Lifetime extension by varying switching frequency of inverters based on junction temperature estimation

Saleki, Araz; Changizian, Mahmudreza; Rezazade, Saman; Bina, Mohammad Tavakoli

doi:10.1109/pedstc.2018.8343806

Cited by 12 publications

(1 citation statement)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A more complex problem is optimization of semiconductor thermal cycles, which exist on longer time scale. Improvement of thermal cycles can be achieved by varying power losses according to the junction temperature by changing switching frequency [29], gate-emitter voltage on IGBT [30] or output power [31]. A complex converter driver is needed in this case and it represents an significant disadvantage of this solution.…”

Section: Introductionmentioning

confidence: 99%

Maximizing power utilization of dual converter using predictive control with steady state cost

Votava

Šmı́dl

Peroutka

et al. 2021

IET Power Electronics

View full text Add to dashboard Cite

Capability to deliver maximum power of a converter is essential for increasing power density that is limiting in many applications. Therefore, control algorithm should guarantee operation within converter thermal limits. This requirement can be formalized by a hard constraint in the cost function of model predictive control (MPC). However, the temperatures of the converter elements are not constant in the steady state, which complicates evaluation of the cost on long prediction horizon. Therefore, the evaluation is simplified utilizing the analysis of steady state behaviour of the model and derived current derating laws calculated in off‐line manner. The derating law is used as the terminal set in MPC which allows using one‐step‐ahead evaluation for efficient real‐time implementation. The steady state analysis also provides coefficients for power loss balancing. The proposed approach is applied to control of dual converter, which has high redundancy of switching elements and, thus, wide space for optimization. It is shown in simulation that the proposed approach has better performance than previously published algorithms, at lower computational cost. Experimental evaluation of the algorithm performed on a converter prototype of rated power of 10 kW shows that the proposed controller is able to safely operate the converter near the thermal limit.

show abstract

Section: Introductionmentioning

confidence: 99%