当前位置: X-MOL 学术IEEE ASME Trans. Mechatron. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Global Optimization of the Hydraulic-Electromagnetic Energy-Harvesting Shock Absorber for Road Vehicles With Human-Knowledge-Integrated Particle Swarm Optimization Scheme
IEEE/ASME Transactions on Mechatronics ( IF 6.4 ) Pub Date : 2021-02-01 , DOI: 10.1109/tmech.2021.3055815
Quan Zhou , Sijing Guo , Lin Xu , Xuexun Guo , Huw Williams , Hongming Xu , Fuwu Yan

This article proposes a human-knowledge-integrated particle swarm optimization (Hi-PSO) scheme to globally optimize the design of the hydraulic-electromagnetic energy-harvesting shock absorber (HESA) for road vehicles. A newly developed k-fold swarm learning framework is the key to the Hi-PSO scheme, which runs k groups (folds) of individual local optimization (using a selected learning cycle), and validation (using the other k-1 testing cycles) with the concept of digital twin introduced into the design of the HESA. It aims to achieve the optimum energy recovery efficiency globally in both learning cycles and testing cycles. Within the learning framework, a nearest-neighborhood particle swarm learning algorithm is developed to incorporate human knowledge (e.g., ISO standards) for local optimization so that the computational load can be reduced through downsizing of the learning spaces. Experiments have been conducted to evaluate the energy recovery and damping performance under both local conditions (duty cycles used for learning) and global conditions (six duty cycles covering the main equivalent amplitudes and frequencies of the suspension's operation). Compared with the conventional PSO algorithm, Hi-PSO is shown to be more robust by achieving a 5.17% higher mean value in 10 trials while achieving the same maximum energy efficiency. The global optimum result is obtained under 20 mm/1.5 Hz condition and achieves an average energy efficiency of 59.07%.

中文翻译:

基于人-知识-集成粒子群优化方案的道路车辆液压-电磁能量收集减振器的全局优化

本文提出了一种人类知识集成的粒子群优化(Hi-PSO)方案,以全局优化道路车辆液压-电磁能量收集减震器(HESA)的设计。新开发的 k-fold swarm 学习框架是 Hi-PSO 方案的关键,它运行 k 组(折叠)个体局部优化(使用选定的学习周期)和验证(使用其他 k-1 测试周期) HESA 的设计中引入了数字孪生的概念。它旨在在学习周期和测试周期中实现全球最佳能量回收效率。在学习框架内,开发了最近邻粒子群学习算法以结合人类知识(例如,ISO 标准)进行局部优化,以便通过缩小学习空间来减少计算负载。已经进行了实验以评估在局部条件(用于学习的占空比)和全局条件(涵盖悬架运行的主要等效振幅和频率的六个占空比)下的能量恢复和阻尼性能。与传统的 PSO 算法相比,Hi-PSO 在 10 次试验中获得了高出 5.17% 的平均值,同时实现了相同的最大能效,显示出更稳健。全局最优结果是在20 mm/1.5 Hz条件下获得的,平均能效为59.07%。已经进行了实验以评估在局部条件(用于学习的占空比)和全局条件(六个占空比,涵盖悬架运行的主要等效振幅和频率)下的能量恢复和阻尼性能。与传统的 PSO 算法相比,Hi-PSO 在 10 次试验中获得了高出 5.17% 的平均值,同时实现了相同的最大能效,显示出更稳健。全局最优结果是在20 mm/1.5 Hz条件下获得的,平均能效为59.07%。已经进行了实验以评估在局部条件(用于学习的占空比)和全局条件(六个占空比,涵盖悬架运行的主要等效振幅和频率)下的能量恢复和阻尼性能。与传统的 PSO 算法相比,Hi-PSO 在 10 次试验中获得了高出 5.17% 的平均值,同时实现了相同的最大能效,显示出更稳健。全局最优结果是在20 mm/1.5 Hz条件下获得的,平均能效为59.07%。10 次试验的平均值高出 17%,同时实现相同的最大能效。全局最优结果是在20 mm/1.5 Hz条件下获得的,平均能效为59.07%。10 次试验的平均值高出 17%,同时实现相同的最大能效。全局最优结果是在20 mm/1.5 Hz条件下获得的,平均能效为59.07%。
更新日期:2021-02-01
down
wechat
bug