Adaptive behavior analysis of WMR based on chaotic phase-space reconstruction

The chaotic theory of phase-space reconstruction is applied to the adaptive behavior analysis and construction of the robot due to the complex nonholonomic and non-linear characteristics of WMR (Wheeled Mobile Robot). The WMR and the operating environment as a whole, a set of control rules for the adaptive behavior which can produce local optimal planning path are obtained by QL (Q_Learning). Then a set of time series of the adaptive behavior are sampled utilizing the density of obstacles as a variable in the running environment and the phase-space is reconstructed. At last the data in the phase-space are discussed and analyzed based on the chaotic time series theory. The simulation results show that the one dimension adaptive time series sampled according to the obstacles density demonstrate the chaotic characteristics. So the chaotic theory can be used in the adaptive behavior analysis and forecast and provides an adequate theoretical basis for the more precise control to WMR.