Najah Yousfi

Optimized fuzzy controller for mobile robot navigation in a cluttered environment

Obstacle avoidance and path planning are the most important problems in mobile robots. In this paper, a fuzzy logic controller has been constructed in order to train an intelligent robot. Gradient method is used to optimize consequences of a Sugeno fuzzy logic controller for the mobile robot navigation, in order to reach a target in a clutterd environment. Not only simulation results are shown in this paper, but also the “real-time” implementation has been realized onto the mini robot Khepera II. Simulation results verify successfully the application of the proposed method to real motion situations.

Path tracking design by input/output Davidson-Cole transfer function

Autonomous vehicle navigation requires the integration of many technologies such as path tracking. In path tracking design, we must take into account the dynamic of actuators in order to reduce overshoots appearing for small displacements. It permits the generation of an optimal movement reference-input giving a minimum path completion time, taking into consideration the maximum velocity, acceleration and jerk, and the bandwidth of the closed-loop on which the input is applied. Different methods were studied. The work presented here focuses on the fractional differentiation prefilter applied to nonvarying plants. A new approach to path tracking using fractional differentiation by direct optimization of an I/O transfer function was developed. Fractional differentiation is used here through a Davidson-Cole filter (DC) whose main property is having no overshoot on the plant output, whatever the values of its constitutive parameters. A simulation on a motor model validates the methodology.

Path tracking design based on I/O Davidson-Cole transfer function for square MIMO systems

Abstract The aim of this paper concerns motion control and robust path tracking design for square MIMO systems. A path tracking approach using fractional differentiation is proposed, based on direct optimization of an I/O Davidson-Cole type transfer function. The main important property of this transfer function is that it does not present overshoot on its output and allows synthesis with only two parameters to obtain an optimal settling time. The SISO approach was presented in previous works. This paper presents an extension for square multivariable systems. Using MIMO-QFT robust synthesis methodology and taking into account the plant uncertainties, the fractional prefilters are deduced. This method specifically deals with the appropriate transformation of the MIMO system to the equivalent MISO system. The optimization of the two prefilter parameters is founded on the plant output error integral minimization, considering the actuators physical constraints and the tracking performance specifications. A numerical example is considered to indicate performances of the used procedure.