Figure - available from: Journal of Intelligent Manufacturing
This content is subject to copyright. Terms and conditions apply.
Source publication
Many Data-driven decisions in manufacturing need accurate and reliable predictions. Due to high complexity and variability of working conditions, a prediction model may deteriorate over time after deployed. Traditional performance evaluation indexes mainly assess the prediction model from a static perspective, which is difficult to meet the actual...
Similar publications
To accommodate the rapid development of the distribution network of China, it is essential to research load forecasting methods with higher accuracy and stronger generalization capabilities in order to optimize distribution system control strategies, ensure the efficient and reliable operation of the power system, and provide a stable power supply...
Citations
... The research on this theme covers various aspects, including online performance evaluation, hybrid federated learning, and explainable AI for fault diagnosis The first two papers on this theme covered data-driven predictive maintenance. Shen et al. (2024) design an online prediction performance evaluation index (OPPEI) to evaluate prediction models in terms of accuracy, degradation speed, and stability. The authors also propose a model maintenance evaluation method based on principal component analysis (PCA) for proactive maintenance. ...
... To this end, the developed predictive model often requires frequent on-site retraining or tuning by human experts. The reason is that in real-world manufacturing, the underlying distribution that generates the data samples of the process, Pr(x, y) = Pr(x) · Pr(y|x), tends to evolve over time Shen et al., 2024). It is inefficient to purely rely on humans to inspect such evolution and maintain the predictive model while interrupting the normal usage. ...
Real-time prediction of future process outputs is critical for the model predictive control of continuous manufacturing processes. It helps identify when and how to adjust the process variables under the disturbances. A lot of recurrent neural network-based predictive models have been developed and validated on the simulated processes. Based on that, some works further consider the existence of multiple operating conditions that can be unforeseen and transitive in real-world manufacturing. However, their designed online learning mechanisms mostly focus on fast local tracking without preserving important old knowledge. Besides, the proactive input data adaptation is largely unexplored. To bridge this gap, a novel self-adaptation mechanism is proposed in this paper. This mechanism can be easily integrated into different choices of predictive model to improve the stability of performance towards various changes in a long period of manufacturing. In the proposed mechanism, the components of adaptive sequence filtering and adaptive input normalization first extract the compact and properly scaled features from the growing multivariate time series subject to delayed output response and non-stationarity. Based on the encoder-decoder network as an exemplary predictive model, the component of adaptive model update consists of a non-Euclidean loss for evaluating sequential predictions and a task-free knowledge consolidation strategy for continual learning-based regularization. The application to an industrial rotary drying process is demonstrated, where data streams are collected from four production lines over 14 months. Extensive comparative study shows the superior performance of proposed mechanism and ablation study further verifies the effectiveness of each individual component.