基于投影寻踪技术的保护性耕作效益评价

    Benifit evaluation of conservation tillage based on projection pursuit

    • 摘要: 为了客观评价保护性耕作的综合效益,该文采用投影寻踪技术,建立了保护性耕作综合效益评价模型,模型的平均绝对误差为0.0393,平均相对误差为1.75%。该文以辽宁和山西试验区为例,通过优化投影方向,将6个样本的8个评价指标投影到低维子空间,利用投影值计算综合效益等级并进行评价。评价结果表明,山西寿阳的免耕碎秆播种玉米综合效益最好,传统耕作播种玉米综合效益最差,6个样本的综合效益从优至劣排序为:山西寿阳免耕碎秆覆盖(B3),山西寿阳深松碎秆覆盖(B2),辽宁凌海免耕碎秆覆盖(A3),辽宁凌海免耕高茬覆盖(A2),辽宁凌海传统耕作(A1),山西寿阳传统耕作(B1)。研究结果表明,与传统耕作相比较,长期的保护性耕作能够较大提高农业综合效益。

       

      Abstract: In order to evaluate objectively the composite achievement of conservation tillage, a conservation tillage evaluation model based on projection pursuit was established. And the check results of the evaluation model indicated that the mean absolute error was 0.0393, and the average relative error was 1.75%. After optimizing the project direction, the multi-dimension data of the eight evaluation indices were synthesized to one dimension, and we could evaluate easily each item of test areas in Shanxi Province and Liaoning Province with the projection data. Results of the evaluation indicated that the order from good to bad of composite achievement of six samples was as follows: no-tillage with crushed stalk mulching in Shouyang test areas (B3), subsoiling with high stubble mulching in Shouyang test areas (B2), no-tillage with crushed stalk mulching in Linghai test areas (A3), no-tillage with high stubble mulching in Linghai test areas (A2), conventional tillage in Linghai test areas(A1), conventional tillage in Shouyang test areas (B1). And the results show that long term conservation tillage is practical and effective for promoting agricultural evaluation.

       

    /

    返回文章
    返回