Study Of Key Technique And Application Of The Optimal Route Based On Multi-scale Road Network

The optimal route planning is an important function module of vehicle intelligent navigation system and logistics transportation system. In order to solve the contradiction between vast traffic navigation data and limited computer resource, the optimal route planning algorithm of large region becomes research hotspot of intelligent transportation system and geographic information system at all times. Aiming at the problem, experts research much and put forward a lot of algorithms. Up to now, the problem can not get a perfect solving yet.The paper researches key technique of the optimal route planning of large region based on multi-scale road network, and the main research contents include below:1. Based on spatial cognitive theory, the rule of people's visual resolution with scale changing is proposed in this paper: aiming at certain scale geographic spatial data, before visual scale deflates (1/ x )1 /( x?1 ) times, the ratio of graphics spare space under non-generalization and graphics spare space under generalization is less than scale changing. When scale deflating multiple exceeds (1/ x )1 /( x?1 ) multiple of original scale, the ratio of graphics spare space under non-generalization and graphics spare space under generalization toboggans. So the paper draws the conclusion: aiming at certain scale geographic spatial data, generalization need perform when visual scale deflates (1/ x )1 /( x?1 ) times in order to ensure visual definition.2. Based on the scale changing rule of spatial data and national essential scale series, consulting geographic information system multi-level data model of Beer S, the paper constructs road network multi-level data model, which makes a multi-scale data system.3. The paper analyzes a lot of typical schemes which are much database much edition scheme, a database much edition scheme, a database an edition scheme, Level of Detail scheme and layered, leveled and partitioned scheme, and it proposes a multi-scale road network scheme and realizes it with an appropriate spatial data model.4. Based on spatial data model which follows Geographic Data File Standard and organizes data based on dataset, section and layer, the paper proposes a physical memory structure suited for incredent updating.5. The paper analyzes many matching methods of cognominal entity which are geometry matching, topology matching, semantics matching, compound matching and road matching, and it establishes correlative relation between adjoining scale road navigation data, which provides technical sustaining for establishing multi-scale road network topology structure.6. The paper put forward a method which dynamic establishes multi-scale road network topology structure and a multi-scale route planning algorithm. The test experiment indicated that this algorithm could meet the requirement of vehicle automatic navigation system.