中文摘要：

应用灰色关联度分析方法确定了与待预测状态量关联度较高的因素,并利用熵理论建立了具有客观权重的组合预测模型。预测区间可有效量化由不确定因素引起的油中溶解气体浓度波动,应用比例系数法和粒子群优化算法建立了一定置信水平下油中溶解气体浓度的区间预测模型,且不受传统区间预测方法中必须服从正态分布的限制。实例结果验证了所提模型的有效性。

英文摘要：

The factors highly related to the variables to be predicted are confirmed by the grey relational analysis and a combination prediction model with objective weight is built based on the entropy theory. Since uncertain factors may influence the dissolved-gas concentration in transformer oil and the prediction interval can effectively quantify its fluctuation,the proportionality coefficient method and particle swarm algorithm are adopted to build an interval prediction model of dissolved gas in transformer oil at a certain confidence level,which,different from the traditional interval prediction method,does not have to obey the normal distribution limitation. The calculative results for an example show the effectiveness of the proposed model.