A Bad Data Identification Method for State Estimation Based on BP Neural Network

Abstract

The present invention proposes a state estimation bad data identification method based on BP neural network. Aiming at the higher requirement of training samples for the BP neural network bad data identification method, a BP neural network model for bad data identification based on the state estimation result is established. Take the online state estimation calculation result section as the sample for training, take the measurement value as the input data, and the state estimation value as the expected output. Through the back-propagation of the error between the input and the output, the connection weights and The threshold is corrected, a neural network based on measurement is trained, and the new measurement section is detected by the trained neural network. When the deviation between the obtained measurement and the predicted value is large, it can be judged as bad data. Because the method directly uses the state estimation calculation result as a sample for training, it provides a sample with high accuracy, thereby improving the recognition accuracy of the bad data by the neural network method.

Description

technical field [0001] The invention belongs to the field of power system operation and automation, and in particular relates to a state estimation bad data identification method based on a BP neural network. Background technique [0002] The operation data of the power grid mainly includes the topology structure of the power grid, model parameters and measurement and collection data. Accurate and reliable power grid operation data is the basis for the intelligence of the dispatching process. The measurement data system of the power system inevitably has errors and errors due to the collection and forwarding of multiple links. The measurement errors cannot be eliminated due to the conditions of the measurement data collection system. Technologies such as improving data redundancy can be processed, and wrong measurement data mainly refers to data that deviates far from the actual measurement data change trajectory, also known as bad data. Certain deviations cannot be handled...

AI Technical Summary

Problems solved by technology

The disadvantage of the neural network method is that it is highly dependent on the training process of the network. The selection and representativeness of the training samples will directly affect the final identification effect.

Images