Machine Learning Algorithms in Heavy Process Manufacturing

In a global economy, manufacturers mainly compete with cost efficiency of production, as the price of raw materials are similar worldwide. Heavy industry has two big issues to deal with. On the one hand there is lots of data which needs to be analyzed in an effective manner, and on the other hand making big improvements via investments in cooperate structure or new machinery is neither economically nor physically viable. Machine learning offers a promising way for manufacturers to address both these problems as they are in an excellent position to employ learning techniques with their massive resource of historical production data. However, choosing modelling a strategy in this setting is far from trivial and this is the objective of this article. The article investigates characteristics of the most popular classifiers used in industry today. Support Vector Machines, Multilayer Perceptron, Decision Trees, Random Forests, and the meta-algorithms Bagging and Boosting are mainly investigated in this work. Lessons from real-world implementations of these learners are also provided together with future directions when different learners are expected to perform well. The importance of feature selection and relevant selection methods in an industrial setting are further investigated. Performance metrics have also been discussed for the sake of completion.

Machine learning and small robot navigation

Developing successful navigation and mapping strategies is an essential part of autonomous robot research. However, hardware limitations often make for inaccurate systems. This project serves to investigate efficient alternatives to mapping an environment, by first creating a mobile robot, and then applying machine learning to the robot and controlling systems to increase the robustness of the robot system. My mapping system consists of a semi-autonomous robot drone in communication with a stationary Linux computer system. There are learning systems running on both the robot and the more powerful Linux system. The first stage of this project was devoted to designing and building an inexpensive robot. Utilizing my prior experience from independent studies in robotics, I designed a small mobile robot that was well suited for simple navigation and mapping research. When the major components of the robot base were designed, I began to implement my design. This involved physically constructing the base of the robot, as well as researching and acquiring components such as sensors. Implementing the more complex sensors became a time-consuming task, involving much research and assistance from a variety of sources. A concurrent stage of the project involved researching and experimenting with different types of machine learning systems. I finally settled on using neural networks as the machine learning system to incorporate into my project. Neural nets can be thought of as a structure of interconnected nodes, through which information filters. The type of neural net that I chose to use is a type that requires a known set of data that serves to train the net to produce the desired output. Neural nets are particularly well suited for use with robotic systems as they can handle cases that lie at the extreme edges of the training set, such as may be produced by "noisy" sensor data. Through experimenting with available neural net code, I became familiar with the code and its function, and modified it to be more generic and reusable for multiple applications of neural nets.

Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

This study presents an approach to combine uncertainties of the hydrological model outputs predicted from a number of machine learning models. The machine learning based uncertainty prediction approach is very useful for estimation of hydrological models' uncertainty in particular hydro-metrological situation in real-time application [1]. In this approach the hydrological model realizations from Monte Carlo simulations are used to build different machine learning uncertainty models to predict uncertainty (quantiles of pdf) of the a deterministic output from hydrological model . Uncertainty models are trained using antecedent precipitation and streamflows as inputs. The trained models are then employed to predict the model output uncertainty which is specific for the new input data. We used three machine learning models namely artificial neural networks, model tree, locally weighted regression to predict output uncertainties. These three models produce similar verification results, which can be improved by merging their outputs dynamically. We propose an approach to form a committee of the three models to combine their outputs. The approach is applied to estimate uncertainty of streamflows simulation from a conceptual hydrological model in the Brue catchment in UK and the Bagmati catchment in Nepal. The verification results show that merged output is better than an individual model output. [1] D. L. Shrestha, N. Kayastha, and D. P. Solomatine, and R. Price. Encapsulation of parameteric uncertainty statistics by various predictive machine learning models: MLUE method, Journal of Hydroinformatic, in press, 2013.

Aquatic weed automatic classification using machine learning techniques

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

HOW FAR do WE GET USING MACHINE LEARNING BLACK-BOXES?

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Safety illusion and error trap in a collectively-operated machine accident

Workplace accidents involving machines are relevant for their magnitude and their impacts on worker health. Despite consolidated critical statements, explanation centered on errors of operators remains predominant with industry professionals, hampering preventive measures and the improvement of production-system reliability. Several initiatives were adopted by enforcement agencies in partnership with universities to stimulate production and diffusion of analysis methodologies with a systemic approach. Starting from one accident case that occurred with a worker who operated a brake-clutch type mechanical press, the article explores cognitive aspects and the existence of traps in the operation of this machine. It deals with a large-sized press that, despite being endowed with a light curtain in areas of access to the pressing zone, did not meet legal requirements. The safety devices gave rise to an illusion of safety, permitting activation of the machine when a worker was still found within the operational zone. Preventive interventions must stimulate the tailoring of systems to the characteristics of workers, minimizing the creation of traps and encouraging safety policies and practices that replace judgments of behaviors that participate in accidents by analyses of reasons that lead workers to act in that manner.

A machine learning approach for genome-wide prediction of morbid and druggable human genes based on systems-level data

Background: The genome-wide identification of both morbid genes, i.e., those genes whose mutations cause hereditary human diseases, and druggable genes, i.e., genes coding for proteins whose modulation by small molecules elicits phenotypic effects, requires experimental approaches that are time-consuming and laborious. Thus, a computational approach which could accurately predict such genes on a genome-wide scale would be invaluable for accelerating the pace of discovery of causal relationships between genes and diseases as well as the determination of druggability of gene products.Results: In this paper we propose a machine learning-based computational approach to predict morbid and druggable genes on a genome-wide scale. For this purpose, we constructed a decision tree-based meta-classifier and trained it on datasets containing, for each morbid and druggable gene, network topological features, tissue expression profile and subcellular localization data as learning attributes. This meta-classifier correctly recovered 65% of known morbid genes with a precision of 66% and correctly recovered 78% of known druggable genes with a precision of 75%. It was than used to assign morbidity and druggability scores to genes not known to be morbid and druggable and we showed a good match between these scores and literature data. Finally, we generated decision trees by training the J48 algorithm on the morbidity and druggability datasets to discover cellular rules for morbidity and druggability and, among the rules, we found that the number of regulating transcription factors and plasma membrane localization are the most important factors to morbidity and druggability, respectively.Conclusions: We were able to demonstrate that network topological features along with tissue expression profile and subcellular localization can reliably predict human morbid and druggable genes on a genome-wide scale. Moreover, by constructing decision trees based on these data, we could discover cellular rules governing morbidity and druggability.

EIGENVALUES of FIBONACCI STOCHASTIC ADDING MACHINE

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

A comment on a nonideal centrifugal vibrator machine behavior with soft and hard springs

This paper concerns a type of rotating machine (centrifugal vibrator), which is supported on a nonlinear spring. This is a nonideal kind of mechanical system. The goal of the present work is to show the striking differences between the cases where we take into account soft and hard spring types. For soft spring, we prove the existence of homoclinic chaos. By using the Melnikov's Method, we show the existence of an interval with the following property: if a certain parameter belongs to this interval, then we have chaotic behavior; otherwise, this does not happen. Furthermore, if we use an appropriate damping coefficient, the chaotic behavior can be avoided. For hard spring, we prove the existence of Hopf's Bifurcation, by using reduction to Center Manifolds and the Bezout Theorem (a classical result about algebraic plane curves).

Seeder-fertilizer machine: energetic demand as a function of vegetal covering handling and manure deposition shank depth in no-tillage system

O trabalho teve por objetivo avaliar a demanda energética de uma semeadora-adubadora, em função do tipo e manejo da cultura de cobertura vegetal e da profundidade da haste de deposição de adubo. Foi utilizado um trator Valtra BM100, instrumentado, para tracionar uma semeadora-adubadora de precisão equipada com quatro fileiras de semeadura espaçadas de 0,9 m para cultura de milho. O experimento foi conduzido em parcelas subsubdivididas, na área experimental do Laboratório de Máquinas e Mecanização Agrícola (LAMMA) da UNESP-Jaboticabal, utilizando duas culturas de cobertura (mucuna-preta e crotalária), três manejos dessas coberturas, sendo dois mecânicos (triturador de palhas e rolo-faca) e um químico (pulverização com herbicida), realizados 120 dias após a semeadura das culturas de cobertura e três profundidades da haste de deposição do adubo (0,11; 0,14 e 0,17 m), perfazendo 18 tratamentos, com quatro repetições, totalizando 72 observações. Foram avaliados os parâmetros velocidade de deslocamento, patinagem, força na barra de tração, força de pico, potência na barra de tração, potência de pico e consumo de combustível. Pôde-se concluir que a força na barra de tração foi menor para as profundidades de 0,11 e 0,14 m da haste sulcadora de adubo, o mesmo ocorrendo para força de pico, potência na barra de tração e consumo volumétrico. O consumo específico foi menor na profundidade de 0,17 m da haste sulcadora de adubo. As culturas de cobertura e seus manejos não interferiram no desempenho das máquinas estudadas.

Analytical method for the design of a machine with aligned and skewed permanent magnets

This work aims to present the design and the evaluation of a standard multi-pole machine with permanent magnets inserted in the rotor by two different geometrical forms: aligned and skewed magnets. The design (new analytical method) was based on a standard 250 W three phase 12-pole induction motor (squirrel cage rotor type), beginning with the original stator constructive data to calculate the magnetic flux density to determine the permanent magnets. In the development of the work, a simple and modular rotor was built reusing the original 12-pole stator (concentrated windings). The machine was evaluated in a laboratory for the purpose of checking the quantity and quality of energy produced with the machine operating as a generator and its start, torque, and performance working as a motor. In conclusion, the modular skewed magnet is an option for electrical machines, for the generation of a reasonable quality, in the context of decentralized generation and a motor with high torque and low energetic consumption.

Inverter-fed linear asynchronous machine

The behaviors of an arc-shaped stator induction machine (the sector-motor) and a disc-secondary linear induction motor are analyzed in this work for different values of the frequency. Variable frequency is produced by a voltage source controlled-current inverter which keeps constant the r.m.s. value of the phase current, also assuring a sinusoidal waveform. For the simulations of the machine developed thrust, an equivalent circuit is used. It is obtained through the application of the one-dimensional theory to the modeling. The circuit parameters take into account the end effects, always present is these kind of machines. The phase current waveforms are analyzed for their harmonic contents. Experimental measurements were carried out in laboratory and are presented with the simulations, for comparison.

New approach of induction machine drive using an adaptive digital PID controller with gain planning based on artificial neural networks

The present work introduces a new strategy of induction machines speed adjustment using an adaptive PID (Proportional Integral Derivative) digital controller with gain planning based on the artificial neural networks. This digital controller uses an auxiliary variable to determine the ideal induction machine operating conditions and to establish the closed loop gain of the system. The auxiliary variable value can be estimated from the information stored in a general-purpose artificial neural network based on CMAC (Cerebellar Model Articulation Controller).

Line based camera calibration in machine vision dynamic applications

The problem of dynamic camera calibration considering moving objects in close range environments using straight lines as references is addressed. A mathematical model for the correspondence of a straight line in the object and image spaces is discussed. This model is based on the equivalence between the vector normal to the interpretation plane in the image space and the vector normal to the rotated interpretation plane in the object space. In order to solve the dynamic camera calibration, Kalman Filtering is applied; an iterative process based on the recursive property of the Kalman Filter is defined, using the sequentially estimated camera orientation parameters to feedback the feature extraction process in the image. For the dynamic case, e.g. an image sequence of a moving object, a state prediction and a covariance matrix for the next instant is obtained using the available estimates and the system model. Filtered state estimates can be computed from these predicted estimates using the Kalman Filtering approach and based on the system model parameters with good quality, for each instant of an image sequence. The proposed approach was tested with simulated and real data. Experiments with real data were carried out in a controlled environment, considering a sequence of images of a moving cube in a linear trajectory over a flat surface.

Identification of synchronous machine parameters using load rejection test data

This work shows a computational methodology for the determination of synchronous machines parameters using load rejection test data. By machine modeling one can obtain the quadrature parameters through a load rejection under an arbitrary reference, reducing the present difficulties. The proposed method is applied to a real machine.