随着汽车智能电网和智能交通的逐步合并，其安全性和稳定性逐渐受到关注，这

With the gradual merger of automotive smart grids and smart transportation, their safety and stability have gradually received attention, which poses challenges to the maintenance and guarantee of existing power equipment. Therefore, this paper proposes a multi-feature fusion electric vehicle smart grid OLTC fault feature selection. This method is mainly based on the feature selection method of multi-feature fusion and improved gray wolf learning algorithm optimization kernel extreme learning machine. First of all, for the defects of traditional OLTC with single fault features and low diagnostic accuracy, this paper will use multi-scale arrangement entropy and energy correlation coefficients to perform feature fusion to construct high-dimensional features; then use kernel principal component analysis to filter the initial matrix to form the initial sub Set; Finally, IGWO is used to optimize the nuclear extreme learning machine while performing feature selection. The test results prove the effectiveness of this method, which can accelerate the merger of smart grid and intelligent transportation of automobiles.

With the gradual merger of automotive smart grid and intelligent transportation, the safety and stability of automobile sitcoms have been paid attention to, which poses a challenge to the maintenance and security of existing power equipment. Therefore, this paper presents a multi-feature fusion of electric vehicle smart grid OLTC fault characteristics selection. This method mainly relies on multi-feature fusion and improved gray wolf learning algorithm to optimize the characteristic selection method of the nuclear limit learning machine. Firstly, in view of the defects of the traditional OLTC fault characteristics which are relatively single and the diagnostic accuracy is low, this paper combines the multi-scale arrangement entropy and energy correlation coefficients to construct the high-dimensional features, then uses the analysis of the nuclear main component to screen the initial matrix to form the initial subset, and finally uses IGWO to optimize the nuclear limit learning machine while selecting the characteristics. The test results prove the effectiveness of this method and can accelerate the integration of smart grid and intelligent transportation.

With the gradual integration of smart grid and intelligent transportation, the security and stability of smart grid are gradually concerned, which challenges the maintenance and guarantee of existing power equipment. Therefore, this paper proposes a multi feature fusion feature selection for OLTC fault of EV smart grid. This method mainly relies on multi feature fusion and improved gray wolf learning algorithm to optimize the feature selection method of kernel limit learning machine. Firstly, aiming at the defects of single fault feature and low diagnosis accuracy of traditional OLTC, this paper combines multi-scale arrangement entropy and energy correlation coefficient to build high-dimensional features; then, the initial matrix is screened by kernel principal component analysis to form an initial subset; finally, igwo is used to optimize the kernel limit learning machine and select features at the same time. The experimental results show that the method is effective and can speed up the combination of smart grid and intelligent transportation.<br>