针对建筑火灾固有的时空混沌特征，基于无线传感网络构建火灾燃烧过程模型，对研究建筑火灾燃烧机理，从而快速、准确地模拟和预测火势蔓延和发展，为及时控制、扑灭火灾提供科学决策,具有重大的理论与实际意义。本课题将无线传感网络与耦合映像格子模型有机结合，研究基于超小波的时变映射函数和表征其复杂时变空间耦合关系的理论与方法，并探索获取时变映射函数最佳系数的优化方法；在此基础上，建立具有时变和普适性的多维无线传感网络耦合映像格子火灾模型及相关网络分布式算法；通过有效提取其时空混沌特性，探索在不同环境条件下火灾模型运行的非线性动力学特征参数及其演化规律，为火灾演化过程提供新的定性、定量评价标准和预测信息；同时还将基于自组织神经网络与支持向量机结合实现对火灾多维信号的分类识别。该项目的研究不仅进一步丰富与完善有关耦合映像格子模型的理论与方法，并为无线传感网络的研究及其应用探索新的途径。

In order to timely control the situation of building fire, as well as to reduce the damage, an appropriate model of fire spread and growth gives scientifically accurate simulation, and is considered to be of great significance. In the case, the model based on inherent spatiotemporal chaos characteristics is thought to be extraordinarily important.This project properly combines wireless sensor networks and coupled map lattice model to theoretically study time-varying mapping function according to beyond wavelet, and the complex spatial coupling relationship concerned. By working on the performance of mapping function, together with the approach to multivariate function, the optimization method to obtain the best coefficient is established. Whereat, the universal multi-dimensional coupled map lattice fire model of time-varying characteristics is built.The model in this project will be applied on nonlinear spatiotemporal evolution to analyze the situation of building fire. By effectively focusing on its spatiotemporal chaos characteristics,nonlinear dynamical characteristics of parameters and their evolution running within different environmental conditions of fire model is determined. This functionally provides quantitative evaluation and prediction information of fire spreading process. Meanwhile, combination of self-organizing neural network and support vector machine materializes classification and identification of fire multidimensional signal. This study enriches theory and provides methods of coupled map lattice model，moreover, it gives a new way for the research of wireless sensor networks.