{"title":"通过机械数据驱动的主动悬架道路预览控制解锁智能车辆的最佳乘坐舒适性","authors":"Cheng Wang, Guanyu Tao, Xiaoxian Cui, Quan Yao, Xinran Zhou, Konghui Guo","doi":"10.1016/j.aei.2025.103592","DOIUrl":null,"url":null,"abstract":"<div><div>Ride comfort is vital for intelligent vehicle adoption, and active suspension systems are central to its improvement. Road preview perception can further enhance suspension performance, yet effectively integrating road data into control strategies remains challenging. This paper introduces DGRL-RP, a mechanism-data-driven road preview control strategy that merges differential geometry analysis with deep reinforcement learning (DRL). First, differential geometry elucidates how future road information influences suspension dynamics, enabling decomposition into mechanism and data-driven modules. In the data-driven module, an expert-guided soft-hard module—which normalizes multi-scale state inputs and enforces actuator limits through combined soft and hard constraints—is integrated with the Twin Delayed Deep Deterministic Policy (TD3-SH) algorithm and a Deterministic Experience Tracking (DET) mechanism to accelerate learning and convergence. Simulation experiments using the root mean square (RMS) of body acceleration as the performance metric demonstrate that DGRL-RP outperforms Model Predictive Control (MPC) by 71.73% and TD3-SH by 65.39%. Moreover, it maintains over 90% comfort optimization across diverse scenarios, illustrating superior control performance and strong generalization. DGRL-RP offers a novel solution for ride comfort optimization and advances active suspension control toward greater intelligence and precision.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"68 ","pages":"Article 103592"},"PeriodicalIF":8.0000,"publicationDate":"2025-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Unlocking optimal ride comfort in intelligent vehicles via mechanism-data-driven active suspension road preview control\",\"authors\":\"Cheng Wang, Guanyu Tao, Xiaoxian Cui, Quan Yao, Xinran Zhou, Konghui Guo\",\"doi\":\"10.1016/j.aei.2025.103592\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Ride comfort is vital for intelligent vehicle adoption, and active suspension systems are central to its improvement. Road preview perception can further enhance suspension performance, yet effectively integrating road data into control strategies remains challenging. This paper introduces DGRL-RP, a mechanism-data-driven road preview control strategy that merges differential geometry analysis with deep reinforcement learning (DRL). First, differential geometry elucidates how future road information influences suspension dynamics, enabling decomposition into mechanism and data-driven modules. In the data-driven module, an expert-guided soft-hard module—which normalizes multi-scale state inputs and enforces actuator limits through combined soft and hard constraints—is integrated with the Twin Delayed Deep Deterministic Policy (TD3-SH) algorithm and a Deterministic Experience Tracking (DET) mechanism to accelerate learning and convergence. Simulation experiments using the root mean square (RMS) of body acceleration as the performance metric demonstrate that DGRL-RP outperforms Model Predictive Control (MPC) by 71.73% and TD3-SH by 65.39%. Moreover, it maintains over 90% comfort optimization across diverse scenarios, illustrating superior control performance and strong generalization. DGRL-RP offers a novel solution for ride comfort optimization and advances active suspension control toward greater intelligence and precision.</div></div>\",\"PeriodicalId\":50941,\"journal\":{\"name\":\"Advanced Engineering Informatics\",\"volume\":\"68 \",\"pages\":\"Article 103592\"},\"PeriodicalIF\":8.0000,\"publicationDate\":\"2025-06-28\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Advanced Engineering Informatics\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S1474034625004859\",\"RegionNum\":1,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Advanced Engineering Informatics","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1474034625004859","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
Unlocking optimal ride comfort in intelligent vehicles via mechanism-data-driven active suspension road preview control
Ride comfort is vital for intelligent vehicle adoption, and active suspension systems are central to its improvement. Road preview perception can further enhance suspension performance, yet effectively integrating road data into control strategies remains challenging. This paper introduces DGRL-RP, a mechanism-data-driven road preview control strategy that merges differential geometry analysis with deep reinforcement learning (DRL). First, differential geometry elucidates how future road information influences suspension dynamics, enabling decomposition into mechanism and data-driven modules. In the data-driven module, an expert-guided soft-hard module—which normalizes multi-scale state inputs and enforces actuator limits through combined soft and hard constraints—is integrated with the Twin Delayed Deep Deterministic Policy (TD3-SH) algorithm and a Deterministic Experience Tracking (DET) mechanism to accelerate learning and convergence. Simulation experiments using the root mean square (RMS) of body acceleration as the performance metric demonstrate that DGRL-RP outperforms Model Predictive Control (MPC) by 71.73% and TD3-SH by 65.39%. Moreover, it maintains over 90% comfort optimization across diverse scenarios, illustrating superior control performance and strong generalization. DGRL-RP offers a novel solution for ride comfort optimization and advances active suspension control toward greater intelligence and precision.
期刊介绍:
Advanced Engineering Informatics is an international Journal that solicits research papers with an emphasis on 'knowledge' and 'engineering applications'. The Journal seeks original papers that report progress in applying methods of engineering informatics. These papers should have engineering relevance and help provide a scientific base for more reliable, spontaneous, and creative engineering decision-making. Additionally, papers should demonstrate the science of supporting knowledge-intensive engineering tasks and validate the generality, power, and scalability of new methods through rigorous evaluation, preferably both qualitatively and quantitatively. Abstracting and indexing for Advanced Engineering Informatics include Science Citation Index Expanded, Scopus and INSPEC.