For early diagnosis and treatment of disease, sensor systems that automatically get the physiological data in the toilet and bathroom have been developed. The data that obtained by these sensors need to be classified individually when two or more people live together. However, most of the existing personal identification methods, it is difficult to protect the user's privacy and reduce the burden. In this paper, we developed a personal identification method using features of floor vibration by walking on the wooden floor. From the waveform measured with the vibration sensor, eight features, peak to peak values and time intervals of each step, were extracted. As a result of personal identification by Mahalanobis' distance using the extracted features, the identification rate was 92%. As the number of registration data was increased, Mahalanobis' distance decreased a small value after the third day.

This paper deals with the effectiveness of the remote experimental system for active learning in robust control. It is usually difficult to understand control theory so as to design and implement control systems only by following general theoretical development without practical system examples; however, it is difficult to introduce practical systems in lectures. In order to solve the problem, we have developed a remote experimental system which can be used over the Internet providing an active learning environment for students who learn robust control theory even outside lecture hours. The developed remote experimental system automatically carries out the experiment when parameter files of designed controllers are uploaded via the Web page, then the control experimental results are posted on the same Web page in about one minute. The developed system was introduced in a lecture, ‘Advanced Automation’, in 2015 and 2016, and used for the control experiment of robust active noise control of duct. By analyzing the usage of the remote experimental system such as the number of designed controllers and the frequency of executions, and comparing the results with the final reports submitted by students, we confirmed that the number of active students who work well with the design problem was increased by introducing the remote experimental system compared with the previous situation before the introduction of the system. Similar results were obtained from the questionnaire results.

In this paper, we address distributed 3-D visual human localization for a visual sensor network and present a passivity-based solution. We first formulate the intended problem as a distributed optimization problem. A passivity-based distributed optimization algorithm is then applied to the problem, and it is pointed out that it ensures asymptotic optimality but the convergence speed is considerably slow. To accelerate the convergence, we present a novel algorithm based on a loop shaping technique and it is shown to maintain convergence to the optimal solution based on passivity. The present algorithm is finally demonstrated through simulation, wherein not only convergence acceleration but also robustness against disturbances are confirmed.

Recently, many devices driven by pneumatic actuators have been developed. In additions, many researches on pneumatic artificial muscles focusing on characteristics such as flexible operation, light weight and high output have been conducted. These artificial muscles were limited in their application range due to their durability and the necessity of an air pressure source. Therefore, in this research, as another approach apart from the improvement of the air pressure source device, we propose hollow artificial muscle with high efficiency and high durability.

This paper focuses on a sensor, which provides information of a shock propagating in flow field such as strength and direction of the shock. In engineering practice of aerospace, automobile, manufacture and so on, the shock wave is formed in high-speed flow and its strength and shape affect the flow properties behind the shock, and finally are responsible for efficiency of manufactured products and productivity of industrial goods. In this study, a sensitive head of a pressure transducer, on which a semi-conductor type strain gauge is glued, is analyzed using FEM and the relation is discussed between the behavior of its diaphragm over which a shock passes and the direction in which the shock is travelling. The numerical analysis offers that the strain distribution of surface of the diaphragm depends on the shock direction and that the direction of maximum principal strain of the diaphragm corresponds to the shock direction or its perpendicular direction. Experimentally, the strains in different directions on the diaphragm are obtained and the direction of maximum principal strain is estimated using Rosette analysis, being compared with the shock direction. As a result, it is possible to estimate the shock direction with adequate accuracy, using a scheme proposed in the present paper.

This paper presents a novel dynamical path following feedback control method for a snake-like robot. The snake-like robot is an undulatory locomotor transforming its periodic changes in shape into its displacement. To achieve work involving movement, e.g., an inspection task, in its working environment, it is necessary to specify its motion quantitatively. Since each of the links of the locomotor has passive wheels at its midpoint, it is not simple to design a feedback control system which enables the locomotor to follow desired paths (or to track desired trajectories) compared with wheel drive vehicles. Especially, the locomotor has singular attitudes in which the driving of its joints cannot be transformed into its movement. In this control method, a virtual link with a virtual steering system at its tip and a virtual axle at its midpoint connected to a top link through a virtual joint at its tip is caused to follow winding paths, which makes it possible to transform the driving torque of its joints into its propulsion force. The kinematical equations of the locomotor are formulated in a curvilinear coordinate system in which a winding path is one coordinate axis. A part of the kinematical equations describing the motion of the virtual mechanical elements only are converted into a chained form which facilitates to design a kinematical path following feedback control system. By a backstepping method, the acceleration required to realize the desired velocity in the kinematical control system is calculated. Based on the dynamical equations of the locomotor derived in Lagrangian mechanics, the torque of the joints of the locomotor is designed so as to realize the acceleration calculated by the backstepping method. The validity of this dynamical control method is verified by simulations in which the locomotor climbs a slope while following a serpenoid curve path.