Bi-LSTM-based pest and disease prediction method and system

The invention discloses a Bi-LSTM-based pest and disease damage prediction method and system, and the method comprises the steps: obtaining historical rice pest and disease damage data, carrying out the arrangement to form a first rice pest and disease damage data set, and for a specific disease of a specific region, enabling the data items of the first rice pest and disease damage data set to comprise an occurrence time period, a pest and disease damage evaluation value and a plurality of weather factor items; calculating the correlation between the pest evaluation values in the first rice pest data set and each weather factor item, and selecting the weather factor items with the first five correlation ranks and the corresponding pest evaluation values to form a second rice pest data set; dividing the second rice pest data set into a training set and a test set, and converting the training set and the test set into an input form adapted to a Bi-LSTM neural network model; the Bi-LSTM neural network model is t