Mi DSpace
Please use this identifier to cite or link to this item: http://hdl.handle.net/UCSP/15668
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorMayhua López, Efrain Tito-
dc.contributor.authorSánchez Mora, Katty Miluska-
dc.description.abstractThe loss of macronutrients and micronutrients from agricultural land is one of the processes that are always present in agriculture. Thus, is necesary use of fertilizers NPK (Nitrogen, Phosphorus and Potassium) to optimize crop yields, increase profitability and minimize losses. Among the NPK fertilizers, nitrogen is the essential nutrient in the process of plant growth. Thus, it is that in agriculture nitrogen is applied to plants during irrigation, so care should be taken in the amounts of water that has been granted to the crop, as its excess leads to a loss of nitrogen and other nutrients by leaching. Given this reality, technology capable of monitoring and measure the Nitrogen levels in soil locally and in real time, so that in this way the necessary amount of nutrients can be given to the crop. In recent years there has been proposed a number of techniques and methods (direct and indirect) for measurement of nitrogen. Direct methods can be in-situ or in laboratory, but they tend to be really expensive and/or dependent on soil conditions. Instead indirect methods, can estimate Nitrogen levels in-situ and in real time, based on measuring other parameters, at the expense of accuracy. Weighing the advantages and disadvantages, is that an indirect method was designed, based on algorithms own from the machine learning area (ML), it will be capable of predicting future values of the Nitrogen levels in soil, after learning a model from a set of observations. These observations are given by parameters such as electrical conductivity, temperature and soil moisture. The values of these parameters were acquired through sensors that exist in the market. For the validation of the method, experimental tests were carried out with real data that were measured in the Santa Gabriela S.A.C. of Santa Rita de Siguas, Arequipa, Peru. From the results obtained from the estimator it has been noted that the most influential variables to obtain the estimated nitrogen levels are the electrical conductivity and temperature, which gives reasonable results within a confidence interval of 99% given by the range of permitted levels for the crop.es_PE
dc.publisherUniversidad Católica San Pabloes_PE
dc.sourceUniversidad Católica San Pabloes_PE
dc.sourceRepositorio Institucional - UCSPes_PE
dc.subjectAgricultural soiles_PE
dc.subjectMachine learninges_PE
dc.titleDiseño de un estimador no lineal para predecir el nivel de nitrógeno en suelo agrícolaes_PE
thesis.degree.nameIngeniero de Telecomunicacioneses_PE
thesis.degree.grantorUniversidad Católica San Pablo. Facultad de Ingeniería y Computaciónes_PE
thesis.degree.levelTítulo Profesionales_PE
thesis.degree.disciplineIngeniería Electrónica y de Telecomunicacioneses_PE
thesis.degree.programEscuela Profesional de Ingeniería Electrónica y de Telecomunicacioneses_PE
Appears in Collections:Tesis Pregrado - Ingeniería Electrónica y de Telecomunicaciones

Files in This Item:
File Description SizeFormat 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.