掲載種別：研究論文（学術雑誌）   出版者・発行元：IEEE Sensors Journal  

The characteristics of the sensor selection method based on various algorithms for state estimation in a high-dimensional system using the Kalman filter were investigated. Sensors were selected based on the error covariance matrix of the Kalman filter. The performance of the sensor selection methods based on different algorithms, semidefinite programming (SDP), approximate convex relaxation, and greedy algorithm, including newly formulated methods, were compared by varying the number of potential sensor locations and the number of sensors to be selected under several noise ratio conditions. Two sensor selection methods for state estimation in a high-dimensional system using a Kalman filter were newly proposed based on the SDP with gain formulation and approximate convex relaxation, and the characteristics of the method including the previously proposed method were compared in a high-dimensional system. Although the condition was limited at n < O(10³) , the approximate convex relaxation method and its randomized method are effective in terms of computational time and objective value for a small-scale problem. The objective value obtained by the greedy method shows the best performance compared to the other methods in almost all investigated conditions. Particularly, the greedy method outperforms other methods when the number of selected sensors is small. In addition, only the greedy method can handle the large-scale problem of n> O(10⁴). Overall, the greedy-based method is found to be favorable in large-scale problems in terms of computation time and the performance of the obtained sensor set.

DOI： 10.1109/JSEN.2024.3388849

Open Access

Web of Science

Scopus

researchmap

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：IFAC Journal of Systems and Control  

We propose a method to design reduced-order output-feedback controllers for fluid flows with the use of data produced by full-order controllers. First, the full-order controller is obtained by combining an ensemble Kalman filter (EnKF) and a model predictive controller (MPC) that are designed based on the Navier–Stokes equations. The full-order controller has high computational complexity and, therefore, is not suitable for real-time implementation. Hence, we use the full-order controller in offline numerical simulations to generate data for data-driven design of the reduced-order controller with low computational complexity. We find a reduced-order subspace of a closed-loop system under the full-order control from the data. This subspace underlies the reduced-order output-feedback controller. The reduced-order state-feedback law is obtained by approximating the full-order MPC with the use of its input/output data. The reduced-order observer is designed for a reduced-order model that is derived by using the Gaussian process regression (GPR). The GPR enables us to design the reduced-order observer which can evaluate uncertainty due to state-dependent residuals of the reduced-order model. We demonstrate the proposed method for a control problem of a flow around a cylinder at the Reynolds number 100. Numerical simulations reveal that the reduced-order controller performs as almost well as the full-order controller for a set of initial states. In addition, robustness of the reduced-order controller to a temporal disturbance that is not considered in the control design is confirmed in the simulations.

DOI： 10.1016/j.ifacsc.2024.100261

Open Access

Web of Science

Scopus

researchmap

記述言語：英語   出版者・発行元：Sensors  

Optimization approaches that determine sensitive sensor nodes in a large-scale, linear time-invariant, and discrete-time dynamical system are examined under the assumption of independent and identically distributed measurement noise. This study offers two novel selection algorithms, namely an approximate convex relaxation method with the Newton method and a gradient greedy method, and confirms the performance of the selection methods, including a convex relaxation method with semidefinite programming (SDP) and a pure greedy optimization method proposed in the previous studies. The matrix determinant of the observability Gramian was employed for the evaluations of the sensor subsets, while its gradient and Hessian were derived for the proposed methods. In the demonstration using numerical and real-world examples, the proposed approximate greedy method showed superiority in the run time when the sensor numbers were roughly the same as the dimensions of the latent system. The relaxation method with SDP is confirmed to be the most reasonable approach for a system with randomly generated matrices of higher dimensions. However, the degradation of the optimization results was also confirmed in the case of real-world datasets, while the pure greedy selection obtained the most stable optimization results.

DOI： 10.3390/s23135961

Web of Science

Scopus

PubMed

出版者・発行元：IEEE Access  

In this paper, the sparse sensor placement problem for least-squares estimation is considered, and the previous novel approach of the sparse sensor selection algorithm is extended. The maximization of the determinant of the matrix which appears in pseudo-inverse matrix operations is employed as an objective function of the problem in the present extended approach. The procedure for the maximization of the determinant of the corresponding matrix is proved to be mathematically the same as that of the previously proposed QR method when the number of sensors is less than that of state variables (undersampling). On the other hand, the authors have developed a new algorithm for when the number of sensors is greater than that of state variables (oversampling). Then, a unified formulation of the two algorithms is derived, and the lower bound of the objective function given by this algorithm is shown using the monotone submodularity of the objective function. The effectiveness of the proposed algorithm on the problem using real datasets is demonstrated by comparing with the results of other algorithms. The numerical results show that the proposed algorithm improves the estimation error by approximately 10% compared with the conventional methods in the oversampling case, where the estimation error is defined as the ratio of the difference between the reconstructed data and the full observation data to the full observation. For the NOAA-SST sensor problem, which has more than ten thousand sensor candidate points, the proposed algorithm selects the sensor positions in few seconds, which required several hours with the other algorithms in the oversampling case on a 3.40 GHz computer.

DOI： 10.1109/ACCESS.2021.3076186

Open Access

Web of Science

Scopus

出版者・発行元：IEEE Sensors Letters  

A vector-measurement-sensor problem for the least squares estimation is considered, by extending a previous novel approach in this letter. An extension of the vector-measurement-sensor selection of the greedy algorithm is proposed and is applied to particle-image-velocimetry data to reconstruct the full state based on the information given by sparse vector-measurement sensors.

DOI： 10.1109/LSENS.2020.2999186

Web of Science

Scopus

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：Energies  

Complexity of online computation is a drawback of model predictive control (MPC) when applied to the Navier-Stokes equations. To reduce the computational complexity, we propose a method to approximate the MPC with an explicit control law by using regression analysis. In this paper, we extracted two state-feedback control laws and two output-feedback control laws for flow around a cylinder as a benchmark. The state-feedback control laws that feed back different quantities to each other were extracted by ridge regression, and the two output-feedback control laws, whose measurement output is the surface pressure, were extracted by ridge regression and Gaussian process regression. In numerical simulations, the state-feedback control laws were able to suppress vortex shedding almost completely. While the output-feedback control laws could not suppress vortex shedding completely, they moderately improved the drag of the cylinder. Moreover, we confirmed that these control laws have some degree of robustness to the change in the Reynolds number. The computation times of the control input in all the extracted control laws were considerably shorter than that of the MPC.

DOI： 10.3390/en13061325

Open Access

Web of Science

Scopus

researchmap

開催年月日： 2023年8月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2023年7月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2022年11月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2022年11月

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2022年9月

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2022年1月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2021年11月

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2020年11月

会議種別：口頭発表（一般）  

researchmap

開催年月日： 2020年9月

記述言語：英語   会議種別：口頭発表（一般）  

researchmap

開催年月日： 2020年3月

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap