Abstract: The paper utilized high-temperature instantaneous gelatinization technology as a substitute for the traditional steaming technology.By measuring several indexes,such as water content,starch gelatinization ratio,enzymatic hydrolyzeα- amino nitrogen content and total fat content,the variations in the process of these indexes were obtained.Then corresponding models were set up based on artificial neural network.It could be concluded that established models are quiet accrate and can disclose the relationship among operating conditions and indexes by evaluation of two types of data(trained data and untrained data).

Key words: amino nitrogen content, artificial neural network, high-temperature instantaneous gelatinization, starch gelatinization ratio