Based on the PB Neural Network of Optimization Design in Lubricant Additives
This article use the Sulphide Isobutene, Five Sulfides Dialkyl, and Star of Phosphorus as the additives, Neopentyl Polyol Ester (NPE) as base oil for screening lubricant formulation. The purpose of this article is screening the lubricant additives formula. Apply the BP neural network method in optimization design. Through the optimization of lubricant additive formula select the best formula for experiment. The selected best formula is Sulphide Isobutene 0.8%(mass percent), Five Sulfides Dialkyl 1.2%(mass percent) , Star of Phosphorus 1.6%(mass percent), relative error is 0.089.After validation experiment，it is conclusion that S-type blends with P-type additive use will acquire good result, and the method of optimal convergence faster, the forecast precision test is satisfied.
Z. J. Chen and L. L. Feng, "Based on the PB Neural Network of Optimization Design in Lubricant Additives", Advanced Materials Research, Vols. 311-313, pp. 218-222, 2011