The invention discloses a prediction Method of Internal Resistance attenuation coefficient for Photovoltaic Cells. The method comprises the following steps: establishing a time series of an evolutionary system of an internal resistance attenuation coefficient of a photovoltaic cell according to a parameter obtained by real-time monitoring, establishing an equation of the internal resistance attenuation coefficient of the photovoltaic cell; performing a phase space reconstruction processing on the time series of the evolution system; processing a battery internal resistance attenuation coefficient equation according to a wavelet network method; predicting and calculating the internal resistance attenuation coefficient of the photovoltaic cells on the time series after the phase space reconstruction. The invention has the advantage of solving the technical problems that a power grid and distributed photovoltaic power generation operation data resources cannot be effectively utilized, and evaluation accuracy and the efficiency of PV are low; the reliability of the model is improved; the utilization rate of the PV is improved; the accuracy of the evaluation is improved; and the technical effect of the reliability and economy of the distribution system after the PV system is improved. The invention is applied to detect a photovoltaic power generation system.