Faults detection in PMSM drive using Artificial Neural Network

Abstract

In this paper, simulation research results of PMSM drive with open phase fault detection are presented. Proposed fault detection system is implemented using two artificial neural networks. One of them is neural model of healthy PMSM and another one generates diagnostic signals. When the fault occurs, the amplitude of current residuals increases and evaluation system returns diagnosis. In proposed system detection time is about 1 ms. Moreover, diagnosis does not depend on load state. Streszczenie. Artykuł przedstawia wyniki badań symulacyjnych napędu PMSM z detekcją przerwy fazy. Proponowany system detekcji uszkodzeń zaimplementowano z użyciem dwóch sztucznych sieci neuronowych. Jedna z nich pełni rolę modelu neuronowego sprawnego PMSM, natomiast druga generuje sygnały diagnostyczne. W przypadku wystąpienia uszkodzenia amplituda residuów prądów wzrasta, a system ewaluacji zwraca diagnozę. Czas detekcji w przedstawionym układzie jest rzędu 1 ms. Ponadto działanie systemu nie zależy od stanu obciążenia (Detekcja uszkodzeń w napędzie z PMSM przy użyciu Sztucznej Sieci Neuronowej).

Fault Diagnosis and Fault-Tolerant Control of PMSM Drives–State of the Art and Future Challenges

“…The analysis of the literature presented in Table VIII shows that the largest number of works related to the design of neural damage detectors of PMSMs concerns the application of the MLP structure. The MLP network is characterized by an extremely simple mathematical description, thanks to which it is used in the case of mechanical damages: bearing damage [121,231], eccentricity [231], and in the case of the damage to the stator electrical circuits [215][216][217][218], [224], [225], [232], supply voltage unbalance [219], [221], [227] or stator phase loss [221], [226], and demagnetization faults [232].…”

Position Estimation at Zero Speed for PMSMs Using Artificial Neural Networks

“…To use ANNs to process unknown functions, a learning process using learning sets is required. ANNs are used in electrical drives to diagnose faults, e.g., using residual evaluation system and the object model [35], or using highly processed network input signals [36], to increase efficiency in the drive, e.g., using ANN as current controllers [37], in order to optimal parameters design [38], or using ANN as speed controller [39]. A probabilistic neural network can be used for estimation the sine and cosine of the shaft position [40], or a polynomial neural network for inductance estimation [41].…”