Computer-aided path-finding for building design.

Abstract

Path-finding research is an investigation of different kinds of paths and the discovery of the interrelationships between each path-finding object and its environment. A path finding problem deals with how an object finds a continuous route from an initial position to a goal position, avoiding collisions with obstacles along the path. The object can be a person or a distribution system such as water pipes, air ducts, or electrical cables; the environment can be an interstitial space, a room, a floor, or a building; and the path can be a person's route or a set of connected ducts. The purpose of this research was to investigate search algorithms, problem representation, and cost evaluation related to the path-finding problem in an architectural environment. The research was conducted from a geometric viewpoint which emphasized the interrelationship between a path-finding object and obstacles. By examining possible paths, the causes of the path-finding problem can be identified, and judgments can be made at the preliminary stages of the architectural design process. A path which has been efficiently and effectively designed represents a better understanding of different systems and better communication between different expertises. The research was categorized into several tasks: investigating different paths in various building types, identifying common characteristics of paths, establishing a theoretical foundation, determining the costs of a path, implementing a path-finding strategy, and testing the strategy. The path-finding problem was modelled through the representation of the path-finding elements, such as the path-finding object, the path, constraints, space, and the source and goal. Based on the model, paths were classified into five types, evaluated by a cost function, and simulated according to a new path-finding strategy. The strategy consists of three parts: constraint representation, constraint retrieval, and constraint manipulation. This strategy was then tested on three cases: an air duct, a maze, and a complicated floor plan. The research solved a path-finding problem by applying an obstacle-oriented problem-solving strategy. Through the interactions that occur among different kinds of interruptions, a path is thus a record of a path-finding object's response to its environment.