Radar Target Count Estimation via Angular Spectrum
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional radar systems for vehicles face challenges in accurately estimating the number of target objects, which is crucial for precise angle estimation and effective safety and convenience technologies, but research in this area is insufficient.
Innovation Solution
An apparatus and method that utilize a radar signal receiver and a controller with a learning unit and estimation unit, employing a neural network to process radar signals, extract feature point information, and parameterize it for accurate estimation of the number of targets by identifying angular spectra and applying consistent weighting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional radar signal processing methods are used, then the system is simple and easy to implement, but the accuracy of target number estimation is insufficient
Solution Approach 1:
The patent introduces an angular spectrum as an intermediary representation between the raw radar signal and the target number estimation. The controller performs Fourier transformation on the radar signal to generate the angular spectrum, which reveals the distribution of target reflections across different angles. This intermediary representation makes the target number estimation more accurate by providing clear spectral peaks corresponding to different targets, while keeping the processing method relatively simple.
Solution Approach 2:
The patent replaces conventional mechanical or algorithmic signal processing methods with a neural network-based learning system. The controller includes a learning unit that uses neural networks to learn the relationship between angular spectra and target numbers from training data, and an estimation unit that applies the learned model to new signals. This substitution significantly improves estimation accuracy while maintaining reasonable system complexity through software-based processing.
2Measurement precision
If more research is conducted on target angle estimation, then angle estimation accuracy improves, but target number estimation accuracy remains insufficient
Solution Approach 1:
The patent segments the target detection process into two distinct functional units: a learning unit that specifically focuses on target number estimation using neural networks, and an estimation unit that applies the learned knowledge. This segmentation allows the system to dedicate specific computational resources and algorithms to target number estimation without compromising angle estimation capabilities, thereby addressing the research focus imbalance by giving target number estimation dedicated attention.
Solution Approach 2:
The patent transitions from analyzing radar signals in the time-domain or simple frequency-domain to the angular spectrum domain through Fourier transformation. This dimensionality change provides a new perspective for target number estimation, where multiple targets appear as distinct peaks in the angular spectrum. By adding this angular dimension analysis, the system can accurately estimate target numbers without losing information about angle estimation, effectively balancing both research focuses.
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
Enables accurate estimation of the number of target objects, reducing false detection and non-detection probabilities, and allowing for precise angle estimation based on the learned information.
Implementation Method 1
a radar signal receiver configured to receive a radar signal that belongs to a detection signal transmitted by a radar and that is reflected by an object on the ground
Data Source
AI summary
An apparatus for estimating the number of targets including a radar signal receiver configured to receive a radar signal that belongs to a detection signal transmitted by a radar and that is reflected by an object on the ground, and a controller configured to learn the number of targets by processing the received radar signal and to estimate the number of targets by processing a newly received radar signal based on the learned information.