This paper focuses on the parameter identification issue of electrochemical double layer capacitors (EDLCs). The superiority of the fractional-order equivalent impedance model is revealed mathematically and electrochemically by analyzing the variation trends of different models and the general electrochemical impedance spectroscopy (EIS) of EDLCs. Since the fractional-order models can be used to describe the long-tailed variation trend of the EIS of EDLCs in the low-frequency band and can reflect the self-charging phenomenon of EDLCs, the accuracy of parameter identification can be ensured. In addition, the idea of Levy flight is introduced and combined with the intelligent algorithms in the parameter identification process, thus accelerating the convergence rate of the parameter identification process. At the same time, the problem of nonconvex fitness function falling into local optimum can be solved. To confirm the effectiveness and superiority of this work, we provided diverse test scenarios. We used three different types of EDLCs, while a series of scenarios with or without Levy flight strategy were included in the tests. In our work, we compared not only the accuracy of the parameter identification but also the convergence rate of the identification process. The test results show that, by applying the proposed scheme, the sum of square error (SSE) between experiments and parameter identification results is less than 1.9%. Moreover, the convergence rate of the parameter identification process was improved. In extreme conditions, the convergence rate is 1287% faster than the schemes without Levy flight strategy.