Model Predictive Control-based Motion Cueing for 4-Degree of Freedom Simulator Platform
Vehicle simulators are widely used for many reasons: user training, vehicle model testing, etc. Generally, vehicle simulator consists of visual simulation system and movement simulation system. Motion cueing is the part of movement simulation system which calculates the platform’s position to produc...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/42534 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Vehicle simulators are widely used for many reasons: user training, vehicle model testing, etc. Generally, vehicle simulator consists of visual simulation system and movement simulation system. Motion cueing is the part of movement simulation system which calculates the platform’s position to produce the sensation of real vehicle movement to the user while also considers the platform’s mechanical restrictions in its calculation and algorithm. This thesis explains about the design of motion cueing system for Institut Teknologi Bandung’s simulator platform which has 4 degrees of freedom (DOF) for its movements. Currently, the implemented motion cueing system uses Classical Washout Filter (CWF) algorithm which suffers from the platform movement limitation that potentially causes false perceived sensation when the platform stops moving in its limits and unavoidable steady-state error which is caused by the algorithm itself. Thus, Model Predictive Control (MPC)-based motion cueing for 4 DOF simulator platform is developed to overcome this problem since MPC involves the plant’s model in its calculation. Previously, there was no MPCbased motion cueing algorithm that was specifically designed for 4 DOF simulator platform. The design process consists of plant modelling, determining cost function and its parameters, solving quadratic problem of cost function, and program design. From the simulation and implementation, it can be inferred that MPC-based motion cueing has lower steady-state error (1% for surge sensation and 5% for sway sensation) and faster settling time (2 seconds for surge and 10 seconds for sway) compared to the CWF-based one for the linear acceleration perception. The performance can also be adjusted further by tweaking the weighting matrix that is used in the cost function. Lastly, the calculation of MPC-based motion cueing includes platform limits thus the platform operation is always within its physical constraints. |
|---|