In this paper we present an efficient method to solve reachability problems for Petri nets based on genetic algorithms and a kind of random search which is called postpone search. Genetic algorithm is one of algorithms developed for solving several problems of optimization. We apply GAs and postpone search to approximately solving reachability problems. This approach can not determine exact solutions, however, from applicability points of view, does not directly face state space explosion problems and can extend class of Petri nets to deal with very large state space in reasonable time. First we describe how to represent reachability problems on each of GAs and postpone search. We suppose the existence of a nonnegative parickh vector which satisfies the necessary reachability condition. Possible firing sequences of transitions induced by the parickh vector is encoded on GAs. We also define fitness function to solve reachability problems. Reachability problems can be interpreted as an optimization ones on GAs. Next we introduce random reachability problems which are capable of handling state space and the number of firing sequences which enable to reach a target marking from an initial marking. State space and the number of firing sequences are considered as factors which effect on the hardness of reachability problems to solve with stochastic methods. Furthermore, by using those random reachability problems and well known dining philosophers problems as benchmark problems, we compare GAs' performance with the performance of postpone search. Finally we present empirical results that GAa is more useful method than postpone search for solving more harder reachability problems from the both points of view; reliability and efficiency.

The problem of genetic algorithm's efficiency has been attracting the attention of genetic algorithm community. Over the last decade, considerable researches have focused on improving genetic algorithm's performance. However, they are generally under the framework of natural evolutionary mechanism and the major genetic operators, crossover and mutation, are activated by the prior probabilities. An operator based on a prior probability possesses randomness, that is, the unexpected individuals are frequently operated, but the expected individuals are sometimes not operated. Moreover, as the evaluation function is the link between the genetic algorithm and the problem to be solved, the evaluation function provides the heuristic information for evolutionary search. Therefore, how to use this kind of heuristic information (present and past) is influential in the efficiency of evolutionary search. This paper, as an attempt, presents a eugenics-based genetic algorithm (EGA) -- a genetic algorithm that reflects the human's decision will (eugenics), and fully utilizes the heuristic information provided by the evaluation function for the decisions. In other words, EGA = evolutionary mechanisms + human's decision will + heuristic information. In EGA, the ideas of the positive eugenics and the negative eugenics are applied as the principle of selections and the selections are not activated by the prior probabilities but by the evaluation values of individuals. A method of genealogical chain-based selection for mutation is proposed, which avoids the blindness of stochastic mutation and the disruptive problem of mutation. A control strategy of reasonable competitions is proposed, which brings the effects of crossover and mutation into full play. Three examples, the minimum problem of a standard optimizing function--De Jong's test function F2, a typical combinatorial optimization problem--the traveling salesman problem, and a problem of identifying nonlinear system, are given to show the good performance of EGA.

This letter presents an efficient design method of multiplierless 2-D state-space digital filters (SSDFs) based on a genetic algorithm. The resultant multiplierless 2-D SSDFs, whose coefficients are represented as the sum of two powers-of-two terms, are attractive for high-speed operation and simple implementation. The design problem of multiplierless 2-D SSDFs described by Roesser's local state-space model is formulated subject to the constraint that the resultant filters are stable. To ensure the stability for the resultant 2-D SSDFs, a stability test routine is embedded in th design procedure.

A design method of 2-D lattice digital filter using the Genetic Algorithm (GA) is proposed. By using the GA. 2-D all-pole lattice filter with the cascade connection of transversal (all-zoro) filter is designed directly from a given desired frequency responce.

In this paper, we present a vision system for a depalletizing robot which recognizes carton objects. The algorithm consists of the extraction of object candidates and a labeling process to determine whether or not they actually exist. We consider this labeling a combinatorial optimization of labels, we propose a new labeling method applying Genetic Algorithm (GA). GA is an effective optimization method, but it has been inapplicable to real industrial systems because of its processing time and difficulty of finding the global optimum solution. We have solved these problems by using the following guidelines for designing GA: (1) encoding high-level information to chromosomes, such as the existence of object candidates; (2) proposing effective coding method and genetic operations based on the building block hypothesis; and (3) preparing a support procedure in the vision system for compensating for the mis-recognition caused by the pseudo optimum solution in labeling. Here, the hypothesis says that a better solution can be generated by combining parts of good solutions. In our problem, it is expected that a global desirable image interpretation can be obtained by combining subimages interpreted consistently. Through real image experiments, we have proven that the reliability of the vision system we have proposed is more than 98% and the recognition speed is 5 seconds/image, which is practical enough for the real-time robot task.

In this study, an extraction method of failure sound signal which is strongly contaminated by noise is investigated by genetic algorithm and statistical tests of the frequency domain for the failure diagnosis of machinery. In order to check the extraction accuracy of the failure signal and obtain the optimum extraction of failure signal, the "existing probability Ps (t*k) of failure signal" and statistical information Iqp are defined as the standard indices for evaluation of the extraction results. It has been proven by practical field data and application of the inspection and diagnosis robot that the extraction method discussed in this paper is effective for detection of a failure and distinction of it's origin in the diagnosis of machinery.

Genetic Algorithm (GA) is the method of approaching optimization problem by modeling and simulating the biological evolution. As the genetic algorithm is rather time consuming, the use of a parallel genetic algorithm can be advantage. This paper describes new methods for fine-grained parallel genetic algorithm using a multiprocessor system FIN. FIN has a VLSI-oriented interconnection network, and is constructed from a viewpoint of fractal geometry so that self-similarity is considered in its configuration. The performance of the proposed methods on the Traveling Salesman Problem (TSP), which is an NP-hard problem in the field of combinatorial optimization, is compared to that of the simple genetic algorithm and the traditional fine-grained parallel genetic algorithm. The results indicate that the proposed methods yield improvement to find better solutions of the TSP.

A number of flexible control mechanisms used in buffer management, congestion control and bandwidth allocation have been proposed to improve the performance of ATM networks by introducing parameters, such as threshold, push-out probability and incremental bandwidth size of a virtual path, which are adjustable by network providers. However, it is difficult to adaptively adjust these parameters, since the traffic in ATM networks is further complicated by accommodating various kinds of services. To overcome the problem, we propose in this paper a control scheme based on the genetic algorithms and the neural estimator. The neural estimator forecasts the future QOS values for each candidate parameter set, and the genetic algorithms select the best one to control the real network. An example of buffer management in an ATM switch is examined in this paper. Simulation results show the effectiveness of the proposed control scheme in adaptively adjusting the parameter set even when the traffic environment and the QOS requirements are dynamically changing.

Genetic algorithms have been shown to be very useful in a variety of search and optimization problems. In this paper, we propose a modified genetic channel router. We adopt the compatible crossover operator and newly designed compatible mutation operator in order to search solution space more effectively, where vertical constraints are integrated. By carefully selected fitness function forms and optimized genetic parameters, the current version speeds up benchmarks on average about 5.83 times faster than that of our previous version. Moreover the total convergence to optimal solutions for benchmarks can be always obtained.

An active contour model which is called Snakes was proposed to extract the border line of an object from an image. This method presents the minimization problem of the energy function defined on the contour curve. The authors obtained an excellent result by applying genetic algorithm to the contour extraction. In this paper, the biphased genetic algorithm, which is a new type of genetic algorithm, is proposed to minimize the energy function of Snakes. The parameters of the genetic algorithm are examined to tune up its local and global search abilities. The biphased genetic algorithm composed of two phases of genetic search is constructed to use both abilities of the exploration and the exploitation properties of the genetic algorithm. The processing results of the biphased genetic algorithm are compared with those of the previous methods, and the advantages of the proposed algorithm are shown by several experiments.

The classical supervised learning algorithms for optimizing multi-layered feedforward neural networks, such at the original back-propagation algorithm, suffer from several weaknesses. First, they have the possibility of being trapped at local minima during learning, which may lead to failure in finding the global optimal solution. Second, the convergence rate is typically too slow even if the learning can be achieved. This paper introduces a new learning algorithm which employs a genetic-type search during the learning phase of back-propagation algorithm so that the above problems can be overcome. The basic idea is to evolve the network weights in a controlled manner so as to jump to the regions of smaller mean squared error whenever the back-propagation stops at a local minimum. By this, the local minima can always be escaped and a much faster learning with global optimal solution can be achieved. A mathematical framework on the weight evolution of the new algorithm in also presented in this paper, which gives a careful analysis on the requirements of weight evolution (or perturbation) during learning in order to achieve a better error performance in the weights between different hidden layers. Simulation results on three typical problems including XOR, 3-bit parity and the counting problem are described to illustrate the fast learning behaviour and the global search capability of the new algorithm in improving the performance of back-propagated network.

We develop a convergence theory of the simple genetic algorithm (SGA) for two-bit problems (Type I TBP and Type II TBP). SGA consists of two operations, reproduction and crossover. These are imitations of selection and recombination in biological systems. TBP is the simplest optimization problem that is devised with an intention to deceive SGA into deviating from the maximum point. It has been believed that, empirically, SGA can deviate from the maximum point for Type II while it always converges to the maximum point for Type I. Our convergence theory is a first mathematical achievement to ensure that the belief is true. Specifically, we demonstrate the following. (a) SGA always converges to the maximum point for Type I, starting from any initial point. (b) SGA converges either to the maximum or second maximum point for Type II, depending upon its initial points. Regarding Type II, we furthermore elucidate a typical sufficient initial condition under which SGA converges either to the maximum or second maximum point. Consequently, our convergence theory establishes a solid foundation for more general GA convergence theory that is in its initial stage of research. Moreover, it can bring powerful analytical techniques back to the research of original biological systems.

Recently, the back propagation method, which is one of the algorithms for learning neural networks, has been widely applied to various fields because of its excellent characteristics. But it has drawbacks, for example, slowness of learning speed, the possibility of falling into a local minimum and the necessity of adjusting a learning constant in every application. In this article we propose an algorithm which overcomes some of the drawbacks of the back propagation by using an improved genetic algorithm.

Genetic algorithms have been shown to be very useful in a variety of search and optimization problems. In this paper, we describe the implementation of genetic algorithms for channel routing problems and identify the key points which are essential to making full use of the population of potential solutions, that is one of the characteristics of genetic algorithms. Three efficient crossover techniques which can be divided further into 13 kinds of crossover operators have been compared. We also extend our previous work with ability to deal with dogleg case by simply splitting multi-terminal nets into a series of 2-terminal subnets. It routes the Deutsch's difficult example with 21 tracks without any detours.

Evolution programs have been shown to be very useful in a variety of search and optimization problems, however, until now, there has been little attempt to apply evolution programs to channel routing problem. In this paper, we present an exolution program and identify the key points which are essential to successfully applying evolution programs to channel routing problem. We also indicate how integrating heuristic information related to the problem under consideration helps in convergence on final solutions and illustrate the validity of out approach by providing experimental results obtained for the benchmark tests. compared with the optimal solutions.

A genetic method to generate a neural network which has both structure and connection weights adequate for a given task is proposed. A neural network having arbitrary connections is regarded as a virtual living thing which has genes representing its connections among neural units. Effectiveness of the network is estimated from its time sequential input and output signals. Excellent individuals, namely appropriate neural networks, are generated through generation iterations. The basic principle of the method and its applications are described. As an example of evolution from randomly generated networks to feedforward networks, an XOR problem is dealt with, and an action control problem is used for making networks containing feedback and mutual connections. The proposed method is available for designing a neural network whose adequate structure is unknown.

This paper deals with an adaptive credit assignment algorithm to select strategies having higher capabilities in the learning classifier system (LCS) based upon the genetic algorithm (GA). We emulate a kind of prizes and incentives employed in the economies with imperfect information. The compensation scheme provides an automatic adjustment in response to the changes in the environment, and a comfortable guideline to incorporate the constraints. The learning process in the LCS based on the GA is realized by combining a pair of most capable strategies (called classifiers) represented as the production rules to replace another less capable strategy in the similar manner to the genetic operation on chromosomes in organisms. In the conventional scheme of the learning classifier system, the capability s(k, t) (called strength) of a strategy k at time t is measured by only the suitableness to sense and recognize the environment. But, we also define and utilize the prizes and incentives obtained by employing the strategy, so as to increase s(k, t) if the classifier provide good rules, and some amount is subtracted if the classifier k violate the constraints. The new algorithm is applied to the portfolio management. As the simulation result shows, the net return of the portfolio management system surpasses the average return obtained in the American securities market. The result of the illustrative example is compared to the same system composed of the neural networks, and related problems are discussed.