汽车安全与节能学报
汽車安全與節能學報
기차안전여절능학보
JOURNAL OF AUTOMOTIVE SAFETY AND ENGERGY
2013年
4期
339-347
,共9页
张继游%门永新%彭鸿%冯擎峰
張繼遊%門永新%彭鴻%馮擎峰
장계유%문영신%팽홍%풍경봉
汽车安全%侧面碰撞%轻量化%试验设计(DOE)%6σ稳健性优化%响应面模型
汽車安全%側麵踫撞%輕量化%試驗設計(DOE)%6σ穩健性優化%響應麵模型
기차안전%측면팽당%경양화%시험설계(DOE)%6σ은건성우화%향응면모형
vehicle safety%side crash%light-mass%design of experiments (DOE)%six sigma robustness optimization%respond surface model
针对某自主品牌多用途汽车(MPV),进行侧面碰撞的轻量化和稳健性优化设计。优化过程中,提出了基于离散设计变量和噪声因素的组合方法。该方法综合了试验设计(DOE)、近似建模、Monte Carlo采样和基于响应面模型的稳健优化技术,考虑了侧面碰撞工艺参数(关键件板厚)和碰撞工况的波动(移动壁障的位置和高度)。进行了3轮优化,分析了其中的灵敏度、确定性和6σ稳健性。结果表明:优化后车身结构的质量减少4.60 kg,侧面碰撞性能的可靠度高于99.97%。因此,该优化方法能满足响应面模型的精度要求。
針對某自主品牌多用途汽車(MPV),進行側麵踫撞的輕量化和穩健性優化設計。優化過程中,提齣瞭基于離散設計變量和譟聲因素的組閤方法。該方法綜閤瞭試驗設計(DOE)、近似建模、Monte Carlo採樣和基于響應麵模型的穩健優化技術,攷慮瞭側麵踫撞工藝參數(關鍵件闆厚)和踫撞工況的波動(移動壁障的位置和高度)。進行瞭3輪優化,分析瞭其中的靈敏度、確定性和6σ穩健性。結果錶明:優化後車身結構的質量減少4.60 kg,側麵踫撞性能的可靠度高于99.97%。因此,該優化方法能滿足響應麵模型的精度要求。
침대모자주품패다용도기차(MPV),진행측면팽당적경양화화은건성우화설계。우화과정중,제출료기우리산설계변량화조성인소적조합방법。해방법종합료시험설계(DOE)、근사건모、Monte Carlo채양화기우향응면모형적은건우화기술,고필료측면팽당공예삼수(관건건판후)화팽당공황적파동(이동벽장적위치화고도)。진행료3륜우화,분석료기중적령민도、학정성화6σ은건성。결과표명:우화후차신결구적질량감소4.60 kg,측면팽당성능적가고도고우99.97%。인차,해우화방법능만족향응면모형적정도요구。
Robustness optimization and light-mass of side crash performance were performed for an owned-brand multi-purpose vehicle (MPV). The optimization method was a combination method based on discrete design variables and noise factors, which including the design of experiments (DOE), the approximation model, the Monte Carlo sampling technique, and the robust optimization based on a respond surface model, and considering with the process parameters of side crash performance (the thicknesses of key parts) and the differences of a crash load case (the position and the height of a movable barrier). A three-stage optimization was used to analyzing the sensitivity, the deterministic, and the 6-sigma robustness for side crash. The results show that the body structure mass is reduced by 4.60 kg, while the reliability of side crash performance is higher than 99.97%, after the optimization. Therefore, the method ensures the accuracy of the approximate model to meet the optimization requirements.