Crack detection of plastic gears using a convolutional neural network pre-learned from images of meshing vibration data with transfer learning

Kien, B. H.; Iba, Daisuke; Ishii, Yunosuke; Tsutsui, Yusuke; Miura, Naruhisa; Iizuka, Takashi; Masuda, Arata; Sone, Akira; Moriwaki, Ichiro

doi:10.1007/s10010-019-00354-5

Cited by 25 publications

(7 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the literature work above, it is simple to realize that the gear health monitoring methods that are based on the vibrational signal processing technique are popular. The results of the research by [4][5][6][7], especially by [4,5], show that the proposed methods are effective for the purposes of determining gear failures. Moreover, the authors of these studies have also indicated that compared to the classic plain vibration-based gear health monitoring techniques, the accuracy of their condition monitoring results was enhanced significantly.…”

Section: Introductionmentioning

confidence: 92%

“…A recent literature review has given that, there are numerous interesting researches related to the topics of gear health monitoring have been introduced. Indeed, to enhance the vibration-based gear health monitoring technique, the proposed methods by [4,5] showed a combination of the vibration data analysis and a convolutional neuron network technique for the recognition of cracks on the plastic gear of a gear meshing. The research by [6] tried to combine acoustic emission and vibration signal processing for monitoring gears in a power gearbox.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Identification of Q value and angular frequency of smart gear antenna using the resonant return loss of the receiving antenna

et al. 2022

View full text Add to dashboard Cite

A wireless smart gear system has been being developed as a gear health monitoring system. This system works based on the magnetic coupling between its two major elements consisting of a monitoring antenna and a smart gear. In operation, the smart gear works as a working gear in the power transmission system while the monitoring antenna monitors this gear continuously. Specifically, since the monitoring antenna is fixed and kept connecting to a network analyzer during the gear operation, it allows the magnetic coupling to occur and remain thoroughly. As a result, the return loss signal of the monitoring antenna becomes a resonant return loss data of the smart gear system. In this paper, the authors propose a method to use this resonant return loss signal for identifying two parameters called the $$Q$$ Q value and the angular frequency of the smart gear side antenna. There are three cases of magnetic coupling that have been considered. These cases are called Open, Short, and Loaded by a 50 [Ω] resistor where the characteristics of the antenna on the smart gear side are changed via its connector by leaving Open or connecting to a Short and a 50 [Ω] resistor modules, respectively. The results indicate that the method is effective for parameter identification of the $$Q$$ Q value and the angular frequency of the smart gear side antenna for all three cases. Moreover, the analysis of the validated $$Q$$ Q value and the angular frequency also indicates the relation between these two characteristic values and the practical conditions of the smart gear. This finding consolidates the effectiveness of our developing wireless smart gear system. Article Highlights This article proposes a parameter identification method for a developing smart gear health monitoring system. By only using the resonant return loss data obtained from the monitoring antenna, there are two characteristic values of the smart gear have been achieved. This article also introduces a completed return loss equation of a magnetically coupled system. This return loss equation is helpful for the computation of characteristic parameters of the magnetically coupled system such as a wireless power transfer system or a coupled antenna system. The results figure out that the identified characteristic parameters and the physical condition of the smart gear side’s antenna circuit are related to each other. This finding affirms the effectiveness of not only the proposed parameter identification method but also the developing smart gear system.

show abstract

Section: Introductionmentioning

confidence: 92%

Section: Introductionmentioning

confidence: 99%

Identification of Q value and angular frequency of smart gear antenna using the resonant return loss of the receiving antenna

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Given their high-speed rotation during operation, directly installing sensors on gears is impractical. Indirect information has been employed, particularly in gear damage detection systems, wherein sensors are placed not on the gears themselves but on the housing or gearbox supporting the rotating shaft bearing, which accommodates the gears [1][2][3][4]. The vibration data obtained from this method encompasses not only gear meshing vibration but also vibrations and noise from other mechanical elements.…”

Section: Introductionmentioning

confidence: 99%

Gear health monitoring using smart gears with printed sensor/antenna circuits by running tests

Fujita,

Iba,

Sakai

et al. 2024

Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2024

View full text Add to dashboard Cite

The Smart Gear System comprises two integral components: the Smart Gear, featuring a sensor and an antenna circuit intricately embedded onto the gear, and a monitoring antenna seamlessly linked to a network analyzer. The proximity of the smart gear antenna to the monitoring antenna induces magnetic coupling between the two antennae, consequently altering the frequency characteristics of the monitoring antenna. Additionally, minute deformations in the gear tooth root result in modifications to the sensor's shape on the gear, thereby causing alterations in the frequency characteristics of the sensor. The overall health status of the gear can be discerned on the monitoring antenna side by monitoring the frequency characteristic variations of the sensor, based on the fundamental principle of magnetic coupling during operational processes. In the context of this investigation, a dynamic evaluation of the Smart Gear System is executed, with the return loss of the monitoring antenna meticulously gauged at periodic intervals throughout the operational phase. This study delves into the correlation between the return loss exhibited by the monitoring antenna and the dynamic physical alterations in the Smart Gear during operation.

show abstract

“…On the other hand, image-based computer vision can be used in defect detection [ 17 ]. Researchers tried to use 2D images of gears to recognize gear defects [ 18 ] besides the gray image transformed by vibration signals [ 19 , 20 ]. Nonetheless, it is difficult to recognize the defects of gears, especially on the tooth surface owing to its complex concave structure.…”

Section: Introductionmentioning

confidence: 99%

Defect detection of gear parts in virtual manufacturing

Wang

Hou

et al. 2023

Vis. Comput. Ind. Biomed. Art

View full text Add to dashboard Cite

Gears play an important role in virtual manufacturing systems for digital twins; however, the image of gear tooth defects is difficult to acquire owing to its non-convex shape. In this study, a deep learning network is proposed to detect gear defects based on their point cloud representation. This approach mainly consists of three steps: (1) Various types of gear defects are classified into four cases (fracture, pitting, glue, and wear); A 3D gear dataset was constructed with 10000 instances following the aforementioned classification. (2) Gear-PCNet+ + introduces a novel Combinational Convolution Block, proposed based on the gear dataset for gear defect detection to effectively extract the local gear information and identify its complex topology; (3) Compared with other methods, experiments show that this method can achieve better recognition results for gear defects with higher efficiency and practicability.

show abstract

Crack detection of plastic gears using a convolutional neural network pre-learned from images of meshing vibration data with transfer learning

Cited by 25 publications

References 6 publications

Identification of Q value and angular frequency of smart gear antenna using the resonant return loss of the receiving antenna

Identification of Q value and angular frequency of smart gear antenna using the resonant return loss of the receiving antenna

Gear health monitoring using smart gears with printed sensor/antenna circuits by running tests

Defect detection of gear parts in virtual manufacturing

Contact Info

Product

Resources

About