关键词: 储油罐, 变位, 遗传算法, 反向传播（BP）网络

Abstract: In view of tank capacity charts are distorted after oil tank dislodge, the integral model is established with micro-element method. The function relationship between oil volume, oil height, vertical angle and crosswise deflection angle is obtained. Dislodgement parameter is determined with least squares method and Matlab software. Also deflection angle is amended with genetic algorithms and BP networks to improve the accuracy. Then tank capacity charts are revised after oil tank dislodge. The model has certain promotion through sample data.

Key words: oil tank, dislodge, genetic algorithms, Back Propagation（BP） networks