Online Outdoor Terrain Classification Algorithm for Wheeled Mobile Robots Equipped with Inertial and Magnetic Sensors

• DOI: 10.3390/electronics12153238
• Publication Date: 2023-07-27T06:07:17Z
• Authors: Peter Sarcevic; Dominik Csík; Richard Pesti; Sara Stančin; Sašo Tomažič; Vladimir Tadic; Juvenal Rodriguez...; József Sárosi; Akos Odry
• Keywords: Feature (linguistics)

Terrain classification provides valuable information for both control and navigation algorithms of wheeled mobile robots. In this paper, a novel online outdoor terrain algorithm is proposed The based on only time-domain features with low computational memory requirements, which are extracted from the inertial magnetic sensor signals. Multilayer perceptron (MLP) neural networks applied as classifiers. tested measurement database collected using prototype system various types. Different datasets were constructed setups processing window sizes, used types, robot speeds. To examine possibilities three types in application, data different sensors alone, pairs fused together. suitable to operate embedded robot. achieved results show that feature set highest efficiencies unknown can be above 98%. It also shown gyroscope higher rates than widely accelerometer. alone cannot effectively but fusing improve performance.