CN203350635U - Dynamic fuzzy control system - Google Patents

Abstract

The utility model discloses a dynamic fuzzy control system. The dynamic fuzzy control system comprises a dynamic fuzzification interface module, a dynamic fuzzy inference module, a defuzzification interface module and a dynamic fuzzy knowledge base module, wherein sampled input is received by the dynamic fuzzification interface module and compared with a reference input, and a deviation value between the sampled input and the reference input is mapped to a dynamic fuzzy input universe to obtain a dynamic fuzzy input; and inference and calculation are carried out by the dynamic fuzzy inference module according to a dynamic fuzzy control rule of the dynamic fuzzy knowledge base module to obtain a dynamic fuzzy control output, the dynamic fuzzy control output is defuzzified by the defuzzification interface module, a determined control output is obtained and output to an execution mechanism, and an object to be controlled is controlled by the execution mechanism. With the dynamic fuzzy control system, the dynamic problem in a non-linear dynamic fuzzy system process control is well solved, and dynamic fuzzy characteristics of the dynamic fuzzy control system is embodied and quality of the control system is improved.

Description

A kind of dynamic fuzzy control system

Technical field

The utility model relates to field of intelligent control, relates in particular to a kind of dynamic fuzzy control system.

Background technology

Traditional automatic control, comprise classical control and modern control, and classical control and the design of the modern controller of controlling all will be based upon on the basis of mathematical models of controlled device, but in actual applications, the influence factor of many systems is a lot, present non-linear, strong coupling, the time characteristic that becomes and lag behind, the mathematical models of its controlled device is very complicated or root is wooden does not just exist, and can not be controlled by traditional control method.

For this reason, people have proposed the intelligent control methods such as fuzzy control, artificial neural network control, for the control problem that solves complicated nonlinear system provides a kind of effective way; Wherein, fuzzy control is because it directly utilizes human expert's experience as control law, do not need to set up accurate mathematical model, make control mechanism and control strategy easy to understand and acceptance, and the control to non-linear complex object has shown the advantage that robustness is good, control performance is high, has now become a kind of important technical of Based Intelligent Control.

Its fuzzy logic part of current fuzzy control is static mostly, does not possess dynamic property, is a kind of static controller; A kind of new dynamic fuzzy controller (Xian Electronics Science and Technology University's journal) has been proposed in prior art, this dynamic fuzzy controller is realized by the input parameter of increase control system and extra fuzzy inference rule, not only increased the complicacy of system, and the control law of its core remains the fuzzy logic principle based on static.

In addition, for mainly relying on people's experience, static error, the fuzzy rule that overcomes fuzzy control greatly, do not there is the shortcomings such as self-learning capability, reach the purpose of optimal design, people combine fuzzy rule to have formed the fuzzy hybrid method with other self studies and optimisation technique again, as fuzzy control, Fuzzy Neural-network Control, fuzzy expert system control, fuzzy genetic algorithm control etc.But these methods are mainly to propose for the static characteristics in control system, ambiguity, self-study habit, can not solve well the dynamic problem in the Non-linear Dynamic Fuzzy System process control.

Therefore, be necessary to propose a kind of dynamic fuzzy control system, make it can solve the dynamic problem in complicated Non-linear Dynamic Fuzzy System process control, again the variation tendency of online tracing object parameters well.

The utility model content

In order to overcome the defect of prior art, the utility model aims to provide a kind of Fuzzy control system that not only has advantages of, and needn't additionally increase input parameter and the control law of control system, realize simply, intuitively, and the dynamic fuzzy control system of energy online tracing object parameters variation.

To achieve these goals, the utility model provides a kind of dynamic fuzzy control system, comprises dynamic fuzzy base module, dynamic mode gelatinization interface module, dynamic fuzzy reasoning module and sharpening interface module; Wherein, described dynamic mode gelatinization interface module connects a measurement mechanism, described dynamic mode gelatinization interface module receives the sampling input quantity of described measurement mechanism, the departure of described sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtain the dynamic fuzzy input quantity, and described dynamic fuzzy input quantity is input to described dynamic fuzzy reasoning module; Described dynamic fuzzy reasoning module connects respectively described dynamic mode gelatinization interface module and described sharpening interface module, the dynamic fuzzy control law provided according to described dynamic fuzzy base module by described dynamic fuzzy reasoning module, and described dynamic fuzzy input quantity is carried out to dynamic fuzzy contains and compositional rule of inference calculates, obtain the control output quantity of corresponding dynamic fuzzy output domain, and described control output quantity is inputed to described sharpening interface module; Described sharpening interface module connects a topworks, turn to definite control output quantity by described sharpening interface module by described control output quantity is clear, finally export described definite control output quantity to described topworks, by described topworks, one controlled device is controlled; Wherein, described dynamic fuzzy base module connects respectively described dynamic mode gelatinization interface module, described dynamic fuzzy reasoning module and described sharpening interface module, rule and relevant various parameters that described dynamic fuzzy base module provides dynamic fuzzy to process for described dynamic mode gelatinization interface module, described dynamic fuzzy reasoning module and described sharpening interface module;

Preferably, described dynamic fuzzy base module comprises dynamic fuzzy database and dynamic fuzzy control rule base, described dynamic fuzzy database comprises with dynamic fuzzy control law and dynamic fuzzy data processes relevant various parameters, mainly comprises the dynamic fuzzy input quantity of change of scale parameter, definition and controls the dynamic fuzzy Linguistic Value of output quantity and the subordinate function of each dynamic fuzzy Linguistic Value; Described dynamic fuzzy control rule base comprises some " IF-THEN " type dynamic fuzzy conditional clause, the Linguistic Value that the former piece of described some " IF-THEN " type dynamic fuzzy conditional clause is the dynamic fuzzy input quantity, the consequent of the dynamic fuzzy conditional clause of described some " IF-THEN " type is the Linguistic Value that dynamic fuzzy is controlled output quantity.

Preferably, the control desired value that described given reference value is described dynamic fuzzy control system.

Preferably, described sampling input quantity is the feedback signal that described measurement mechanism gathers from described controlled device.

Concrete, according to the description of above-mentioned utility model, the dynamic fuzzy control method of this utility model, comprise the steps:

(1) described dynamic mode gelatinization interface module receives the sampling input quantity of described measurement mechanism, and the departure of described sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtains the dynamic fuzzy input quantity;

(2) described dynamic fuzzy reasoning module, according to described dynamic fuzzy input quantity and dynamic fuzzy control law, is carried out reasoning and decision-making according to dynamic fuzzy implication computing rule and dynamic fuzzy reasoning compositional rule, obtains dynamic fuzzy and controls output quantity;

(3) described sharpening interface module is controlled described dynamic fuzzy that output quantity is clear turns to definite control output quantity, and export described definite control output quantity to topworks, by described topworks, according to described definite control output quantity, controlled device is controlled.

Preferably, described step (1) specifically comprises:

1. described dynamic mode gelatinization interface module receives the sampling input quantity that described measurement mechanism detects, and described sampling input quantity and given reference input are compared, and obtains departure;

2. described departure is carried out to change of scale, make described departure be mapped to the desired domain scope of dynamic fuzzy control inputs;

3. the departure that transforms to the domain scope is carried out to the dynamic fuzzy processing, make it become the dynamic fuzzy amount, and mean by corresponding dynamic fuzzy set.

Preferably, before being mapped on dynamic fuzzy input domain by the departure of described sampling input quantity and given reference input, need to determine respectively the subordinate function of basic domain, dynamic fuzzy domain, dynamic fuzzy Linguistic Value and the described dynamic fuzzy Linguistic Value of described departure and described control output quantity, wherein:

The actual change scope that described basic domain is described departure and described control output quantity;

Described dynamic fuzzy domain is that dynamic fuzzy is controlled desired variation range, comprises dynamic fuzzy input domain and dynamic fuzzy output domain corresponding to described control output quantity that described departure is corresponding;

Described dynamic fuzzy Linguistic Value is described departure and the value of described control output quantity on described dynamic fuzzy domain, with Dyna Mic Fuzzy Sets is incompatible, mean, the corresponding dynamic fuzzy set of described each dynamic fuzzy Linguistic Value, and described dynamic fuzzy set is determined by the subordinate function of described dynamic fuzzy Linguistic Value.

Preferably, described step (2) further comprises:

At first, subordinate function according to described each Linguistic Value of dynamic fuzzy input quantity, by the former piece of each " IF-THEN " type dynamic fuzzy conditional clause and the dynamic fuzzy vector corresponding to consequent of each " IF-THEN " type dynamic fuzzy conditional clause in described dynamic fuzzy control rule base, contain by dynamic fuzzy the Research on Dynamic Fuzzy Relationship matrix that computing rule obtains every rule, and the Research on Dynamic Fuzzy Relationship matrix of described every rule is done to union, obtain total Research on Dynamic Fuzzy Relationship matrix;

Then, according to the dynamic fuzzy reasoning compositional rule, by described dynamic fuzzy input quantity and described Research on Dynamic Fuzzy Relationship matrix, carry out compose operation, obtain dynamic fuzzy and control output quantity.

Preferably, described step (3) further comprises, described sharpening interface module is according to described dynamic fuzzy base module, utilize the clarification method in Dynamic Fuzzy Logic that described dynamic fuzzy control output quantity is converted to the clear amount that is actually used in control, i.e. described definite control output quantity.

Compared with prior art, the beneficial effects of the utility model are as follows:

1, the utility model is by adopting the control method based on Dynamic Fuzzy Logic, the departure of the sampling input quantity of measurement mechanism input and given reference input is mapped on dynamic fuzzy input domain according to the Dynamic Fuzzy Logic theory, and carry out reasoning according to the dynamic fuzzy control law and obtain dynamic fuzzy and control output quantity, to after its sharpening, controlled device be controlled, overcome the defect that existing control system can not solve the dynamic problem in the Non-linear Dynamic Fuzzy System process control well, thereby can embody better the dynamic fuzzy characteristic of control system, and the quality of raising control system.

2, dynamic fuzzy control system of the present utility model, being based on the Dynamic Fuzzy Logic theory is controlled, not only there is Fuzzy control system, and needn't increase extraly input parameter and the control law of control system, realize simple, directly perceived, not only can solve the dynamic problem in complicated Non-linear Dynamic Fuzzy System process control, again the variation tendency of online tracing object parameters well.

The accompanying drawing explanation

The structured flowchart that Fig. 1 is the utility model dynamic fuzzy control system;

Fig. 2 is the system chart that the utility model dynamic fuzzy control system is controlled in high water tank;

The process flow diagram that Fig. 3 is the utility model dynamic fuzzy control method;

Fig. 4 is that the high water tank that is applied to that the utility model is implemented is controlled dynamic fuzzy control method process flow diagram.

The symbol list:

100-dynamic fuzzy control system, 101-dynamic fuzzy base module, 102-dynamic mode gelatinization interface module, 103-dynamic fuzzy reasoning module, 104-sharpening interface module, 200-topworks, 300-water tank, 400-measurement mechanism.

Embodiment:

Referring to the accompanying drawing that the utility model embodiment is shown, hereinafter will describe in more detail the utility model.Yet the utility model can be with realizations such as multi-form, specifications, and should not be construed as the restriction of the embodiment be subject in this proposition.On the contrary, it is abundant and complete open in order to reach proposing these embodiment, and makes more relevant those skilled in the art person understand scope of the present utility model fully.In these accompanying drawings, for clearly visible, may zoom in or out relative size.

Please refer to Fig. 1 to Fig. 4, as shown in Figure 1, a kind of dynamic fuzzy control system 100 that the utility model provides, comprise dynamic fuzzy base module 101, dynamic mode gelatinization interface module 102, dynamic fuzzy reasoning module 103 and sharpening interface module 104; This dynamic mode gelatinization interface module 102 connects a measurement mechanism 400, this dynamic fuzzy reasoning module 0103 connects respectively dynamic mode gelatinization interface module 102 and sharpening interface module 104, this sharpening interface module 104 connects a topworks 200, and this dynamic fuzzy base module 101 connects respectively dynamic mode gelatinization interface module 102, dynamic fuzzy reasoning module 103 and sharpening interface module 104; Wherein: dynamic mode gelatinization interface module 102 receives the sampling input quantity of measurement mechanism 400, the departure of sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtain the dynamic fuzzy input quantity, and this dynamic fuzzy input quantity is input to dynamic fuzzy reasoning module 103; By the 103 pairs of dynamic fuzzy input quantities of dynamic fuzzy reasoning module, through dynamic fuzzy reasoning, calculate, the dynamic fuzzy that obtains corresponding dynamic fuzzy output domain is controlled output quantity, and dynamic fuzzy is controlled to output quantity inputs to sharpening interface module 104; And by sharpening interface module 104, dynamic fuzzy is controlled to output quantity is clear turns to definite control output quantity, and export the control output quantity of determining to topworks 200, by 200 pairs of controlled devices of topworks, controlled; Dynamic fuzzy base module 101 provides the control law of dynamic fuzzy processing and relevant various parameters for this dynamic mode gelatinization interface module 102, dynamic fuzzy reasoning module 103 and sharpening interface module 104.

Wherein, dynamic fuzzy base module 101 comprises dynamic fuzzy database and dynamic fuzzy control rule base, the dynamic fuzzy database comprises with dynamic fuzzy control law and dynamic fuzzy data processes relevant various parameters, mainly comprises the dynamic fuzzy input quantity of change of scale parameter, definition and controls the dynamic fuzzy Linguistic Value of output quantity and the subordinate function of each dynamic fuzzy Linguistic Value; The dynamic fuzzy control rule base comprises some " IF-THEN " type dynamic fuzzy conditional clause, the Linguistic Value that the former piece of this some " IF-THEN " type dynamic fuzzy conditional clause is the dynamic fuzzy input quantity, the consequent of the dynamic fuzzy conditional clause of " IF-THEN " type is the Linguistic Value that dynamic fuzzy is controlled output quantity.

And given reference input is that of dynamic fuzzy control system 100 controls desired value, the sampling input quantity is the feedback signal that measurement mechanism gathers from controlled device.

Be illustrated in figure 2 the dynamic fuzzy control method of this dynamic fuzzy control system, utilize 100 pairs of controlled devices of the dynamic fuzzy control system shown in Fig. 1 to carry out dynamic fuzzy control, comprise the steps:

(1) dynamic mode gelatinization interface module 102 receives the sampling input quantity of described measurement mechanism, and the departure of described sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtains the dynamic fuzzy input quantity;

Step (1) specifically comprises:

1. dynamic mode gelatinization interface module 102 receives the sampling input quantity that measurement mechanisms 400 detect, and will sample input quantity and given reference input compare, and obtains departure;

2. departure is carried out to change of scale, make departure be mapped to the desired domain scope of dynamic fuzzy control inputs; And the basic domain and the dynamic fuzzy domain that now need to determine departure and control output quantity, and for carrying out the quantizing factor of change of scale;

3. the departure that transforms to the domain scope is carried out to dynamic mode gelatinization processing, make originally departure accurately become the dynamic fuzzy amount, and mean by corresponding fuzzy set, and now need to determine departure and control the dynamic fuzzy Linguistic Value of output quantity and the subordinate function of dynamic fuzzy Linguistic Value.

(2) dynamic fuzzy reasoning module 103, according to described dynamic fuzzy input quantity and dynamic fuzzy control law, is carried out reasoning and decision-making according to dynamic fuzzy implication computing rule and dynamic fuzzy reasoning compositional rule, obtains dynamic fuzzy and controls output quantity;

Step (2) further comprises:

At first, subordinate function according to each Linguistic Value of dynamic fuzzy input quantity, by the former piece of each " IF-THEN " type dynamic fuzzy conditional clause and the dynamic fuzzy vector corresponding to consequent of each " IF-THEN " type dynamic fuzzy conditional clause in the dynamic fuzzy control rule base, contain by dynamic fuzzy the Research on Dynamic Fuzzy Relationship matrix that computing rule obtains every rule, and the Research on Dynamic Fuzzy Relationship matrix that is the dynamic fuzzy control law by each rule does union, obtain total Research on Dynamic Fuzzy Relationship matrix;

Such as: order

for domain

on dynamic fuzzy set,

for domain

on dynamic fuzzy set, suppose that dynamic fuzzy contains

be

upper Research on Dynamic Fuzzy Relationship description, by "

be " and the dynamic fuzzy control law " if be

?

be " dynamic fuzzy set that obtains

may be defined as:

In formula, " o " is the compose operation symbol;

for the dynamic fuzzy implication relation, if there is n bar dynamic fuzzy control law to participate in reasoning in rule base, $(\stackrel{\←}{R},\stackrel{\→}{R})=\underset{i=1}{\overset{n}{\∪}}(\stackrel{\←}{R_i},\stackrel{\→}{R_i});$

Then, according to the dynamic fuzzy reasoning compositional rule, by described dynamic fuzzy input quantity and described Research on Dynamic Fuzzy Relationship matrix, carry out compose operation, obtain dynamic fuzzy and control output quantity;

During with maximum-minimum compose operation

calculating formula be:

(3) sharpening interface module 104 is controlled described dynamic fuzzy that output quantity is clear turns to definite control output quantity, and export described definite control output quantity to topworks 200, by described topworks 200, according to described definite control output quantity, controlled device is controlled;

In Dynamic Fuzzy Logic, clarification method commonly used has average maximum membership degree method, median method and method of weighted mean, and the present embodiment is calculated by the computing method of method of weighted mean, gets the dynamic fuzzy controlled quentity controlled variable

the weighted mean value of degree of membership as its clear value, have:

$(\stackrel{\←}{z_0},\stackrel{\→}{z_0})=\frac{\underset{(\stackrel{\←}{a},\stackrel{\→}{a})}{\overset{(\stackrel{\→}{b},\stackrel{\←}{b})}{\∫}}(\stackrel{\←}{z},\stackrel{\→}{z})(\stackrel{\←}{C^{\′}}\left(\stackrel{\←}{z}\right),\stackrel{\→}{C^{\′}}\left(\stackrel{\→}{z}\right))d(\stackrel{\←}{z},\stackrel{\→}{z})}{\underset{(\stackrel{\←}{a},\stackrel{\→}{a})}{\overset{(\stackrel{\←}{b},\stackrel{\→}{b})}{\∫}}(\stackrel{\←}{C^{\′}}\left(\stackrel{\←}{z}\right),\stackrel{\→}{C^{\′}}\left(\stackrel{\→}{z}\right))d(\stackrel{\←}{z},\stackrel{\→}{z})}$

Be discrete situation for domain, can be write as following calculating formula:

$(\stackrel{\←}{y_0},\stackrel{\→}{y_0})=\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}(\stackrel{\←}{y_i},\stackrel{\→}{y_i})(\stackrel{\←}{B^{\′}}\left(\stackrel{\←}{y_i}\right),\stackrel{\→}{B^{\′}}\left(\stackrel{\→}{y_i}\right))}{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}(\stackrel{\←}{B^{\′}}\left(\stackrel{\←}{y_i}\right),\stackrel{\→}{B^{\′}}\left(\stackrel{\→}{y_i}\right))}$

Wherein, the actual change scope that basic domain is described departure and described control output quantity; The dynamic fuzzy domain is that dynamic fuzzy is controlled desired variation range, comprises the dynamic fuzzy input domain that departure is corresponding and controls dynamic fuzzy output domain corresponding to output quantity; The dynamic fuzzy Linguistic Value is departure and the value of control output quantity on its dynamic fuzzy domain, the incompatible expression of available Dyna Mic Fuzzy Sets; The corresponding dynamic fuzzy set of each dynamic fuzzy Linguistic Value, and dynamic fuzzy set is determined by the subordinate function of dynamic fuzzy Linguistic Value.

The dynamic fuzzy control method of the present embodiment is by adopting the control method based on Dynamic Fuzzy Logic, the departure of the sampling input quantity of measurement mechanism and given reference input is mapped on dynamic fuzzy input domain according to the Dynamic Fuzzy Logic theory, and carry out reasoning according to the dynamic fuzzy control law and obtain dynamic fuzzy and control output quantity, to after its sharpening, controlled device be controlled, overcome the defect that existing control system can not solve the dynamic problem in the Non-linear Dynamic Fuzzy System process control well, thereby can embody better the dynamic fuzzy characteristic of control system, improve the quality of control system.

Application examples:

As shown in Figure 3 to Figure 4, the liquid level that is applied to water tank 300 for the utility model dynamic fuzzy control system 100 is controlled, this control system be applied on water tank 300 comprises dynamic fuzzy control system 100, measurement mechanism 400, topworks 200 and water tank 300, this dynamic fuzzy control system 100 comprises dynamic fuzzy base module 101, dynamic mode gelatinization interface module 102, dynamic fuzzy reasoning module 103 and sharpening interface module 104, these dynamic fuzzy control system 100 1 ends connect measurement mechanism 400, one end connects topworks 200, this topworks 200 comprises the control devices such as PLC or frequency converter, this topworks 200 connects and controls high water tank, and measurement mechanism is measured water tank 300 middle water level height by a liquid level sensor, gather height of water level by this measurement mechanism, dynamic mode gelatinization interface module 102 receives the sampling input quantity of measurement mechanism, and relatively draw the input deviation amount with given reference input, suppose that a certain moment t measurement mechanism is 6dm to the input deviation amount e of dynamic mode gelatinization interface module 102 inputs, dynamic mode gelatinization interface module 102 is mapped to this departure on the dynamic fuzzy domain, for

on the point

press the method for single-point dynamic mode gelatinization, the Dyna Mic Fuzzy Sets after the gelatinization of departure e dynamic mode is combined into

dynamic fuzzy reasoning module 103 can be according to the subordinate function of the dynamic fuzzy Linguistic Value in dynamic fuzzy database 101, each dynamic fuzzy conditional statement former piece in the dynamic fuzzy control rule base and dynamic fuzzy vector corresponding to consequent is done to dynamic fuzzy and contain minimum computing and obtain the Research on Dynamic Fuzzy Relationship matrix

then according to maximum-minimum compositional rule, incite somebody to action

with

the synthetic dynamic fuzzy value that obtains dynamic fuzzy control output quantity, that is:

This dynamic fuzzy is controlled output quantity for a dynamic fuzzy set on the output domain, need carry out sharpening by the sharpening interface module, and the present embodiment is according to average weighted method principle, obtains dynamic fuzzy output controlled quentity controlled variable and is

it is mapped to output domain [20,20] on, output frequency variable quantity u while obtaining corresponding to a certain moment input signal e as+6dm be (2.5) * 20/4=-12.5Hz, requires the frequency minimizing 12.5Hz of variable-frequence governor, and trend less and less is arranged.

Then, the output interface of dynamic fuzzy control system 100 (D/A device) is controlled output quantity u by dynamic fuzzy and is converted to the desired simulating signal of topworks 200, and then controls the water level in water tank 300.

As shown in Figure 4, concrete is applied to for this dynamic fuzzy control system 100 control method that cistern system carries out dynamic fuzzy control, and its concrete steps comprise:

Step 1: dynamic mode gelatinization interface module 102 receives the sampling input quantity of measurement mechanism, and the departure of this sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtains the dynamic fuzzy input quantity;

Concrete, this dynamic fuzzy control system that is applied to the control of cistern system liquid level can design rotating speed change automatically along with the variation of water level that the output frequency that makes frequency converter is motor based on PLC, PLC obtains through interrupting sampling the current water level value that liquid level sensor transmits, the water level value of this water level value and setting is compared and obtains the water level deviation, and this water level departure is carried out to dynamic mode gelatinization processing, the water level departure is mapped on corresponding dynamic fuzzy input domain, obtains corresponding dynamic fuzzy subset; This current water level input quantity of sampling, the water level value of this setting is given reference input, this dynamic fuzzy subset is the dynamic fuzzy input quantity.

Wherein, the basic domain of water level deviation e is by the decision of sampling input quantity, and dynamic fuzzy is controlled the frequency variation decision of the basic domain of output quantity u by the actual output of frequency converter, and the dynamic fuzzy input quantity corresponding input domain

and dynamic fuzzy is controlled output quantity

corresponding output domain

according to the dynamic fuzzy control law, determine; The basic domain of water level deviation and dynamic fuzzy control output quantity is set to respectively to [10 ,+10] and [20 ,+20], and both dynamic fuzzy domains all are defined as discrete limited domain, are defined as respectively 7 grades and 9 grades, that is:

$(\stackrel{\←}{X},\stackrel{\→}{X})=\{(\stackrel{\←}{-3},\stackrel{\→}{-3}),(\stackrel{\←}{-2},\stackrel{\→}{-2}),(\stackrel{\←}{-1},\stackrel{\→}{-1}),(\stackrel{\←}{0},\stackrel{\→}{0}),(\stackrel{\←}{1},\stackrel{\→}{1}),(\stackrel{\←}{2},\stackrel{\→}{2}),(\stackrel{\←}{3},\stackrel{\→}{3})\}$

$(\stackrel{\←}{Y},\stackrel{\→}{Y})=\{(\stackrel{\←}{-4},\stackrel{\→}{-4}),(\stackrel{\←}{-3},\stackrel{\→}{-3}),(\stackrel{\←}{-2},\stackrel{\→}{-2}),(\stackrel{\←}{-1},\stackrel{\→}{-1}),(\stackrel{\←}{0},\stackrel{\→}{0}),(\stackrel{\←}{1},\stackrel{\→}{1}),(\stackrel{\←}{2},\stackrel{\→}{2}),(\stackrel{\←}{3},\stackrel{\→}{3}),(\stackrel{\←}{4},\stackrel{\→}{4})\}$

Deviation

quantizing factor K _e=3/10=0.3, the frequency variation of frequency converter output

scale factor K _u=20/4=5; By this quantizing factor, deviation basic domain [10 ,+10] in numerical value can be mapped to the desired input domain of dynamic fuzzy controller; And, by this scale factor, dynamic fuzzy can be controlled to output quantity

be transformed in the basic domain [20 ,+20] of frequency variation of frequency converter output; Dynamic fuzzy input domain

with the output domain

in each grade corresponding deviation e and control the numerical value of certain limit in the basic domain of output quantity u respectively; In the present embodiment, deviation

by quantizing factor K _etransform to the input domain, K _ebe 0.3; Be specially the input deviation e certain recorded in basic domain constantly by following rule and transform to dynamic fuzzy input domain

in:

If n is the gear number be divided into after continuous deviation discretize in [x, x] scope, form the element of input domain, get n=3 in the utility model, comprise two kinds of situations:

Situation 1: when | K _eduring e|<n, n _jequal | K _e| the round of e;

Situation 2: when | K _eduring e|>=n, n _j=n;

For example, deviation e ∈ [x, x] is+during 6dm, | K _e| e=0.3 * (+6)=1.8, the grade in corresponding dynamic fuzzy input domain "

"; Deviation e is-during 10dm, | K _e| e=0.3 * (6)=-3, the grade in corresponding input domain "

"; And shown in the dynamic mode gelatinization of deviation e processes in the following example: if the deviation e of control inputs amount be-10dm, after the quantizing factor conversion, and adopt the single-point dynamic fuzzy set, it is mapped to dynamic fuzzy input domain, can obtain the dynamic fuzzy input quantity and be

Step 2: determine respectively dynamic fuzzy control inputs amount and the dynamic fuzzy Linguistic Value of dynamic fuzzy control output quantity and the subordinate function of each dynamic fuzzy Linguistic Value;

By dynamic fuzzy control inputs amount

control output quantity with dynamic fuzzy the dynamic fuzzy Linguistic Value of getting all is defined as 5 DF subsets

wherein,

expression " negative large ",

expression " negative little ", expression " zero ",

expression " just little ",

mean " honest ";

The water level deviation

each dynamic fuzzy Linguistic Value, each dynamic fuzzy subset is defined on dynamic fuzzy input domain, because this dynamic fuzzy input domain is discrete domain, to the membership function value of the dynamic mode gelatinization of each Linguistic Value regulation as following table 1:

Table 1 water level deviation

the subordinate function of Linguistic Value

Dynamic fuzzy is controlled output quantity

each dynamic fuzzy Linguistic Value, each dynamic fuzzy subset definition is on dynamic fuzzy output domain, the membership function value of its dynamic mode gelatinization is listed in table 2:

The frequency variation of table 2 frequency converter output

the subordinate function of Linguistic Value

Wherein, in table 1 and table 2, the first row represents respectively the different brackets in dynamic fuzzy input domain and dynamic fuzzy output domain, and first row is the representation level deviation respectively

each dynamic fuzzy Linguistic Value and dynamic fuzzy control output quantity

each dynamic fuzzy Linguistic Value, the center section of form represents the degree of membership of each dynamic fuzzy Linguistic Value corresponding to corresponding domain;

(3) according to the subordinate function of dynamic fuzzy Linguistic Value, dynamic fuzzy control law former piece in the dynamic fuzzy control rule base and dynamic fuzzy vector corresponding to consequent are contained to the Research on Dynamic Fuzzy Relationship matrix that minimum computing obtains every rule by dynamic fuzzy, and the Research on Dynamic Fuzzy Relationship matrix of each rule is done to union, obtain total Research on Dynamic Fuzzy Relationship matrix;

Wherein, according to expert and operating personnel's experience, determine the dynamic fuzzy control rule base that water level is controlled, the present embodiment comprises five rules:

If

be

?

be

If

be

?

be

If

be ?

be

If

be ?

be

If

be

?

be

The Research on Dynamic Fuzzy Relationship that as above five rules form for:

$(\stackrel{\&LeftArrow;}{R},\stackrel{\&RightArrow;}{R})=\underset{i=1}{\overset{5}{\&cup;}}(\stackrel{\&LeftArrow;}{R_i},\stackrel{\&RightArrow;}{R_i})$

In formula,

former piece and the consequent of each conditional statement, i.e. dynamic fuzzy input quantity

control output quantity to dynamic fuzzy

research on Dynamic Fuzzy Relationship; Wherein:

$(\stackrel{\&LeftArrow;}{R_1},\stackrel{\&RightArrow;}{R_1})=(\stackrel{\&LeftArrow;}{A_1},\stackrel{\&RightArrow;}{A_1})\&times;(\stackrel{\&LeftArrow;}{B_5},\stackrel{\&RightArrow;}{B_5})={(\stackrel{\&LeftArrow;}{A_1},\stackrel{\&RightArrow;}{A_1})}^T^(\stackrel{\&LeftArrow;}{B_5},\stackrel{\&RightArrow;}{B_5})$

$=\left[\begin{array}{c}(\stackrel{\&LeftArrow;}{1},\stackrel{\&RightArrow;}{1})\\ (\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\end{array}\right]^\left[\begin{array}{ccccccc}(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})& (\stackrel{\&LeftArrow;}{1},\stackrel{\&RightArrow;}{1})\end{array}\right]$

$=\left[\begin{array}{ccccccc}(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})& (\stackrel{\&LeftArrow;}{1.0},\stackrel{\&RightArrow;}{1.0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})& (\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})\\ (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})& (\stackrel{\&LeftArrow;}{0.0},\stackrel{\&RightArrow;}{0.0})\end{array}\right]$

In like manner, can be by calculating respectively $(\stackrel{\&LeftArrow;}{A_2},\stackrel{\&RightArrow;}{A_2})\&times;(\stackrel{\&LeftArrow;}{B_4},\stackrel{\&RightArrow;}{B_4}),$ $(\stackrel{\&LeftArrow;}{A_3},\stackrel{\&RightArrow;}{A_3})\&times;(\stackrel{\&LeftArrow;}{B_3},\stackrel{\&RightArrow;}{B_3}),$ $(\stackrel{\&LeftArrow;}{A_4},\stackrel{\&RightArrow;}{A_4})\&times;(\stackrel{\&LeftArrow;}{B_2},\stackrel{\&RightArrow;}{B_2}),$ With $(\stackrel{\&LeftArrow;}{A_5},\stackrel{\&RightArrow;}{A_5})\&times;(\stackrel{\&LeftArrow;}{B_1},\stackrel{\&RightArrow;}{B_1}),$ Obtain

with

again above-mentioned each Research on Dynamic Fuzzy Relationship is asked to union, thereby obtain total Research on Dynamic Fuzzy Relationship matrix

(4) according to the dynamic fuzzy reasoning compositional rule, by the dynamic fuzzy input quantity

with the Research on Dynamic Fuzzy Relationship matrix the synthetic dynamic fuzzy that obtains is controlled output quantity;

Now suppose $(\stackrel{\&LeftArrow;}{A^{\&prime;}},\stackrel{\&RightArrow;}{A^{\&prime;}})=(\stackrel{\&LeftArrow;}{A_1},\stackrel{\&RightArrow;}{A_1})=\left[(\stackrel{\&LeftArrow;}{1},\stackrel{\&RightArrow;}{1})(\stackrel{\&LeftArrow;}{0.5},\stackrel{\&RightArrow;}{0.5})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})(\stackrel{\&LeftArrow;}{0},\stackrel{\&RightArrow;}{0})\right]$

According to dynamic fuzzy reasoning maximum-minimum compositional rule, by

with

research on Dynamic Fuzzy Relationship is synthetic to be calculated, and obtains dynamic fuzzy control output quantity to be:

?

(5) dynamic fuzzy is controlled to output quantity and carried out the sharpening processing, the control output quantity that obtains determining, and definite control output quantity exports topworks 200 to by this, the control output quantity definite according to this by topworks 200 controlled controlled device;

Dynamic fuzzy is controlled the control output quantity that output quantity obtains determining after sharpening is processed, processor in dynamic fuzzy control system 100 is passed to PLC again, and to export to frequency converter by PLC be topworks 200, by regulating the output frequency of frequency converter, thereby change motor speed, control the height of water level in water tank;

For example, the dynamic fuzzy obtained in step (4) is controlled output quantity

a dynamic fuzzy set on output controlled quentity controlled variable domain, wherein "

" corresponding output domain in grade "

", according to average maximum membership grade principle, choose controlled quentity controlled variable and be

level, be mapped to output domain [20,20] by it upper, and obtaining output frequency variable quantity u is (4) * 20/4=20Hz, requires the frequency of frequency converter to increase 20Hz, and increasing trend is arranged.

The dynamic fuzzy control method of the present embodiment is passed through based on the dynamic fuzzy control law, the departure of the Water in Water Tank place value that measurement mechanism 400 is gathered and given water level value is mapped on dynamic fuzzy input domain according to the Dynamic Fuzzy Logic theory, and carry out reasoning, calculate the dynamic fuzzy output quantity according to the dynamic fuzzy control law, the definite output quantity of output after sharpening interface module 104 sharpenings, and controlled by 400 pairs of water tank 300 water levels of topworks; Therefore, the dynamic fuzzy control method of the present embodiment does not need to know the mathematical model of controlled device, be easy to realize to thering is the dynamic fuzzy control of uncertain non-linear object, and can reflect preferably the variation tendency of control inputs amount and output quantity, thereby can effectively process the dynamic problem in the Non-linear Dynamic Fuzzy System process control, further improve the quality of control system.

In addition, we it should further be appreciated that, the utility model is not limited with this embodiment, and it is example that PLC or frequency converter not merely be take in the topworks that the utility model relates to, can also be other the control device that can realize accurate control, thereby realize the accurate control to measurand; And the application of this dynamic fuzzy control system also is not limited only to the dynamic fuzzy of the cistern water level of the present embodiment proposition and controls, and can also be applied to other control field, thereby further improve quality and the widespread use thereof of this dynamic fuzzy control system.

Obviously, those skilled in the art can carry out various changes and distortion and not break away from spirit and scope of the present utility model the utility model.Like this, if these modifications of the present utility model and distortion belong in the scope of the utility model claim and equivalent technologies thereof, the utility model also is intended to comprise these changes interior.

Claims (4)

1. a dynamic fuzzy control system, is characterized in that, comprises dynamic fuzzy base module, dynamic mode gelatinization interface module, dynamic fuzzy reasoning module and sharpening interface module; Wherein:

Described dynamic mode gelatinization interface module connects a measurement mechanism, described dynamic mode gelatinization interface module receives the sampling input quantity of described measurement mechanism, the departure of described sampling input quantity and given reference input is mapped on dynamic fuzzy input domain, obtain the dynamic fuzzy input quantity, and described dynamic fuzzy input quantity is input to described dynamic fuzzy reasoning module;

Described dynamic fuzzy reasoning module connects respectively described dynamic mode gelatinization interface module and described sharpening interface module, the dynamic fuzzy control law that described dynamic fuzzy reasoning module provides according to described dynamic fuzzy base module, and described dynamic fuzzy input quantity is carried out to dynamic fuzzy contains and compositional rule of inference calculates, the dynamic fuzzy that obtains corresponding dynamic fuzzy output domain is controlled output quantity, and described dynamic fuzzy control output quantity is inputed to described sharpening interface module;

Described sharpening interface module connects a topworks, by described sharpening interface module, described dynamic fuzzy is controlled to output quantity is clear turns to definite control output quantity, and export described definite control output quantity to described topworks, by described topworks, one controlled device is controlled;

Described dynamic fuzzy base module connects respectively described dynamic mode gelatinization interface module, described dynamic fuzzy reasoning module and described sharpening interface module, rule and relevant various parameters that described dynamic fuzzy base module provides dynamic fuzzy to process for described dynamic mode gelatinization interface module, described dynamic fuzzy reasoning module and described sharpening interface module.

2. dynamic fuzzy control system according to claim 1, it is characterized in that, described dynamic fuzzy base module comprises dynamic fuzzy database and dynamic fuzzy control rule base, and described dynamic fuzzy database comprises with dynamic fuzzy control law and dynamic fuzzy data processes relevant various parameters; Described dynamic fuzzy control rule base comprises some " IF-THEN " type dynamic fuzzy conditional clause, the Linguistic Value that the former piece of described some " IF-THEN " type dynamic fuzzy conditional clause is the dynamic fuzzy input quantity, the consequent of the dynamic fuzzy conditional clause of described some " IF-THEN " type is the Linguistic Value that dynamic fuzzy is controlled output quantity.

3. dynamic fuzzy control system according to claim 1, is characterized in that, described given reference value is that of described dynamic fuzzy control system controls desired value.

4. dynamic fuzzy control system according to claim 1, is characterized in that, described sampling input quantity is the feedback signal that described measurement mechanism gathers from described controlled device.

Images