基于连接模型的局部优化算法在水域污染监测数据融合系统中的应用

Connectionist model based local optimization algorithm for large-scale water pollution monitoring data fusion systems

针对水域污染监测数据融合系统中存在的困难 ,讨论了基于“连接”模型的局部优化算法及其应用 .该模型采用非“抑制”连接 ,极大地减少了节点“连接”数和扇出数 ;各个节点只和相邻节点通过“连接”传递信息 ,竞争输出 ,保证了局部最优 ,同时为实现分布式计算提供了方便 .在此模型的基础上本文用局部优化及其改进算法对一个水域污染监测问题进行了仿真研究 ,理论分析和计算结果表明 ,局部最优及其改进算法在保证搜索准确性的同时极大地减少了计算量 ,是解决水域污染监测问题的有力工具 .

Aiming at the difficulties existing in large-scale water pollution monitoring systems, a connectionist model based local optimization algorithm and its application are discussed in this paper. With just the excitatory connections the connectionist model drastically reduced the storage for links and the fanouts of the nodes. Based on the competitive activation mechanism, the local optimization algorithm and its improvement-partial resettling algorithm, realize the dynamically changing functional relationships between disorders and appropriate multiple-winners-take-all behavior. As an illustrative example, the connectionist model is introduced to the water pollution monitoring data fusion system. Computer simulation results show that the local optimization algorithm and the partial resettling algorithm greatly save the computation time, as well as ensure that the most probable disorders can be founded.