An improved method of supercapacitor aging prediction based on Prophet under actual working condition

Supercapacitors have a specific application in power supplement and energy recycling of electrical vehicles. The actual service life of supercapacitors under different working conditions may be inconsistent with the nominal life measured in the laboratory. This paper proposes a method for supercapacitor aging prediction with data from buses under actual working conditions. Aiming at the low sampling frequency and accuracy of the monitoring data, the limited memory least square method is used to identify the parameters along with the penalty function method. Considering the periodic change of aging indicators, a life prediction method based on an improved Prophet algorithm based on Fourier analysis is proposed. Finally, the forward chain method is used to verify the results, the prediction error of C and ESR is lower than 2.5% within 150 days