Matteo MacchiniStudent meeting - May 2014 Motion control design for the new BWS Matteo Macchini Technical student BE-BI-BL Supervisor: Jonathan Emery
Matteo Macchini Outline Student meeting - May 2014 Beam Wire Scanner overview Motor selection and sizing procedure Motor control Tuning of the controllers Simulations
Matteo Macchini Beam Wire Scanner Student meeting - May 2014 Purpose: Evaluate the profile of the beam into the accelerators
Matteo Macchini Beam Wire Scanner Student meeting - May 2014 Electromechanical system Motor Position/speed sensors Thin wire Scintillator Photomultiplier
Matteo Macchini Desired wire motion Student meeting - May 2014 Three phases “Constant” acceleration Constant speed (beam crossing) “Constant” deceleration
Matteo Macchini Motor selection and sizing Student meeting - May 2014 Air-gap thickness Torque to inertia ratio Vacuum compatibility Radiation tolerance Temperature tolerance Torque ripple Specifications
Matteo Macchini Permanent Magnet Synchronous Motor Student meeting - May 2014 Model: PARKEM k Y Air gap thickness: 0.3mm ÷ 0.75mm Inertia: kg*m 2 Max torque: 32.8Nm
Matteo Macchini PMSM input Student meeting - May 2014 Three phase sine wave Flux = LC of the 3ph fluxes Motion is achieved keeping the flux spin Torque = k*(Flux)x(Theta) So MAX torque ↔ Flux ⊥ Theta
Matteo MacchiniStudent meeting - May 2014 Three phase static ↕ CLARKE ↕ Two phase static ↕ PARK ↕ Two phase rotating Clarke – Park decomposition
Matteo MacchiniStudent meeting - May 2014 Simulated system Overall system Simplified system
Matteo Macchini Motor control Student meeting - May 2014 FOC – Field Oriented Control Control method T = k*i q (if i d =0) Double feedback control Current Quadriture (αT) Direct(αF) Speed Space Vector Modulation Controllers
Matteo Macchini PID controllers Student meeting - May 2014
Matteo Macchini PID controllers Student meeting - May 2014 PID(s)PID(s)G(s)G(s) ue r y + -
Matteo Macchini PID tuning Student meeting - May 2014 THEORETICAL APPROACHES Transfer function computation and simulation Pole-zero compensation … unpopular and not very efficient
Matteo Macchini PI control and tuning Student meeting - May 2014 EXPERIMENTAL APPROACHES Open loop methods Ziegler-Nichols Cohen-Coon Cooper Closed loop methods Z-N closed loop ATV Adaptive methods PSO “Modern” methods H ∞ controller H 2 controller μ controller
Matteo Macchini Particle Swarm Optimization Student meeting - May 2014 Particle filter based on the behaviour of swarms of birds Parametric optimization in order to follow local-global optimums Looking for MIN of a quality function (IAE)
Matteo Macchini Results classic tuning (so far…) Student meeting - May 2014 Current loopSpeed loop
Matteo Macchini Results PSO (so far…) Student meeting - May 2014 Current loop Speed loop i=1 i=4 i=7i=10 i=4 i=7 i=10
Matteo Macchini In the future… Student meeting - May 2014 NEXT GOALS Space Vector PWM implementation Position loop Complete system simulation See what happens with the real system!