Abstract-In this paper a robust fault estimation method, based on sliding mode observers, is proposed for a collection of agents undertaking a shared task and exchanging only relative information over a communication network. Since the 'system of systems' formed by the agents is not observable with respect to relative sensing information, by appropriate transformations and scalings of the inputs and outputs of the actual system, a meaningful observable subsystem is created. For this new subsystem, after modal decomposition based on the associated Laplacian, decoupled sliding mode observers, depending only on the individual node level dynamics of the network, can be created exploiting an existing design philosophy. These collectively form a centralized fault estimation scheme for the original system.

SUMMARYA general anti-windup (AW) compensation scheme is provided for a class of input constrained feedback-linearizable nonlinear systems. The controller considered is an inner-loop nonlinear dynamic inversion controller, augmented with an outer-loop linear controller, of arbitrary structure. For open-loop globally exponentially stable plants, it is shown that (i) there always exists a globally stabilizing AW compensator corresponding to a nonlinear generalization of the Internal-Model-Control (IMC) AW solution; (ii) important operator theoretic parallels exist between the AW design scheme for linear control and the suggested AW design scheme for nonlinear affine plants and (iii) a more attractive AW compensator may be obtained by using a nonlinear state-feedback term, which plays a role similar to the linear state-feedback term in linear coprime factor-based AW compensation. The results are demonstrated on a dual-tank simulation example.

Polynomial chaos and Gaussian process emulation are methods for surrogate-based uncertainty quantification, and have been developed independently in their respective communities over the last 25 years. Despite tackling similar problems in the field, to our knowledge there has yet to be a critical comparison of the two approaches in the literature. We begin by providing a detailed description of polynomial chaos and Gaussian process approaches for building a surrogate model of a black-box function. The accuracy of each surrogate method is then tested and compared for two simulators used in industry: a land-surface model (adJULES) and a launch vehicle controller (VEGACONTROL). We analyse surrogates built on experimental designs of various size and type to investigate their performance in a range of modelling scenarios. Specifically, polynomial chaos and Gaussian process surrogates are built on Sobol sequence and tensor grid designs. Their accuracy is measured by their ability to estimate the mean, standard deviation, exceedance probabilities and probability density function of the simulator output, as well as a root mean square error metric, based on an independent validation design. We find that one method does not unanimously outperform the other, but advantages can be gained in some cases, such that the preferred method depends on the modelling goals of the practitioner. Our conclusions are likely to depend somewhat on the modelling choices for the surrogates as well as the design strategy. We hope that this work will spark future comparisons of the two methods in their more advanced formulations and for different sampling strategies.

Abstract-The application of two evolutionary optimization methods, namely, differential evolution and genetic algorithms, to the clearance of nonlinear flight control laws for highly augmented aircraft is described. The algorithms are applied to the problem of evaluating a nonlinear handling quality clearance criterion for a simulation model of a high-performance aircraft with a delta canard configuration and a full-authority flight control law. Hybrid versions of both algorithms, incorporating local gradient-based optimization, are also developed and evaluated. Statistical comparisons of computational cost and global convergence properties reveal the benefits of hybridization for both algorithms. The differential evolution approach in particular, when appropriately augmented with local optimization methods, is shown to have significant potential for improving both the reliability and efficiency of the current industrial flight clearance process.

Physiological simulators which are intended for use in clinical environments face harsh expectations from medical practitioners; they must cope with significant levels of uncertainty arising from non-measurable parameters, population heterogeneity and disease heterogeneity, and their validation must provide watertight proof of their applicability and reliability in the clinical arena. This paper describes a systems engineering framework for the validation of an in silico simulation model of pulmonary physiology. We combine explicit modelling of uncertainty/variability with advanced global optimization methods to demonstrate that the model predictions never deviate from physiologically plausible values for realistic levels of parametric uncertainty. The simulation model considered here has been designed to represent a dynamic in vivo cardiopulmonary state iterating through a mass-conserving set of equations based on established physiological principles and has been developed for a direct clinical application in an intensive-care environment. The approach to uncertainty modelling is adapted from the current best practice in the field of systems and control engineering, and a range of advanced optimization methods are employed to check the robustness of the model, including sequential quadratic programming, mesh-adaptive direct search and genetic algorithms. An overview of these methods and a comparison of their reliability and computational efficiency in comparison to statistical approaches such as Monte Carlo simulation are provided. The results of our study indicate that the simulator provides robust predictions of arterial gas pressures for all realistic ranges of model parameters, and also demonstrate the general applicability of the proposed approach to model validation for physiological simulation.