完全自适应第二代小波及在循环冲击类故障特征提取中的应用

项目名称： 完全自适应第二代小波及在循环冲击类故障特征提取中的应用

项目编号： No.51205130

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 机械工程学科

项目作者： 张龙

作者单位： 华东交通大学

项目金额： 25万元

中文摘要： 从振动信号中提取周期性冲击响应成分并确定其强度和发生频率是诊断循环冲击类故障的关键。第二代小波变换具有运算效率高、小波基函数构造灵活等优点，在机械故障诊断领域具有良好的应用前景。目前已有的自适应第二代小波方法不能实现提升算子（包括预测算子和更新算子）阶数和系数的同步自适应设计，限制了小波基函数自适应设计的自由度。本项目以循环冲击类故障特征提取为目标，研究完全自适应第二代小波理论和方法，实现提升算子阶数和系数的同步自适应设计，改善自适应第二代小波在循环冲击类故障特征提取中的应用效果。项目将在第二代小波完全自适应设计理论和循环冲击类故障特征优化指标等方面开展研究，并拟提出一种可变染色体长度协同进化算法用于最终实现第二代小波的完全自适应设计。项目研究可充分增强第二代小波在循环冲击类故障特征提取中的优势，提高故障诊断的精准度和早期诊断能力，对小波理论与故障诊断技术的发展均具有积极意义。

中文关键词： 第二代小波；提升小波；自适应小波；故障诊断；遗传算法

英文摘要： The key to detect faults featured by cyclic impacts is to find the impact responses and determine their intensity and frequency by processing collected vibrations. The second generation wavelet has the advantages of high computational efficiency and flexible construction of wavelet basis, thus with a promising prospect in the field of machinery fault diagnosis. The available adaptive frameworks of the second generation wavelet are unable to design the order and coefficients of lifting operators (including the prediction operator and update operator) synchronously, limiting the degree of freedom for wavelet basis design. Aiming to extract fault features of cyclic impacts, a fully adaptive wavelet transform based on lifting scheme is to be explored, which is able to achieve synchronous design of the orders and coefficients of the lifting operators, with the purpose of improving the feature extraction accuracy. Specifically, the philosophy of the fully adaptive second generation wavelet and optimization indicators for evaluating the extracted features will be investigated. Furthermore, a co-evolutionary algorithm with variable chromosome length will be proposed for the final implementation of the fully adaptive design of basis function in the context of the second generation wavelet. The research achievements will

英文关键词： Second generation wavelet；Lifting wavelet；Adaptive wavelet；Fautl diagnosis；Genetic algorithm