Rémy Guyonneau

Model-based approach for fault diagnosis using set-membership formulation

This paper describes a robust model-based fault diagnosis approach that enables to enhance the sensitivity analysis of the residuals. A residual is a fault indicator generated from an analytical redundancy relation which is derived from the structural and causal properties of the signed bond graph model. The proposed approach is implemented in two stages. The first stage consists in computing the residuals using available input and measurements while the second level leads to moving horizon residuals enclosures according to an interval consistency technique. These enclosures are determined by solving a constraint satisfaction problem which requires to know the derivatives of measured outputs as well as their boundaries. A numerical differentiator is then proposed to estimate these derivatives while providing their intervals. Finally, an inclusion test is performed in order to detect a fault upon occurrence. The proposed approach is well suited to deal with different kinds of faults and its performances are demonstrated through experimental data of an omni-directional robot.

A visibility information for multi-robot localization

This paper proposes a set-membership method based on interval analysis to solve the pose tracking problem for a team of robots. The originality of this approach is to consider only weak sensor data: the visibility between two robots. The paper demonstrates that with this poor information, without using bearing or range sensors, a localization is possible. By using this boolean information (two robots see each other or not), the objective is to compensate the odometry errors and be able to localize in an indoor environment all the robots of the team, in a guaranteed way. The environment is supposed to be defined by two sets, an inner and an outer characterizations. This paper mainly presents the visibility theory used to develop the method. Simulated results allow to evaluate the efficiency and the limits of the proposed algorithm.

Guaranteed interval analysis localization for mobile robots

This paper presents a set membership method (named Interval Analysis Localization (IAL)) to deal with the global localization problem of mobile robots. By using a LIDAR (LIght Detection And Ranging) range sensor, the odometry and a discrete map of an indoor environment, a robot has to determine its pose (position and orientation) in the map without any knowledge of its initial pose. In a bounded error context, the IAL algorithm searches a set of boxes (interval vector), with a cardinality as small as possible that includes the robot’s pose. The localization process is based on constraint propagation and interval analysis tools, such as bisection and relaxed intersection. The proposed method is validated using real data recorded during the CAROTTE challenge, organized by the French ANR (National Research Agency) and the French DGA (General Delegation of Armament). IAL is then compared with the well-known Monte Carlo Localization showing weaknesses and strengths of both algorithms. As it is shown in this paper with the IAL algorithm, interval analysis can be an efficient tool to solve the global localization problem. Graphical Abstract

An Image Processing Method Based on Features Selection for Crop Plants and Weeds Discrimination Using RGB Images

In the context of computer vision applied to precision agriculture, this paper presents an imaging system based on shape and intensity features, extracted from RGB images, for the discrimination between crop plants and weeds. A segmentation method with many constraints to overcome light acquisition conditions is used and coupled with morphological filtering suitable for denoising segmented images. A SVMs classifier based on a polynomial kernel function is implemented and a k-folds cross validation process is used to evaluate the performance of the SVMs classifier usable in 2 different configurations. On a training dataset, these 2 configurations are evaluated for the performance of classification in terms of true and false positive rates, according to ROC curves and area under curves. On a test dataset, these 2 configurations are exploited, giving both a relevant classification rate.

LiDAR-only based navigation algorithm for an autonomous agricultural robot

The purpose of the work presented in this paper is to develop a general and robust approach for autonomous robot navigation inside a crop using LiDAR (Light Detection And Ranging) data. To be as robust as possible, the robot navigation must not need any prior information about the crop (such as the size and width of the rows). The developed approach is based on line extractions from 2D point clouds using a PEARL based method. In this paper, additional filters and refinements of the PEARL algorithm are presented in the context of crop detection. A penalization of outliers, a model elimination step, a new model search and a geometric constraint are proposed to improve the crop detection. The approach has been tested over a simulator and compared with classical PEARL and RANSAC based approaches. It appears that adding those modification improved the crop detection and thus the robot navigation. Those results are presented and discussed in this paper. It can be noticed that even if this paper presents simulated results (to ease the comparison with other algorithms), the approach also has been successfully tested using an actual Oz weeding robot, developed by the French company Naio Technologies.