内容简介：

In many applications, the states of multiple objects (called targets) are needed to be estimated and tracked from sensor measurements. The multi-target tracking (MTT)problem is especially challenging when in addition to the states, the number of objectsare unknown and vary with time, and the measurements include missed detections and false alarms. Such problems arise in many application areas including aerospace, defense, field robotics, security and surveillance, machine learning, and biomedical research. The most widely known class of solutions to the MTT problem are Bayesian multi-object filters that generalize classical paradigms such as Kalman/particle filters to multi-object systems.

The MTT problem is even more challenging in a distributed setting with multiple intercommunicating fusion centers (for example, a network of radars or a team of autonomous vehicles) each of which can only access part of the sensor data. In this context, the fusion centers have to exchange data and cooperate in order to solve the MTT problem. Distributed solutions are preferable in terms of robustness and scalability with respect to the problem size, but pose significant challenges in terms of information fusion. This is a fertile area of research and many aspects and applications have not been explored yet.

This seminar provides a general introduction to the distributed MTT problem and presents recent research achievements.

主讲人简介：

Giorgio Battistelli received the Laurea degree in electronic engineering and the Ph.D. degree in robotics from the University of Genoa, Genoa, Italy, in 2000 and2004, respectively.

From 2004 to 2006, he was a Research Associate with the Dipartimento di Informatica, Sistemistica e Telematica, University of Genoa. Since 2006 he has been with the University of Florence, Florence, Italy, where he is currently an Associate Professor of automatic control with the Department of Information Engineering. In 2014, he received the national scientific qualification (abilitazione scientifica nazionale) as Full Professor of Automatic Control. His current research interests include adaptive and learning systems, real-time control reconfiguration, linear and nonlinear estimation, hybrid systems, sensor networks, data fusion, and large-scale systems. He is the co-recipient of three awards at the International Conference on Information Fusion in the years 2009, 2012, and 2013, and of the best paper award at the 10th Workshop: Sensor Data Fusion: Trends, Solutions, Applications in the year 2015.

Dr. Battistelli was a member of the editorial boards of the IFAC Journal Engineering Applications of Artificial Intelligence and of the IEEE Transactions on Neural Networks and Learning Systems. He is currently an Associate Editor of the IFAC Journal Nonlinear Analysis: Hybrid Systems, and a member of the conference editorial boards of IEEE Control Systems Society and the European Control Association.