Trend prediction method and prediction system for price limit instruction book

The invention relates to a trend prediction method and prediction system for a price limit instruction book, and the method comprises the following steps: obtaining the data of the price limit instruction book, carrying out the preprocessing of the data of the price limit instruction book, and obtaining the preprocessed data; constructing a stacked residual gated recurrent unit neural network, employing the preprocessed data for training the neural network, and obtaining the trained stacked residual gated recurrent unit neural network, wherein the stacked residual gated recurrent unit neural network comprises multiple residual blocks based on a GRU layer, and the multiple residual blocks based on the GRU layer are arranged in a stacked mode; and performing trend prediction on the price limit instruction book by using the trained stacked residual gating circulation unit neural network. According to the method, the stacked residual gating circulation unit neural network model is provided and is used for predict