Dual EKF Inertial Navigation Algorithm Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional inertial navigation systems using solid state sensors require complex algorithms and excessive computational resources, are prone to errors, and have inflexible error correction methods that are not adaptable to varying vehicle configurations and dynamics.
Innovation Solution
A navigation system that selects between different algorithms based on sensor input quality and availability, using an Extended Kalman Filter (EKF) with a measurement correction module, time propagation module, and state propagation module, and dynamically adjusts parameters such as accelerometer adjustment factors and sensor usage to compute navigation state information more accurately and robustly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional algorithms are used with solid state inertial sensors, then navigation state estimation can be achieved, but computational complexity and resource requirements become excessive
Solution Approach 1:
The navigation algorithm is divided into multiple modes (first navigation algorithm and second navigation algorithm) that can be selectively executed based on sensor input quality. This segmentation allows the system to use simpler algorithms when appropriate, reducing computational complexity while maintaining accuracy when needed.
Solution Approach 2:
The system dynamically changes algorithm parameters and selection based on sensor input quality metrics. By adjusting which algorithm mode is active based on real-time sensor performance, the system optimizes the balance between computational resources and navigation accuracy.
2Reliability
If feedback signals are used to stabilize navigation state information, then navigation accuracy improves, but sensor errors corrupt the navigation state
Solution Approach 1:
The system uses feedback signals from sensors to continuously update and stabilize navigation state information. The feedback mechanism allows the system to maintain accuracy over extended periods by constantly adjusting based on sensor measurements.
Solution Approach 2:
The system dynamically changes algorithm selection and parameters based on sensor input quality assessment. When sensor errors are detected, the system switches to algorithms less affected by those specific sensors, thereby mitigating the harmful effects of sensor errors while maintaining feedback-based stabilization.
3Reliability
If conventional error correction methods are used, then sensor errors are guarded against, but the methods are inflexible and cannot be adjusted for various vehicle configurations
Solution Approach 1:
The system implements dynamic algorithm selection where the navigation algorithm changes based on real-time sensor quality assessment. This dynamic approach allows the system to adapt error correction methods to current operating conditions and vehicle configurations, providing both reliability and flexibility.
Solution Approach 2:
The system changes algorithm parameters and selection based on vehicle configuration and sensor performance. This allows error correction methods to be dynamically adjusted for different vehicle types, sensor configurations, and operating conditions, making the system both reliable and adaptable.
4Device complexity
If resource-constrained embedded DSPs are used, then system cost and size are reduced, but computational resources are insufficient for complex algorithms
Solution Approach 1:
The navigation system is segmented into multiple algorithm modes with different computational requirements. The first navigation algorithm uses fewer computational resources while the second provides higher accuracy when resources are available. This segmentation enables deployment on resource-constrained embedded DSPs while maintaining the option for higher accuracy when needed.
Solution Approach 2:
The system dynamically changes algorithm selection based on available computational resources and sensor quality. This allows the system to operate within the constraints of embedded DSPs while maintaining navigation accuracy through intelligent algorithm selection rather than relying on always-executing computationally intensive algorithms.
Data Source
AI summary
A system and method for more accurately and robustly estimating navigation state of a vehicle by adaptively processing signals from an inertial sensor assembly and other sensors. A navigation system receives signals from two or more sensors to evaluate and correct the attitude estimated by an Extended Kalman Filter (EKF). The navigation system selects sensor signals from the other sensor assemblies and processes the selected sensor signals in conjunction with estimates from the inertial navigation module to obtain more accurate estimates of the attitude. The parameters and conditions for using certain sensor signals may be adjusted based on the characteristics and configuration of the vehicle.