Industrial time series, as a kind of data that responds to production process information, can be analyzed and predicted for effective monitoring of industrial production processes. There are problems of data shortage...
详细信息
Industrial time series, as a kind of data that responds to production process information, can be analyzed and predicted for effective monitoring of industrial production processes. There are problems of data shortage and algorithm cold start in industrial modelingprocess caused by complex working conditions, change of data acquisition environment, and short running time of equipment. As a result, the accuracy of the existing data-driven industrial time series prediction algorithm is greatly limited. To address the aforementioned problems, we propose a new time series prediction method for industrial processes under limited data based on dynamic transfer learning in this work. This method aims to effectively use historical data of similar equipment or working conditions rather than discard them to help establish an industrial time series prediction model with limited target data. In this method, first, historical data are divided into multiple batches, and then a new multisource transfer learning framework with dynamic maximum mean difference loss is established according to the distribution distance between each batch of historical data and the limited target data at the current moment. The framework also combines multitask learning methods to establish multistep prediction model for online learning in industrial processes. Compared with other commonly used methods, experiments on two real-world datasets of solar power generation prediction and heating furnace temperature prediction demonstrate the effectiveness of the proposed method.
Folding rudders are widely used in multiple aircraft platforms. Now the simulation analysis of folding rudders focuses on the calculation of structural strength and impact load, and lacks the analysis of the dynamic p...
详细信息
Accurate reconstruction of unknown external forces from measurable responses is critical for ensuring structural safety and minimizing maintenance costs of aircraft structures. This paper presents a novel multitask-tr...
详细信息
Accurate reconstruction of unknown external forces from measurable responses is critical for ensuring structural safety and minimizing maintenance costs of aircraft structures. This paper presents a novel multitask-transfer-learning method for random-force frequency identification that accounts for modeling and measurement uncertainties. A data-driven convolutional neural network (CNN) model is utilized to capture the relationship between the power spectral densities of external forces and measured responses, addressing the inherent ill-posedness of traditional model-driven force identification methods. To shorten the frequency-dependent training time in the full frequency domain, a transfer-learning strategy is implemented, fine-tuning hyperparameters from a CNN model trained at one source frequency to another target frequency. Furthermore, an iterative dimensionwise collocation method based on nonprobabilistic interval modeling is introduced to quantify the uncertain boundaries of external loads caused by multisource uncertainties. By incorporating a multitask-learning framework, the process of establishing CNN models for collocated samples is accelerated, reducing the computational effort for uncertainty quantification. The proposed method is validated through both numerical and experimental examples, demonstrating its accuracy, robustness, and computational efficiency for force identification in the full frequency domain, even under conditions of insufficient measurements, measurement noises, and material dispersions.
DAB converters under triple phase shift (TPS) modulation can broaden the zero voltage switching (ZVS) range to improve efficiency and noise robustness. Conventionally, the piecewise approach and harmonic approach are ...
详细信息
ISBN:
(纸本)9781665475396
DAB converters under triple phase shift (TPS) modulation can broaden the zero voltage switching (ZVS) range to improve efficiency and noise robustness. Conventionally, the piecewise approach and harmonic approach are two commonly used approaches to build analytical models for ZVS conditions under this modulation strategy. However, both two approaches fail to achieve good modeling accuracy as well as low computational cost simultaneously due to heavy human dependence. To solve this problem, this digest proposes a data-driven modeling approach for ZVS analysis (DM-ZVS) for non-resonant DAB converters under the TPS modulation strategy. The data-driven modelingprocess is conducted with the random forest algorithm automatically using ZVS performance data from simulation tools, greatly mitigating human dependence to improve accuracy and computational efficiency. With the trained data-driven classification model of ZVS, the optimal TPS modulation parameters can be found to ensure the full ZVS range. A design case is given, and 1 kW hardware experiments comprehensively validate the feasibility of the proposed DM-ZVS.
The objective of this study is to integrate two main workflows seamlessly;i.e. hydraulic fracture modeling and dynamic reservoir simulation process. A smooth coupling is developed and implemented for unconventional re...
Steady-state tire models fail to capture elastic hysteresis effects, leading to significant modeling errors under transient conditions. In trajectory tracking tasks, such errors may lead to controller overresponse or ...
详细信息
Steady-state tire models fail to capture elastic hysteresis effects, leading to significant modeling errors under transient conditions. In trajectory tracking tasks, such errors may lead to controller overresponse or underresponse, seriously compromising vehicle safety and stability. To address this challenge, this study proposes a data-driven framework for discovering interpretable, physics-informed tire dynamic models under transient conditions. First, transient data are extracted using the Latin hypercube sampling, targeting the relatively scarce data in transient scenarios. Next, the sampling process in candidate library sampling-sparse identification of nonlinear dynamics is enhanced by incorporating physics-informed priors during candidate feature expansion, thereby improving model interpretability. High-frequency feature selection is then performed on the nonzero sparse coefficients to construct the candidate function library and identify the transient tire model. Finally, the identified model is integrated into a nonlinear model predictive controller for trajectory tracking and validated via both real vehicle experiments and hardware-in-the-loop simulations. The experimental results demonstrate significant improvements in tire force prediction accuracy, trajectory prediction, and tracking performance with the proposed transient tire model, compared to steady-state models. These improvements are observed under transient conditions, including slalom maneuvers and emergency double-lane changes on the ice-snow road.
In the task of system analysis for VSG cluster, aggregation modeling method is widely used for simplification. However, there are inevitable errors occur from the process of cluster aggregation. To improve the accurac...
详细信息
In the task of system analysis for VSG cluster, aggregation modeling method is widely used for simplification. However, there are inevitable errors occur from the process of cluster aggregation. To improve the accuracy of VSG cluster modeling, a data-physical driven modeling method is presented. At first, the equivalence between aggregation error and black box modeling issue is analyzed. Secondly, a hybrid model structure is proposed, which consists of single machine aggregation model and deep neural network based aggregated-error model. Then, to illustrate the modeling procedure, test cases are studied under large disturbance and multi-operating points conditions. The simulation results confirm that the proposed method can provide satisfactory modeling accuracy. (c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://***/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the The 2nd International conference on Power Engineering, ICPE, 2021.
Space launch failure is an uncommon but extremely important real-world problem. This paper proposes a Hierarchical Timed Colored Petri nets (HTCPN) simulation and validation method for the emergency response plan in r...
详细信息
A fractional-order model of a photovoltaic (PV) system is proposed in this paper. The system identification approach is used to develop an effective dynamical model for a PV system. A real PV module and a boost conver...
详细信息
A fractional-order model of a photovoltaic (PV) system is proposed in this paper. The system identification approach is used to develop an effective dynamical model for a PV system. A real PV module and a boost converter are used to gather the experimental input-output data for the identification process. The black box modeling is applied to the system identification to obtain the transfer function, without the requirement to perform any mathematical analysis. The identification process is based on the least squares criteria for minimizing model output error, and the Levenberg-Marquardt algorithm is used for optimizing the model parameters. The proposed fractional-order model (FOM) is investigated using MATLAB to study the frequency response and the model's stability. Simulation results verify the effectiveness and advantages of the FOM in comparison to identified integer-order model.
A description of the PRIM-AES software package (NPP simulator program) for modeling neutronic and thermal-hydraulic processes in transient modes at power units with VVER reactors is presented. The package includes a d...
详细信息
A description of the PRIM-AES software package (NPP simulator program) for modeling neutronic and thermal-hydraulic processes in transient modes at power units with VVER reactors is presented. The package includes a dynamic-simulation environment for the implementation of complex algorithms for automatic control systems based on SimInTech CAD systems and the TIGR-M program for combined neutronic and thermal-hydraulic calculation of normal operation and emergency modes, which is a development of the TIGR-1 software package. The modernization of the thermal-hydraulic module of the TIGR-1 software package is described, which is carried out to extend the applicability of the code in the volume of the power unit. The structure of a new software-computing package with a description of its constituent modules and the interface for data exchange between them has been developed and presented. The goals and objectives of the application of the developed software package are determined, which allow assessing the relevant criteria of dynamic stability for stationary modes of normal operation and transient modes of normal operation, including power maneuvering modes, and operation limitation modes in the case of violation of safe operation conditions and emergency modes of the second category involving failures of the main equipment.
暂无评论