Tecnura

13 - 26

Authors: José Antonio Valero Medina, Beatriz Elena Alzate Atehortúa

Context: Nowadays, the images of the Earth surface and the algorithms for their classification are widely available. In particular, the algorithms are promising in the differentiating of cotton crops stages, but it is necessary to establish the capabilities of the different algorithms in order to identify their advantages, and disadvantages. Method: This paper describes the assessment process in which the Support Vector Machines (SVM) and random-forest technique (decision trees) are compa-red with the maximum likelihood estimation when differentiating the stages of cotton crops. A RapidEye satellite image of a geographic area in the municipa-lity of San Pelayo, Cordoba (Colombia), is used for the study. Using a set of sampling polygons, a ran-dom sample of 6000 pixels was taken (2000 training and 4000 for validating the classifications.) Confu-sion matrices, and R (data processing and analysis software) were used during the validation process Results: The maximun likelihood estimation presen-ted a correct classification percentage of 68.95%. SVM correctly classified 81.325% of the cases and the decision trees correctly classified 78.925%. The confidence test for the classifications showed non-overlapping intervals, and SVM obtained the highest values. Conclusions: It was possible to confirm the supe-riority of the technique based on support vector machines for the proposed verification zones. How-ever, this technique requires a number of classes that comprehensively represent the variations of the im-age (in order to guarantee a minimum number of support vectors) to avoid confusion in the classifi-cation of non-sampled areas. This was less evident in the other two classification techniques analysed.

37 - 52

Authors: Álvaro Ervey Pulido Aponte, Hernán Mauricio Rivera Escobar, Jhon Jairo Espitia Pardo

Introduction: In Colombia, bovine ruminal content (BRC) is generally poured into water courses and open fields and, given its high organic load, its final disposal represents an environmental, economic, and social threat. To date, the implementation of te-chnologies that reduce the impact is limited. This is due, on the one hand, to the scarce characterization of the BRC, which is required for the construction of mathematical models of anaerobic digestion that explain the reduction of this bio-solid; and on the other hand, to the restricted application of control and supervision systems focused on the generation of by-products with potential use. Aim: To document the methods used in the develo-pment of control and monitoring system ot tempera-ture and pressure, for a discontinuous bio-digester at laboratory scale with biomass BRC Methods: The implementation of a distributed signal acquisition system and a web server interface were carried out; additionally, a dynamic model corres-ponding to the behavior of the anaerobic digestion system for this BRC biomass was identified Results: Using gravimetry and calorimetry, it was possible to characterize variables associated with the anaerobic digestion system with BRC. As for the control, the on-off controller kept the gas pressure in the desired range and the PI controller guaranteed the desired temperature of the process. Conclusion: It was possible to generate biogas from biomass BRC from basic schemes of characteriza-tion, supervision, and control of the process

65 - 78

Authors: Edgar Orlando Ladino Moreno, Germán Ricardo Santos Granados, César Augusto García-Ubaque

Context: Describing mathematically the hydrau-lic behavior in control structures leads to nonlinear coupled differential equations, which do not have analytical solution in the majority of engineering problems. However, it is possible to obtain approxi-mate solutions from the finite volume method (FVM). This method converts a continuous medium with in-finite variables into a discrete medium with establi-shed geometries and certain boundary conditions. Methods: Using the Ansys Fluent 17.0 software, a 1:70 scale 2D numerical model was developed wi-thout considering the effect of stacks and abutments on the weir under permanent flow conditions of the control structure and the radial gate system of the El Quimbo dam. The hydrodynamic analysis of the structure was carried out under free surface condi-tions and with different hydraulic loads, for openings of 704 meters above sea level, 706.9 meters above sea level, 709.4 meters above sea level, 712 meters above sea level and 724.6 meters above sea level. Subsequently, the numerical model was validated based on the comparison between the numerical re-sults and the theoretical data made by both Ingetec and the tests in the physical model developed by the National University of Colombia in Manizales. Results: For a return period of 100 years, the numeri-cal model estimated a sheet of 0.0306 m correspon-ding to 2,143 m in the El Quimbo structure. With this arrangement, 80.7996 Liters per second entered the control volume and 80.7324 Liters per second left, gi-ving rise to a variation of 0.08316%. The largest water sheet variation of the numerical model with respect to the observed sheet is 3.1% (0.258 m), which co-rresponds to the increasing probable maximum for an opening of gates with elevation (724.6 m) and with a discharge of 290 lps. Similarly, the higher speeds developed by the flow in the landfill originate at the exit of the deflector, corresponding to the abscissa K0 + 5,19 m. According to the numerical model, it was determined that the impact of the jet in the channel is made with a maximum speed of 6.15 m/s. Conclusions: The relation between the hydraulic load with respect to the design load must not exceed 1.33, which agrees with that recommended by the USBR. Also, the cavitation condition found in the fast can be considered incipient cavitation. In land-fill design, the H/HD ratio directly affects the appea-rance of cavitation phenomena on the structure, the development of high velocities in the fast, the deflec-tor, and the discharge capacity of the landfill.