BP Neural Network in the Quality Evaluation of Tanyang Congou Black Tea

For a long time, sensory evaluation method has been usually used to assess the quality of tea. But it affects the accuracy of the assessment because of the sensitivity of the human sense organs which are vulnerable to external disturbances. For the tea sensory evaluation deficiencies, MATLAB neural network was used in this research and a BP Neural Network model was built based on the sensory evaluation factor in a sample of Tanyang Congou Black Tea. The results showed that BP Neural Network model can accurately reflect the nonlinear relationship between the grade of Tanyang Congou Black Tea and sensory evaluation factors. The tasters who have knowledge of tea assessment can correctly assessed the grade, overcome subjective error caused by human and provides a more efficient, objective and impartial method for the quality accessment of the tea.