| 湿式离合器出油口甩出油温度预测方法 |
| |
| 引用本文： | 鲍伟,曹将.湿式离合器出油口甩出油温度预测方法[J].中国公路学报,2019,32(5):162-171. |
| |
| 作者姓名： | 鲍伟 曹将 |
| |
| 作者单位： | 合肥工业大学 电气与自动化工程学院, 安徽 合肥 230009 |
| |
| 基金项目： | 国家自然科学基金项目（51405122）；中央高校基本科研业务费专项资金项目（JZ2016YYPY0035） |
| |
| 摘 要： | 为了实现对湿式离合器出油口甩出油温度传感器的冗余校验和自我诊断,提出了一种基于粒计算约简的离合器出油口甩出油温度的模糊预测方法。首先分析出油口甩出油温度的影响因素,将主要影响因素作为预测输入量,并采用模糊推理理论预测当前离合器出油口甩出油温度。在设计模糊预测方法的过程中,通过分析实车数据得到车辆行驶时离合器处于高滑摩功率过程和低滑摩功率过程的不同特性,分别确定相对应的隶属度函数和模糊预测规则,从而进一步提高出油口甩出油温的预测精度。为了提高模糊预测算法的实时性,基于模糊预测规则创建模糊决策表,模糊输入量和模糊输出量分别作为决策表的条件属性集与决策属性集。利用粒计算理论对模糊决策表的条件属性集进行属性约简,通过削减冗余信息有效降低模糊输入量和模糊预测规则的个数。最后利用实车采集的数据对比分析约简前后模糊预测算法的单步运行时间和预测误差等性能指标。试验结果表明:基于粒计算约简的模糊预测算法能够有效保障预测精度,同时拥有更少的模糊预测规则数和模糊输入量,有效解决了模糊预测算法占用资源较多以及实用性较差的问题。
|
| 关 键 词： | 汽车工程 湿式双离合器 模糊推理理论 出油口甩出油温 粒计算 属性约简 |
| 收稿时间： | 2018-03-21 |
Predicting Oil Temperature at Wet Dual Clutch Oil Outlet |
| |
| BAO Wei,CAO Jiang.Predicting Oil Temperature at Wet Dual Clutch Oil Outlet[J].China Journal of Highway and Transport,2019,32(5):162-171. |
| |
| Authors: | BAO Wei CAO Jiang |
| |
| Affiliation: | School of Electrical Engineering and Automation, Hefei University of Technology, Hefei 230009, Anhui, China |
| |
| Abstract: | In order to realize the redundancy check and self diagnosis of the oil temperature sensor at the wet dual clutch oil outlet, a fuzzy prediction method based on granular computing reduction was proposed. First, the factors that influence oil temperature at the wet dual clutch oil outlet were analyzed. These factors were then used as input for fuzzy reasoning to predict oil temperature. Different clutch characteristics under high and low friction power were obtained through analyzing actual vehicle data. This was utilized to determine the membership function and fuzzy prediction rule to enhance prediction accuracy. To improve real-time performance of the algorithm, fuzzy rules were applied to create a decision table, using fuzzy input and output quantities as conditional and decision attribute sets, respectively. Granular computing was employed to reduce the conditional attribute set of the fuzzy decision table, which effectively reduced the number of inputs and prediction rules. Finally, the fuzzy prediction algorithm's performance before and after reduction was analyzed using data collected from the vehicle. The results show that the fuzzy prediction algorithm that can effectively guarantee prediction accuracy based on granular computing reduction has fewer fuzzy prediction rules and inputs than the ordinary fuzzy prediction algorithm. This resolves the constraint of ordinary fuzzy prediction algorithm being too resource intensive for practical application. |
| |
| Keywords: | automotive engineering wet dual clutch fuzzy reasoning theory oil temperature at oil outlet granular computing attribute reduction |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《中国公路学报》浏览原始摘要信息 |
|
点击此处可从《中国公路学报》下载全文 |
|
