A multiscale adaptive framework based on convolutional neural network:Application to fluid catalytic cracking product yield prediction

Since chemical processes are highly non-linear and multiscale,it is vital to deeply mine the multiscale coupling relationships embedded in the massive process data for the prediction and anomaly tracing of crucial process parameters and production indicators.While the integrated method of adaptive signal decomposition combined with time series models could effectively predict process variables,it does have limitations in capturing the high-frequency detail of the operation state when applied to complex chemical processes.In light of this,a novel Multiscale Multi-radius Multi-step Convolutional Neural Network(Msrt Net)is proposed for mining spatiotemporal multiscale information.First,the industrial data from the Fluid Catalytic Cracking(FCC)process decomposition using Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)extract the multi-energy scale information of the feature subset.Then,convolution kernels with varying stride and padding structures are established to decouple the long-period operation process information encapsulated within the multi-energy scale data.Finally,a reconciliation network is trained to reconstruct the multiscale prediction results and obtain the final output.Msrt Net is initially assessed for its capability to untangle the spatiotemporal multiscale relationships among variables in the Tennessee Eastman Process(TEP).Subsequently,the performance of Msrt Net is evaluated in predicting product yield for a 2.80×10^(6) t/a FCC unit,taking diesel and gasoline yield as examples.In conclusion,Msrt Net can decouple and effectively extract spatiotemporal multiscale information from chemical process data and achieve a approximately reduction of 30%in prediction error compared to other time-series models.Furthermore,its robustness and transferability underscore its promising potential for broader applications.

Modeling and Optimization of a Fractionation,Absorption,and Stabilization System in an Industrial Fluid Catalytic Cracking Process

Fluid catalytic cracking(FCC)is a vitally important refinery process.The fractionation,absorption,and stabilization system in the FCC process is a significant way to obtain key products,and its parameters will directly affect the quality of the products.In this work,using industrial data from an actual FCC process,a model of the FCC fractionation,absorption,and stabilization system was developed using process simulation software.The sequence quadratic program algorithm was then used to identify the parameters of each tower,increasing the accuracy of the simulation results.Next,using this improved model,a sensitivity analysis was performed to examine the effects of different operating conditions.The pattern-search method was then used to optimize the operating parameters of the system.The results showed that the optimized model has good prediction accuracy,and using the model,it was found that changing the operation parameters could result in a 1.84%improvement in economic benefits.As such,the developed model was demonstrated to be usefully applicable to the optimization of the process operation of an FCC fractionation,absorption,and stabilization system.

A new generic reaction for modeling fluid catalytic cracking risers

A new generic reaction in the form of PC_i→PC_m+[i,m]→PC_m+λi,m coke+surplusage has been proposed for describing the catalytic cracking behavior of petroleum narrow cuts or pseudo-components(PCs),where the rate constant formula is derived from the transition state theory and the coking amount is correlated to the properties of the intermediate substance [i,m].In composing the cracking reaction network for feedstock and product oils,only the product PC m of the proposed generic reaction is used,which together with a criterion for excluding exothermic reactions,distinctly reduces the number of reactions in the network.With the proposed cracking reaction scheme coupled with special pseudo-components,a predictive one-dimensional steady state model for fluid catalytic cracking risers is formulated in the sense that for a given riser and given catalyst,the model parameters are independent of stock oils,product schemes and other operational conditions.The great correlating and predicting capability of the resulted model is tested with production data in different scenarios of four commercial risers.

LVP Modeling and Dynamic Characteristics Prediction of A Hydraulic Power Unit in Deep-Sea

A hydraulic power unit （HPU） is the driving ＂heart＂ of deep-sea working equipment. It is critical to predict its dynamic performances in deep-water before being immerged in the seawater, while the experimental tests by simulating deep-sea environment have many disadvantages, such as expensive cost, long test cycles, and difficult to achieve low-temperature simulation, which is only used as a supplementary means for confirmatory experiment. This paper proposes a novel theoretical approach based on the linear varying parameters （LVP） modeling to foresee the dynamic performances of the driving unit. Firstly, based on the varying environment features, dynamic expressions of the compressibility and viscosity of hydranlic oil are derived to reveal the fluid performances changing. Secondly, models of hydraulic system and electrical system are accomplished respectively through studying the control process and energy transfer, and then LVP models of the pressure and flow rate control is obtained through the electro-hydraulic models integration. Thirdly, dynamic characteristics of HPU are obtained by the model simulating within bounded closed sets of varying parameters. Finally, the developed HPU is tested in a deep-sea imitating hull, and the experimental results are well consistent with the theoretical analysis outcomes, which clearly declare that the LVP modeling is a rational way to foresee dynamic performances of HPU. The research approach and model analysis results can be applied to the predictions of working properties and product designs for other deep-sea hydraulic pump.

Diving dynamics identification and motion prediction for marine crafts using field data

Ensuring accurate parameter identification and diving motion prediction of marine crafts is essential for safe navigation,optimized operational efficiency,and the advancement of marine exploration.Addressing this,this paper proposes an instrumental variable-based least squares(IVLS)algorithm.Firstly,aiming to balance complexity with accuracy,a decoupled diving model is constructed,incorporating nonlinear actuator characteristics,inertia coefficients,and damping coefficients.Secondly,a discrete parameter identification matrix is designed based on this dedicated model,and then a IVLS algorithm is innovatively derived to reject measurement noise.Furthermore,the stability of the proposed algorithm is validated from a probabilistic point of view,providing a solid theoretical foundation.Finally,performance evaluation is conducted using four depth control datasets obtained from a piston-driven profiling float in Qiandao Lake,with desired depths of 30 m,40 m,50 m,and 60 m.Based on the diving dynamics identification results,the IVLS algorithm consistently shows superior performance when compared to recursive weighted least squares algorithm and least squares support vector machine algorithm across all depths,as evidenced by lower average absolute error(AVGAE),root mean square error(RMSE),and maximum absolute error values and higher determination coefficient(R2).Specifically,for desired depth of 60 m,the IVLS algorithm achieved an AVGAE of 0.553 m and RMSE of 0.655 m,significantly outperforming LSSVM with AVGAE and RMSE values of 8.782 m and 11.117 m,respectively.Moreover,the IVLS algorithm demonstrates a remarkable generalization capability with R2 values consistently above 0.95,indicating its robustness in handling varied diving dynamics.

DISOPE distributed model predictive control of cascade systems with network communication

A novel distributed model predictive control scheme based on dynamic integrated system optimization and parameter estimation （DISOPE） was proposed for nonlinear cascade systems under network environment. Under the distributed control structure, online optimization of the cascade system was composed of several cascaded agents that can cooperate and exchange information via network communication. By iterating on modified distributed linear optimal control problems on the basis of estimating parameters at every iteration the correct optimal control action of the nonlinear model predictive control problem of the cascade system could be obtained, assuming that the algorithm was convergent. This approach avoids solving the complex nonlinear optimization problem and significantly reduces the computational burden. The simulation results of the fossil fuel power unit are illustrated to verify the effectiveness and practicability of the proposed algorithm.

CFD SIMULATION OF FLUID CATALYTIC CRACKING IN DOWNER REACTORS

A mathematical model has been developed for the simulation of gas-particle flow and fluid catalytic cracking in downer reactors. The model takes into account both cracking reaction and flow behavior through a four-lump reaction kinetics coupled with two-phase turbulent flow. The prediction results show that the relatively large change of gas velocity affects directly the axial distribution of solids velocity and void fraction, which significantly interact with the chemical reaction. Furthermore, model simulations are carried out to determine the effects of such parameters on product yields, as bed diameter, reaction temperature and the ratio of catalyst to oil, which are helpful for optimizing the yields of desired products. The model equations are coded and solved on CFX4.4.

A novel pore-fracture dual network modeling method considering dynamic cracking and its applications

Unconventional reservoirs are normally characterized by dual porous media, which has both multi-scalepore and fracture structures, such as low permeability or tight oil reservoirs. The seepage characteristicsof such reservoirs is mainly determined by micro-fractures, but conventional laboratory experimentalmethods are difficult to measure it, which is attribute to the dynamic cracking of these micro-fractures.The emerging digital core technology in recent years can solve this problem by developing an accuratepore network model and a rational simulation approach. In this study, a novel pore-fracture dualnetwork model was established based on percolation theory. Fluid flow in the pore of two scales, microfracture and matrix pore, were considered, also with the impact of micro-fracture opening and closingduring flow. Some seepage characteristic parameters, such as fluid saturations, capillary pressure, relative permeabilities, displacement efficiency in different flow stage, can be predicted by proposedcalculating method. Through these work, seepage characteristics of dual porous media can be achieved.

融合GA-Attention-LSTM算法的温室樱桃环境参数预测与裂果预警

针对温室环境因素对樱桃的影响,设计一套大樱桃温室环境自动监测装置,用来采集温室内的环境参数值为樱桃裂果提供数字化预警支持及防治方案。基于采集的环境参数值,首先使用相关性分析得出与棚内裂果具有强相关性的环境参数特征;其次使用滑动窗口方法将输入的环境特征生成时间序列矩阵形式;随后提出一种融合GA-Attention-LSTM算法的预测模型,实现精准预测棚内的环境参数的功能;最后通过SPSS数据分析软件来分析不同大棚的环境参数和裂果率。所提的融合GA-Attention-LSTM算法的预测模型的平均绝对误差为0.112,均方误差为0.087,相比于LSTM网络模型高出12.80%和9.72%,对环境参数的预测精度更高,同时得出一套科学的樱桃环境参数值范围,为预测模型对樱桃裂果数字化预警提供有力支持。

Soft Sensor for Inputs and Parameters Using Nonlinear Singular State Observer in Chemical Processes

Chemical processes are usually nonlinear singular systems.In this study,a soft sensor using nonlinear singular state observer is established for unknown inputs and uncertain model parameters in chemical processes,which are augmented as state variables.Based on the observability of the singular system,this paper presents a simplified observability criterion under certain conditions for unknown inputs and uncertain model parameters.When the observability is satisfied,the unknown inputs and the uncertain model parameters are estimated online by the soft sensor using augmented nonlinear singular state observer.The riser reactor of fluid catalytic cracking unit is used as an example for analysis and simulation.With the catalyst circulation rate as the only unknown input without model error,one temperature sensor at the riser reactor outlet will ensure the correct estimation for the catalyst circulation rate.However,when uncertain model parameters also exist,additional temperature sensors must be used to ensure correct estimation for unknown inputs and uncertain model parameters of chemical processes.

储粮仓仓壁动态侧压力的树模型预测方法

针对储粮仓卸料时仓壁动态侧压力难以准确预测的问题,利用机器学习方法中的树模型建立了仓壁动态侧压力预测模型。首先,分析了仓壁动态侧压力的主要影响因素为筒仓的结构尺寸、贮料的物理参数及测点位置。利用收集的496组仓壁动态侧压力数据,构建机器学习预测模型的数据集。然后,基于树模型,建立了仓壁动态侧压力的决策树(decision tree,DT)预测模型,在此基础上,利用Bagging算法和Boosting算法,建立了仓壁动态侧压力的随机森林(random forest,RF)预测模型和梯度提升树(gradient boosting decision tree,GBDT)预测模型。通过对比3种预测模型在测试集的均方误差(mean-square error,MSE)、决定系数和相对误差,表明GBDT预测模型的泛化性能最优。最后,通过开展模型试验和数值模拟,对GBDT预测模型进行验证,结果表明拟合良好。同时,根据树模型的分枝原理,判断出仓壁动态侧压力影响因素的重要性,得到对于贮料的物理参数,密度的重要性排第一;对于筒仓的结构尺寸,卸料口尺寸排第一。因此,在进行储粮仓设计时,建议优先考虑仓内散体物料的密度和仓的卸料口尺寸。

融合深度学习与过程机理的FCC装置关键参数软测量模型

产品产率作为催化裂化(FCC)装置的关键参数,构造其软测量模型对提升装置效益具有重要的现实意义,而原料与催化剂性质的缺失往往使得产率软测量模型性能迅速恶化。为此,以基于半监督学习的深度置信网络-极限学习机(DBN-ELM)算法为基础,将工艺过程机理模型与数据驱动模型集成,提出了可用于预测商业催化裂化装置产品产率的软测量混合建模方法。此外,还提出基于流程模拟的灵敏度分析-相关系数矩阵(SA-CCM)策略用于软测量模型主要输入变量的选择。结果表明,混合模型相比于数据驱动模型具有更优的模型性能,即预测精度提升43.9%、数据相关性(皮尔森系数)提升29.3%。这说明所提出的产率软测量混合建模方法使得模型的预测性能提高,能较好地适应原料与催化剂性质的变化。

特重冰区特高压直流孤立档导线脱冰动力响应参数预测模型

特重冰区特高压直流线路导线覆冰设计厚度达60~80 mm,多采用孤立档。该文针对特重冰区特高压直流六分裂导线,建立有限元模型,模拟研究不同档距、高差比、覆冰厚度、脱冰率和风速条件下孤立档导线脱冰动力响应,并分析其响应特征。给出导线冰跳高度包络线、有风情况下导线脱冰后横向摆幅参数和纵向不平衡张力的定义。结合利用数值模拟结果建立的数据集和BP(back propagation)神经网络机器学习算法,建立以档距、高差比、覆冰厚度、脱冰率和风速为输入,导线冰跳高度包络线、最大横向摆幅、最小横向摆幅、脱冰前平衡位置、孤立档两端纵向不平衡张力为输出的预测模型,为特重冰区特高压直流线路塔头设计提供依据。

永磁同步电机预测控制参数畸变影响分析

当永磁同步电机(PMSM)预测控制系统在高温等复杂环境中工作时,受电机齿槽效应、边端效应、饱和效应以及外部干扰的影响,电机参数如定子电阻、d轴电感、q轴电感、永磁体磁链常会发生畸变。针对上述问题,建立了PMSM的有限控制集模型预测电流控制(FCS-MPCC)模型,对不同参数畸变的影响进行分析,计算给出由畸变导致的d、q轴预测电流偏差。对参数畸变范围进行分区设置并依次施加在模型中。在不同畸变程度下,实验分析不同参数畸变对PMSM动、静态性能的影响,包括d、q轴电流、预测电流偏差、电磁转矩以及转速。

船用橡胶隔振器参数反演及预测模型应用

以BE型橡胶隔振器为研究对象,根据隔振器的额定参数,通过ABAQUS与ISIGHT联合仿真反演橡胶材料的超弹本构参数和黏弹本构参数,利用反演参数建立橡胶隔振器的预测模型。结果表明,预测模型垂、纵、横3个方向静刚度和3个方向动刚度的误差均小于7%;使用此预测模型研究橡胶隔振器频率、振幅和预载荷对动刚度的影响时,仿真结果与同类橡胶器件的试验结果基本一致;将此预测模型应用于系统级振动模拟时,仿真结果能较好地吻合振动试验结果。

基于CT技术的水泥稳定碎石微裂程度控制模型

为探究水泥稳定碎石材料内部细观空隙特征变化对微裂程度的影响,借助CT无损检测技术提取微裂后水泥稳定碎石试件空隙特征参数,采用灰熵关联度探讨水泥稳定碎石空隙参数与微裂程度关联性,建立空隙数量与微裂程度灰色关系模型。结果表明:延长振动时间,面空隙率呈现先增大后减小趋势、微裂程度呈现先增长后降低趋势。微裂后水泥稳定碎石试件相比未微裂试件空隙数量有所增加,且微裂作用对面积在0~0.1 mm^(2)范围内的空隙影响最为显著。圆度在0~0.2范围内空隙占比随振动时间延长逐渐增大,微裂后空隙分形维数均大于未微裂试件,且材料内部空隙形状趋于复杂化。空隙数量与微裂程度关联度最高。基于上述结果建立的微裂程度GM(1,2)预测模型可为水泥稳定碎石微裂技术细观层面研究和应用提供理论参考。

基于Weibull时间函数的采煤沉陷区地表动态沉陷预测模型

煤矿地表开采沉陷是一个复杂的时空过程,是矿山开采沉陷重点解决的问题。本文将时间参数引入到地表移动变形的动态预计中,分析了Weibull模型及其描述地表动态沉陷规律的适用性,并进行了参数分析,建立了地表动态沉陷预测模型。进行了实例分析。研究结果表明,参数n和k是该函数模型的重要参数,对描述地表动态移动变形量较为敏感;通过运用本文提出的地表动态移动变形预计模型,计算得到的地表下沉值相对误差均在7.0%以内,预计结果与实测结果较为吻合,预测模型可靠,预测结果可为相关减损措施的制定提供参考。

基于学习型模型预测控制的智能车辆路径跟踪控制

路径跟踪控制是智能车辆的一项关键技术。然而,现有车辆跟踪控制方法多依赖于较为精确的车辆控制模型,而实际的车辆控制系统多存在建模误差、参数摄动以及外部扰动等,显著影响路径跟踪控制精度。本文针对性地提出一种考虑车辆未建模动态的智能车辆学习型路径跟踪控制方法。首先建立车辆标称模型,并采用线性预言模型对车辆未建模动态进行近似补偿,以提高车辆模型的精度;然后基于扩展卡尔曼滤波原理实现对未建模动态参数的学习更新;之后构建考虑系统未建模动态的学习型模型预测控制器(LMPC);最后通过CarSim和Matlab/Simulink设计多工况多组别联合仿真试验,验证所提方法在提高路径跟踪精度方面的有效性。

An approach on dynamic earthquake prediction by georesistivitymeasurements

Experiences on earthquake prediction accumulated by the Chinese scientists in the last 20 years were synthetically analyzed. A prediction program was set up to demonstrate the development of the georesistivity anomaly by using of the dynamic image, accordingly the earthquake prone area can be recognized. By revising the DYBS Ⅰ, which was developed in 1989, and adding some latest achievements, we worked out a software on earthquake prediction by the geoelectric method the DYBS Ⅱ. Some new feature of DYBS Ⅱ are: the anomalous area may be determined by the space distribution and its time variation of geoelectric parameters; The dynamic process that is associated with the development of earthquake anomaly can be displayed on the computer screen; Technique for the prediction of an impending earthquake was included too. Some results of the Tangshan earthquake were presented at the end of this paper.

Aircraft parameter estimation using a stacked long short-term memory network and Levenberg-Marquardt method

To effectively estimate the unknown aerodynamic parameters from the aircraft’s flight data,this paper proposes a novel aerodynamic parameter estimation method incorporating a stacked Long Short-Term Memory(LSTM)network model and the Levenberg-Marquardt(LM)method.The stacked LSTM network model was designed to realize the aircraft dynamics modeling by utilizing a frame of nonlinear functional mapping based entirely on the measured input-output data of the aircraft system without requiring explicit postulation of the dynamics.The LM method combines the already-trained LSTM network model to optimize the unknown aerodynamic parameters.The proposed method is applied by using the real flight data,generated by ATTAS aircraft and a bio-inspired morphing Unmanned Aerial Vehicle(UAV).The investigation reveals that for the two different flight data,the designed stacked LSTM network structure can maintain the efficacy of the network prediction capability only by appropriately adjusting the dropout rates of its hidden layers without changing other network parameters(i.e.,the initial weights,initial biases,number of hidden cells,time-steps,learning rate,and number of training iterations).Besides,the proposed method’s effectiveness and potential are demonstrated by comparing the estimated results of the ATTAS aircraft or the bio-inspired morphing UAV with the corresponding reference values or wind-tunnel results.