Hybrid Modeling Method Research Articles

A Hybrid Modeling Method Based on Expert Control and Deep Neural Network for Temperature Prediction of Molten Steel in LF

A Hybrid Modeling Method Based on Expert Control and Deep Neural Network for Temperature Prediction of Molten Steel in LF

Flexural vibration of multistep rotating Timoshenko shafts using hybrid modeling and optimization techniques

In this paper, the flexural vibration of uniform and non-uniform rotating shafts based on the Timoshenko beam theory is investigated. Considering shear deformation and gyroscopic effects to build an exact framework of a solution, the fourth-order differential equation of vibration is solved by an analytical method. The overall transfer matrix of the system formed by a series of flexible distributed in addition to lumped elements is achieved. The results obtained by the distributed lumped modeling technique (DLMT) are verified with other techniques. The damped frequencies obtained by the hybrid modeling method for a multistep gas turbine rotor system using the Euler–Bernoulli and Timoshenko beam theories are compared with the results of the transfer matrix method (TMM). It is shown that the presented method provides highly accurate results, while with no compromise, it can be simply and effectively employed for complex systems. It is also shown how the Hooke and Jeeves and the ant colony optimization (ACO) besides the direct enumeration method can be applied to obtain the damped natural frequencies of complicated vibrational systems such as the gas turbine rotors. The comparison between the frequency and damping ratio values obtained by the optimization and direct enumeration methods shows less than 1% error for different cases.

Inverse Design and Optimization of Low Specific Speed Centrifugal Pump Blade Based on Adaptive POD Hybrid Model

To improve the prediction accuracy of the surrogate model and reduce the calculation cost for hydraulic optimization design of centrifugal pump impeller, an inverse design and optimization method based on adaptive proper orthogonal decomposition (APOD) hybrid model was proposed. Initial samples were designed by perturbing blade control parameters of the original model. The samples were classified using the K-means clustering algorithm, and the adaptive samples were selected according to the category of the objective sample. The snapshot set is composed of blade shape parameters and the CFD flow field data in impeller, which is decomposed into a linear combination of orthogonal bases by the proper orthogonal decomposition (POD) method to predict the objective parameters. According to the objective load distribution, the low specific speed centrifugal pump was inversely designed by using the APOD model, and its initial blade was obtained. And then, the flow field corresponding to disturbed blade shape was predicted using the APOD method, so as to evaluate the gradient of the objective function to design variables. Finally, the initial blade was optimized by the gradient descent method. The results show that the APOD hybrid model method can be employed to accomplish the blade inverse design and the flow field prediction in the optimization design of centrifugal impeller, which significantly reduces the numerical calculation cost and improves the accuracy of the flow field prediction.

Parameter optimal identification and dynamic behavior analysis of nonlinear model for the solution purification process of zinc hydrometallurgy

<p style='text-indent:20px;'>Impurity removal is a momentous part of zinc hydrometallurgy process, and the quality of products and the stability of the whole process are affected directly by its control effect. The application of dynamic model is of great significance to the prediction of key indexes and the optimization of process control. In this paper, considering the complex coupling relationship of stage II purification process, a hybrid modeling method of mechanism modeling and parameter identification modeling was proposed on the basis of not changing the actual production process of lead-zinc smeltery. Firstly, the overall nonlinear dynamic mechanism model was established, and then the deviation between the theoretical value and the actual detected outlet ion concentration was taken as the objective function to establish the parameter identification optimization model. Since the built model is nonlinear, it may pose implementation problems. On the premise of deriving the gradient vector and Hessian matrix of the objective function with respect to the parameter vector, an optimization algorithm based on the steepest descent method and Newton method is proposed. Finally, using the historical production data of a lead-zinc smeltery in China, the model parameters were accurately inversed. An intensive simulation validation and analysis of the dynamic characteristics about the whole model shows the accuracy and the potential of the model, also in the perspective of practical implementation, which provides the basis for the optimal control of system output and the guidance for the optimal control of zinc powder addition.</p>

A machine learning-based surrogate model for the rapid control of piping flow: Application to a natural gas flowmeter calibration system

With the notion of intelligent pipelines gaining ground in natural gas pipe networks, the requirements for fully automated daily operations are increasing. The flow control problem, especially in the complex station-level process piping network (SLPPN), has become the main challenge. This paper proposed a machine learning-based surrogate model that used the precise SLPPN flow control. A hybrid modeling method that uses data-driven method to surrogate numerical model was adopted to reveal the flow characteristics in the SLPPN, retaining the accuracy of the numerical simulation model while increasing its calculation speed simultaneously. Then, a two-step controller combining open-loop optimal control and closed-loop feedback control was designed to replace the classic proportional-integral-differential (PID) controller. First, genetic algorithms (GA) were used to determine the optimal control scheme of the surrogate model for fast simulation. Second, the follow-up PID controller was employed to compensate for the gap between the setpoint and the flow after the first control step. The controller was verified using the actual calibration station of a natural gas flowmeter in Guangzhou, China. The results showed that this two-step control method based on the surrogate model outperformed the conventional PID control technique regarding control duration and time, highlighting its potential in automating and intellectualizing the natural gas transportation industry.

Visualized calculation of regional power grid power data based on multiple linear regression equation

Abstract The power generation of distributed power generation systems in regional power grids is mainly affected by external environmental factors, and there are volatility and intermittent power generation issues. This paper establishes a relatively simple multiple linear regression prediction model to predict grid-connected photovoltaic power generation systems. The article uses a hybrid modelling method to filter more important weather forecast input variables through a step-by-step linear regression method to realize the visual calculation of power data. The research results show that the multiple linear regression equation methods proposed in the article have pervasive application value in the actual application of solar photovoltaic power consumption data.

A Hybrid Model Method for Accurate Surface Deformation and Incision Based on FEM and PBD

In some simulations like virtual surgery, an accurate surface deformation method is needed. Many deformation methods focus on the whole model swing and twist. Few methods focus on surface deformation. For the surface deformation method, two necessary characteristics are needed: the accuracy and real-time performance. Some traditional methods, such as position-based dynamics (PBD) and mass-spring method (MSM), focus more on the real-time performance. Others like the finite element method (FEM) focus more on the accuracy. To balance these two characteristics, we propose a hybrid mesh deformation method for accurate surface deformation based on FEM and PBD. Firstly, we construct a hybrid mesh, which is composed of a coarse volume mesh and a fine surface mesh. Secondly, we implement FEM on coarse volume mesh and PBD on fine surface mesh, and the deformation of fine surface mesh is constrained by the displacement of the coarse volume mesh. Thirdly, we introduced a small incision process for our proposed method. Finally, we implemented our method on a simple deformation simulation and a small incision simulation. The result shows an accurate surface deformation performance by implementing our method. The incision effect shows the compatibility of our proposed method. In conclusion, our proposed method acquires a better trade-off between accuracy and real-time performance.

A robust hybrid predictive model of mixed oil length with deep integration of mechanism and data

Accurate estimation of mixed oil length is highly required in multi-product pipelines because it can guide the operator to correctly handle the mixed oil segment and effectively reduce the loss of petroleum product quality. In previous study, a hybrid model combined with machine learning algorithm with existing mechanism has been developed and has good predictive accuracy. Unfortunately, due to incorrect measurement and improper recording, outliers are widely present in industrial datasets and may render the predictive performance of the previous model quite disappointing, while the effect of outliers on predictive models for the mixed oil length is rarely discussed. In order to deal with such issues, this paper first proposes a way to define the outlier sample and explicitly studies its impact on the performance of the predictive model for mixed oil prediction. Subsequentially, various new hybrid modeling methods are developed driven by both operation data (exploited by the Gradient Boosting Decision Tree algorithm) and the mechanism (based on the Austin-Palfrey equation) in different arrangements. Extensive experiments are conducted on real-life transportation pipelines, and the results show that with the clean training set, the R2 index of the proposed serial-parallel hybrid model (SPHM) is 0.96, which is higher than that of mechanism model and the existing hybrid model. Even with all the outliers added, advantage in prediction accuracy of the SPHM is still noticed, demonstrating feasibility and robustness of the hybrid modeling approach for prediction of mixed oil length.

Photovoltaic power prediction based on hybrid modeling of neural network and stochastic differential equation

With the increasing influence of new energy power system, the prediction of Photovoltaic (PV) output power becomes more and more important In this paper, it is the first time to put forward a hybrid modeling method combining long–short term memory recurrent neural network (LSTM) and stochastic differential equation (SDE). This method realizes the prediction of PV output power in different seasons and overcomes the uncertainty of PV power generation. Wavelet analysis and automatic encoder are used to decompose data and extract important features. According to the detailed signal sequence and the approximate signal sequence, the LSTM prediction model is established. Meanwhile, the mathematical model of SDE is established according to the detailed signal sequence. Finally, the output sequences of the two models are reconstructed by wavelet transform. This hybrid model can not only realize the point prediction of PV output power according to the predicted mean value, but also achieve the interval prediction under different confidence levels according to the randomness. In this paper, the proposed method is applied to predict the PV output power of CHINT photovoltaic power generation system with installed capacity of 10MW in different seasons, and the weather forecast data with errors of ±10%, ±20% and ±30% are used. Experimental results prove the effectiveness of the method. In the summer model considering forecast errors within ±20% of weather forecast data, the RMSEs of BP neural network, LSTM and convolutional neural network (CNN) are 5.9468, 5.6762 and 5.8004 respectively. However, the RMSE of the mean prediction with the confidence level of 90% under the proposed method is 4.4647. With this method, the results of interval prediction and point prediction of PV output power can provide better decision support for the stable and safe operation of PV grid connection. They have higher reference value for energy dispatching departments.

DC model for SiC MOSFETs using artificial neural network optimized by artificial bee colony algorithm

A DC model for silicon carbide (SiC) metal–oxide–semiconductor field effect transistors (MOSFETs) is proposed in this paper using a hybrid modeling method based on the artificial neural network and artificial bee colony (ABC) algorithm. A multi-layer perceptron neural network using the Levenberg–Marquardt (LM) method is applied to model SiC MOSFETs based on the data provided by the datasheet. The search strategy of artificial bees is improved based on the standard ABC, which enhances the search ability of the standard ABC. In view of the sensitivity of the LM method to the initial value, the improved ABC algorithm is adopted to help the neural network find initial weights and biases, which improves the accuracy of the modeling results. Comparing the modeling results with the I–V curves in the datasheet, the accuracy of the DC model is verified under different temperatures. In addition, the small signal parameters gm and gd that are not exposed in the training process also fit well with the datasheet, which fully demonstrates the feasibility of this hybrid modeling method.

Three-dimensional modeling of T-wave generation and propagation from a South Mid-Atlantic Ridge earthquake.

A three-dimensional (3D) hybrid modeling method is used to study the generation and propagation of T waves in the ocean triggered by a Southern Mid-Atlantic Ridge earthquake. First, a finite-element method model named SPECFEM3D is used to propagate seismic waves in the crust and acoustic waves in the ocean for the T-wave generation in a 200 × 50 km area near the epicenter. A 3D parabolic equation (PE) method is then used to propagate the T waves in the ocean for about 850 km further to the hydrophone stations deployed by the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) near Ascension Island. All of the simulations considered the realistic bathymetry and water sound speed profile. The SPECFEM3D results suggest that T waves with clear modal features could be generated by the concentration of reflected head waves in two depressions 40 km away from the epicenter. To compare with the hybrid modeling method for calculating T-wave propagation losses and arrival azimuths at the CTBTO hydrophones, point source simulations using the 3D PE model from the T waves source locations, identified with SPECFEM3D, were also implemented. The advantages and limitations of each approach are discussed.

A Hybrid Modeling Method Based on Linear AR and Nonlinear DBN-AR Model for Time Series Forecasting

A Hybrid Modeling Method Based on Linear AR and Nonlinear DBN-AR Model for Time Series Forecasting

Multimethodology for diagnosis and intervention in a prosthetics and orthotics factory in Brazil

AbstractThis paper develops an approach that articulates methods characterized as soft and hard Operational Research. This approach is known in the literature as Multimethodology, Hybrid Modeling, or Mixing Methods. As a Multi‐Paradigm Hybrid Study, the present research provides an interface between the qualitative and quantitative modeling of the problem addressed. In this way, it facilitates the engagement of managers with analysts in the search for solutions to problems. The applied objective is to improve the performance of production processes at a prosthetics and orthotics factory located in Rio de Janeiro, Brazil. It applies a multimethodology to support evaluation and intervention using knowledge mapping and a discrete‐event simulation model. Lessons learned from integrated workshops enabled the devising of an alternative improvement scenario for orthotics manufacturing. Theoretical results are validated through effective intervention, providing an increase in production capacity and reduction of the manufacturing lead time. This was made possible through the understanding and engagement of stakeholders during the modeling process.

Research on Adaptive Hybrid Energy Consumption Model Based on Data Driven under Variable Working Conditions

As to the nonlinear and time-varying problems of the energy consumption model, this paper proposes an adaptive hybrid modeling method. Firstly, the recursive least squares algorithm with adaptive forgetting factor based on fuzzy algorithm and recursive least squares algorithm is used to identify the simplified mechanism energy consumption model, which solves the data saturation phenomenon and the weights of the “old and new” data during the online identification process and guarantees the adaptability of the mechanism model. Secondly, because there is a deviation between the identified model and the simplified mechanism energy consumption model, the deviation compensation model of mechanism model is established through kernel partial least squares algorithm and the model updating strategy with sliding window, which is used to update the deviation compensation model, and then the adaptive hybrid model is established by combining with the mechanism model identified online and updated deviation compensation model. Finally, the effectiveness, generalization and adaptability of the model are verified by the actual operating data of a single working condition and variable working conditions. And comparing with the mechanism model and the data model, The comparison results show that the adaptive hybrid model has higher calculation accuracy with adaptation.

Reverse Design and Manufacturing of a Double Wolf Head Model Based on 3D Printing and Five-Axis Machining

The design and machining technology of curved and irregular surfaces has always been an important subject in the field of engineering. In the product design process, sequential design and reverse design approaches each have their own characteristics in the field of modeling. This paper puts forward a sequential/reverse hybrid modeling method based on combined Geomagic Wrap, Geomagic Design X and UG in order to improve the efficiency of modeling. The method takes into consideration the characteristics of both sequential and reverse designs and brings out their advantages. With the “double wolf head” model of complex structures as the object of study, the “double wolf head” is re-developed and remanufactured based on 3D printing and five-axis machining combined with reverse design, and data from the model are collected using a three-dimensional scanner and Geomagic Wrap. Reverse engineering to parametric modeling is generated using Geomagic Design X; this solid model is then imported to the 3D printing and slicing software Cura and the multi-axis programming software hyperMILL, respectively, before machining, and the 3D printing model and the five-axis machining model are obtained. This method explores a new approach to part modeling, which vastly improves the machining precision and accuracy of model data, accurately perceives the real size of an object and is of great significance to shorten product development cycle times.

Application of a Hybrid Modeling Method for Generating Synthetic Ground Motions in Fennoscandia, Northern Europe

ABSTRACT We present a modeling technique for generating synthetic ground motions, aimed at earthquakes of design significance for critical structures and ground motions at distances corresponding to the engineering near field, in which real data are often missing. We use dynamic modeling based on the finite-difference approach to simulate the rupture process within a fault, followed by kinematic modeling to generate the ground motions. The earthquake source ruptures were modeled using the 3D distinct element code (Itasca, 2013). We then used the complete synthetic program by Spudich and Xu (2002) to simulate the propagation of seismic waves and to obtain synthetic ground motions. In this work, we demonstrate the method covering the frequency ranges of engineering interests up to 25 Hz and quantify the differences in ground motion generated. We compare the synthetic ground motions for distances up to 30 km with a ground-motion prediction equation, which synthesizes the expected ground motion and its randomness based on observations. The synthetic ground motions can be used to supplement observations in the near field for seismic hazard analysis. We demonstrate the hybrid approach to one critical site in the Fennoscandian Shield, northern Europe.

Joint modeling using nonrigid cross-sections for beam-based analysis of a car body

A new joint modeling method for a beam-based analysis of a car body applied during the concept design phase is presented. The body-in-white (BIW) of a car mainly consists of thin-walled beams with closed cross-sections, the joints of which show a significant flexibility owing to higher-order deformation modes. The joint flexibility effect cannot be evaluated using conventional beam theories. Although a hybrid modeling method using shell elements for a joint region, condensed in the form of a superelement, can be employed, its increase in accuracy for the joint flexibility is marginal. This is because the interface of the shell-modeled joint and that of conventional beam elements is rigidly connected, and thus the flexibility of the shell elements of the joint disappears at the interface. In this investigation, a thin elastic layer is introduced to account for a flexible connection between the interfaces of the shell-modeled joint and those of the beam elements. By employing an absorption parameter, the stiffness of the thin layer is set such that the thin layer can represent not only the accurate flexibility of the shell-modeled joint, but also higher-order deformations such as distortion and warping of thin-walled beams that are connected to the joint. To determine the absorption parameter of a thin layer, a surrogate model is built with respect to the geometric parameters of the cross-section of a connected beam, joint angle, and wall thickness. The effectiveness of the proposed beam-joint modeling method is shown by solving two-beam joint problems with various cross-sections and the BIW of a passenger car.

Improved hybrid modeling method with input and output self-tuning for gas turbine engine

Gas-path model (GPM) plays a key role in the application of sensor fault tolerant control for the gas turbine engine (GTE). However, the modeling accuracy of traditional physics-based GPM is restricted by modeling inaccuracies and measurement uncertainty of sensors, etc., making fault-tolerant control of sensors difficult. In this paper, an improved input and output self-tuning hybrid modeling (IOSTHM) method is proposed for improving the GPM modeling accuracy of a dual shaft turbofan GTE. The proposed IOSTHM consists of an input self-tuning model (ISTM) and an output self-tuning model (OSTM). In the framework of ISTM, a fundamental traditional physics-based model (PBM) is constructed by using the component level modeling method firstly. Based on the PBM and the actual measurement of rotor speed of low-pressure shaft (N1) which has high measurement accuracy, the ISTM is conducted by tuning the measurement of fuel flow in real-time according to the deviation of N1 between engine and PBM, aiming to deal with the measurement uncertainty of fuel flow. Based on the ISTM, the IOSTHM is constructed by combining with an OSTM which is constituted by a bank of residual learning models to reduce the modeling inaccuracies further. The effectiveness of the proposed IOSTHM is evaluated with the verification of simulated flight data and actual ground test data. Both flight and ground verification results reveal that the proposed hybrid model IOSTHM achieves the best modeling performance when compared with other models such as traditional PBM, ISTM, and OSTM, it shows the superiority of this methodology.

Digital Twin Modeling of a Solar Car Based on the Hybrid Model Method with Data-Driven and Mechanistic

Solar cars are energy-sensitive and affected by many factors. In order to achieve optimal energy management of solar cars, it is necessary to comprehensively characterize the energy flow of vehicular components. To model these components which are hard to formulate, this study stimulates a solar car with the digital twin (DT) technology to accurately characterize energy. Based on the hybrid modeling approach combining mechanistic and data-driven technologies, the DT model of a solar car is established with a designed cloud platform server based on Transmission Control Protocol (TCP) to realize data interaction between physical and virtual entities. The DT model is further modified by the offline optimization data of drive motors, and the energy consumption is evaluated with the DT system in the real-world experiment. Specifically, the energy consumption error between the experiment and simulation is less than 5.17%, which suggests that the established DT model can accurately stimulate energy consumption. Generally, this study lays the foundation for subsequent performance optimization research.

A hybrid simulation model to study the impact of combined interventions on Ebola epidemic.

Pandemics have been recognized as a serious global threat to humanity. To effectively prevent the spread and outbreak of the epidemic disease, theoretical models intended to depict the disease dynamics have served as the main tools to understand its underlying mechanisms and thus interrupt its transmission. Two commonly-used models are mean-field compartmental models and agent-based models (ABM). The former ones are analytically tractable for describing the dynamics of subpopulations by cannot explicitly consider the details of individual movements. The latter one is mainly used to the spread of epidemics at a microscopic level but have limited simulation scale for the randomness of the results. To overcome current limitations, a hierarchical hybrid modeling and simulation method, combining mean-field compartmental model and ABM, is proposed in this paper. Based on this method, we build a hybrid model, which takes both individual heterogeneity and the dynamics of sub-populations into account. The proposed model also investigates the impact of combined interventions (i. e. vaccination and pre-deployment training) for healthcare workers (HCWs) on the spread of disease. Taking the case of 2014-2015 Ebola Virus Disease (EVD) in Sierra Leone as an example, we examine its spreading mechanism and evaluate the effect of prevention by our parameterized and validated hybrid model. According to our simulation results, an optimal combination of pre-job training and vaccination deployment strategy has been identified. To conclude, our hybrid model helps informing the synergistic disease control strategies and the corresponding hierarchical hybrid modeling and simulation method can further be used to understand the individual dynamics during epidemic spreading in large scale population and help inform disease control strategies for different infectious disease.