Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Diff-MTS: Temporal-Augmented Conditional Diffusion-based AIGC for Industrial Time Series Towards the Large Model Era

Lei Ren, Haiteng Wang, Yuanjun Laili

TL;DR

Results show that Diff-MTS facilitates the generation of industrial data, contributing to intelligent maintenance and the construction of industrial large models, and performs substantially better in terms of diversity, fidelity, and utility compared with the GAN-based methods.

Abstract

Industrial Multivariate Time Series (MTS) is a critical view of the industrial field for people to understand the state of machines. However, due to data collection difficulty and privacy concerns, available data for building industrial intelligence and industrial large models is far from sufficient. Therefore, industrial time series data generation is of great importance. Existing research usually applies Generative Adversarial Networks (GANs) to generate MTS. However, GANs suffer from unstable training process due to the joint training of the generator and discriminator. This paper proposes a temporal-augmented conditional adaptive diffusion model, termed Diff-MTS, for MTS generation. It aims to better handle the complex temporal dependencies and dynamics of MTS data. Specifically, a conditional Adaptive Maximum-Mean Discrepancy (Ada-MMD) method has been proposed for the controlled generation of MTS, which does not require a classifier to control the generation. It improves the condition consistency of the diffusion model. Moreover, a Temporal Decomposition Reconstruction UNet (TDR-UNet) is established to capture complex temporal patterns and further improve the quality of the synthetic time series. Comprehensive experiments on the C-MAPSS and FEMTO datasets demonstrate that the proposed Diff-MTS performs substantially better in terms of diversity, fidelity, and utility compared with GAN-based methods. These results show that Diff-MTS facilitates the generation of industrial data, contributing to intelligent maintenance and the construction of industrial large models.

Diff-MTS: Temporal-Augmented Conditional Diffusion-based AIGC for Industrial Time Series Towards the Large Model Era

TL;DR

Results show that Diff-MTS facilitates the generation of industrial data, contributing to intelligent maintenance and the construction of industrial large models, and performs substantially better in terms of diversity, fidelity, and utility compared with the GAN-based methods.

Abstract

Industrial Multivariate Time Series (MTS) is a critical view of the industrial field for people to understand the state of machines. However, due to data collection difficulty and privacy concerns, available data for building industrial intelligence and industrial large models is far from sufficient. Therefore, industrial time series data generation is of great importance. Existing research usually applies Generative Adversarial Networks (GANs) to generate MTS. However, GANs suffer from unstable training process due to the joint training of the generator and discriminator. This paper proposes a temporal-augmented conditional adaptive diffusion model, termed Diff-MTS, for MTS generation. It aims to better handle the complex temporal dependencies and dynamics of MTS data. Specifically, a conditional Adaptive Maximum-Mean Discrepancy (Ada-MMD) method has been proposed for the controlled generation of MTS, which does not require a classifier to control the generation. It improves the condition consistency of the diffusion model. Moreover, a Temporal Decomposition Reconstruction UNet (TDR-UNet) is established to capture complex temporal patterns and further improve the quality of the synthetic time series. Comprehensive experiments on the C-MAPSS and FEMTO datasets demonstrate that the proposed Diff-MTS performs substantially better in terms of diversity, fidelity, and utility compared with GAN-based methods. These results show that Diff-MTS facilitates the generation of industrial data, contributing to intelligent maintenance and the construction of industrial large models.
Paper Structure (22 sections, 16 equations, 5 figures, 6 tables, 2 algorithms)

This paper contains 22 sections, 16 equations, 5 figures, 6 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Industrial Multivariate Time Series Synthesis
  4. Diffusion Models
  5. Temporal-augmented Conditional Adaptive Diffusion Model for ITMS Generation
  6. Denoised Diffusion Probabilistic Model
  7. Conditional Ada-MMD Diffusion Method
  8. Temporal Decomposition Reconstruction UNet
  9. Input Embedding Module
  10. Temporal Decomposition Reconstruction UNet Module
  11. Training and Sampling Procedure
  12. Experiment
  13. Experimental Data and Setup
  14. Dataset
  15. Evaluation Metrics
  16. ...and 7 more sections

Figures (5)

  • Figure 1: Illustration for diffusion process and reverse process. During diffusion process, noise is gradually added to the original signal to become Gaussian noisy signal. In the reverse process, the noisy signal is recovered to the original signal by estimating the added noise.
  • Figure 2: Overview of the Temporal Decomposition Reconstruction UNet Model.
  • Figure 3: Structure of the UNet encoder and decoder.
  • Figure 4: PCA visualization (1$^{\rm st}$ row) and t-SNE visualization (2$^{\rm nd}$ row) on FD001 dataset. Each column presents the two visualizations of the four methods. Blue represents the original samples, and brown represents the synthesis samples. Two points overlapping more means that the two data types are more similar.
  • Figure 5: Several visualizations of prediction results. Black curves represent the original time series. The gold, blue and green curves represent the synthetic time series of Diff-MTS, DiffWave and TimeGAN, respectively.