基于不确定性测度的机械零部件再制造加工质量控制理论与方法研究

发布时间：2018-03-28 22:06

本文选题：再制造加工　切入点：不确定性　出处：《合肥工业大学》2015年博士论文

【摘要】：再制造工程是构建循环经济、实现社会可持续发展的最佳实现形式,是二十一世纪应对世界环境污染和资源短缺的有效技术之一。“再制造产品质量不低于、甚至高于原型新品”是再制造工程高速发展的重要保障。但是,当前再制造产品质量存在的问题与不足,已经成为制约再制造产业进一步发展的瓶颈。由于再制造过程自身的特殊性和复杂性,其不确定性要远远高于传统正向制造过程,而这些更高的不确定性最终影响着再制造产品的质量,制约着再制造产品的服役安全性能。对再制造过程不确定性的研究,已成为再制造产业规模化发展亟需解决的关键问题之一。鉴于此,本文通过探索再制造加工过程不确定性内涵及其测度理论,研究不确定环境下机械零部件再制造加工质量控制方法,为保障再制造产品质量提供理论、方法和技术支持,论文的主要研究内容如下：(1)综述再制造基本理论、再制造过程不确定性、再制造质量控制等国内外相关研究现状,分析再制造质量控制研究的不足,阐述本论文研究背景、研究意义和主要研究内容。(2)探究再制造加工系统的不确定特征及其来源,建立再制造加工系统分析模型,对其不确定性进行多层次、多维度解析：选择“熵”作为不确定性测度基本尺度,构建再制造加工系统不确定性概念模型及测度基础模型。(3)研究再制造基础要素的随机不确定性及模糊不确定性测度模型,基于积空间P×M探索测度尺度融合方法,给出定量化描述再制造基础要素整体不确定度的复合熵模型；通过非线性拟合模型,探究再制造基础要素不确定性与再制造加工质量属性之间的耦合机理,并利用再制造曲轴加工实例,验证了方法的有效性。(4)建立基于GERT图的再制造工艺路线模型,考虑加工工艺难度及加工状态多样性要素,构建基于广义信息熵的再制造加工过程不确定性测度模型；定义再制造加工过程控制力指数,实现对再制造加工过程控制能力的评估,以曲轴再制造加工过程为例,对其加工过程控制有效性进行了评估。(5)利用质量熵度量再制造产品质量波动,并以质量波动最小为控制目标,建立再制造质量预测控制系统基本框架模型；利用小波Elman神经网络实施再制造质量建模预测,采用了自适应学习速率进行网络训练,并给出了训练网络的动态递推梯度下降算法；最后将预测模型应用于再制造发动机曲轴质量预测。(6)设计开发了再制造加工质量控制原型系统,构建了系统概念模型,给出了系统总体架构和系统实现的关键使能技术。
[Abstract]:Remanufacturing engineering is the construction of circular economy, the best form to realize the sustainable development of society, the world is twenty-first Century in response to the environmental pollution and resource shortage. "One of the effective techniques for remanufacturing product quality is not less than, even higher than the new prototype" is an important guarantee for development of remanufacturing engineering in high speed. However, the current existence of the remanufacturing product quality the problems and shortcomings, which has become a bottleneck for further development of the remanufacturing industry. Due to the particularity and complexity of manufacturing process itself, the uncertainty is much higher than the traditional manufacturing process, but these higher uncertainty ultimately affects the quality of remanufactured products, restricting the performance of service safety. Study on remanufacturing process uncertainty, has become one of the key issues of remanufacturing industry scale development needs to be solved. In view of this, this article through To explore the manufacturing process uncertainty connotation and measurement theory, study the uncertain environment of mechanical parts remanufacturing process quality control method, in order to provide theoretical support for remanufacturing product quality, support method and technology, the main research contents of this paper are as follows: (1) the basic theory of remanufacturing review, remanufacturing process uncertainty, remanufacturing the quality control of domestic and international relevant research, analysis of remanufacturing research quality control, this paper expounded the research background, research significance and main research contents. (2) to explore the uncertainty characteristics and source of remanufacturing system, establishing the analysis model of remanufacturing system, multi-level of uncertainty, many dimension analysis: entropy as the uncertainty measure of the basic scale, construction of remanufacturing system uncertainty conceptual model and measure model based on business. (3) The random uncertainty made the basic factors and fuzzy uncertainty measure model, product space P * M fusion method based on exploration measure scale, given the quantitative description of remanufacturing based on elements of the whole uncertainty model of composite entropy; through nonlinear fitting model, explore the remanufacturing coupling mechanism between basic elements of uncertainty and re manufacturing quality the attribute, and the use of remanufactured crankshaft machining example, verify the effectiveness of the method. (4) based on the GERT diagram of the remanufacturing process routings model, considering the difficulty of processing and processing state diversity elements, construction process and manufacture of generalized information entropy uncertainty measurement model based on the definition of manufacturing process control; force index, evaluation of remanufacturing process control ability, the crankshaft remanufacturing process as an example, the process control is effective in the assessment Estimated. (5) to measure the remanufacturing quality fluctuation by quality entropy, and to the quality fluctuation as the control target, the establishment of quality control system of the basic framework of prediction model; implementation of remanufacturing quality prediction model based on wavelet Elman neural network, the adaptive learning rate training network, and gives the dynamic recursive network training push the gradient descent algorithm; prediction model is applied to the engine crankshaft remanufacturing quality prediction. (6) the design and development of remanufacturing process quality control system, constructs the concept model of the system, given the key to achieve the system overall architecture and enabling technologies.

【学位授予单位】：合肥工业大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TH16

【相似文献】