Fuzzy motion controllers and hybrids

作     者：Gerdelan, Anton 

作者单位：Massey University 

学位级别：doctor

导师姓名：Professor Ken Hawick

授予年度：2011年

主      题：Fuzzy logic Motion control systems Computer animation Computer simulation Computer algorithms 

摘      要：This thesis describes implementations of motion control systems that are based on fuzzy logic; fuzzymotion controllers. The controllers are used by to drive a variety of simulated vehicles and computeranimatedcharacters. The problem of heading towards a destination whilst simultaneously avoidingstatic and dynamic obstacles is addressed with fuzzy motion controllers. For situations where a levelabove reactive motion control is required, such as path-planning behaviour or traffic rule following,then hybrid algorithms are proposed; combining fuzzy motion controllers with other navigationalgorithms. Consideration is given to behavioural level of detail models, with transition between behaviouralmodels of different complexity based on the proximity, or visual importance of charactersto the *** controllers have a set of fuzzy rules, or a “rule base that defines the inference of the *** is no assurance that hand-calibrated rule bases are optimal, and indeed that calibrationbased on fixed test environment will apply well to a dynamic environment. Special considerationis given to genetic-fuzzy systems, which use a genetic algorithm to automatically calibrate a rulebase. Various architectures for genetic-fuzzy system are proposed and evaluated including dynamicsystems, which have the ability to learn “on the fly, rather than in fixed experiment scenarios. A relationshipbetween genetic algorithm parameters and time-efficient fitness improvement is *** requirements of training more complex “cascading fuzzy systems are discussed. Distributedand parallel training models are also considered.A new, modular agent middleware is proposed, which is the underpinning software that perceivesthe complex environment, feeds inputs into the fuzzy motion controllers, and effects output actionsfor each character and vehicle. The middleware model is successfully used to drive a range ofvehicles and characters used in *** problem of evaluating motion