High temperature superconductor (HTS) technology enables a significant reduction in the size and weight of MW-class generators for direct-drive wind turbine systems and reduce the cost of clean energy relative to conv...High temperature superconductor (HTS) technology enables a significant reduction in the size and weight of MW-class generators for direct-drive wind turbine systems and reduce the cost of clean energy relative to conventional copper an permanent-magnet-based generators and gearbox. Using MAXWELL, we studied MW class superconducting synchronous machines. By comparison the weight, we concluded that HTS wind turbine with rotor iron is the heaviest and HTS wind turbine without rotor iron and stator teeth is the lightest. By comparison the flux density, HTS wind turbine without rotor iron is the least and HTS wind turbine without rotor iron and stator teeth is the largest. By comparison the cost, HTS wind turbine with rotor iron is the highest and the other two is almost the same. HTS wind turbine without rotor iron and stator teeth is the best type.展开更多
风电机组在齿轮箱油温过高时会导致机组限功率运行,影响机组发电效率。传统应对风机高温降容状态多采用阈值判断,反应迟缓,加剧风机齿轮箱劣化趋势。利用贝叶斯网络对风机高温降容状态进行评估,为提取并准确合理地利用机组数据采集与监...风电机组在齿轮箱油温过高时会导致机组限功率运行,影响机组发电效率。传统应对风机高温降容状态多采用阈值判断,反应迟缓,加剧风机齿轮箱劣化趋势。利用贝叶斯网络对风机高温降容状态进行评估,为提取并准确合理地利用机组数据采集与监视控制系统(supervisory control and data acquisition system,SCADA)各个相关状态参数之间的耦合特性,通过vine-Copula模型对机组各个状态参数进行相关性分析,建立更符合机组实际运行状态的贝叶斯概率图形网络,实现对机组高温降容状态的评估。通过交叉熵算法对模型输出结果进行评价,发现与朴素贝叶斯模型相比,vine-Copula贝叶斯网络评估结果更为精确可靠,所建模型更符合机组实际运行工况,能够为现场的运维人员制定准确合理的运行和维护方案提供参考。展开更多
发电机、齿轮箱等风机部件的高温降容状态是表征风电机组亚健康状态的良好指标,其评估的准确性直接影响了后期人员、设备、资金等多种资源的投入的多少,以及运维方案的最终效果。为了尽可能真实客观反映风机高温降容状态,提出了一种基...发电机、齿轮箱等风机部件的高温降容状态是表征风电机组亚健康状态的良好指标,其评估的准确性直接影响了后期人员、设备、资金等多种资源的投入的多少,以及运维方案的最终效果。为了尽可能真实客观反映风机高温降容状态,提出了一种基于随机森林和长短时记忆网络自编码(Long Short Term Memory-Autoencoder,LSTM-Aec)算法相结合的智能评估方法,该方法首先采用随机森林算法对SCADA系统采集的数据进行特征约简,再利用LSTM-Aec算法对风机高温降容状态进行评估检测。测试结果表明,基于该方法的风机高温降容状态评估的精确率和准确率分别达0.9809和0.9558,整体优于未经过特征约简的RNN-Aec算法和未经过特征约简的LSTM-Aec算法的评估检测方法以及传统分类算法。展开更多
文摘High temperature superconductor (HTS) technology enables a significant reduction in the size and weight of MW-class generators for direct-drive wind turbine systems and reduce the cost of clean energy relative to conventional copper an permanent-magnet-based generators and gearbox. Using MAXWELL, we studied MW class superconducting synchronous machines. By comparison the weight, we concluded that HTS wind turbine with rotor iron is the heaviest and HTS wind turbine without rotor iron and stator teeth is the lightest. By comparison the flux density, HTS wind turbine without rotor iron is the least and HTS wind turbine without rotor iron and stator teeth is the largest. By comparison the cost, HTS wind turbine with rotor iron is the highest and the other two is almost the same. HTS wind turbine without rotor iron and stator teeth is the best type.
文摘风电机组在齿轮箱油温过高时会导致机组限功率运行,影响机组发电效率。传统应对风机高温降容状态多采用阈值判断,反应迟缓,加剧风机齿轮箱劣化趋势。利用贝叶斯网络对风机高温降容状态进行评估,为提取并准确合理地利用机组数据采集与监视控制系统(supervisory control and data acquisition system,SCADA)各个相关状态参数之间的耦合特性,通过vine-Copula模型对机组各个状态参数进行相关性分析,建立更符合机组实际运行状态的贝叶斯概率图形网络,实现对机组高温降容状态的评估。通过交叉熵算法对模型输出结果进行评价,发现与朴素贝叶斯模型相比,vine-Copula贝叶斯网络评估结果更为精确可靠,所建模型更符合机组实际运行工况,能够为现场的运维人员制定准确合理的运行和维护方案提供参考。
文摘发电机、齿轮箱等风机部件的高温降容状态是表征风电机组亚健康状态的良好指标,其评估的准确性直接影响了后期人员、设备、资金等多种资源的投入的多少,以及运维方案的最终效果。为了尽可能真实客观反映风机高温降容状态,提出了一种基于随机森林和长短时记忆网络自编码(Long Short Term Memory-Autoencoder,LSTM-Aec)算法相结合的智能评估方法,该方法首先采用随机森林算法对SCADA系统采集的数据进行特征约简,再利用LSTM-Aec算法对风机高温降容状态进行评估检测。测试结果表明,基于该方法的风机高温降容状态评估的精确率和准确率分别达0.9809和0.9558,整体优于未经过特征约简的RNN-Aec算法和未经过特征约简的LSTM-Aec算法的评估检测方法以及传统分类算法。