Injection-Based Online Capacitance Monitoring With Optimal Selection of Injection Frequency in Buck Converters

The aging process of aluminum electrolytic capacitors leads to a reduction in capacitance, which affects the safe operation of switching power converters. This article proposes an online capacitance monitoring method of Buck converters with high capacitance estimation accuracy. The proposed method does not ask for specific operating condition of the circuit, and does not need to model complex circuit dynamics in time domain. With the measurement errors considered, the article mathematically proves that the minimum error of capacitance estimation can be achieved with maximum sensitivity. Without interrupting the operation of the original system, a small disturbance with a certain frequency is injected to the duty cycle of the switches, to generate additional information for capacitance estimation. To achieve optimal sensitivity for capacitance estimation, the injection frequency is selected as the characteristic frequency based on the transfer function for capacitance estimation. Different candidates of transfer functions are also compared and evaluated to achieve minimum capacitance estimation error. Simulation and hardware experiments verify that optimal parameter sensitivity of capacitance can be achieved at the characteristic injection frequency, and the capacitance can be accurately estimated based on the proposed method.