Model Predictive Control is an important tool in self-driving cars. It is used to force the vehicle to follow a given trajectory generated by the path planning module. In this project, I wrote a simple MPC class in Python & TensorFlow based on the C++ implementation provided in Udacity's self-driving car nano-degree program.
The following plots show the results from the Python/TensorFlow and C++/IPOPT implementations, respectively.