基于RBF神经网络的三维温度场重建算法

Three-dimensional Temperature Field Reconstruction Algorithm Based on RBF Neural Network

声学法测量温度场是目前很具发展前景的一种温度场测量方法,而重建算法是实现声学法温度场重建的关键。提出了一种基于径向基函数(Radical Basis Function,RBF)神经网络的三维温度场重建算法。通过对被测温度场进行三维离散余弦变换(Discrete Cosine Transform,DCT),再利用RBF神经网络良好的函数逼近能力,实现DCT低阶次项系数向量与声波路径平均温度向量间的映射关系,最后通过逆离散余弦变换实现被测温度场的重建。进行了对模拟温度场的重建仿真,结果表明,该算法具有温度场重建精度高、速度快等特点。

Temperature fiekt acoustic measurement is a promising temperature field measurement method at present, and reconstruction algorithm is essential to temperature field image. This paper presented a new algorithm based on RBF neural network to reconstruct the three-dimensional temperature field. The algorithm used three-dimension discrete cosine transform （DCT） on tem- perature field and established a mapping relation between low order term coefficient vector and sound wave path average tempera- ture vector then implemented the mapping relation using radical basis function （RBF） neural network that has strong function fitting ability. The three-dimensional temperature field was reconstructed by using inverse three-dimension discrete cosine transform. Simulation results show that the algorithm features high precision and high-speed.