Automatically Based On The Knowledge Of The Linear Water System Elements Comprehensive Study

On the basis of the researching Map Cognition Theory, this thesis attempts to discuss how to solve the automatic selection of the two typical distributing linear drainage features by using the theory and method of the Artificial Intelligence (AI) and Knowledge Engineering including knowledge representation, reasoning and decision based knowledge. The main researches of this thesis are:1. Discuss the map generalization theory based on Map Cognition Theory. Analyze the challenges of the automated map generalization. Discuss the automated map generalization strategy based on knowledge.2. Discuss the knowledge formalization method of river selection, two kinds of spatial knowledge data models are put forward and the rules of river selection are summed up. Meanwhile, a river system construction method based on master river identification is described. The approach of multiple criteria decision is utilized to identify the master river of each level among the river system. Then, construct the full river system tree structure.3. Put forward the reasoning framework of the river selection based on the constructed information of river system and the rules of river selection. Further more, this thesis set up the reasoning framework to preserve the entireness character of the whole river system.4. This thesis analyzes the character of the complex ditch network, and imports the Dynamic Segmentation theory to organize the ditch network data and its spatial knowledge. Put forward a method based grid to detect the network density.5. This thesis sets up a ditch network selection modal based decision of spatial knowledge. This modal integrates five factors of the ditch selection such as the attribute, length, link condition, spatial relation between ditches and network density. Then, utilizes the multiple criteria decision to evaluate each ditch by these five factors, and make alternated selection to preserve the whole network character.