基于动态贝叶斯网络的无人机巡检输电线路故障诊断研究

发布时间：2018-03-16 14:13

本文选题：输电线路　切入点：无人机巡检　出处：《华北电力大学(北京)》2016年硕士论文　论文类型：学位论文

【摘要】：当今社会,人们对于电力能源的需求越来越大,做好保证电力资源安全运行的工作尤为重要。输电线路是传输电力能源的主要途径,当输电线路因为部件故障而使电力资源不能正常的传输时,必然会造成巨大的经济损失。因此输电线路安全可靠的运行是保证电力资源稳定传输的关键所在,对输电线路及各个部件的故障检测及诊断工作就显得极为重要。本文所做的研究工作主要是针对无人机巡检输电线路时,对各个部件巡检采集数据进行故障类型的分类、诊断以及对潜在故障的预测。首先,利用故障模式影响分析技术对输电线路系统进行定义,并对输电线路各个部件的故障模式、运行状态和巡检方式进行了详细介绍,为后面进行输电线路的故障诊断和预测提供了理论基础；其次,研究了小样本数据下的输电线路故障分类诊断方法,采用三种不同的算法对分类器的模型参数进行优化选择,并通过改进粒子群算法,解决了分类器参数的寻优问题,然后通过仿真实验验证了改进粒子群算法的优越性；再次,提出了基于数据扩展的动态贝叶斯网络预测方法,通过动态贝叶斯网络与隐马尔科夫模型的转换建立预测模型,并对预测模型进行修改,使预测模型能够在数据量增加、节点间的相互关联变得复杂之后,降低预测模型中因节点间的相互关联造成的预测误差积累的问题,提高模型的预测精度：最后,建立了输电线路故障诊断系统,通过获取的数据对输电线路的运行状态进行评估,并且系统还会对输电线路的潜在故障进行预测,得出整体输电线路的薄弱环节,使巡检维修工作更有针对性。
[Abstract]:In today's society, the demand for power energy is increasing, so it is particularly important to ensure the safe operation of power resources. Transmission line is the main way to transmit power energy. When the transmission line is unable to transmit the power resources normally due to the fault of the components, it will inevitably cause huge economic losses. Therefore, the safe and reliable operation of the transmission line is the key to ensure the stable transmission of the power resources. It is very important to detect and diagnose the faults of transmission lines and components. The research work in this paper is mainly aimed at classifying the fault types of the data collected from the patrol inspection of transmission lines by UAV. First of all, the fault mode influence analysis technology is used to define the transmission line system, and the fault mode, operation state and inspection mode of each part of the transmission line are introduced in detail. It provides a theoretical basis for fault diagnosis and prediction of transmission lines. Secondly, the method of fault classification and diagnosis for transmission lines with small sample data is studied, and three different algorithms are used to optimize the model parameters of the classifier. By improving the particle swarm optimization algorithm, the optimization problem of classifier parameters is solved, and then the superiority of the improved particle swarm optimization algorithm is verified by simulation experiments. Thirdly, a dynamic Bayesian network prediction method based on data expansion is proposed. Through the transformation of dynamic Bayesian network and hidden Markov model, the prediction model is established, and the prediction model is modified to make the prediction model become more complicated after the amount of data increases and the interrelation between nodes becomes complex. In order to reduce the accumulation of prediction errors caused by the correlation between nodes in the prediction model, the prediction accuracy of the model is improved. Finally, the transmission line fault diagnosis system is established. Through the obtained data to evaluate the operation state of the transmission line, and the system will also predict the potential fault of the transmission line, draw the weak link of the whole transmission line, make the inspection and maintenance work more targeted.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM75

【相似文献】