Douglas Eduardo Zampieri

Pearson's Correlation Coefficient for Discarding Redundant Information in Real Time Autonomous Navigation System

Lately, many applications for control of autonomous vehicles are being developed and one important aspect is the excess of information, frequently redundant, that imposes a great computational cost in data processing. Based on the fact that real-time navigation systems could have their performance compromised by the need of processing all this redundant information (all images acquired by a vision system, for example), this work proposes an automatic image discarding method using the Pearsons correlation coefficient (PCC). The proposed algorithm uses the PCC as the criteria to decide if the current image is similar to the reference image and could be ignored or if it contains new information and should be considered in the next step of the process (identification of the navigation area by an image segmentation method). If the PCC indicates that there is a high correlation, the image is discarded without being segmented. Otherwise, the image is segmented and is set as the new reference frame for the subsequent frames. This technique was tested in video sequences and showed that more than 90% of the images can be discarded without loss of information, leading to a significant reduction of computational time necessary to identify the navigation area.

Robust horizon finding algorithm for real-time autonomous navigation based on monocular vision

Navigation of an Autonomous Vehicle is based on its interaction with the environment, through information acquired by sensors. The perception of the environment is a major issue in autonomous and (semi)-autonomous systems. This work presents the embedded real-time visual perception problem applied to experimental platform. In this way, a robust horizon finding algorithm that finds the horizon line was proposed and applied to generate the navigable area. It permits to investigate dynamically only a small portion of the image (road) ahead of the vehicle. From a dynamic threshold search method based on Otsu segmentation and Hough transform, this system was robust to illumination changes and does not need any contrast adjustments.

Recurrent neural network approaches for biped walking robot based on zero-moment point criterion

The main objective of this paper is to use a recurrent neural network (RNN) to determine the trunk motion for a biped-walking machine, based on the zero-moment point (ZMP) criterion. ZMP criterion can be used to plan a stable gait for a biped-walking machine that has a trunk (inverted pendulum). So, a RNN is trained to determine a compensative trunk motion that makes the actual ZMP get closer to the planned ZMP. In this context, an identification scheme is presented to obtain the vector of parameters of the RNN. A first order standard back-propagation with momentum (BPM) is used to adjust free parameters for the network. Artificial neural network brings up important features for function approximation. This was the main reason to use an RNN to determine the trunk motion. The proposed scheme is simulated on a 10-degree-of-freedom biped robot. The results confirm the convergence of the proposed scheme, proving this is a new way to solve this classical problem in the biped-walking machine area.

Design of a vehicular suspension controller by static output feedback

A static output feedback controller is applied to the design of an active suspension model for a quarter-car vehicle, using only the suspension travel displacement and velocity as output variables. The feedback design proposed makes use of recent linear matrix inequalities results from the literature, also allowing to consider the presence of uncertain parameters. Simulations are carried out taking into account the model uncertainty. The results are shown to be as good as the classical linear quadratic regulator which supposes full state availability and precisely known parameters.

Real-time dynamic power management based on Pearson's Correlation Coefficient

An autonomous robotic platform should be able to perform long-range and long-endurance missions, which energy limitation is one of the most important challenges. Studies show that motion is not the only power consumer. Management of all power resources is therefore important for these systems. Moreover, many applications for control of autonomous platform are being developed and one important aspect is the excess of information, frequently redundant, that imposes a great computational cost in data processing and consequently power consummation. In our previous works, we have proposed a visual-perception system based on an automatic image discarding method as a simple solution to improve the performance of a real-time navigation system. In this paper, we propose a new environment observer method based on Pearsons Correlation Coefficient. This monocular-vision system permits that some logical components may be shut down to save processor energy consumption, and/or to make the CPU available for running concurrent processes. Nevertheless, this method may be extended to other sensors and components. Our real-time perception system has been evaluated from real data obtained by our intelligent vehicle. It is not based on previous knowledge of the environment neither camera calibration.

Multiobjective weighting selection for optimization-based control design

A methodology for the multiobjective design of controllers is presented, motivated by the problem of designing an active suspension controller. This problem has, as a particular feature, the possibility of being defined with two design variables only. The multiobjective controller is searched inside the space of optimal controllers defined by a weighted cost functional. The weightings are taken as the optimization variables for the multiobjective design. The method leads to (local) Pareto-optimal solutions and allows the direct specification of controller constraints in terms of some primary objectives which are taken into account in the multiobjective search.

A Nonlinear Control Design for Tracked Robots with Longitudinal Slip

Abstract This paper presents an adaptive control strategy for a tracked mobile robot, in which the longitudinal slip of the left and right tracks are described by two unknown parameters. It is assumed that the kinematic model of the tracked robot is approximated by the one of a differential wheeled robot. An adaptive nonlinear feedback control law that compensates for the longitudinal slip is proposed to achieve a given trajectory tracking objective. Asymptotic stability of the close-loop system is ensured using an appropriate Lyapunov function. Numerical results show the benefits of the proposed approach.

An integrated control for a biped walking robot

The main objective of this work is to present and discuss some results of an integrated control system for a biped robot machine in the dynamic gait. We divided the integrated control system in two sub-systems: a control of the trajectories for the legs and the Automatic Generator of Trajectory. We designed the Automatic Generator of Trajectory by employing a neural network, which updates online the conditions of trajectory for the trunk, from the evolution of the gait, with the objective to reduce the magnitude of the resultant force and moment. We consider that this scheme is a new and interesting approach to generate online the trajectory for the trunk, giving so reflexes for the biped-walking robot.

Optimal active suspension design via convex analysis

In this note the design of active suspension control is approached in a Youla parametrization setting, leading to dynamic output feedback controllers calculated by convex optimization. The controllers so designed have the following advantages over the usual ones: (i) they do not require the availability of the full states vector; (ii) the optimization problem may be cast in a true multiobjective fashion; (iii) the method reveals the limits of performance of the physical system under the given costs and constraints. An example is provided in a two-degree-of-freedom quarter-car model with H∞, optimization criterion.

Sistemas Classificadores Aplicados à Navegação de Robôs Autônomos

RESUMO: A geração de trajetórias para robôs móveis pode ser macro classificada em duas grandes categorias, tendo o ambiente (área de trabalho) como elemento principal: conhecida e determinística; ou desconhecida e dinâmica. Enquanto a navegação de robôs por ambiente conhecido e determinístico pode-se resumir à aplicação de algoritmos de busca com critérios de otimização, a navegação por ambientes desconhecidos e dinâmicos exige a utilização de algoritmos adaptativos. Neste artigo analisamos a navegação de robôs em ambiente desconhecidos utilizando-se aprendizado evolutivo com sistemas classificadores.