摘要
在风机齿轮箱故障诊断过程中,针对由于故障数据稀疏导致模型建立困难的问题,提出一种使用改进人工蜂群算法(IABC)优化Elman神经网络的故障诊断模型。该模型通过建立齿轮箱正常状态下的温度模型,采用残差分析,得到齿轮箱的故障状态,降低了建立模型的复杂度。采用IABC对Elman神经网络的相关参数进行优化,解决了Elman网络收敛速率慢、易陷入局部最优的问题。在IABC算法中,观察蜂阶段采用动态搜索策略,实现搜索能力和开发能力的平衡;在侦查蜂阶段,通过引入混沌变量扰动,增大种群多样性,进而达到全局最优。通过华北某风电场历史数据进行实验,结果表明,IABC与Elman神经网络的结合可对风机齿轮箱故障状态进行识别,且诊断正确率较高,可应用于实际故障诊断。
In the process of fault diagnosis of wind turbine gearbox,aiming at the difficulty of model establishment due to sparse data of faults,this paper proposes a fault diagnosis model using improved artificial bee colony algorithm (IABC) to optimize Elman neural network. The model is to analyze the residual error through the establishment of the temperature model under the normal state of the gearbox to obtain the fault state of the gearbox, which reduces the complexity of building the model. In order to solve the problem of slow convergence speed and easy to fall into the local optimal in Elman network, IABC is used to optimize the relevant parameters of Elman neural network. In the IABC algorithm, the observational bee phase adopts a dynamic search strategy to achieve a balance between search capabilities and development capabilities. In the detection bee phase, chaotic variable disturbances are introduced to increase the diversity of the population and thus to achieve global optimization. The historical data of the actual wind field are selected to carry out the experiment. The results show that the combination of IABC and Elman neural network can identify the fault status of the wind turbine gearbox which can be applied to actual fault diagnosis, and it has high diagnostic accuracy.
作者
林涛
杨欣
蔡睿琪
张丽
刘刚
廖文喆
Lin Tao;Yang Xin;Cai Ruiqi;Zhang Li;Liu Gang;Liao Wenzhe(Institute of Artificial Intelligence and Data Science,Hebei University of Technology,Tianjin 300130,China)
出处
《可再生能源》
CAS
北大核心
2019年第4期612-617,共6页
Renewable Energy Resources
基金
河北省科技计划项目(17214304D)
关键词
风机齿轮箱
故障诊断
ELMAN神经网络
人工蜂群算法
the gearbox of wind turbine
fault diagnosis
Elman neural network
artificial bee colony algorithm
