Research Article
Modeling
S. Sharifi; N. Hafezi; M. H. Aghkhani
Abstract
IntroductionEfficient use of energy in paddy production can lower greenhouse gas emissions, safeguard agricultural ecosystems, and promote the growth of sustainable agriculture. Meanwhile, intelligent agriculture has come to the aid of farmers and policy-makers by harnessing cutting-edge technologies, ...
Read More
IntroductionEfficient use of energy in paddy production can lower greenhouse gas emissions, safeguard agricultural ecosystems, and promote the growth of sustainable agriculture. Meanwhile, intelligent agriculture has come to the aid of farmers and policy-makers by harnessing cutting-edge technologies, which will lead to sustainable welfare and the comfort of human society in the present and the future. Therefore, this study aimed to analyze energy consumption and production, as well as model and optimize the yield of two paddy cultivars using Artificial Bee Colony (ABC) and Genetic Algorithms (GA).Materials and MethodsExtensive research data was collected by thoroughly examining documentary and library resources, as well as conducting face-to-face questionnaires with 120 paddy farmers and farm owners in Rezvanshahr city, located in the province of Guilan, Iran, during the 2019-2020 production year. The farms consisted of 80 high-grading and 40 high-yielding paddies. The independent variables were machinery, diesel and gasoline fuels, electricity, seed, compost and straw, biocides, fertilizers, and labor. The dependent variable was paddy yield per hectare [of the farm area]. In the first step, energy consumption and production were calculated by multiplying the variables by their corresponding coefficients. In the second step, all the variables that maximize paddy yield were entered into MATLAB software. An artificial bee colony (ABC) algorithm with a novel and straightforward elitism structure was utilized to enhance the fitness function of the genetic algorithm (GA). The Sphere, Repmat, and Unifrnd functions were employed to determine the objective function, define the position of the bee colony, and quantify the position of the bee colony, respectively. In each generation, 900 new solutions were created, and the algorithm iterated 200 times. For the genetic algorithm, the population was defined as a double vector with a size of 100.Results and DiscussionThe findings revealed that the Hashemi (high-grading) paddy cultivar had an average energy consumption and production of 55.973 and 30.742 GJ·ha-1, respectively. The Jamshidi (high-yielding) paddy cultivar had an average energy consumption of 54.796 GJ·ha-1 and double the energy production of the Hashemi at 62.522 GJ·ha-1. In both cultivars, agricultural machinery consumed the highest amount of energy, while straw consumed the lowest amount. The average energy consumption of tractors in the Hashemi and Jamshidi cultivars was 25.111 and 25.865 GJ·ha-1, respectively, accounting for 44.862% and 47.202% of the total average consumed energy. This undoubtedly demonstrates the significant effect of this input and reflects the operators' skill and experiential knowledge. The evaluation indexes, including R², RMSE, MAPE, and EF, as well as statistical comparisons such as mean, STD, and distribution, consistently demonstrated that the ABC algorithm provided the essential conditions for the fitness function. The results of the bee-genetic algorithm optimization revealed that the majority of the consumed resources could be effectively managed on the farm to closely match optimal conditions. Through this optimization, energy consumption in the Hashemi and Jamshidi cultivars was reduced by 53.96% and 39.41%, respectively.ConclusionGiven its impressive performance and potential for minimizing energy consumption, the ABC-GA algorithm offers an opportunity for policymakers in energy resource management and rice industry managers to develop innovative strategies for significantly reducing energy usage in rice production. This approach could lead to more sustainable and efficient practices in the agricultural sector.
Research Article
Design and Construction
M. Teimorzadeh; J. Baradaran Motie; A. Rohani; Y. Selahvarzi
Abstract
IntroductionNeglecting the water requirements of trees can result in inefficient irrigation practices, leading to either water wastage or drought stress. Effective irrigation management necessitates precise information on the quantity and pattern of water consumption by trees. To achieve optimal irrigation, ...
Read More
IntroductionNeglecting the water requirements of trees can result in inefficient irrigation practices, leading to either water wastage or drought stress. Effective irrigation management necessitates precise information on the quantity and pattern of water consumption by trees. To achieve optimal irrigation, a reliable method for quantifying plant water needs is crucial, ensuring that trees avoid drought stress. Current methods for assessing tree water requirements often focus on specific components, such as stems or leaves. These techniques typically require manual intervention, which is time-consuming and resource-intensive, thereby restricting their application mainly to research environments.Materials and MethodsA sap-flow meter device was developed to generate a heat pulse in a tree trunk at 15-minute intervals. The device comprises measuring probes, a processing unit, and a data logger. For a comprehensive evaluation, device probes were positioned on the trunk of a Ficus benjamina tree within a controlled environment at two distinct heights. The resulting sap flow through the vascular tissue was then compared to data obtained using the lysimetric method. The Ficus benjamina tree, with a trunk diameter of 3.5 cm and a height of 196 cm, was prepared during the summer of 2022. By measuring the rate of heat pulse dissipation and applying heat transfer principles, sap flow is estimated under the assumption that heat transfer occurs primarily through the sap flow within the vascular tissue. This estimation was achieved using the heat ratio method (HRM).The trunk was triple drilled with holes of 1.5 mm in diameter and 25 mm in depth. Following drilling, the probes were inserted into these holes (Figure 1). To prevent heat transfer from the probes to the surrounding environment, the trunk was wrapped with glass wool insulation. To assess the reliability of the device, the lysimetric method was employed to measure tree transpiration. For this purpose, the soil surface of the pot was covered with cellophane to ensure that evaporation and weight loss of the pot occurred exclusively through the tree's leaves. Hourly measurements of the pot's weight were taken using a digital scale. Changes in the pot's weight indicate the amount of water evaporated, which corresponds to the water transpired by the tree through its vascular tissue.Results and DiscussionThe results showed that the sap-flow meter device slightly overestimates the tree's water consumption compared to the values obtained using the lysimetric method. Sap flow and transpiration follow a similar trend, both escalating throughout the day and reaching their highest levels in the early afternoon. This value reached 17.98 ml h-1 for sap flow and 16 ml h-1 for transpiration (by lysimetric method), followed by a rapid decrease in the late afternoon as the air cooled down. In addition, the results of device measurements showed that spraying water on the leaves lowers both the rate and volume of sap flow. When the canopy becomes wet, the evaporation of water from the leaf surface leads to a drop in the temperature, which in turn significantly slows down the flow of sap.The v1/v2 ratio is not constant over time, making it crucial to choose the right starting point for measurements to ensure effective data acquisition during the device's operational cycle. It is essential to measure (by the device) the difference between temperature probes 40 seconds after heat pulse generation. The sap flow and transpiration followed a similar trend during the experiments. The sap flow and transpiration increased throughout the day, peaking in the early afternoon. On the first day, sap flow reached 17.98 ml h-1, while the second day recorded an even higher rate of 19.75 ml h-1. Correspondingly, the transpiration measured using the lysimetric method peaked at 16 ml h-1, followed by a rapid decline in the late afternoon.ConclusionThe results obtained from the developed device indicate several key findings. Sap flow and transpiration exhibit a similar trend during the test period, with the estimated sap flow value being approximately 30% higher than that obtained using the lysimetric method. The device effectively demonstrated the impact of surface irrigation; spray irrigation influences the sap flow rate such that when the canopy becomes wet, the sap flow rate decreases significantly. Additionally, sap flow and transpiration are positively correlated with air and canopy temperatures, and negatively correlated with relative humidity. Following calibration, the results show that the heat pulse method can accurately and effectively measure sap flow in the vascular tissue of trees.
Research Article
Design and Construction
E. Vahedi Tekmehdash; H. Navid; H. Ghasemzadeh; H. Karimi; M. Javani Holan
Abstract
IntroductionThe livestock sector excels in the production of dairy and meat products. These products, serving as vital sources of animal protein, hold a significant position in household diets. The significance of these two products in the food basket has heightened awareness around animal health. Regularly ...
Read More
IntroductionThe livestock sector excels in the production of dairy and meat products. These products, serving as vital sources of animal protein, hold a significant position in household diets. The significance of these two products in the food basket has heightened awareness around animal health. Regularly tracking rumination time serves as a vital and insightful measure to obtain information about the rest and overall health of an animal. This information enables prompt intervention for health or nutritional issues, allowing for earlier management adjustments and veterinary care to effectively combat the onset of disease. In the past, rumination was usually monitored through visual observation by on-site staff or through videos recorded by cameras installed on the livestock. Nowadays, the growing scale of livestock farms makes it impractical to effectively monitor the animals individually. The traditional method of visual observation demands the continuous presence of livestock professionals and is extremely time-consuming. Currently, sensors and digital technologies have become important tools for accurate animal husbandry, enabling real-time monitoring of rumination. A review of the research in the field of precision animal husbandry shows that many efforts are being made to develop precision monitoring sensors to overcome the mentioned problems. Continuous and automatic monitoring of animal behavior through sensors can offer valuable insights into nutrition, reproduction, health, and overall well-being of dairy cows.Materials and MethodsIn this research, an accelerometer-based sensor was developed and used in the precision agriculture laboratory of Tabriz University, Iran. The sensor was installed in three different positions on the cow's body to collect data. Important factors were calculated from the raw data, and the modeling was done using the logistic regression method. The logistic regression model was trained to distinguish rumination from the other cow's behaviors. The developed model was evaluated using the receiver operating characteristic (ROC) curve, and three other evaluation criteria: precision, sensitivity, and F-score. Finally, the performance of the final model and sensor was evaluated in the field for a few days.Results and DiscussionAfter calculating the evaluation criteria for different calculation factors, four optimal factors were finally selected from the 50 arrays. Muzzle mode was found to be the best place to install the sensor. Logistic regression was the best modeling method for binary classification between rumination and other behaviors. The evaluation criteria of the model in the proposed sensor are the highest, and the values of sensitivity 88%, accuracy 94%, and F-score 91% were obtained through logistic regression analysis. The final test results of the model revealed that the sensor demonstrated an impressive detection capability of 89.47%. Furthermore, the developed system exhibited strong alignment with the actual field observations, highlighting its effectiveness and reliability. Finally, the results of the current study were compared with other studies in the literature.ConclusionThis study investigated recording and monitoring rumination behavior using an accelerometer, which can help prevent financial losses in cattle farms. After examining different mounting locations of the sensor, it was found that the muzzle position provided more accurate detections than the other mounting locations. The final model was created using the statistical factors and the calculation of the evaluation criteria. The results showed that the proposed model provided more correct diagnoses and achieved the optimal solution.AcknowledgmentWe would like to express our gratitude to the Khalat Poushan Cattle Farming Complex of the University of Tabriz, Iran, its professors and staff for supporting this project, and for their commitment to promote animal husbandry science.
Research Article
Design and Construction
N. Loveimi; A. Azizi; A. Kaab; A. Neisi
Abstract
IntroductionSubsoiling is a critical tillage operation for many crops, particularly sugarcane, due to the impact of agricultural machinery traffic and its significance in managing heavy-textured and compacted soils. Given the extensive size of sugarcane fields and the time-intensive nature of subsoiling ...
Read More
IntroductionSubsoiling is a critical tillage operation for many crops, particularly sugarcane, due to the impact of agricultural machinery traffic and its significance in managing heavy-textured and compacted soils. Given the extensive size of sugarcane fields and the time-intensive nature of subsoiling operations, the application of intelligent control techniques for monitoring and managing these processes is of considerable importance. Currently, subsoiling operations are monitored using manual gauges. This approach involves collecting a limited number of samples per hectare, typically after the operation is completed, which makes it nearly impossible to implement real-time corrections. To address this limitation, the development and implementation of a depth measurement system offer a promising solution. Such a system enables real-time observation of working depth by both the operator, via an on-screen display, and by a remote observer through an online platform. This capability allows for immediate adjustments during the operation, ensuring greater precision and efficiency. Furthermore, by integrating recorded depth data with geospatial information, it becomes possible to generate detailed maps illustrating depth variations across the field. These maps can serve as valuable tools for further evaluations, such as performance monitoring in areas where subsoiling depth deviates from the desired range, either being too shallow or excessively deep. This technological advancement has the potential to significantly enhance the accuracy and effectiveness of subsoiling operations in modern agricultural practices.Materials and MethodsThis study focused on the design, development, and evaluation of a depth measurement system for a subsoiler attached to a track-type tractor, specifically tailored for sugarcane fields. The system not only provided real-time depth display but also recorded the location and transmitted it online. The research employed three distinct depth measurement techniques and was conducted using a randomized complete block design with split plots. The main plots are the three depth measurement techniques: based on the angles of the driving profiles of the subsoiler shanks (T1), the laser distance measurement method (T2), and the ultrasonic distance measurement method (T3), and sub-plots are depth ranges at three levels: 0-30 cm (R1: surface range), 30-60 cm (R2: mid-range), and 60-90 cm (R3: deep range). Initially, we calculated the absolute difference between the depths recorded by the system and those measured manually with a rod at each location. Following this, we analyzed key statistical indicators, including the average, standard deviation, and the minimum and maximum of errors, for comparison.Results and DiscussionThe results showed that the depth measurement error was significantly influenced by the technique employed. The angle technique yielded the lowest average error of 1.91 cm, while the ultrasonic technique resulted in the highest average error of 3.83 cm. Across all depth ranges, statistical indicators for depth error were significant. Specifically, within these ranges, the deep range exhibited an average depth error of 2.33 cm, and the surface range had an average error of 3.65 cm. Statistical analysis revealed that only indices related to minimum and maximum errors for interactions between factors were significant. The lowest minimum error value (0.05 cm) was observed with the angle technique at deeper depths, whereas the highest minimum error (0.34 cm) occurred with ultrasonic measurements at shallower depths on surfaces. Similarly, maximum errors followed this trend: The lowest maximum error (3.21 cm) was associated with angle measurements at deeper depths, while ultrasonic measurements on surfaces yielded a higher maximum error (8.63 cm). Both laser and ultrasonic techniques consistently demonstrated greater errors across all three depth ranges compared to angle-based methods. This discrepancy may be attributed to inaccuracies inherent in rangefinders when their beams encounter obstacles like clods or pits during field operations. Notably, as working depths increased across all measurement techniques, errors in depth measurement decreased significantly due to reduced vibrations from subsoiler devices at greater depths, thereby minimizing vibration-related inaccuracies.ConclusionThe results indicate that the depth measurement technique based on the angles of the driving profiles of subsoiler shanks exhibits superior accuracy in determining the working depth of subsoilers mounted on tractors, particularly during sugarcane field operations. The laser distance meter technique ranked second in terms of accuracy, while the ultrasonic distance meter method demonstrated the least precision. Notably, as working depths increased, reduced vibrations during operation were observed, leading to enhanced accuracy in depth calculations across all techniques. This improvement is attributed to decreased mechanical disturbances at greater depths. Overall, measurements within deeper ranges achieved higher levels of accuracy compared to those at shallower surface ranges. This trend suggests that operational conditions and device stability play significant roles in optimizing measurement accuracy.
Research Article
Agricultural systems engineering (greenhouse, fish farming, mushroom production)
M. Zangeneh; E. Godini
Abstract
IntroductionIn recent years, the lack of adequate regional assessment and classification has led to unequal investments and policies, resulting in polarization and disparities in the development of agricultural units. However, since agricultural products are produced, distributed, and consumed nationwide, ...
Read More
IntroductionIn recent years, the lack of adequate regional assessment and classification has led to unequal investments and policies, resulting in polarization and disparities in the development of agricultural units. However, since agricultural products are produced, distributed, and consumed nationwide, analyzing production conditions across different agricultural systems can provide valuable insights for supply chain managers. A comprehensive evaluation of production system development across the country can enhance product quality, stabilize the supply chain, reduce costs, and improve overall efficiency and performance. These improvements are vital for advancing the agricultural sector and strengthening national competitiveness. In many regions, agriculture forms the backbone of the local economy, making regional equity and balanced development essential for sustainable agricultural growth.Materials and MethodsThis research was conducted with the aim of evaluating the development levels of different provinces of the country in the field of edible mushroom cultivation. The approach of this research is descriptive-analytical. The statistical population includes 31 provinces of Iran, and the required data are based on the results of the 2016 and 2021 censuses of the Statistical Center of Iran. Following the initial review, indicators that emphasize the aspects of human power, infrastructure, performance, waste, economy, and energy were collected. Weight estimation of indicators was done using Shannon's entropy method. The TOPSIS method was used to assess and rank the provinces based on their level of development within the mushroom cultivation industry. The ranking operation was done using eight different index groups: infrastructure, consumption of inputs, value of consumption inputs, types of products and waste, value of payment types, value of product categories, value of energy consumption, and the number of employees and payments to them.Results and DiscussionResults show that in 2016, the provinces were classified into three levels: relatively deprived of development, medium development, and relatively developed. Apart from Alborz province, which was placed at a relatively developed level, other provinces were placed at lower levels. By 2021, all provinces had made significant progress compared to 2016, elevating their development status so that none were classified as relatively deprived. Furthermore, the number of provinces categorized as relatively privileged surged from just one in 2016 to eight by 2021. The findings revealed that the smaller, non-industrialized provinces exhibited greater development compared to their larger, industrialized counterparts.ConclusionThe results showed that Alborz province had the highest level of development, and Semnan province had the lowest level of development of this industry in the country. The level of development and ranking of edible mushroom cultivation units in the provinces was obtained by using different categories of indicators and the TOPSIS multi-criteria decision-making method. To enhance the production and productivity of cultivated edible mushrooms, it is essential to advance cultivation techniques and technologies through the expansion of research initiatives, educational programs, and extension activities.
Research Article
Bioenergy
S. R. Mousavi Seyedi; M. Askari; S. M. R. Miri
Abstract
IntroductionIn Asia, two-wheeled agricultural tractors predominantly use single-cylinder two-stroke diesel engines, which are characterized by high fuel consumption and substantial air pollution. At the same time, the severe environmental impacts of energy production from diminishing fossil fuel reserves ...
Read More
IntroductionIn Asia, two-wheeled agricultural tractors predominantly use single-cylinder two-stroke diesel engines, which are characterized by high fuel consumption and substantial air pollution. At the same time, the severe environmental impacts of energy production from diminishing fossil fuel reserves are increasingly evident. Therefore, it is essential to develop sustainable and clean energy sources to meet these needs. Biodiesel is an alternative fuel that can be blended with conventional diesel to help reduce environmental pollution. In this study, diesel-biodiesel blends produced from rapeseed, soybean, and palm oil were evaluated for their effects on engine performance metrics, including power (P), torque (T), and specific fuel consumption (SFC). Furthermore, the emissions of pollutants (NOx, HC, CO, and CO₂) from these fuels were measured and modeled using linear and non-linear regression, as well as the adaptive neuro-fuzzy inference system (ANFIS).Materials and MethodsTo leverage the benefits of palm oil biodiesel, known for its high calorific value, along with the low kinematic viscosity of biodiesel derived from soybean and rapeseed oils, pure diesel was blended with 10% and 20% mixtures of rapeseed, soybean, and palm biodiesel, as well as 10% and 20% combinations of all three biodiesels. These nine fuel blends were tested at four engine speeds (1800, 2100, 2400, and 2700 rpm) under full load conditions. The diesel-biodiesel blends were produced at Sari Agricultural Sciences and Natural Resources University (SANRU) and transported to the engine laboratory at Tarbiat Modares University in Tehran, Iran, for detailed analysis. A total of 36 treatments were evaluated using a randomized complete block design (RCBD), incorporating four engine speeds and nine fuel types. The measured outputs included engine power, torque, specific fuel consumption, and pollutant emissions such as NOx, HC, CO, and CO₂. The collected data were used as input for modeling through both linear and non-linear regression in SPSS software, as well as ANFIS in MATLAB software.Results and DiscussionThis study evaluated nine diesel-biodiesel blends derived from palm, rapeseed, and soybean oils using a diesel engine in a controlled laboratory setting. Tests were carried out at four engine speeds—1800, 2100, 2400, and 2700 rpm—under full load conditions to assess engine performance and exhaust emissions. The results showed that for all tested fuel blends, power, specific fuel consumption, and pollutant emissions increased with engine speed, while torque decreased. Based on the findings, a blend containing 20% palm biodiesel can be used as an alternative fuel in diesel engines without requiring any modifications. The modeling results indicated that non-linear regression provided better accuracy than linear regression. However, ANFIS demonstrated a much higher correlation between actual and predicted values, with R² exceeding 0.98 for both performance parameters and emissions, compared to R² values below 0.47 for linear regression and below 0.92 for non-linear regression. The ANFIS model achieved its highest and lowest R² values at 0.99 for specific fuel consumption (SFC) and 0.98 for power (P), respectively; substantially higher than those from linear regression, which yielded 0.47 for torque (T) and 0.00 for power. Non-linear regression resulted in R² values of 0.92 for SFC and 0.60 for carbon monoxide (CO), still lower than those achieved by ANFIS. Overall, the highest R² value recorded was 0.7525 for torque, and the lowest was 0.6112 for power.ConclusionSingle-cylinder diesel engines, which have high fuel consumption and contribute to air pollution, are commonly used in two-wheel agricultural tractors across Asia. One approach to reducing the environmental impact of fossil fuels is to use biodiesel in these engines without requiring any modifications. The results of this study showed that a 20% blend of palm biodiesel can be an effective alternative fuel for diesel engines, with no need for engine modification. Furthermore, the modeling results indicated a significantly higher correlation (R² > 0.98) between actual and predicted values of performance variables and emissions using ANFIS, compared to linear regression (R² < 0.47) and non-linear regression (R² < 0.92). Therefore, ANFIS can be effectively used to accurately predict engine performance and emission parameters.
Research Article
Modeling
M. Boroun; M. Ghahderijani; A. A. Naseri; B. Beheshti
Abstract
IntroductionEnergy analysis offers significant benefits by establishing a foundation for resource conservation, quantifying the energy consumed at each stage of production, identifying processes that require minimal energy input, and supporting sustainable management practices. In sustainable agricultural ...
Read More
IntroductionEnergy analysis offers significant benefits by establishing a foundation for resource conservation, quantifying the energy consumed at each stage of production, identifying processes that require minimal energy input, and supporting sustainable management practices. In sustainable agricultural systems, maximizing the productivity of input energies is a key objective. This study aims to assess energy consumption patterns within the sugar industry and to compare the optimization of energy consumption indicators using two meta-heuristic algorithms, ultimately seeking to enhance resource efficiency and promote sustainable production methods.Materials and MethodsThis study evaluated energy efficiency and environmental impacts in sugarcane-based sugar production at Dehkhoda Sugarcane Agro-Industry Company (in Khuzestan Province, Iran), during the 2019-2020 agricultural cycle. Data collection integrated field questionnaires, expert interviews, operational records from the facility, and national agricultural databases (Ministry of Agriculture Jihad statistics and energy balance sheet). Energy flow were analyzed using MATLAB statistical software and the Equinonet database, with comparative optimization through genetic algorithms and imperialist competitive algorithms to identify efficiency improvements.Results and DiscussionThe results showed that, for the majority of indicators evaluated, the imperialist competitive algorithm outperformed the genetic algorithm in optimizing energy consumption. In addition to reducing the environmental impacts of this profitable industry in the country, it has a high potential for energy savings. The total energy input reduction with the genetic algorithm was 17.05%, while the imperialist competitive algorithm achieved a higher reduction of 26.40%. Natural gas consumption decreased by 3.82% using the genetic algorithm, and by 27.60% with the imperialist competitive algorithm. Direct energy savings were 16.97% for the genetic algorithm and 27.48% for the imperialist competitive algorithm. Soil acidification reduction was 23.03% with the imperialist competitive algorithm and 19.19% with the genetic algorithm, compared to conditions before optimization.ConclusionIn general, it can be concluded that, given the growing demand for sugar production and related industries, as well as the high efficiency of the sugar production sector, it is advisable to utilize expert knowledge and apply meta-heuristics methods to optimize energy consumption and available inputs with the aim of reducing harmful environmental impacts.
Research Article
The relationship between machine and soil
N. Farhadi; A. Mardani; A. Hosainpour; B. Golanbari
Abstract
Introduction The formation of ruts induced by vehicle traffic poses a significant challenge for agricultural soils due to soil compaction both at the surface and deeper layers. This phenomenon compromises vehicle performance increases energy consumption, and leads to long-term environmental degradation, ...
Read More
Introduction The formation of ruts induced by vehicle traffic poses a significant challenge for agricultural soils due to soil compaction both at the surface and deeper layers. This phenomenon compromises vehicle performance increases energy consumption, and leads to long-term environmental degradation, such as soil erosion and fertility reduction. To enhance vehicle performance and reduce soil damage, it is crucial to accurately predict how factors such as vehicle speed, vertical load, and the number of passes impact rut depth. The findings of this study hold significant practical implications, facilitating the development for the creation of more efficient agricultural practices, while simultaneously minimizing environmental impact. The complexity of these interactions necessitates using machine learning models, especially artificial neural networks (ANNs), to predict rut depth based on input parameters. In this study, two machine learning models, namely the multilayer perceptron (MLP) and the radial basis function (RBF) networks, were employed to predict rut depth.Materials and MethodsExperiments were conducted using a soil bin that allows for precise control of independent parameters, measuring 24 meters in length, 2 meters in width, and 0.8 meters in depth. The soil used was agricultural soil, comprising 35% sand, 22% silt, and 43% clay, with a moisture content of 8%. The tests included three independent parameters: vertical load (2, 3, and 4 kN), forward speed (1, 2, and 3 km h-1), and number of wheel passes (up to 15). Two types of traction devices, including a rubber wheel and a track wheel, were tested. A caliper was used to measure the rut depth after each pass with an accuracy of 0.02 mm. The data collected from soil bin tests were used to train neural network models in MATLAB 2021-b software. The MLP model had a topology with two hidden layers and included three inputs and one output. In the RBF model, the network topology had a single hidden layer. The trial-and-error method was used to adjust the hyperparameters of the neural networks, including the number of neurons in the hidden layers, the learning rate, and momentum for the MLP network, as well as the spread rate and regularization rate for the RBF network.Results and DiscussionExperimental data confirmed that increasing the vertical load and the number of passes resulted in deeper ruts. Conversely, an increase in speed led to a reduction in rut depth, particularly during the initial pass. Both artificial neural network (ANN) models accurately predicted rut depth, with the multilayer perceptron (MLP) neural network outperforming the radial basis function (RBF) neural network. Specifically, the root mean square error (RMSE) for the optimal MLP model, which utilized a learning rate of 0.001 and a momentum of 0.67, was 0.10. In contrast, the optimal RBF model, with an expansion rate of 0.23456, yielded an RMSE of 0.12. The findings indicate that the MLP artificial neural network model surpasses the RBF neural network model in terms of accuracy and overall performance. However, the RBF neural network exhibits a faster response time, making it particularly suitable for real-time applications.ConclusionThis study demonstrates the efficacy of machine learning techniques, particularly artificial neural networks (ANNs), in predicting rut depth caused by off-road vehicle traffic. Both multilayer perceptron (MLP) and radial basis function (RBF) neural networks exhibited robust predictive capabilities, with the MLP model providing slightly superior accuracy and the RBF model offering better computational efficiency. These findings highlight the potential of machine learning in modeling complex interactions between soil and vehicles, which can enhance vehicle performance, mitigate soil erosion, and guide the design of off-road vehicles. Future research directions could include investigating additional soil parameters, various vehicle configurations, and the real-world implementation of autonomous off-road vehicles to promote more environmentally sustainable operations.
