【作者】 薛俊芳；

【导师】 邱长华；

【作者基本信息】 哈尔滨工程大学 ， 机械设计及理论， 2008， 博士

【摘要】 对机电产品拆卸过程进行规划是实现产品面向拆卸回收设计的有效途径,是评价产品绿色特性的重要手段,是提高产品实际拆卸过程综合效益的必要前提。产品的拆卸过程规划问题已成为产品面向拆卸回收设计中的核心问题。由于拆卸序列规划问题在数学上属于NP-完全型,即随着产品复杂程度的提高,判断出合适的产品拆卸序列是异常困难的。本研究的主要目的,在于探索具备较好的搜索策略,在适当图模型表达方式的基础上提出能够高效率地求解产品拆卸序列的方法。本文主要研究内容包含如下六个方面:(1)建立了拆卸层次信息图(DHIG)模型,将产品拆卸序列的解空间成功映射为可以执行的简单图形空间,即将产品的拆卸序列规划问题转化成为对该加权有向图中具备最优值的路径的搜索和寻优问题。构建产品的DHIG模型的过程,实际上是对产品中有关信息的收集和整理过程,其中,最主要的就是对产品中各零件间优先约束关系的判定,从而确定出DHIG中节点的拆卸可行性。本文设计和开发了基于CAD系统的产品中零件间拆卸优先约束矩阵的自动生成系统。(2)提出一种基于图解的蚁群优化算法(ACO),以实现基于DHIG模型对产品拆卸序列的搜索和寻优。不同于传统的拆卸分析与规划研究,这一求解算法将最优拆卸序列的搜索进程和DHIG模型的构建进程并行起来,为将研究方法应用于解决实际工程问题提供了基础。针对拆卸序列规划过程中节点之间极易产生非平衡竞争的特性,改进了基本ACO中的概率选择方式、信息素更新机制等,使改进后的ACO具有较好的效能。讨论了算法的特点,并和其它智能拆卸序列优化方法进行了比较和分析。(3)对于具有较多零件的产品,采用了基于灰色聚类的子拆卸体划分方法,将产品划分为若干个适当规模的子拆卸体,该方法可有效降低DHIG的复杂性,提高拆卸过程建模和规划的效率。(4)在基于DHIG模型和蚁群优化算法相结合的拆卸过程规划的基础上,针对蚁群搜索到的少数几条拆卸序列(方案),提出了根据信息熵进行权重向量计算的拆卸方案综合评价方法。(5)对基于粗糙集理论的零件合并专家知识获取方法进行了研究,以逐步扩充零件合并的设计知识库。(6)基于所阐述的有关基本理论和方法,设计开发了基于蚁群优化算法的拆卸过程规划原型系统,并对其中的有关CAD系统二次开发的技术进行了研究。更多还原

【Abstract】 The research on disassembly process planning for electromechanical product is a necessary part of product’s design for disassembly & recycling, which can envaluate the product’s green characteristic and enhance the integrative benefit in practical disassembly process. The disassembly process planning has become the core of product’s design for disassembly & recycling.In mathematics, the Disassembly Sequence Planning (DSP) is a NP-Complete problem. With the increasing of complexity of products, it is very difficult to find out proper disassembly sequence solution. The work presented in this proposal is to seek a method which, based on an appropriate graph model, have the better searching strategy and could solve the DSP problem with a high efficiency.The main contents are as following:(1) Disassembly Hierarchy Information Graph (DHIG) is presented. Based on this graph model, the solution space of product’s disassembly sequences is mapped onto an executable and simple graph space. Namely, the DSP problem is transformed into a problem of searching optimum path in the directed and weighted graph. In technology, the process of fulfilling DHIG is a process collecting relevant product’s data & information for disassembly sequences planning. In this process, determination of constraint relations between components is the most important, which is vital to disassembly feasibility of nodes in DHIG. A Disassembly Precedence Restriction Matrix (DPRM) extracting system is designed and developed based on 3 dimension CAD platforms.(2) An Ant Colony Optimization (ACO) algorithm based on the graph is applied to realize the search and optimization for disassembly sequence which is symbolized as a path on DHIG. And different from traditional methods, this method combines the fulfilling process of DHIG with searching process for optimum path, which make a solid base for applying this method into engineering. In the process of disassembly sequence planning, there is always non-balanced competition between nodes. Aiming to this property, transition probabilities and pheromone update rule are improved, which enhance the efficiency of ACO. The characteristics of the algorithm are discussed and the comparisons to other intelligent disassembly optimization algorithms are analyzed.(3) Toward the products which have more parts, the method based on grey-clustering for subassembly identification is presented to effectively reduce the complexity of DHIG and further improve the efficiency of disassembly process modeling and planning.(4) Based on the ACO algorithm associated with DHIG model for disassembly sequences planning, a synthetical evaluating method for little disassembly scheme searched by ant colony is researched, which calculates the weight vectors according to the information entropy.(5) Based on the rough theory, an expert knowledge extraction approach about the component integration is studied, which can extend the design knowledge base step by step.(6) According to the proposed basic approaches, a prototype system of disassembly process planning based on ACO is designed and developed. Meanwhile, several technical issues about CAD system’s secondary development, which involved in the system, are studied.更多还原