变电站集控智能告警系统分析-analysis of substation centralized control intelligent alarm system.docxVIP

贵州大学硕士学位论文Ab 贵州大学硕士学位论文 Abstract - -II- Abstract A large number of alarm information will occur when there are some fault arising in substation, and for some reason, such as signal transmission channel subject to interference and so on, it is impossible to omit some important information. In addition, some are unreliable in these alarm information. Under the circumstances, it is difficult for operator to seize the alarm substance and determine equipment failure in a very short period of time. Therefore, it is a significant task for the operator to rapidly handle failure and ensure safe and stable operation of power grids, which a large number of alarm information is automatic processed and equipment fault is diagnosed with artificial intelligence methods. In this paper, the work done is on this background. Firstly, the paper gives a brief introduction of neural network and makes analytical methods with RBF neural networks. It analysis the structure、the theoretical basis and its learning algorithm of the RBF neural network, and RBF neural network fault tolerance performance have been studied. Secondly, this paper presented a practical method for enhancing fault tolerance performance of RBF neural network. Thirdly, this paper used relational knowledge representation of the substation that based on relational database before the establishment of the network model. Then, we introduce an example of Guiyang 220kV Wudang substation. The network diagnosis module is categorized by the main transformers 、 busbars and lines as the core equipment in this substation, the related protection information and the related switching information as the input of the network, and fault type as output which corresponding to the device. Finally, the effect of the system are analyzed and examined by the example, which the practical way of substation intelligent alarm handing is used based on reconstructed RBF neural network. Key words: RBF neural network ； Substati