The Mechanism Research Of Visualization Experiment Approach For Solving System Global Shortest Path

System global shortest path is one of the classic nonlinear combinatorial optimization problems. The minimum Steiner tree theory is the basis theory of shortest path, and designing effective algorithms for it has important theoretical significance and wide application value. A variety of heuristic algorithms, which are used in solving the minimum Steiner tree problem are easy to fall into local optimum. A visualization experiment approach, which used in solving system global shortest path, succeeds in generating Steiner points automatically. It has made up for the SteinLib, in which Steiner points are not included in the objective function. However, it takes a long time to form membrane path, and sometimes it is difficult to forming stabilized system shortest path when the number of given points are increased and irregular distribution. The promotion of visualization experiment approach in engineering is restricted by this bottleneck. On the foundation of visualization experiment, the research mechanism of visualization experiment approach for solving system global shortest path will be further explored in this article. In view of the instability and inaccuracy of experiment approach, Experiment - Geometry Algorithm (EGA) will be constructed to prepare for improving and perfecting visualization experiment approach. It is used to satisfy project actual needs better.Based on the physical visualization experiment ,we use experiment approach - algorithm - experiment approach - project application as our research route , analysis and research visualization experiment unit, experiment process, as well as experiment results detailed, explore the relationship between construction of the Steiner minimal tree and the distributed shape of given points, as well as the relationship between the location and the number of Steiner points and the distributed shape of given points, use the Delaunay triangular net's basic property, take advantage of the GeoSteiner algorithm's thought as well as Melzak algorithm thought, propose a new algorithm for the Steiner minimal tree problem-EGA. We aim to find an algorithm in this paper, which can be used with visualization experiment unit, and can make up for the experiment defect that it is difficult to form membrane when given points are increased and distributed irregularly.In this paper, simple graphical examples have demonstrate the feasibility of EGA, and confirmed the approximate proportion of EGA and the visualization experiment is more than 90%, the usability and the validity of EGA are proved, too. Massive examples proved that EGA can solve the project problem with the visualization experiment unit, and lay the foundation for the next step that the improvement and consummation of visualization experiment approach.