Research Article
Research on the Condition Monitoring of Transmission and Transformation Equipment Based on Improved Support Vector Machine in the Internet of Things
@INPROCEEDINGS{10.1007/978-3-030-00410-1_35, author={Chao Fu and Qing Lv and Chong Li and Yun Feng and Xiao-li Li}, title={Research on the Condition Monitoring of Transmission and Transformation Equipment Based on Improved Support Vector Machine in the Internet of Things}, proceedings={IoT as a Service. Third International Conference, IoTaaS 2017, Taichung, Taiwan, September 20--22, 2017, Proceedings}, proceedings_a={IOTAAS}, year={2018}, month={10}, keywords={Internet of Things Power transmission and transformation equipment On-line monitoring Support vector machines (SVM)}, doi={10.1007/978-3-030-00410-1_35} }- Chao Fu
Qing Lv
Chong Li
Yun Feng
Xiao-li Li
Year: 2018
Research on the Condition Monitoring of Transmission and Transformation Equipment Based on Improved Support Vector Machine in the Internet of Things
IOTAAS
Springer
DOI: 10.1007/978-3-030-00410-1_35
Abstract
The realization of smart grid is based on the real-time command of important operation parameters of power transmission and transformation equipment. The Internet of things has powerful capabilities of information collection and interactive, which can be used as the support for the condition monitoring of transmission and transformation equipment in the smart grid environment. This paper takes the power grid equipment as the center, takes intelligent on-line monitoring of equipment as the direction, starting from the equipment condition assessment and fault types, carried out the research about of grid equipment real-time state monitoring and fault diagnosis under the environment of Internet of things. Paper mainly includes: constructing the status evaluation framework and real-time evaluation model of power grid equipment from the angle of the status value of on-line monitoring of IoT, using support vector machine (SVM) for power transmission and transformation equipment condition monitoring, choosing a suitable kernel function by comparing the linear kernel function, polynomial kernel function, the radial basis kernel function and multilayer perceptron kernel function of multiple parameters; by analyzing the traditional cross-validation method, this paper proposed the improved cross validation (K-CV) method, and we use Actual data of power grid field as the sample, finally obtain the fault classification result by constant parameter optimization. The experimental result shows that the support vector machine based on improved 10-CV cross-validation method in the Internet of things is able to monitor the condition of transmission and transformation equipment more rapidly and accurately.
