MLE in bearings-only parameter estimation

A Binder-compatible repo with a requirements.txt file.
Access this Binder at the following URL:

https://mybinder.org/v2/gh/orl1k/mle-in-bearing-only-estimation/HEAD

Project structure:

Problem Statement:

Find parameters of uniform and rectilinear object's motion given physical system, maneuvering observer
Common object trajectory example:

Observer's data is formed of bearing (angle) measurements with normal distributed error
Example of real data vs noised data:

Implemented 4 methods to solve the problem:

• maximum likelihood using lev-mar algorithm and embedded in scipy
• dynamic maximum likelihood
• n bearings

Convergence of maximum likelihood lev-mar:

Algorithm comparsion:

For more results check folder tests.
For more information check research paper:

https://github.com/orl1k/mle-in-bearing-only-estimation/blob/main/images/research.pdf