Trajectory-control using deep System Identification and Model Predictive Control for Drone Control under Uncertain Load
- Resource Type
- Conference
- Authors
- Mahe, Antoine; Pradalier, Cedric; Geist, Matthieu
- Source
- 2018 22nd International Conference on System Theory, Control and Computing (ICSTCC) System Theory, Control and Computing (ICSTCC), 2018 22nd International Conference on. :753-758 Oct, 2018
- Subject
- Components, Circuits, Devices and Systems
Computing and Processing
Robotics and Control Systems
Signal Processing and Analysis
Drones
Neural networks
Predictive models
Training
Task analysis
Trajectory
Mathematical model
identification
model predictive control
neural networks
learning
- Language
Machine learning allows to create complex models if provided with enough data, hence challenging more traditional system identification methods. We compare the quality of neural networks and an ARX model when use in an model predictive control to command a drone in a simulated environment. The training of neural networks can be challenging when the data is scarce or datasets are unbalanced. We propose an adaptation of prioritized replay to system identification in order to mitigate these problems. We illustrate the advantages and limits of this training method on the control task of a simulated drone.