Intelligent Predictive Control Based on State Analysis of Visual Field for Robot Path Tracking

An intelligent predictive control model based on state analysis of visual field is presented for path tracking of the automated guided vehicle(AGV) with vision navigation.An optimal conversion strategy of error states is used to replace the quadratic cost function and to describe the objective of predictive control,so difficulties in parameter selection of pure algebra optimizing methods can be avoided.A synchronous control approach is proposed to completely eliminate two path errors of the ideal rectification state and to keep an error-free tracking state eventually.For other states,an iterative algorithm with assumption-prediction-adjustment is used to realize a harmonious state conversion.Numerical simulations and tests demonstrate that a realizable speed difference output can always be generated to eliminate two path errors synchronously, quickly and smoothly at different error states and velocities.Furthermore,with its low computational complexity,the proposed algorithm can satisfy the demand from realtime processing of embedded control systems.