Cutting condition monitoring is essential for managing the operation of machine tools in manufacturing. The t-distributed stochastic neighbor embedding (t-SNE) method has been adopted to learn the underlying relationship between cutting parameters in a manifold. This article focuses on cutting force embedding in the t-SNE, termed Ft-SNE, for applications in feature selection to achieve better performance for condition monitoring of machine tools. In the method, the measured vibration signals are used to build the fundamental feature space in the high-dimensional space with cutting force signals being introduced to improve its separability. Further analysis is performed to optimize the classification performance of the Ft-SNE by selecting the appropriate weighting coefficient by minimizing the Kullback–Leibler (KL) divergence. The experiments were implemented on a three-axis vertical machining center. Vibration signals and cutting force signals under different cutting conditions were collected from the vertical machining center. The visualization results of the Ft-SNE and t-SNE were analyzed in low-dimensional space. By introducing the tangential cutting force, the classification performance is improved with classification evaluation indicators being around 98%, with an improvement of over 41% compared to the traditional t-SNE for cutting condition classification in machining process.
23 July 2024 in IEEE Journal of Emerging and Selected Topics in Power Electronics
The space vector pulse width modulation has been widely used in AC motors powered by the two-level three-phase inverters. However, the significant switching frequency harmonics are generated in the power spectrum of the output line voltage. To suppress the sideband harmonics near the switching frequency and its integer multiples, a novel cost function using the variance of all higher-order harmonics in the power spectrum is established and an optimization method for the selection of the switching frequency is proposed. According to an optimization case study, the optimal switching frequency distribution for the harmonic diffusion is the beta distribution. Furthermore, this paper is concerned with the optimization parameter selection including the number of the discrete switching frequency and the shape parameters of beta distribution to improve the harmonic dispersion effect in the line voltage spectrum. Simulation and experimental results demonstrate that a significant effect of the sideband harmonic dispersion is achieved when the number of discrete switching frequency is 2 × 10 1 ~ 5 × 10 3 and the beta distribution shape parameters are set as 0.09 ~ 0.21. This provides the theoretical basis for the selection of the switching frequency and hence reduce the harmonics near the switching frequency in motor drive systems.
This paper presents an active acoustic excitation method for leak detection of buried gas pipelines based on cavity resonance reflection. The principles of gas leakage in pipelines are analyzed, including the gas passage model and the gas cavity model. The principle of Helmholtz resonator is employed to establish the cavity model. For the cavity model, the relationships between cavity resonance frequency, acoustic impedance, sound pressure amplification and leakage damage size are derived. The resonant effect of the gas cavity on the acoustic signal is considered in this study to solve the problem that the echo signal after long distance propagation and reflection becomes very weak. Numerical simulations are conducted to demonstrate the relationships between acoustic reflection coefficient of the leak hole size, cavity volume and pipe wall thickness. In order to verify the effectiveness of the proposed method, a pipeline experimental rig with a length of 100 m is constructed. Sound waves are generated by a speaker and reflected echoes are received by a microphone. The cavity resonance reflection and echo characteristics of different leak hole size, different transmitting acoustic frequency and different cavity volume are analyzed. The EMD algorithm is used to decompose and reconstruct the echo signals to eliminate the noise interference in the pipeline system. An echo time-distance conversion method is used to visualize the locations of the leak hole and welds. Experimental results show that the proposed method can effectively detect the leak holes and welds in the pipeline.
14 May 2024 in Mechanical Systems and Signal Processing
Tunnels are an essential component of modern transportation infrastructure, and their structural health is critical to traffic safety, which can be seriously affected by tunnel lining cavities. In this paper, an acoustic-based detection approach for assessing the integrity of tunnel linings is studied. By tapping the tunnel lining surface, acoustic signals are sampled and analyzed using a novel feature parameter extraction algorithm-the energy-frequency cepstral coefficient, which uses wavelet packet decomposition to obtain energy distribution statistics in the frequency domain of the signal, and constructs a signal-dependent filter bank to achieve the cepstral coefficient extraction. Compared with the traditional Mel filter bank, this method can adaptively adjust the resolution of the filter bank according to the frequency characteristics of the classified samples. This allows for higher frequency resolution in regions where the energy distribution is concentrated. As a result, the extracted feature parameters achieve both dimensional compression and superior information retention. Experimental results show that the proposed energy-frequency cepstral coefficient feature outperforms the traditional Mel-frequency cepstral coefficient feature, resulting in a higher accuracy of tunnel lining detection. The convolutional neural network model achieves an accuracy of 99.2%, with a 78.9% reduction in error rate compared with the traditional Mel-frequency cepstral coefficient feature parameters. Additionally, a particle swarm optimization support vector machine model is trained to achieve an accuracy rate of 99.6% and an error rate reduction of 76.5%.