Failure Detection by signal similarity measurement of Brushless DC motors

Abstract

During the last years the Brushless DC (BLDC) motors are gaining popularity as a solution for providing mechanical power, starting from low cost mobility solutions like the electric bikes, to high performance and high reliability aeronautical Electro- Mechanical Actuators (EMAs). In this framework, the availability of fault detection tools suited for these types of machines appears necessary. There is already a vast literature on this topic, but only a small percentage of the proposed techniques are developed to a Technology Readiness Level (TRL) sufficiently high to be implementable in industrial applications. The investigation on the state of the art carried out during the first phase of the present work, tries to collect the articles which are closer to a possible implementation. This choice has been influenced by the author experience when dealing with fault detection papers, which often are oriented towards a more academic public and do not concentrate on the implementation. The methodology used in this work to compile the state of the art has been the Systematic Literature Review (SLR) and it is still not diffused in the engineering world. For this reason a dedicated description has been inserted in the respective chapter of the thesis. From this study, some characteristics needed for the fault detection on electric machine have been listed and a new technique for demagnetisation detection on BLDC motors has been proposed. In the second part of the thesis, it is presented an algorithm to detect demagnetisation based on the dissimilarity between the voltages of the various electric turns of the motor due to this failure. The exposed method presents the advantages of not needing domain transforms or previous knowledge of the motor (made exception for the number of pole-pairs). Furthermore the proposed indicators are fast to be computed and require only the acquisition of motor phases voltages for a mechanical turn. The hypotheses made about the effect of a demagnetisation with Finite Element Method (FEM) have also been confirmed through simulations analysis and the proposed method to detect demagnetisation has been validated with experimental tests on a real motor. 2 Applications and Limitations The presented indicators have been studied, simulated and experimented only on an outrunner, low power BLDC motor. Anyway it is not excluded that, with some adaptation, they could be used on any BLDC motor or also on different types of motors; indeed this is an argument for a future work. Another important consideration is that, in order to detect demagnetisation, the motor should have a number of pole pairs greater than 2. This because the algorithm compares the electric turns between them and it is obviously necessary to have more than one. Another characteristic is that it can only detect partial demagnetisation. The demagnetisation of all the magnets to the same level, although very improbable, would not cause those differences in the voltage signals needed for fault detection. Various tests have been executed both at fixed and variable speed. In the first case it was possible to define a threshold to discern between the healthy and the demagnetised motor, while in the second case, even if the indicators are still separated, it was not possible to define a fixed threshold. Hence, if no classification algorithms are used (Support Vector Machine (SVM), Neural Network (NN), Artificial Intelligence (AI), etc.), the indicator shall be computed when the motor is running in steady state conditions. 3 Advantages The method of fault detection by using the proposed indicators has the main advantage of being straightforwardly applicable with no need of extra hardware. Another important characteristic to be highlighted is that the only previous needed knowledge of the motor is the number of pole-pairs. Also the intermediate data are easy to understand as they represent physical variables of the motor in the time domain. Thanks to this, also no domain transformations for frequency analysis are needed, saving computation time. The algorithm to compute the indicators is composed by few steps, it is fast to execute and does not need complex programming or libraries. Indeed the execution time for the PC implementation is already very low and an optimised implementation in a lower level programming language could easily fit in a microcontroller and be executed at even higher speed, permitting both real time monitoring and punctual testing during maintenance. Furthermore it uses only few and easily obtainable data, which makes it suitable for every industrial implementation and interesting for further academic researches. Having a maximum theoretical value for the indicator is also an important advantage, because it permits to evaluate a motor without previous knowledge of the same; indeed a healthy motor should have an ixc value always very close to this maximum value. It is worth to notice that the proposed indicators have been validated with experimental tests in various conditions, showing both good performances and space for further improvements. Finally, although it is true that constant speed is required for a correct analysis, it is needed for just a mechanical turn, i.e. for few milliseconds. For example if the motor is running at 3000 RPM, a complete turn is executed in 20 ms.