Component Content Soft-sensor Based on Neural Networks in Rare-earth Countercurrent Extraction Process

Throught fusion of the mechanism modeling and the neural networks modeling,a compo- nent content soft-sensor,which is composed of the equilibrium calculation model for multi-component rare earth extraction and the error compensation model of fuzzy system,is proposed to solve the prob- lem that the component content in countercurrent rare-earth extraction process is hardly measured on-line.An industry experiment in the extraction Y process by HAB using this hybrid soft-sensor proves its effectiveness.

Melt Index Prediction by Neural Soft-Sensor Based on Multi-Scale Analysis and Principal Component Analysis

Prediction of melt index （MI）, the most important parameter in determining the product＇s grade and quality control of polypropylene produced in practical industrial processes, is studied. A novel soft-sensor model with principal component analysis （PCA）, radial basis function （RBF） networks, and multi-scale analysis （MSA） is proposed to infer the MI of manufactured products from real process variables, where PCA is carried out to select the most relevant process features and to eliminate the correlations of the input variables, MSA is introduced to a^quire much more information and to reduce the uncertainty of the system, and RBF networks are used to characterize the nonlinearity of the process. The research results show that the proposed method provides promising prediction reliability and accuracy, and supposed to have extensive application prospects in propylene polymerization processes.

MIMO Soft-sensor Model of Nutrient Content for Compound Fertil- izer Based on Hybrid Modeling Technique

In compound fertilizer production, several quality variables need to be monitored and controlled simultaneously. It is very diifficult to measure these variables on-line by existing instruments and sensors. So, soft-sensor technique becomes an indispensable method to implement real-time quality control. In this article, a new model of multi-inputs multi-outputs （MIMO） soft-sensor, which is constructed based on hybrid modeling technique, is proposed for these interactional variables. Data-driven modeling method and simplified first principle modelingmethod are combined in this model. Data-driven modeling method based on limited memory partial least squares（LM-PLS） al.gorithm is used to build soft-senor models for some secondary variables.then, the simplified first principle model is used to compute three primary variables on line. The proposed model has been used in practicalprocess; the results indicate that the proposed model is precise and efficient, and it is possible to realize on line quality control for compound fertilizer process.

Component Content Soft-Sensor Based on Hybrid Models in Countercurrent Rare Earth Extraction Process

In consideration of the online measurement of the component content in rare earth countercurrent extraction separation process, the soft sensor method based on hybrid modeling was proposed to measure the rare earth component content. The hybrid models were composed of the extraction equilibrium calculation model and the Radial Basis Function (RBF) Neural Network (NN) error compensation model; the parameters of compensation model were optimized by the hierarchical genetic algorithms (HGA). In addition, application experiment research of this proposed method was carried out in the rare earth separation production process of a corporation. The result shows that this method is effective and can realize online measurement for the component content of rare earth in the countercurrent extraction.

Neural Networks Based Component Content Soft-Sensor in Countercurrent Rare-Earth Extraction

The equilibrium model for multicomponent rare earth extraction is developed using neural networks, which combined with the material balance model could give online prediction of component content in countercurrent rare earth (extraction) production. Simulation experiments with industrial operation data prove the effectiveness of the hybrid soft-(sensor).

A data-derived soft-sensor method for monitoring effluent total phosphorus

The effluent total phosphorus(ETP) is an important parameter to evaluate the performance of wastewater treatment process(WWTP). In this study, a novel method, using a data-derived soft-sensor method, is proposed to obtain the reliable values of ETP online. First, a partial least square(PLS) method is introduced to select the related secondary variables of ETP based on the experimental data. Second, a radial basis function neural network(RBFNN) is developed to identify the relationship between the related secondary variables and ETP. This RBFNN easily optimizes the model parameters to improve the generalization ability of the soft-sensor. Finally, a monitoring system, based on the above PLS and RBFNN, named PLS-RBFNN-based soft-sensor system, is developed and tested in a real WWTP. Experimental results show that the proposed monitoring system can obtain the values of ETP online and own better predicting performance than some existing methods.

Method of Soft-Sensor Modeling for Fermentation Process Based on Geometric Support Vector Regression

The soft-sensor modeling for fermentation process based on standard support vector regression(SVR) needs to solve the quadratic programming problem(QPP) which will often lead to large computational burdens, slow convergence rate, low solving efficiency, and etc. In order to overcome these problems, a method of soft-sensor modeling for fermentation process based on geometric SVR is presented. In the method, the problem of solving the SVR soft-sensor model is converted into the problem of finding the nearest points between two convex hulls (CHs) or reduced convex hulls (RCHs) in geometry. Then a geometric algorithm is adopted to generate soft-sensor models of fermentation process efficiently. Furthermore, a swarm energy conservation particle swarm optimization (SEC-PSO) algorithm is proposed to seek the optimal parameters of the augmented training sample sets, the RCH size, and the kernel function which are involved in geometric SVR modeling. The method is applied to the soft-sensor modeling for a penicillin fermentation process. The experimental results show that, compared with the method based on the standard SVR, the proposed method of soft-sensor modeling based on geometric SVR for fermentation process can generate accurate soft-sensor models and has much less amount of computation, faster convergence rate, and higher efficiency.

Forward heuristic breadth-first reasoning based on rule match for biomass hybrid soft-sensor modeling in fermentation process

Biomass is a key parameter in fermentation process, directly influencing the performance of the fermentation system as well as the quality and yield of the targeted product. Hybrid soft-sensor modeling is a good method for on-line estimation of biomass. Structure of hybrid soft-sensor model is a key to improve the estimating accuracy. In this paper, a forward heuristic breadth-first reasoning approach based on rule match is proposed for constructing structure of hybrid model. First, strategy of forward heuristic reasoning about facts is introduced, which can reason complex hybrid model structure in the event of few known facts. Second, rule match degree is defined to obtain higher esti- mating accuracy. The experiment results of Nosiheptide fermentation process show that the hybrid modeling process can estimate biomass with higher accuracy by adding transcendental knowledge and partial mechanism to the process.

Water quality soft-sensor prediction in anaerobic process using deep neural network optimized by Tree-structured Parzen Estimator

Anaerobic process is regarded as a green and sustainable process due to low carbon emission and minimal energy consumption in wastewater treatment plants(WWTPs).However,some water quality metrics are not measurable in real time,thus influencing the judgment of the operators and may increase energy consumption and carbon emission.One of the solutions is using a soft-sensor prediction technique.This article introduces a water quality soft-sensor prediction method based on Bidirectional Gated Recurrent Unit(BiGRU)combined with Gaussian Progress Regression(GPR)optimized by Tree-structured Parzen Estimator(TPE).TPE automatically optimizes the hyperparameters of BiGRU,and BiGRU is trained to obtain the point prediction with GPR for the interval prediction.Then,a case study applying this prediction method for an actual anaerobic process(2500 m^(3)/d)is carried out.Results show that TPE effectively optimizes the hyperparameters of BiGRU.For point prediction of CODeff and biogas yield,R^(2)values of BiGRU,which are 0.973 and 0.939,respectively,are increased by 1.03%–7.61%and 1.28%–10.33%,compared with those of other models,and the valid prediction interval can be obtained.Besides,the proposed model is assessed as a reliable model for anaerobic process through the probability prediction and reliable evaluation.It is expected to provide high accuracy and reliable water quality prediction to offer basis for operators in WWTPs to control the reactor and minimize carbon emission and energy consumption.

基于沙地猫群优化–最小二乘支持向量机的动态NOx排放预测

针对火电机组频繁调峰导致机组燃烧状态不稳,进而导致锅炉出口NOx浓度波动范围大的问题,提出一种基于沙地猫群优化(sand cat sarm optimization,SCSO)的最小二乘支持向量机(leastsquaressupportvectormachine,LSSVM) NOx动态预测模型。首先利用k近邻互信息计算时间延迟的同时筛选辅助变量。然后,基于SCSO算法进行输入变量阶次的选择。使用包含辅助变量时间延迟和阶次的信息作为模型的输入,SCSO算法优化最小二乘支持向量机参数,建立动态NOx排放最小二乘支持向量机预测模型(SCSO-LSSVM动态软测量模型)。最后将模型与未加入迟延的LSSVM模型,加入迟延的LSSVM模型和粒子群优化算法(particle swarm optimization,PSO)优化最小二乘支持向量机参数的动态软测量模型进行对比验证。结果表明,相较于其他模型,该文建立SCSO-LSSVM动态软测量模型均方根误差、平均绝对误差、平均绝对误差最小,预测精度最高,而且在NOx浓度剧烈波动时也能够较好地预测NOx浓度,具有很好的动态特性。

基于GAN的软测量缺失数据生成方法研究

针对工业过程中传感器数据缺失造成软测量模型精度低的问题,提出一种基于生成对抗网络(generative adversarial nets,GAN)的传感器缺失数据生成方法。利用孤立森林算法检测出传感器数据的缺失区域;利用缺失数据属性特征训练条件生成对抗网络(conditional generative adversarial nets,CGAN),在CGAN的输入条件中添加随机序列作为附加信息迭代送入CGAN中生成数据,并借助WGAN-GP(wasserstein generative adversarial nets gradient penalty)成本函数提高网络训练的稳定性;针对缺失区域检测结果引入采样器,将采样的数据填补进缺失区域,形成完整数据集,以提高软测量模型精度。以镍闪速炉温度传感器数据为目标变量进行软测量建模,验证所提出的提高软测量模型精度方法的可行性与有效性。

基于蝙蝠算法优化ESN的氯乙烯质量分数软测量模型预测

为解决氯乙烯因其精馏过程具有较强的非线性,无法实现对氯乙烯质量分数实时测量的问题,提出一种基于蝙蝠算法(bat algorithm,BA)优化回声状态网络(echo state network,ESN)的软测量模型BA-ESN.首先,通过对氯乙烯精馏过程的分析,选取模型的辅助变量,并将归一化处理后的数据作为模型输入变量;其次,由于回声状态网络中的权值和阈值都是随机产生的,影响其泛化能力,故采用蝙蝠算法对回声状态网络的输出权值进行优化,从而提高ESN模型的收敛速度;最后,将BA-ESN模型预测氯乙烯质量分数的预测结果与ESN模型和BP模型的预测结果进行对比.仿真结果表明:BA-ESN模型的预测精度较高,泛化能力和鲁棒性都较好,能够满足氯乙烯精馏过程实时测量的要求.

基于MF-EI法的水闸底板脱空参数反演

为了更好地解决水工结构以损伤诊断为目标的传感器布置问题,以软基水闸为研究对象、水闸底板脱空为损伤类型,针对传统有效独立法未考虑结构损伤信息的缺陷,将模态柔度作为结构损伤敏感特征量,并引入结构的模态柔度修正有效独立法,提出基于模态柔度-有效独立法(modal flexibility-effective independent method,简称MFEI)的水闸底板脱空参数反演方法。采用模态置信准则、Fisher信息阵、条件数准则和奇异值分解比准则4种评价指标,分别对距离系数-有效独立法(distance coefficient-effective independence method,简称DC-EI)、有效独立-总位移法(effective independence-total displacement method,简称EI-TD)和MF-EI法进行评价,并将3种方法得到的布置方案应用到不同工况下软基水闸底板脱空动力学反演中,对比分析3种方法各工况脱空参数的反演结果。结果表明:MF-EI法测得的模态矩阵具有更好的正交性、可观测性和模态扩阶性,同时保留了对软基水闸底板脱空敏感的损伤信息,3种工况下的脱空参数反演识别精度分别提高至97.5%,96.1%和75.9%,是一种较为有效的水闸底板脱空参数反演方法。

3D打印硅橡胶研究进展

柔性硅橡胶弹性体具有生物相容性、电绝缘性、耐水性、耐高低温等特性,在电子、医疗、航空航天等高端领域应用广泛。硅橡胶弹性体是3D打印领域研究最多的高分子材料之一。硅橡胶3D打印主要采用墨水直写或材料挤出、嵌入式固化打印和立体光刻、喷墨打印、粉末床烧结等技术,应用于柔性传感器、康复鞋垫、软机器人、光学透镜、柔性电子等器件的加工制备。但目前还存在打印精度不高、成本贵、打印材料缺乏以及复杂结构难以实现等问题。近年来人们通过开发新型打印技术,以及聚合物分子设计和材料创新,解决硅橡胶打印加工难题,提高3D打印制件的性能,拓展制件应用功能。本文综述了近年来3D打印硅橡胶技术的研究进展。

基于形状记忆合金的软体指套设计与精确控制

提出了一种基于形状记忆合金(SMA)弹簧驱动的软体指套。设计了基于霍尔传感器的传感系统,以精确感知指套弯曲状态。通过分析人手的结构和运动特征,建立了指套的运动学模型。最后,搭建实验平台,开展传感器标定实验分析,完成锯齿状速度曲线跟踪。实验结果表明:软体指套通过高精度的位置信息和经典PID控制器,能够实现精确运动控制。

基于即时学习的改进条件高斯回归软测量

基于数据驱动的在线软测量是当前工业智能化感知的重要研究方向。在算法实际部署中,过程模态切换以及数据漂移都会导致软测量性能下降,传统自适应方法又存在模型单一、模态遗忘等不足。针对上述问题提出一种基于即时学习的样本时空加权条件高斯回归(STWCGR)软测量算法。该方法用概率密度估计和条件概率计算实现软测量建模和预测:首先根据即时学习思想通过样本时空混合加权方法筛选局部建模数据,然后结合高斯混合回归思想累积局部单高斯概率密度模型对数据分布进行拟合,最后引入预测动量更新和模态更新策略提高预测稳定性并赋予模型对新工况的学习适应能力。通过仿真实验验证了所提方法在预测精度、稳定性以及新模态适应能力上的有效性。

基于慢特征分析与最小二乘支持向量回归集成的草酸钴合成过程粒度预报

草酸钴合成过程是钴湿法冶炼的关键单元操作,其粒度分布是重要的质量指标,然而难以在线实时测量。同时,草酸钴合成过程通常存在非线性、多约束和慢时变特征。因此,提出一种集成慢特征分析(slow feature analysis,SFA)与最小二乘支持向量回归(least square support vector regression,LSSVR)的草酸钴粒度预报模型对草酸钴合成过程质量指标实现在线测量。在该方法中,首先,SFA方法可以有效地捕获过程的慢特征向量,解决慢时变问题;然后,利用LSSVR方法建立慢特征与粒度之间的非线性关系模型,进而实现质量指标在线预报。最后,应用非线性的数值案例以及草酸钴合成过程数据,验证该方法的有效性。实验结果显示:相较于单一的径向基函数神经网络(radial basis function neural network,RBFNN)、LSSVR预测模型以及SFA与NN相结合的预报模型,所提方法在数值案例中的预测精度分别提升了13.31%、2.26%、1.72%;在草酸钴合成过程中的预测精度分别提升了13.27%、9.96%、8.92%。

用于间歇过程的拓扑时间卷积网络软测量方法

在间歇过程中,变量的相关性和时序性通常演变为变量间的交叉相关性。在间歇过程的数据驱动软测量建模中,捕捉交叉相关性可以提升模型的透明度,因此提出基于拓扑引导时间卷积网络的软测量方法。首先,使用条件格兰杰因果关系检验构建变量关系的拓扑结构,并使用自注意力机制完善拓扑图。然后,使用图注意力机制以每个节点为中心构建子图,并通过时间卷积捕捉子图中变量间的交叉相关性。最后,使用图解释器对模型预测逻辑的物理一致性做出评估。基于青霉素发酵过程的质量预测进行实验,实验结果验证了所提方法的有效性和优越性,证明了模型的预测逻辑符合过程机理。

助训练策略下的多模型软测量建模

由于复杂工业过程中存在强非线性、多阶段耦合以及有标签样本数量偏少的情况,传统的全局软测量模型难以精确描述整个过程。为此,提出一种助训练策略下的多模型软测量建模方法。该方法采用模糊C均值聚类算法挖掘样本集中的相似性样本并建立若干子模型;通过引入助训练策略,形成基于主、辅学习器的协同训练框架,并设计置信度评估机制淘汰误差样本的同时扩充子模型的建模空间;进而将模糊隶属度作为D-S证据理论的概率分配函数计算出子模型权重,对子模型的输出进行融合以得到最终的模型预测结果。通过对脱丁烷塔工业过程的实际数据进行建模仿真,结果表明此模型具有良好的预测性能。

带FIR滤波的非线性滑动平均动态软测量模型

非线性滑动平均(NMA)模型能有效描述工业过程的动态特性,是一种典型的动态软测量模型.而受限于模型复杂度,NMA模型的输入时序边界相对较窄,难以适应带有大滞后或强测量噪声的动态工业过程.针对该问题,本文将NMA模型与结构简单、输入时序边界宽的FIR滤波器相结合,构造一种非线性、强抗干扰的软测量建模策略.并设计层白化结构来避免二者间的参数耦合现象,采用Adam算法进行同步优化,提高模型的预测精度及训练效率.最后,利用数值仿真和硫回收过程建模实验,验证所提模型的预测精度以及模型设计的合理性.