Machinery signal separation using non-negative matrix factorization with real mixing

A big challenge in detecting damage occurs when the sound of a machine mixes with the sound of another machine. This paper proposes the separation of mixed acoustic signals using Non-negative Matrix Factorization (NMF) method for fault diagnosis. The NMF method is an effective solution for finding hidden parameters when the number of observations obtained by the sensor is less than the number of sources. The real mixing process is done by placing two microphones in front of the machine. Two microphones will be used as sensors to capture a mixture of four machinery signals. Performance testing of signal separation is done by comparing baseline signals with estimated signals through the mean log spectral distance (LSD) and the mean square error (MSE). The smallest spectral distance between the estimated signal and the baseline signal is found in Ŝ2 with an average LSD of 1.26. The estimated signal Ŝ2 is the closest to the baseline signal with MSE of 1.15 x 10-2. The pattern of bearing damage in the male screw compressor can be identified from the spectrum of estimated signal through harmonic frequencies as in the estimated signal Ŝ3 which is seen at 11x fundamental frequency, 12x fundamental frequency, 15x fundamental frequency, and 16x fundamental frequency.

Download Full-text

Music Signal Separation Using Supervised Robust Non-Negative Matrix Factorization with β-divergence

International Journal of Circuits, Systems and Signal Processing ◽

10.46300/9106.2021.15.16 ◽

2021 ◽

Vol 15 ◽

pp. 149-154

Author(s):

Feng Li ◽

Hao Chang

Keyword(s):

Prior Knowledge ◽

Matrix Factorization ◽

Signal Separation ◽

Separation Performance ◽

Effective Strategy ◽

Music Signal ◽

Training Sequences ◽

Conventional Methods ◽

Music Signals ◽

Non Negative Matrix Factorization

We propose a supervised method based on robust non-negative matrix factorization (RNMF) for music signal separation with β-divergence called supervised robust non-negative matrix factorization (SRNMF). Although RNMF method is an effective method for separating music signals, its separation performance degrades due to has no prior knowledge. To address this problem, in this paper, we develop SRNMF that unifying the robustness of RNMF and the prior knowledge to improve such separation performance on instrumental sound signals (e.g., piano, oboe and trombone). Application to the observed instrumental sound signals is an effective strategy by extracting the spectral bases of training sequences by using RNMF. In addition, β-divergence based on SRNMF be extended. The results obtained from our experiments on instrumental sound signals are promising for music signal separation. The proposed method achieves better separation performance than the conventional methods.

Download Full-text

DETECTION OF PUMP FAULTS BASED ON SOUND SIGNALS USING NON-NEGATIVE MATRIX FACTORIZATION

Jurnal Teknologi ◽

10.11113/jt.v82.13272 ◽

2020 ◽

Vol 82 (2) ◽

Author(s):

Anindita Adikaputri Vinaya ◽

Fitri Nurmaulidah ◽

Dhany Arifianto ◽

Qurrotin Ayunina Maulida Okta Arifianti

Keyword(s):

Matrix Factorization ◽

Instantaneous Frequency ◽

Local Spectrum ◽

Machine Condition ◽

Statistical Parameters ◽

Full Spectrum ◽

Spectral Distance ◽

A Value ◽

Recording Process ◽

Non Negative Matrix Factorization

Maintenance is very closely related to the performance of the production process. An alternative method that can be used to determine the damage to the engine is from the analysis of the sound pattern produced. If the sound source is more than one, then there will be signal mixing, and it will be a challenge in detecting damage to the engine. In this study, mixed signals will be separated. Separation of mixed sound signals was done using non-negative matrix factorization (NMF) method. Overall this study is aimed at detecting unbalance, misalignment, and bearing faults at pumps with microphones as sensors. The pumps used in this study were three pumps, where each pump had different conditions (unbalance, misalignment, and bearing fault). All three pumps have 3000 rpm. In this study, the recording process was carried out for 5 s. In this study, we also compare the location of the instantaneous frequency in full spectrum and corresponding frequency in local spectrum, and the distance between the spectra via the log spectral distance from the baseline signal and the estimated signal. Based on the instantaneous frequency approach, no error was found because of the instantaneous frequency suitability of the unbalanced machine condition with the estimated signal in the mixing configuration of three sources with two sensors. From the log spectral instance (LSD) results, the smallest value was obtained the smallest value in estimation 2, which tends to approach the unbalance condition with the LSD value of 1.0889. The most significant relative error is the estimated misalignment signal with a value of 11.2. However, overall damage can still be identified based on the pattern formed and some statistical parameters.

Download Full-text

A Novel Signal Separation Method Based on Improved Sparse Non-Negative Matrix Factorization

Entropy ◽

10.3390/e21050445 ◽

2019 ◽

Vol 21 (5) ◽

pp. 445

Author(s):

Huaqing Wang ◽

Mengyang Wang ◽

Junlin Li ◽

Liuyang Song ◽

Yansong Hao

Keyword(s):

Frequency Distribution ◽

Matrix Factorization ◽

Vibration Signal ◽

Separation Method ◽

Signal Separation ◽

Reference Vector ◽

Time Frequency ◽

Fault Features ◽

Time Frequency Distribution ◽

Non Negative Matrix Factorization

In order to separate and extract compound fault features of a vibration signal from a single channel, a novel signal separation method is proposed based on improved sparse non-negative matrix factorization (SNMF). In view of the traditional SNMF failure to perform well in the underdetermined blind source separation, a constraint reference vector is introduced in the SNMF algorithm, which can be generated by the pulse method. The square wave sequences are constructed as the constraint reference vector. The output separated signal is constrained by the vector, and the vector will update according to the feedback of the separated signal. The redundancy of the mixture signal will be reduced during the constantly updating of the vector. The time–frequency distribution is firstly applied to capture the local fault features of the vibration signal. Then the high dimension feature matrix of time–frequency distribution is factorized to select local fault features with the improved SNMF method. Meanwhile, the compound fault features can be separated and extracted automatically by using the sparse property of the improved SNMF method. Finally, envelope analysis is used to identify the feature of the output separated signal and realize compound faults diagnosis. The simulation and test results show that the proposed method can effectively solve the separation of compound faults for rotating machinery, which can reduce the dimension and improve the efficiency of algorithm. It is also confirmed that the feature extraction and separation capability of proposed method is superior to the traditional SNMF algorithm.

Download Full-text