Intelligent forecast-based maximum power point tracing method of photovoltaic multi-peak value

The invention relates to an intelligent forecast-based maximum power point tracing method of a photovoltaic multi-peak value. The intelligent forecasting-based maximum power point tracing method comprises the following steps of S1, building a photovoltaic array kinematic horizontal-projectile model by combining similarity of a photovoltaic array current-voltage characteristic curve and mass pointhorizontal-projectile motion track; S2, solving the photovoltaic array kinematic horizontal-projectile model by employing an improved particle swarm optimization method to achieve maximum power forecast (intelligent forecast), and acquiring the maximum power point and a voltage of the maximum power point; S3, fitting the photovoltaic array current-voltage characteristic curve by the photovoltaic kinematic horizontal-projectile model, and locally tracing the maximum power point by employing a voltage closed-loop control method after forecast to acquire a control signal, wherein the fitting might not be accurate; and S4