PHD Filter Intensity Update Using Sensor Track ID Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current unmanned aircraft systems (UAS) face challenges in accurately tracking multiple intruder aircraft for self-separation and collision avoidance due to limitations in detecting and correlating measurements across different sensors, leading to inefficiencies in updating intensities within probabilistic hypothesis density (PHD) filters.
Innovation Solution
The method involves using track identifiers (IDs) provided by sensors to correlate and update intensities in a PHD filter, allowing for accurate tracking of multiple objects by matching sensor-specific IDs with measurement statistics and applying statistical distance tests to determine updates, thereby improving the accuracy and efficiency of tracking multiple objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional PHD filter intensity updating methods are used without sensor track ID correlation, then the filtering process is simpler, but tracking accuracy deteriorates due to incorrect intensity updates from uncorrelated sensor measurements
Solution Approach 1:
The patent applies preliminary action by pre-establishing track ID correlations between sensors before the PHD filter updating process. The system pre-processes sensor measurements to correlate track IDs across different sensors, so that when intensity updates are needed, the correlation structure is already in place. This eliminates the need for complex real-time correlation calculations during filter updating, thereby improving tracking accuracy without proportionally increasing filtering complexity.
Solution Approach 2:
The patent uses track ID as an intermediary element to bridge sensor measurements and PHD filter intensities. Instead of directly correlating all sensor measurements with all track intensities (which would be computationally complex), the track ID serves as a mediator that efficiently links specific measurements to their corresponding intensities. This intermediary approach maintains high tracking accuracy while reducing the computational burden of the filtering process.
2Reliability
If sensor measurements are processed without track ID correlation, then processing speed is faster, but tracking reliability deteriorates due to incorrect measurement-to-track associations
Solution Approach 1:
The system performs preliminary track ID correlation processing on sensor measurements before they enter the main PHD filter processing pipeline. By pre-establishing the association between track IDs and sensor measurements, the main processing loop can operate efficiently without repeated correlation calculations, thus maintaining high processing speed while ensuring reliable tracking associations.
Solution Approach 2:
The patent implements self-service by having the sensor processing system automatically generate and attach track ID correlation information to its measurements. This self-generated metadata enables the PHD filter to perform reliable intensity updates without requiring complex external correlation logic, thereby maintaining processing speed while improving tracking reliability through accurate measurement-to-track associations.
3Measurement precision
If all sensor measurements are used to update all track intensities, then comprehensive tracking is achieved, but computational efficiency deteriorates due to unnecessary calculations
Solution Approach 1:
The patent applies segmentation by dividing the intensity updating process into distinct segments based on track ID correlations. Instead of having all sensor measurements update all track intensities (a dense, computationally expensive operation), the system segments the updates so that each measurement only updates the specific intensity associated with its track ID. This segmented approach maintains comprehensive tracking coverage while dramatically improving computational efficiency by eliminating unnecessary calculation pairs.
Data Source
AI summary
In one embodiment, a method of tracking multiple objects with a probabilistic hypothesis density filter is provided. The method includes obtaining measurements corresponding to a first object with at least one sensor, the at least one sensor providing one or more first track IDs for the measurements. A Tk+1 first predicted intensity is generated for the first object based on a Tk first track intensity. A Tk+1 measurement from a first sensor of the at least one sensors is obtained, the first sensor providing a second track ID for the Tk+1 measurement. The second track ID is compared to the one or more first track IDs, and the Tk+1 first predicted intensity is selectively updated with the Tk+1 measurement based on whether the second track ID matches any of the one or more first track IDs to generate a Tk+1 first measurement-to-track intensity for the first object.