Disassembly Planning is the core of product's disassembly &recycle research. As a NP-Complete problem, with the increasing of product's complexity, it will be very difficult to find out proper solution. The goal of this research is to solve disassembly planning problem with good strategy and efficiency, therefore, to assist engineers in analyzing product's performance on disassembly and recycle.Firstly, Disassembly Feasibility Information Graph (DFIG) is presented, and disassembly planning problem is mapped onto DFIG as an optimal path searching problem. With the characters of complete theory and concise structure, DFIG model could represent all disassembly solutions by simple solutions space, and meanwhile, build the"one to one correspondence"between paths on DFIG and product's disassembly solutions. Moreover, this conversion also offers a general model for solving disassembly planning problem by meta-heuristics。Then, an ACO based hybrid heuristic algorithm for disassembly planning problem is provided. Some unique characters, e.g. the suicide-termination mechanism of ant, and intelligent improvement for DFIG model during ant colony's stochastic search, could make the stochastic search more efficiently. And different from traditional methods, this hybrid algorithm uses a concurrent mechanism between ant colony's stochastic search and DFIG model's constructing, which make a solid base for applying this method into engineering. Considering the fact that disassembly analysis &planning in engineering applications is very dependent on engineers'knowledge and experience, the idea of building disassembly precedence relationships by experienced engineers is proposed. This could enhance constraint information in DFIG and thus, speed up the optimization and ensure final solution more feasible in practice. Based on the method mentioned above, an approach to solve selective disassembly problem also is given.Finally, based on the proposed method, a disassembly analysis &planning system on CAD platform has been developed. In engineering applications, collaboratively work with product R&D group in industrial partner, performance of our proposed method/tool is testified, and more applications in product development are discussed. |