Particle Filtering Navigation Using Measurement Correlation on FPGA
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems for aircraft face challenges in correcting inertial drift due to measurement errors, particularly when using box-regularized particle filters, which require significant computing resources that are not feasible with current FPGA technology, and are sensitive to terrain non-linearities and ambiguities.
Innovation Solution
A box-regularized particle filtering method with random jamming processes based on an Epanechnikov kernel, implemented using a computing unit of the FPGA type, which generates random modifications to state intervals to reduce computing requirements while maintaining accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If box-regularized particle filtering is used to correct inertial drift, then navigation accuracy is improved, but computing resource requirements increase beyond current FPGA capabilities
Solution Approach 1:
The particle filtering process is segmented into distinct functional blocks (prediction block, update block, selection block) that can be independently implemented and optimized on FPGA hardware. Each block processes specific aspects of the filtering algorithm, allowing parallel execution and reducing overall computing resource requirements while maintaining navigation accuracy.
Solution Approach 2:
The invention modifies the particle filtering parameters and mathematical operations to be more suitable for FPGA implementation. This includes using fixed-point arithmetic instead of floating-point, optimizing the number of particles, and adjusting the regularization parameters to achieve acceptable accuracy with reduced computational complexity that fits within current FPGA capabilities.
2Extent of automation
If terrain correlation navigation is used to correct inertial drift, then autonomous navigation is improved, but sensitivity to terrain non-linearities and ambiguities increases
Solution Approach 1:
The system implements feedback mechanisms where the particle filter continuously compares predicted terrain measurements with actual telemetry sensor measurements. This feedback loop allows the system to detect and correct for terrain non-linearities and ambiguities by adjusting particle weights and selecting particles that are consistent with observed terrain, thereby improving reliability while maintaining autonomous navigation.
Solution Approach 2:
The system performs preliminary terrain characterization and stores relief map data before navigation. By having preprocessed terrain information available, the system can more effectively correlate predicted positions with actual terrain features, reducing sensitivity to non-linearities and ambiguities during the navigation process.
Data Source
AI summary
Disclosed is a box-regularized particle filtering process which includes an Epanechnikov kernel smoothing step. For this purpose, the process uses a special method for generating random numbers that follow an Epanechnikov probability density function. The process can be performed autonomously in a navigation system using correlation measurement, in particular on board an aircraft such as an aircraft, a flying drone or any self-propelled aerial carrier.