Aiming at the impact of rapidly changing waveform modal caused by DC nonlinear load, such as sudden rise or drop waveform of voltage and current waveform, current oscillation, or stepped change on DC electricity meters, this paper employs a mechanism-based modeling approach. Firstly, mathematical models for the signal preprocessing unit, signal conversion unit, power measurement unit, and energy measurement unit of the DC electricity meter are established. Secondly, mathematical models for eight typical single waveform modes of current and voltage, as well as three typical combined waveform modes of current and voltage, are developed. Subsequently, a simulation analysis is conducted on the influence of 11 dynamic energy signal waveform modes on the dynamic error of the DC electricity meter. Finally, the waveform mode which has the greatest impact on the dynamic error of the DC electricity meter is given as follows: in the single waveform mode, the dynamic error of the DC electricity meter is the largest when the peaked waveform mode is input, reaching -5.16×10-2. In the combined waveform mode, the dynamic error is the largest for the waveform mode labeled as number 9, reaching -1.05×10-3, as well as the type of power filter that most significantly affects the dynamic error, is identified. The conclusions of this paper can provide a theoretical basis for establishing testing and evaluation conditions for the dynamic error of DC electricity meters.