Resumo em Inglês:

ABSTRACT The combine harvester is a widely used piece of agricultural equipment in modern agriculture, and the seed loss rate is one of the important indexes used to measure its operational performance, so the monitoring of the seed loss rate is crucial for adjusting the operational parameters of combine harvesters and improving the quality of grain harvesting. Aiming at the problems of the slow response speed and low monitoring accuracy of the existing domestic seed loss rate monitoring models, this paper proposed a rice seed loss rate monitoring method based on the whale optimization algorithm-back propagation neural network (WOA-BP). The loss rate monitoring device consisted of a piezoelectric ceramic sensor module, charge amplification circuit, band-pass filter circuit, analog-to-digital (AD) converter, main control unit, etc. The WOA-BP algorithm, which has a high accuracy and fast response speed, was used to classify and count the signals to realise seed loss rate monitoring. The indoor test results showed that the relative errors of the monitoring results are less than 8.5% under the condition of a conveying speed of 1.3-2.1m/s, and the relative errors showed an increasing trend as the proportion of straw increased.

Resumo em Inglês:

ABSTRACT At present, there are serious problems in sorghum harvesting, such as lodging entanglement and the loss of broken stems due to the lack of an external dividing device for the cutting table and holding device. Based on the physical and mechanical characteristics of sorghum plants, an integrated outer divider was developed. The main structural parameters and working parameters of the outer divider were determined. A comparative test of the working quality of the outer divider was carried out. The results show that the working quality of the cutting table with the outer divider is obviously better than that without the outer divider and that the harvest loss of the cutting table is effectively reduced. The Box–Behnken experimental design method was used to investigate the effects of the forward speed, rotation speed of the grain lifter and dividing angle of the outer divider on the lodging and broken stem loss rates during sorghum harvesting. The regression mathematical model and response surface of the lodging and broken stem loss rates and the analysis factors were established, and the optimal working parameters of the outer divider were determined as follows: the dividing angle of the outer divider was 20°, the forward speed was 0.8 m/s, and the rotating speed of the grain lifter was 330 rpm. Under these parameters, the loss rate of the fall was 1.08%, and the loss rate of broken stems was 1.05%, which met the requirements of sorghum cutting tables

Resumo em Inglês:

ABSTRACT The anti-clogging device of a no-till seeder is an important component that affects the seeding quality. In one year two crop area of China, a no-tillage seeder for corn generally includes a passive anti-clogging device, and there are often problems with straw winding around working parts due to the low straw cleaning rate. To solve these problems, a support cutting anti-clogging device with vibration for a no-till seeder was designed in this study to efficiently cut wheat stalks and remove them to the sides of the seedbed. Through theoretical calculations and kinematic analysis, the main structural parameters of the device were limited to small ranges of values: the mounting angle of the vibrating knife was in the range θ1 = 70–82°, the amplification was in the range l1 = 14–24 mm, and the frequency was in the range ωm = 240–340 rad/min. To analyse the effects of the main parameters on the average straw cutting rate, soil bin experiments were carried out using an orthogonal multinomial regressive experimental design with three factors and three levels. The optimal structural parameters derived in this way were as follows: mounting angle for the vibrating knife θ1 = 70.7°, amplification l1 = 19 mm, and frequency ωm = 330.2 rad/min. The results of adaptability and comparison tests showed that the average straw cutting rate was 93.30% when the proposed anti-clogging device with vibration was used. Compared to traditional devices, the cutting rate for the device was increased by 12.82%, and the metrics were superior to those of a traditional anti-clogging device. This research can serve as a reference for designing anti-clogging devices for no-till seeders.

Resumo em Inglês:

ABSTRACT Computer vision systems combined with machine learning techniques have demonstrated success as alternatives to empirical methods for classification and selection. This study aimed to classify tomatoes based on their colorimetric characteristics, which influence consumer purchasing potential, using the decision tree algorithm. Tomatoes were categorized into two classes based on ripeness: Higher Purchasing Potential (20 fruits) and Lower Purchasing Potential (40 fruits). Images were captured in the RGB color model and converted to HSI and CIELab models. Principal component analysis was employed to evaluate the influence of colorimetric characteristics within each class, and the decision tree algorithm was applied to classify the fruits into the respective categories. Tomatoes in the Higher Purchasing Potential class were primarily influenced by red intensity and chromaticity a and b, while tomatoes in the Lower Purchasing Potential class were influenced by green intensity and hue. The decision tree achieved an accuracy of 83.6% and an F1-score of 90.9%, demonstrating its potential for classifying tomatoes based on colorimetric characteristics linked to consumer preferences.

Resumo em Inglês:

ABSTRACT Due to the complex distribution of greenhouse seedbeds and the limited area of greenhouses, the accuracy of the positioning of modular planting of crops is required to be high. Moreover, modern greenhouses are mostly vertical planting structures. Most of the existing positioning modes only support plane positioning, and cannot determine the specific position of the seedbed in space through three-dimensional space positioning. In order to spray pesticides on various crops planted vertically in a modern greenhouse, a greenhouse fixed-point spraying system based on Ultra-Wideband (UWB) indoor positioning was designed. The system uses UWB indoor positioning technology to cope with the complex environments in modern greenhouses. Four base stations and one label are used to obtain the position coordinates of vertically planted crops. This information is processed through an embedded kernel, and then the upper computer sends instructions remotely. The motor drives a screw to rotate and move the nozzle to the location of the crop to complete the pesticide spraying. Experiments show that the real-time accuracy in the coordinates collected by this system is below 10 cm, which makes fixed-point spraying feasible in a modern greenhouse.

Resumo em Inglês:

ABSTRACT Emitter clogging is the main limitation of drip irrigation systems operating with wastewater. This paper aimed to employ discriminant analysis (DA) to generate classification functions that characterize aquaculture wastewater (AW) dilutions in well water (WA), delivered through non-self-compensating drippers. Five AW dilutions in WA were tested (D1: 100% AW; D2: 75% AW + 25% WA; D3: 50% AW + 50% WA; D4: 25% AW + 75% WA; and D5: 100% WA) to investigate the clogging susceptibility of three non-self-compensating dippers: TS (1.6 L h-1), SL (1.6 L h-1), and NJ (1.7 L h-1) after 160 h of operation. Three hydraulic performance evaluations of the drippers were performed in this period. During the same interval, the quality attributes of the AW dilutions in WA were also quantified. The statistical analyses included correlation matrix and DA. The correlation matrix identified 188 variables with significant correlations. Discriminant functions were constructed for each dripper using DA. These functions revealed Mg2+ as the most significant variable. The classification matrix of these functions achieved a 100% success rate.

Resumo em Inglês:

ABSTRACT The selection of materials and substrates is essential for optimizing the thermal performance of green roofs. However, there has been limited research on green roof characteristics under subtropical conditions. Therefore, this study aims to evaluate the internal and substrate temperatures of six green roof prototypes and one control prototype. Prototypes with clay tiles (control), clay substrates with and without vegetation, sandy substrates with and without vegetation, and organic matter substrates with and without vegetation are evaluated. The experimental design involves randomized blocks and the internal and substrate temperatures are monitored. The vegetated sandy substrate prototype exhibits the highest thermal performance, with internal temperatures 0.6 ℃ lower than those of other green roof prototypes and 1.7 ℃ lower than that of the control with clay tiles. This is attributed to the high porosity of the sandy substrate, which enhances thermal insulation. To provide optimal thermal performance, the substrate must have a water retention capacity that is sufficient to guarantee vegetation development, but not excessive so that it constantly increases the thermal conductivity owing to substrate saturation.

Resumo em Inglês:

ABSTRACT During blueberry picking operations, changes in the blueberry harvester’s body posture (BHBP) caused by undulating farmland surfaces significantly affect the operational stability of the harvesting device. To mitigate these effects, an active suspension (AS) control system was developed. First, considering the coherence and time lag of the four-wheel tracks, the filtered white noise method was chosen to generate road excitations. Subsequently, the virtual prototype model of the blueberry harvester (BH) was built in ADAMS with the nonlinear properties of the passive suspension. A fuzzy proportion integral differential (PID) control algorithm and decoupled control strategy were employed to design the AS control system. Finally, ADAMS-MATLAB co-simulations and field tests were conducted. The results indicate that the measured farmland road profile conforms to standard D-level road excitations according to ISO8608, and the co-simulation model accurately predicts the BH’s dynamic response. Under AS control, the vertical acceleration, pitch acceleration, and roll acceleration of the BH decreased by 38.52%, 37.39%, and 34.29%, respectively, during simulations compared to the passive suspension. In field tests, these reductions were 36.51%, 33.84%, and 30.21%, respectively. The AS control system proposed in this study significantly improves the stability of the BH under farmland road excitation, effectively mitigating equipment impacts and wear caused by machine jolts while enhancing harvesting performance. This research addresses the gap in applying AS technology within the field of harvesting machinery, offering a novel technical approach for the development of vehicle control systems for harvesting machinery targeting blueberries and other shrub crops. It holds considerable theoretical value and practical significance.

Resumo em Inglês:

ABSTRACT In light of the problems of large operation resistance and small soil fragmentation during the harvesting operations of existing cassava harvesters, a long- and short-toothed digging shovel was designed. A virtual simulation soil trough model of cassava ridge soil particles was established using the discrete element method, and the Hertz–Mindlin with JKR contact model was employed to simulate the operation quality of the long- and short-toothed digging shovel and the original digging shovel. In the movement and force analysis of the digging shovel, the angle of entry, the advance speed of the machine, and the height of the digging adjustment were the test factors. The response surface test was conducted on the digging rate and the damaged cassava rate. The results of the experimental field trial showed that the average digging rate of harvested cassava increased by 2.56%, and the average rate of damaged harvested cassava decreased by 1.54%, compared with the original digging shovel. The digging operation process was stable and met the requirements of cassava harvesting field operations. The results of this study may inform future studies on the design and improvement of a cassava harvester.

Resumo em Inglês:

ABSTRACT Traditional methods for assessing crop water status have limited practicality for field applications. Conversely, remote sensing via remotely piloted aircraft (RPA) offers a promising alternative, though its effectiveness requires validation in specific studies. This study utilized RPA-derived imagery to support strategies for monitoring water deficit (WD) in corn crops and enabling yield prediction. The AG 1051 corn genotype underwent deficit irrigation levels (20, 40, 60, 80, and 100% of crop evapotranspiration—ETc) at different phenological stages: initial (E1), vegetative growth (E2), flowering (E3), and physiological maturity (E4) in Iguatu, Ceará, Brazil. We measured ten vegetation indices (VIs) and ear and biomass yield. Pearson's correlation analysis revealed that, during E2, the normalized difference vegetation index (NDVI) (r = 0.98) and green leaf index (GLI) (r = 0.99) were the most reliable for distinguishing water stress levels. These indices were also effective in predicting yield in E2 through regression analysis. The findings demonstrate that vegetation indices derived from RPA imagery provide a robust method for assessing water conditions and forecasting yield.

Resumo em Inglês:

ABSTRACT Precision control of tillage depth is crucial for optimizing soil preparation in tea plantations. This study presents an adaptive real-time tillage depth control system based on Linear Active Disturbance Rejection Control (LADRC) to address precision challenges in the tillage depth of electric rotary tillers in tea plantations. The system, constructed using body posture sensors, control units, and hybrid stepper motors, integrates sensor data and LADRC technology to drive the stepper motor, enabling precise tillage depth control. The displacement sensor signals were collected, and the actual tillage depth was compared with the target values, allowing adjustments to achieve closed-loop control of the rotary tiller. Field experiments at speeds of 0.5 km/h and 0.8 km/h with tillage depths of 80 mm and 100 mm demonstrate the system’s effectiveness. The LADRC system achieved a standard deviation of 3.2 mm, outperforming fuzzy PID (10.5 mm) and sliding mode control (5.9 mm). The rate of depth variation was reduced by 44.8% and 68.9% compared to the fuzzy Proportional Integral Derivative (PID) and SMC, respectively. These results confirm that the LADRC-based system effectively minimizes interference during rotary-tiller operation, ensuring the stability and reliability of tillage depth control.

Resumo em Inglês:

ABSTRACT The integration multispectral sensors with machine learning algorithms has demonstrated increasing efficacy in the classification of various maize morphophysiological characteristics. The hypothesis of this study is that maize plants subjected to different irrigation management practices exhibit distinct spectral behaviors, allowing for their classification through machine learning modeling. Thus, the objective of this study is to classify maize hybrids in different irrigation management practices using multispectral images. This involves identifying the most effective machine learning algorithms and inputs variables that enhance model performance for accurate classification. The experiment was conducted at the experimental facility of the Federal University of Mato Grosso do Sul, in Chapadão do Sul – MS. Seven hybrids were evaluated: H1 (AS 1868), H2 (DKB 360), H3 (FS 615 PWU), H4 (K 7510 VIP3), H5 (NK 520 VIP3), H6 (P 3858 PWU), and H7 (SS 182E VIP3). These hybrids were subjected to irrigation and non-irrigation management practices. Sixty days after crop emergence, images were captured in the blue (475 nm, B_475), green (550 nm, G_550), red (660 nm, R_660), red edge (735 nm, RE_735), and near-infrared (790 nm, NIR_790) bands using the Sensefly eBee RTK fixed-wing Remotely Piloted Aircraft, equipped with a Parrot Sequoia multispectral sensor and RTK (Real-Time Kinematics) technology. Through the collected band data, the ESRI ArcGIS 10.5 geographic information system software was used to calculate 41 vegetation indices (VIs). Data were analyzed using machine learning techniques, testing six algorithms: Logistic Regression (RL), REPTree (DT), J48 Decision Trees (J48), Random Forest (RF), Artificial Neural Networks (ANN) and Support Vector Machine (SVM). Three accuracy metrics were utilized to evaluate the algorithms in the classification of irrigation management: correct classifications (CC), Kappa coefficient and F-Score. The ANN and RF algorithms demonstrated better accuracy in classifying maize hybrids with respect to irrigation management. The use of Vegetation Indices (IVs) and Spectral Bands + Vegetation Indices (SB+IVs) enhanced performance of these algorithms.

Resumo em Inglês:

ABSTRACT The potential of oilseed radish at two sowing dates for use as a raw material in biogas (biomethane) production based on laboratory anaerobic digestion with the addition of inoculum with a 60-day incubation period was investigated. The results of biogas productivity were compared with traditional cruciferous species used for bioenergy purposes. In the spring-sown variants, the achievable level of ground bioproductivity of oilseed radish was set at 25.17 t ha−1 of raw and 3.20 t ha−1 of dry matter, which provided a biomethane yield (SMY) of 320.07 ± 31.39 LN kg−1ODM and an indicator of methane accumulation intensity (Rm(ef)) of 130.76 ± 10.20 LN kg−1ODM d−1 with an appropriate biochemical portfolio of the formed biomass. During the summer sowing period, the average bioproductivity of oilseed radish was 18.42 t ha−1 in raw weight and 2.81 t ha−1 in dry matter, which provided an SMY of 262.97 ± 24.64 LN kg−1ODM and an Rm(ef) of 122.22 ± 3.62 LN kg−1ODM d−1 with its appropriate biochemical composition. The maximum level of biomethane production from oilseed radish was achieved with spring sowing under the conditions of 2021, resulting in the following technological parameters of productivity: MS 55.84 ± 9.39%, SMY 359.25 ± 11.24 LN kg−1ODM, Rm(ef) 138.15 ± 1.78 LN kg−1ODM d−1, Rm(full) 31.51 ± 1.69 LN kg−1ODM d−1, t50 4.12 ± 0.34 days, and λ 1.74 ± 0.17 days.

Resumo em Inglês:

ABSTRACT The operating costs and investment in microgeneration plants are high, and the maximum efficiency of the process should be targeted such that farmers can obtain the best financial returns. In this study, the efficiency of a microbiogas plant was estimated using a methodological analysis that involved a biodigester mass balance and plant energy balance. The plant had two biodigesters and an engine generator. The biodigesters under study were considered low tech. The results showed 65% utilization of the volatile solids present in the substrate by the biodigesters, producing biogas with an average methane concentration of 57%. The equipment necessary for the operation of the system consumed 19.45% of the energy produced, and most of this energy was used for the transportation of substrates. The energy balance of the plant indicated an energy efficiency of 27.24% during the production process. The hydraulic retention time, pH, intermediate/partial alkalinity ratio, and organic loading rate indicate that reactor A is underutilized, has a low nutrient supply, and may result in low methane production. These indicators suggest that biogas production and quality can be improved by better distribution of substrates among the digesters.

Resumo em Inglês:

Due to the complex terrain and variable soil conditions in hilly areas, traditional potato harvesters often face challenges such as high excavation resistance, insufficient soil disturbance, and high crop loss rates. This paper, based on soil mechanics and the working principles of potato harvesters, uses EDEM (Discrete Element Method) discrete element software to simulate different combinations of blade angle differences, and systematically studies their on-excavation performance. The results show that the angle difference Δ1 between blade angles α1 and α2, and the angle difference Δ2 between α2 and α3, significantly affect excavation performance. Proper blade angle design can significantly reduce excavation resistance, enhance soil disturbance effects, and reduce power consumption. Specifically, blade pressure fluctuates with increasing angle difference; forward resistance significantly increases under conditions of large angle differences; and within a certain range of Δ2, the blade operates with higher efficiency and lower power consumption. In summary, rational optimization of blade angle design can improve harvest efficiency and crop integrity.

Resumo em Inglês:

ABSTRACT To enhance the automation and efficiency of combine harvesters, this paper proposes a predictive control method based on Long Short-Term Memory (LSTM) neural networks. The method integrates multi-sensor data fusion using an Extended Kalman Filter (EKF) to improve speed measurement accuracy. By considering feeding volume, operational performance indicators, and critical component speeds, an LSTM-based model predicts the optimal operation speed. The predicted speed is then regulated through an incremental proportional-integral-derivative (PID) control control system. Simulation and field experiments validate the effectiveness of the proposed approach, demonstrating improved speed stability and work efficiency. The results indicate that the system enhances operational performance and reduces manual intervention, contributing to the advancement of intelligent agricultural machinery.

Resumo em Inglês:

ABSTRACT A two-wheel tractor-drawn garlic seeder was developed. The seeder was composed of a multiple adjustment mechanism, transmission and mulching devices, a furrow opener, a ground roller, and a depth adjustment device. The problem of multiple seeds being placed in a single hole was solved through the vibration and change in volume of the seed-picking spoon. The problems with the seed orientation and sowing depth were overcome by the multi-stage cam adjustment mechanism and the depth adjustment device, respectively. After the key components of the new garlic seeder were successfully developed, product manufacturing, simulation verification, and testing were performed. The results showed that an average correct seed orientation rate of 90.97%, a missing-seed rate of 1.12%, and a single-hole repetition rate of 1.16% were achieved. The depth adjustment device ensured the required garlic sowing depth, and its sowing efficiency was 25–45 times that of manual planting, effectively meeting the needs of growers.

Resumo em Inglês:

ABSTRACT Sugarcane field renewal is usually carried out using conventional soil tillage, followed by sugarcane planting or by soybean cultivation or cover cropping. However, efforts have often been made to reduce the intensive operations involved in conventional soil preparation. The objective of this study was to evaluate the physical quality of the soil under conventional tillage (CT) and no-tillage (NT) systems after a sugarcane harvest cycle, before soil preparation and after four months of cultivation of Crotalaria juncea intercropped with Urochloa ruzizienses cv. Xaraés. This study was conducted in the municipality of Juti, Mato Grosso do Sul, Brazil, on a dystrophic psammitic Red Latosol. A randomized block design was used in a split-plot scheme with thirty replications. The main plots consisted of two soil management systems, and the subplots corresponded to evaluation periods (before and after soil preparation and cover crop cultivation). The assessed attributes were soil bulk density (BD), macroporosity (Ma), microporosity (Mi), and total porosity (TP), as well as soil penetration resistance (PR) and soil moisture (SM). The CT promoted greater reductions in soil bulk density and penetration resistance, as well as increases in total porosity and macroporosity in deeper soil layers, compared to NT, which showed improvements particularly in the surface layer. In areas with no chemical or biological limitations in the soil profile, the adoption of NT combined with the cultivation of Crotalaria juncea intercropped with Urochloa ruzizienses cv. Xaraés, is recommended to improve the physical quality of the soil for sugarcane cultivation.

Resumo em Inglês:

ABSTRACT omputer vision systems have proven to be a promising alternative for assessing fruit quality attributes in a non-invasive, instantaneous, and accurate manner. This study aimed to use colorimetric characteristics extracted from digital images to estimate tomato fruit firmness through multivariate modeling. Images of 80 tomato fruits at four ripening stages were acquired using two digital cameras, enabling the extraction of average intensity values for the red, green, blue, and near-infrared bands, followed by the calculation of colorimetric indices. Reference firmness values were measured using a digital fruit penetrometer. Colorimetric indices were employed to estimate fruit firmness using principal component regression. Principal component analysis enabled the dimensionality to be reduced to a single principal component (explanatory percentage of the data variance of 97.06%), which was used to generate firmness estimation equations. The application of the model to the validation dataset yielded an R2 = 0.937 and a mean standard error (SE) of 2.05 N, demonstrating that the protocol based on colorimetric characteristics extracted from digital images is suitable for estimating tomato firmness.

Resumo em Inglês:

ABSTRACT Soil compaction alters soil structure, reduces soil air permeability, and adversely impacts crop growth. However, the rapid field detection of soil compaction remains a significant technical challenge. To address this issue, this study describes the development of an unsaturated soil bulk density measurement system based on soil air permeability, which enables the estimation of soil compaction through the measurement of soil air permeability. The system uses a specially designed soil probe that injects air into the soil at a constant pressure (P0) and flow rate (Q0). The steady-state air pressure (ΔP0) and flow rate (ΔQ0) at a radius r0 were recorded to calculate the soil gas permeability coefficient (ka). Laboratory experiments demonstrated a significant relationship between the soil gas permeability coefficient (ka), soil bulk density (ρ), and soil moisture content (w): soil compaction, which increases the bulk density, markedly reduces ka, while a higher soil water content further diminishes air permeability. Using experimental data, a bulk density estimation model was constructed with ka and w as independent variables. Field validation showed that the model achieved an accuracy of R2=0.72, thus demonstrating its reliability. This method is operationally straightforward, and allows for the rapid assessment of soil compaction. It provides a robust tool for precision field management and enhanced crop productivity.

Resumo em Inglês:

ABSTRACT Technological advances in soybean (Glycine max L.) cultivation, particularly with Xtend technology cultivars, require appropriate and conscious management practices. This study aimed to identify canonical interrelationships among the physical, physiological, and nutritional parameters of soybean seeds carrying the DMO gene, using dicamba applied in both pre- and post-emergence stages. The experiment was conducted in Augusto Pestana, RS, Brazil, during the 2022/2023 growing season. Treatments consisted of eight dicamba application timings (pre- and post-emergence) across five Xtend cultivars. After harvesting, seed physical traits (diameter, circumference, perimeter, and area), germination, seedling viability, and seed nutritional components were evaluated. Seed diameter, perimeter, and area varied among cultivars, and reductions in these traits were associated with lower nutritional quality. Dicamba applications during the vegetative stages caused less damage to the physiological and nutraceutical quality of the seeds.

Resumo em Inglês:

ABSTRACT Cherries are popular with consumers due to their high nutritional value. Their appearance characteristics are essential criteria for consumer purchasing and are closely related to ripeness classification. However, the existing deep learning methods for cherry appearance quality detection face a trade-off between accuracy and computational efficiency, while manual sorting remains labor-intensive and costly. To address these issues, we introduce an improved regular network (RegNet), aiming to accurately detect the visual quality of fruit. This is achieved by using different combinations of deep features and classifier models. The method involves extracting deep features from 12 RegNet models of varying parameter scales and combining these features with nine classifiers to assess cherry appearance quality. The proposed method achieves an accuracy of 98.7% in detecting the quality of cherry fruit, with a significant reduction in the training time. Experiments demonstrate that our method can rapidly and accurately evaluate cherry quality, offering a practical solution for automatic sorting systems.

Resumo em Inglês:

ABSTRACT At present, in the field of cherry recognition, problems such as dense fruit growth and frequent occlusions of branches and leaves affect the accuracy of the recognition process. To address these issues, this paper proposes a lightweight cherry detection model based on an improved YOLOv8n (CHERRY-YOLO). By introducing a FasterNet block module based on partial convolution (PConv) and replacing the bottleneck module in YOLOv8n's C2f with a FasterNet block module, a new C2f-Faster module is formed. As the module reduces the number of parameters and calculations, it also ensures high detection accuracy. By adding a convolutional block attention module (CBAM) between the SPPF module and C2f-Faster module in the backbone network area, the ability to extract target features is enhanced before multiscale feature map fusion, further improving the target detection accuracy. The experimental results show that CHERRY-YOLO has more advantages than the original YOLOv8n model. Its precision, mAP@0.5, mAP@0.5:0.95, and recall increase by 1.3%, 1.0%, 0.6%, and 0.4%, respectively. Moreover, the number of parameters, computational size, and weight decrease by 21.27%, 20.99%, and 20.63%, respectively. The results demonstrate that CHERRY-YOLO provides important technical support for automatic cherry picking.

Resumo em Inglês:

ABSTRACT The western region of Paraná is an important center for agricultural production, particularly in the pork chain. This study estimated the energy incorporated in this chain by analyzing three sectors: grain production system, feed mill, and pig breeding system. The physical dimensions of the systems and the energy incorporation rates were evaluated, resulting in energy units. The results showed that the grain production system and the feed mill had an energy return on investment (EROI) below one, indicating a negative return on the energy invested, whereas the energy balance (BE) reflected a net gain. The embodied energy (EI) was 3.42 MJ kg-1 for soybean, 1.89 MJ kg-1 for corn, and 0.17 MJ kg-1 for feed. In the pig breeding system, the average EROI was 4.68, with no energy return, and the BE was -4,205.50 GJ lot-1, with an EI of 43.11 MJ kg-1 in the feed produced.

Resumo em Inglês:

ABSTRACT Beans are a legume that is widely grown and consumed globally, being the staple food for humans in developing countries. Nitrogen (N) is the most limiting nutrient for yield and foliar analysis is crucial to ensure a balanced nitrogen fertilization. However, conventional methods are time-consuming, requiring new technologies to optimize the supply of N. In this work, the performance of two deep learning models in the classification of leaf nitrogen in beans using RGB images was evaluated and compared. The BRS Estilo was used in a greenhouse in a completely randomized design with 4 doses of nitrogen (0, 25, 50 e 75 kg N ha-1) and 12 reps. The image bank was composed of 4 subfolders, each containing 500 images of 224x224 pixels obtained from plants grown under different doses of N. Matlab© R2022b were used for processing of the models. The performance of ResNet-50 was superior when compared to CNN, with an accuracy test of 84%, while the value observed for CNN was 82%. The use of images combined with deep learning can be a promising alternative to slow laboratory analyses, optimizing the estimation of leaf N and providing a rapid intervention by the producer to achieve higher productivity.

Resumo em Inglês:

ABSTRACT Babassu (Orbignya speciosa Mart.) is an oil palm that can be used as an alternative in biodiesel production. However, few studies have compared the babassu and soybean biodiesel production. Therefore, the objective of this study was to compare the performance and emissions of babassu and soybean biodiesels under different engine loads. Four babassu and soybean blends (B10, B15, B20, and B100) were used in a 3-kVA generator engine at resistive loads ranging between 500 and 2500 W. The measured parameters were the specific fuel consumption (SFC), exhaust gas temperature (EGT), nitrogen oxides (NOx), hydrocarbon (HC), and carbon monoxide (CO). The results showed that the SFC of the babassu biodiesel was 9% lower than that of soybean B20 at a 2500-W load. Interestingly, the NOx emissions and EGT decreased for B100 at all tested loads for babassu biodiesel compared with those for soybean biodiesel. Conversely, the soybean biodiesel exhibited lower HC emissions than babassu biodiesel at all loads. The B20 babassu biodiesel blend exhibited better performance in terms of HC and CO emissions. Therefore, babassu biodiesel can be used as an alternative in biodiesel production.

Resumo em Inglês:

ABSTRACT This paper addresses the issues of insufficient accuracy and energy efficiency balance in the weed detection models that are currently used in agricultural weeding robots. To tackle this problem, we propose an enhanced YOLOv11-SNEW algorithm based on the YOLOv11 framework, which is specifically designed for weed detection in real agricultural settings. In the proposed algorithm, the backbone network is replaced with the lightweight ShuffleNetV2, the NAM attention mechanism is introduced, the EMSCP optimisation module is incorporated, and the loss function is improved. Experimental results demonstrate that the YOLOv11-SNEW model excels in weed detection, with an average recognition accuracy of 93.5%, a recall rate of 90.1%, and an mAP50 value of 91.4%. These metrics represent significant improvements in detection accuracy, recall rate, and mAP50 compared to the original YOLOv11 model and other comparative models, with a significant reductions in the number of parameters and computational effort. The model also exhibits greater robustness in complex environments, with a noticeable reduction in detection leakage. This approach provides a more efficient and precise solution for weed detection in agricultural production, thereby promoting the development of precision agriculture.

Resumo em Inglês:

ABSTRACT Under the cultivation mode of walnut dwarfism and dense planting, existing walnut vibration harvesting machinery has the problems of poor adaptability and low harvesting efficiency. A vibration walnut harvester was designed, paying particular attention to the design of the trunk clamping mechanism, the symmetrical double eccentric vibration mechanism, and the hydraulic control system; determination of the optimal harvesting parameters was also undertaken. A forced vibration dynamics model of the tree was developed and the fruit shedding dynamics equation was built. The key factors influencing harvesting performance were determined, including the mass of the eccentric block, the vibration frequency, and the clamping height. In order to determine the optimal working parameters, a three-factor, three-level field performance test was conducted, resulting in a regression model for fruit picking rate and tree vibration acceleration. The test results showed that, when the mass of the eccentric block was set to 63.2 kg, the vibration frequency was 8.7 Hz and the clamping height was 83.2 mm. A walnut picking rate of 90.42% was achieved and the tree vibration acceleration reached 55.02 m/s2, which meets the operational requirements for walnut harvesting. This study provides a theoretical foundation for improving the efficiency of vibratory harvesting of walnut trees.

Resumo em Inglês:

ABSTRACT This study aimed to evaluate the effect of different drying temperatures on the physicochemical characteristics of Brazil nut kernels during storage. Nuts were dried at five temperatures (40, 50, 60, 70, and 80 °C) until reaching a moisture content of about 4.0% (wb). Kernels were then stored under laboratory conditions (27.77 ± 1.77 °C and 58.27 ± 8.57% RH) for 0, 2, 4, 6, 8, and 10 months. At each storage period, water, protein, ash, lipid, mineral, pH, titratable acidity, total soluble solids, and colorimetric parameters were analyzed. Results showed that protein content remained constant throughout storage, with values of 13.7 and 14.4% at 40 and 80 °C, respectively. Higher ash content was recorded at 70 and 80 °C (3.6 and 3.8%). Lipid content was the lowest at 60 °C (65.3%). The highest macro- and micromineral concentrations were observed for nitrogen (24.5 g kg⁻1), iron (62.4 g kg⁻1), and zinc (48.64 g kg⁻1). Based on physicochemical characteristics, Brazil nut kernels remained viable for consumption throughout ten months of storage.

Resumo em Inglês:

ABSTRACT Image-based systems are designed to approximate human color perception, accounting for nonlinear nature of vision. Non-destructive methods predicting parameters traditionally obtained through destructive analysis enable larger sample sizes to be assessed with sufficient accuracy. This study developed a model to estimate lycopene and β-carotene levels in fresh and processed tomato varieties using non-destructive spectrophotometric parameters. Models showed strong agreement with observed values, yielding correlation coefficients of 0.71 for lycopene and 0.79 for β-carotene, with minimal associated error. These findings demonstrate that carotenoid levels can be reliably estimated through fruit color, reducing need for destructive analyses.

Resumo em Inglês:

ABSTRACT Drum speed is a key parameter controlling threshing intensity: excessively high speeds cause a high proportion of broken grains (due to uneven impact), while excessively low speeds leave more unthreshed grains. It is difficult to balance these two effects by adjusting only the drum speed of a threshing device. Since threshing teeth structure affects the impact force on soybeans, optimising tooth shape is essential to address speed-related defects. This study presents a design for arc-shaped threshing teeth based on logarithmic spiral self-similarity, which maintains a relatively constant contact angle with the crop along the tooth surface. This makes the impact force smoother and more uniform (eliminating large fluctuations), simplifies the tooth profile, balances the impact force distribution, and effectively addresses the trade-off between the proportions of broken and unthreshed grains. Bench tests were conducted with drum speed as the influencing factor and the proportions of broken and unthreshed grains as evaluation indicators. Results showed that under optimal parameters, 15° arc-shaped teeth achieved a proportion of broken grains of 0.61% (0.68% lower than traditional teeth) and a proportion of unthreshed grains of 2.22% (only 0.1% higher). It is concluded that 15° arc-shaped teeth improve threshing device performance.

Resumo em Inglês:

ABSTRACT Nitrogen (N) is a key factor for achieving high grain yield in irrigated rice, and remote sensing tools, such as the Normalized Difference Vegetation Index (NDVI), can support optimization of N application during the growing season. This study aimed to evaluate the use of NDVI for estimating biomass, shoot N accumulation, and grain yield, as well as to assess its potential for guiding N fertilization in rice. Field experiments were conducted in the state of Rio Grande do Sul, Brazil, during the 2021/22 and 2022/23 growing seasons, using the genotype IRGA 424 RI and five N fertilization levels. NDVI, measured throughout the crop cycle, detected differences in plant growth associated with N availability. Strong relationships were observed between NDVI values and the evaluated parameters, indicating that the index can predict plant N status and estimate crop yield. Based on NDVI readings obtained with a proximal vegetation sensor (GreenSeeker), a model was developed to classify crop biophysical parameters and to guide site-specific N fertilization during the growing season.

Resumo em Inglês:

ABSTRACT The contribution of photovoltaic energy has steadily increased within the energy matrix worldwide. Within this renewable energy system, optimization is there is continually pursued, which can be achieved through various approaches, including advancements in photovoltaic cell technology, impedance adjustments, inverter efficiency, collected energy storage (either on- or off-grid), cell cooling, module cleaning, and solar position tracking, which was the focus of this study. In remote areas without electrical grid access, motorized systems may not be feasible. A tracking prototype was developed whose structure is classified as chronological due to its fixed rotation, manual as it requires daily adjustment, single-axis azimuthal, and "analog" as it lacks motors, sensors, or algorithms. The results showed that the developed tracking prototype achieved a 9.69% energy production increase during the research period, with peaks of 35.51% on sunny days, and a significant efficiency improvement during the early morning and late afternoon. Statistical analysis results revealed a significant energy production difference between panels. In terms of economy, the prototype proved unviable when compared to the cost of grid electricity.

Resumo em Inglês:

ABSTRACT Hydrological data are typically sequential and often correspond to large datasets, and not all application programs allow their retrieval, storage, and manipulation. This technical article aims to contribute to this subject by reporting the applicability of the Data Storage System of the Hydrologic Engineering Center (HEC-DSSVue) using 5-min rainfall datasets collected from 10 self-recording rain gages (2014–2023) installed in Southern Brazil. Although the HEC-DSSVue is commonly used to integrate databases into other HEC models (HEC-HMS and HEC-RAS), its application in managing hydrological gaging networks remains undocumented. Our methodology comprised six steps conducted within the HEC-DSSVue: i) acquiring all files generated from each hydrological campaign, ii) addressing irregular-interval time series, iii) evaluating file management functionalities, iv) identifying data visualization functionalities, v) assessing mathematical operations, and vi) examining edition and export functions. The main conclusion was that the HEC-DSSVue is a powerful tool for efficiently managing datasets from hydrological gage networks and supporting data manipulation, thereby making routine tasks more efficient and less susceptible to user-introduced errors.

Resumo em Inglês:

ABSTRACT The influence of the airflow field around the conditioning roll of a mower conditioner on the forage harvesting process cannot be ignored. Full factorial simulation tests of the airflow field of a single conditioning roll and double conditioning roll were carried out, respectively, by taking the roll type, roll rotational velocity, and roll clearance as factors. The results show that the average airflow velocity around the conditioning rolls has a linear positive correlation with the roll velocity and an exponential negative correlation with the distance from the centre of the conditioning rolls in the single roll case. In the case of double rolls, the airflow around the conditioning rolls increase with the velocity of the rolls and decrease with the increase of the clearance between the rolls. The results of the field verification test show that, compared with the measured values for the prototype, the average error of the average velocity of the airflow around the single roll, predicted by the regression model, is 6.35%. The average error of the predicted velocity of the back-feeding airflow of the double roll is 5.86%. The results are reliable and could provide a reference for the optimised design of the conditioning roll.

Resumo em Inglês:

ABSTRACT The rotary buryig blade roller (RBBR) is a tillage tool that can both break up soil and bury straw. However, in previous experiments, it was found that the configuration of rotary blades on RBBR will significantly affect its tillage quality. To achieve the optimal structure of RBBR, this study optimized the configuration of rotary blades on RBBR. Additionally, the tillage performance of different RBBRs was evaluated using the Discrete Element Method (DEM). The results showed that after removing loose soil, the optimized RBBR (SR-8-1) achieved a flatter bottom of tilled layer compared to the previous RBBR (SR-6 and SR-9). Specifically, the calculated bottom evenness after tillage decreased by 40.49% and 40.65%, respectively. Furthermore, upon measuring tillage depth, stability coefficient of tillage depth increased by 5.12% and 3.84%. The exported average axial force from the software decreased by 65.72% and 80.39%. The comparison of field test and simulation showed that SR-8-1 was superior to SR-6 in all tillage parameters except power. Therefore, due to the advantages of SR-8-1 in terms of tillage performance, it is better able to meet the tillage needs than other structures.