改进的高速铁路安全状态协同监测数据处理方法研究

Study on improved data-processing method of high-speed vehicles’ safety monitoring based on Cooperative-sensing

摘要:

将无线传感器网络协同感知方法,应用于高速铁路固定设施状态监测,具有诸多优势,而协同感知数据的处理方法尤为关键。数据融合方法是协同感知方法在数据处理层面的理想选择。在分布式融合场景下,改进的动态误差平方求权系数法,可以充分利用无线节点的存储能力,使得数据波动性减小、一致性增强；在集中式融合场景下,将求取权系数的相关矩阵法从标量推广至多维向量,使得多维向量数据的融合方法得到丰富。仿真实验表明,改进方法下数据准确性、稳定性均优于原有算法。同时,分析比较了选取不同距离函数指数对多维向量相关矩阵法的影响。

Abstract:

Method based on cooperative-sensing has many advantages on safety monitoring of high-speed vehicles, meanwhile, to handle cooperative-sensing data is crucial. Data fusion is a good way on cooperative-sensing data-processing level. Under distributed circumstances, improved dynamic square-error way on calculating weight-coefficient was proposed. As a result, data volatility is reduced , and the consistency is enhanced. Under concentrated circumstances, correlation matrix method was extended to vector aspect. Finally, simulation analysis testifies that the performance has promoted by new algorithms, and algorithm capability with different exponent of distance function is compared as well.