Search

Faults during machining processes generate negative impacts on productivity, product quality and scrap rate. In recent years, the research of leveraging deep learning algorithms for developing fault diagnostics approaches has been actively conducted. However, the approaches have not been widely adopted by industries yet due to their inadaptability in addressing varying working conditions throughout machining process lifecycles. To overcome the limitation, this paper presents a novel deep transfer learning enabled adaptive diagnostics approach. In the approach, firstly, a Convolutional Neural Network (CNN) is designed to perform diagnostics on machining processes. Then, a transfer learning strategy is incorporated into the CNN to enhance the approach's adaptability for different machining conditions via the following steps: (1) Input datasets from machining conditions are optimally aligned to facilitate cross-domain data reuse; and (2) Weights of the trained CNN are regularized to minimize feature distribution mismatches to implement domain transfer learning. Based on the steps, the CNN can be adaptively applied across the conditions, and thereby re-training processes for the CNN from scratch can be alleviated. The developed approach was validated and benchmarked based on different parameters and settings. In the experiments, comparative results indicate that the approach achieved 94% in accuracy, which was significantly higher than other approaches without transfer learning mechanisms.

To achieve zero-defect production during computer numerical control (CNC) machining processes, it is imperative to develop effective diagnosis systems to detect anomalies efficiently. However, due to the dynamic conditions of the machine and tooling during machining processes, the relevant diagnosis systems currently adopted in industries are incompetent. To address this issue, this paper presents a novel data-driven diagnosis system for anomalies. In this system, power data for condition monitoring are continuously collected during dynamic machining processes to support online diagnosis analysis. To facilitate the analysis, preprocessing mechanisms have been designed to de-noise, normalize, and align the monitored data. Important features are extracted from the monitored data and thresholds are defined to identify anomalies. Considering the dynamic conditions of the machine and tooling during machining processes, the thresholds used to identify anomalies can vary. Based on historical data, the values of thresholds are optimized using a fruit fly optimization (FFO) algorithm to achieve more accurate detection. Practical case studies were used to validate the system, thereby demonstrating the potential and effectiveness of the system for industrial applications.

The remarkable interface characteristics between micro-structure surface and water have raised great interests. The static and dynamic interface characteristics between functional micro-structure and water exhibit the wettability and drag reduction of the surface micro-structure separately. The computational fluid dynamics (CFD) model was built to simulate the special characteristics of water drops on micro-groove surfaces. Four kinds of micro-groove surfaces, which were V-groove, rectangular groove, sawtooth groove and dome-groove, were investigated. The reduction of fluid resistance on these micro-groove surfaces were found during the simulation. Good drag reduction efficiency on the V-groove surface can be achieved, which proves that the hydrophobic micro-groove surfaces have broad application prospect. Besides this, the simulation witnessed that the structural parameters of grooves have a significant influence on the drag reduction behaviours. The drag reduction mechanisms were also elaborated. Finally, the wettability properties of the proposed micro-groove surfaces were investigated experimentally to validate the simulation results.

In order to treat the performance of nano-electro-mechanical (NEM) switches in low vacuum, molecular dynamics model of NEM switches is established and the influence of air damping on NEM switches is investigated. When the pressure in low vacuum is from 100Pa to 1000Pa, we get the gas molecule in low vacuum belong to free molecule flow. Tersoff-Brenner potential function is used to describe the interaction of carbon nanotube atoms. L-J potential function is adopted to state the interaction between carbon nanotube and the base ground. Results show that in low vacuum condition, when the same amplitude change of intensity of pressure, the lower intensity of pressure change will take a more clearly effect for the nano structure than the higher intensity of pressure change. Results also show when the length of carbon nanotube from 200 circles to 50 circles, the beam vibrating amplitude increased gradually. When the length is 50 circles, the vibrating amplitude reach the maximum. With the beam length keep to decrease, the vibrating amplitude becomes weak crossly.

Metal-bonded ball-headed diamond wheel with small diameter has been widely used in the ultra-precision grinding of optical parts, especially in the parts with complex surface. In-position precise dressing technique of ball-headed diamond wheel is the key technology to improve grinding precision. The effect of mechanical errors on the profile accuracy of wheel dressing were theoretically analyzed firstly. Afterwards a ball-headed diamond wheel dressing equipment based on electrical discharge principle has been designed and built. Orthogonal experiments have been carried out on the dressing equipment, and the dressing quality of wheels under different dressing electrical parameters were analyzed in terms of wheel profile accuracy and wheel surface micro-topography. The influence of dressing electrical parameters on the dressing quality were investigated and the best electrical parameter combination was obtained. Experimental result showed that the wheel profile accuracy can reach upward 0.8μm after dressing, and the protrusion heights of diamond grain were uniformity and the protrusion effect was better than without dressing.

In this paper, the response to the first harmonic component (2f) of the electrostatic force in single terminal driven electrostatic comb-drive and parallel-plate drive was used as a signal to extract device parameters, namely, the Q-factor and resonant frequency instead of the fundamental (1f) resonance response. It is shown that the difficulty in motional measurement due to electrical cross-talk (parasitics) using 1fmeasurement can be overcome with a higher signal-to-noise ratio of the 2fsignal. Both atmospheric (low-Q) and reduced pressure environment were investigated using off-chip electronics and lock-in amplifier. The measurements were done on the electrostatic comb-drive and capacitive parallel plate sensing plates that form the two core modules of a yaw rate sensor (dual-axis resonator). The effects of AC and DC bias voltages on the measured response have been investigated. Experimental amplitude and phase response data have been analysed using the Lorentzian curve-fit, Resonance Curve Area (RCA) method, the half-power bandwidth method (3 dB) and the Nyquist plot for data fitting and determination of the Q-factor and resonance frequency.

A concise approach for the implementation of an anti-swing algorithm in fuzzy crane control is proposed. The algorithm is based on heuristics and can be conveniently incorporated into the existing fuzzy crane controller through creating proper displacements on the fuzzy rule plane. No modification to existing fuzzy rules is required.