Khaled Belarbi

Genetic algorithm for the design of a class of fuzzy controllers: an alternative approach

A simple, easy to implement alternative method for designing fuzzy logic controllers (FLCs) with symmetrically distributed fuzzy sets in a universe of discourse is introduced. The design parameters include the parameters of the membership functions of the inputs and outputs and the rule base. The method is based on a network implementation of the FLC with real and binary weights with constraints. Due to the presence of the binary weights the backpropagation technique cannot be used. The learning problem is cast as a mixed integer constrained dynamic optimization problem and solved using the genetic algorithm (GA). The crossover and mutation are slightly disrupted in order to cope with the constraints on the binary weights. Training of the controller is carried out in a closed-loop simulation with the controller in the loop.

Design of Mamdani fuzzy logic controllers with rule base minimisation using genetic algorithm

This paper presents a design procedure for Mamdani fuzzy logic controller including rule base minimisation. The rules are modelled with binary weights on which constraints are imposed in order to ensure consistency. A genetic algorithm is used for finding stabilising controllers that minimise the number of rules. The cost function includes a stability/performance coefficient which insures that stable, performance satisfying controllers are given the highest possible fitness. The number of fuzzy sets for the input and the control variables are set by the user and the design procedure is concerned only with the rule base and the distribution of the fuzzy sets in the universes of discourses. Two examples were studied: the control of the pole and cart system and the control of the concentration in CSTR. In both cases, the fuzzy sets were isosceles triangles evenly distributed, in the universe of discourses.

A Stable Model-Based Fuzzy Predictive Control Based on Fuzzy Dynamic Programming

A stable model based fuzzy predictive controller based on fuzzy dynamic programming is introduced. The objective of the fuzzy predictive controller is to drive the state of the system to a terminal region where a local stabilizing controller is invoked, leading to a dual mode strategy. The prediction horizon is fixed and specified. The stability of the controlled system is studied using the value function as a Lyapunov function. Guaranteed stability is obtained under conditions on the terminal region, the local control law and the membership functions of fuzzy goal and constraints therein. The solution procedure is based on dynamic programming with branch and bound.

Nonlinear predictive control of a mobile robot: a solution using metaheuristcs

Abstract The basic features of model-based predictive control (MBPC) make it an interesting candidate for the control of mobile robots. However, fast solution procedures remain a challenge for nonlinear MBPC problems such as the one arising in mobile robot control. Metaheuristics are general purpose heuristics which have been successful in solving difficult optimization problems in a reasonable computation time. In this work, we present a comparison between the uses of three different heuristics, namely particle swarm optimization (PSO), ant colony optimization, and gravitational search algorithm for the solution of the nonlinear MBPC for a mobile robot tracking trajectory with dynamic obstacle avoidance. The computation times obtained show that PSO is a feasible alternative for real-time applications. The MBPC based on the PSO is applied to controlling a LEGO mobile robot with encouraged results.

Real-time application of a fuzzy adaptive control to one level in a three-tank system

In this experimental work, a fuzzy combined control method has been described and applied to test the ability of the artificial intelligence technique in controlling nonlinear systems in real-time applications. A direct feedback linearization ideal control law is approximated with a Takagi–Sugeno fuzzy inference system. The adaptation law of the Takagi–Sugeno controller parameters is computed based on a fuzzy approximation term of the control error. This approximation is computed using a Mamdani fuzzy system. The experiment is applied on one level in a three-tank system and a test of the controller ability against perturbations is considered. The obtained results were compared to those leaded by a classical proportional–integral controller. The basic idea of this work is the use of the control error (between the actual control signal and the perfect control signal) instead of the tracking error (between the output of the one level in a three-tank system and the reference signal). As our approach is model free, that is, we do not use the mathematical model of the three-tank system, in the application part of this work, the output of the constructed Takagi–Sugeno fuzzy controller is injected to the real three-tank system.

Single and Multi Objective Predictive Control of Mobile Robots

In this work, we present a comparison between the use of a simple and multi objective MBPC in robots control for tracking trajectories and obstacle avoidance. Two cases were considered, in the first each robot has its own MPC controller where in the second a single two- objectives MPC controller is used for both robots. In the second case; two approaches were proposed to solve the multi objective optimization problem arising in the MOMPC: the multi objective Particle Swarm Optimization (MOPSO) and weighted sum method. The simulation results show that the robots movement is more stable by the MOPSO-NMPC than the PSO-NMPC. Computation times as expected are shorter PSO-NMPC; however MOPSO-NMPC although more time consuming is still feasible.

Accelerated micro particle swarm optimization for the solution of nonlinear model predictive control

Purpose n n n n nRapid solution methods are still a challenge for difficult optimization problems among them those arising in nonlinear model predictive control. The particle swarm optimization algorithm has shown its potential for the solution of some problems with an acceptable computation time. In this paper, we use an accelerated version of PSO for the solution of simple and multiobjective nonlinear MBPC for unmanned vehicles (mobile robots and quadcopter) for tracking trajectories and obstacle avoidance. The AµPSO-NMPC was applied to control a LEGO mobile robot for the tracking of a trajectory without and with obstacles avoidance one. n n n n nDesign/methodology/approach n n n n nThe accelerated PSO and the NMPC are used to control unmanned vehicles for tracking trajectories and obstacle avoidance. n n n n nFindings n n n n nThe results of the experiments are very promising and show that AµPSO can be considered as an alternative to the classical solution methods. n n n n nOriginality/value n n n n nThe computation time is less than 0.02 ms using an Intel Core i7 with 8GB of RAM.

Nonlinear model predictive control of quadcopter

In this paper, a quadcopter is controlled by a nonlinear model predictive controller, NMPC, for trajectory tracking in presence of an external perturbation. The nonlinear model predictive control was basically confined to slow processes. Applications to fast processes such as robots are rare because the time for the solution may exceed the sampling period. Metaheuristics have been used for solving many difficult problems. In this work, we consider the application of the Particle Swarm Optimization algorithm to the NMPC optimisation problem model applied for the quadcopter tracking trajectory with presence of an external perturbation. Results show that NMPC-PSO provides a fast solution and can be used in real time.

An adaptive fuzzy control scheme with application in flexible spacecraft control

We introduce a direct fuzzy adaptive control for the large angle rotational movement and vibration suppression of a flexible spacecraft. A Takagi Sugeno inference system is used to approximate a feedback linearization control law. The adaptation mechanism is based on estimate of the error between the ideal unknown control signal and the actual control signal using a Mamdani fuzzy inference system. Simulation results show precise trajectory control and vibration suppression of the flexible spacecraft.

Nonlinear model predictive control based on state dependent Riccati equation

In this work, we introduce an approach to nonlinear model predictive based on the so-called state dependent Riccati equation, SDRE. In this approach, the model is first cast in a form similar to the linear state space representation. Then the algebraic Riccati equation is constructed based on a similarity with the linear quadratic regulator to obtain stable NMPC. The method requires the solution of the Riccati equation at each sampling period. Simulation results are quite encourageing.