基于多源征兆和并网运行状态的大型风电机组早期故障诊断研究

The fault diagnosis of wind turbine generator system (WTGS) is the guarantee of WTGS running in health state. However, the early fault diagnosis becomes very difficult and the fault evolution trend is not clear while the WTGS is on grid. Based on our early study, it was dicovered that the reason is relating to the fact that only one fault sign is used when study the fault diagnosis and the WTGS on grid running state is not considered when analyse the fault evolution trend.In this project, we plan to study the imitatation of the faults such as the main shaft deformation, bearing wear, magnetic crack distortion,windings (and the inside wiring) insulation effectiveness loss, magnet steel dropping of WTGS generator and the detection method of the corresponding multiple source fault signs such as the current,vibration and temprature. Test and simulate under the WTGS running state of grid short circuit fault, unbalanced three phase voltage and low voltage ride-through; and then catch the relationship between the WTGS early stage fault evolution mechnism under the influence of different on grid WTGS running state and the mutiple fault signs.Define the fault sign factor (FSF) and the WTGS running state influence factor (RSIF). Draw the characteristic parameters and determine the fault threshold, and build the synthetical model of early stage fault diagnosis based on the FSF and RSIF using the information blend technology. Detect and collect the spot running effective data of the WTGS in the wind farm, and using the data to evaluate the accuracy of the fault diagnosis result and study the early stage fault evolution trend and scientific maintain strategy of WTGS. Obtain the theoretical system of the early stage fault diagnosis and fault trend analysis. The study results will provide theoretical support for the on line health evaluation of WTGS and provide effective guarantee method for the safe and economic production of wind power of wind farm.

风电机组故障诊断是机组健康运行的保障，但并网运行时存在早期故障诊断困难和故障演化趋势不明确问题。基于我们前期研究发现其原因与使用的故障征兆信息单一、且没考虑机组运行状态有关。本项目计划研究风电机组电机的主轴变形、轴承磨损和磁隙畸变、绕组（及内部接线）绝缘失效、磁钢脱落故障的模拟及其与电流、振动和温度多源故障征兆的检测方法，在电网短路、三相不平衡和机组低压穿越状态下试验和仿真，扑捉机组在不同工况影响下早期故障演化机理及其与多源征兆的关系，定义征兆表现因子和工况影响因子；提取特征参数，确定故障阈值，并根据征兆表现因子和工况影响因子利用信息融合技术建立早期故障综合诊断模型；检测和收集风电场机组运行有效数据，评估故障诊断准确度，研究早期故障演化趋势和机组科学维修策略；最终得到风电机组早期故障诊断和故障趋势分析理论体系。研究成果将为风电机组在线健康评估提供理论依据，为风电场安全经济生产提供有效方法。

我国目前风电行业已跨入世界竞争行列；风电机组的运营环境恶劣和来自电网的干扰冲击会导致机组设备事故，国家对风电机组各阶段健康跟踪和后期维护也提出了更高的要求，而早期故障诊断是完成这些要求的重要技术基础。.主要研究内容：研究风电机组电机主轴变形、轴承磨损和磁隙畸变、绕组绝缘失效、磁钢脱落故障的模拟及其与电流、振动和温度多源故障征兆的检测方法，在电网短路、三相不平衡和机组低压穿越状态下的试验和仿真，捕捉机组在不同工况影响下早期故障演化机理及其与多源征兆的关系，定义征兆表现因子和工况影响因子；提取特征参数，确定故障阈值，并根据征兆表现因子和工况影响因子利用信息融合技术建立早期故障综合诊断模型；检测和收集风电场机组运行有效数据，评估故障诊断准确度，研究早期故障演化趋势和机组科学维修策略；得到风电机组早期故障诊断和故障趋势分析理论体系。.取得的重要研究成果：完成了对风电机组的发电机早期故障及故障演化的模拟实验与仿真分析；定义并对征兆表现因子和工况影响因子进行了量化计算，找出了不同工况对故障演化过程的影响；建立了机组电机轴承和主轴的非线性疲劳载荷损伤累积模型，找出了早期故障到复杂故障的演化关键点；完成了从征兆信息中提取特征参数并确定了自适应故障阈值；建立了相关向量机子故障诊断模型和利用信息融合技术的基于多源故障征兆信息并考虑不同运行工况影响的在线的早期故障及其演化趋势的综合诊断模型，并得到了验证。.已发表学术期刊33篇，其中SCI 期刊1篇，EI中文期刊2篇，核心期刊22篇；基金第一标注20篇，第二标注7篇，还有5篇录用论文；发表会议论文7篇，EI收录3篇，第一标注4篇，第三标注2篇；专利5项，已授权2项，已受理3项；软件著作权3项；培养研究生23名；研制故障检测与诊断系统一套，获得省级科技进步奖（二等）一项。.研究成果对风电机组健康运行、电力可靠生产具有重要理论意义；为现场维修人员排除故障提供了可靠依据。