Spectral Merging of Sonar Tracks via Harmonic Comb Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data merging techniques face challenges in handling multiple sensors with varying noise levels and absence of data, leading to complexity in implementing algorithms, especially when considering multiple proximity criteria, and often require operator intervention or training data.
Innovation Solution
A statistical approach for data merging that automatically groups initial tracks based on azimuth and frequency criteria, using a method that iteratively merges tracks in azimuth and refines results through spectral analysis to identify real objects, without requiring training data or operator intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple proximity criteria (azimuth and frequency) are considered for data merging, then the accuracy of object identification is improved, but the complexity of implementing the merging algorithms increases excessively
Solution Approach 1:
The patent segments the data merging process into two distinct steps: first spatial association based on azimuth proximity to form merged tracks, then frequency analysis to identify harmonic combs. This segmentation allows each step to focus on a single criterion, reducing overall algorithmic complexity while maintaining the benefits of multi-criteria analysis.
Solution Approach 2:
The patent introduces merged tracks as an intermediary structure between initial sensor tracks and final object identification. These merged tracks serve as intermediate results that combine spatial information before frequency analysis is applied, facilitating a stepwise processing approach that reduces computational complexity.
2Ease of operation
If deterministic approaches or fuzzy approaches are used for data merging, then operator intervention or training data is required, but automation and ease of operation are reduced
Solution Approach 1:
The patent implements self-service through automatic harmonic comb identification in the frequency domain. The system autonomously identifies characteristic frequency patterns of marine vessels without requiring operator intervention or external training data, enabling fully automated maritime domain awareness operations.
Solution Approach 2:
The patent replaces manual operator judgment and fuzzy logic systems with objective statistical signal processing methods. By using harmonic comb analysis based on physical principles of vessel noise generation, the system eliminates the need for operator intuition or trained neural networks.
3Quantity of substance
If sensors with different noise levels are integrated, then comprehensive data coverage is improved, but the difficulty of detecting and measuring increases due to varying measurement qualities
Solution Approach 1:
The patent transforms the data representation by transitioning from time-domain signals to frequency-domain spectral analysis. This parameter transformation converts varying noise levels into distinguishable spectral patterns, allowing heterogeneous sensor data to be integrated effectively through harmonic comb identification rather than direct time-domain comparison.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method effectively merges data from multiple sensors by considering multiple proximity criteria, reducing computational complexity and eliminating the need for training data or operator intervention, while providing accurate characterization of objects through spectral analysis.
Implementation Method 1
a second step of frequency analysis of the merged tracks which performs the selection of the merged tracks whose spectrum takes the form of a harmonic comb characteristic of a single real object
Data Source
Figure 1
Figure 2~2b
Figure 3
AI summary
The present invention relates to the field of data merging, the term "data" being taken within the broad sense of "observation data". It relates notably to the data obtained on the basis of the processing of sonar, radar or optronic signals. The subject of the invention is a method for carrying out the spectral merging and characterization of tracks, each track consisting of the data relating to the evolution over time of the bearing position and of the Doppler frequency of an assumed object. This method comprises: a first step of merging the initial tracks which correspond substantially to identical observation directions (azimuths), a second step of frequency analysis of the tracks resulting from merging the initial tracks and of characterizing the merged tracks whose spectrum takes the form of a harmonic comb characteristic of a single real object. The invention finds for example its application in data merging and echo classification, for narrowband passive sonar applications.