Search

This paper addresses the problems of establishing structural stability and the L2 gain of 2D mixed continuous-discrete-time systems. The first contribution is to show that Lyapunov functions quadratically dependent on the frequency are exact for establishing structural stability. This is particularly important since the existing works that exploit Lyapunov functions provide a much larger upper bound on the dependence on the frequency or other parameters. The second contribution is to propose a novel linear matrix inequality (LMI) necessary and sufficient condition for establishing the existence of such Lyapunov functions. It is shown, analytically and through several examples, for both best and worst cases, that the numerical complexity of this novel condition is much smaller than that of the existing methods. The third contribution is to show that the proposed methodology can be used to establish upper bounds on the L2 gain, in particular, deriving a novel necessary and sufficient LMI condition based on Lyapunov functions quadratically dependent on the frequency. Lastly, the paper presents the generalization of the proposed methodology to non-mixed 2D systems.

This paper considers the transmission of an infinite sequence of messages (a streaming source) over a packet erasure channel, where every source message must be recovered perfectly at the destination subject to a fixed decoding delay. While the capacity of a channel that introduces only bursts of erasures is well known, only recently, the capacity of a channel with either one burst of erasures or multiple arbitrary erasures in any fixed-sized sliding window has been established. However, the codes shown to achieve this capacity are either non-explicit constructions (proven to exist) or explicit constructions that require large field size that scales exponentially with the delay. This work describes an explicit rate-optimal construction for admissible channel and delay parameters over a field size that scales only quadratically with the delay., arXiv admin note: text overlap with arXiv:1903.07434

Semi-active suspension systems have become a widespread tool to improve the handling and comfort of vehicles. These systems require adjustable dampers as well as supplementary sensing elements. In addition to displacements and accelerations, some of the best performing approaches require knowledge of the semi active damper force. Since this variable can be difficult and expensive to measure, several estimation methods have been proposed. In this article, two Linear-Parameter-Varying H ∞ (LPV- H ∞ ) filters are developed to estimate the Semi-Active (SA) damper force, considering two different combinations of sensing elements: the first configuration is more expensive, but potentially more accurate and reliable; whereas the second configuration is cheaper and arguably less reliable. Thanks to the use of LPV- H ∞ theory, both filters are designed to account for the main nonlinear phenomena of SA dampers (i.e. saturation, hysteresis, etc.), as well as being quadratically stable, robust to the road disturbances and optimized to reduce the estimation error in a specified frequency band. Simulations and experimental data are used to assess the proposed estimators as well as a typical inverse-dynamics estimation approach. The results show that while both of the proposed estimators yield a good degree of accuracy, there are indeed fundamental differences depending on the available sensing elements; a conclusion which could be crucial to appropriately define the instrumentation of semi-active suspension systems.

In this article, we study constrained online convex optimization (OCO) in the presence of feedback delays, where a decision maker chooses sequential actions without knowing the loss functions and constraint functions a priori . The loss/constraint functions vary with time and their feedback information is revealed to the decision maker with delays, which arise in many applications. We first consider the scenario of delayed function feedback, in which the complete information of the loss/constraint functions is revealed to the decision maker with delays. We develop a modified online saddle point algorithm suitable for constrained OCO with feedback delays. Sublinear regret and sublinear constraint violation bounds are established for the algorithm in terms of the delays. In practice, the complete information (functional forms) of the loss/constraint functions may not be revealed to the decision maker. Thus, we further examine the scenario of delayed bandit feedback, where only the values of the loss/constraint functions at two random points close to the chosen action are revealed to the decision maker with delays. A delayed version of the bandit online saddle point algorithm is proposed by utilizing stochastic gradient estimates of the loss/constraint functions based on delayed bandit feedback. We also establish sublinear regret and sublinear constraint violation bounds for this bandit optimization algorithm in terms of the delays. Finally, numerical results for online quadratically constrained quadratic programs are presented to corroborate the efficacy of the proposed algorithms.

A numerical method is a mathematical tool designed to solve numerical problems. The implementation of a numerical method in a programming language is called a numerical algorithm. This chapter provides a basic review of some of the commonly used iterative numerical algorithms for solving nonlinear problems. It briefly goes through the algorithms of Gauss–Seidel, predictor‐corrector, Newton’s method, and the particle swarm optimization algorithm. The chapter also summarizes the approximate errors for Gauss–Seidel and Newton’s methods. It clearly shows that the error of Newton’s method decreases quadratically (square of the error) whereas that of the Gauss–Seidel method decreases linearly. Newton’s method is a very powerful method as the convergence of the solution is quadratic, which means that the difference between the root and the approximation is squared at each iteration step.

This paper deals with the robust numerical method for solving the singularly perturbed semilinear partial differential equation with the spatial delay. The quadratically convergent quasilinearization technique is used to linearize the semilinear term. It is formulated by discretization of the solution domain and then replacing the differential equation by finite difference approximation that in turn gives the system of algebraic equations. The method is shown to be first-order convergent. It is observed that the convergence is independent of the perturbation parameter. Numerical illustrations are investigated on model examples to support the theoretical results and the effectiveness of the method.

This paper investigates the problem of observer-based $$H_\infty $$ control for one-sided Lipschitz nonlinear systems subject to parametric uncertainties and external disturbances. In order to relax some conservatisms and limitations of the traditional Lipschitz condition, the one-sided Lipschitz and quadratically inner-bounded conditions are used. On the contrary to the methods proposed in the literature, our method allows for uncertainty in the input matrix B, as well as the dynamic and output matrices A and C. To derive design conditions in terms of LMIs, the well-known Young’s relation is employed for handling the bilinear terms naturally arising in observer-based controller design. Finally, two examples are presented to demonstrate the validity of the theoretical results.

We consider a parametric family of quadratically constrained quadratic programs (QCQP) and their associated semidefinite programming (SDP) relaxations. Given a nominal value of the parameter at which the SDP relaxation is exact, we study conditions (and quantitative bounds) under which the relaxation will continue to be exact as the parameter moves in a neighborhood around the nominal value. Our framework captures a wide array of statistical estimation problems including tensor principal component analysis, rotation synchronization, orthogonal Procrustes, camera triangulation and resectioning, essential matrix estimation, system identification, and approximate GCD. Our results can also be used to analyze the stability of SOS relaxations of general polynomial optimization problems., 23 pages, 3 figures

In this paper, we propose a smoothing Levenberg-Marquardt method for the symmetric cone complementarity problem. Based on a smoothing function, we turn this problem into a system of nonlinear equations and then solve the equations by the method proposed. Under the condition of Lipschitz continuity of the Jacobian matrix and local error bound, the new method is proved to be globally convergent and locally superlinearly/quadratically convergent. Numerical experiments are also employed to show that the method is stable and efficient.

Anisotropic fiber-reinforced composites are used in lightweight construction, which is of great industrial relevance. During mold filling of fiber suspensions, the microstructural evolution of the local fiber arrangement and orientation distribution is determined by the local velocity gradient. Based on the Folgar–Tucker equation, which describes the evolution of the second-order fiber orientation tensor in terms of the velocity gradient, the present study addresses selected states of deformation rates that can locally occur in complex flow fields. For such homogeneous flows, exact solutions for the asymptotic fiber orientation states are derived and discussed based on the quadratic closure. In contrast to the existing literature, the derived exact solutions take into account the fiber-fiber interaction. The analysis of the asymptotic solutions relying upon the common quadratic closure shows disadvantages with respect to the predicted material symmetry, namely, the anisotropy is overestimated for strong fiber-fiber interaction. This motivates us to suggest a novel normalized fully symmetric quadratic closure. Two versions of this new closure are derived regarding the prediction of anisotropic properties and the fiber orientation evolution. The fiber orientation states determined with the new closure approach show an improved prediction of anisotropy in both effective viscous and elastic composite behaviors. In addition, the symmetrized quadratic closure has a simple structure that reduces the effort in numerical implementation compared to more elaborated closure schemes.

We propose Ellipsoidal One-Class Constraint Acquisition (EOCCA), a fast and scalable algorithm for the acquisition of constraints for Mixed-Integer Quadratically Constrained Programming (MIQCP) models from data. EOCCA acquires a well-formed MIQCP model using solely the examples of the feasible solutions to this model. It combines x-means partitioning, standardization, and principal components analysis to preprocess the training set and then wraps the preprocessed data into several hyper-ellipsoids expressed using MIQCP constraints. These MIQCP constraints are projected back to the space of the original training set, and their further use does not require data preprocessing. Experimental evaluation shows that EOCCA scores better than a state-of-the-art algorithm in terms of fidelity of the acquired constraints to ground-truth constraints and achieves this in few orders of magnitude shorter time. We demonstrate the practical use case of EOCCA in a fully automated workflow of modeling and optimization of a rice farm using real-world data.

Under the assumption that uncertain coefficients corresponding to each constraint are perturbed in an ellipsoidal set, we consider the problem of maximizing the perturbation radius of the ellipsoidal set associated to a robust convex quadratically constrained quadratic programming problem to maintain some properties of a pre-decision. To this end, a fractional programming problem is first formulated to solve the problem, and then equivalently reformulated into linear conic programs over positive semi-definite, second-order cones that are solvable in polynomial time. Numerical experiments in connection with the robust Markowitz’s portfolio selection problem are provided to demonstrate the proposed concept of sensitivity analysis. Additionally, certain numerical results are also presented to compare the efficiency of direct solutions of the proposed linear conic programs with that of a bisection method for the corresponding fractional programming problem.

In this paper, we construct the utility-based optimal hedging strategy for a European-type option in the Almgren-Chriss model with temporary price impact. The main mathematical challenge of this work stems from the degeneracy of the second order terms and the quadratic growth of the first order terms in the associated HJB equation, which makes it difficult to establish sufficient regularity of the value function needed to construct the optimal strategy in a feedback form. By combining the analytic and probabilistic tools for describing the value function and the optimal strategy, we establish the feedback representation of the latter. We use this representation to derive an explicit asymptotic expansion of the utility indifference price of the option, which allows us to quantify the price impact in options' market via the price impact coefficient in the underlying market.

This paper presents a quadratically constrained multiobjective quadratic fractional programming model (MOQFPM) and proposed a methodology to obtain a best preferred solution with the help of parametric functions and using fuzzy goal programming. In the initial stage, we obtain a non-fractional optimization model from the multi-objective quadratic fractional programming model by assigning a vector of parameters to fractional functions. Then, in the next stage, we use fuzzy goal programming approach to obtain the best preferred solution for the decision maker to the optimization problem by finding membership functions and aspiration levels of each objective function. This methodology proposes an efficient method to obtain Pareto-optimal solution of MOQFPM.

In this paper we consider the weighted nonlinear complementarity problem (denoted by wNCP) which contains a wide class of optimization problems. We introduce a family of new weighted complementarity functions and show that it is continuously differentiable everywhere and has several favorable properties. Based on this function, we reformulate the wNCP as a smooth nonlinear equation and propose a nonmonotone Levenberg–Marquardt type method to solve it. We show that the proposed method is well-defined and it is globally convergent without any additional condition. Moreover, we prove that the whole iteration sequence converges to a solution of the wNCP locally superlinearly or quadratically under the nonsingularity condition. In addition, we establish the local quadratic convergence of the proposed method under the local error bound condition. Some numerical results are also reported.

The paper considers nonconvex separable quadratic optimization problems subject to inequality constraints. A sufficient condition is given for finding the value and the point of the global extremum of a problem of this type by calculating the Lagrange dual bound. The peculiarity of this condition is that it is easily verified and requires from the Hessian matrix of the Lagrange function only that its region of positive definiteness is not empty. The result obtained for the dual bound also holds for the bound obtained using SDP relaxation. References Shor, N. Z., Stetsenko, S. I. (1989). Quadratic extremal problems and nondifferentiable optimization. Naukova Dumka, Kiev. Berezovskyi, O. A. (2017). Zero duality gap in quadratically constrained quadratic programming. Mathematical and computer modelling. Series: Physical and mathematical sciences, 15, 20-25. Nesterov, Y., Wolkowicz, H., Ye, Y. (2000). Semidefinite programming relaxations of nonconvex quadratic optimization. Handbook of semidefinite programming, Springer, New York, 361-419. DOI doi.org/10.1007/978-1-4615-4381-7_13 Berezovskyi, O. A. (2016). Exactness criteria for SDP-relaxations of quadratic extremum problems. Cybernetics and Systems Analysis, 52(6), 915-920. DOI doi.org/10.1007/s10559-016-9893-3

Explicit model predictive control is an established methodology for the offline determination of the optimal control policy for linear discrete time-invariant systems with linear constraints. Nevertheless, nonlinearities in the form of quadratic constraints naturally appear in process models or are imposed for stability purposes in model predictive control formulations. In this manuscript, we present the theoretical developments and propose an algorithm for the exact solution of explicit nonlinear model predictive control problems with convex quadratic constraints. Our approach is based on a second-order Taylor approximation of Fiacco’s Basic Sensitivity Theorem, which allows for the existence and the analytic derivation of the optimal control actions. The complete exploration of the parameter space is founded on an active set strategy, which employs a pruning criterion to eliminate infeasible active sets. Based on that, the optimal map of solutions is constructed along with the corresponding control actions. The proposed strategy is applied to an explicit nonlinear model predictive control problem with an ellipsoidal terminal set, and comparisons with approximate solutions are drawn to demonstrate the benefits of the presented approach. Furthermore, as a practical application, the optimal operation of a chemostat in the presence of disturbances is exhibited.