Fluidized bed granulation is widely used in the production of traditional Chinese medicine granules，but the traditional granulation process relies on the operator to observe the material state in the fluidized bed，and manually adjust the air volume and liquid injection speed according to experience，which has the problems of time-consuming，labor-intensive and insufficient intelligent level. Therefore，this paper constructed a fluidization state value feedback control system based on near infrared spectroscopy，and evaluated its control performance. The near infrared spectrum of the fluidized bed granulation process was collected in real time to analyze the movement state of the material and calculate the fluidization state value index. Then，the system automatically adjusted the peristaltic pump speed in the spray system according to the fluidization state value index and the preset control rules，thereby realizing the dynamic control of the spray speed and achieving the fluidization state value real-time feedback control. The experimental results show that the feedback control system can effectively prevent the occurrence of bed collapse when the fluidization state deteriorates. When the fluidization state value is good，the system can speed up the liquid spraying，thereby saving the granulation time and improving the production efficiency. Compared with the manual control mode，the automatic control system reduced the liquid injection time by more than 20%. In terms of fluidization state value control，the system is equivalent to the manual control，while it shows significant advantages in the optimization of liquid injection time. In addition，the control system shows good applicability and stability under the interference conditions of changing process conditions such as inlet air temperature，inlet air volume，atomization pressure，etc. Therefore，the fluidization state value feedback control system of fluidized bed granulation process based on near infrared spectroscopy established in this paper has stability and reliability，and can provide strong technical support for the intelligent manufacturing of fluidized bed granulation.