Recently, multi-objective optimization by use of the genetic algorithms (GAs) has been getting a growing interest as a novel approach to the problem. Population based search of GA is expected to find Pareto optimal solutions of the multi-objective optimization problem in parallel. In order to achieve this goal, it is an intrinsic requirement that the evolution process of GA maintains well the diversity of the population in the Pareto optimality set. In this paper, the authors propose to utilize the Thermodynamical Genetic Algorithm (TDGA), a genetic algorithm that uses the concepts of the entropy and the temperature in the selection operation, for multi-objective optimization. Being combined with the Pareto-based ranking technique, computer simulation shows that TDGA can find a variety of Pareto optimal solutions efficiently.
First, a new type of models for trajectory methods to solve optimization problems is considered. In this new model, a velocity of the trajectory is given by a convolution integral form with all gradients of the minimization function on the trajectory for the past time. The new trajectory method can be called a gradient method with the optimizer's “memory” with respect to the past gradient information, and the model can be transformed into the second order differential equation model whose trajectory tides over trapping into local optima under a suitable initial velocity. Next, in order to solve quadratic programming problems with variables constrained on the closed interval [0, 1] 's, the gradient method with “memory” is realized by neural networks as operational circuits composed of neurons, each of which has two integral elements. The trajectory by realized neural networks has possibility to overcome trapping into local minima, while the Hopfield type with first order differential equation model traps into them. Last, the numerical simulation results for simple test problems demonstrate properties of these presented neural networks.
Today, many automobile traffic jams always occur on expressways. A traffic jam burdens drivers with congestion, and causes air pollution. In order to avoid a traffic jam, the traffic information should be offered. Ordinary method which detects the traffic jam requires many traffic detectors on the roads, but these systems are very large and expensive to construct. By using the characteristic lines method, it is possible to estimate the traffic density between two detecting points. This method can be applied on condition that the property of road is uniform, but the property of practical road is not uniform. There are some bottleneck points like tunnels that cause traffic jams. This paper reports the method that determines the change of traffic density at the entrance of bottleneck. As a result, the characteristic lines method is also useful on the long road which includes a bottleneck section.
This paper deals with an identification method based on a model with an automatic choosing function (ACF) for nonlinear systems. A full data region or a whole domain is divided into some subdomains and the unknown nonlinear function to be estimated is approximately described by a linear equation on each subdomain. These linear equations are united into a single one by the ACF smoothly, and thus the resulting model becomes linear in the parameters. Hence these parameters are easily evaluated by the linear least squares method. Moreover the structure of the model by the ACF expansion is properly determined by the genetic algorithm, where the AIC by Akaike is utilized as an objective function. Numerical experiments are carried out to demonstrate the effectiveness of this approach.
Based on the idea of special purpose observers, a new approach to the observer-based stabilization of a class of nonlinear systems by means of I/O linearization is proposed. The class of the system is assumed to be I/O linearizable and Complete Uniformly Local Weakly Observable with an Input-to-State Stable internal dynamics. The strategy considered in this paper is as follows. At first the system of interest is cascaded with a chain of integrators to extend its relative degree to system order, and then observer-based I/O linearization control law is designed to yield stability for the enlarged closed loop system. It is shown that for implementing such a specific control law the linear structured observer is valid.
In this paper, we describe the attitude control of a tumbler system (pendulum system) with three joints. This tumbler system has two purposes. One is experimental verification of various control strategies and another is the use as a model of the walking robot. It's advantage as the device for the experimental verification is its ability to treat from the SIMO system to the MIMO system. It's importance as a model of the walking robot exists in the biped locomotion on the unleveled land where one's feet are unstable. As the first step, we consider a pendulum system with fixed knee joints. Then, we design a stabilizing compensator for this pendulum system. The H∞ loop-shaping design procedure is applied in the design of the stabilization compensator. We herein report the experimental results of a control system.