Search

Background: Experiments were conducted to analyze environmental parameters and growth-related factors to identify the most influential factors in estimating pig’s body weight (PBW) using artificial neural networks (ANNs) and multiple linear regression (MLR)-based models. Methods: Four different ANNs, including Cascade Forward Back-propagation (CFBP), Layer Recurrent (LR), Feed Forward Back-propagation (FFBP) and Elman (EL), and MLR models were developed to estimate the body weight of pigs. The current research was conducted for 92 days during the two experimental periods (2021–2022) in a pig barn. Result: The Levenberg-Marquardt training function, gradient descent weight and bias learning function, tan-sigmoid transfer function, and two hidden layers with 16 neurons in each layer were shown to be the most effective architecture of the FFBP model in predicting PBW. According to the sensitivity analysis, length of pig (LP) was the most influential factor in estimating the PBW for MLR/ANN models. However, the environmental parameters along with growth-related factors could not always be the same association with PBW. Therefore, further research on viable alternative breeds with different management conditions may be considered to evaluate MLR and ANN model’s performance.

Forecasting solar irradiance is of utmost importance in supplying renewable energy efficiently and timely. This paper aims to experiment five variants of recurrent neural networks (RNN), and develop effective and reliable 5-minute short term solar irradiance prediction models. The 5 RNN classes are long-short term memory (LSTM), gated recurrent unit (GRU), Simple RNN, bidirectional LSTM (Bi-LSTM), and bidirectional GRU (Bi-GRU); the first 3 classes are unidirectional and the last two are bidirectional RNN models. The 26 months data under consideration, exhibits extremely volatile weather conditions in Jinju city, South Korea. Therefore, after different experimental processes, 5 hyper-parameters were selected for each model cautiously. In each model, different levels of depth and width were tested; moreover, a 9-fold cross validation was applied to distinguish them against high variability in the seasonal time-series dataset. Generally the deeper architectures of the aforementioned models had significant outcomes; meanwhile, the Bi-LSTM and Bi-GRU provided more accurate predictions as compared to the unidirectional ones. The Bi-GRU model provided the lowest RMSE and highest R2 values of 46.1 and 0.958; additionally, it required 5.25*10-5 seconds per trainable parameter per epoch, the lowest incurred computational cost among the mentioned models. All 5 models performed differently over the four seasons in the 9-fold cross validation test. On average, the bidirectional RNNs and the simple RNN model showed high robustness with less data and high temporal data variability; although, the stronger architectures of the bidirectional models, deems their results more reliable.

In this study, the main objectives are to show the performance of deep convolutional neural network in identifying individual pig and investigate the accuracy level of CNN using four datasets made with pig’s face in different growing period. Firstly, the datasets were captured in an experimental pig barn at a different time. Secondly, the datasets were filtered similar images using the structural similarity index measure (SSIM) for data preparation. Finally, face image classification is performed by employing a deep convolutional neural network (DCNN) namely ZFNet model. The results have shown that individual pig identification is outperformed while using the same age dataset in training and testing stage with an accuracy rate above 97%. The model performed better in a combined dataset which is a combination of all individual data. For future recommendation, it would be beneficial to perform the effectiveness on a large scale of pigs, and a network model should be considered unsupervised learning in case of ageing classification.

Fertilizer misapplications have induced widespread environmental deteriorations, climatic catastrophes, and economic losses; meanwhile, the Precision Agriculture (PA) endorsements have been influential in alleviating these issues. This study intended to tackle the fertilizer consumption inefficiencies by utilizing non-destructive remote sensing technologies, soil macronutrient distribution analysis, and a deep learning model. Specifically, an Unmanned Air Vehicle (UAV) was used in a cornfield to capture the plant's reflectance information for retrieving the Normalized Difference Vegetation Index (NDVI) during the vegetative and reproductive growth stages. Consequently, the field's soil samples were examined for their Nitrogen, Phosphorus, Potassium, and Carbon (NPKC) macronutrient constituencies. Finally, a Convolutional Neural Network-Regression model was developed to predict infield NPKC spatiotemporal variations in soil using the NDVI measurements. The deep learning model effectively determined the surpluses or shortages of the NPKC macronutrients within the cornfield throughout the growth stages. The model performed vigorously with R2 values of 0.93, 0.92, 0.98, and 0.83 in predicting N, P, K, and C levels in soil, respectively.

In previous studies, the stability and maturity of forced aerated composts were examined from the physicochemical analysis inside the reactors. The objective of this study is to check the maturity and effects of forced aerated composts after its addition to soil and plants. The composts were prepared from fresh chicken manure and sawdust in cylindrical reactors and aerated at 0.75, 0.50, and 0.25 L min−1 kg−1 organic matter (OM) for 1 month. Afterward, it was mixed with soil and coded as treatment T1 (0.75 L min−1 kg−1 OM aeration rate + soil), T2 (0.50 L min−1 kg−1 OM aeration rate + soil), T3 (0.25 L min−1 kg−1 OM aeration rate + soil), and CS (only soil), respectively. The soil properties (pH, electrical conductivity (EC), organic matter content (OM), loss of organic matter content, bulk density (BD), water holding(WH)) and plant growth (shoot length (cm), leaves’ area (cm2), fresh and dry weight (g)) were checked for 4 months. The results showed a significant increase (at P

An experiment was conducted to evaluate modelling relationships between pig’s body surface temperature and ambient environment including inside and outside of pig barn. For this purpose, four different artificial neural network (ANN), including Feed Forward Back-propagation (FFB), Layer recurrent (LR), Elman (EL) and Cascade Forward Back-propagation (CFB) with different learning algorithms, transfer functions, hidden layers and neuron in each layer, and multi-linear regression (MLR) models have been performed to predict body temperature of pig. Six two-month-old pigs were studied over a period of 92 days during two years (2017–2018) to develop and evaluate the ANN and MLR models. The performance of the models in predicting pig’s body temperature was determined using statistical quality parameters, including coefficient of determination (R2), root mean square error (RMSE) and mean absolute percentage error (MAPE). The FFB model with the Levenberg-Marquardt training function, Gradient descent weight and bias learning function, Log-sigmoid transfer function and two hidden layers with 20 neurons was found as the best model. Sensitivity analysis indicated that the temperature-humidity index (THI) inside the room is the most influential factor in predicting pig’s body temperature in the MLR/ANN models.

In this study, the possibility of enhancing chilling stress tolerance of pepper plant (Capsicum annuum Linnaeus) during early growth stages was investigated using DME combustion gas in controlled greenhouse. The experiment was conducted to determine the performance of DME combustion gas when used as a fuel for DME burner for raising temperature and CO2 concentration in greenhouse and also to examine its effects on morphology parameters of hot pepper in winter season. To analyze the relationship between the application of DME and morphology parameters of pepper plant, three treatments (DME-1, DME-2 and DME-3) for three controlled greenhouses were assayed. DME-1 and DME-2 treatments consisted of average DME flow quantity in duct were 17.4 m3/min and 10.2 m3/min, respectively to greenhouse-1 and greenhouse-2 and no DME gas was supplied to greenhouse-3 which was left as control (DME-3). Morphology parameters such as plant height, leaf area and leaf area index (LAI), net assimilation rate (NAR), relative growth rate (RGR), fresh weight and dry weight were measured for eight weeks for each treatment and analyzed using completely randomized designs through analysis of variance with a significance level of P < 0.05. Although DME-1, DME-2 and DME-3 received same crop management practices and controlled environmental factors, the highest changes (p < 0.05) of plant height, leaf area, LAI and fresh weight were found from the DME-1 treatment, followed by DME-2. A comparison of relative growth rates among the treatments indicated more rapid relative growth rate of morphology parameters at vegetative phase of plant implying better yield. Therefore endorsed quantify of DME combustion gas for a specified crop can be applied to greenhouse to improve the plant growth and enhance yield with mitigating chilling stress in winter season.

The experiment was conducted to determine the performance of DME combustion gas when used as a fuel for DME burner for raising temperature and CO2 concentration in greenhouse and also to examine its effects on chlorophyll content, and fresh and dry weight of lettuce and Chinese cabbage. DME-1 and DME-2 treatments consisted of average DME flow quantity in duct were 17.4 m3 min-1 and 10.2 m3 min-1 respectively to greenhouse-1 and greenhouse- 2 and no DME gas was supplied to greenhouse-3 which was left as control (DME-3). DME supply times were 0.5 hr day-1, 1 hr day-1, 1:30 hrs day-1 and 2 hrs day-1 on week 1, 2, 3, and 4 respectively. Chlorophyll content and fresh and dry weight of lettuce and Chinese cabbage were measured for each treatment and analyzed through analysis of variance with a significance level of P

Regression analysis to predict growth indices of plant is essential for understanding the relationship between the total leaf area, production of fresh weight and dry matter, and expansion of the plant growth. An experiment was conducted to develop regression models for estimating leaf area, fresh weight, and dry weight from measurements of plant height at the vegetative phase of hot pepper (Capsicum annuum Linnaeus) grown in biodegradable pots in a greenhouse. Five models were evaluated and compared: linear regression model, two-order polynomial regression model (P. order 2), three-order polynomial regression model (P. order 3), four-order polynomial regression model (P. order 4), and power regression model. The models were compared using the coefficient of determination (R2), Pearson’s correlation coefficient (r), root mean square error (RMSE), relative standard error (RSE), and mean absolute percentage error (MAPE). Power regression involving plant height demonstrated the highest R-square among the other models with minimum error estimate for the expected leaf area (R2 > 0.96, r > 0.98, RMSE 0.98, r > 0.99, RMSE 0.97, r > 0.98, RMSE

During tillage operations, draft force contributes hugely to the operational cost. Accurate knowledge of draft force (DF) requirement for tillage implements is a useful tool for machinery selection, matching of implements to tractor, etc. The goal of this research was to measure the draft force of a compact disc harrow and analyze the effect of soil moisture content (SMC) and machine speed on draft force. Also, another aim is to develop a model equation to predict draft force requirement of compact disc harrow. A EWX 10000F winch was used to provide the draft (pull) force which was powered by a 24 V DC battery. The draft force was measured using a load cell which transmitted the measured values to an AND ad 4532B indicator. The draft force of the compact disc harrow was measured at three different levels of SMC: low SMC (5–7%, 9–11%, 13–15%); medium SMC (25–27%, 29–31%, 33–35%); and High SMC (40–42%, 44–46%, 48–50%) with machine speed (MS) of 0.06 m/s, 0.09 m/s, and 0.12 m/s. The results obtained were statistically analyzed and it showed that the lowest draft force occurred at 26% SMC with a force of about 658 N. A multi-variable regression equation was developed after the regression analysis. The regression equation DF = 466.91 + 6.25*(SMC) + 2952.1*(MS) was selected as the best model for predicting draft force. This equation was verified using the Bland–Altman plot, the t test analysis, and the standard error analysis. The results of the verification methods proved that the suggested model can be used for predicting draft force requirement of a compact disc harrow.

Indoor air temperature (IAT) and indoor relative humidity (IRH) are the prominent microclimatic variables, still, potential contributors that influence the homeostasis of livestock animals reared in closed barns. Further, predicting IAT and IRH encourages farmers to think ahead actively and to prepare the optimum solutions. Therefore, the primary objective of the current literature is to build and investigate extensive performance analysis between popular ML models in practice used for IAT and IRH predictions. Meanwhile, multiple linear regression (MLR), multilayered perceptron (MLP), random forest regression (RFR), decision tree regression (DTR), and support vector regression (SVR) models were utilized for the prediction. This study used accessible factors such as external environmental data to simulate the models. In addition, three different input datasets named S1, S2, and S3 were used to assess the models. From the results, RFR models performed better results in both IAT (R2 = 0.9913, RMSE = 0.476, MAE = 0.3535) and IRH (R2 = 0.9594, RMSE = 2.429, MAE = 1.47) prediction among other models particularly with S3 input datasets. In addition, it has been proven that selecting the right features from the given input data builds supportive conditions under which the expected results are available. Overall, the current study demonstrates a better model among other models to predict IAT and IRH of a naturally ventilated swine building containing animals with fewer input attributes.

Linear partial least square and non-linear support vector machine regression analysis with various preprocessing techniques and their combinations were used to determine the soluble solids content of hardy kiwi fruits by a handheld, portable near-infrared spectroscopy. Fruits of four species, namely Autumn sense (A), Chungsan (C), Daesung (D), and Green ball (Gb) were collected from five different areas of Gwangyang (G), Muju (M), Suwon (S), Wonju (Q), and Yeongwol (Y) in South Korea. The dataset for calibration and prediction was prepared based on each area, species, and in combination. Half of the dataset of each area, species, and combined dataset was used as calibrated data and the rest was used for model validation. The best prediction correlation coefficient ranges between 0.67 and 0.75, 0.61 and 0.77, and 0.68 for the area, species, combined dataset, respectively using partial least square regression (PLSR) method with different preprocessing techniques. On the other hand, the best correlation coefficient of predictions using the support vector machine regression (SVM-R) algorithm was 0.68 and 0.80, 0.62 and 0.79, and 0.74 for the area, species, and combined dataset, respectively. In most cases, the SVM-R algorithm produced better results with Autoscale preprocessing except G area and species Gb, whereas the PLS algorithm shows a significant difference in calibration and prediction models for different preprocessing techniques. Therefore, the SVM-R method was superior to the PLSR method in predicting soluble solids content of hardy kiwi fruits and non-linear models may be a better alternative to monitor soluble solids content of fruits. The finding of this research can be used as a reference for the prediction of hardy kiwi fruits soluble solids content as well as harvesting time with better prediction models.

An experiment was conducted to find out the most influential factors affecting pig’s body temperature (PBT). For this purpose, eight environmental parameters and three growth related factors were considered as variables. Among these factors, seven environmental parameters, including temperature, CO2, temperature-humidity index inside and outside the pig’s barn and relative humidity inside the barn were taken as input variables for artificial neural networks (ANN) and multiple linear regression (MLR) models due to their good correlation (r ³ 0.5) with PBT. The results showed that ANN and MLR models had the lowest R2 values (0.81 and 0.69, respectively) and the highest RMSE (1.17 and 1.48, respectively) when they were run without temperature-humidity index; however, the maximum R2 (0.90 and 0.75, respectively) and minimum RMSE (0.92 and 1.40, respectively) were found without relative humidity. Based on the results, the temperature-humidity index could represent an important indicator in registering early warning signs of PBT status alternations.

An experiment was conducted to evaluate the performance of temperature model (T model), relative humidity model (H model), temperature-humidity model (TH model), and temperature-humidity index model (THI model) in predicting pig’s body surface temperature (PBT). Infrared Sensor (IR) was used to measure PBT at different locations: left side (LS), right side (RS), forehead (FH) and back side (BS). Ambient environmental parameters inside the room such as temperature (ART), relative humidity (RRH) and CO2 concentration were measured using livestock environment management system (LEMS). THI model was selected as the best model in making more accurate prediction in both training (R2=0.72, RMSE=0.80, RSE=0.26 and MAPE=2.08) and validation (R2=0.74, RMSE=1.10, RSE=0.40 and MAPE=2.80) stages. For more precise modeling, apart from temperature and humidity data other environmental factors inside pig’s barn (CO2 concentration, wind speed, air pressure etc.) as well as growth factors (body weight, feed intake etc.) may be included in models.

Solar renewable energy (SRE) applications are substantial in eradicating the rising global energy shortages and reversing the approaching environmental apocalypse. Hence, effective solar irradiance forecasting models are crucial in utilizing SRE efficiently. This paper introduces a partially amended hybrid model (PAHM) by the implementation of a new algorithm. The algorithm innovatively utilizes bi-directional gated unit (Bi-GRU), autoregressive integrated moving average (ARIMA) and naive decomposition models to predict solar irradiance in 5-min and 60-min intervals. Meanwhile, the models’ generalizability strengths would be tested under an 11-fold cross-validation and are further classified according to their computational costs. The dataset consists of 32 months’ solar irradiance and weather conditions records. A fundamental result of this study was that the single models (Bi-GRU and ARIMA) outperformed the hybrid models (PAHM, classical hybrid model) in the 5-min predictions, negating the assumptions that hybrid models oust single models in every time interval. PAHM provided the highest accuracy level in the 60-min predictions and improved the accuracy levels of the classical hybrid model by 5%, on average. The single models were rigorous under the 11-fold cross-validation, performing well with different datasets; although the computational efficiency of the Bi-GRU model was, by far, the best among the models.

This study elucidated and attempted to design and development of arsenic and iron removal unit for drinking water. The process of reducing iron and arsenic from water has been experimented by developing a unit in the Department of Farm Structure at Bangladesh Agriculture University, Mymensingh. In order to test the performance of the developed unit, arsenic contaminated water with four different concentrations like 0.05, 0.10, 0.15 and 0.20 mg/l were prepared in the laboratory and passed into the developed unit and the respective output concentrations were found to be 0.00, 0.01, 0.025 and 0.05 mg/l. The input and output concentrations of arsenic were tested in the chemical testing laboratory under the Bangladesh Institute of Nuclear Agriculture, Mymensingh. Iron contaminated water were collected from four selected tube wells of local Mymensingh and were also passed into the developed unit with four input concentrations like 0.18, 0.1532, 0.179 and 0.133 mg/l and the respective output concentrations were found to be 0.10, 0.1021, 0.11 and 0.09 mg/l. The concentrations of Iron were tested in the chemical testing laboratory under the Soil Resource Development Institute, Dhaka. The results have revealed that iron and arsenic content brings to allowable limit. The developed unit has the capacity to remove Arsenic and Iron and help to eradicates hazardous problem of people.

This simulation study aimed to predict the effects of potential change of temperature and carbon dioxide on winter rice (Boro rice) production in Bangladesh. We assessed the effects of temperature and carbon dioxide on yield of Boro rice using DSSAT model for six major rice growing Agro- Ecological Zones (AEZs). The sensitivity analysis included variation of maximum and minimum temperature at 2 0 C and 4 0 C and different levels of CO 2 concentrations (50 ppm, 100 ppm and 200 ppm) above the value reported in 2005 of IPCC Fourth Assessment Report (379 ppm). The model predicted that the change of maximum temperature drastically reduced rice yield at all selected locations. The average value of yield reductions (average percentage change of rice yield for 6 locations) of Boro rice were 6% and 16% for increasing maximum temperature at 2 0 C and 4 0 C respectively. Minimum temperature had also negative impacts on Boro rice yields that reduced yield by 4% for 2 0 C and 8% for 4 0 C. However, a slight increase in rice yield due to increase in CO 2 concentration could not overcome the negative effect of temperature. Such yield reductions under changing climatic conditions could significantly affect the rice production in Bangladesh in future. DOI: http://dx.doi.org/10.3329/brj.v17i1-2.20897 Bangladesh Rice j. 2013, 17(1&2): 15-25

Due to an unfortunate oversight, Table 8 has been shown erroneously. The authors apologize for this mistake.

Impact of climate change on development phases of rice is one of the major concern issues due to yield reduction. In the present study, development phases of winter rice (boro rice) of BR3 and BR14 varieties for the years 2008, 2030, 2050 and 2070 have been simulated for 12 major rice growing Agro-Ecological Zones (AEZs) of Bangladesh using DSSAT model. Available data on soil and hydrologic characteristics of these locations, and typical crop management practices for boro rice were used. The weather data required for the model were generated for the selected years and for the selected locations using the regional climate model PRECIS. Compared to 2008, the model predicted average reductions in development phases in days of BR3 variety for the selected 12 locations were 1 in Emergence (E) to End Juvenile (EJ) phase, 1 in Panicle Inition (PI) to End Leaf Growth (ELG) and 1 in Grain Filling Phase (GFP) phase during 2030; 4 in E to EJ, 1 in EJ to PI, 3 in PI to ELG and 2 GFP phase for the year 2050 and 7 in E to EJ, 3 in EJ to PI, 5 in PI to ELG and 3 GFP in 2070. Similar results were also found for BR14 variety for the selected locations in Bangladesh. Therefore, developing new rice varieties with considering climatic adaptive, proper growing period should be maintained.