Abstract: According to the characteristic of actual measurement viscosity,temperature and blending ratio of Kazakhstan’s crude oil,a mathematical model that viscosity changes along with temperature and blending ratio is established.Meanwhile Particle Swarm Optimization is adopted to compute the viscosity- temperature of the mathematical model.It id indicated that the optimization method can quickly search the optimal regression parameters of the model through establishing and computing of the three algorithm examples.It is pointed out that the form of commingled crude’s viscosity-temperature mathematical model should be further studied to find out a suitable optimization method for model computation for the purpose of making wider range of model application and higher fitting accuracy.