Predictive-delay control based on real-time feedback scheduling

Abstract

Usually in the context of integrated control and real-time scheduling, quality of control improvement is mainly based on the dynamic measurement process and the system task scheduling parameters (i.e., sampling period, execution time). In this paper, a new feedback controller based on delay prediction is proposed to overcome the degradation of quality for multi-controller systems due to scheduling delays. A statistical analysis is performed to highlight the correlation between scheduling artifacts (delays and jitters) and quality of the control. It is stated that the input–output latency has a significant influence on the quality of control. Hence, it is proposed to reduce control impairments by using a prediction of the response time to update the output. A case study consisting of the control of three servo-motors is used to practically illustrate our statement. The consistency and effectiveness of the improvement are checked through the case of the control of three inverted pendulums.