The invention provides a cotton production process optimization method based on hierarchical clustering of big data, which comprises the following steps: data preprocessing is carried out on the obtained production monitoring original data; the key parameters describing the parameter attributes for the pre-processed raw data are determined; numerical statistics of the determined key parameters isused to obtain the numerical distribution grouping of the distribution statistical attributes. According to the numerical distribution grouping obtained from distribution statistics, the attribute data in the sample are mapped into each attribute grouping interval to form a new data set. The optimization objective is determined and the production process parameters are optimized. Cotton processingprocess analysis is combined. It can be used for the analysis and adjustment of process parameters so as to optimize cotton production, improve cotton quality and safeguard the production safety of enterprises.