摘要
为了在多传感器数据融合中实现各传感器输出误差分布归一化,根据模糊集合理论的模糊可能性方法,给出了从概率分布到可能性分布的截断三角变换,该变换可用于所有单峰对称概率分布。应用这个变换,建立了传感器测量误差的模糊集合模型。对于三种情况的传感器测量误差及分布,分别给出了对应的模糊可能性分布模型。同时,该可能性分布也可用于计算测量结果的置信区间。
In order to execute assimilated of probability distribution for fusion of various sensor measurements, a fuzzy representation, based on the possibility theory, could be preferred. A truncated triangular probabilitypossibility transformation to be applied to any unimodal and symmetric probability distribution was proposed. By applying this transformation, a fuzzy model of sensors measurement date was built. Three main situations of sensor with different probabilistic model had been considered, and for each situation, an adapted fuzzy modeling had been proposed. Moreover, this possibility distribution could also be built by the knowledge of confidence intervals associated to the measurement
出处
《中国测试》
CAS
2009年第1期37-39,共3页
China Measurement & Test
关键词
传感器测量
截断三角变换
模糊模型
概率分布
分析
Sensor measurement
Truncated triangular transformation
Fuzzy model
Probability distribution
Analysis
