针对以往船舶压载水仿真系统仅能用于某些特定船舶的问题，提出基于管网有限元法的参数化建模方法。为实现对压载水调节方案的优化，在仿真系统中加入基于改进非劣分层遗传算法（Non-dominated Sorting Genetic Algorithm Ⅱ, NSGA-Ⅱ）的自动调载算法和基于反向传播（Back Propagation, BP）神经网络的调载方案搜索方法。通过C#、MATLAB和结构化查询语言（Structured Query Language, SQL）等混合编程手段开发一套船舶压载水仿真系统。经不同船舶上的应用表明，该系统具有通用性，能够根据目标自动生成合理的调载方案。
In view of the problem that the previous ship ballast water simulation system can only be used for certain specific ships, the parametric modeling method based on the pipe network finite element method is proposed. In order to realize the optimization of ballast water regulating scheme, the automatic regulating algorithm based on Non-dominated Sorting Genetic Algorithm Ⅱ (NSGA-Ⅱ) and the regulating scheme search method based on Back Propagation (BP) Neuron Network are added to the simulation system. A ship ballast water simulation system is developed by the hybrid programming methods such as C#, MATLAB, Structured Query Language (SQL), etc. The application in different ships shows that the system possesses the generality and can automatically generate the reasonable regulating scheme according to the target.