FRNT05 Nonlinear Control Systems and Servo Systems
Lecture 13: Control Systems with
discontinuities
YIANNIS KARAYIANNIDIS, ASSOCIATE PROFESSOR www.yiannis.info
AUTOMATIC CONTROL, FACULTY OF ENGINEERING.
[email protected] Outline
• Time optimal control
• Sliding mode control
Sliding set
• Sliding set:
Sliding set
Example
Example: Sliding set
Example: Sliding dynamics
Sliding mode control
Sliding mode control
Sliding mode control
Equivalent control
Sliding mode control
Sliding mode control
Example
- Chattering → wear on
mechanical devices
Example
• Contninuous control
Friction
Lubrication regimes
For low velocity: friction increases
with decreasing velocity
Stribeck (1902)
Hysteresis and friction
• Dynamics are important also outside sticking regime
• Hess and Soom (1990)
• Experiment with unidirectional motion
• Hysteresis effect!
Classical friction models
Advanced Friction Models
• Karnopp model
• Armstrong's seven parameter model
• Dahl model
• Bristle model
• Reset integrator model
• Bliman and Sorine
• LuGre model (Lund-Grenoble)
See PhD-thesis by Henrik Olsson
https://lucris.lub.lu.se/ws/portalfiles/portal/4768278/8840259.pdf
Friction models with extended state
Dahl’s model
LuGre model
Contact between bristles
Friction and control
• Friction compensation
– Lubrication High-frequency mechanical
– Dither vibration: used to avoid sticking
– Integral action Integral term compensate for
– Non-model based control slowly varying disturbances
– Model-based friction compensation
– Adaptive friction compensation
Friction and control
• Friction compensation • To be useful for control the model should
– Lubrication be:
– Dither – sufficiently accurate,
– Integral action – suitable for simulation,
– Non-model based control – simple, few parameters to determine.
– Model-based friction compensation – physical interpretations, insight
– Adaptive friction compensation
• Simple models should be preffed.
• If no stiction occurs the v=0-models are
not needed.
Adaptive friction compensation
Velocity Control – Results
Quantization
Linear Model of Quantization
• e: Noise independent of u with rectangular
distribution over the quantization size with
rectangular distribution with variance
• Digital signals have specific number of bits
(accuracy and range of signals) (e.g. 8 bits, 64 bits)
• Quantization in A/D and D/A converters
• Quantization of parameters • It works if quantization level is small
• Roundoff, overflow, underflow in operations compared to the variations in u
For prediction of limit cycles
• Use describing function analysis
Sliding mode control
Sliding mode control
Quantization: Describing function
• Recall the deadzone nonlinearity
