论文部分内容阅读
通过分析与溢油污染程度有关的影响因素,首次构建了海上石油平台溢油污染程度评价指标体系。针对模型无样本的难题,对评价指标进行分级,利用Rand函数在各分级标准内随机生成足够数量的训练和测试样本,建立了较合理的网络结构,构建了石油平台溢油污染等级BP网络模型。研究结果表明BP网络模型具有很强的泛化能力,能够用于评判未知样本,具有较强的实用性。
By analyzing the influencing factors related to the pollution degree of oil spill, this paper constructed the evaluation index system of oil pollution degree of offshore oil platform for the first time. Aiming at the problem of no sample in the model, the evaluation index is graded, a sufficient number of training and testing samples are randomly generated in each grading standard by using Rand function, a more reasonable network structure is established, and a BP network model of oil platform pollution level is constructed . The results show that BP network model has strong generalization ability, which can be used to judge unknown samples and has strong practicability.