Prediction of surface roughness based on a hybrid feature selection method and long short-term memory network in grinding

Guo, Weicheng; Wu, Chongjun; Zhou, Ding; Zhou, Qidi

doi:10.1007/s00170-020-06523-z

Cited by 111 publications

(30 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then it (the model) analyzes these signals in the time and frequency domains. Guo et al [7] developed a system for predicting surface roughness due to grinding. The system analyzes the grinding force, vibration, and AE signals in the time and frequency domains.…”

Section: Literature Reviewmentioning

confidence: 99%

“…When a machining operation continues in a given smart manufacturing environment, as seen in Figure 1, sensors collect signals. The signals are generally processed in the time [5], frequency [6][7][8], and time-frequency [5,6,9] domains to extract the underlying features. In some cases, alternative processing methods, such as fractal-based [10], Approximate Entropy (ApEn)-based, and Sampling Entropy (SampEn)-based [11] methods, are used.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Time Latency-Centric Signal Processing: A Perspective of Smart Manufacturing

Ura

Ghosh

2021

Sensors

View full text Add to dashboard Cite

Smart manufacturing employs embedded systems such as CNC machine tools, programable logic controllers, automated guided vehicles, robots, digital measuring instruments, cyber-physical systems, and digital twins. These systems collectively perform high-level cognitive tasks (monitoring, understanding, deciding, and adapting) by making sense of sensor signals. When sensor signals are exchanged through the abovementioned embedded systems, a phenomenon called time latency or delay occurs. As a result, the signal at its origin (e.g., machine tools) and signal received at the receiver end (e.g., digital twin) differ. The time and frequency domain-based conventional signal processing cannot adequately address the delay-centric issues. Instead, these issues can be addressed by the delay domain, as suggested in the literature. Based on this consideration, this study first processes arbitrary signals in time, frequency, and delay domains and elucidates the significance of delay domain over time and frequency domains. Afterward, real-life signals collected while machining different materials are analyzed using frequency and delay domains to reconfirm its (the delay domain’s) significance in real-life settings. In both cases, it is found that the delay domain is more informative and reliable than the time and frequency domains when the delay is unavoidable. Moreover, the delay domain can act as a signature of a machining situation, distinguishing it (the situation) from others. Therefore, computational arrangements enabling delay domain-based signal processing must be enacted to effectively functionalize the smart manufacturing-centric embedded systems.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Time Latency-Centric Signal Processing: A Perspective of Smart Manufacturing

Ura

Ghosh

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…They used the improved Pareto particle swarm algorithm to optimize the two parameters of production efficiency and roughness. In the grinding experiments of C-250 maraging steel, Guo et al [ 21 ] collected force signal and acoustic signal characteristics. A long short-term memory (LSTM) network algorithm is proposed based on a time series to predict workpiece roughness.…”

Section: Introductionmentioning

confidence: 99%

Prediction of Surface Roughness of Abrasive Belt Grinding of Superalloy Material Based on RLSOM-RBF

et al. 2021

View full text Add to dashboard Cite

It is difficult to accurately predict the surface roughness of belt grinding with superalloy materials due to the uneven material distribution and complex material processing. In this paper, a radial basis neural network is proposed to predict surface roughness. Firstly, the grinding system of the superalloy belt is introduced. The effects of the material removal process and grinding parameters on the surface roughness in belt grinding were analyzed. Secondly, an RBF neural network is trained by reinforcement learning of a self-organizing mapping method. Finally, the prediction accuracy and simulation results of the proposed method and the traditional prediction method are analyzed using the ten-fold cross method. The results show that the relative error of the improved RLSOM-RBF neural network prediction model is 1.72%, and the R-value of the RLSOM-RBF fitting result is 0.996.

show abstract

“…In traditional machining techniques, such as grinding and milling, vibrations created in the milling process contribute significantly to machining accuracy and quality [5][6][7][8], Vibration phenomenon in milling can lead to poor machining surface quality and impair machine life. Shtehin et al [9] have carried out experimental research on low frequency vibration when the spherical milling cutter is machining bevels.…”

Section: Introductionmentioning

confidence: 99%

Effect of Milling Vibration on The Machining Performance of 7075-T651 Aluminum Alloys

Guo

Wang

Liu

et al. 2021

Preprint

View full text Add to dashboard Cite

Milling of 7075-T651 is widely used in aerospace industry, however the process vibration restricts the machining performance in milling process. This paper puts forward a study on the effect of vibration on machining performance in milling to improve the machining quality. According to the characteristics of end milling, the process vibration is calculated and added based on the unformed chip thickness model of milling, and a milling simulation model considering vibration is established. Applying the finite element model and milling experiments, the simulation model is verified, the results proves the accuracy of the FEM models in predicting the milling force and milling temperature. Furthermore, the effect of milling vibration on machining performance is studied by numerical simulation, in which the relationship between amplitude-frequency characteristics and milling force-temperature fluctuation.

show abstract

Prediction of surface roughness based on a hybrid feature selection method and long short-term memory network in grinding

Cited by 111 publications

References 38 publications

Time Latency-Centric Signal Processing: A Perspective of Smart Manufacturing

Time Latency-Centric Signal Processing: A Perspective of Smart Manufacturing

Prediction of Surface Roughness of Abrasive Belt Grinding of Superalloy Material Based on RLSOM-RBF

Effect of Milling Vibration on The Machining Performance of 7075-T651 Aluminum Alloys

Contact Info

Product

Resources

About