IFS And Genetic Algorithm Studying In Simulation Of Plant

Fractal’s main application area is the natural landscape simulation, plantsimulation is the main branch of the natural landscape simulation, plant has a typicalcomplex self-similar structure in nature. On the basis of fractal graphics, we canexplore the decipline of plant growth by simulating the natural scenery. This paperintroduces the basis of fractal, and describes the methods of fractal graphics: IFS,DLA, ETA and L-System. The plant graph generated by IFS has a strong sense oftexture, rich in texture, shading and so on. the program that can generate the plantgraph by IFS is simple enough. But the plant simulation of traditional IFS isdifferent from the process of plants’ natural evolution.Little difference between the morphology of the plant graphics generated by theIFS, and the morphology is diversificated even the same plant. The IFS withprobability and the IFS with parameters are simulating the plant by the geometricfeatures, but lack of the description of the plant’s evolution and physiologicalcharacteristics. Although the method of generating plant graph by IFS is simple, thesensible way to caculate the IFS code is complex. Therefore, we simulate the plantmorphology by combine the IFS and GA, considering the IFS code as chromosomeof GA and the affine transformation of IFS code as gene.In such conditions, we canselect and cross the gene with genetic operation, and evaluate the generated plantgraph by siting the sensible fitness funtion. In the simulation part, it shows thegenetic process and evolution of natural plants. In this paper, we can not onlysimulate the plant graph which is has the characteristic of fractal, but also conformthe natural law.The generated plant graphics by GA and IFS combination are not only with richmorphology, but also reflect the effects of genetic and environmental factors duringthe process of plant evolution. Then it can perform the reproduction and evolution ofplants.This method is certainly valuable for reproduction of natural landscapes,virtual animation, botany etc. Finally we prove this algorithm’s feasibility with someexperiments of plants simulation.