Case analysis of fault diagnosis results of line protection in secondary system of smart substation

作者： GOZ Electric 时间：2024-06-19 09:38:06 阅读：15

    The line protection faults of the secondary system of a smart substation are mainly divided into protection devices 1/2 and measurement and control network links. The input signals mainly include the merging unit and the line protection receiving intelligent terminal GOOSE disconnection. The PSO algorithm is applied to the SVM model, and the optimal γ value is accurately determined to be 1.9282, and the optimal C value is determined to be 0.01. The fault information processing model after optimization proves that its accuracy rate actually meets 100% by the ten-fold cross-validation of the training samples. With the help of the trained PSO-SVM model, multiple test samples are given fault diagnosis, and the judgment results of professional maintenance personnel are used as the fault determination standard. The fault diagnosis results show that the diagnosis model has an accuracy rate of up to 97%. If there is no PSO optimization method to assist, the accuracy of the model is 94%, which proves that PSO has played an important role in the optimization of the SVM model.

    Analysis of the diagnosis results of transformer protection faults in the secondary system of smart substations The transformer protection faults of the secondary system of smart transformers mainly include merging units and intelligent terminals. The input unit mainly includes the merged unit SVM overall warning and GOOSE overall warning. The PSO algorithm is applied to the SVM model, and the optimal γ value is accurately determined to be 1.4497, and the optimal C value is determined to be 0.01. After the optimization, the fault information processing model is proved to have an accuracy of 100% by the ten-fold cross validation of the training samples. The PSO-SVM model is applied to the transformer protection fault of the secondary system of the smart substation. The detection result has a very high diagnostic accuracy of 98.1%, which proves that the model is extremely correct, reliable and applicable in the fault diagnosis of the secondary system of the smart substation.

    This study optimizes the research based on the shortcomings of the current smart substation secondary system fault detection, analyzes and explains the smart substation secondary system fault detection technology based on SVM in detail, creates a fault diagnosis data set, and a diagnosis model. The conclusions include the following points.

(1) This paper uses a more typical actual case to analyze, and at the same time constructs a data set corresponding to the status of all faults in the secondary system of the smart substation. By analyzing the diagnosis results, it is known that the data set has better training properties.

(2) This paper also uses the efficient small sample regression performance of the SVM algorithm to create a diagnostic model that meets the fault diagnosis needs of the secondary system of the smart substation. At the same time, the PSO algorithm is used to determine the optimal value of the SVM data information (i.e., optimization), which effectively prevents the deviation of the selection of SVM data information and further improves the original advantages of the SVM fault diagnosis model.

(3) After analyzing the diagnostic results, it can be seen that the diagnostic model described in this study has shown significant accuracy in the fault diagnosis of the line protection and transformer protection of the secondary system of the smart substation, which proves that the diagnostic technology is of great application value.