A knowledgebase web platform for collaborative physical system modeling and simulation

Engineering study involves concepts, principles and phenomena that are difficult to interpret due to people’s cognitive constraints. Learning through experiments is acknowledged to be one of the most effective ways of gaining new knowledge. Computer simulation involves applying scientific theories and performing data analysis on the collected data. Many simulation platforms provide visualization tools for observing phenomena and data analysis, and these are all characteristics of experiments. Hence, computer simulation can be used as a complementary tool to understand concepts, principles and phenomena other than actual physical experiments. Today’s Internet has the ability and capacity to share and publish large quantities of digital content. The Internet is becoming the most common channel for learning and sharing knowledge. The most popular places where people learn and share knowledge on the Internet is World Wide Web (WWW). However, there are not many simulation tools that support simulation capability and experiments sharing capability over the Web. In addition, traditional web document types such as text, image, and video are not able to describe abstract terms that are usually expressed or demonstrated using simulation experiments. A web-based platform, called Proteus (http://www.visualphysics.net/pweb), has been developed to provide a place where students and educators can easily create and share their computer models of physical systems. It comes with a web-based modeling and simulation tool called ProteusGWT (http://www.visualphysics.net/ProteusGWT). ProteusGWT uses an intuitive component-oriented approach to the modeling of physical systems spanning multiple domains including systems containing mechanical, hydraulic, thermal, control, electrical, electronic, electric power or process-oriented subcomponents. It synthesizes state-of-the-art web technologies, computational methods for physical systems modeling and simulation to create a computing environment that is widely deployable and scalable. Anyone with a computer or browser-enabled device will be able to use it. Hence, anyone can contribute to this platform. As the platform grows, it could turn out to be a reference for all kinds of physical-system models on the Internet. Together with this platform, a new web document format, named ModelicaJSON that can represent and allow editing of simulation models. Currently, there is no standard web document format for simulation experiments. Defining the structure of this new web document format that can be used as a web standard is critical. Users have been looking for proper simulation languages to better describe complex composite physical systems. Modelica is a language specially developed for physical system modeling and simulation with numerous advanced features such as object-oriented design, equation-based and a-causal modeling, multiple inheritances and algorithms. It has been developed in an international effort with the aim of creating a unified modeling language for complex physical systems spanning multi-domains. It is an open standard. There are many free and commercial Modelica libraries that are developed with it. Hence, Modelica has been selected as a reference for the development of this new web document format. The new web document format which can represent simulation models and enable them to be edited via the web-based graphical tool-ProteusGWT also implements key modeling features of the Modelica language. The main contributions of this research is the development of a knowledgebase web platform for physical modeling and simulation models. The web platform acts as a repository for simulation models to facilitate better sharing and collaboration. The thesis also proposes an intuitive and interactive way of presenting information on the web by providing a web-based, easy-to-use, easily accessible modeling and simulation tool called ProteusGWT. It is the first web-based modeling and simulation tool capable of modeling and simulating multi-domain physical systems that can also include mechanical, electrical, fluid, heat, magnetic and control elements. It proposes a complete implementation of a Modelica parser which can be extended to become a full featured Modelica compiler in future. Another contribution of this thesis is the new web document format that can be used to describe simulation models of physical systems. It puts forward a new standard format for representing a particular type of information on the web, i.e., information on the composition and inner workings of a machine, a mechanism, or a physical phenomenon. Based on the structure of the new web document format, a Modelica MySQL database of Modelica standard library 3.1 is produced to allow easy searching and data manipulation and which can be used for future research. The thesis also includes a detailed review of existing literature on web-based modeling and simulation.DOCTOR OF PHILOSOPHY (MAE