基于DSP的中药丸料湿度密度在线检测系统

设计与研制一种能在现代化中药制药生产线上在线连续测量中药丸料湿度与密度的微波检测系统。系统包括三个组成模块：微波信号源模块，微波谐振腔，数据处理模块。其中，微波谐振腔是专门设计用于中药丸料湿度与密度检测的中心开通孔的金属谐振腔，当中药丸料通过微波谐振腔的中孔时，丸料密度及湿度均令谐振腔的谐振频率产生偏移和功率衰减，数据处理模块据此计算中药丸料的湿度与密度。为提高检测精度，采用模糊聚类算法对实验数据进行分组，对金属谐振腔的频偏和功率衰减特性进行建模，并利用DSP设计并实现了中药丸料湿度与密度的在线检测。现场调试表明：该检测系统适宜于现代化中药制药生产线上的湿度连续在线测量，其检测精度达到93％以上。     

基于微波谐振腔的湿度传感器

构建了含水混合物介电特性模型，设计了基于此模型的开路同轴谐振腔传感器，并指出影响其传感特性的两个主要参数是保护盖介电常数和空载谐振频率。鉴于此，针对保护盖材料及空载谐振频率对传感器传感特性的影响进行了全面分析和仿真。加工了具有不同谐振频率的微波谐振腔及不同材料的端口保护盖。实验结果表明，该模型能够很好地指导传感器的设计，同时选取Al2O3作为保护盖材料及具有空载谐振频率为2．5GHz的谐振腔，具有较好的测量效果。     

截止波导介质腔介电常数测量理论与方法研究

本文提出了用截止波导介质谐振腔测量微波材料相对介电常数和微波损耗的方法.利用开波导法给出了腔中TE0ml谐振模的电磁场,导出了利用此模测量介电常数和微波损耗的公式;通过测量频率和谐振曲线,就能算出材料的复介电常数,并对谐振模TE0ml进行了讨论;比较了这种测量方法与原有的短路金属板介质谐振腔法的优劣,结果表明采用截止波导介质谐振腔测量材料的微波损耗时更有优越性.     

基于正交实验分析的面齿轮磨削表面粗糙度的研究

对面齿轮磨削齿面粗糙度进行正交实验分析,采用多元非线性回归分析法建立了砂轮转速、进给速度、磨削深度的回归模型,分析了砂轮转速、进给速度、磨削深度等磨削参数对面齿轮磨削齿面粗糙度的影响状况,最后通过数学模型对单因素实验组数据进行实测值与预测值的比较,显示6组数据预测值与实际测量值的最大误差为11.13%。说明相对误差不大,此模型具有一定的精度。     

微波谐振腔含水率传感器的设计方法

构建了含水混合物介电特性模型,设计了基于此模型的开路同轴谐振腔传感器,并指出影响其传感特性的两个主要参数是保护盖介电常数和空载谐振频率.针对此问题,利用MATLAB对保护盖介电常数和空载谐振频率对传感器传感特性的影响进行了仿真和分析.加工了具有不同谐振频率的含水率传感器及不同介电常数的保护盖.实验结果表明,本文提出的模型能够很好地指导传感器的设计,同时选取Al2O3制作的保护盖和空载谐振频率为2.5G的同轴谐振腔具有较好的测量效果.     

测量气体折射率的微波谐振腔频率标定方法

利用微波谐振腔实现对气体折射率的测量,需在真空环境中精确确定测量系统的零点频率,即微波谐振腔零点频率。针对微波谐振腔定标中真空度难以保证的现状,提出了一种常规大气环境下腔体零点频率的标定方法。通过精确测量大气环境的温度、湿度、压强参数得到高精度实时折射率,最后根据得到的高精度实时折射率对腔体谐振零点频率进行调节,以实现谐振腔的零点频率标定。实验结果表明,该方法具有误差小、调节方便的特点,且能够保证整个微波测量系统的测量精度。     

基于微扰法的气体折射率测试仪的设计

折射率是气体的重要物理性质,精确测量气体的折射率具有重要意义。以腔体微扰理论为原理,利用微波谐振腔的谐振频率随腔内介质的折射率变化而发生偏移的特性,通过测量微波谐振腔谐振频率的偏移量获得气体的折射率,由于该方法采用3 cm微波谐振频率,其频率的偏移量可以精确测量出来,获得的折射率精度很高。通过实验表明该系统具有反应速度快、稳定性好、时效性好的特点。在海上蒸发波导实际测试中,可精确测量出大气结构参数,为蒸发波导在线测量,实现雷达超视距探测和波导电磁盲区补盲措施提供可靠依据。     

基于ANN和SVM的质量预测方法研究

针对随机抽取的质量数据序列的特点,提出时序空间(Time Seauence Space,TSS)的概念,将人工神经网络(Artificial Neural Network,ANN)和支持向量机(Support VectorMachine,SVM)回归模型引入质量数据预测领域.并给出了相应的过程和算法.使用均方误差对拟合精度进行检验,用相对误差对预测精度进行检验.结果表明,相对于传统的多项式回归模型,人工神经网络和支持向量机回归模型的拟合精度较高,相对误差较小,泛化能力较强,可以作为质量数据的预测方法.     

PSO优化SVM在木材干燥建模中的应用

在木材干燥计算机控制过程中,基准模型化是此控制过程的必要环节。为提高此建模预测精度,针对SVM木材干燥基准模型的参数进行研究。利用粒子群优化算法中的粒子位置和速度优化此模型参数,并对木材含水率进行预测。仿真实验表明,PSO算法在优化SVM木材干燥基准模型参数方面表现出良好的性能,预测结果具有很高的精度,此模型具有较好的泛化能力和预测能力。     

基于多模型建模的木材含水率测量方法研究

针对单一模型无法全面描述木材含水率的复杂非线性特性问题,本文提出一种多模型建模的方法测量木材含水率,该方法先用模糊C均值聚类算法将含水率等效电阻、进风口、出风口温度等数据分成具有不同聚类中心的子集,每一子集依据样本数分别采用径向基网络、支持向量积训练得出子模型,再用模糊聚类后产生的隶属度将各子模型的输出加权求和得到木材含水率测量的模型。通过实例验证了本方法与RBFNN建模对于木材含水率的检测,具有更好的泛化结果和测量精度。     

[成形质量的闭环与自适应控制技术]热轧带钢出口凸度数据驱动建模及智能化预测分析

提出一种基于热轧现场生产数据和智能算法的新型带钢出口凸度预测模型，该模型采用差分进化算法对支持向量机的惩罚因子和核函数宽度进行优化。确定了支持向量回归模型的最佳参数组合，采用大量实际生产数据对模型进行训练并将其用于带钢出口凸度预测。该模型结构简单、容易实现，其整体性能用平均绝对误差（MAE）、平均绝对百分误差（MAPE）、均方根误差（RMSE）和决定系数R2来评价。预测值和实际值的比较验证了所提出模型的可行性。     

Support vector machine modelling applied to benchmark data set for two-phase Coriolis mass flow metering

An earlier paper introduced a dataset of Coriolis meter mass flow and density errors, induced by the effects of two-phase (gas/liquid) flow, as a benchmark for which various error correction strategies might be developed. That paper further presented a series of error correction models based on neural nets. The current paper presents an alternative analysis of the same data set, using a support vector machine (SVM) approach. The analysis demonstrates that, for the benchmark data set, more accurate models are generated than those developed using neural nets. More specifically, it is found that a linear SVM model provides better performance than non-linear SVM. This improved performance may arise from over-fitting by both non-linear SVM and neural nets on this relatively small data set.     

基于SVM的多传感器信息融合算法

支持向量机(Support Vector machine,简称SVM)是一种基于结构风险最小化原理,具有很高泛化性能的学习算法。针对工业多传感器测控系统中,被测系数与相关参数之间存在有较大的非线性和模糊关系,提出了一种基于支持SVM的多传感器信息融合模型及算法。为小样本、非线性、高维数一类多传感器信息融合问题的建模提供了一种有效的途径。通过对“纸张水份在线测量系统”应用表明,基于SVM的多传感器信息融合模型及算法在测量精度和推广性能上都具有一定的优越性。     

Discharge prediction for rectangular sharp-crested weirs by machine learning techniques

The stage-discharge relationship of a weir is essential for posteriori calculations of flow discharges. Conventionally, it is determined by regression methods, which is time-consuming and may subject to limited prediction accuracy. To provide a better estimate, the machine learning models, artificial neural network (ANN), support vector machine (SVM) and extreme learning machine (ELM), are assessed for the prediction of discharges of rectangular sharp-crested weirs. A large number of experimental data sets are adopted to develop and calibrate these models. Different input scenarios and data management strategies are employed to optimize the models, for which performance is evaluated in the light of statistical criteria. The results show that all three models are capable of predicting the discharge coefficient with high accuracy, but the SVM exhibits somewhat better performance. Its maximum and mean relative error are respectively 5.44 and 0.99%, and 99% of the predicted data show an error below 5%. The coefficient of determination and root mean square error are 0.95 and 0.01, respectively. The model sensitivity is examined, indicative of the dominant roles of weir Reynolds number and contraction ratio in discharge estimation. The existing empirical formulas are assessed and compared against the machine learning models. It is found that the relationship proposed by Vatankhah exhibits the highest accuracy. However, it is still less accurate than the machine learning approaches. The study is intended to provide reference for discharge determination of overflow structures including spillways.     

基于AWLS-SVM的污水处理过程软测量建模

针对污水处理过程建模中样本数据可能存在的测量误差对模型性能的影响,提出一种自适应加权最小二乘支持向量机(AWLS-SVM)回归的软测量建模方法。该方法基于最小二乘支持向量机模型,根据样本拟合误差,并结合改进的指数分布赋权规则,自适应地为每个建模样本分配不同的权值,以降低随机误差对模型性能的影响;同时采用一种全局优化算法——混沌粒子群模拟退火(CPSO-SA)算法对最小二乘支持向量机的模型参数进行优化选择,以提高模型的泛化能力。仿真实验表明,AWLS-SVM模型的预测精度及鲁棒性能优于LS-SVM和WLS-SVM。最后,应用AWLS-SVM方法建立污水处理过程出水水质关键参数的软测量模型,获得了较好的效果。     

Performance evaluation of two different neural network and particle swarm optimization methods for prediction of discharge capacity of modified triangular side weirs

Technical design of side weirs needs high accuracy in predicting discharge coefficient. In this study, discharge coefficient prediction performance of multi-layer perceptron neural network (MLPNN) and radial basis neural network (RBNN) were compared with linear and nonlinear particle swarm optimization (PSO) based equations. Performance evaluation of the model was done by using root mean squared error (RMSE), coefficient of determination (R²), mean absolute error (MAE), average absolute deviation (δ) and mean absolute relative error (MARE). Comparison of the results showed that both neural networks and PSO based equations could determine discharge coefficient of modified triangular side weirs with high accuracy. The RBNN with RMSE of 0.037 in test data was found to be better than MLPNN with RMSE of 0.044 and multiple linear and nonlinear PSO based equations (ML-PSO and MNL-PSO) with RMSE of 0.043 and 0.041, respectively. However, due to their simplicity, PSO based equations can be sufficient for use in practical cases.     

Regression analysis, support vector machines, and Bayesian neural network approaches to modeling surface roughness in face milling

This study examines the influence of cutting speed, feed, and depth of cut on surface roughness in face milling process. Three different modeling methodologies, namely regression analysis (RA), support vector machines (SVM), and Bayesian neural network (BNN), have been applied to data experimentally determined by means of the design of experiment. The results obtained by the models have been compared. All three models have the relative prediction error below 8%. The best prediction of surface roughness shows BNN model with the average relative prediction error of 6.1%. The research has shown that, when the training dataset is small, both BNN and SVR modeling methodologies are comparable with RA methodology and, furthermore, they can even offer better results. Regarding the influence of the examined cutting parameters on the surface roughness, it has been shown that the feed has the largest affect on it and the depth of cut the least.     

A wrapper approach-based key temperature point selection and thermal error modeling method

A wrapper approach-based key temperature point selection and thermal error modeling method is proposed to concurrently screen the optimal key temperature points and construct the thermal error model. This wrapper approach can strengthen the intrinsic relation between the key temperature points and the thermal error model to ensure the strong prediction performance. On the whole, the least squares support vector machine (SVM) is used as the basic thermal error modeling method and the binary bat algorithm (BBA) is used as the optimization algorithm. The selection status of temperature points and the values of hyperparameters γ and σ2 of SVM are coded in separate binary parts of the artificial bat’s position vector of BBA. The cost function is designed by balancing the prediction error and the number of key temperature points. For verification, the thermal error experiment was conducted on a horizontal machining center. Feeding the collected experimental temperature data and thermal error data to the proposed method, three optimal key temperature points were screened out and the corresponding optimal hyperparameters were simultaneously searched. To verify the superiority of the proposed method, the prediction performance comparison analysis was conducted with the conventional filter-based method. Specifically, in the conventional method, the key temperature points were screened by combining fuzzy c means (FCM) clustering and correlation analysis, and the multiple linear regression (MLR), the backpropagation neural network (BPNN), and the SVM were used to build the thermal error model, respectively. Comparison results showed that the prediction accuracy of the proposed method increased by up to 44.0% compared to the conventional method, which suggests the superior prediction performance of the proposed method.     

基于多翼型特征的非奇异变形感知方法研究

针对传统变形感知方法在复杂翼型结构中常见的病态、奇异等问题，提出了一种基于多翼型特征的非奇异变形重构模型。依据Timoshenko梁变形理论，采用依存插值技术离散单元位移场，建立理论截面应变与测量应变的最小二乘变分函数，推导单元节点变形与测量应变的积分重构模型。该模型的位置无关性有效消除评估截面选取不当引起的奇异，增强重构模型在复杂翼型结构中的适用性。同时，针对应变传感器服役期间常见的环境扰动，以重构精度与鲁棒性为评估指标，建立自适应多目标粒子群优化模型。实验结果表明，提出的重构模型整体测量精度较高，在机翼变形量小于20 mm范围内最大绝对误差为0.26 mm,最大相对均方根误差为0.42%;当变形量增大时，绝对误差随之增大，但相对均方根误差不超过3.5%。因此基于多翼型特征的非奇异变形重构模型能够满足机翼实时重构需求，有效扩展变形感知方法在复杂结构中的应用价值。     

Time series modeling of endocardial border motion in ultrasonic images comparing support vector machines, multilayer perceptrons and linear estimation techniques

This paper presents a novel study of endocardial boundary motion tracking from sequences of echocardiogram images using neural network and linear estimation techniques. Contrary to the majority of previous studies theoretically analyzing endocardial motion physical parameters, a time series modeling approach is herein adopted. Such a modeling approach is demonstrated, by extensive experimentation, to be very efficient in terms of endocardial border motion tracking performance. The tracking performance of the different modeling techniques involved is evaluated quantitatively by defining suitable error measures as well as qualitatively. A thorough experimental investigation shows the importance of the highly correlated nature of endocardial contour motion within a cardiac cycle. Moreover, it shows that its short term dynamics can be almost equally well captured by support vector machines (SVM) for non-linear regression, multilayer perceptrons (MLP) and two matrix-parameters vector autoregressive (VAR) process models. Longer term dynamics, however, can be described more effectively using SVM for non-linear regression rather than MLPs. Additionally, the latter is shown to describe more effectively longer term dynamics than the VAR modeling approach. Such results are important for modeling the endocardial motion process, aiming at introducing improved adaptive focusing of ultrasonic scanners in order to enhance the quality of heart ultrasonic images.