Towed Antenna Shape Estimation Using Dynamic Kalman Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the shape of a towed acoustic antenna using Kalman filtering are limited by restrictive assumptions, such as a homogeneous, thin, flexible cylindrical body, which are not valid under real conditions, especially when the towing point is not at the head of the acoustic section and during cornering maneuvers.
Innovation Solution
A method that sets the deterministic part of the state equation of the Kalman filter to zero, uses a mechanical model to describe the behavior of the towed antenna, and updates the transition matrix based on known tow point movements and physical properties, allowing for estimation of the antenna's shape even when the tow point is at a different interface and during non-straight maneuvers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Kalman filtering is used with the assumption of a homogeneous, thin, flexible cylindrical body, then shape estimation is possible, but the method is only valid for special cases and cannot handle real conditions where the towing point is not at the head and during cornering maneuvers
Solution Approach 1:
The patent changes the fundamental parameters of the Kalman filter by setting the deterministic part of the state equation to zero and updating the transition matrix based on actual tow point movements and mechanical models. This allows the filter to adapt to real conditions including arbitrary tow point positions and cornering maneuvers, while maintaining shape estimation accuracy through dynamic parameter adjustment based on measured data.
2Ease of manufacture
If the towing point is assumed to be at the head of the acoustic section, then Kalman filtering can be applied, but this assumption is not valid when the towing point is at the interface between the vehicle and the towing cable
Solution Approach 1:
The patent introduces a mechanical model as an intermediary between the Kalman filter and the physical system. This mechanical model accounts for the actual towing configuration with the towing point at the vehicle-cable interface, allowing the filter to process data from this realistic setup rather than requiring the simplified head-mounted configuration. The mechanical model transforms the complex real-world dynamics into a form suitable for filtering while maintaining accuracy.
3Measurement precision
If the deterministic part of the state equation is set to zero and the transition matrix is updated based on mechanical models, then accurate shape estimation under real conditions is achieved, but the complexity of the filtering process increases
Solution Approach 1:
The patent performs preliminary computation of the transition matrix using a mechanical model before the Kalman filtering process. By pre-calculating the transition matrix based on known tow point movements and system dynamics, the complex computations are done in advance, allowing the actual filtering process to use these pre-computed values and reducing real-time computational complexity while maintaining high estimation accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method and an apparatus for estimating the shape of an acoustic trailing antenna (12), wherein the shape is estimated using Kalman filtering. First of all, this involves the deterministic component (u(ki)) of the state equation of the Kalman filter being set to zero. In addition, successive discrete times (ki) are predefined and, at the predefined times (ki), a respective estimated shape for the trailing antenna (12) is described by a model-based state vector (x)(ki). In this case, the model-based state vectors (x)(ki) are ascertained by the estimated timing response of a mechanical model (24) of the trailing antenna (12) and by movements in a traction point (16) for the trailing antenna (12) that are assumed to be known. The discrepancy in the respective current model-based state vector (x)(ki) is determined for one or more preceding model-based state vectors (x)(ki-1), is considered as a current matrix (F)(ki), and the current transition matrix (F)(ki) is periodically updated for the Kalman filtering with the ascertained matrices (F)(ki).