Research On Computer-Aided Creative Design Method Based On Planar Mechanism Path Synthesis

In the information era with rapid development of science and technology, new vital force is being brought to traditional subjects by information technology, different subjects keep on supplementing, crossing over and infiltrating mutually, making their scopes of theoretical research and fields of application expanded continuously. In order to keep up with the revolution in information industry and the development of other subjects, traditional mechanics is going through profound changes in many aspects like research objects, contents, theories and methods, among which the Mechanical Creative Design (MCD) is a fairly active field. Using the relevant theoretical research achievements for reference, the path synthesis and creative design theory of planar mechanism is deeply researched in this dissertation, in which new methods of computational intelligence and creative design theory are introduced and a new approach to the theoretical research on path synthesis and creative design of planar mechanism is presented. By introducing Artificial Immune System (AIS) algorithms into the path synthesis of planar mechanism, the current direct synthesis methods and indirect synthesis ones are improved. In the former one, revised artificial immune network algorithm (ai-Net) for multimodal optimization is adopted to solve the problem of the similarity between candidate solutions, which is ubiquitous in current optimization methods. By means of the new method with good global convergence and robustness, a large number of candidate solutions with large deversity can be obtained; In the latter one, the excellent performance of immune-computing algorithms in pattern recognition is used to develop a new indirect path synthesis framework based on immune-computing, in which several different algorithms and models are applied to the clustering, classification and retrieval of path curves. Referring the creative strategy of Yan's mechanism creative design method and applying the path synthesis methods mentioned above to mechanism design, a path-feature based planar mechanism creative design method is proposed. Furthermore, two important supporting technologies, viz. the isomorphic identification and the automatic sketching methods for mechanism kinematic chain, are discussed in detail. The former one can be transformed into a degree-reducible Traveling Salesman Problem (TSP) firstly, and then ant algorithm and immune algorithm are adopted to solve it easily. The latter one, which can be seen as the arrangement problem of the basic loops in the graph of mechanism kinematic chain in essence, can be solved by a new approach to automatic sketching of mechanism kinematic chain presented, in which graph theory is introduced to generate the basic loops and their proper order. In addition, how to generate mechanical kinematic scheme according to qualitatively described path features is also studied in this dissertation. In the new approach proposed to kinematic scheme generation based on multi-granularity co-evolutionary function reasoning, classified functions are used to describe the kinematic characters of mechanical units, each mechanism and mechanical system is treated as a motion-transforming function, ground on which series-connected mechanical kinematic schemes are generated by means of function reasoning method. Using object-oriented programming language, a prototype system for computer-aided mechanism path creative design is developed, by means of which the design of the shedding mechanism of simple loom is successfully accomplished. The case study indicates that the proposed methods are feasible and effective.