Research team:

• Prof. Teresa Orłowska-Kowalska, PhD, DSc, Eng (project manager),
• Mateusz Dybkowski, PhD, DSc, Eng.
• Grzegorz Tarchała, PhD, DSc, Eng. 
• Prof. Czesław T. Kowalski, PhD, DSc, Eng.
• Marcin Wolkiewicz, MSc, Eng.
• Piotr Sobański, MSc, Eng.
• Kamil Klimkowski MSc, Eng. 

The project focused on modern electric drives with induction motors, which are integral parts of complex industrial processes and mass-produced mechatronic devices. Such subsystems face very stringent requirements regarding their dynamic and static properties, reliability and predictability of operation, and, more recently, increased safety in installations responsible for human health and safety, such as traction drives, automotive drives, electric drive systems in aviation, mining, nuclear power, etc. Such drives must have appropriate system and software safeguards against the possibility of damage to their components, including the converter, motor, and sensors. Also crucial is the ability to detect and localize faults (FDI, Fault Detection and Isolation) in their initial phase, so that the power supply and control system can be appropriately reconfigured and continue operating (FTC, Fault Tolerant Control), or – in justified cases – bring the drive of a complex device to a safe stop before an emergency shutdown occurs. Furthermore, for certain types of faults, the vector control algorithm for such a drive system can compensate for the impact of the fault by utilizing redundant information from state variable estimators or by reconfiguring the system and applying other control structures, such as scalar, speed-sensorless, or torque control. If the diagnostic system determines that the fault type poses a threat to the correct operation of the drive, the system must be stopped safely for the controlled process, and the fault is eliminated according to established procedures. The main goal of the project was to develop and test, through simulation and experimental studies, new methods for detecting, isolating, and compensating electrical and mechanical faults in drive systems with induction motors. Analyzed were faults in the frequency converter's power electronic switches, faults in the sensors measuring electrical and mechanical quantities (current, voltage, speed/position), and incipient motor winding faults (short-circuits in the stator winding and incipient cracks in the rotor bars or cage ring). The impact of selected faults on the operation of a drive controlled by vector methods (DRFOC and DTC-SVM), with angular velocity measurement, and in sensorless mode was examined. Diagnostic signals were also selected to enable not only the detection of selected faults but also the differentiation of faults causing similar symptoms. Based on these signals, fault detection and diagnostic algorithms were developed, based on algorithmic methods and artificial intelligence, implemented using software tools. Original diagnostic algorithms were developed for the detection of faults in the transistors of a two-level voltage inverter. These algorithms are characterized by very fast operation (less than one cycle of the current supplying the motor stator winding) and simple implementation. They were shown to be resistant to other types of faults, such as short circuits in the stator of an induction motor. The proposed voltage inverter circuit reconfiguration method, along with a proprietary modulation method, ensures continued proper operation of the drive system after a transistor failure in the voltage inverter bridge. In the event of a failure of the current and speed measuring sensors, the FTC system proposed the use of a state variable estimator (MRAS type) to reproduce the angular speed, with increased resistance to changes in the induction motor parameters. Algorithmic and neural detectors were developed to detect failures of the aforementioned measuring sensors in the vector control structures of the induction motor. Methods for compensating sensor failures were developed, using sensorless control, hardware and analytical redundancy, and a change in the control system topology. A complete diagnostic system was also developed to detect and compensate for failures of all analyzed sensors in a single drive cycle, enabling its continued operation. Regarding failures of the induction motor's electrical circuits, methods were developed for detecting damage to the stator windings (early stage of a short circuit) and the rotor (breakage of one or two rotor cage bars), using spectral analysis of signals available in the internal structure of the vector drive control. All developed methods for detecting and compensating faults in the voltage inverter, measuring sensors, and induction motor windings were tested on developed and implemented test benches with a floating-point signal processor controlling the electric drive in the DRFOC and DTC-SVM structures.

The results of the work were presented at international conferences (ICIT, IECON, ISIE, PEMC, ELECTRIMACS, CPE-POWERENG, EPNC) and national ones (SENE, SME). The most valuable results of the work were published in international journals (IEEE Transactions on Industrial Electronics, Automation, Mathematics and Computers in Simulation, Archives of Electrical Engineering), and in the monograph "Advanced Control of Electrical Drives and Power Electronic Converters" (Elsevier). Some of the research results carried out within the project were included in the defended doctoral theses of the project scholarship holders (Piotr Sobański, "Diagnostics and control methods in drives with induction motors in emergency states of a two-level frequency converter" – 2017 and Kamil Klimkowski, "Analysis of drive systems with induction motors with fault-tolerant measuring sensors" – 2017.