditional method to a fixed road segment to obtain more detailed information;(2)Comparing with the network kernel density algorithm proposed by the previous scholars,the linear kernel density algorithm improves the operation efficiency under the premise of meeting the scale requirement,and the algorithm is relatively simple,easy to expand,easy to use,and can meet the needs of scholars with weak programming foundation.(3)The linear kernel density algorithm is based on the grid idea,and the spatial discrete point data is associated with its closest linear unit,and the practicability is enhanced.