Turbidity is an important index of water quality and the requirement of turbidity measurement is increasingly stringent. At present

the main method of turbidity measurement is photoelectric detection such as scattering method and transmission method. Design a new online turbidity sensor by integrating three turbidity measurement methods together which are scattering method

transmission method and ratio method. This turbidity sensor follows international standard ISO7027 and adds temperature and light to overcome the interferences. Least square support vector regression（ LS-SVR） used for model turbidity and experiments are carried out to analyze the method. Experimental results show that this method can effectively improve the detection precision and antiinterference capability.