In recent years, manufacturing has focused on the combination of “Real (Measurement)” and “Virtual (Simulation)” such as Digital Twin. Data assimilation is one of the methods to realize the Digital Twin. On the other hand, “Physical modeling” is often used for complex plant models because of high readability, modifiability, and multi-domain support. In our privious work we developed Data assimilation library using Modelica Language. We extended this library to Particle Filter and apply to inverted pendulum and assimilated measurement data to simulation.