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
24 巻, 3 号
選択された号の論文の4件中1~4を表示しています
論文
  • ―未達率指標の特性解析と基点在庫方策との比較
    上野 信行, 角本 清孝, 奥原 浩之
    2011 年 24 巻 3 号 p. 43-53
    発行日: 2011年
    公開日: 2011/06/15
    ジャーナル フリー
    In the previous paper, we proposed Mass Customization Production Planning & Management System (MCPS) based on unfulfilled-order-rate by using Advance Demand Information which is called ‘Naiji System’ as an unique corporation between a maker and suppliers in Japan, and 3 indicators to estimate the unfulfilled-order-rate. Applying these indicators to the model, we investigated the behavior of unfulfilled-order-rate at the final period in the planning horizon.
    In this paper, we propose a new model for purchasing, and investigate the unfulfilled-order-rate at each period and the impact to the total inventory. We find that the total inventories become 5.9%-20.0% decreases by using SOn rather than by using SOn(0). And we enhance a base-stock policy to a new one with multi-period. We prove that the MCPS model for purchasing by using SOn(0) is equivalent to the base-stock policy with multi-period under the specified condition. Under this condition, the proposed model by using SOn decreases inventories more than the base-stock policy with multi-period.
  • 横井 裕一, 引原 隆士
    2011 年 24 巻 3 号 p. 54-60
    発行日: 2011年
    公開日: 2011/06/15
    ジャーナル フリー
    Parametrically excited pendulum inherently demonstrates a conversion from forcing vibration to rotational motion. The onset of periodic rotation strongly depends on the initial condition. We propose a control method to start up a parametric pendulum into periodic rotation based on an external force input with time delay. The feasibility of proposed method is verified numerically and experimentally. This paper advocates that the proposed method is suitable for crossing over a separatrix which governs the dynamics from initial conditions.
  • ―プレフィルタ統合型学習則の改善
    野口 慎, 杉本 謙二
    2011 年 24 巻 3 号 p. 61-72
    発行日: 2011年
    公開日: 2011/06/15
    ジャーナル フリー
    This paper proposes an improved learning scheme for MIMO-FEL (Muliti-input multi-output feedback error learning) of prefilter-integration-type. Compared with an existing scheme, which is based on the gradient method and easily implemented, the proposed scheme does not require positive realness of the closed loop system. The paper shows the parameters converge to true values under a PE condition, and verifies its effectiveness in terms of simulation.
研究速報
feedback
Top