ECT图像重建正则化参数选取新方法

电容层析成像图像重建是一不适定反问题。此种情况下，仅使用最小二乘法不能保证获得满意的介质分布图像重建结果，因此广泛使用TIkhonov正则化算法来产生适当的解。正则化参数的合适选取对图像重建至关重要，其对重建质量和计算时间都有影响。本文提出了一种基于最平坦斜率的Tikhonov正则化参数选择方法，并针对2种典型介质分布，将基于此方法计算的正则化参数同L-曲线法在电容测量数据无噪声和施加噪声情况下的图像重建结果进行了比较。     

基于少量声波飞行时间数据的温度场重建

为依据少量声波飞行时间数据较高精度地重建温度场,提出了一种基于径向基函数和奇异值分解的声学CT温度场重建新算法.采用新算法对单峰和双峰温度场模型进行了仿真数据重建,重建结果表明,与高斯函数正则化重建算法、代数重建算法相比,新算法的重建精度有明显改善.采用新算法对实验室内的均匀温度场和加热温度场进行了实测数据重建,重建结果与被测温度场一致,且均匀温度场的重建均方根百分误差仅为0.31%.由于新算法重建速度快、重建精度高、抗干扰能力较强,可望用于复杂温度场的在线重建.     

基于U曲线法的油气润滑ECT系统图像重建病态优化

采用U曲线法确定油气润滑ECT系统图像重建中的正则化参数,分析正则化处理后灵敏度矩阵的病态性;通过LBP算法和Tikhonov正则化算法分别对油气润滑ECT系统管道截面进行第一次图像重建;对第一次重建图像的灰度分布矩阵进行门限滤波阈值的优化,并对管道截面进行二次图像重建。结果表明:相较于L曲线法,U曲线法选取的正则化参数在削弱灵敏度矩阵病态程度方面的作用显著;第一次图像重建中,图像重建质量有较大改善;门限滤波阈值优化后的二次图像重建中,图像重建质量进一步提高。研究表明U曲线法确定的正则化参数和门限滤波阈值优化有助于提高油气润滑ECT系统的图像重建质量。     

基于两相流电容层析成像系统的正则化Landweber算法

针对电容层析成像系统的不适定性问题,提出了一种正则化Landweber阈值迭代改进算法。该算法将正则化项引入初值等式约束中,得到带约束的正则化问题,并在Landweber迭代算法的过程中设置门限阈值,在每次迭代中都自适应的调节阈值参数,最终用带约束的正则化Landweber阈值迭代算法得到最终的重建图像结果。数值仿真实验表明,正则化Landweber阈值迭代算法在改善图像重建质量和提高收敛速度方面均有明显的优势。     

基于小波变换的正则化盲图像复原算法

提出了一种将小波变换和自适应正则化方法相结合的盲图像复原算法。该算法先对退化后的图像进行小波分解,得到图像在不同子频段的信息;然后针对各个子频段内图像的频率和方向特性,使用不同的自适应正则化复原方法,在图像的低频子频段进行去模糊;高频子频段则进行抑制噪声和保边缘特征;最后通过小波逆变换得到复原后的图像。实验结果表明, MSE减少了1.60,信噪比增量为1.76,算法性能和复原效果相对空间自适应正则化方法,都有一定的提高。     

电容层析成像系统图像重建新算法的研究

快速而又具在一定精度的图像重建算法是电容层析成像（ＥＣＴ）技术的关键。本文提出一种用多元线性回归法建立ＥＣＴ系统正向模型,用正则化法获得图像重建这一不适宜逆问题稳定解的图像重建新算法。仿真结果表明该方法能快速而较高精度地重建出两相流断层图像。     

声学法炉内温度场与速度场协同测量方法研究

实时监测炉内燃烧温度场和烟气速度场是保证锅炉安全、经济运行的重要手段,声学法测物理场被认为是一种非侵入性和有效的测量方法。本文提出了一种基于声波法的炉内温度场和烟气速度场的协同测量新方法,建立了基于径向基函数的多物理场重建模型,采用Tikhonov正则化算法求解不适定问题,同时考虑了声波的折射效应对物理场的重建影响。采用典型的炉内物理场模型进行了数值模拟,模拟结果表明,本文方法能够很好的协同重建温度场和速度场。当考虑声线弯曲时能够显著提高各物理场的重建质量。算法具有较好的适应性和良好的抗噪性能,重建精度较高,标准均方根误差在10%以下。模拟实验平均计算时间为31.4 s,可保证炉内声学测量的实时性。声学法协同测量多物理场可为优化炉膛燃烧过程提供依据。     

声学法深海热液温度场测量及重建算法研究

介绍了利用声学方法测量深海热液温度场的基本原理。对单峰温度场模型和双峰温度场模型,采用最小二乘法进行重建仿真,给出了重建温度场的三维图和等温线图,计算了重建温度场的绝对误差、相对误差和均方根误差。比较了增加声学测量路径后单峰温度场模型的重建结果。仿真结果表明:最小二乘法具有较高的重建精度,增加声学测量路径能有效地提高温度场重建精度。湖试实验结果表明,最小二乘法可基本还原单峰温度场。     

基于RBF神经网络的温度场重建算法研究

在声学法锅炉炉膛温度场测量中，重建算法是实现炉膛温度场重建的关键。本文提出一种基于径向基函数神经网络的复杂温度场重建算法。该算法首先对被测温度场用离散余弦变换，建立离散余弦变换低阶次项DCT系数向量与声波路径平均温度向量的映射关系，然后利用RBF神经网络良好的函数逼近能力实现该映射关系，并通过正交最小二乘法进行学习和训练，实现被测温度场的重建。本文对3种原型温度场进行了重建，并在40dB、30dB和20dB等3种不同噪声水平下进行了重建实验。仿真及初步实验结果表明，该算法具有温度场重建精度高、速度快、抗干扰能力强的特点。     

自适应卡尔曼滤波在无刷直流电机系统辨识中的应用

为了有效抑制量测噪声特性变化对系统辨识精度的影响以获得准确的无刷直流电机模型,提出了一种采用自适应卡尔曼滤波算法的无刷直流电机系统辨识方法。通过计算新息理论方差的极大似然最优估计,并将其引入卡尔曼滤波算法中修正滤波增益来抑制量测噪声特性变化对辨识结果的影响,使该滤波算法实现对模型参数的准确估计,提高辨识精度。实验结果表明,在量测噪声特性变化的情况下,该算法能够准确跟踪实际量测噪声特性的变化,参数估计平滑,相对于目前系统辨识广泛采用的带有遗忘因子的递推最小二乘算法,输出误差的均方根值减小了73.5%。该算法简单易行,计算量小,辨识结果可以很好地描述系统行为,便于在工程实践中应用。     

基于统计最优的近场声全息理论与仿真实验

对统计最优平面近场声全息（SOPNAH）技术进行研究,提出了一种吉洪诺夫（Tikhonov）正则化方法并选择最优正则化参数λ。通过仿真实验验证了结合吉洪诺夫正则化方法的SOPNAH能够对噪声源进行准确定位及实现声场重建,并且重建的精度比传统的方法更高。     

Adaptive regularization parameter optimization in output-error-based finite element model updating

In finite element (FE) model updating, regularization methods are required to alter the ill-conditioned system of equations towards a well-conditioned one. The present study addresses the regularization parameter determination when implementing the Tikhonov regularization technique in output-error-based FE model updating. As the output-error-based FE model updating results in a nonlinear least-squares problem which requires iteration for solution, an adaptive strategy that allows varying value of the regularization parameter at different iteration steps is formulated, where the optimal regularization parameter at each iteration step is determined based on the computationally efficient minimum product criterion (MPC). The performance of MPC in output-error-based FE model updating is examined and compared with the commonly used L-curve method (LCM) and the generalized cross validation (GCV) through numerical studies of a truss bridge using noise-free and noise-corrupted modal data. It is shown that MPC is effective and robust in determining the regularization parameter compared with the other two methods, especially when noise-corrupted data are used. The adaptive strategy is more efficient than the fixed strategy that uses a constant value of the regularization parameter throughout the iteration process.     

正则化自适应匹配追踪电能质量数据重构方法

为了有效改善传统方法应用在电能质量信号的采集和压缩方面所面临的资源浪费以及重构性能较差等问题,提出了一种基于正则化自适应匹配追踪的电能质量数据重构方法,此方法利用压缩感知理论,对电能质量信号进行采样和压缩并行处理。首先,对感知矩阵中的原子进行一次挑选并且计算相关系数,将挑选出的原子索引值存入至候选集中。然后,在电能质量信号稀疏度K不作为先验条件的前提下,对候选集中原子的数目进行自适应地调节,并运用正则化的处理过程完成支撑集的二次挑选,用步长逐步逼近信号的稀疏度进而准确重构出电能质量原始信号。仿真实验结果表明,信号的重构精度高于98.2%,并且能够保存原始电能质量信号大部分能量,重构信噪比高,均方误差小。     

自适应SA-PSO优化的威布尔混合分布参数估计方法及应用

针对采用传统参数估计方法得到的模型拟合误差较大的问题,建立多重威布尔混合分布参数估计的非线性最小二乘模型,并提出基于模拟退火（SA）思想的自适应粒子群（PSO）算法进行求解。在PSO算法优化过程中,采用自适应方法调整惯性权重和加速因子,加快其收敛速度;引入模拟退火机制,根据Metropolis准则确定最优粒子的取舍,改善其全局搜索能力。将该方法应用到某型柴油机喷油器失效分布的参数估计中,并与图解法、基于Levenberg-Marquardt的非线性最小二乘法、标准PSO算法、自适应PSO算法求解的结果进行比较,分析所提方法的优化性能及精度。结果表明,该方法能够有效提高多重威布尔混合分布模型参数估计的精度和效率。     

A hybrid regularization method combining Tikhonov with total variation for electrical resistance tomography

Electrical resistance tomography (ERT) is a promising measurement technique in industrial process imaging. However, image reconstruction in ERT is an ill-posed inverse problem. Regularization methods have been developed to solve the ill-posed inverse problem. Since the penalty term is a form of L2-norm, Tikhonov regularization method guarantees the stability of the solution, but it always makes the image edge oversmoothed. Total variation (TV) regularization method has good ability of preserving image edges. A hybrid regularization method, which combines Tikhonov with TV regularization method, is proposed to get better reconstructed images. The choice of the adaptive weighted parameter between TV and Tikhonov penalty term has been discussed in detail. In the proposed hybrid regularization method, the function of conductivity gradients is used as the adaptive weighted parameter to control automatically the weighting between the penalty terms from TV and Tikhonov regularization. For the model with sharp edges, the proportion of the penalty term from TV regularization is increased to preserve the edges, while for the model with smooth edges, the proportion of penalty term from Tikhonov regularization is increased to make the solution stable and robust to noise. Both simulation and experimental results of Tikhonov, TV and hybrid regularization method are shown respectively, which indicates that the hybrid regularization method can improve the reconstruction quality with sharp edges and is more robust to noise, and it is applicable for models with different edge characteristic.     

An adaptive non-convex hybrid total variation regularization method for image reconstruction in electrical impedance tomography

The image reconstruction of conductivity distribution in electrical impedance tomography (EIT) is a seriously ill-posed inverse problem. To cope with the problem, it is recognized that the regularization method is an effective approach. In this paper, an adaptive non-convex hybrid total variation (ANHTV) regularization method is proposed to reconstruct the conductivity distribution in EIT. The iterative reweighted least squares algorithm and the iterative alternating direction method of multipliers algorithm are developed to solve the ANHTV-based inverse model in the image reconstruction. Besides, all the parameters utilized in the inverse model are adaptively selected. To validate the advantage of the proposed method, extensive numerical simulation and experimental work have been carried out. Also, qualitative and quantitative comparisons with two convex TV-based regularization methods are conducted. The results show that the proposed method is more advantageous in terms of staircase effect suppression, edge information preservation and noise resisting in the image reconstruction.     

Ultrasonic tomography based temperature distribution measurement method

The temperature distribution information plays an important role in industrial applications. Owing to the advantages such as non-intrusive sensing and low cost, the ultrasonic tomography (UT) is considered to be a promising method for temperature field visualization. The ultrasonic time-of-flight (TOF) measurement and the reconstruction algorithm are crucial for practical applications of the UT measurement. In this paper, a dual-threshold measurement method is proposed to ensure a high-quality TOF measurement. In view of the inaccurate nature of the reconstruction model and TOF data, a new reconstruction method that integrates the advantages of the Tikhonov regularization method and the least squares support vector machine (LSSVM) is proposed to improve the reconstruction quality. The experimental results were compared against thermocouple measurements and the results show that the temperature distribution can be reconstructed with the error of 1.3%, which validates the feasibility and effectiveness of the proposed Tikhonov-LSSVM reconstruction algorithm.     

变权重指数平滑法在冷轧过程控制中的应用

冷轧机组在轧制过程中很多状态都是在时刻变化的。轧辊直径、温度、粗造度、轧件来料厚度、硬度的变化、机架间张力的波动、各个调节器特性的变动、仪表零点的漂移等都会影响模型计算的精度。故在轧制过程中引入自适应自学习技术是提高模型计算精度的一种很好的选择。冷轧过程控制系统自适应算法通常采用指数平滑法,为了提高自适应速度往往采用变权重指数平滑法。