https://tede.unioeste.br/handle/tede/528 Wed, 17 Jun 2026 03:27:22 GMT 2026-06-17T03:27:22Z https://tede.unioeste.br/handle/tede/8310 Title: Investigação das blendas de sebo bovino no desempenho e emissões em um motor ciclo diesel. Autor: Mathias, Marco Antonio Primeiro orientador: Machado Junior, Waldir Mariano Abstract: The growing demand for energy and environmental concerns have made clean energy production a priority in the 21st century. Biodiesel emerge as a promising alternative to reduce atmospheric pollution caused by the burning of fossil fuels. The use of residual raw materials, such as animal tallow, represents a sustainable solution by transforming waste into energy, reducing environmental impacts and contributing to the circular economy. Bovine tallow is currently the second main raw material used in biodiesel production in Brazil, although there are still few studies on its performance and emissions in diesel engines. The present study aimed to evaluate the specific fuel consumption and emissions of a single-cylinder diesel engine operating with different blends of beef tallow biodiesel (B10, B20, B30 and B100) under five different loads. Specific fuel consumption (SFC), exhaust gas temperature (EGT), and emissions of NOₓ, HC, CO and CO₂ were analyzed. It was observed that the specific consumption of the biodiesel blends was higher than that of fossil diesel due to their lower calorific value; however, a tendency for reduction was seen at higher loads. The results indicated good engine performance and lower NOₓ emissions compared to conventional diesel, confirming the potential of animal tallow biodiesel as a renewable and cleaner environmental alternative. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Dissertação Mon, 15 Dec 2025 00:00:00 GMT https://tede.unioeste.br/handle/tede/8310 2025-12-15T00:00:00Z https://tede.unioeste.br/handle/tede/8262 Title: Avaliação do Desempenho Térmico de Cobertura Fotovoltaica com Filme OPV Autor: Nunes, Camila Topanott Primeiro orientador: Santos, Reginaldo Ferreira Abstract: This study evaluated the thermal performance of three roof prototypes under a humid subtropical climate (Cascavel–PR): Prototype A (polycarbonate, EPS, and steel), Prototype B (same configuration with OPV film), and Prototype C (metal sandwich panel). The prototypes were installed in an open area on the UNIOESTE campus and instrumented with sensors for internal and external air temperature, radiant temperature, and solar radiation. Data were recorded at one-minute intervals over 30 days (September 5 to October 5, 2025), processed using a Hampel filter, and analyzed for the daytime period between 08:00 and 18:00. Thermal indicators were calculated to compare system performance, including internal–external temperature differences, thermal amplitudes, damping factor, time lags between radiation and internal temperature peaks, maximum heating and cooling rates, and mean radiant and operative temperatures. The results showed lower thermal amplitudes in Prototype A and intermediate behavior in Prototype B. The OPV film reduced thermal oscillations compared to Prototype C but maintained higher average internal temperatures. Since conductive heat transfer differences were minimal, the observed contrasts are mainly attributed to optical–radiative effects. It is concluded that, under the tested configuration, OPV acts as a partial thermal barrier, attenuating short-term thermal variations without reducing the average internal thermal load. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Dissertação Fri, 28 Nov 2025 00:00:00 GMT https://tede.unioeste.br/handle/tede/8262 2025-11-28T00:00:00Z https://tede.unioeste.br/handle/tede/8261 Title: DESENVOLVIMENTO DE ALGORITMO DE INTELIGÊNCIA ARTIFICIAL PARA DETECÇÃO DE FALHAS DE PLANTIO : UMA APLICAÇÃO AO CULTIVO DE MILHO Autor: Rosa, Rodrigo da Primeiro orientador: Santos, Reginaldo Ferreira Abstract: This ѕtᴜdу рrороѕеѕ the dеѕіgn оf an algorithm anchored in Artificial Intelligence, aiming to identify gaps in the planting process of energy crops, especially focusing on corn production.To achieve this, the Vision Transformer (ViT) architecture was employed, accompanied by transfer learning techniques, to train a model capable of classifying visual records captured by drones into three classificatory groups: plant, soil, and straw.The dataset was structured from authentic images of fields, provided by partners in the agricultural sector, and the training was carried out on the Google Colab platform, utilizing specialized libraries such as PyTorch and Hugging Face.The performances obtained evidenced by metrics such as accuracy and the F1-score indicate that the model proved effective in identifying irregularities in the planted spacing.Analyzing the results, a promising potential of AI is perceived as a strategic tool for intelligent management in precision agriculture, boosting operational efficiency and well-founded decision-making in the field. Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Dissertação Tue, 19 Aug 2025 00:00:00 GMT https://tede.unioeste.br/handle/tede/8261 2025-08-19T00:00:00Z https://tede.unioeste.br/handle/tede/8141 Title: Manejo da irrigação através da evapotranspiração em estágios fenológicos do níger (Guizotia abyssinica) Autor: Zitha, Agostinho Rodrigues Primeiro orientador: Santos, Reginaldo Ferreira Abstract: The objective was to evaluate the impact of water deficit during the phenological stages of Niger (Guizotia abyssinica). Two experiments were conducted at the State University of Western Paraná (UNIOESTE), in Cascavel, Paraná, Brazil: one in a protected environment and the other in open field conditions. The experimental design was completely randomized, with eight treatments and four replications (32 plots), as follows: T1-(Vegetative), T2-(Reproductive), T3-(Grain Formation), T4-(Vegetative and Reproductive), T5-(Vegetative and Grain Formation), T6-(Reproductive and Grain Formation), T7-(None of the Stages), and T8-(Vegetative, Reproductive, and Grain Formation). The substrate consisted of red soil and vermicompost humus placed in PVC pots. Irrigation was based on reference evapotranspiration (ETo), estimated using the Thornthwaite, Camargo, and Penman-Monteith methods adapted to the region. The cultivation in the protected environment was conducted under a high tunnel structure without temperature control. Irrigation was applied locally. In the field, sowing took place in an open area without climate control. In both environments, the plants’ water demand was adjusted using the crop coefficient (Kc), considering the phenological stages. The variables analyzed included plant height, number of leaves and capitula, stem and capitulum diameter, fresh and dry biomass (shoot and root), and yield. Statistical analysis was performed using the F test at a 5% significance level with the Sisvar software. In the greenhouse, T8 was significant for: root length (17 cm) and stem length (144.50 cm), fresh mass (3.49 g) and dry mass (1.80 g) of roots, fresh mass (40.37 g) and dry mass (8.89 g) of stems, stem diameter (8.85 mm), number of capitula (20.25), and number of leaves (35.50). Capitulum diameter showed no significant differences. Among the combined treatments, T4 stood out, and among the isolated ones, T1 was the most prominent. In the open field, T8 also showed the best results, particularly for: fresh mass (4.23 g) and dry mass (0.85 g) of roots, fresh mass (46.6 g) and dry mass (8.19 g) of stems, stem diameter (8.53 mm), number of capitula (33), and number of leaves (49). For certain variables, no significant differences were observed: stem and root length, number and weight of seeds, one thousand-grain weight, capitulum diameter, and productivity indicators. Among the combined treatments, T6 had the highest values. Among the isolated treatments, T3 showed the highest mean. T3 proved to be sensitive to water stress, and continuous irrigation enhanced growth and productivity. "This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001" Publisher: Universidade Estadual do Oeste do Paraná Tipo do documento: Dissertação Thu, 07 Aug 2025 00:00:00 GMT https://tede.unioeste.br/handle/tede/8141 2025-08-07T00:00:00Z