This paper focuses on the experimental identification of a proton exchange membrane water electrolyzer models. Two types of approaches are investigated: temporal and frequency domains identification techniques. The obtained models are validated and compared. A discussion is then provided to highlight the implications of using these identified models, particularly with respect to the choice of operating points, the impact of aging on model complexity, and their relevance for control and diagnostic purposes.

This paper is dedicated to the control of hydrogen production with an experimental proton exchange membrane water electrolyzer in the context of renewable energy sources. Two control laws, $i \mathbf{P}$ and PI controllers have been evaluated under several scenarii including renewable energy sources variations. A discussion of the performances of the controllers allows to formulate several open issues for the control of PEMWE in a renewable energy context.

An active model-free sensor fault tolerant control approach is presented in this paper. The proposed method is based on a model-free controller that has demonstrated an effective ability to work without any analytical model knowledge. The active fault tolerant control procedure has three stages: firstly, the model-free controller is designed using an ultra-local model; secondly, this ultra-local model is used to detect and estimate the sensor fault; thirdly, the obtained estimation is used to adapt the control law according to the sensor fault. The aim of the proposed active fault tolerant control procedure is to ensure that the regulated output, but not the measured one, tracks the desired trajectory despite the occurrence of a sensor fault. The developed method is validated via numerical simulations for both stable and unstable linear systems. The performances of the developed active fault tolerant control procedure for unstable systems are evaluated with and without saturation of the control input.

In this paper, we present two ILC schemes. The first one is a PD-type iterative learning control with an initial state algorithm to solve the trajectory tracking problem for nonlinear systems with uncertainties and without satisfying the classical resetting condition. λ-norm method is used to prove the asymptotic stability of the closed loop system and the simulation results on perturbed nonlinear system have been given. The second approach is a simple P-type iterative learning fuzzy control scheme to solve the trajectory tracking problem for MIMO nonlinear systems. The control design is applicable to deal with a class of nonlinear systems without satisfying the global Lipschitz continuity condition, for which a fuzzy logic term is added to cope with unknown parameters. In addition, the swarm optimization algorithm is used to design the optimum iterative learning fuzzy control (ILFC). Using Lyapunov theory, the asymptotic stability of the closed loop system is guaranteed over the whole finite time. Finally, an illustrative example on two-link manipulator is provided to illustrate the effectiveness of the proposed controller.

This paper presents an iterative learning scheme (PD-type) to solve the trajectory tracking problem for repetitive uncertain nonlinear systems. This scheme consists of two parts, the first is an iterative learning controller and the second is an algorithm which gives us the initial state at each trial. λ-norm method is used to prove the asymptotic stability of the closed loop system. Finally, we apply this controller scheme on perturbed nonlinear system to show its effectiveness.

This paper proposes a novel P-type iterative learning fuzzy control with optimal gains for a class of Multi Input Multi Output (MIMO) nonlinear systems. The control design is very simple, in the sense that we use just a proportional learning action. Another advantage of this proposed controller is that, the global Lipschitz condition is not required for nonlinear systems. Thus, to approximate the unknown nonlinear function, we use a fuzzy logic term. In addition, the swarm optimization algorithm is used to design the optimum iterative learning fuzzy control (ILFC), in the sense that the tracking errors converge at the fastest rate. To prove the asymptotic stability of the closed loop system over the whole finite time, Lyapunov theory is used. Finally and to illustrate the effectiveness of the proposed control scheme, simulation results are presented.

This article explores in detail the influence of the sensor position on the pitch/yaw dynamics modeling and on the autopilot design and performance for a 155mm canard-guided spin-stabilized projectile which incorporates a nose-mounted course correction fuse (CCF) for trajectory correction. A complete and exact nonlinear model is given and used for computing a q-LPV model necessary for the controller synthesis. Using this linearized model, the influence of the sensor position on the load factor-related open-loop dynamics is highlighted. The H∞loop-shaping design approach, which permits to obtain a high-performance, robust, fixed structure and fixed order controller for any operating point, is presented. The necessity, for the controller synthesis, of considering the actual sensor position in the projectile nose and of calculating the load factor feedback signals at the center of gravity (CG) using the load factors actually measured at the CCF, in order to cope with this important practical constraint, is demonstrated.