This paper studies a method of designing robust output estimators for discrete-time linear periodically time-varying (LPTV) systems with uncertainties. The key idea is to use not the well-known lifting technique but that called cycling for dealing with LPTV systems in the estimator synthesis. Robustness for uncertainties in the estimation is evaluated with the separator-type robust stability theorem through such cycling-based treatment of systems. An advantage of our cycling-based approach, compared to the lifting-based approach, is that we can easily introduce restrictions on the coefficients of estimators in the synthesis for predetermining the estimator period regardless of the system period.
This paper presents a stereo-vision-based approach for sea-bottom docking of autonomous underwater vehicles (AUVs) for battery recharging. According to the intended application, a unidirectional docking station was designed in which the AUV has to dock from a specific direction. Real-time relative pose (position and orientation) estimation was implemented utilizing three-dimensional model-based matching to the actual target and a real-time multi-step genetic algorithm. Using the proposed approach, we conducted the experiments in which an AUV docked to a simulated underwater battery recharging station in the sea near Wakayama City, Japan. The experimental results confirmed the functionality and potential of the proposed approach for sea-bottom docking of AUVs. Although similar sea trials were reported previously, detailed discussions and performance analyses were not presented, especially regarding the relations among pose estimation, output control voltage, and photographic records. The analyses confirmed that the successful docking was realized and that the method has tolerance against turbulence applied to a remotely operated vehicle near the docking station.
The neural network is one of the most successful machine learning models. However, the neural network often requires large amounts of well-balanced training data to ensure prediction accuracy. Meanwhile, human learners can generalize a new concept from even a small quantity of biased examples, simultaneously enlarging knowledge with an increase in experience. As a possible key factor in the ability to generalize, human beings have cognitive biases that effectively support concept acquisition. In this study, to narrow the gap between human and machine learning, we have implemented human cognitive biases into a neural network in an attempt to imitate human learning to enhance performance. Our model, named loosely symmetric neural network, has shown superior performance in a breast cancer classification task in comparison with other representative machine learning methods.