卫星遥感农业监测系统中权重优化的实例修改算法

针对播种机设计过程需求获取的复杂性和可用性低等问题,采用遥感等技术动态监测土壤、农作物生长过程中的一些参数,以优化设计需求特征,并以此为基础,提出改进的混合权重实例修改算法。该算法以相似实例检索为前提,以相似实例集中的特征相似度、问题-解特征间的关联度、问题特征间的自相关度为因子,通过矩阵乘法来构造解特征混合权重矩阵,进行实例修改。将其应用到铲式播种机的数字化设计系统中,以实现智能设计。研究结果表明新算法能有效提高机械设计结果的可用性和准确度。

Case Adaptation Algorithm Based on Weight Optimization in Satellite Remote Sensing Monitoring System for Agriculture

To solve the problems such as the complicated requirement acquisition and low requirement usability in the design process of intelligent planters, some parameters of crops and soil are monitored dynamically by use of the remote sensing and other technologies to optimize the design requirement features, and then a new case adaptation algorithm with improved hybrid weight is presented in this paper. In the new algorithm, to perform case adaptation using the principle of feature-based case adaptation (FCA), a hybrid weighted matrix is constructed by multiplying the similarity matrix of similar cases, the correlation coefficient matrix of problem-solution (P-S) features, and the pair-wise comparison matrix of problem-problem features on the basis of the retrieval of similarity cases. The proposed algorithm is applied to the digital design system of a spade planter to perform intelligent planter design. The results of the experiment indicate that the new case adaptation algorithm can achieve better adaptation performance than that of other adaptation methods, as evaluated in terms of the accuracy and the usability of the mechanical design results.