中文摘要：

针对作物叶水势的连续测量问题，提出一种软测量建模新方法。该方法结合土壤一植物一大气连续体理论，选取8个易于获取的作物微环境信息作为辅助变量，利用自适应神经模糊推理系统（ANFIS）实现对作物叶水势的软测量建模。通过对实验数据的Matlab仿真表明，该模型方法简单有效，具有较高的软估计精度。

英文摘要：

Aiming at the problem of continuous measurement for leaf water potential, a new soft-sensing mod- eling method is proposed. By combining the theory of SPAC, the new mehod choosed eight accessible microenvironment information as the correlation instrumental variables, then an 8 input-loutput ANFIS model for leaf water potential is established. Simulation results have verified the efficiency of the method and high estimating precision.