Scholarlink

<p>Raised bed planting offers numerous benefits on soil health, drainage, water savings, and improved economic and energy use efficiency. The dimension of raised bed especially top width vary for different crops according to specific growth habits, space need, and irrigation requirement of the crop, which require specialized bed former suitable for multiple crops. Considering these views in mind, the present study was undertaken on development and performance evaluation of variable width raised bed former under soil bin and field conditions. The soil bin experiments were conducted according to factorial completely randomized design with three replications to examine the effect of soil moisture content (11, 13 and 15% on wet basis), working width (0.70, 0.80 and 1.0&#xa0;m) and forward speed (0.42, 0.56 and 0.69&#xa0;m s<sup>− 1</sup>) on draft force of test rig. The soil bin data were used to build regression and radial basis function (RBF) neural network models to predict the specific draft and optimizing the input parameters using RBF neural network-particle swarm optimization (PSO) technique. Both, regression and RBF neural network models predicted the specific draft of test rig with good accuracy (R<sup>2</sup> value &gt; 0.98) in both training and testing phases. In soil bin, actual specific draft of 4002.4&#xa0;N m<sup>− 1</sup> was observed against the predicted value of 4056.6&#xa0;N m<sup>− 1</sup> with optimal input parameters given by RBF neural network-PSO technique. A variation of ± 7.2% was observed between predicted and actual value of specific draft for tractor operated variable width raised bed former in the field conditions with similar input settings. The findings of the study suggest that use of modern techniques such as neural network and particle swarm optimization might be a good approach for optimizing the input parameters and diminishing the energy expenditure during the field operation of tillage machinery.</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

Development and performance evaluation of variable width raised bed former with optimal parameters predicted by RBF neural network-PSO technique

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月12日

查看全文
  • Chetankumar Prakash Sawant,
  • Bhaskar Bharat Gaikwad,
  • Ajit Pralhad Magar,
  • Abhijit Khadatkar,
  • Neeraj Kumar,
  • K. V. Ramana Rao,
  • Ved Prakash Chaudhary

摘要

Raised bed planting offers numerous benefits on soil health, drainage, water savings, and improved economic and energy use efficiency. The dimension of raised bed especially top width vary for different crops according to specific growth habits, space need, and irrigation requirement of the crop, which require specialized bed former suitable for multiple crops. Considering these views in mind, the present study was undertaken on development and performance evaluation of variable width raised bed former under soil bin and field conditions. The soil bin experiments were conducted according to factorial completely randomized design with three replications to examine the effect of soil moisture content (11, 13 and 15% on wet basis), working width (0.70, 0.80 and 1.0 m) and forward speed (0.42, 0.56 and 0.69 m s− 1) on draft force of test rig. The soil bin data were used to build regression and radial basis function (RBF) neural network models to predict the specific draft and optimizing the input parameters using RBF neural network-particle swarm optimization (PSO) technique. Both, regression and RBF neural network models predicted the specific draft of test rig with good accuracy (R2 value > 0.98) in both training and testing phases. In soil bin, actual specific draft of 4002.4 N m− 1 was observed against the predicted value of 4056.6 N m− 1 with optimal input parameters given by RBF neural network-PSO technique. A variation of ± 7.2% was observed between predicted and actual value of specific draft for tractor operated variable width raised bed former in the field conditions with similar input settings. The findings of the study suggest that use of modern techniques such as neural network and particle swarm optimization might be a good approach for optimizing the input parameters and diminishing the energy expenditure during the field operation of tillage machinery.