Scholarlink

<p>Forage harvesting is a critical agricultural practice that ensures prompt crop collection to maintain nutritional quality and economic viability. However, operating heavy machinery on sensitive field areas can lead to soil compaction, which may subsequently reduce crop yields. This study presents a comprehensive evaluation of advanced optimization techniques for capacitated multi-objective route planning in agricultural fields. Two algorithms including <i>Non-Dominated Sorting Genetic Algorithm II (NSGA-II)</i> and a novel <i>Stochastic Greedy Algorithm (SGA)</i> are compared within a parallel computing environment designed to significantly reduce computation time. By leveraging an innovative approach to three-dimensional capacitated route planning, the proposed method eliminates traditional block-based around the obstacle for the first time, thereby expanding the solution space and enabling the exploration of track combinations across entire fields. Experimental results demonstrate that the SGA reduced non-working distance by up to 41%, fuel consumption by 13%, and soil compaction stress length by 5% in a small field (19 tracks), and by 36%, 12%, and 30%, respectively, in a large field (82 tracks). NSGA-II also improved performance compared to conventional methods but was less efficient in larger, more complex fields. These findings indicate that advanced optimization strategies can substantially enhance operational efficiency and reduce both resource consumption and soil compaction in capacitated agricultural operations. Future work will focus on extending the optimization framework to support multiple harvesters operating simultaneously, enabling coordinated route planning and further reducing total field coverage time.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Three-dimensional capacitated route planning optimization using parallel computing for agricultural field involving obstacle

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • Erfan Khosravani Moghadam,
  • Michael Nørremark,
  • Kun Zhou,
  • René Søndergaard Nilsson,
  • Kenneth Guldbrandt Lausdahl,
  • Claus Aage Grøn Sørensen

摘要

Forage harvesting is a critical agricultural practice that ensures prompt crop collection to maintain nutritional quality and economic viability. However, operating heavy machinery on sensitive field areas can lead to soil compaction, which may subsequently reduce crop yields. This study presents a comprehensive evaluation of advanced optimization techniques for capacitated multi-objective route planning in agricultural fields. Two algorithms including Non-Dominated Sorting Genetic Algorithm II (NSGA-II) and a novel Stochastic Greedy Algorithm (SGA) are compared within a parallel computing environment designed to significantly reduce computation time. By leveraging an innovative approach to three-dimensional capacitated route planning, the proposed method eliminates traditional block-based around the obstacle for the first time, thereby expanding the solution space and enabling the exploration of track combinations across entire fields. Experimental results demonstrate that the SGA reduced non-working distance by up to 41%, fuel consumption by 13%, and soil compaction stress length by 5% in a small field (19 tracks), and by 36%, 12%, and 30%, respectively, in a large field (82 tracks). NSGA-II also improved performance compared to conventional methods but was less efficient in larger, more complex fields. These findings indicate that advanced optimization strategies can substantially enhance operational efficiency and reduce both resource consumption and soil compaction in capacitated agricultural operations. Future work will focus on extending the optimization framework to support multiple harvesters operating simultaneously, enabling coordinated route planning and further reducing total field coverage time.