Using the Fuzzy Toolbox in MATLAB

4 Steps to teach you how to use MATLAB Fuzzy Control Toolbox

MATLAB Fuzzy Control Toolbox provides a very convenient way for the design of fuzzy controller, through which we do not need complicated fuzzy, fuzzy inference and inverse ambiguity operation, only need to set the corresponding parameters, we can quickly get the controller we need, and the modification is very convenient. Based on the fuzzy controller design step, the fuzzy controller is designed with MATLAB Toolbox.

First we enter the fuzzy in the command window of MATLAB, the carriage return will come out such a window.

Below we are all in such a window for the design of the fuzzy controller.

1. Determine the structure of fuzzy controller: that is, according to the specific system to determine the input and output amount.

Here we can choose the standard two-dimensional control structure, that is, the input is the error E and the error change EC, the output is the control amount U. Note that the variables here are also exact quantities. The corresponding fuzzy amount is E,EC and u, we can choose to add the input (add Variable) to realize the double-entry single-out control structure.

2. Fuzziness of input and output variables: the exact amount of input and output is transformed into a fuzzy set of corresponding language variables.

First we need to determine the fuzzy subset that describes the language value of the input and output variables, such as {NB,NM,NS,ZO,PS,PM,PB}, and set the domain of the input and output variables, for example, we can set the error E (at this time as the fuzzy amount), the error change EC, the control of the U's domain are { -3,-2,-1,0, Then we select the appropriate membership function for the fuzzy language variable.

In the Fuzzy Control Toolbox, we can complete these steps in member Function edit. First we open the member Function edit window.

Then define the scope of the input and output variables, add the membership function, take e As an example, set the scope of the domain is [-3 3], add the number of membership function is 7.

Then, according to the design requirements of these membership functions are modified, including the corresponding language variables, membership function type.

3. Fuzzy inference decision-making algorithm design: fuzzy inference according to fuzzy control rules, and the decision-making fuzzy output amount.

The first step is to determine the fuzzy rule, the expert experience. For our two-dimensional control structure and the corresponding input fuzzy set, we can make 49 fuzzy control rules (in general, these rules are readily available, many textbooks have).

After the formulation, a fuzzy control rule matrix is formed, and then the fuzzy input quantity is calculated according to the corresponding fuzzy inference algorithm, and the output is determined.

4. The ambiguity of the output fuzzy quantity: The output of the fuzzy controller is a fuzzy set, and an exact exact quantity is judged by the inverse ambiguity method, where the fuzzy square

method, we choose the center of gravity method here.

5. Then Export to disk, you can get one. FIS file, this is the fuzzy controller you are designing.

Source: http://blog.sina.com.cn/s/blog_408540af0100avnv.html

Using the Fuzzy Toolbox in MATLAB