8 应用 这里给出两个实例来说明在多传感器跟踪中利用模糊逻辑进行数据相关的可行性。输入变量为位置和速度误差。假设它们有高斯pdf。图3中给出了高斯pdf的最优隶属(梯形部分)。为了把它表示成辨识框架,我们用几个模糊标记近似,如第6部分所讨论的图8和图9中所示辨识框架。相关器的输出为决策变量。决策函数是在革新函数的基础上定 8 Applications Two examples are given here to illustrate the feasibility of using fuzzy logic for data correlation in multisensor tracking. Input variables are position and speed errors. Suppose they have Gaussian pdf. Figure 3 shows the optimal membership of the Gaussian pdf (trapezoidal part). To represent it as a recognition framework, we approximate it with a few ambiguities, as shown in Figure 8 and Figure 9, discussed in Part 6. The output of the correlator is a decision variable. The decision function is based on the innovation function