Abstract: In order to realize real-time and accurate detection of water turbidity in the water treatment process, the turbidity detection system was designed based on infrared light scattering and the turbidity forecasting model was put forward based on clustering grey fusion. The infrared light emitting diode with 890 nm wavelength was used as the light emitting device, the photosensitive diode was used as the receiver, and the response time of the detector was short, and the zero error was small. The data collected by the sensor was processed by the method of grey prediction algorithm and cluster fusion. The data processed by the cluster fusion were as the input data of the grey predictive control, and the output data of the grey predictive control and the fusion data were compared and analyzed. Data tracking and operation were carried out through the actual project. The average error of the measured value and the output value of the turbidity prediction is 0.008 7 NTU. Grey fusion algorithm is superior to the single grey prediction algorithm, to ensure that the water quality turbidity parameters are stable and meet the requirements of water quality, and ensures that the water quality turbidity parameters are more stable and meet the requirements of water quality.