Leopoldo Acosta

Ping-pong player prototype

In this paper, a low-cost robot capable of playing ping-pong against human opponent using a vision system to detect the ball is presented. In the subsequent sections, the three main subsystems of the robot, i.e., the vision system, mechanical structure, and the control systems, are described. A prototype has been designed with lightweight and resistant materials to increase the response time and accurateness of the shot. One of the important features of this system is that it uses only one camera to detect the ball, thus reducing the computational time and hardware requirements. To detect the location of the ball, the robot combines the information about the ball and about the shadow it casts on the table. The expert module control defines the game strategy. Orienting the bat in order to return the ball to the desired position on the table does this. In these experiments, the success rate in returning balls was greater than 80%.

Obstacle avoidance for a mobile robot: A neuro-fuzzy approach

In this paper, a neuro-fuzzy approach is presented in order to guide a mobile robot. This task could be carried out specifying a set of fuzzy rules taking into account the different situations found by the mobile robot. This set of fuzzy rules could be optimised according to different criteria. However, the approach shown in this paper, is able to extract a set of fuzzy rules set from a set of trajectories provided by a human. These trajectories guide the mobile robot towards the target in different cases. Thus, it has been possible to obtain the rules and membership functions automatically, whereas other approaches need a previous definition of the rules and membership functions. In order to verify that the obtained behaviour is satisfactory, the neuro-fuzzy approach has been implemented in two mobile robots.

A self-tuning neuromorphic controller: application to the crane problem

Abstract This paper is concerned with the design and application of a self-tuning controller, aided by means of neural network s (NN). The structure of the controller is based on the use of neural networks as an implicit self-tuner for the controller. The aim o f this approach is to take advantage of the learning properties of the neural networks to increase the performance of the self-tuning. The a pplication of this technique is performed on an overhead crane. The control objective is to suppress undesirable oscillations during op eration of the crane. First, some simulations were carried out, as well as a comparison with a standard self-tuning method, that demon strate the advantages of this method. After this, a real-time implementation on a scale prototype of a crane was done to verify th e applicability of the method.

A New Low Cost System for Autonomous Robot Heading and Position Localization in a Closed Area

A low cost system for the localization of mobile indoor robots is presented. The system is composed of an emitter located on a wall and a receptor on top of the robot. The emitter is a laser pointer acting like a beacon, and the receptor is a cylinder made out of 32 independent photovoltaic cells. The robots position and orientation are obtained from the moments when the laser crosses each cell.

An Application of a Neural Self-Tuning Controller to an Overhead Crane

A neural network-based self-tuning controller is presented. The scheme of the controller is based on using a multilayer perceptron, or a set of them, as a self-tuner for a controller. The method proposed has the advantage that it is not necessary to use a combined structure of identification and decision, common in a standard self-tuning controller. The paper explains the algorithm for a general case, and then a specific application on a nonlinear plant is presented. The plant is an overhead crane which involves an interesting control problem related to the oscillations of the load mass. The method proposed is tested by simulation in different conditions. A comparison was made with a conventional controller to evaluate the efficiency of the algorithm.

A robotic system based on neural network controllers

Abstract In this paper, a control algorithm based on neural networks is presented. This control algorithm has been applied to a robot arm which has a highly nonlinear structure. The model based approaches for robot control (such as the computed torque technique) require high computational time and can result in a poor control performance, if the specific model-structure selected does not properly reflect all the dynamics. The control technique proposed here has provided satisfactory results. A decentralised model has been assumed here where a controller is associated with each joint and a separate neural network is used to adjust the parameters of each controller. Neural networks have been used to adjust the parameters of the controllers, being the outputs of the neural networks, the control parameters.

Design of a neural network based self-tuning controller for an overhead crane

In the process industry, the use of overhead crane systems for the transportation of material is very common. These are nonlinear systems that present undesirable oscillations during the motion, especially at arrival. The paper presents a self-tuning controller based on neural networks for the anti-swing control problem of the crane. The scheme of the controller is based on using neural networks as self-tuners for the parameters of a state feedback controller. The aim of this approach is to take advantage of the ability to learn of the neural networks and to use them in place of an identifier in the conventional self-tuner scheme. One of the main advantages of this method is that the training of the networks is done online using a backpropagation algorithm. The algorithm was implemented and tested by means of different simulations carried out with the crane.

On the Design and Implementation of a Neuromorphic Self-Tuning Controller

This paper deals with the design and implementation of a neural network-based self-tuning controller. The structure of the controller is based on using a neural network, or a set of them, as a self-tuner for a controller. The intention of this approach is to take advantage of the ability to learn of the neural networks and to use them in place of an identifier in the conventional self-tuner scheme. The work is divided into two main parts. The first one is dedicated to the design of the self-controller. And the second is an application of the algorithm on a nonlinear system: an overhead crane. Some simulations were carried out to verify the efficiency of the self-tuner and then a real-time implementation on a scale prototype was performed.

Solving the forward kinematics problem in parallel robots using Support Vector Regression

Abstract The Stewart platform, a representative of the class of parallel manipulators, has been successfully used in a wide variety of fields and industries, from medicine to automotive. Parallel robots have key benefits over serial structures regarding stability and positioning capability. At the same time, they present challenges and open problems which need to be addressed in order to take full advantage of their utility. In this paper, we propose a new approach for solving one of these key aspects: the solution to the forward kinematics in real-time, an under-defined problem with a high-degree nonlinear formulation, using a popular machine learning method for classification and regression, the Support Vector Machines. Instead of solving a numerical problem, the proposed method involves applying Support Vector Regression to model the behavior of a platform in a given region or partition of the pose space. It consists of two phases, an off-line preprocessing step and a fast on-line evaluation phase. The experiments made have yielded a good approximation to the analytical solution, and have shown its suitability for real-time application.

Using Kinect on an Autonomous Vehicle for Outdoors Obstacle Detection

An accurate method to detect obstacles and dangerous areas is the key to the safe performance of autonomous robots. Time of flight sensors can report their existence through the emission, reflection, and measurement of wave patterns, but large wavelength light projection is often unreliable in outdoors environments, due to solar radiation contamination. In this paper, a specific Microsoft Kinect arrangement on a robotic vehicle is proposed, such that outdoors detection is possible. The main contribution of this paper is the description of a sequence of filtering techniques, which translate the depth image provided by the sensor into definite obstacle projections in the navigability map used by the vehicle. A series of experiments proves that the Kinect device is more accurate at detecting obstacles using this procedure than a camera pair using two different stereovision techniques.