مجله کنترل
سال سیزدهم شماره 3 (پاییز 1398)

In this paper, an optimization problem for the observer trajectory in the bearings-only surface moving target tracking (BOT) is studied. The BOT depends directly on the observability of the target's position in the target/observer geometry or the optimal observer maneuver. Therefore, the maximum lower band of the Fisher information matrix is opted as an independent criterion of the target estimator. First, modeling of the optimal control problem of the observer path is presented based on the orthogonal Chebyshev polynomial. Then, a control law for the observer direction, which is independent of the initial conditions, is obtained using the direct numerical optimization. The advantages of the proposed model include maximization of the total maneuver time, calculation of the control law at the start time of the maneuver, and high flexibility in applying the tracking constraints of the observer's motion. The efficiency of the proposed algorithm is compared with the conventional path optimization methods using the Monte Carlo. In addition, the performance of the algorithm is evaluated in different scenarios for target tracking, including remote, near, moving, and stationary, and its reliability is investigated. It is also applied in the surface submarine tracking problem using sonar.

This paper presents a new approach to estimate and to enlarge the domain of attraction for a planar continuous-time piecewise affine system. Various continuous Lyapunov functions have been proposed to estimate and to enlarge the system’s domain of attraction. In the proposed method with a new vision and with the aids of a discontinuous piecewise quadratic Lyapunov function, the domain of attraction at the origin is enlarged by designing a state feedback controller. This paper shows that the continuity of the Lyapunov function on the boundaries, increases the conservativeness in estimating the domain of attraction, and gives more powerful search ability to the domain of attraction estimation algorithm by relaxing this continuity condition. The simulation results show the superiority of the proposed method so that using this method the larger estimation of the domain of attraction is obtained than continuous one.

There are many reports of using the Kinect to detect hand and finger gestures after release of device by Microsoft. The depth information is mostly used to separate the hand image in the two-dimension of RGB domain. This paper proposes a method in which the depth information plays a more dominant role. Using a threshold in depth space first the hand template is extracted. Then in 3D domain the perpendicular vector to the hand surface is found. Using the rotation matrix all the rotations along three axes are compensated in a way that the camera z- coordinate lies perpendicular to hand surface. Then the resulted 3d image is translated to a distance of 80 to 100 cm from the Kinect. Wavelet transform with a new descriptor, called Circular Descriptor are used to extract required features. A trained MLP neural network in conjunction with a SVM is used to classify the signs. Empirical results show an average accuracy of 96.7 % with a two seconds delay for online recognition of Persian Sign Language.

In this paper we introduce some stability criteria of nonlinear hybrid systems with time delay described by impulsive hybrid fuzzy system of differential equations. Firstly, a comparison principle for fuzzy differential system based on a notion of upper quasi-monotone nondecreasing is presented. Here, for stability analysis of fuzzy dynamical systems, vector Lyapunov-like functions are defined. Then, by using these functions together with the new comparison theorem, we will get results for some concepts of stability (eventual stability, asymptotic stability, strong stability and uniform stability) for impulsive hybrid fuzzy delay differential systems. Furthermore, theorems for practical stability in terms of two measures are introduced and are proved. Finally, an illustrating example for stability checking of a differential system with fuzziness and time delay is given. Then, by introducing an applied example in Pharmacokinetics, we bridge theoretical concepts to the application of research in real world.

In this paper, an ANFIS+PID hybrid control policy has been addressed to control a 6-degree-of freedom (6-DOF) robotic manipulator. Then its error convergence has been also evaluated. The ability to formulate and estimate the system uncertainties and disturbances along with system dynamics and rejecting the disturbances effect are some advantages of the proposed method in   comparing with the conventional ANFIS structures. The error convergence could not be proved in the ordinary ANFIS structures. But in the proposed method, the error convergence of the robot manipulator can be established under considering some mathematical conditions. The proposed control law is realized via parallel combination of ordinary ANFIS network and PID controller. The suggested method has been successfully applied in a 6-DOF robot manipulator system. Furthermore, in presence of uncertainties and  external  disturbances  error  convergence  would  be  justified  using  the Lyapunov-like  theorem and Barbalat lemma.

This paper investigates Itô-type stochastic linear quadratic controller design for uncertain model of vehicle suspension. The Itô stochastic model of quarter-car is constructed considering parametric stochastic perturbations in stiffness and damping characteristics of suspension. To tackle with uncertainties of model, a stochastic optimal control law is obtained applying stochastic Hamilton-Jacobi-Bellman equation. By means of Itô lemma and stochastic extension of Lyapunov method, stochastic stability of the closed-loop system is guaranteed. The stochastic optimal controller is designed for a general form of Itô uncertain model which is comprised multi-dimensional multiplicative perturbations and then it is implemented on perturbed model of vehicle suspension. Furthermore, it is shown that the separation principal does not hold for the system with state multiplicative noise; therefore, the synthesized observer-based controller guarantees the stability of augmented dynamic consists of system and estimation error dynamics. A simulation study is performed to evaluate the effectiveness of stochastic optimal control approach in satisfying objectives of active suspension. To this end, time and frequency responses of ride comfort and road holding characteristics are demonstrated for two specific road cases including sinusoidal bump and ISO random profile.

The aim of this paper is to propose a novel method for controlling a class of parameter-varying systems by controlling interval observer. The interval observers, that are applicable for the systems with uncertainty, estimate bounds on the states instead of estimating the states. It has been shown that by designing appropriate control inputs for controlling the bounds of the interval observer, the states of the system can be also controlled using the same control inputs. For this purpose, a suitable interval observer is firstly designed for the parameter-varying system and the required conditions, for which the dynamical system consisting of lower and upper bounds on error is monotone, are presented. Then, a novel controller is designed for stabilizing the interval observer such that the bounds on the states are stabilized and thereby the states are also stabilized. The proposed controller is based on adaptive sliding mode control method that is utilized to tackle the effects of variations in some parameters of the interval observers as well as the existing disturbances in the system. By choosing an appropriate Lyapunov function, the conditions and areas for the stability of the observer are determined. Simulation results, obtained by applying the method to a sample system, show the effectiveness of the proposed method.