Adaptive Attitude Estimation Using a Hybrid Model-Learning Approach
arXiv:2207.06903 · doi:10.1109/tim.2022.3205003
Abstract
Attitude determination using the smartphone's inertial sensors poses a major challenge due to the sensor low-performance grade and variate nature of the walking pedestrian. In this paper, data-driven techniques are employed to address that challenge. To that end, a hybrid deep learning and model based solution for attitude estimation is proposed. Here, classical model based equations are applied to form an adaptive complementary filter structure. Instead of using constant or model based adaptive weights, the accelerometer weights in each axis are determined by a unique neural network. The performance of the proposed hybrid approach is evaluated relative to popular model based approaches using experimental data.
References in corpus (4)
- Denoising IMU Gyroscopes with Deep Learning for Open-Loop Attitude Estimation
- Neural Networks Versus Conventional Filters for Inertial-Sensor-based Attitude Estimation
- RoNIN: Robust Neural Inertial Navigation in the Wild: Benchmark, Evaluations, and New Methods
- DANAE: a denoising autoencoder for underwater attitude estimation