Guided Wave Radar Interface Detection with Sliding Window Peaks

The patent applies dynamics by transitioning from a static threshold-based detection method to a dynamic sliding window algorithm. The sliding window continuously adapts to local signal characteristics by comparing each point with its neighboring points within the window, allowing the detection threshold to vary dynamically based on local signal conditions rather than using a fixed global threshold. This enables accurate peak detection even when echo amplitudes fluctuate due to mixed or emulsified product layers.

The patent changes the detection parameter from a fixed amplitude threshold to a dynamic reference based on local signal statistics. The sliding window algorithm calculates the maximum value within the window and uses it as a dynamic reference point, changing the detection criterion from absolute amplitude comparison to relative comparison within the local context. This parameter transformation allows the system to adapt to varying echo amplitudes caused by emulsified interfaces.

The patent applies segmentation by dividing the continuous received signal into discrete sliding window segments. Each window processes a small portion of the signal independently, comparing points only within that local segment. This segmentation approach breaks down the complex task of analyzing the entire signal into simpler, manageable local comparisons, reducing overall computational complexity while maintaining high detection precision.

The patent uses partial action by implementing a sliding window that processes only a portion of the signal at a time rather than analyzing the entire signal simultaneously. The window moves through the signal in steps, performing partial detections at each position. This partial processing approach reduces computational load compared to global signal analysis while still achieving accurate peak detection through cumulative local assessments.