A5015061267- Neural Networks and Applications
- Fuzzy Logic and Control Systems
- Fault Detection and Control Systems
- Adaptive Control of Nonlinear Systems
- Metaheuristic Optimization Algorithms Research
- Natural Language Processing Techniques
- Advanced Vision and Imaging
- Robotic Path Planning Algorithms
- Topic Modeling
- Control and Dynamics of Mobile Robots
- Iterative Learning Control Systems
- Evolutionary Algorithms and Applications
- Robotics and Sensor-Based Localization
- Robotics and Automated Systems
- Spectroscopy and Chemometric Analyses
- Image Processing Techniques and Applications
- Mathematical and Theoretical Epidemiology and Ecology Models
- Advanced Control Systems Optimization
- Cellular Automata and Applications
- Vehicle Dynamics and Control Systems
- Evolutionary Game Theory and Cooperation
- Dynamics and Control of Mechanical Systems
- Advanced Text Analysis Techniques
- Industrial Vision Systems and Defect Detection
- Neural Networks Stability and Synchronization
Instituto Politécnico Nacional
2015-2024
Mathematics Research Center
2024
Centro de Estudios Universitarios Adolfo López Mateos
2007-2011
Mexicali Institute of Technology
2007
Universidad del Centro de México
2007
Universidad La Salle
2001-2006
Universidad La Salle
2005-2006
La Salle University
2005
In this article we present a flatness-based controller that solves the trajectory tracking problem for an unmanned four-rotor aircraft. A simple version of dynamical system obtained from suitable coordinate transformation Euler–Lagrange model is used. This model, combined with input saturation nonlinearity, also avoids inconvenient singularities. The corresponding error shown to be globally asymptotically stable but only locally exponentially stable, while maintains motions inside feasible...
<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> Since the fuzzy cerebellar model articulation controller (FCMAC) uses linguistic variables, it is highly intuitive and easily comprehended. Despite FCMAC's good local generalization capability for approximating nonlinear functions fast learning, a normal FCMAC requires huge memory, its dimension increases exponentially with number of inputs. In order to overcome memory explosion problem, this...
In this paper, a deep neural network based model for set of small-scale magnetorheological dampers (MRD) is developed where relevant parameters that have physical meaning are inputs to the model. An experimental platform and 3D-printing rapid prototyping facility provided different conditions including MRD filled with two MR fluids, which were used train Deep Neural Network (DNN), core proposed Testing results indicate could forecast hysteretic response load various configurations.
Single image depth estimation works fail to separate foreground elements because they can easily be confounded with the background. To alleviate this problem, we propose use of a semantic segmentation procedure that adds information estimator, in case, 3D Convolutional Neural Network (CNN)-segmentation is coded as one-hot planes representing categories objects. We explore 2D and models. Particularly, hybrid 2D-3D CNN architecture capable obtaining at same time. tested our on SYNTHIA-AL...
Nowadays, more than half of the world’s population lives in urban areas, and this number continues increasing. Consequently, there are scientific publications that analyze health problems people associated with living these highly urbanized locations. In particular, some recent work has focused on relating people’s to quality quantity green areas. context, considering huge amount land area large cities must be supervised, our seeks develop a deep learning-based solution capable determining...
This work presents a model based on Deep Neural Networks for the prediction of apparent personality. It can quantify personality traits with Five-Factor (Big Five) from Portrait image. In order to evaluate effectiveness this approach, new corpus 30,935 portraits their associated trait was extracted an existing resource videos (First Impressions, ChaLearn) tagged redundant pairwise comparisons ensure consistency. We propose several models using Convolutional automatically extract features...
The normal proportional derivative (PD) control is modified to a new dual form for the regulation of ball and plate system. First, analyze this controller, novel complete nonlinear model system obtained. Second, an asymptotic stable PD with compensation developed. Finally, experimental results are provided verify effectiveness proposed methodology.
Traditionally, approaches based on neural networks to solve the problem of disambiguation meaning words (WSD) use a set classifiers at end, which results in specialization single words-those for they were trained. This makes impossible apply learned models not previously seen training corpus. paper seeks address generalization WSD order it through deep without limiting method fixed words, with performance close state-of-the-art, and an acceptable computational cost. We explore different...
In this paper, the vision regulation of a quadrotor is introduced. The objective to maintain vehicle in desired and constant position. proposed strategy consists three steps: 1) camera employed as retro obtain position, 2) optical flow Lucas Kanade method applied linear velocity, 3) position velocity are by proportional integral derivative control (PID) with nested saturation for regulation. addition, comparison between monocular stereo visions showed.
Hierarchical fuzzy neural networks can use less rules to model nonlinear system with high accuracy. But the normal training method for hierarchical is very complex. In this paper we modify backpropagation approach and employ a time-varying learning rate that determined from input-output data structure. Stable algorithms premise consequence parts of are proposed. The calculation does not need any prior information such as estimation modeling error bounds. new simple, train each sub-block...
Tools that perform pattern recognition analysis of crimes, comprising at the same time forecasting, clustering, and recommendations on real data such as patrolling routes, are not fully integrated; modules developed separately, thus, a single workflow providing all steps necessary to pe rform this has been reported. In paper, we propose forecasting criminal activity in particular region by using supervised classification; then, use information automatically cluster find important hot spots;...
In this work we describe a system composed of deep neural networks that analyzes characteristics customers based on their face (age, gender, and personality), as well the ambient temperature, with purpose generating personalized signal to potential buyers who pass in front beverage establishment; faces are automatically detected, displaying recommendation using learning methods. order present suitable digital posters for each person, several technologies were used: Augmented reality,...