中文摘要：

以信息连续分布的智能空间内传感器优化布置为研究目标，以温度信息为例对智能空间进行势场模拟，在等势线上选取曲率较大、密集的点作为兴趣点（POI），提取POI初集；提出了一种基于质量阈值聚类的自适应二分法（简称QT算法），对兴POI初集进行优化；通过仿真和实测实验验证，在减少传感器数目的同时能够合理的规划区域中各被监测点，使得各传感器节点的效能得到很好的利用；所提出的方法同样适合智能空间中其它连续的信息场，如湿度、风速、电磁污染和VOCs等。

英文摘要：

This paper addressed the sensor - optimization of continuously - distributed - information for smart space. Poten- tial for temperature information was simulated, and POIs were selected in equipotential lines of large curvature or high density, thus the primary POI - Set was formed ; A novel Quality - Threshold - Clustering - Recursive - Bisection algorithm＇ （QT） was proposed for the Optimization of POI - Set ; Simulations and experiments show that minimum sensors may give reasonable results for sensor utilization and information acquisition; Method proposed may similarly be used for other information fields, including humidity, wind field, electromagnetic pollution ,VOCs and etc.