In this paper, we propose a new global optimization model which is described as discrete time-varying inertial gradient dynamics. The proposed model has a autonomous damping term adjustment structure which takes the search history into consideration. In the proposed model, its dynamics destabilizes autonomously when its search point approaches to the best point in the search history. Thus, stagnation of search around the neighborhood of the best point is restrained, in consequence, global search is continued. Furthermore, in this paper, we analytically explain a characteristic of the search trajectory generated by the inertial gradient system. The characteristic is the implementation of intensive and diverse search in the attracting region for the inertial gradient system. Consequently, we propose a multi-point type search model in which search points are attracted to a promising region, where objective function value is small, by a coupling structure in order to make the attracting region the promising region. We confirm effectiveness of the proposed model through numerical experiments using several benchmark problems.

View full abstract

In this study, we propose a distributed cooperative scheduling method, and apply the method into a machine maintenance scheduling problem in re-entrant production systems. As one of the distributed cooperative scheduling methods, we focus on Lagrangian decomposition and coordination (LDC) method, and formulate the machine maintenance scheduling problem with LDC so as to improve computational efficiency by decomposing an original scheduling problem into several sub-problems. The derived solutions by solving the decomposed dual problem are converted into feasible solutions with a heuristic procedure applied in this study. The proposed approach regards maintenance as job with starting and finishing time constraints, so that product and maintenance schedule can realize proper maintenance operations without losing productivity. We show the effectiveness of the proposed method in several simulation experiments.

View full abstract

In this paper, the optimal control problem of probabilistic Boolean networks (PBNs) is discussed. A PBN is one of the significant models in biological networks. Although some control methods of PBNs have been proposed so far, it is necessary to compute the state transition diagram with 2ⁿ nodes for a given PBN with n states. To avoid this computation, an integer programming-based approach is proposed. In the proposed method, PBNs are transformed into a linear system with binary variables, and the optimal control problem is reduced to an integer linear programming problem, which can be computed relatively easier than the existing methods.

View full abstract

In real flocks, it was revealed that the angular density of nearest neighbors shows a strong anisotropic structure of individuals by very recent extensive field studies [Ballerini et al, Proceedings of the National Academy of Sciences USA, 105, pp. 1232-1237 (2008)]. In this paper, we show this structure of anisotropy also emerges in an artificial flock simulation, namely, Boid simulation. To quantify the anisotropy, we evaluate a useful statistics, that is to say, the so-called γ-value which is defined as an inner product between the vector in the direction of the lowest angular density of flocks and the vector in the direction of moving of the flock. Our results concerning the emergence of the anisotropy through the γ-value might enable us to judge whether an optimal flock simulation seems to be realistic or not.

View full abstract

We have investigated several characteristics of keystroke dynamics in Japanese free text typing. We performed experiments on 189 subjects, representing three groups according to the number of letters they could type in five minutes. In this experiment, we extracted the feature indices from the keystroke timing for each alphabet single letter and for two-letter combinations composed of consonant and vowel pairs in Japanese text. Taking into account two identification methods using weighted Euclidean distance (WED) and Vector Disorder (VD), we proposed their hybrid model for individual identification based on keystroke data in Japanese free text typing. By evaluating the personal identification for the three groups, its high performance was confirmed in proportion to the typing level of the group.

View full abstract

In this paper, we propose a learning method of recurrent spiking neural networks by using particle swarm optimization (PSO). The existing learning method by using the gradient based method sometimes falls into local minima. The proposed method aims to find the global optimum regardless of initial solutions. Since PSO can treat optimization problems in which the objective function is a non-differentiable function, in this paper we formulate learning problems with such objective functions and propose a learning method based on PSO to solve them. The proposed method makes it possible to treat not only the learning problem in terms of firing instants which the conventional method treats, but also one in terms of the number and frequency of firings.

View full abstract

Advanced driver assistance systems (ADAS) such as a forward obstacle collision warning system (FOCWS) and a night vision enhancement system (NVES) aim to decrease driver's mental workload and enhance vehicle safety by provision of useful information to support driver's perception process and judgment process. On the other hand, the risk homeostasis theory (RHT) cautions that an enhanced safety and a reduced risk would cause a risk compensation behavior such as increasing the vehicle velocity. Therefore, the present paper performed the driving simulator experiments to discuss dependence on the NVES and emergence of the risk compensation behavior. Moreover, we verified the side-effects of spontaneous behavioral adaptation derived from the presentation of the fuel-consumption meter on the risk compensation behavior.

View full abstract

DNA molecules autonomously change their forms from the single strand to the double helix by specific binding between complementary sequences according to the Watson-Crick base pairing rule. This paring rule allows us to control connections among molecules and to construct various structures by sequence design. Further, the motion of constructed structures can also be designed by considering sequential bindings. Recently, the feasibility to utilize the programmed DNA structural change for information processing was studied. In the present paper, we report an efficient synthetic chain reaction based on autonomous binding of DNA to realize a computing system, which enable us to implement computational intelligence in vitro.

View full abstract

This paper presents a fully decentralized control inspired by plasmodium of true slime mold and its validity using a soft-bodied amoeboid robot. The notable features of this paper are twofold: (1) the robot has truly soft and deformable body stemming from real-time tunable springs and a balloon, the former is utilized as an outer skin of the body and the latter serves as protoplasm; and (2) a fully decentralized control using coupled oscillators with completely local sensory feedback mechanism is realized by exploiting the long-distance physical interaction between the body parts induced by the law of conservation of protoplasmic mass. Experimental results show that this robot exhibits truly supple locomotion without relying on any hierarchical structure. The results obtained are expected to shed new light on design scheme for autonomous decentralized control system.

View full abstract

Recently, synthesis of gene networks having desired functions has become of interest to many researchers because it is a complementary approach to understanding gene networks, and it could be the first step in controlling living cells. There exist several periodic phenomena in cells, e.g. circadian rhythm. These phenomena are considered to be generated by gene networks. We have already proposed synthesis method of gene networks based on gene expression. The method is applicable to synthesizing gene networks possessing the desired cyclic expression pattern sequences. It ensures that realized expression pattern sequences are periodic, however, it does not ensure that their corresponding solution trajectories are periodic, which might bring that their oscillations are not persistent. In this paper, in order to resolve the problem we propose a synthesis method of gene networks possessing the desired cyclic expression pattern sequences together with their corresponding solution trajectories being periodic. In the proposed method the persistent oscillations of the solution trajectories are realized by specifying passing points of them.

View full abstract

This paper considers an estimation problem of protein networks for phenomena in cell. By assuming that protein concentrations concerning the phenomena are obtained, an approach to solve the estimation problem is proposed. The proposed approach is based on the least-squares method for state space models. A 6-dimensional protein network is estimated for cell cycle in budding yeast to demonstrate the efficacy of the proposed approach. Moreover robustness analysis and numerical simulations are performed by deriving nonlinear equations which describe a dynamical behavior of the estimated protein network. From these theoretical and numerical results, robustness of the estimated protein network is discussed.

View full abstract

In the DNA-based molecular computation, calculation procedures described by the hardware design can provide an effective physical random source for the theoretically unbreakable encryption system. To realize its practical application, we present a new information processing idea.

View full abstract