Applying Wavelet Decomposition and Reconstruction to Solve Conflict Between Macroscale and Microscale Simulation Step of Buck Converter

摘要

To accurately describe the operation of power electronic converters, the switching device must be depicted by a microscale model based on quantum mechanics, which makes the step selection in simulation analysis difficult. Choosing the step suitable for the switching device will result in long calculation time and nonconvergence, but the accessible step for the circuit will result in the switching device characteristics' noncoincidence with actual. In order to solve the conflict between macroscale and microscale simulation steps, wavelet decomposition and reconstruction are applied in the simulation model. First, the method of coarse-to-fine scale transformation is confirmed. According to the Nyquist-Shannon sampling theorem and the partition rule of dynamic sampling grid of discrete wavelet decomposition and reconstruction, the simulation step should be selected smaller than a quarter of the switching time. Second, the macroscale model of circuit and the microscale model of switching device are combined based on wavelet decomposition and reconstruction. Third, in the simulation of the combined model, the exact solution of the switching process is obtained from the imprecise solution by using discrete wavelet decomposition and reconstruction. The operation of the switching device is precisely reflected. Finally, the simulation result is verified to be in good agreement with the experiment. The solution proposed in this article can be expanded to the simulation analysis of other power electronic converters.

关键词