An optical analogue computing system which can discriminate shapes and/or sizes of small objects and/or letters has been developed using by a Multiplexed Matched Spatial Filter (MMSF) and the optical system composed on an optical rail base for making hologram automatically. The system discriminates the shapes and/or sizes of particles and/or letters instantaneously and simultaneously.
The non-destrutive inspection of cracks in concrete structures has been attempted using electromagnetic wave, infrared ray, sound wave and so on. Inspection using electromagnetic wave sometimes did not detect cracks, since concrete structures contain iron bars. The infrared ray was hard to use in tunnels, because difference in temperature was hardly observed before and behind the cracks. It was also difficult to detect cracks using an ultrasonic wave because of the reflected waves strictly superposing on each other. This paper proposes two non-destructive inspection methods using ultrasonic wave which detect the cracks in concrete structures. One is based on a multi-reflected-wave model, the other a stationary wave model. The methods enable not only detection of the cracks but also an accurate measurement of their positions. These methods should be selected depending on the sensor to be used and the object to which the inspection is applied.
This paper considers robust stabilization problem for plants with uncertainties/nonlinearities satisfying a certain class of IQC conditions. The restricted class of IQC conditions is still fairly wide. We show this problem reduces to feasibility of nonconvex matrix inequalities. Owing to their structure, they can be solved, by combining LMI and a covering-type algorithm, with reasonable computational effort compared with BMI. Illustrative numerical examples applied to gain-phase uncertainties are given.
In this paper, we perform the seismic control experiment of a flexible structure in order to investigate applicability of the model set identification methods to robust controller design. From the viewpoint of application to robust control, we identify model sets based on nominal models represented by rational transfer functions, in contrary to most existing methods based on affinely parameterized functions which tend to be high order functions. Also, we derive a less conservative error bound which takes account of low correlation properties of input signals. Based on the model set identification method and the H∞ controller design method using LMIs, we show that a closed-loop system which satisfies given design specifications is obtained systematically.
This paper studies the minimization of the l1 norm of a closed loop map with time domain constraints of the transient response for a fixed input. A computational method to reduce the problem to finite dimensional linear programs is proposed. The framework uses the DA method which existed for the standard l1 control without time domain constraints. Additional advantages such as no use of rank interpolation condition and no assumptions on FIR feasible solution are inherited to the time domain constraints version. Examples are included to show the efficacy of the proposed method.
This paper gives necessary and sufficient conditions for multiple input discrete-time linear systems to be controllable with input constraints due to saturation and positiveness. These input constraints are very important because they are often encountered in many practical control applications such as optimal drug administration problem, tracer kinetics in medical system, electrically heated oven system, population dynamics, or ecology, etc. Furthermore a simple method to test whether a given system is controllable or not with input constraints is proposed based on the Jordan canonical form by using the elimination method of Gauss and transforming the state and input variables into some particular form. The particular form proposed here contains more informations about the structure of controllable systems with input constraints than that obtained earlier. The results obtained here are expected to become basic frameworks to solve many practical control problems mentioned above.
ILQ optimal servo design method has been generalized so far in many ways. Noting that a general control structure of robust servo systems has a free feedforward parameter, such a free parameter is introduced here in the usual ILQ design method for further generalization. As a result, the proposed method inherits two characteristics of an existing ILQ design method; one is that we can specify directly a desired closed loop transfer function from reference input to output in a larger class than usual; the other is that design results are expressed analytically in terms of system parameters and design specifications. Moreover, the proposed method can decouple those servo systems of type one that cannot be decoupled by state feedback and hence inapplicable by the usual ILQ design. Finally we clarify the class of transfer functins which can be decoupled in general servosystems. The class shows there is essentially no difference between ILQ servo systems and general servo systems.
This paper considers the linear quadratic optimal control problem for linear implicit systems based on the dissipation inequality. We derive a necessary and sufficient condition for the dissipativity with respect to a quadratic supply rate in terms of a linear matrix inequality (LMI) condition with an equality constraint. Based on this constrained LMI condition, the optimal control law is given by an implicit algebraic constraint among system variables. We also show that the present constrained LMI condition easily reduces to an LMI without any equality constraints.
In this paper, we demonstrate a computational complexity analysis of the global optimization for the Matrix Product Eigenvalue Problem (MPEP). A global optimization scheme for the MPEP is compared with other nonconvex optimization problems such as Convex Multiplicative Programming (CMP) and branch and bound methods for solving Bilinear Matrix Inequality (BMI) problems. We show that convex subproblems for the MPEP global optimization algorithm are solved more efficiently than those of the BMI branch and bound methods. Numerical experiments illustrate that the MPEP global optimization algorithm achieves less number of iterations and CPU-time than the BMI branch and bound methods even in the total computational complexity.
We discuss the controller design strategy for a class of 3-state and 2-input nonholonomic affine systems. We propose an algorithm of the chained form transformation for 3-state and 2-input symmetric affine systems, and design a controller. We also consider 3-state and 2-input affine nonholonomic systems with a drift term. By applying the algorithm of the chained form transformation to an affine system with a drift term, we obtain the chained form with a drift term which is named expanded chained form. We design a controller based on this expanded chained form. Nextly we propose the coordinate and input transformation to convert a class of 3-state and 2-input affine nonholonomic systems with a drift term into the time-state control form. Finally, an asteroid sample-return robot is considered as an example. To demonstrate the validity of the proposed controllers, simulation and experiment have been carried out.
This paper deals with applicability of an autonomous decentralized algorithm for a loading problem. This problem is an optimization problem for loading several kinds of boxes on three pallets in a motor truck by taking account of safety, availability of capacity and easiness of carrying in and out. The goal of the problem is to attain the most favorite arrangement of boxes on three pallets for various constraints. In the problem formulation, some constraints are treated explicitly, and the others are included in the objective as the penalty functions. This loading problem has two important requests. One is to level the top of boxes loaded, in consideration of safety. The other is to load boxes up to the height of a truck, in consideration of easiness of carrying in and out. Two algorithms are proposed for the respective requests. Numerical results are shown to investigate the performance of these algorithms.
In this paper, we discuss on new Coevolutionary Genetic Algorithm for Constraint Satisfaction. Our basic idea is to explore effective genetic information in the population, i.e., schemata, and to exploit the genetic information in order to guide the population to better solutions. Our Coevolutionary Genetic Algorithm (CGA) consists of two GA populations; the first GA, called “H-GA”, searches for the solutions in a given environment (problem), and the second GA, called “P-GA”, searches for effective genetic information involved in the H-GA, namely, good schemata. Two kinds of the fitness evaluation methods for the P-GA are introduced. We then applied our CGA to Constraint Satisfaction Problems (CSPs) incorporating a stochastic “repair” operator for the P-GA to raise the consistency of schemata with the (local) constraint conditions in CSPs. Various computer simulations on general CSPs, dynamic CSPs and cluster-structured CSPs elucidate the effectiveness of our approach.
This paper is about humanitarian demining activities on anti-personnel mines, such as the problems face during mines detecting and the difficulties of conducting such activities. In present time, there are about 80 million of unmanaged ant-ipersonnel mines left in conflict ended countries throughout in the world. As result, many noncombatants die during peacetime due to the mines and a great amount of moneys is spent on mines management. Through this paper, we suggest some technical solutions on this problem.
Two different concepts of “symmetry” are found in the literature of control systems theory, one related to reciprocity and the other to group symmetry. For robust stabilization of symmetric plants in either sense, symmetric controllers are known to be more effective. This paper shows a common mechanism of this phenomenon.
This paper concerns the number of zeros of a polynomial inside a given circle in the left half plane. Marden provided an algorithm to compute the number of zeros of a polynomial in a unit circle centered at the origin by using only the coefficients of the given polynomial. In this short article, it will be shown that the result can easily be extended to the case where the circle, with its center and radius arbitrary, is located on the left half complex plane.
In this study, a variable damper comprising electro-rheological fluid is equipped to the joint of pneumatic rubber artificial muscle manipulator. Some experiments have confirmed that the ER damper can improve the stability of manipulator in the high gain control.