A Example >0.7 probability of guard c 61<0.7 0、7 61 meanof estimated 8, guard boundar 1>0.7 has-ball/+ 0.5 P(true->false I estimate 1=0. 0.5 Hybrid Mode Estimation and Gaussian Filtering with Hybrid HMM Models 16412/6.834 Lecture,15 March2004 0.5 true 09 Concurrent Probabilistic Hybrid Automata PHA composition 1,2,B1,62 failed ball: tre false Gaussian filtering Merge the difference equations in component models e.g., equation solver Track sequences of full mode assignments as before Compute the transition probability component-wise Hybrid Mode Estimation and Gaussian Filtering with Hybrid HMM Models 6412/6.834 Lecture,15 March2004 desireeHybrid Mode Estimation and Gaussian Filtering with Hybrid HMM Models 16.412 / 6.834 Lecture, 15 March 2004 25 Example 0.7 T1 meanof estimated T1 guard boundary probability of guard c 0.7 T1  0.7 T1 ! 0.7 T1 ! 0.7 T1  ( ) 1 ( ) 1 ˆ , i t i xt C  P(true -> false | estimate i) = 0.1* + 0.5* Hybrid Mode Estimation and Gaussian Filtering with Hybrid HMM Models 16.412 / 6.834 Lecture, 15 March 2004 26 Concurrent Probabilistic Hybrid Automata z PHA + composition 1 2 1 2 T ,T ,T ,T   µ µ T ball: Gaussian filtering: z Merge the difference equations in component models ƒ e.g., equation solver z Track sequences of full mode assignments as before z Compute the transition probability component-wise