Scheduled Sampling Training Framework for ANN-Based PID Control

Proportional-Integral-Derivative (PID) controllers are extensively used in industrial control applications due to their simplicity and effectiveness in various control tasks. In recent years, there has been a growing emphasis on the integration of Artificial Neural Networks (ANNs) with control theory. This integration aims to harness the versatility of ANNs to facilitate controller design in dynamic environments. Additionally, it opens the possibility of transfer learning for these models, allowing knowledge gained from one control scenario to be applied to others' thereby enhancing adaptability and efficiency in control applications. In this work-in-progress paper, we propose a training framework that addresses the key challenges of modeling PIDs as ANNs, specifically the discrepancy between training and inference behaviors, known as exposure bias, commonly encountered in text summarization models. To tackle this, we integrate scheduled sampling, a technique devised for text sequence generation tasks, with an online control simulation environment.