基于无线传感器网络的河流中稳态扩散污染源定位研究

The Localization of Pollution Source in Steady Diffusion State in River Based on Wireless Sensor Networks

针对河流中污染源定位问题,首先分析了河流污染源扩散模型,给出了一种处理完全吸收边界,不完全反射边界以及完全反射边界的通用河流污染源稳态扩散模型。改进了在边界约束下以测量值与理论值之差的平方和为目标的非线性最小二乘算法,并提出了一种新的最小二乘污染源定位算法。该算法弥补了直接非线性最小二乘算法在数值计算过程中稳定性较差的缺点。最后,仿真研究了浓度测量噪声,节点漂移误差和反射系数误差对定位性能的影响。仿真结果表明:已知信息的误差越大,则定位均方根误差越大;传感器节点个数越多,估计精度越高,但当节点增加到一定数量时,继续增加节点对定位精度的影响会减小。另外,仿真结果验证了非线性最小二乘算法的优越性,说明了算法在河流污染源定位应用中的有效性。

To solve the localization problem of the pollution source in river,the diffusion process is analyzed first. In steady diffusion state,a general diffusion model is proposed,which can apply to the diffusion process with complete absorption boundary,perfect reflection boundary and imperfect reflection boundary,respectively. A nonlinear least square based localization method is improved with the boundary constrain and the target of minimizing the sum of the square of the differences between the estimation and the measurement of the concentration computing. Then,to improve the numerical stability and robustness,a new objective function of localization is proposed. Simulations are conducted to the localization performances under different levels of measurement noise,node position drift and re￣flection coefficient uncertainties. The simulation results show that the more the related information error is,the larger the root mean square error of the position estimation is,and that the more sensor nodes are,the higher the accurate of position estimation is. However,when the number of nodes is up to a certain value,the localization accuracy is improved hardly. The results validate the superiority of the improved nonlinear least square method and its efficiency in the application of the localization of pollution source in river.