Fault diagnosis method for ANPC three-level inverter

Abstract

The invention relates to a fault diagnosis method for an ANPC three-level inverter. The fault diagnosis method comprises the following steps: constructing an ANPC three-level inverter circuit simulation model, and acquiring upper, middle and lower bridge arm voltage signals of each phase through a voltage sensor; extracting a wavelet packet energy spectrum as a signal feature, performing dimension reduction on the extracted data to obtain a final signal feature vector, and dividing the final signal feature vector into training data and test data; establishing a joint optimization echo state network JO-ESN, optimizing parameters of an echo state network ESN reserve pool, and training the joint optimization echo state network JO-ESN by using the training data to obtain a trained JO-ESN model; and testing the JO-ESN model by using the test data, and analyzing a diagnosis result. According to the invention, the wavelet packet is used to replace the traditional wavelet and Fourier transform to extract the fault signal, the integrity of the signal is effectively retained, and the LDA is used to reduce the dimension after the fault signal is extracted, so that the fault diagnosis time can be effectively reduced.

Description

technical field [0001] The invention relates to the technical field of ANPC inverter fault analysis, in particular to a fault diagnosis method for an ANPC three-level inverter. Background technique [0002] With the rapid development of power electronics technology in my country's new energy power generation, high-speed electrified railway and other fields, multi-level inverters have been widely used. Neutral point clamped (NPC) inverter is one of the most mature multilevel inverters, with simple structure, suitable for driving medium-voltage motors such as fans, compressors, and conveyors. But the uneven use of semiconductor devices in NPC inverters will lead to uneven power loss in switches, which limits the rated power of NPC inverters. To solve this problem, an active neutral point clamped (ANPC) inverter is proposed, which replaces the clamping diodes with active switches to equalize the power loss between the switches. [0003] Due to the large number of switching de...

AI Technical Summary

Problems solved by technology

The echo state network (ESN) uses a reserve pool composed of randomly connected neurons to replace the hidden layer of the traditional neural network, which can reduce the amount of computation and improve the convergence speed of the network, but its reserve pool parameters are randomly set by humans, which is uncertain. , often cannot reach the optimal value

Images