载荷谱编制中极值载荷确定方法及应用

发布时间：2018-12-14 22:26

【摘要】：载荷谱是对零部件进行疲劳试验的基础，它的真实性直接影响零部件疲劳寿命预测的准确度。目前已有一些领域（如汽车、风力发电、航空航天等）的载荷谱编制方法已经发展的比较成熟，并形成了相应的行业标准和专门的通用处理软件。工程机械领域的载荷谱编制工作起步晚，其理论计算结果与试验测试结果之间仍存在较大偏差。为研究以装载机为代表的工程车辆耐久服役性能的室内快速评价、理论预测方法，本文结合国家自然基金资助项目（基金号：50805065），在借鉴参考国内外相关研究成果的基础上，对载荷谱编制的关键技术—分布函数的选择与极值载荷的确定方法进行了探讨，主要的研究内容和结论分述如下：（1）探索了建立混合分布模型的通用方法。根据装载机的循环作业特点与每一作业段的实际载荷分布特性，提出了基于混合分布理论建立最佳分布函数的方法。建立了多备选方案多目标属性的评估矩阵，通过MTOPSIS优化方法求解出不同准则的权重系数，然后利用多准则决策技术对多目标属性进行量化处理，并利用优化方法确定出混合分布模型中基本函数的个数。（2）建立了极值载荷最佳分布模型，，给出了外推的极值载荷近似表达式。考虑载荷样本的随机性，并结合统计学理论，建立了极值载荷的统计学模型，对比分析了不同类型分布函数的外推风险值（Value at Risk），筛选出最佳的拟合函数，基于POT理论，在阀值已知的基础上，利用反推法确定出极值载荷的经验计算公式。（3）提出了阀值与分布参数的改进算法。综述了目前不同领域常用的阀值确定方法理论基础和应用范围，分析了阀值高低对偏差、方差之间的影响规律，推导出基于自助采样法（Bootstrap）求解阀值的过程，针对工程车辆载荷随机性、不可重复性与难预测性等特点，提出利用贝叶斯理论将统计分布函数的参数视为随机变量，以实测数据为基础及时对原参数进行更新修正，通过对比修正前后的拟合效果，验证了考虑贝叶斯理论后计算结果的合理性。（4）以轮式装载机变速箱输出法兰为例进行极值载荷分布特性分析，并验证本文提出理论方法的有效性与准确性。对比分析了三种不同工况下极值分布的离散程度、阀值的差异和极值载荷的大小，并对这种现象产生的原因进行了细致的分析。本文系统的阐述了工程车辆极值载荷确定方法的基本原理及具体实现过程，为形成系统的工程车辆载荷谱编制理论、制定关键零部件室内快速检测方法和标准提供基础理论具有重要的学术意义与实际应用价值。
[Abstract]:Load spectrum is the basis of fatigue test of parts, and its authenticity directly affects the accuracy of fatigue life prediction. At present, there are some fields (such as automobile, wind power, aerospace, etc.) load spectrum compilation method has been developed more mature, and formed the corresponding industry standards and specialized general processing software. The compilation of load spectrum in the field of construction machinery started late, and there is still a big deviation between the theoretical calculation results and the test results. In order to study the indoor rapid evaluation of durable service performance of engineering vehicles represented by loaders, the theoretical prediction method is used in this paper, combined with the project funded by the National Fund for Nature (fund number: 50805065), on the basis of referring to the relevant research results at home and abroad. The selection of distribution function and the method of determining extreme load are discussed in this paper. The main research contents and conclusions are as follows: (1) the general method of establishing mixed distribution model is explored. According to the characteristics of the Loader's cyclic operation and the actual load distribution of each working section, a method of establishing the optimal distribution function based on the mixed distribution theory is proposed. The evaluation matrix of multi-objective attribute is established, and the weight coefficients of different criteria are calculated by MTOPSIS optimization method. Then, the multi-objective attribute is quantified by multi-criteria decision technology. The number of basic functions in the mixed distribution model is determined by the optimization method. (2) the optimal distribution model of extreme load is established, and the approximate expression of extrapolated extreme load is given. Considering the randomness of load sample and combining with statistical theory, the statistical model of extreme load is established, and the extrapolation risk value (Value at Risk), of different distribution function is compared and analyzed. The best fitting function is obtained based on POT theory. On the basis of the known threshold value, the empirical formula of extreme load is determined by using the backstepping method. (3) an improved algorithm of threshold and distribution parameters is proposed. This paper summarizes the theoretical basis and application range of threshold determination methods in different fields, analyzes the influence of threshold values on deviations and variances, and deduces the process of solving threshold values based on self-help sampling method (Bootstrap). In view of the randomness, unrepeatability and unpredictability of the loads of engineering vehicles, the Bayesian theory is used to treat the parameters of the statistical distribution function as random variables, and the original parameters are updated and revised timely on the basis of the measured data. By comparing the fitting results before and after the correction, the rationality of the calculation results after considering Bayesian theory is verified. (4) taking the gearbox output flange of wheel loader as an example, the extreme load distribution characteristics are analyzed, and the validity and accuracy of the theoretical method proposed in this paper are verified. The discreteness degree of extreme value distribution, the difference of threshold value and the magnitude of extreme load under three different working conditions are compared and analyzed, and the causes of this phenomenon are analyzed in detail. In this paper, the basic principle and realization process of the method for determining the extreme load of engineering vehicles are systematically expounded, so as to form the theory of compiling the load spectrum of engineering vehicles. It is of great academic significance and practical application value to establish the basic theory of indoor rapid inspection method and standard for key parts.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH123

【参考文献】