About the Takei Laboratory
Thank you for visiting the Takei Laboratory homepage. Our laboratory is engaged in research that uses the power of information science to elucidate electromagnetic and acoustic phenomena and related physical phenomena that are used in our living environment and product design.
What is Simulation?
In addition to the two pillars of theory and experimentation that have been used as research tools in the field of science and engineering, the third pillar of ``simulation'' is attracting attention. A feature of this simulation is that it uses computers to clarify physical phenomena such as electromagnetic phenomena and the interaction of multiple different physical phenomena such as electromagnetism, heat, and sound.
In order to clarify physical phenomena, it is necessary to solve equations that govern physical phenomena, often partial differential equations, but these partial differential equations can only be solved under simple conditions such as spherical symmetry or cylindrical symmetry.
In other words, in order to understand the physical phenomena that occur in complex structures and spaces such as motors, the inside of the human body, and the living environment, it is necessary to numerically solve partial differential equations using computers and software.
To be more specific, by replacing the partial differential equation with a linear equation (simultaneous linear equations) using a mathematical procedure and solving this linear equation using a matrix calculation algorithm, we can indirectly and approximately solve the partial differential equation.
This is called numerical analysis. Numerical analysis was born before the advent of computers, and has evolved along with the development and spread of computers.
Numerical analysis and simulation
There are various options for how to take this step forward, but we are focusing on "advanced physical/mathematical modeling" that reproduces complex and large-scale structures and spaces, and parallel technology that makes it possible to perform calculations. We are trying to achieve this through ``advanced algorithms'' such as . We present this as a research theme.
In this way, simulation methods based on physics, mathematics, and information science can obtain solutions with physical accuracy equivalent to experiments under correct modeling and boundary and initial condition settings. However, numerical analysis only "approximately solves" partial differential equations, and the solutions always contain errors. since simulations that reproduce real phenomena are based on numerical analysis, they are naturally subject to errors.
However, in general, as the scale of a linear equation becomes larger, the amount of calculation and memory increases, and the calculation itself becomes more difficult due to the impact on the convergence of the iterative solution of linear equations. In other words, there is a trade-off between obtaining highly accurate solutions and ease of calculation (ease of finding solutions). This is the difficult part of simulation.
To achieve highly accurate simulation
To achieve highly accurate simulation
1. Accurate understanding of the phenomenon of interest
2. Implementation of sophisticated numerical modeling to handle the phenomenon through simulation
3. Development of algorithms that can calculate solutions with high resolution in both space and time,
and technology for using parallel computers such as supercomputers
is required. Highly accurate simulation cannot be achieved overnight, and requires continuous study and improvement of the three points mentioned above.
This process is very messy, and all the students and researchers working on the research have to overcome each obstacle in front of them,
one by one, and ultimately reach the goal of realizing a high-precision simulation method. I'll get there.
Finally
Regardless of the field of research, great effort is required to accomplish research without exception.
At first glance, simulation research seems light and easy to approach (I thought the same way before I started research).
However, once you start working on it, you will quickly realize that it is profound and has many difficulties.
Simulation is a new research field that can be called a comprehensive science that spans engineering, science, and information science.
Therefore, compared to research fields where theory and experiment are the main methods, the accumulation of research results is small,
and this is thought to be the reason why there are many difficulties in proceeding with research.
However, the flip side of the many difficulties is that we can expect great research results.
Also, the fact that it has a very large ripple effect on various fields is what makes the research so rewarding.
I would be happy if all the students who viewed this page felt that ``simulation looks interesting'' or ``I want to try it.''
Please come and visit the faculty's laboratory. In that case, please make an appointment to the email address below and Takei will be able to accommodate you.
In addition, after being assigned to a laboratory, whether you wish to get a job after graduating from an undergraduate degree or wish to proceed to a master's or doctoral course,
you will be able to play an active role in society and become a person with a promising future. Takei will train you through seminars and research meetings.
If you are a company viewing this page and are interested in simulation technology, please feel free to contact us.
We also welcome requests for joint research and commissioned research.
