改进层次分析法在生猪养殖工艺评估的应用

Application of improved analytic hierarchy process in evaluation of pig breeding technology

针对养猪业养殖工艺评估问题,基于层次分析法(AHP)建立相对规范的生猪养殖技术评价体系模型。引入智能算法思想对层次分析法中一致性问题做出解决方案。试验将采用模拟退火算法(SA)与粒子群算法(PSO)相结合的方式优化一致性。首先构建分层递进的层级模型,其次初始化粒子和给定起始温度和退火系数利用模拟退火粒子群算法更新粒子状态优化判断矩阵一致性参数。试验结果表明:初始一致性参数0.2319,0.1273,0.1304,0.1304,粒子群优化参数为0.0947,0.0981,0.0992,0.0995,经本文方法优化后为0.0857,0.0973,0.0981,0.0987。该方法能够有效避免单一粒子群算法陷入局部最优问题,并且能得到更好的一致性指标。说明此种方法可以有效评估分析出生猪养殖工艺中较重要指标,充分反映专家评估意愿,减少个人主观因素的影响,能较为理想地得出评估结果。

To pig breeding technology evaluation problem,establishing a relatively standard model for evaluation system of pig breeding technology based on AHP.Introduce the idea of intelligent algorithm to solve the consistency problem in the analytic hierarchy process.The experiment will use a combination of simulated annealing algorithm(SA)and particle swarm optimization(PSO)to optimize consistency.Firstly,build a hierarchical model of hierarchical progression.Secondly,initialize the particles and the given starting temperature and annealing coefficients.Use simulated annealing particle swarm algorithm to update the particle state optimization judgment matrix consistency parameters.The test results show that the initial consistency parameters 0.2319,0.1273,0.1304,0.1304.The particle swarm optimization parameters are 0.0947,0.0981,0.0992,0.0995.After optimized by the method in this paper,it is 0.0857,0.0973,0.0981,0.0987.This method can effectively avoid the single particle swarm optimization algorithm falling into the local optimal problem,and can get better consistency index.It shows that this method can effectively evaluate and analyze the more important indicators in the pig breeding process,fully reflect the expert’s willingness to evaluate,reduce the impact of personal subjective factors,and can obtain the evaluation results more ideally.