Application of Python and open hardware tools for machinery fault diagnosis
Application of Python and open hardware tools for machinery fault diagnosis

Summary/Abstract: The requirements for the quality and reliability of operation of industrial machinery are constantly growing. This necessitates an increase in the level of automated analysis and prediction of failures thus transforming the emergency repair into a planned one. Significant progress has been made worldwide in this direction in recent years, but still commercially available tools are too costly for researchers. The paper discusses an ongoing research on building a low cost data acquisition system for condition assessment of electric motors employing non-invasive methods for detecting failures. It is well known that the changes in the load and various types of damage lead to the occurrence of additional harmonics inthe spectrum of the current signal of the machine and it has a noticeable impact on its profile. This method is called a Motor Current Signature Analysis (MCSA). A custom built software based on the Python programming language is used in the paper to perform MCSA by using two methods. The first one consists of decomposing the signal spectrum using a Fast Fourier Transform (FFT) and the second one employs the Park Vector Approach (PVA) in which the vector pattern degenerates from its idealistic circular shape for the healthy machine to a distorted ellipse for a faulty machine. While still currently requiring a computer for solving the problem directions for building hardware based on Field Programmable Gate Array (FPGA) device are given. This would allow for performing the entire analysis in true real time ensuring a fast and reliable response and thus achieving longer machine life and increased production which has a serious economic effect.

• Details
• Contents