Continuous-Time Prediction of Industrial Paste Thickener System With Differential ODE-Net

Yuan, Zhaolin; Li, Xiaorui; Wu, Di; Ban, Xiaojuan; Wu, Nan; Dai, Hong-Ning; Wang, Hao

doi:10.1109/jas.2022.105464

Cited by 11 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…See the next section for the dimensional analysis. Time series data contains a lot of uncertain information, and the forecasting effect of applying a single model is often not very satisfactory [25]. Therefore, in order to improve the accuracy of prediction, the LSTM and prophet models are combined.…”

Section: Prophet-lstm Combined Model Constructionmentioning

confidence: 99%

Time Series Forecasting Fusion Network Model Based on Prophet and營mproved LSTM

Liu¹,

Yu²,

Zhao³

et al. 2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

Time series forecasting and analysis are widely used in many fields and application scenarios. Time series historical data reflects the change pattern and trend, which can serve the application and decision in each application scenario to a certain extent. In this paper, we select the time series prediction problem in the atmospheric environment scenario to start the application research. In terms of data support, we obtain the data of nearly 3500 vehicles in some cities in China from Runwoda Research Institute, focusing on the major pollutant emission data of non-road mobile machinery and high emission vehicles in Beijing and Bozhou, Anhui Province to build the dataset and conduct the time series prediction analysis experiments on them. This paper proposes a P-gLSTNet model, and uses Autoregressive Integrated Moving Average model (ARIMA), long and short-term memory (LSTM), and Prophet to predict and compare the emissions in the future period. The experiments are validated on four public data sets and one self-collected data set, and the mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) are selected as the evaluation metrics. The experimental results show that the proposed P-gLSTNet fusion model predicts less error, outperforms the backbone method, and is more suitable for the prediction of time-series data in this scenario.

show abstract

Section: Prophet-lstm Combined Model Constructionmentioning

confidence: 99%

Time Series Forecasting Fusion Network Model Based on Prophet and營mproved LSTM

Liu¹,

Yu²,

Zhao³

et al. 2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

“…Moreover, ODE-Recurrent Neural Network (ODE-RNN) (Rubanova, Chen, and Duvenaud 2019) inserts an ODE-Net module between adjacent RNN updates for modeling the continuous-time evolution of hidden states. In order to improve the longterm prediction, previous studies such as (Demeester 2020;Yuan et al 2022) incorporate the advantages of recurrent neural networks and differential equations to tackle the unit root problem and propose the Time-Aware RNN. The abovementioned models are capable of tackling the problem of unevenly sampling though they still have limitations in modeling stochastic systems because their internal states evolution is essentially deterministic.…”

Section: Related Workmentioning

confidence: 99%

ODE-RSSM: Learning Stochastic Recurrent State Space Model from Irregularly Sampled Data

Yuan

Ban

Zhang

et al. 2023

AAAI

View full text Add to dashboard Cite

For the complicated input-output systems with nonlinearity and stochasticity, Deep State Space Models (SSMs) are effective for identifying systems in the latent state space, which are of great significance for representation, forecasting, and planning in online scenarios. However, most SSMs are designed for discrete-time sequences and inapplicable when the observations are irregular in time. To solve the problem, we propose a novel continuous-time SSM named Ordinary Differential Equation Recurrent State Space Model (ODE-RSSM). ODE-RSSM incorporates an ordinary differential equation (ODE) network (ODE-Net) to model the continuous-time evolution of latent states between adjacent time points. Inspired from the equivalent linear transformation on integration limits, we propose an efficient reparameterization method for solving batched ODEs with non-uniform time spans in parallel for efficiently training the ODE-RSSM with irregularly sampled sequences. We also conduct extensive experiments to evaluate the proposed ODE-RSSM and the baselines on three input-output datasets, one of which is a rollout of a private industrial dataset with strong long-term delay and stochasticity. The results demonstrate that the ODE-RSSM achieves better performance than other baselines in open loop prediction even if the time spans of predicted points are uneven and the distribution of length is changeable. Code is availiable at https://github.com/yuanzhaolin/ODE-RSSM.

show abstract

“…Ding and Jiang provided an RFID-based production data analysis method for production control in IoT-enabled smart job shops [28]. Yuan et al established an integrated, deep-learning, continuous time network structure that consists of a sequential encoder, a state decoder and a derivative module to learn the deterministic state-space model from thickening systems [29].…”

Section: Data Acquisition and Analysis Of Im Workhopsmentioning

confidence: 99%

Data Acquisition Network Configuration and Real-Time Energy Consumption Characteristic Analysis in Intelligent Workshops for Social Manufacturing

Zhang

et al. 2022

Machines

View full text Add to dashboard Cite

To achieve energy-saving production, one critical step is to calculate and analyze the energy consumption and energy efficiency of machining processes. However, considering the complexity and uncertainty of discrete manufacturing job shops, it is a significant challenge to conduct data acquisition and energy consumption data processing of manufacturing systems. Meanwhile, under the growing trend of personalization, social manufacturing is an emerging technical practice that allows prosumers to build individualized services with their partners, which produces new requirements for energy data processing. Thus, a real-time energy consumption characteristic analysis method in intelligent workshops for social manufacturing is established to realize data processing and energy efficiency evaluation automatically. First, an energy-conservation production architecture for intelligent manufacturing processes is introduced, and the configuration of a data acquisition network is described to create a ubiquitous manufacturing environment. Then, an energy consumption characteristic analysis method is proposed based on the process time window. Finally, a case study of coupling-part manufacturing verifies the feasibility and applicability of the proposed method. This method realizes a combination of social manufacturing and real-time energy characteristic analysis. Meanwhile, the energy consumption characteristics provide a decision basis for the energy-saving control of intelligent manufacturing workshops.

show abstract

Continuous-Time Prediction of Industrial Paste Thickener System With Differential ODE-Net

Cited by 11 publications

References 39 publications

Time Series Forecasting Fusion Network Model Based on Prophet and營mproved LSTM

Time Series Forecasting Fusion Network Model Based on Prophet and營mproved LSTM

ODE-RSSM: Learning Stochastic Recurrent State Space Model from Irregularly Sampled Data

Data Acquisition Network Configuration and Real-Time Energy Consumption Characteristic Analysis in Intelligent Workshops for Social Manufacturing

Contact Info

Product

Resources

About