システム制御情報学会論文誌
Online ISSN : 2185-811X
Print ISSN : 1342-5668
ISSN-L : 1342-5668
最新号
選択された号の論文の5件中1~5を表示しています
論文
  • Gou Nakura
    2019 年 32 巻 2 号 p. 47-54
    発行日: 2019/02/15
    公開日: 2019/05/15
    ジャーナル オープンアクセス

    While H filtering theory for stochastic continuous-time systems driven by Poisson processes has been presented by B. Song et al.(2015), H smoothing theory for the systems and the relationship between H filtering and smoothing have not been yet fully investigated. In this paper, we study the H state estimation (filtering and smoothing) problems for a class of linear continuous-time systems driven by Wiener and Poisson processes on the finite time interval. In order to derive H state estimators, we adopt unified stochastic variational approach which has not been found in previous work by any other researchers.

  • Kenta Hanada, Takayuki Wada, Izumi Masubuchi, Toru Asai, Yasumasa F ...
    2019 年 32 巻 2 号 p. 55-62
    発行日: 2019/02/15
    公開日: 2019/05/15
    ジャーナル オープンアクセス

    Multi-agent systems over noisy networks with multi-input/multi-output linear symmetric agents are considered. The information is assumed to be sent from an agent to its neighbors via multi-channels. The communication graph of each channel is allowed to be time varying and a different topology. The entire network which is defined by the union of all graph is assumed to be connected in an undirected graph case or weakly connected and balanced in a directed one. The aim of this study is to establish a stopping rule of a stochastic averaging consensus under a noisy and time-varying network. The convergence analysis reveals an explicit relation between the number of iterations and the closeness of the consensus. The results are illustrated through numerical examples.

  • Kanako Mikami, Kunihiko Oura
    2019 年 32 巻 2 号 p. 63-68
    発行日: 2019/02/15
    公開日: 2019/05/15
    ジャーナル オープンアクセス

    This paper discusses “visual attention” related hemodynamics and potentials during performance of modified paced visual serial addition test (mPVSAT). In researches for human cognitive function, “attention” is regarded as the basis of all mental abilities or human actions. The mPVSAT task, as our revised version of standardized test concerning attention, also needs higher-order attention to solve the test. Then, in order to make clear the mechanism of attentional function, brain activity is evaluated by event related hemodynamics (ERH) and event related potentials (ERP) in our study. From the results, some typical areas with high processing speed or large volume of blood flow are shown. Judgement and recognition process tend to be slow while complicated stimulus in visual attention task.

  • 杉浦 翔, 村田 忠彦, 原田 拓弥
    2019 年 32 巻 2 号 p. 69-78
    発行日: 2019/02/15
    公開日: 2019/05/15
    ジャーナル オープンアクセス

    In this paper, we propose a synthetic reconstruction method (SR method) to allocate an income attribute to each worker in synthetic populations. A synthetic population is a population synthesized based on statistics. We assign an income attribute to each worker in individual households using statistics in Japan. In order to add that attribute, we first prepare a synthetic population of households with members synthesized by our previously proposed SR method. Then we assign job attributes such as a working status, a type of employment, a type of industry, and a size of enterprise according to four statistics using our SR method. After determining the working status, we assign job attributes besides the working status for each worker. We determine the monthly income of each worker based on his/her job attributes. To see the validity of allocated monthly income, we compare the average income of each type of industry in the synthesized population with the statistics of the average income of each type of industry that is not used in the synthesizing procedure. The result showed that the error over all industries becomes 0.8% to 10.3% in allocating income to each worker in a synthesized population.

  • 中島 裕貴, 飯間 等
    2019 年 32 巻 2 号 p. 79-86
    発行日: 2019/02/15
    公開日: 2019/05/15
    ジャーナル オープンアクセス

    Metaheuristics are promising methods for combinatorial optimization problems. However, existing metaheuristics cannot necessarily find optimal solutions or near-optimal solutions. Meanwhile, in the field of machine learning, reinforcement learning theoretically enables an agent to learn optimal actions and has attracted attention. Thus, we previously proposed a general metaheuristic framework based on the reinforcement learning. However, this framework is not superior in finding good solutions in a short time because it generates a relatively large number of infeasible solutions. In this paper, we propose an extension of the metaheuristic framework to find better solutions. In addition, we propose a concrete method based on the metaheuristic framework for solving the knapsack problem as a case study. The effectiveness of our proposed method is verified through experiments of applying the proposed method and other methods to the knapsack problem.

feedback
Top