区间两相模糊多目标模型在种植结构优化中的应用——以辽宁省大连市为例

Application of in inexact two-phase fuzzy multi-objective programming method to crop area planning

针对基于"Max-min"算子的区间模糊多目标规划仅采用一或两个控制变量放松所有目标和模糊约束会造成某些约束过满意而某些约束不满意的情况,本文引入两相模糊规划,构建了区间-两相模糊多目标规划模型,并以辽宁省大连市种植结构优化为例进行研究。结果表明,该模型引入多个控制变量放松每个不确定目标和约束条件,且要求它们分别不小于"Max-min"算子中相应目标和约束条件的隶属度,更充分地利用了约束资源,保证了求解的有效性,减少了农业灌溉用水量;另外区间形式的最优解及4种不同情景的优化方案为决策者提供了决策空间,更真实地反映输入参数的不确定性对配置结果的影响。

The interval fuzzy multi-objective linear programming based on Max-min operators is employed for only one or two control variables to relax the value of all objective functions and fuzzy constraints. However, this approach is not valid in many practical problems as some model objectives or fuzzy constraints are not satisfied and others are over-satisfied. To solve this problem, the two-phase fuzzy method is introduced to ensure the obtained solution is always efficient, and an interval two-phase fuzzy multi-objective programming model is proposed. Crop area planning is an important measure to save water in agriculture. Influenced by climatic and socioeconomic factors, it exhibits some complex characteristics, such as multi-objective and uncertainty. This combined model is applied to optimizing the cropping structure in Dalian City, in which parameter uncertainty is described as interval number and stipulation uncertainty as fuzzy relations. The scenario analysis is introduced to quantify the effect of food security policy on crop configuration area and water saving level on irrigation water use. The results are compared with those from the un-renewed model based on Max-min operators. It showed that the proposed model, which introduces multiple control variables to relax every objective and uncertainty constraint under different levels, can make better use of each constrained resource and provides a more satisfactory level in searching for optimal solutions. The optimization results for 2010 and the four scenarios are all interval numbers, which provide decision-making with a flexible decision space and reflect the impact of parameter uncertainty on the results of this model.