Submission Deadline: 31 March 2027 View: 57 Submit to Special Issue
Prof. Weiwei Bai
Email: baiweiweidl@dlmu.edu.cn
Affiliation: Navigation College, Dalian Maritime University, Dalian, China
Research Interests: data driven, physical driven, modeling, control
Prof. Xin Hu
Email: huxinsea@163.com
Affiliation: School of Mathematics and Statistics Science, Ludong University, Yantai, China
Research Interests: data-physical driven modeling, disturbance observer, ship motion control
Prof. Andrea D'Ariano
Email: andrea.dariano@uniroma3.it
Affiliation: Dipartimento di Ingegneria, Università degli Studi Roma Tre, via della Vasca Navale, Roma, Italy
Research Interests: intelligent transportation systems, transportation & logistics, operations research, air traffic control
Recently, data-driven modeling and control methods have received much attention in both theory and practice. Different from traditional approaches, these methods do not require any prior knowledge. As we all know, physics-driven methods infer the characteristics of the research object through analyzing its working mechanism and describe the causal relationships between variables with appropriate mathematical expressions in combination with functional requirements. The characteristics of physics-based methods are that unknown dynamics can be predicted through physical mechanisms. However, data-driven methods do not depend on the strict analysis of the working mechanism of the object. Based on a large number of experiments and test data, data processing algorithms are used to analyze the mapping between the data and system dynamics. The characteristic of this method is to extract the correlations between variables based on data samples.
In practical applications, physics-based methods are mainly faced with problems such as inaccurate physical models, difficulties in constructing physical models, and so on. In order to achieve a breakthrough in the performance of mechanism models, the model must be improved, and data-driven methods can provide assistance in discovering new variables and explaining new phenomena. Data-driven methods are often affected by the quality and quantity of samples, and it is difficult to adapt to situations in which the target scenario changes greatly. Therefore, when building empirical models based on data, the guidance of physics-based methods is also needed, such as introducing prior conditions based on physical knowledge constraints to avoid abnormal phenomena, so as to enhance the adaptability of empirical models. Therefore, in the study of practical problems, combining data and physics-based methods will help to improve the overall performance of the method and enhance its application effectiveness.
The main purpose of this Special Issue is to provide a forum for global researchers to share the latest achievements and discuss the future challenges in data-physical-driven modeling, control, and their applications. The submitted papers are expected to present original ideas and potential contributions to theory and practice. The topics include, but are not limited to, the following research areas:
* data-physical-driven modeling
* data-physical-driven fuzzy modeling
* data-physical-driven machine learning
* data-physical-driven deep learning and control
* Reinforcement learning control design
* data-physical-driven fault diagnosis
* data-physical-driven locally weighted learning
* Applications to marine systems and other related areas
Login
Register
Submit a Paper
Propose a Special lssue