Adaptive Confirmation Method for Programmed Remote Control Results in Centralized Control Stations Based on Multi-source Evidence Fusion

To address the limitations of fixed thresholds and discrete position signals in programmed remote-control confirmation, an adaptive method based on time-series features observed before and after remote-control execution is proposed. A robust baseline representing the operating condition before command execution is established using the median and median absolute deviation. The new steady state following command execution is identified using sliding-window statistics, and multi-source evidence from status signals, analog measurements, secondary criteria, video recognition, and topology consistency is fused with data-quality factors. Hard-constraint gating is combined with soft-evidence confidence fusion to prevent explicit contradictory signals from being masked. In 1,000 synthetically constructed scenarios, the proposed method achieved an overall accuracy of 98.50% with no false confirmations, demonstrating the feasibility of combining position-based hard constraints with adaptive comparison of measurements taken before and after remote-control execution.