The invention discloses a photovoltaic cell parameter recognition method based on FFRLS (forgetting factor recursion least square), and the method comprises the following steps: building a recursion least square model of a photovoltaic cell, and determining a to-be-recognized parameter; initializing a parameter estimation value, a forgetting factor and a covariance matrix; obtaining the real-time output voltage and current of the photovoltaic cell, and updating the parameter estimation value and a target function value; outputting the parameter estimation value at this moment when a target function meets a preset threshold value, wherein the parameter estimation value is taken as the optimal parameter value of the photovoltaic cell. According to the invention, the method enables a four-parameter model of the photovoltaic cell to be converted into the recursion least square model, shields the impact on the model from the internal structure of the photovoltaic cell and system errors, and obtains the optimal parameter value through iteration. The method is easy to implement, can reduce the calculation burden, reduces the memory, occupied by data, of a computer, and improves the recognition speed. The forgetting factor can highlight the effect of data, the effect of gradually forgetting old data, enables the model to be higher in precision, and enables the stability of data to be better.