随着对GIS中的空间对象模型和自然地理特征表达的研究深入,模糊空间对象被提出。针对模糊空间对象表达的特点,提出了一种基于模糊神经网络的模糊空间对象生成方法。该方法将模糊技术与神经网络相结合,利用神经网络的学习能力调整模糊隶属函数和模糊规则,使系统具备自适应的特性。实验表明,这种基于模糊神经网络的生成模糊空间对象的方法比传统方法大大的提高了成果的精度。
With the deep research on the spatial objects model of GIS and the representation of natural geographical feature, this paper put forward the fuzzy spatial objects. Referring to the characteristics of the representation of fuzzy spatial objects, it also brought forward a generation method of fuzzy spatial objects based on fuzzy neural networks. By combining the fuzzy technology and neural networks, utilizing the learning ability to enhance the fuzzy Membership Function and fuzzy rules, the system will be self - adaptive. From the experiments of this paper, by comparing with the traditional fuzzy objects generation, the method in this paper improved the accuracy of results.