نوع مقاله : مقاله پژوهشی
نویسندگان
1 گروه مهندسی مکانیک ماشین‌های کشاورزی، دانشکده کشاورزی و منابع طبیعی، دانشگاه تهران، کرج، ایران
2 مرکز تحقیقات کشاورزی و آموزش کشاورزی و منابع طبیعی استان سمنان، سازمان تحقیقات، آموزش و ترویج کشاورزی، سمنان، ایران
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
This study aimed to analyze the energy and economic aspects of rose production in Damghan County and model the optimal use of inputs using machine learning algorithms. The required data were collected through questionnaires and interviews. The results showed that the total energy consumed in rose production was 43,438 megajoules per hectare, with electricity accounting for 79.6 percent of the energy consumption. The energy efficiency index was 0.02 kg/megajoule, which indicates a significant energy loss in the production system. On the other hand, the economic analysis indicated a net profit of 44.395 million tomans per hectare and a benefit-to-cost ratio of 2.07, which indicates the appropriate economic justification for rose production despite high energy consumption. Gradient Booster (GBR), Enhanced Gradient Booster (XGBR), and Random Forest (RFR) algorithms were used to model and predict energy consumption and costs. The results showed that the GBR model with a coefficient of determination (R2) of 0.99 and a minimum error of 251.97 has the best performance in predicting energy and costs. Also, sensitivity analysis using the SHAP method revealed that animal manure and electricity have the greatest impact on energy consumption, while water management and chemical fertilizers play a key role in economic profitability. The results showed that optimizing energy consumption in rose production is possible by reducing electricity and fertilizer consumption, and the use of machine learning is also suggested as an efficient tool in predicting and managing agricultural inputs.
کلیدواژه‌ها [English]
EXTENDED ABSTRACT
Medicinal plants, especially rose (Rosa damascena), play an important role in the agricultural economy of Iran due to their extensive applications in the perfumery, food, and pharmaceutical industries. With the increasing global demand for plant products, it is essential to optimize energy consumption and analyze the economic aspects of the production of these plants. Previous studies have shown that agricultural production systems are often faced with high energy waste and costs, but few studies have addressed the modeling of these parameters with modern tools such as machine learning. This study aimed to analyze the energy and economic aspects of rose production in Damghan County and provide predictive models based on machine learning algorithms.
In this study, energy modeling and economic analysis of rose production in Damghan County was conducted using machine learning algorithms. The energy consumption pattern and economic and cost analysis of rose production were examined. Information was collected through a specialized questionnaire and interview, and simple random sampling was used to determine the sample size. Three algorithms were used for modeling: gradient boosting regression (GBR), enhanced gradient boosting regression (XGBR), and random forest regression (RFR).
The total energy consumption in production was 43438.83 megajoules per hectare. Electrical energy and nitrogen fertilizer accounted for the largest share of energy consumption, accounting for about 79.60 and 9.92 percent of the total energy consumption, respectively. The energy ratio, energy efficiency, energy intensity, and net energy added were 0.08, 0.021 kg/MJ, 47.20 MJ/kg, and net energy was -39574 megajoule, respectively. The results of economic analysis showed that the total production costs of rose are 44.395 million tomans per hectare. Animal fertilizer, with a share of 33.15 percent, and human labor, with a share of 29.47 percent of the total costs, accounted for the largest share of the production costs. For energy, GBR, XGBR, and RFR were the best models with R2 values of 0.99, 0.97, and 0.92, respectively, and for the economic production model, GBR, XGBR, and RFR were the best models with R2 values of 0.99, 0.96, and 0.88.
This study shows that rose production in the study area is economically profitable and justifiable, although energy consumption pattern analysis indicates significant energy waste in rose production. The application of machine learning methods in this study demonstrated the high ability of this technology to predict and optimize resource consumption. The results showed that intelligent management of key inputs such as water resources, irrigation systems, and fertilizers can help reduce energy consumption, reduce costs, and increase profits. These findings emphasize the need to apply advanced technical solutions and adopt modern and mechanized management methods to achieve sustainable and efficient rose production, in a way that maintains economic justification while optimizing resource consumption.
The idea and design of this study were carried out by Farzane Bahadori and Seyed Omid Davodalmousavi. The identification method, software and validation were carried out by Shahin Rafiee and the research and data and resource collection and writing of the first draft were carried out by Seyed Omid Davodalmousavi. The project management was carried out by Farzane Bahadori and Seyed Omid Davodalmousavi. All authors have read and agreed to the published version of the manuscript.
All data generated or analyzed during this study are available upon request from the corresponding author.
This study was carried out by the Forest and Rangeland Research Department, Agricultural and Natural Resources Research and Education Center of Semnan Province, Agricultural Research, Education and Extension Organization, Semnan, Iran, to whom we are grateful. The authors would like to thank all the participants in the present study.
The authors avoided data fabrication, falsification, plagiarism and misconduct.
The author declares no conflicts of interest.
)