Voltage source inverters (VSI) with modern power-switching elements are often used to control industrial AC motors. However, the non-linearities of the inverters, such as dead time, turn-on and turn-off switching delay times and voltage drops, are often behind the distortion of the phase currents of the controlled motor. The current distortions can be suppressed by appropriately calculated non-linear functions, which represent the compensation voltages and are consequently added to the control values of the current regulators in the field-oriented control (FOC) algorithm. An artificial neural network (ANN) was designed to identify the non-linear functions of the compensation voltages, which is presented in this paper. Only signals available in the FOC algorithm are used as ANN inputs. The learning process of the neural network takes place online during the running of the motor control algorithm. The learning pattern is generated in each step of the control algorithm from the control errors of the current controllers and the previous ANN outputs. It is not necessary to know the VSI parameters when learning the neural network. The proposed ANN and back-propagation learning algorithm were implemented on one core of the AURIX microcontroller TC397. The proposed strategy was validated through experiments on a real permanent magnet synchronous motor (PMSM), and experimental results prove the effectiveness of the ANN.

The challenges of fault detection and condition monitoring in powertrain systems have become increasingly prominent, particularly with the widespread adoption of fail-operational systems. These systems are pivotal in diverse sectors, including the robotics, automotive industry, and various industrial applications. A critical attribute of such systems lies in their capability to identify non-standard behaviour of the system. This study describes a inovative conditional convolutional autoencoder-based fault detection algorithm for the permanent magnet synchronous motor. The study compares a train process of conditional convolutional autoencoder with a classical convolutional autoencoder. The presented autoencoder structure was designed to be implementable into the target microcontroller AURIX TC397 while providing sufficient recognition capabilities of the interturn short-circuit. Autoencoders are trained on data obtained during healthy motor operation and subsequently used to detect interturn short-circuit faults on the experimental dual three-phase permanent magnet synchronous motor with the possibility of emulating an interturn short-circuit fault. The paper provides insights into the achieved autoencoder inference times and the sensitivity in detecting the fault.

The paper aims to solve the distortion problem of the inverter output voltages that cause harmonic deformation of the phase currents and ripple of dq - currents of the three-phase permanent magnet synchronous motor (PMSM). The inverter non-linearities adversely affect the effectiveness of the PMSM control algorithm. The compensation strategy is based on the neural network and knowledge of the three-phase PMSM model structure and its parameters. The input data for the neural network consist of the normed values and detected polarities of the phase currents and rotor position information. As a result, the proposed artificial neural network (ANN) can extract non-linear functions from the measured data to compensate for the VSI output voltages. The ANN is designed to learn online while the PMSM is running. The back-propagation algorithm is used for neural network learning. The proposed strategy was implemented in an AURIX TC397 microcontroller and validated by experiments on a real PMSM. The presented results demonstrate the effectiveness of the proposed solution.

The problem of condition monitoring and fault detection in powertrain systems becomes more critical with the increasing use of fail-operational systems. These systems are essential in the automotive industry, robotics, and other industrial applications. One of the critical features of such a system is recognizing the fault and suppressing its influence. The paper describes a feed-forward artificial neural network-based diagnostic of interturn short-circuit faults in a dual three-phase permanent magnet synchronous motor. The paper focuses on using multi-layer perceptron network (MLPN) and convolutional neural network (CNN) for interturn short-circuit detection and, more importantly, their real implementation into the automotive AURIX TC397 microcontroller. The paper presents the achieved neural network inference times as well as data preprocessing computation time. The behavior of the artificial neural networks (ANNs) is tested on an experimental configurable multi-phase permanent magnet synchronous motor (PMSM) with the possibility to emulate interturn short-circuit fault using prepared winding taps. The paper includes the essential aspects that should be respected during ANN design and implementation into the microcontroller.

In this study, model of a multi-phase permanent magnet synchronous motor is demonstrated. The model is designed to allow simulations of inter-turn short circuit faults of various severity. The considered multi-phase motor is connected as a dual three-phase system. In this case, stator windings are divided into two electrically separated sub-systems. Nevertheless, the magnetic coupling between sub-systems needs to be taken into account in general. The motor model is designed to take into account the exact stator coils distribution. In this case motor uses combination of serial and parallel connections. Model is implemented in MATLAB/Simulink using SimScape environment. The paper also shows a comparison of the behavior of a real PMS motor and derived model behavior.

In this paper, a new compensation strategy design of the dead-time effects and other VSI non-linearities is presented for a dual three-phase PMSM. The strategy is based on an estimation of compensation voltages for each motor sub-system by the proposed disturbance observer. This observer is based on the Kalman filter algorithm and the knowledge of the motor model structure and its parameters. The proposed approach is verified by experiments on the test bench which consists of the experimental dual three-phase PMSM and set of three-phase power stages controlled by one Tricore AURIX TC275 microcontroller. The achieved results are compared with the results of the standard compensation strategy.

Permanent magnet synchronous machines (PMSM) are widely used for their high efficiency and power density. High machine reliability is increasingly required nowadays. PMSMs can't be simply disconnected from the power source unlike asynchronous machines if any fault appears in the system. Many motor faults lead to malfunction of the whole system. This limitation can be solved by using multi-phase structures instead of commonly used three-phase structures. Multi-phase machines have many advantages in terms of high reliability. The power density of properly designed multi-phase system is also higher. The multi-phase PMSM can operate even during a fault, under certain conditions, depending on the motor parameters. The control algorithm must be capable to detect the fault and apply proper control method according to the detected fault.This paper demonstrates the behaviour of the dual three-phase PMSM motors under various inter-turn short circuit faults. An experimental PMSM was prepared, having stator windings with multiple taps. These winding taps are utilised to emulate inter-turn short circuit faults. The winding short-circuiting is realized by a solid-state relay. The fault influence is analysed on the experimental machine.Three compensation strategies are applied to reduce the fault influence. All compensation strategies are based on field weakening of the damaged motor part. The motor behaviour during the fault without compensation method is compared with the behaviour when using compensation strategies. Realised experiments demonstrate that a properly constructed dual three-phase PMSM under control with compensation strategies can continuously operate under the fault condition.

Importance of the condition monitoring and predictive maintenance in motion systems is growing up as motion systems quantum and their complexity (number of axes, performance parameters) increases with increasing the automation of huge range of human activities and manufacturing processes. Probability of failures increases with the system complexity. Many faults and indication of their propagation in the electric drives would require additional sensors or hardware, higher bandwidth and sampling frequencies of feedback sensors, high computing power etc. for development of sophisticated methods to detect specific faults with good sensitivity, robustness and reliability under any operating condition. This paper presents an approach to the condition monitoring and prognosis applicable into the existing systems. These methods use the information available in the traditional electric drives - especially the information from the individual sensors in a voltage source inverter (VSI) and/or an electric motor. Condition indicators for these methods are based on application specific operating states or actions, which generates typical patterns in the signals. The condition monitoring is based on observing the deviations of these patterns between the healthy system and the system with fault propagating. The implementation strategy is described in the paper and some demonstration examples are shown as well.

In this paper, two novel approaches to compensate VSI nonlinearities and dead-time effects are presented. The voltage disturbance observers are based on the Kalman filter algorithm. Their design is focused on suppressing the effect of the quantization noise and other speed measurement inaccuracies on the estimates of the disturbance voltage. In the one case, the rotor velocity is estimated by angle tracking observer and enters to the proposed observer instead of the measured speed. In the latter case, the observer is extended by another state. As a result, the observer is able to estimate the disturbance voltages and rotor velocity. The efficiency of the proposed approaches is verified by experiments on the real vector-controlled PMSM under different operating conditions.

In this paper, a novel adaptive compensation strategy to reduce dead-time effects and nonlinearity of the VSI is presented. These undesirable properties of the inverter are necessary to compensate because they cause 5th and 7th harmonic distortion of the phase current, torque pulsation and generally reduce the effectiveness of control algorithm. The observer which estimates the dq-currents and the value of the voltage error by only one parameter is designed based on the harmonic analysis and algorithm of the Extended Kalman filter. Subsequently, the compensating voltages are determined from estimated voltage error and the polarity of the estimated currents. The results of simulations and real experiments demonstrate the effectiveness of the proposed approach.

This paper is focused on stator winding fault detection of a permanent magnet synchronous motor (PMSM). Open phase fault and interturn fault in one phase are considered. Differences between stator resistances are considered as a fault symptom. The estimation of stator resistances is based on Extended Kalman Filter (EKF) with three phase PMSM electrical model. The fault detection method was verified offline on measured experimental data from the PMSM test platform with emulated faults.

Model predictive control is a modern method which was used in many different applications. One of the applications are electrical machines. Designing a controller for a synchronous drive has proven itself to be a challenging task. Many authors try to substitute one or more controllers in the cascade structure. Nevertheless, this approach calculates only the suboptimal trajectory, not the optimal trajectory. This paper presents a design of a nonlinear model predictive controller as one multiple-input multiple-output (MIMO) controller considering the system as one single system. The simulation results are shown and discussed in detail - especially the process of currents and their dynamics.