Development of a Prediction Model for Tire Tread Pattern Noise Based on Convolutional Neural Network with RMSProp Algorithm
Bibliographic record
Abstract
Tire tread pattern noise is a major source of road noise generated by motor vehicles. Recently, noise control technology has been developing, and low-noise motor vehicles, such as electric vehicles and hybrid vehicles, have been commercialized. The importance of low-noise tires has increased since regulations R117 for tire noise and R51.03 for motor vehicle noise have been strengthened. To evaluate the tire noise in the development stage of motor vehicles, finished products of tires are required; hence, financial and time costs should be invested. Therefore, it is highly useful to predict tire noise levels in the early stages. Recently, a technology to predict the tire pattern noise using a supervised training method of artificial neural network (ANN) has been developed. The tire tread depth is estimated using the shading of the full image of the actual tire, and the leading edge of the contact patch is calculated using tire contact patch images. This method creates an artificial intelligence learning model by scanning the entire tire image with the leading edge, making input factors by Gaussian curve fitting of the tread profile spectrum and air volume velocity spectrum according to the tire rotation speed and evaluating the vehicle road noise. However, because this method requires finished products of tires, it is difficult to use it for prediction in the early stage. In this study, a convolutional neural network based on the unsupervised training method was developed to predict the tire tread pattern noise. The prediction results of applying two learning algorithms, SGD and RMSProp, to the CNN model showed that the RMSProp algorithm displayed a good predictive power in the CNN model. The tire pattern image to be designed was used as the input of the CNN model. The pattern noises of 28 tires were measured in coast-down condition of R117 on the ISO10844 certified road surface, and pattern images were scanned. The tread pattern noises and pattern images for 24 tires were used for the ANN and CNN, and trained ANN and CNN models were used for the verification of the remaining four untrained tires. Two training models were successfully developed and verified for the prediction of tire tread pattern noise. The trained CNN model can be used to predict the tire tread pattern noise in the early stage using only drawn tire images. Furthermore, the ANN model can be used to predict the pattern noise of actual tires in the developing stage, and it was verified by the actual mold design.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".