Start with defining your goal. How should the result look like? Is it enough to write some scripts without documentation, which produce some diagrams, or should it have a professional level, because you want to sell the solution or submit it as PhD thesis?
Do you have the needed Matlab skills to implement this according to the defined goal? If not, start with Matlab's Onramp tutorial and some smaller projects.
You find a large variety of different "finite element software" methods. E.g. the grid can be fixed or adaptive, the code can be easy to read or highly optimized, etc. "The software" might mean, that you write the calculator by your own, or that your Matlab code creates the initial values only to create files used as inputs for professional programs.
"Magnetic field" can include relativistic effects e.g. for calculating the electro-magnetic fields on the surface of a magnetar. Maybe quantum fields matter, if you simulate the near-field of oscillating Cooper pairs in superconductors. Or it is a simple macorscopic simulation.
"calculation of the coordinates at the point of impact of the particles with the wall of the rotating cylindrical electrode" - It is really hard to follow you. On one hand you give far too few details to understand, what you want to do, and on the other hand some points contain too many details.
A "graphical representation" can be a real-time 3D-animation during the simulation, a post processed 3D animation, a set of 2D diagrams of the coordinates, or a bar graphic of the statistical distribution of some events.
So your point to start from is the decision, what you want to achieve. There is a large field of different approachs, which match the given keywords.
Currently your seem to be in the state of defining, which problem you want to solve. Then there is not relation to Matlab yet, because this phase is solved best with pencil, paper and coffee.