pdf download - Software and Computer Technology - TU Delft
pdf download - Software and Computer Technology - TU Delft
pdf download - Software and Computer Technology - TU Delft
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
4.5 Entropy Gain Model-Based Fault Diagnosis<br />
Figure 4.2: 4-inverter example<br />
are already known; the input x is true <strong>and</strong> the output y is false. The considered points are referred to<br />
by a, b <strong>and</strong> c.<br />
The prerequisite for the entropy calculation is the assignment of an a priori probability to each<br />
health variable. This a priori probability is the probability that h=0, without having made any<br />
observations. The 4-inverter system has 4 health variables, with a priori probabilities:<br />
p(¬h A ) = 0.01<br />
p(¬h B ) = 0.01<br />
p(¬h C ) = 0.01<br />
p(¬h D ) = 0.01 (4.8)<br />
The interpretation of, for example p(¬h A ) = 0.01, is not that r<strong>and</strong>omly picking n components yields<br />
0.01 ∗ n broken components. The meaning depends on the model of the system. If no observations<br />
have been made, all single faults <strong>and</strong> multiple faults are possible. There are 4 health variables. This<br />
means that if x <strong>and</strong> y are not observed, the number of possible diagnoses is 2 4 = 16. This information,<br />
the existence of 16 possible diagnoses, can be stored in 4 bits. The corresponding outcome of<br />
the LYDIA diagnostic engine is as follows:<br />
(0.960596) hA = true, hB = true, hC = true, hD = true<br />
(0.00970299) hA = false, hB = true, hC = true, hD = true<br />
(0.00970299) hA = true, hB = false, hC = true, hD = true<br />
(0.00970299) hA = true, hB = true, hC = false, hD = true<br />
(0.00970299) hA = true, hB = true, hC = true, hD = false<br />
...<br />
The firstly listed diagnosis, that states that all inverters are healthy, has highest probability. Specific<br />
diagnoses are referred to as health vectors. The correct functioning of all inverters is denoted<br />
by the health vector ¯h = {1,1,1,1}, <strong>and</strong> its probability is calculated by using the formula:<br />
p({h A ,h B ,h C ,h D }) = p(hA) ∗ p(hB) ∗ p(hC) ∗ p(hD) (4.9)<br />
This formula assumes that the variables are independent: the correct or incorrect functioning of<br />
one component has no effect on the health of any other component. The interpretation of, p(¯h =<br />
{1,1,1,1}) = 0.970596, is that diagnosing the 4-inverter system at an arbitrary moment, without<br />
any observations being made until that moment, yields a probability of 0.97% that the entire system<br />
is healthy. p(¯h = {0,1,1,1}) = 0.00970299 means that there is a probability of 0.01% that hA=0.<br />
Also, there is probability of 0.01% that hB=0, etcetera.<br />
The a priori probabilities of the health vectors allow for the calculation of entropy. The entropy<br />
(H) is defined as:<br />
H = −∑ p i log p i , (4.10)<br />
44