NPC three-level inverter multi-class fault diagnosis method based on reduced-order observer

Abstract

The invention discloses an NPC three-level inverter multi-class fault diagnosis method based on a reduced-order observer, and belongs to the technical field of fault diagnosis. The method comprises the following steps of establishing a hybrid logic dynamic model, performing decoupling by utilizing matrix transformation, decomposing a sensor fault and a system state, establishing a reduced-order observer, defining a current form factor and a fault diagnosis adaptive threshold value, performing fault diagnosis, and then performing fault positioning. According to the method, the sensor fault and the system state are decomposed by adopting matrix transformation, so that the inverter switching tube fault and the sensor fault can be diagnosed at the same time; according to the method, the accuracy and robustness of fault diagnosis are improved by adopting the current form factor and the self-adaptive threshold value; according to the method, a switch tube fault detection characteristic value, a sensor fault detection characteristic value and a fault positioning characteristic value are defined according to characteristics of three-phase output current and a sensor fault estimation value, and multi-class fault diagnosis is carried out on the NPC three-level inverter.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis, in particular to a multi-type fault diagnosis method for an NPC three-level inverter based on a reduced-order observer. Background technique [0002] In the photovoltaic power generation system, the inverter plays an important role. NPC three-level inverters have many advantages such as low loss, low harmonic content of output voltage and current waveforms, small device stress, and large output capacity, and are widely used in photovoltaic microgrids. Although the NPC three-level inverter has many of the above advantages, due to the complex and changeable environment of the photovoltaic power generation system, the power devices and sensors in the inverter are prone to failure, which seriously threatens the safety of the photovoltaic power generation system. The complex structure of the flat inverter has too many power electronic devices, which leads to an increase in the failure rate du...

AI Technical Summary

Problems solved by technology

The main idea of ​​this type of method is to establish a mathematical model of the inverter, compare the estimated system output with the measurement information to obtain the residual, and analyze the residual to realize the fault diagnosis of the converter. Related patent documents such as "Based on interval sliding mode Inverter Open Circuit Fault Diagnosis Method Based on Observer", "A Model-Based Parametric Fault Diagnosis Method for Multilevel Inverter" (Application Publication No. CN108649600A), etc. These methods have high mathematical model requirements and poor robustness The problem

[0008] In summary, the existing technology has problems such as high signal processing complexity, long diagnostic cycle, large amount of calculation, difficulty in establishing and maintaining the knowledge base, high requirements for mathematical models, and poor robustness.

Images