[关键词]
[摘要]
通过把自适应线性神经元(adaline)网络与自由曲面的生成原理相结合,提出了一种生成分形曲面的新方法.给出了对自由曲面分形的各种分形方法的数学模型,详细介绍了如何通过设置神经网络可调参数的数值来控制和调整分形曲面形状的方法,实现了在控制神经网络可调参数的情况下,改变确定自由曲面各型值点的线性组合关系,生成可预测、可控制和可调整的分形曲面.
[Key word]
[Abstract]
In this paper, a new method of generating fractal surface is presented by combining adaptive linear neural networks with the principle of generating free surface. The mathematical model of various fractal methods on free surface and the methods of controlling and adjusting the fractal surface shape according to different parameters of neural networks are put forward. Thus, the predictable, controllable and adjustable fractal surface can be generated with changing linear combination relation of all control points of the definite free surface.
[中图分类号]
[基金项目]
国家自然科学基金资助项目(69873038);国家杰出青年科学基金资助项目(69425005);教育部骨干教师资助项目;浙江省教育厅科研基金资助项目(20000031)
