Training method of day length change forecasting model and day length change forecasting method and device

The invention discloses a training method of a day length change forecasting model, and a day length change forecasting method and device. The training method comprises: fitting a first historical day length change sequence based on a preset fitting algorithm, and determining a first fitting residual sequence and a day length extrapolation value of a first target time point based on an obtained first fitting curve, the first fitting residual sequence comprising fitting residuals corresponding to a plurality of first historical time points; determining a fitting residual corresponding to the first target time point based on the daily length extrapolation value and the daily length actual value of the first target time point; performing phase-space reconstruction on the first fitting residual error sequence on the basis of an embedded dimension matched with the first fitting residual error sequence and delay time to obtain first phase-space representation; and taking the first phase space representation as a tr