Jiayi Lu, Boao Zhang, Rui Cao, Bin Sun, Rui Wang, Zhaowen Pang, Zexiang Tong, Shichun Yang, Yaoguang Cao
{"title":"基于压力的智能座椅系统:一种驾驶员行为识别的新方法","authors":"Jiayi Lu, Boao Zhang, Rui Cao, Bin Sun, Rui Wang, Zhaowen Pang, Zexiang Tong, Shichun Yang, Yaoguang Cao","doi":"10.1016/j.measurement.2025.117588","DOIUrl":null,"url":null,"abstract":"<div><div>As intelligent vehicles become increasingly prevalent, ensuring their operational safety has emerged as a paramount challenge. Vehicles equipped with advanced intelligent systems offer the potential to enhance safety by continuously monitoring and analyzing the driver’s condition and behavior in real time, thereby aligning vehicle control more closely with the driver’s intentions. While visual detection methods within the cockpit are often impeded by obstructions such as other passengers or interior elements, the driver’s seat provides a unique advantage due to its constant, unobstructed contact with the driver throughout the driving process. This positioning enables the seat to detect subtle changes in its morphology, offering a reliable means to map the driver’s state with minimal interference. This paper introduces a novel pressure-based intelligent seat system integrated with an array of polyvinylidene fluoride (PVDF) sensors and proposes a tailored Pressure Mat Vision Transformer (ViT) model to accurately classify the driver’s operational behaviors. The system demonstrates the capability to predict the precise depth of throttle, brake, and clutch pedal engagement, as well as the steering wheel’s control state. Validation of the model’s performance was conducted using a custom-built dataset. Furthermore, the feasibility of the pressure recognition technology and the efficacy of the fine-tuned architecture for real-time application were substantiated through engineering deployment on an operational platform. The development of this intelligent seat represents a significant advancement in addressing safety concerns related to human–machine interactions and establishes an innovative framework for leveraging data in the evolution of next-generation intelligent vehicles.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"253 ","pages":"Article 117588"},"PeriodicalIF":5.2000,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A pressure-based intelligent seat systems: A novel approach to driver behavior recognition\",\"authors\":\"Jiayi Lu, Boao Zhang, Rui Cao, Bin Sun, Rui Wang, Zhaowen Pang, Zexiang Tong, Shichun Yang, Yaoguang Cao\",\"doi\":\"10.1016/j.measurement.2025.117588\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>As intelligent vehicles become increasingly prevalent, ensuring their operational safety has emerged as a paramount challenge. Vehicles equipped with advanced intelligent systems offer the potential to enhance safety by continuously monitoring and analyzing the driver’s condition and behavior in real time, thereby aligning vehicle control more closely with the driver’s intentions. While visual detection methods within the cockpit are often impeded by obstructions such as other passengers or interior elements, the driver’s seat provides a unique advantage due to its constant, unobstructed contact with the driver throughout the driving process. This positioning enables the seat to detect subtle changes in its morphology, offering a reliable means to map the driver’s state with minimal interference. This paper introduces a novel pressure-based intelligent seat system integrated with an array of polyvinylidene fluoride (PVDF) sensors and proposes a tailored Pressure Mat Vision Transformer (ViT) model to accurately classify the driver’s operational behaviors. The system demonstrates the capability to predict the precise depth of throttle, brake, and clutch pedal engagement, as well as the steering wheel’s control state. Validation of the model’s performance was conducted using a custom-built dataset. Furthermore, the feasibility of the pressure recognition technology and the efficacy of the fine-tuned architecture for real-time application were substantiated through engineering deployment on an operational platform. The development of this intelligent seat represents a significant advancement in addressing safety concerns related to human–machine interactions and establishes an innovative framework for leveraging data in the evolution of next-generation intelligent vehicles.</div></div>\",\"PeriodicalId\":18349,\"journal\":{\"name\":\"Measurement\",\"volume\":\"253 \",\"pages\":\"Article 117588\"},\"PeriodicalIF\":5.2000,\"publicationDate\":\"2025-04-25\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Measurement\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0263224125009479\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, MULTIDISCIPLINARY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Measurement","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0263224125009479","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MULTIDISCIPLINARY","Score":null,"Total":0}
A pressure-based intelligent seat systems: A novel approach to driver behavior recognition
As intelligent vehicles become increasingly prevalent, ensuring their operational safety has emerged as a paramount challenge. Vehicles equipped with advanced intelligent systems offer the potential to enhance safety by continuously monitoring and analyzing the driver’s condition and behavior in real time, thereby aligning vehicle control more closely with the driver’s intentions. While visual detection methods within the cockpit are often impeded by obstructions such as other passengers or interior elements, the driver’s seat provides a unique advantage due to its constant, unobstructed contact with the driver throughout the driving process. This positioning enables the seat to detect subtle changes in its morphology, offering a reliable means to map the driver’s state with minimal interference. This paper introduces a novel pressure-based intelligent seat system integrated with an array of polyvinylidene fluoride (PVDF) sensors and proposes a tailored Pressure Mat Vision Transformer (ViT) model to accurately classify the driver’s operational behaviors. The system demonstrates the capability to predict the precise depth of throttle, brake, and clutch pedal engagement, as well as the steering wheel’s control state. Validation of the model’s performance was conducted using a custom-built dataset. Furthermore, the feasibility of the pressure recognition technology and the efficacy of the fine-tuned architecture for real-time application were substantiated through engineering deployment on an operational platform. The development of this intelligent seat represents a significant advancement in addressing safety concerns related to human–machine interactions and establishes an innovative framework for leveraging data in the evolution of next-generation intelligent vehicles.
期刊介绍:
Contributions are invited on novel achievements in all fields of measurement and instrumentation science and technology. Authors are encouraged to submit novel material, whose ultimate goal is an advancement in the state of the art of: measurement and metrology fundamentals, sensors, measurement instruments, measurement and estimation techniques, measurement data processing and fusion algorithms, evaluation procedures and methodologies for plants and industrial processes, performance analysis of systems, processes and algorithms, mathematical models for measurement-oriented purposes, distributed measurement systems in a connected world.