An approach to intelligent robot motion planning and tracking in known and
static environments is presented in this paper. This complex problem is
divided into several simpler problems. The first is generation of a collision
free path from starting to destination point, which is solved using a
particle swarm optimization (PSO) algorithm. The second is interpolation of
the obtained collision-free path, which is solved using a radial basis
function neural network (RBFNN), and trajectory generation, based on the
interpolated path. The last is a trajectory tracking problem, which is solved
using a proportional-integral (PI) controller. Due to uncertainties, obstacle
avoidance is still not ensured, so an additional fuzzy controller is
introduced, which corrects the control action of the PI controller. The
proposed solution can be used even in dynamic environments, where obstacles
change their position in time. Simulation studies were realized to validate
and illustrate this approach.