Gaps and steps between platforms and trains reduce the accessibility and mobility of people with wheelchairs in railway transportations. Using an experimental platform, the observations are performed how gaps and steps influence their capabilities for manual wheelchair or electric powered wheelchair users with spinal cord injury. A quantity of Normalized Driving Force (NDF) is introduced to evaluate the manual wheelchair user's abilities in the case of getting on or off the trains. Three types of electric powered wheelchairs are also tested under the same experimental conditions as the manual wheelchair. The dynamic wheelchair driving force is measured by using a torque meter equipped on a wheelchair to analyze the required force when getting on the trains. To improve practical accessibility of such people, an assistive device for boarding the trains is designed and its effect is verified.
To measure a position by GPS (Global Positioning System), observing more than three satellites is usually needed. If there are not obstacles near a receiver, more than three satellites are generally observed and the position is located. However, in an urban area or a forest, more than three satellites can not be frequently acquired because of buildings and trees. Then we suggest a new algorithm for positioning using only three satellites. When only three satellites are observed, the data for positioning is not enough. To solve the problem, we use consecutive data and measure the position. Other methods using only three satellites usually need past data of the receiver such as the position or external data by the reference station or devices other than the receiver. However our method does not need other data and has only to fix a receiver for a few seconds. Therefore our method is useful for pedestrians. When consecutive data is used, noise of pseudo range influences the positioning accuracy. Then we use carrier phase in the L2 signal and make the noise less. In this paper, we show the effectiveness of our method using a dual frequency receiver and received data from three satellites whose elevation angles are high.
In this paper, we present a modified dynamic programming (DP) method. The method is basically the same as the value iteration method (VI), a representative DP method, except the preprocess of a system's state transition model for reducing its complexity, and is called the dynamic programming on reduced models (DPRM). That reduction is achieved by imaginarily considering causes of the probabilistic behavior of a system, and then cutting off some causes with low occurring probabilities. In computational illustrations, VI, DPRM, and the real-time Q-learning method (RTQ) are applied to elevator operation problems, which can be modeled by using Markov decision processes. The results show that DPRM can compute quasi-optimal value functions which bring more effective allocations of elevators than value functions by RTQ in less computational times than VI. This characteristic is notable when the traffic pattern is complicated.
We propose an approach of automated co-evolution of the optimal values of attributes of active sensing (orientation, range and timing of activation of sensors) and the control of locomotion gaits of a simulated snake-like robot (Snakebot) that result in a fast speed of locomotion in a confined environment. The experimental results illustrate the emergence of a contactless wall-following navigation of fast sidewinding Snakebots. The wall-following is accomplished by means of differential steering, facilitated by the evolutionary defined control sequences incorporating the readings of evolutionary optimized sensors.
This paper addresses a problem of finding an optimal dynamic quantizer for systems including discrete-valued signal constraints. In the previous works of the authors, an optimal dynamic quantizer has been derived, and its effectiveness has been illustrated. However, since the former results have been derived under a strong assumption, the quantizer can be applied to limited cases. To overcome this drawback, we provide a dynamic quantizer based on a multirate sampling technique. First, we formulate an optimization problem for a class of multirate sampling type dynamic quantizers, and then we give a closed-form solution as an extension of the former single-rate sampling case. Finally, the validity of the proposed dynamic quantizer is demonstrated by an experiment with a cart system.
This paper proposes a novel approach to the Hamilton-Jacobi equation (HJE). Through a functional analytic formulation, a stabilizing solution of the HJE is derived from a fixed-point of a nonlinear mapping on a function space. A sufficient condition for the existence of such a fixed-point is derived quantitatively. For computing the fixed-point, a numerical method based on optimization techniques is introduced. Based on these results, the nonlinear H∞ control problem is reconsidered. Simple numerical examples show that the resulting controllers have enough performance and computed solutions can approximate their analytical counterparts almost uniformly in the entire region under consideration.
To maintain stable operation of semiconductor fabrication lines, statistical process control (SPC) methods are recognized to be effective. However, in semiconductor fabrication lines, there exist a huge number of process state signals to be monitored, and these signals contain both normally and non-normally distributed data. Therefore, if we try to apply SPC methods to those signals, we need one which satisfies three requirements: 1) It can deal with both normally distributed data, and non-normally distributed data, 2) It can be set up automatically, 3) It can be easily understood by engineers and technicians. In this paper, we propose a new SPC method which satisfies these three requirements at the same time. This method uses similar rules to the Shewhart chart, but can deal with non-normally distributed data by introducing “effective standard deviations”. Usefulness of this method is demonstrated by comparing false alarm ratios to that of the Shewhart chart method. In the demonstration, we use various kinds of artificially generated data, and real data observed in a chemical vapor deposition (CVD) process tool in a semiconductor fabrication line.
This paper describes a method for sweet pepper picking robots working in greenhouses to distinguish the pepper fruits from the leaves. The fruits of the sweet pepper plant are recognized by image processing techniques, using a parallel stereovision system installed in the robot. However, as fruits and leaves of the sweet pepper have almost the same color, it is very difficult to recognize fruits using only color information. In this paper, we propose a new method using reflections of LED lights. The fruits of the sweet pepper are more reflective than the leaves; therefore, we can identify fruits by assuming that the more reflective parts are the fruits. We perform experiments using the improved image processing algorithm in a greenhouse, and the algorithm indeed improves the recognition ability of the robots.