GNSS Error Simulation for Precise Vehicle Navigation Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle navigation systems face significant GNSS errors due to factors like ionospheric delays, multipath errors, and terrain interference, leading to inaccuracies of up to 30 meters, which hinder precise vehicle operation.
Innovation Solution
A method involving a machine learning model that simulates GNSS errors based on real-world data, including satellite orbits, atmospheric conditions, and terrain models, to enhance training and testing of vehicle operation systems, allowing for accurate trajectory planning with less than one-meter precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS data is used for vehicle navigation, then location determination is provided, but location accuracy deteriorates due to ionospheric delays, multipath errors, and terrain interference
Solution Approach 1:
The patent creates a simulated environment that copies real-world conditions (ionospheric delays, multipath errors, terrain interference) into a virtual space. This allows the system to train and test navigation algorithms using synthetic GNSS data that replicates actual error conditions without exposing physical vehicles to hazardous real-world scenarios, thereby improving location accuracy through comprehensive error condition simulation.
Solution Approach 2:
The simulation system acts as an intermediary between real-world GNSS error conditions and the navigation system development process. By mediating through virtual environments, the system can study and compensate for harmful factors like ionospheric delays and multipath errors without direct physical exposure, enabling safer and more accurate navigation algorithm development.
2Adaptability or versatility
If real-world training data is used for vehicle operation systems, then system training is comprehensive, but training safety and data processing complexity increase
Solution Approach 1:
The patent generates synthetic training data by copying and transforming real-world scenarios into virtual environments. This approach maintains training comprehensiveness by covering diverse conditions (urban canyons, open roads, various weather conditions) while reducing data processing complexity through automated generation and standardized formats, eliminating the need for manual data collection and processing infrastructure.
Solution Approach 2:
The system performs preliminary action by pre-generating and pre-processing training data in virtual environments before actual vehicle operation. This allows comprehensive training scenarios to be prepared in advance with controlled complexity, enabling safe and efficient training without the overhead of real-time data processing during actual vehicle operation.
3Productivity
If simulated GNSS data is generated using machine learning models, then training efficiency improves, but model training complexity increases
Solution Approach 1:
The patent uses machine learning models to copy and transform real GNSS data into simulated data that replicates various error conditions. This significantly improves training efficiency by enabling rapid generation of diverse training scenarios without requiring physical data collection, while the models are trained once on representative data and then deployed for efficient simulation.
Solution Approach 2:
The machine learning models perform preliminary training on representative real-world data to learn patterns of GNSS errors, then use these learned patterns to efficiently generate simulated data for broader training scenarios. This preliminary action reduces the complexity of ongoing model training while maintaining high productivity through rapid synthetic data generation.
Data Source
AI summary
A computer that includes a processor and a memory, the memory including instructions executable by the processor to input real world coordinates of a real world location and a real world time, number and orbit data of GNSS satellites, terrain model data, and atmospheric model data to a first machine learning model to generate a simulated GNSS location and heading based on the real world location and a simulated GNSS error in location and heading in the real world coordinates. The first machine learning model can be trained based on an acquired GNSS location, an acquired GNSS time, an acquired number and orbit data of the GNSS satellites, an acquired GNSS DOP from real world systems, the terrain model data, and the atmospheric model data. Simulated GNSS data including the simulated GNSS location and heading can be generated to perform one or more of training and testing of a simulated vehicle operation system.