In recent years, Personal Mobility (PM) has attracted attention as a new means of moving in behalf of foot or car. PM is small electric powered vehicle that runs at low speed, and hence has highly affinity with pedestrians. Therefore, it is suitable to use PM in a space called 'shared space' where pedestrians, cars, and so on coexist. Additionally, research on autonomous driving has also been actively carried out in order to enhance its convenience. In autonomous driving of PM in shared space, it is important to establish a motion control method for safely and comfortably avoiding pedestrians, cars and so on. Therefore, in this paper, we propose a motion control method of PM with trajectory prediction that enables safe and comfort obstacle avoidance. In this method, firstly, the future trajectory of the obstacle is predicted. Next, this predicted trajectory of the future is represented as an oblique prism on the three dimensional space-time temporal space consisting of x, y and time axes and the path and speed of the PM is controlled so as to avoid this oblique prism. Thus, it is possible to control the path and speed of the PM to avoid various obstacles safely based on prediction. From various simulation results, we show the effectiveness of the proposed method.

This paper is concerned with the problem of constructing secure communication systems using chaotic synchronization of partial differential equations. We particularly use chaotic vibration of the wave equation with van der Pol boundary condition. In general, a secure communication system consists of a synchronization system and modulation and demodulation components. We adopt chaotic vibration of the wave equation to construct the synchronization system. The proposed method has two features. First the synchronization system is easy to construct. Second, vector-valued signals are transmitted in the secure communication system.

In a Lithium Ion Battery process, it is desired to find a causal factor that reduces the stability of the process. In this paper, we propose the mean replacement contribution. Each representative value is replaced with an average value one by one, and the difference between those Mahalanobis distances is used as a contribution. The proposed method was compared with T² contribution, Q contribution, reconstruction contribution (RBC), and random forest importance using the sensor signal of the coating device. As a result, the mean replacement contribution was able to identify the causal factor most accurately based on the knowledge of the engineer. Normally, an engineer needs a lot of man-hours to perform analysis, so that the proposed method can reduce the time.

This paper is aimed at stabilization of a cascade system composed of two subsystems. The first subsystem (ODE subsystem) is represented by a system of linear ordinary differential equations with a strictly upper triangular structure. The other subsystem (PDE subsystem) is described by a one-dimensional parabolic partial differential equation. The PDE subsystem is independent from the ODE subsystem, whereas the ODE subsystem depends on the PDE subsystem. The control input may affect both the subsystems. We develop a recursive procedure for designing stabilizing control laws based on the idea of the forwarding approach. The stability of the closed-loop system under the resulting control law is theoretically proved and numerically demonstrated.

This paper addresses observer-based scaled H_∞ control design with state estimation performance optimized. The simultaneous design of observer and state-feedback gains in H_∞ control framework is, in general, formulated in terms of Bilinear Matrix Inequality (BMI), which is not tractable compared to Linear Matrix Inequality (LMI). Similarly, the simultaneous design of H_∞ controllers and scaling matrices for multiple uncertainty blocks is also formulated in terms of BMI. On these issues, under some particular plant structures, we have proposed a tractable design method of the simultaneous design of observer and state-feedback gains as well as scaling matrices with some conservatism admitted due to the structural constraints for the matrices introduced in the so-called “dilation” procedure; however, observer performance was not optimized, and thus there is no guarantee on it. On this issue, this paper proposes two-step design, viz., control performance is firstly optimized using the existing method, and then observer performance is optimized under a slightly relaxed control performance imposed as a constraint. This approach expects the controllers' states faithfully work as the estimated plants' states while control performance is almost kept as optimized. This property is illustrated in this paper by the lateral-directional flight controller for a research airplane MuPAL-α validated through hardware-in-the-loop simulation tests.

When an unmanned aircraft autonomously flies in an extreme environment such as a disaster site, it is necessary to cope with environmental changes such as the influence of disturbance and the movement from outdoors to indoors. In this research, we are aiming to develop supervisor-type navigation and control system for unmanned aircraft to adapt to various surrounding environments. The research objective of this paper is to develop a system that recognizes the surrounding environment, which is one of the functions of the supervisor. By knowing the state of the environments where the unmanned aircraft is flying by environment recognition system, it is possible to issue an instruction for correction of the sensor configuration and control model to the navigation system and control system. In this paper, we first design deep learning models such as feed forward neural network(FFNN) or long short-term memory (LSTM) using LIDAR and GNSS sensor data for environment recognition. Designed system is verified by real time environment recognition experiment. From experiments, it was confirmed that ambient environment recognition can be performed with a F1-measure of about 0.94 using sensor data of LIDAR and GNSS. Moreover, it is considered that the performance of the surrounding environment recognition system can be further improved depending on the structure of the learning model and the training data used for deep learning.