Engormix/Animal Feed/Technical articles

Study of Real-Coded Hybrid Genetic Algorithm (RGA) to Find Least-Cost Ration for Non-pregnant Dairy Buffaloes

Published on: 7/26/2022
Author/s : Ravinder Singh Kuntal 1, Radha Gupta 2, Duraisamy Rajendran 3 & Vishal Patil 1 / 1 Department of Mathematics, Jain University, Bengaluru, 562112, India; 2 Department of Mathematics, Dayananda Sagar University, Bengaluru, India; 3 ICAR-National Institute of Animal Nutrition and Physiology, Adugodi, Bengaluru, 560030, India.
Summary

In Mandya District of Karnataka, the cost of milk per liter was more in case of buffaloes compared to local cows due to high fat content and high nutritive value of buffaloes’ milk than its counterpart. Based on earlier research, it was clear that the productivity of the buffaloes maintained by different dairy farms was lower. Therefore, there is a need to focus on two important aspects of dairy farming: One to increase the milk productivity of buffaloes and the other one to minimize the diet cost by upgrading the scientific dairy farming practices. Though several techniques are in use for animal diet formulation but a successful application of soft computing technique to improve the quality of the solution is always preferred as the rigidity of the functions in LPP can be easily handled. Therefore, to meet the nutrient requirements at lowest cost, we have developed a hybrid real-coded genetic algorithm (RGA) for formulating the least-cost ration for dairy buffaloes. This technique is better than old conventional techniques, in the sense that it does not break if the inputs are modified and provides better results over complex problem even if it is linear programming model. The linear programming model is developed from primary data collected from NIANP and as per the standards of ICAR. Next, the developed algorithms RGA and Hybrid RGA are executed and compared with other least-cost feed formulation techniques in non-pregnant dairy buffalo weighing 450 kg and yielding 10 L milk with 6% of fat content as a model and considering standard nutrient requirement on dry matter basis. Further, goal programming model (GP model) has been developed as there are two high priority objectives (out of eight goals), i.e., least-cost and dry matter intake, to be achieved simultaneously, if possible. This GP model is also solved by hybrid RGA showing that four goals out of eight are fully achieved. It could be concluded that real-coded genetic algorithm (RGA) with hybrid function can effectively be used to economize the total mixed ration cost such that the feed requirements of the animals are met without any nutrients deficiency.

Keywords: Dairy feed, Least cost, Real-coded genetic algorithm, Goal programming.

           

Abstract published in Soft Computing for Problem Solving. Advances in Intelligent Systems and Computing, vol 817. Springer, Singapore. https://doi.org/10.1007/978-981-13-1595-4_30.

 
Author/s :
 
Views22Comments 0StatisticsShare