Navigation Filter Using Closed-Form Expressions for Vehicle State Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current state-of-the-art navigation filters for vehicles, such as lander vehicles on celestial bodies, face issues like divergence, high computational cost, and limited accuracy in estimating the vehicle's state during braked descent and landing, particularly due to the need for precise initial estimates, high precision IMUs, and lengthy calibration processes.
Innovation Solution
A novel navigation filter using closed-form expressions to directly calculate the vehicle's state, eliminating the need for recursive algorithms and reducing computational costs, while providing robust error estimation and improved accuracy, allowing for real-time autonomous state estimation without initialization values or extensive calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If recursive filtering algorithms (Kalman Filter) are used for state estimation, then the navigation filter can continuously update state estimates, but the computational cost becomes excessively high and the filter may diverge
Solution Approach 1:
The patent replaces the mechanical recursive filtering process with a direct closed-form mathematical solution. Instead of iteratively updating state estimates through recursive calculations, the invention uses a direct algebraic formula that computes the state estimate in a single step, eliminating the computational burden of repeated matrix operations while maintaining estimation accuracy
Solution Approach 2:
The patent inverts the traditional approach by not starting with the recursive update equations and trying to solve them, but rather deriving a direct closed-form solution that bypasses the iterative process entirely. This inversion transforms the problem from one requiring continuous computation to one with a direct analytical solution
2Adaptability or versatility
If Kalman Filter algorithms are used with measurement update cycles, then the filter can adapt to new measurements, but the filter requires precise initial estimates and may diverge if initialization is incorrect
Solution Approach 1:
The patent replaces the iterative measurement update mechanism with a direct closed-form solution that inherently handles measurements without requiring iterative convergence. The direct algebraic approach eliminates the risk of divergence that plagues recursive filters when initialization is incorrect, while still allowing continuous adaptation to new measurements through direct computation
3Measurement precision
If high precision IMUs are used to improve state estimation accuracy, then the measurement quality increases, but the device cost and complexity increase significantly
Solution Approach 1:
The patent changes the mathematical parameters and approach of the filtering algorithm from recursive iterative methods to direct closed-form expressions. This parameter change in the computational methodology allows for more effective utilization of standard IMU measurements, extracting maximum information without requiring ultra-high precision sensors, thereby reducing device complexity while maintaining estimation accuracy
Solution Approach 2:
The patent substitutes the hardware-dependent approach (relying on high precision IMUs) with a software/mathematical approach (closed-form solution). This substitution demonstrates that improved algorithms can compensate for standard sensor precision, reducing the need for expensive high-precision IMUs while maintaining state estimation accuracy
4Speed
If recursive filtering algorithms are used for real-time navigation, then continuous state updates are possible, but the processing time and computational resources required become prohibitive
Solution Approach 1:
The patent substitutes the time-consuming recursive computation process with a direct closed-form calculation. By replacing iterative matrix operations with a single algebraic solution, the system achieves real-time processing speeds without the computational delay inherent in recursive filtering, significantly reducing the time lost to computations while maintaining continuous state estimation capability
Data Source
AI summary
The invention concerns an innovative Device and Method to estimate the state of a moving vehicle overflying a certain terrain. The device is composed of a camera oriented toward the terrain, an inertial measurement unit, a device for the processing of images and a device called "Navigation Filter". This filter uses an innovative method to obtain the state estimates of the vehicle. Unlike the methods used in state of the art systems, here only robust and flexible expressions are used, producing accurate state estimates, with no possibility of divergence, with no need for initial state estimates or of a high computational power. The method calculates at first some parameters describing geometrical relationships among points of the trajectory and others on the terrain. These parameters are combined with the estimates of the accelerations to get the estimates of the velocity at a given time and of the gravity acceleration vector. By integration of these estimates, velocity and position profiles are obtained. The state is expressed in a reference system fixed with respect to the terrain. In case digital map of the terrain is available, it becomes possible to estimate the state also with respect to the reference system used in the map.