基于蚁群优化的风电机组绝缘可靠性灰色预测

目前风电机组可靠性分析主要是针对风电机组整体系统,缺乏对关键部件可靠性的深入研究,且需要处理大量数据,增加了可靠性测试的成本和难度,因此,在深入研究绝缘主要故障机理的基础上,将两种基于小样本数据的灰色模型引入到绕组绝缘可靠性预测中.为提高预测精度,以两种灰色模型为基础,以最小偏差平方和为标准提出了基于蚁群算法的绝缘可靠性优化模型.预测结果表明,绝缘可靠性满足威布尔分布.利用实测数据进行验证,两者的符合度较好,证明利用所提模型可以有效缩短可靠性测试周期,降低实验成本.

Grey prediction of insulation reliability for wind turbines based on ant colony optimization algorithm

The present reliability analysis of wind turbines mainly focuses on the entire system of wind turbines, while the lack in deep study on key components and requirement in dealing with large amounts of data will increase the reliability testing cost and difficulty. Based on the deep investigation on the main fault mechanisms of insulation, two grey models based on small sample data were introduced into the prediction of winding insulation reliability. To improve the prediction accuracy, the optimization model for insulation reliability based on ant colony algorithm was proposed on the basis of two grey models and through taking the least square sum of deviation as a criterion. The predicted results show that the insulation reliability meets the Weibull distribution. Through performing the verification with the actually measured data, the predicted results are in well agreement with the experimental data. It is proved that the proposed model can shorten the testing cycle of reliability, and reduce the testing cost.