摘要
随着汽车的普及和人们生活质量的提高，消费者对汽车热舒适性的要求也越来越高。车内热舒适性作为汽车舒适性一个很重要的方面，已经成为各大汽车厂家和消费者关注的焦点以及研究的热点。车内热环境不同于建筑室内热环境，其空间狭小，空气流动性差，汽车的移动性使汽车受到的太阳辐射随时发生变化，进而会对车内热环境产生影响。汽车车内人体的热舒适性不仅跟所处热环境有关，还跟人体的主观感受有关，且人体不同部位对热环境的敏感程度也不一样。因此，需要研究适合于汽车乘员舱的热舒适评价模型及试验方法。
本文参考国内外在汽车热舒适性方面取得的一些研究结果，结合现有的试验条件，开展人体热舒适性试验。通过对夏季暴晒和空调制冷阶段的某款SUV车内热环境进行测定，得到车内空气温度、相对湿度和平均辐射温度等参数的变化情况；并设计了人体热舒适性主观问卷调查，获得制冷期间人体的主观热舒适感受。通过对数据进行处理、分析，研究车内热环境的特点及人体整体和局部热舒适性。并采用PMV-PPD和当量温度T eq评价指标对人体的整体和局部热舒适性进行客观评价。
利用支持向量机建立了基于客观热环境参数的人体局部热舒适性预测模型。对试验数据进行处理，选定训练集和测试集，在完成样本数据归一化处理后，选取径向基核函数为预测核函数，通过网格寻优方法对核函数参数进行参数寻优，以空气温度、相对湿度、平均辐射温度和人体新陈代谢量为自变量，人体
局部热舒适性为因变量，分别建立人体头部、躯干、上肢和下肢的人体热舒适性预测模型。预测结果表明，支持向量机方法可以用于人体局部热舒适预测。
研究人体局部热舒适性和整体舒适度之间的关系。利用支持向量机分类的思想，对试验样本进行训练集和测试集分类、归一化、核函数选定、参数寻优等处理，任取三个局部热感建立整体舒适度的预测模型，然后使用所有人体的局部热感觉对整体舒适度进行预测。从预测结果可知，使用所有人体的热感觉能够更好的预测人体的整体舒适度，且使用所有人体建立预测模型的分类准确率都在85%以上，满足工程应用，说明了支持向量机方法的适用性。
关键词：汽车，热舒适性，主观评价，支持向量机，预测模型
ABSTRACT
With the popularization of automobiles and the improvement of people's quality of life, consumers are increasingly demanding the thermal comfort of automobiles. As an important aspect of automobile comfort, thermal comfort has become the focus and research focus of automobile manufacturers and consumers. The thermal environment inside the vehicle is different from the thermal environment in the building, its space is narrow, the air fluidity is poor, and the moving property of the vehicle makes the solar radiation of the vehicle change at any time, which will have a
n impact on the thermal environment inside the vehicle. The thermal comfort of the car body is not only related to the thermal environment, but also to the subjective feeling of the human body, and the sensitivity of different parts of the body to the thermal environment is different. Therefore, it is necessary to study the thermal comfort evaluation model and test method suitable for the vehicle passenger compartment.
This article refers to some domestic and foreign research results in the automotive thermal comfort, combined with the existing test conditions, to carry out human thermal comfort test. Through the measurement of the thermal environment of a certain SUV vehicle during the summer exposure and air-conditioning cooling stages, the changes in the parameters such as the air temperature, relative humidity, and average radiation temperature in the vehicle were obtained; and a subjective questionnaire survey of human thermal comfort was taken to obtain the subjective thermal comfort of the human body during the cooling period. Through processing and analysis of data, the characteristics of the thermal environment in the vehicle and the overall and local thermal comfort of the human body are studied. Using the PMV-PPD and the equivalent temperature evaluation index to objectively evaluate the overall and local thermal comfort of the human body.The prediction model of local thermal comfort of human body based on objective thermal environment parameters is establ
ished by support vector machine (SVM). The test data is processed, and the training set and test set are selected. After normalizing the sample data, the radial basis kernel function is selected as the prediction kernel function, and the parameters of the kernel function parameters are optimized through the grid optimization method. Taking the air temperature, relative humidity, average radiation temperature and body metabolism as independent variables, and the local thermal comfort as the dependent variable, human thermal comfort
prediction models for the human head, trunk, upper limbs, and lower limbs were established. The prediction results show that the support vector machine method can be used to predict the local thermal comfort of the human body.
Study the relationship between local thermal comfort and overall comfort. Using the concept of support vector machine classification, the training samples and test set classification, normalization, selection of kernel functions, parameter optimization, etc. were performed on the test samples, and any three local thermal sensations were used to establish the overall comfort prediction model.Then use all the human body's local thermal sensation to predict overall comfort. From the prediction results, it can be seen that the use of thermal sensations from all parts of the human body can better predict the overall comfort of the human body. The accuracy of classification using all human body b
uilding prediction models is above 85%, which satisfies engineering applications and illustrates the applicability of the support vector machine approach.汽车辐射
Keywords:Automobile, Thermal comfort, Subjective evaluation, Support vector machine, Prediction model
目录
中文摘要.......................................................................................................................................... I 英文摘要........................................................................................................................................ II 1 绪论.. (1)
1.1研究背景及意义 (1)
1.2国内外研究现状 (1)
1.2.1 车内人体热舒适性试验研究 (2)
1.2.2 人体热舒适性评价方法 (4)
1.3课题研究目的及意义 (5)
1.4课题研究的主要内容 (6)
2 车内热环境分布测试与人体热舒适性主观评价 (8)
2.1车内热环境分布测试 (8)
2.1.1 测试系统介绍 (9)
2.1.2 试验方法 (12)
2.2人体热舒适性主观评价 (13)
2.2.1 试验人员 (13)
2.2.2 人体热舒适问卷调查 (13)
2.2.3 试验方法 (15)
2.3车内热环境测试结果与分析 (15)
2.3.1 车内壁面温度分析 (15)
2.3.2 车内人体局部周围温度分析 (18)
2.3.3 车内黑球温度、湿度、空气温度和太阳照度分析 (19)
2.4车内人体热舒适性分析 (21)
2.4.1 局部热舒适性分析 (21)
2.4.2 整体舒适度分析 (24)
2.5本章小结 (24)
3 车内人体热舒适性客观评价模型与分析 (26)
3.1PMV-PPD评价指标及计算结果分析 (26)
3.2EQT评价指标及计算结果分析 (28)
3.3本章小结 (31)
4 基于SVM的人体局部主观热舒适性评价模型的预测 (32)
4.1支持向量机方法 (32)
4.1.1 支持向量机 (32)
4.1.2 核函数及参数 (34)
4.2试验数据预处理 (35)
4.2.1 归一化 (35)
4.2.2 训练集和测试集选取 (35)
4.2.3 核函数参数寻优 (35)
4.3局部热舒适性预测模型的建立及分析 (38)
4.3.1 局部热舒适性预测结果分析 (39)
4.3.2 局部热舒适性预测模型误差分析 (41)
4.4本章小结 (42)
5 基于局部主观热舒适性评价的整体舒适度预测 (43)
5.1部分人体部位热感觉对整体舒适度的预测 (43)
5.1.1 数据预处理 (43)
5.1.2 部分人体部位热感觉对整体舒适度的预测模型建立与分析 (44)
5.1.3 模型分析 (47)
5.2所有人体局部热感觉对整体舒适度的预测 (47)
5.2.1 数据预处理 (48)
5.2.2 所有人体局部热感觉对整体舒适度的预测模型建立与分析 (48)
5.3本章小结 (49)
6 总结和展望 (50)
6.1总结 (50)
6.2后续展望 (51)
致谢 (52)
参考文献 (53)
附录：A.作者在攻读学位期间发表的论文目录 (57)
B.作者在攻读学位期间参加的科研项目 (57)