Published papers

本研究では多自由度マニピュレータを用いて環境及びオペレータ特性の不確かさと通信遅延に対するロバスト性能を統合的に保証した制御系設計を行なう．まず2 台のマニピュレータに対してインピーダンス整形により非干渉・線形化を行い，環境及びオペレータの動特性をインピーダンスモデルとして表現し，これらを統合化した一般化プラントを構成する．次に不確かさを考慮したμ-Synthesisに基づく制御系設計法を提案する．最後に制御実験により提案法の有効性を示す．

本稿ではH∞DIA制御を鋼板磁気浮上系に応用し, モデル変動に対するロバスト性を制御実験により評価することを目的とする．制御系設計では, 外乱や不確かさに対する重みを設計し, システムの状態方程式と組み合わせることで一般化プラントを構成し, これに対してコントローラを設計する．提案法であるH∞DIA制御が十分なモデル変動に対するロバスト性を有することを制御実験により明らかにした.

本研究ではGIMC構造を不安定なメカトロニクスシステムである磁気浮上システムに応用し，制御実験による有効性の検証を行う．更にGIMC構造における切替制御の具体的な設計指針と実装方法を提案する.まずGIMC構造に基づく制御系設計の原理と設計手順を述べ，その切替の実装方法について提案を行う．次にH∞混合感度問題を用いて2つの補償器を設計し，それらを基にGIMC構造制御系を構成する．最後に提案法が制御性能とロバスト性を両立していることを実験的に示す．

This paper deals with a control for teleoperation with different configurations and communication delays. We propose a synchronized control with individual gains and power scaling in the task space. In this method, the end-effector motion and force relationship between the master and slave robots can be specified freely in the task space and the control gains can be independently selected appropriately for the master and slave robots. The delay-independent asymptotic stability of the origin of the position and velocity errors is proven by using passivity of the systems and Lyapunov stability methods. The proposed control law achieves synchronization with power scaling. Several experimental results show the effectiveness of our proposed method.

This paper deals with a new GIMC structure considering a communication delay by using the Smith Predictor and its application to a mechatronic system. First, we stabilize the unstable mechatronic system by a PD controller and define the stabilized system as a new augmented plant. We design the proposed GIMC structure using the Smith Predictor based on H-infinity controllers for the new stabilized augmented plant. Finally, the proposed structure is evaluated experimentally and the effectiveness of the proposed approach is proven. In addition, we apply the proposed structure to a virtual networked control mechatronic system.

本論文では磁気軸受のロータ回転時に発生する不釣合い振動を，回転速度に同期した周波数を持つ周期的外乱とみなし，その周期的外乱を抑制するために，H∞ DIA制御問題における一般化プラントに周波数重み関数を導入することで，任意の回転速度における周波数でコントローラ上にピークゲインを付加した．切替え制御による性能検証では，2 種類のコントローラを設計し，切替えることで良好な不釣合い振動の減衰効果と切替え時の過渡応答特性を有することを確認した．

本論文では補償器の初期状態を考慮したH∞ DIA 制御を磁気軸受に適用して過渡応答特性を検証する．まず磁気軸受の数学モデルの導出と補償器の初期状態を考慮したH∞ DIA制御系設計について述べる．次にこの補償器の初期状態を考慮したH∞ DIA補償器を異なる初期状態の不確かさに対する重みＮを持つ補償器を設計する．その補償器を用いて，初期値応答と外乱応答を計測し，Nの大きさと時間応答の変化を検証する．

This paper proposes an observer-based consensus control strategy for multi-agent system (MAS) with communication time delay. The condition of stability for MIMO agents is derived by Lyapunov theorem. It gives systematic design procedure under assumed unidirectional network. Furthermore, new consensus control law using observers is proposed for the networked MAS with communication delays. Experimental results show effectiveness of our proposed output consensus approaches.

本稿では線形システムとして表現される複数の同一のダイナミクスをもつエージェントの合意問題に対し，新規の合意制御則の設計法を提案する．この設計法はLyapunovの安定定理に基づき，従来法に比べ設計の見通しが良い．つぎに状態推定を含む出力合意問題に対して，情報交換を行うエージェントの数に依存しない状態推定方法を用いた合意制御則を提案し，更に積分器を加えた合意制御則を提案する．最後に提案制御則をフォーメーション制御に応用し，制御実験により提案法の有効性を検証する．

本稿ではGIMC 構造によって得られる推定誤差信号によるオンライン故障検出法の提案と, それを用いた耐故障制御系の構築をおこなう. 故障検出フィルタに関して外乱による影響の最小化問題を新たに定式化し, それに対する最適な故障検出フィルタを導出することで, 外乱と故障の影響を区別する. また磁気浮上系に対し外乱および故障信号を含めたモデルを導出し, 最適故障検出フィルタの設計を行う. 最後に検証実験を行い, 設計した故障検出フィルタの有効性を確認する.

本稿では通信遅延を有する異構造テレオペレーションに対して安定性を保証した制御手法を提案する．まずマスタ及びスレーブロボットに対してフィードバック受動化を行い， 次に通信遅延とスケーリング要素を有する同調制御則を提案する．リアプノフの安定法を用いることで, 提案テレオペレーションの位置誤差が遅延とスケーリング要素に対して独立に漸近安定となることを示す. 最後に構造とスケールの異なる2台の2自由度アームを用いた実験により提案手法の有効性を示す.

This paper deals with an environment estimation based predictive control for teleoperation with communication time delay. In this predictive control framework, the delayed position tracking and non-delayed force reflection are attempted. The goal of this paper is to propose a new control method which guarantees the stability, the convergence of position tracking error to zero and the small force tracking error theoretically. Proposed control structure is the 3ch architecture with environment estimation based force prediction. In this proposed method, the master controller implements force control with environmental force prediction while the slave controller implements position and force control. The theoretical analysis shows the stability, the convergence of position error to zero and the small force tracking error. Experimental results show the effectiveness of the proposed method.

This paper describes a bilateral control of nonlinear teleoperation with time varying communication delay. The proposed method are simple PD-type controllers which are independent of the rate of time delay change and depend on the upper bound of round-trip delay. The proposed control strategy is independent of parameter uncertainties of the model of the robots and the operator and remote environment. The delay-dependent stability of the origin is shown via Lyapunov stability theorem. Furthermore the proposed strategy also achieves master-slave position coordination and bilateral static force reflection. Several experimental results with wireless communication and the Internet show the effectiveness of our proposed strategy.

This paper deals with GIMC (Generalized Internal Model Control) structure based Anti-Windup (AW) control system design. First, we setup a L2 performance criterion based on GIMC structure for Anti-Windup control problems. The design problem is formulated as an optimization problem with a parameter Q. Then we calculate the optimization problem and find an optimal Q. Finally, the effectiveness of proposed method is shown by experimental results for a magnetic suspension system. © 2009 IEEE.

This paper addresses the bilateral control of nonlinear teleoperation with time varying delays. The proposed control strategies are two types of simple PD-type controllers which directly connect the master and slave robots by position and velocity signals over the delayed communication. For the velocity control, the first controller has a time varying D-gains which depend on the rate of change of delays and the second one has a constant D-gains which are designed under stability condition. Moreover the both controllers have explicit position feedback/feedforward control. Using Lyapunov-Krasovskii function, the delay-depend stability of the origin is shown for the range of the gains. Furthermore the proposed framework enforces master-slave position coordination and static force reflection. Several simulation and experimental results show the effectiveness of our proposed framework.

This paper deals with robust stabilization of running self-sustaining two-wheeled vehicle with varying running speed and mass variations. Recently, there have been a lot of research results about stabilization of two-wheeled vehicle. These researches have achieved the stable running only by the steering control. However, an actual two-wheeled vehicle is stabilized by not only the steering control, but the control of a vehicle's center of gravity. We have proposed the stabilization control method of two-wheeled vehicle in the state of stillness, and have shown the effectiveness of proposed approach. In this research, we construct the control system for the running stabilization of two-wheeled vehicle. H∞ control problem is used to design the controller to achieve running stability even if the running speed and the mass of two-wheeled vehicle would change. The experimental results show effectiveness of proposed approach for robustness.

This paper deals with cooperative target-enclosing problem for swarms of vehicles with dynamic network-topology. First, we introduce a dynamic network topology that depends on relative distance between the vehicles. Secondly, we propose the target-enclosing strategy based on consensus seeking with dynamic network-topology. In proposed strategy, at least one vehicle has to acquire the states of target-object. Furthermore, we prove that convergence speed of each vehicle is independent of network-topology. To analyze the convergence of target-enclosing behavior under dynamic network topology, algebraic graph theory and matrix theory are utilized. Finally, numerical simulation results and experimental results are provided that demonstrate the effectiveness of the proposed method.

This paper discusses cyber attacks and faults diagnosis method for networked electrical power systems. A　power system is divided into some areas, then phase angles and angular frequencies are estimated by Kalman　filter in each area. Then we propose a diagnosis method for the power grid by using state and output fault　diagnosis matrices which are composed of the estimate value and output matrix. These diagnosis matrices can　immediately reflect an existence or non-existence of abnormal signal. Finally we show simulation results to analyze effectiveness of the proposed method.

In this paper, the optimal energy management considering battery characteristics for smart grid and micro grid systems is discussed. The energy storage system is very important for the energy peak shift and peak cut for micro grid and smart grid system because of these backgrounds. However, optimal battery formulation and energy management to suppress by using expensive energy storage system and considering battery characteristics are still in a developmental stage. First, the optimal battery placement for smart grid system with Lithium-ion battery is dealt with for lifetime extending. The battery choice index for this optimal battery determination problem to formulate optimal energy management is discussed. Second, the smart grid system with distributed battery system for grid frequency control considering distributed control and battery characteristics is dealt with. Optimal controllers for each distributed battery systems are designed to reduce battery load and to expend battery lifetime are designed. The effectiveness of the proposed method is shown via simulations by comparing with conventional methods.