Modelling formation and growth of H2SO4-H2O aerosols: Uncertainty analysis and experimental evaluation

An aerosol dynamics model, AERO2, is presented, which describes the formation of H₂SO₄-H₂O aerosol in a smog chamber. The model is used to analyse how the uncertainties on four input parameters are propagated through an aerosol dynamics model. The input parameters are: the rate of the reaction between SO₂ and OH (k₁), the ratio between the nucleation rate used in AERO2 and that derived from classical nucleation theory (t_n), the H₂SO₄ mass accommodation coefficient () and a measure of the turbulence intensity in the reactor (k_e). Uncertainties for these parameters are taken from the literature. One of the results of the analysis is that AERO2 and aerosol dynamics models in general can only predict upper bounds for the total number (N_tot) and total volume (V_tot) concentrations of the particles. The uncertainties on N_tot and V_tot are mainly due to the uncertainties on k₁, and t_n. An uncertainty factor of 20–100 still remains when the uncertainty on k₁, is reduced to ±5%. Aerosol measurements from three smog chamber experiments have therefore been used, in an attempt to reduce the uncertainty on k₁ and t_n. Values for k₁ are obtained in the reduced range 7.8 × 10⁻¹³ to 1.0 × 10⁻¹² cm³ s⁻¹, which is within the range found in the literature. For t_n, values in the range 10⁴–10⁷ are obtained, which is close to the upper bound of the range in literature. These values for t_n are in marked contrast with a recent set of experiments on nucleation in H₂SO₄-H₂O mixtures, which suggests a value for t_n of at most 10⁻⁵.     

On input‐to‐state stability‐based design for leader/follower formation control with measurement delays

This paper addresses the design of low‐level controllers for leader–follower formations of nonholonomic vehicles in the presence of bounded measurement delays. The concept of input‐to‐state stability is extended to encompass the effect of bounded delays and restrictions on the input. A method is proposed to integrate a Smith predictor in a backstepping design on the basis of nested saturations and nonlinear small‐gain assignment, which allows for time delays in the feedback loop. Robustness analysis under uncertain bounded time delays is provided, and design tradeoffs resulting from the use of bounded controls are discussed. Illustrative simulations are shown to validate the design and robustness analysis in the context of a simple leader–follower trailing control problem. Copyright © 2012 John Wiley & Sons, Ltd.     

Locally linear discriminant embedding: An efficient method for face recognition

In this paper an efficient feature extraction method named as locally linear discriminant embedding (LLDE) is proposed for face recognition. It is well known that a point can be linearly reconstructed by its neighbors and the reconstruction weights are under the sum-to-one constraint in the classical locally linear embedding (LLE). So the constrained weights obey an important symmetry: for any particular data point, they are invariant to rotations, rescalings and translations. The latter two are introduced to the proposed method to strengthen the classification ability of the original LLE. The data with different class labels are translated by the corresponding vectors and those belonging to the same class are translated by the same vector. In order to cluster the data with the same label closer, they are also rescaled to some extent. So after translation and rescaling, the discriminability of the data will be improved significantly. The proposed method is compared with some related feature extraction methods such as maximum margin criterion (MMC), as well as other supervised manifold learning-based approaches, for example ensemble unified LLE and linear discriminant analysis (En-ULLELDA), locally linear discriminant analysis (LLDA). Experimental results on Yale and CMU PIE face databases convince us that the proposed method provides a better representation of the class information and obtains much higher recognition accuracies.     

Web proxy cache replacement scheme based on back-propagation neural network

Web proxy caches are used to reduce the strain of contemporary web traffic on web servers and network bandwidth providers. In this research, a novel approach to web proxy cache replacement which utilizes neural networks for replacement decisions is developed and analyzed. Neural networks are trained to classify cacheable objects from real world data sets using information known to be important in web proxy caching, such as frequency and recency. Correct classification ratios between 0.85 and 0.88 are obtained both for data used for training and data not used for training. Our approach is compared with Least Recently Used (LRU), Least Frequently Used (LFU) and the optimal case which always rates an object with the number of future requests. Performance is evaluated in simulation for various neural network structures and cache conditions. The final neural networks achieve hit rates that are 86.60% of the optimal in the worst case and 100% of the optimal in the best case. Byte-hit rates are 93.36% of the optimal in the worst case and 99.92% of the optimal in the best case. We examine the input-to-output mappings of individual neural networks and analyze the resulting caching strategy with respect to specific cache conditions.     

A bacterial colony growth algorithm for mobile robot localization

Achieving robot autonomy is a fundamental objective in Mobile Robotics. However in order to realize this goal, a robot must be aware of its location within an environment. Therefore, the localization problem (i.e.,the problem of determining robot pose relative to a map of its environment) must be addressed. This paper proposes a new biology-inspired approach to this problem. It takes advantage of models of species reproduction to provide a suitable framework for maintaining the multi-hypothesis. In addition, various strategies to track robot pose are proposed and investigated through statistical comparisons. The Bacterial Colony Growth Algorithm (BCGA) provides two different levels of modeling: a background level that carries on the multi-hypothesis and a foreground level that identifies the best hypotheses according to an exchangeable strategy. Experiments, carried out on the robot ATRV-Jr manufactured by iRobot, show the effectiveness of the proposed BCGA.

Controller design for flexible structure vibration suppression with robustness to contacts

Model-based feedback control of vibration in flexible structures can be complicated by the possibility that interaction with an external body occurs. If not accounted for, instability or poor performance may result. In this paper, a method is proposed for achieving robust vibration control of flexible structures under contact. The method uses robust linear state feedback, coupled with a state estimation scheme utilizing contact force measurement. Uncertain contact characteristics are modelled by a sector-bounded non-linear function, such that state feedback gains can be synthesized using a matrix inequality formulation of the Popov stability criterion. A separation theorem is used to establish a robust H₂ cost bound for the closed loop system. Experimental results from a multi-mode flexible structure testbed confirm that vibration attenuation and stability can be maintained over a broad range of contact characteristics, in terms of compliance and clearance.     

Optimal experimental design and some related control problems

This paper traces the strong relations between experimental design and control, such as the use of optimal inputs to obtain precise parameter estimation in dynamical systems and the introduction of suitably designed perturbations in adaptive control. The mathematical background of optimal experimental design is briefly presented, and the role of experimental design in the asymptotic properties of estimators is emphasized. Although most of the paper concerns parametric models, some results are also presented for statistical learning and prediction with nonparametric models.     

Supervisory control of modular systems with global specification languages

The paper presents sufficient conditions for modular (supervisory) control synthesis to equal global control synthesis. In modular control synthesis a supervisory control is synthesized for each module separately and the supervisory control consists of the parallel composition of the modular supervisory controls. The general case of the specification that is indecomposable and not necessarily contained in the plant language, which is often the case in practice, is considered. The usual assumption that all shared events are controllable is relaxed by introducing two new structural conditions relying on the global mutual controllability condition. The novel concept used as a sufficient structural condition is strong global mutual controllability. The main result uses a weaker condition called global mutual controllability together with local consistency of the specification. An example illustrates the approach.     

Control of frictional dynamics of a one-dimensional particle array

Control of frictional forces is required in many applications of tribology. While the problem is approached by chemical means traditionally, a recent approach was proposed to control the system mechanically to tune frictional responses. We design feedback control laws for a one-dimensional particle array sliding on a surface subject to friction. The Frenkel-Kontorova model describing the dynamics is a nonlinear interconnected system and the accessible control elements are average quantities only. We prove local stability of equilibrium points of the un-controlled system in the presence of linear and nonlinear particle interactions, respectively. We then formulate a tracking control problem, whose control objective is for the average system to reach a designated targeted velocity using accessible elements. Sufficient stabilization conditions are explicitly derived for the closed-loop error systems using the Lyapunov theory based methods. Simulation results show satisfactory performances. The results can be applied to other physical systems whose dynamics is described by the Frenkel-Kontorova model.     

Adaptive robust posture control of a parallel manipulator driven by pneumatic muscles

Rather severe parametric uncertainties and uncertain nonlinearities exist in the dynamic modeling of a parallel manipulator driven by pneumatic muscles. Those uncertainties not only come from the time-varying friction forces and the static force modeling errors of pneumatic muscles but also from the inherent complex nonlinearities and unknown disturbances of the parallel manipulator. In this paper, a discontinuous projection-based adaptive robust control strategy is adopted to compensate for both the parametric uncertainties and uncertain nonlinearities of a three-pneumatic-muscles-driven parallel manipulator to achieve precise posture trajectory tracking control. The resulting controller effectively handles the effects of various parameter variations and the hard-to-model nonlinearities such as the friction forces of the pneumatic muscles. Simulation and experimental results are obtained to illustrate the effectiveness of the proposed adaptive robust controller.