基于奇异值分解的炉膛三维温度场声学重建仿真研究

Research on Simulation of Furnace Three-Dimensional Temperature Field Reconstruction by Acoustics Based on Singular Value Decomposition

为了实现炉内温度场的实时在线监测,阐述了基于声学理论的三维温度场重建原理,介绍了奇异值分解(singular value decomposition,SVD)算法。采用最小二乘法和SVD算法分别对炉膛火焰分布的几种典型模型:单峰模型、双峰模型以及四峰模型,进行了仿真重建,并对两种算法的仿真结果做了比较。仿真得到了稳定的重建结果,给出了不同温度场分布模型及其重建图像,分析了重建误差。仿真结果表明,两种算法均可用于炉膛三维温度场重建,并且具有一定的抗干扰能力,与最小二乘法相比,奇异值分解算法具有更好的稳定性。

In order to realize the real-time monitoring of temperature field in furnace, the principle of three-dimensional temperature field reconstruction based on acoustic theory was analyzed. Singular value decomposition(SVD) algorithm was introduced. Three typical temperature field models, such as one-peak model, two-peak model and four-peak model, were simulated with only a few acoustic data by least squares method and SVD algorithm respectively. Simulation results acquired by these two algorithms were compared. The simulation acquired a stable reconstruction. In addition, the different temperature field models as well as their reconstruction images were given and reconstruction errors were analyzed. Simulation results indicate that these two algorithms are able to be used to reconstruct furnace three-dimensional temperature field and have good anti-disturbance ability. Compared with least squares method, SVD algorithm has better stability.