Published papers

デシメーションを用いた同定法を，大型宇宙構造物の同定実験に適用している。そのとき，デシメーションで用いるフィルタを加速度データの周波数特性から決定する方法を提案している。同定実験で求められた振動特性より制御系設計用モデルを構築し，そのモデルに対して現代制御理論を適用して制御系を設計している。

従来の勾配法に基づく適応フィルタのタップ係数の調整アルゴリズムを修正し，有色性入力に対しても有効に動作する調整アルゴリズムを提案している。提案法を符号検出器に適用し，その有効性を示している。

除振マウント支持の半導体露光装置の周波数特性を部分空間法により計算した後，同定された周波数応答から物理パラメータを算出する方法を提案している。提案法の有効性を実験データを用いて明らかにしている。

線形離散時間システムのインパルス応答を最小二乗法によって同定する問題について考察している。同定入力信号が滑らかな場合，対応する入力相関行列が悪条件になるが，そのとき入力相関行列の対角要素に小さな正数を加えて悪条件を回避する，正則化法を提案し，その効果を固有値分解に基づいて解析している。また，最適な正数の決定法も提案している。

制御対象に加わる目標値や外乱値が事前に既知であれば（すなわち,予見できれば），制御性能の向上が期待できる。本論文では，制御対象が離散時間表現されるとき，このような予見情報を制御系で有効に利用する方法を提案し，数値シミュレーション例を通して，その有効性を明らかにしている。

Satellite communications (SATCOM) systems play important roles in wireless communication systems. In the future, they will be required to accommodate rapidly increasing communication requests from various types of users. Therefore, we propose a framework for efficient resource management in large-scale SATCOMsystems that integrate multiple satellites. Such systems contain hundreds of thousands of communication satellites, user terminals, and gateway stations; thus, our proposed framework enables simpler and more reliable communication between users and satellites. To manage and control this system efficiently, we formulate an optimization problem that designs the network structure and allocates communication resources for a large-scale SATCOM system. In this mixed integer programming problem, we allow the cost function to be a combination of various factors so that SATCOM operators can design the network according to their individual management strategies. These factors include the total allocated bandwidth to users, the number of satellites and gateway stations to be used, and the number of total satellite handovers. Our numerical simulations show that the proposed management strategy outperforms a conventional strategy in which a user can connect to only one specific satellite determined in advance. Furthermore, we determine the effect of the number of satellites in the system on overall system performance.

In this paper, we propose an optimization framework for the resource and network management in large-scale satellite communications (SATCOM) systems with stochastic and time-varying communication requests. Because further expansions of SATCOM systems are expected in the near future, it is of practical importance to develop management methodologies for efficiently and robustly operating large-scale SATCOM systems. To address this problem, in this paper, we formulate a SATCOM network optimization problem within the framework of the chance-constrained model predictive control. In this problem, we require that a joint chance constraint on the bandwidth loss rate is satisfied with a user-specified performance and probability. Because the joint chance constraint is nonlinear in design variables, we relax the constraint to a linear deterministic one for reformulating the original problem into a relaxed deterministic problem. We numerically verify that the proposed method outperforms a baseline methodology and therefore allows us to efficiently manage large-scale SATCOM systems.

This paper presents a method to obtain a robust optimal path in an environment with time-varying safety features, such as in the lunar polar region. In designing the path for planetary exploration rovers, we must consider various safety conditions, such as terrain hazards, illumination, and communication visibility to the Earth. Some of the safety features are time-varying, and the optimal path should be searched for both the spatial direction and the temporal direction. In addition, there is no guarantee that all of the sequences will be successfully executed on time due to misoperation, failures, or trouble. Therefore, a path that is robust against the delay must be planned so as to guarantee safety even when behind schedule. The authors propose an algorithm called “Robust Spatio-Temporal Path Planner for the Planetary Rover (ROBUST-STP3R)” to obtain a path that is robust against schedule delay in a time-varying environment. This method defines a cost function that consists of the distance as well as the region type cost. To add robustness against schedule delays, the authors consider a weighted summation of the time-varying region type cost with regard to the temporal direction. The effectiveness of the proposed method is demonstrated through use in a simulated lunar polar exploration exercise.

In this paper, we propose a design method of state feedback controllers for discrete-time linear systems whose state-space matrices contain stochastic parameters. For such systems, if the parameters fluctuate according to the normal distribution, it is impossible to strictly guarantee the “conventional deterministic control performance” due to the unlimited parameter variations. Then, in our problem setup, we consider incorporating the “stochastic stability” into the conventional deterministic H2 performance. We derive a design condition on state feedback controllers which achieve deterministic H2 performance and stochastic stability simultaneously in terms of linear matrix inequalities (LMIs) via extended LMI. The effectiveness of the proposed method is illustrated through a numerical example.

We propose Absum, which is a regularization method for improving adversarial robustness of convolutional neural networks (CNNs). Although CNNs can accurately recognize images, recent studies have shown that the convolution operations in CNNs commonly have structural sensitivity to specific noise composed of Fourier basis functions. By exploiting this sensitivity, they proposed a simple black-box adversarial attack: Single Fourier attack. To reduce structural sensitivity, we can use regularization of convolution filter weights since the sensitivity of linear transform can be assessed by the norm of the weights. However, standard regularization methods can prevent minimization of the loss function because they impose a tight constraint for obtaining high robustness. To solve this problem, Absum imposes a loose constraint; it penalizes the absolute values of the summation of the parameters in the convolution layers. Absum can improve robustness against single Fourier attack while being as simple and efficient as standard regularization methods (e.g., weight decay and L1 regularization). Our experiments demonstrate that Absum improves robustness against single Fourier attack more than standard regularization methods. Furthermore, we reveal that robust CNNs with Absum are more robust against transferred attacks due to decreasing the common sensitivity and against high-frequency noise than standard regularization methods. We also reveal that Absum can improve robustness against gradient-based attacks (projected gradient descent) when used with adversarial training.