The invention concerns a lateral velocity estimator based on a novel neural network that is designed with a structure in accordance with kinematics principles.
The new neural network estimates lateral velocity with high accuracy by a generalization of kinematics equations and can be trained with little data. The kinematics structure on which it is based improves the robustness of lateral velocity estimation, ensuring appropriate accuracies even when the vehicle type varies.
The invention uses the measurement of yaw rate, longitudinal and lateral acceleration, and longitudinal velocity and does not require an Inertial Measurement Unit (IMU).
Climate-neutral and smart citiesAddressing target audiences and expressing needs
- Collaboration
- Business Angels
- Venture Capital
Investors (Automotive OEMs, Suppliers) willing to adopt the technology.
- Public or private funding institutions
- Research and Technology Organisations
- Private Investors
R&D, Technology and Innovation aspects
The invention is demonstrated and Patented (TRL about 8). Needs investors, adopters, and deployment on production vehicles.
The neural network architecture is modular, flexible, and customizable. It can be easily fine-tuned to different types of vehicles with little effort. It can self-adapt to varying operating conditions.
The neural network is customizable at two levels: hyperparameters-modular configuration, and training. The latter requires very little training data because the networks have very few efficient parameters.
The solution improves the quality of vehicle control and driver assistance functions (including self-driving).
