分布式协同解耦进化算法在无线传感器网络覆盖与数据传输的研究

The computation time and efficiency when optimizing the coverage and data transmission problems in large scale wireless sensor networks are the bottleneck in the application of wireless sensor networks. Combining the advantages of evolutionary computation algorithms on global optimization and the research of decoupling techniques for large scale optimization problems, this project proposes a novel cooperative decoupling evolutionary computation algorithm. By analyzing the population distribution and the statistics of fitness values, and the relations among the candidated variables, the algorithm will decouple the variables in the considered problem, thus dividing the complex large scale optimization into several smaller problems. It is beneficial for searching local regions in a fast speed and guiding the algorithm to perform global search to the solution space. Based on the analysis to the feature of dividing the optimization scale for the algorithm, the project will propose an adaptive parallel implementation of the proposed cooperative decoupling evolutionary computation algorithm in the distributed environment for further enhancing the optimization speed and efficiency. This project will eventually present a new way of solving the coverage and data transmission problems in wireless sensor networks: a distributed parallel cooperative decoupling evolutionary computation algorithm, for meeting the need of no blind converage region, information transmissibility, and energy conservation, and provide systemetic test. Innovations are expected to be made in the theoretical research and application simulation.

大规模无线传感器网络覆盖与数据传输优化中的计算时间和效率问题是制约无线传感器网络应用发展的重要瓶颈。结合进化计算算法在全局搜索方面的优势以及对大规模优化问题解耦技术的研究，本课题提出一种新型的协同解耦进化计算算法，该算法依据优化过程中对种群的分布和适应值的统计结果以及对待优化问题变量关系的分析，把待解问题的变量进行解耦，从而把复杂的大规模优化问题分割为多个较小的问题协同优化，利于快速搜索局部区域并引导算法对解空间的全局搜索。通过对算法在优化规模分割特征上的研究，制定在分布式环境下实现协同解耦进化计算算法的自适应并行实现方式，进一步提高算法优化的速度和效率。本课题最终将给出一个求解无线传感器网络覆盖与数据路由传输优化问题的新途径：分布式协同解耦进化计算算法，满足覆盖无盲区、信息可传输性和节能的要求，并进行系统测试验证。预期将在理论研究和应用仿真中取得创新成果。

大规模无线传感器网络覆盖与数据传输优化中的计算时间和效率问题是制约无线传感器网络应用发展的重要瓶颈。结合进化计算算法在全局搜索方面的优势以及对大规模优化问题解耦技术的研究，本课题提出一种新型的协同解耦进化计算算法，该算法依据优化过程中对种群的分布和适应值的统计结果以及对待优化问题变量关系的分析，把待解问题的变量进行解耦，从而把复杂的大规模优化问题分割为多个较小的问题协同优化，利于快速搜索局部区域并引导算法对解空间的全局搜索。本课题已经完成了进化计算算法的协同解耦优化策略的研究并发表了2篇论文。完成了分布式协同解耦进化计算算法自适应并行实现的研究并发表3篇论文。完成了在无线传感器网络中应用进化计算算法的协同解耦优化策略的研究，给出一个求解无线传感器网络覆盖与数据路由传输优化问题的新途径：分布式协同解耦进化计算算法，满足覆盖无盲区、 信息可传输性和节能的要求并发表了5篇论文。