• Home
  • Textbooks
  • Fuzzy Logic with Engineering Applications
  • Automated Methods for Fuzzy Systems

Fuzzy Logic with Engineering Applications

Timothy Ross

Chapter 7

Automated Methods for Fuzzy Systems - all with Video Answers

Educators


Chapter Questions

Problem 1

Earlier in this chapter the recursive least squares algorithm was demonstrated using training set Z of Table 7.1; however, only one cycle was performed using the input data tuples. Perform an additional cycle with the input data-tuples to determine if the predicted output changes.

Check back soon!

Problem 2

Earlier in this chapter the recursive least squares algorithm was demonstrated using training set $Z$ of Table 7.1.
(a) Modify the input membership functions centers to the following values and develop a fuzzy model for the Z of Table 7.1 using the RLS algorithm (perform two cycles). Note that the remaining rule-base parameters are the same as those used in the text, e.g.,

$$
\begin{aligned}
& \sigma_j^i=2 \quad \text { and } \quad \hat{\theta}(0)=\left[\begin{array}{l}
0.3647 \\
8.1775
\end{array}\right] \\
& c_1^1=1 \quad c_1^2=3 \\
& c_2^1=3 \quad c_2^2=6
\end{aligned}
$$

(b) In part (a) the input membership function centers were slightly different than those used in the text but the spreads were the same values as those used in the text. Now change the spreads to the following values, $\sigma_1^1=3, \sigma_1^2=2, \sigma_2^1=1, \sigma_2^2=2$, and the remaining values should be the same as in part (a). Develop a fuzzy model for the $Z$ of Table 7.1 using the RLS algorithm (perform two cycles.)

Check back soon!

Problem 3

Using the software provided on the publisher's website (see the preface), improve the output of the recursive least squares model presented in Example 7.1 by modifying the input membership function parameters of Table 7.4 using the gradient method.

Check back soon!

Problem 4

Earlier in the chapter an example using the gradient method was presented using only one time step; provide the second and third time steps.

Check back soon!

Problem 5

Using the clustering method develop a fuzzy model for the input-output data presented in Table 7.2 .

Check back soon!

Problem 6

In Example 7.1 an initial rule was specified for the application of the MLFE algorithm. Change these values for the input membership function centers to $x_1=12,140 \mathrm{psi}$ and $x_2=6510 \mathrm{psi}$ and the output membership function center to $\boldsymbol{d}_L=3.22360$ inch. Also incorporate two test factors, one for $x_1$ and another for $x_2$, equal to 2.1 and 5 respectively. Keeping all other values the same as in Example 7.1 develop a fuzzy model using MLFE. Will the predictability of the developed fuzzy model be improved by changing the test factor for $x_2$ ? If so, does increasing or decreasing the test factor improve the output MLFE?

Check back soon!

Problem 7

Using the batch least squares algorithm improve the fuzzy model produced in Problem 7.4.

Check back soon!