Post by account_disabled on Feb 20, 2024 3:26:31 GMT
To carry out the supervision of this machine, it was decided to use a multivariate analysis. To this end, after analyzing different types of signals, the aim is to choose the most appropriate and sensitive ones to monitor tool wear. Initially it was decided to use the following signals: Total current consumed by each motor, delivered by the drives. Angular speed of the three motors. Vibration corresponding to the rotation of the crankshaft. After an analysis of the problem, two more types of signals were added: Measurement of the sound produced by the machine. Surface temperature of the five machined supports.
Each of these signals must be sampled for 60 seconds for each of the 1,100 crankshafts that the tool machines. We will have to analyze the evolution of the signals throughout the different machined parts. Likewise, it is convenient to capture the signals in the case in which Asia Mobile Number List the machine works in a vacuum, in order to eliminate any type of noise or behavior not associated with the inherent machining process. The analysis carried out is divided into two parts: 1) Individual Analysis. We first proceed to analyze each of the signals individually. Time domain. – Visual inspection: Waveforms, peaks, continuity, trend. – Temporal segmentation: the temporal signal must be divided into fragments corresponding to the machining by each tool plate. – Periodicity analysis.
Statistical analysis: globally and by sections, mean, variance and higher order moments. Frequency domain. – Decomposition of the spectrum into various frequency intervals. – Main frequencies and values. – Repetitiveness. 2) Joint Analysis. After the individual analysis of each signal, the influence between the different types of signals must be analyzed, in order to find relationships between them. Time domain. – Variable-variable correlation. – Input-output model. Frequency domain. – Comparison of spectra. - Transfer function. Analysis of results . Monitoring the tool requires analyzing the dynamic behavior of the motors' response, due to the attack of each of the blades on the piece. Figure 3 – Evolution of the maximum current. In this way, based on the captured signals, the behavior of certain parameters (maximum current, average, frequencies) of each blade must be analyzed as the successive pieces are machined.
Each of these signals must be sampled for 60 seconds for each of the 1,100 crankshafts that the tool machines. We will have to analyze the evolution of the signals throughout the different machined parts. Likewise, it is convenient to capture the signals in the case in which Asia Mobile Number List the machine works in a vacuum, in order to eliminate any type of noise or behavior not associated with the inherent machining process. The analysis carried out is divided into two parts: 1) Individual Analysis. We first proceed to analyze each of the signals individually. Time domain. – Visual inspection: Waveforms, peaks, continuity, trend. – Temporal segmentation: the temporal signal must be divided into fragments corresponding to the machining by each tool plate. – Periodicity analysis.
Statistical analysis: globally and by sections, mean, variance and higher order moments. Frequency domain. – Decomposition of the spectrum into various frequency intervals. – Main frequencies and values. – Repetitiveness. 2) Joint Analysis. After the individual analysis of each signal, the influence between the different types of signals must be analyzed, in order to find relationships between them. Time domain. – Variable-variable correlation. – Input-output model. Frequency domain. – Comparison of spectra. - Transfer function. Analysis of results . Monitoring the tool requires analyzing the dynamic behavior of the motors' response, due to the attack of each of the blades on the piece. Figure 3 – Evolution of the maximum current. In this way, based on the captured signals, the behavior of certain parameters (maximum current, average, frequencies) of each blade must be analyzed as the successive pieces are machined.
