Research activities

Featured research:

  1. Model predictive control of a tower crane

  2. Model predictive control of a VRLA battery

  3. Flexible Lyapunov function based model predictive direct current control

  4. Current Reference Governor of Permanent Magnet Synchronous Machine

 

Model predictive control of a tower crane

Cranes are complex, non-linear, under-actuated and oscillatory systems that are widely
used in transportation systems to transfer heavy loads. During transfer of the heavy loads, the main problem is an oscillatory motion of the payload.

We develop different model predictive control algorithms for controlling the crane system subject to constraints.

 

 

We model the crane as three dynamically coupled subsystems and treat the coupling among them as a variation in the system’s parameters. Then the MPC problem can be solved in a distributed way.

When developing a distributed MPC control law our main concern is stability and recursive feasibility.

Proposed solutions: 

  • dual-mode control law
  • sequence of 1-step controllable sets

We test our algorithms in simulation and experimentally on a scaled model of a tower crane

 

 

References

Šandor Ileš et al. “Real-time Predictive Control of 3D tower crane”. In: Industrial
Electronics (ISIE), 2014 IEEE 23rd International Symposium on. IEEE. 2014,
pp. 224–230

Šandor Ileš et al. “Stabilizing model predictive control of a gantry crane based on flexible
set-membership constraints”. In: IFAC-PapersOnLine 48.23 (2015), pp. 248–253

 

Other related topics:

  • Friction compesation
  • Linear matrix inequalities based crane control
  • Reference governor in crane control

References

Matuško, Jadranko; Ileš, Šandor; Kolonić, Fetah; Lešić, Vinko. Control of 3D Tower Crane based on Tensor Product Transformation with Neural Friction Compensation. // Asian journal of control. 17 (2015) , 2; 443-458

Ileš, Šandor; Matuško, Jadranko; Kolonić, Fetah. Constrained Reference Tracking based on Homothetic Sets. // Automatika : časopis za automatiku, mjerenje, elektroniku, računarstvo i komunikacije. 56 (2015),  4; 411-423

Model predictive control of a VRLA battery

A fast charging of a battery often comes at the cost of a reduced battery lifetime.

We deal with the problem of fast charging considering constraints on the battery voltage, temperature and current.

For VRLA battery we developed a model predictive controller with proven convexity, recursive feasibility and stability of the battery charging optimization problem.
 

 

References

Goran Kujundžić et al. “Optimal charging of valve-regulated lead-acid batteries based on model predictive control”. In: Applied Energy 187 (2017), pp. 189–202

Flexible Lyapunov function based model predictive direct current control

 

We deal with model predictive direct current control for a generator side converter to minimize the switching losses.

We propose using a flexible Lyapunov function to minimize the switching losses during tranzients and steady state and guarantee stability and recursive feasibility.

References

Tin Bariša et al. “Flexible Lyapunov function based model predictive direct current control of permanent magnet synchronous generator”. In: Power Electronics and Motion Control Conference (PEMC), 2016 IEEE International.
IEEE. 2016, pp. 98–103

 

Current Reference Governor of Permanent Magnet Synchronous Machine

 

We propose using a reference governor to augment classical field oriented control to handle voltage contraints of permanent magnet synchronous machines.

We proposed using an octagon to approximate the quadratic voltage constraint.

The proposed reference governor has proven recursive feasibilty and it is implemented as process in the loop simulation on Texas Instruments F28335 microcontroller.

References

Tino Jerčić et al. “Current reference governor of permanent magnet synchronous machine”. In: Electrical Machines (ICEM), 2016 XXII International Conference on. IEEE. 2016, pp. 1022–1028