JPS6295674A - Fuzzy membership s function circuit - Google Patents

Abstract

PURPOSE:To obtain an S function circuit which is optimum to a fuzzy logic circuit which works at the limits of a digital computer and in a current mode. CONSTITUTION:As S function shows the value '0' until the input S reaches a certain level and leads continuously from the value '0' in a fixed slope to reach finally alpha (=1) at a break point SB. As shown in the figure, a current SB related to the break point is delivered from the 1st current source 33 and a limit difference circuit 39 consisting of an FET subtracts an input current S from the current SB. The 2nd current source 36 delivers a current which shows the grade alpha and a subtracting means 37 subtracts the output current SB(-)S of a limit difference circuit 35 from the output current alpha of the source 36. The output 1-(SB(-)S) thus obtained is led to an output terminal 32 via a means which contains an FET 38 and cuts off the negative current.

Description

DETAILED DESCRIPTION OF THE INVENTION A marginal difference 5BeS between the input current S and the value S associated with the break point is calculated and this marginal difference is subtracted from the current α representing a given grade.

That is, α-(SBeS). Negative current is cut by the diode acting element.

Examples using MOS FET are shown in FIG. 9 and FIG. 1O.

Examples using bipolar transistors are shown in FIGS. 1I and 12.

Examples in which the gradient can be switched are shown in FIGS. 17 and 18.

Table of Contents (1) Background of the invention (1,1) Technical field (1,2) Limitations of digital computers and new fuzzy logic circuits operating in current mode (1,3) Concepts of membership function circuits and fuzzy control systems (1st
(Fig. 2) (1, 4) Concept of fuzzy system with learning function (Fig. 3) (2) Outline of the invention (2, 1) Purpose of the invention (2, 2) Structure and effects of the invention (3) Description of Examples (3,1) Various types of membership functions and their definitions (Figure 4) (3,2) Z function circuit (Figure 5.6.7.8) (3,
3) S function circuit (Chapter 9. to, Figure 11.12) (
3, 4) Arbitrary setting of slope during use (14th and 15th
Figure) (3,5) Gradient switching control (No. 15.16.17.18)
Figure) (3,8) Programmable multi-membership function circuit (Figure 19.20.21) (3,7) MIN circuit and MAX circuit (Figure 22.23.2)
4.25°26, 27. (3, 8) Simplified programmable Φ multi-membership function circuit (Fig. 29.30) (3, 9) Expanded programmable Φ multi-membership function circuit (31.32. Figure 33) (3, 10) S-function circuit applicable to crisp sets (Figures 34 and 35) (3, 11) Upgradient function circuit applicable to crisp sets (Figures 36 and 37) (3, 12 ) Programmable multi-dispute membership function circuit applicable to crisp sets (Figure 38) (1) Background of the invention (1,1) Technical field This invention deals with membership function circuits that are essential for constructing a new fuzzy control system. It relates to function circuits, especially S function circuits.

(1,2) Limitations of Digital Computers and New Fuzzy Logic Circuits Operating in Current Mode Fuzzy logic is a logic that deals with fuzziness or "ambiguity."

Ambiguity surrounds human thoughts and actions.

So, if we could quantify and theorize this kind of ambiguity.

It should be possible to apply it to the design of social systems such as traffic control, emergency, and applied medical systems, as well as robots created to imitate humans. Since the concept of fuzzy sets was proposed by A. Zadeh in 1965, research on fuzzy logic has been conducted as a means of dealing with ``ambiguity'' from this perspective. However, much of this research is based on software and software using digital computers.
At present, the focus is on application to systems.

Digital computers perform calculations based on binary logic consisting of 0 and 1, and although the calculation processing is extremely precise, it is necessary to add an A/D conversion circuit to input analog quantities; There is a problem in that when trying to process a huge amount of information, it takes a long time to obtain the final result. Furthermore, programs for applying fuzzy logic must be extremely complex, and complex processing requires a large digital computer, which is not economical.

In the first place, fuzzy logic is a logic that handles continuous values (0, 1) in the interval from 0 to 1, so it has the aspect that it is not compatible with digital computers based on binary logic. Also, fuzzy logic deals with ambiguous quantities with a wide range.

No exactness is required for calculations performed by digital computers. This is why it is desirable to realize new circuits suitable for handling fuzzy logic.

In order to meet such demands, one inventor has already proposed a number of fuzzy logic circuits that operate in current mode (for example, Japanese Patent Application No. 57121/1983). The fuzzy logic circuits proposed by the inventor include a marginal difference circuit, a logical complementary circuit, a marginal sum circuit, a marginal product circuit, an OR (MAX) circuit, and an AND (WIN) circuit.

There are absolute difference circuits, implication circuits, equal circuits, etc., and all of these circuits operate in current mode. All the fuzzy logic circuits mentioned above are composed of one or more marginal difference circuits and addition (
It has the characteristic that it is constructed by a combination of (subtraction) circuits. In current mode, addition and subtraction can be done by simple wiring.
(wired sum or wired subtract), it can be said that all the above-mentioned fuzzy logic circuits basically have a fuzzy marginal difference circuit as their only constituent unit. Therefore, in the circuit design of a fuzzy logic circuit that operates in current mode,
It is also advantageous in many respects in the fabrication of ICs.

(1, 3) Concept of membership function circuit and fuzzy control system (FIGS. 1 and 2) Fuzzy set A is characterized by membership function μ^(X). The membership function μA(x) represents the degree to which the variable X belongs to the fuzzy set A, and this degree is represented by continuous values [0, 11 in the interval from 0 to 1. An example of the membership function μA(X) is shown in FIG.

The membership function circuit is a circuit that outputs a value μA (x) representing the degree to which X belongs to the fuzzy set A when a variable X of a certain value is given as an input.

An example of the concept of a fuzzy control system using fuzzy logic circuits and membership function circuits as described above is shown in FIG.

As an example of the application of fuzzy control, it is considered to automate the control of systems that have traditionally been operated or controlled by humans based on their extensive experience and intuition. The control system that humans have used is extremely complex, but if we simplify it, we can understand it as a combination of several or many empirical rules. This rule of thumb is
If (the state, etc.) is XX, then set ΔΔ (the state, etc.) to 0.'' To make this rule of thumb a little more complicated, roo is XX.

And (or) If OX is ×O, then set ΔΔ to 0.'' It becomes more general. The propositional form of this general empirical rule is called a control law in fuzzy control systems.

According to the usage of the feedback control system, the output e of the controlled system and its deviation Δe are taken as control inputs, and the control output given to the controlled system is taken as ΔU.

In FIG. 2, as an example of the control law, if the control law Ire is a small negative value and Δe is a small positive value, then ΔU
``Make it a small positive value'' is given. This control law 1.

It is expressed as e −N S andΔe−PS−e ΔumPs. Here, NS is a small negative value (nega-ti
ve small), ps is a positive small value (posi
tivesmall), and and means "ka" respectively.

As control law 2, "If e is a small positive value and Δe is a small negative value, then set ΔU to a small negative value". This is expressed as follows.

e −P S and ΔeI=INS → Δu−N
S and some or many other 1lilJ
Rules have been set.

In determining whether "e is a small negative value" in Control Law 1, to what degree can it be said that the given control force e - e o is a small negative value?

The answer to the question is the membership function IA<MS
It is given by the function IA>. Membership function I
A is obtained from a membership function circuit (not shown) and represents the degree to which the control force e belongs to a "set of small negative values." In Figure 2, as the membership function IA,
A triangular function is given that has a peak at a certain negative value of e, and according to this function IA, a certain control input e −
The degree to which eRho-0,2 belongs to this set is 0.8.

Similarly, the membership function IB<MS function IB> representing the degree to which the control force Δe belongs to the "set of small positive values" is shown in FIG. This function IB is also Δ
It has a triangular shape that reaches a peak when e is a certain positive value. According to this membership function IB output from a membership function circuit (not shown), the degree to which a certain control input Δe−Δeosm+o,t belongs to this set is 0.6.

The "and" condition of "e is a small negative value and Δe is a small positive value" in control law 1 is generally calculated by fuzzy logical AND (old N). Specifically, this calculation MIN selects the smaller of the two variables. Therefore, the value of the membership function IA mentioned above is 0
．． 8 and the IB value 0.6, 0.6 is obtained as representing the calculation result of old N.

The command to "make ΔU a small positive value" in control law 1 is also given by the membership function original command 1>. The function representing this original command 1 is also: A triangular shape having a peak value of 1 when ΔU is a certain positive value is shown as an example.

The function representing original command 1 is generated from a membership function generation circuit (path shown).

The "if" in control law 1 is executed, for example, by multiplication. The value 0.6 is obtained by the old N operation described above. When the function of the original command 1 is multiplied by this 0.6, a triangular function command 1> with a peak value of 0.6 is created.

The "if" operation may be performed using the old N. In this case, a trapezoidal function as shown by the broken line will be obtained as the command 1.

Similarly, in control law 2, the given control human power e
By applying this control law 2 to Δe and Δe, a command 2> is created. Other commands may be created as well by application of other control laws.

Generally, a plurality of control laws are set for one controlled system as described above. Each command derived from these control laws is used to finally obtain the control output ΔU. Therefore.

Fuzzy disjunction (W
AX) is calculated. The graph ``Inference Results'' shown in FIG. 2 shows the MAX calculation results for ``Command 1'' and ``Instruction 2''. Among these, the solid line graphs indicate those in which multiplication is used as the "if" condition of each control law, and the broken line graphs indicate those in which the calculation of the old N is performed as the "if" condition.

Using such inference results, the control output ΔU is finally determined. This is defuzzification (def'u
zzlHcation). Each of the above-mentioned operations, including the generation of membership functions, is performed in a state that includes "ambiguity" according to fuzzy logic.
At this stage, the control output Δ has one fixed value.
U must be determined.

Defuzzification can be performed, for example, by taking a weighted average of a function representing the inference result, that is, by determining the position of the center of gravity. In this example, the final control output ΔU−Δu g =
+0 and 1 are determined.

If the old N were performed as an "if" operation, almost the same result would be obtained.

Defuzzification may be performed by first determining the position of the center of gravity of <Command 1> and the position of the center of gravity of Command 2>, and further taking a weighted average of these two positions.

It is preferable that the membership function 1.1 etc. is variable. That is, in the process of continuing to control the controlled system using the control output ΔU determined as described above, it is monitored whether the control is being performed appropriately. If optimal control is not being performed, the membership function (its value or graph shape) is changed in order to find a membership function that allows optimal control. This is generally called the "learning function."

(1, 4) Concept of a fuzzy system with a learning function (Figure 3) Figure 3 shows a fuzzy system with a learning function as described above.
1 schematically shows an example of a system;

Some physical input, such as the above-mentioned control input or key input data, is normalized as necessary by the input conversion circuit 11 or converted into a signal in an appropriate form. This conversion circuit 11 may be unnecessary in some cases.

The membership function circuit group 12 is provided with a large number of membership function circuits with variable parameters, and one or more predetermined ones are selected according to the input signal from the conversion circuit 11, and one or more predetermined ones are selected according to the input signal from the conversion circuit 11. A signal representing the membership function is output.

On the other hand, a circuit 15 is provided for generating one or more membership functions. These circuits 12 and 15
The membership function output from fuzzy logic network 1
3, where calculations are performed according to predetermined fuzzy logic, and the calculation results are output. It is preferable that the logic of this circuit network 13 and the parameters of membership function generation circuit 15 can also be changed as necessary.

The fuzzy information output from the fuzzy logic circuit network 13 may be output as is, but in some cases, some kind of decision is made by the above-mentioned defuzzification circuit 14, and this becomes the output.

This output may be used for various purposes, such as being displayed or as the control output ΔU described above.

The output of the fuzzy logic network 13 or defuzzification circuit 14 is compared to a reference (reference, standard) input. This reference input represents the correct answer for learning,
For example, it may be provided by a trained expert, digital-computer simulation, etc.

The control and storage circuit 16 responds to the above comparison result.

In order to make the deviation zero, the shape and parameters of each membership function of the membership function circuit group 12 and the membership function generation circuit 15 are changed, and the types and connections of the logic functions in the fuzzy logic circuit network 13 are changed. or change it.

In this way, by learning, this fuzzy system is adjusted and changed so that it always generates the correct output (correct answer).

(2) Summary of the Invention (2,1) Purpose of the Invention The purpose of the present invention is to provide a circuit for obtaining the mempage knob function used in the systems described in <1.3) and (1,4) above. Another object of the present invention is to provide a membership function circuit, particularly an S function circuit, which is suitable for the fuzzy logic circuit operating in current mode as described in (1, 2) above.

(2, 2) Structure and effect of the invention The S-function circuit according to the invention includes a first current source that outputs a current SB having a value related to a break point, and an input current S from the output current SB of the first current source. The marginal difference to subtract 5
a limit difference circuit that calculates BeS; a second current source that outputs a current representing a predetermined grade α in fuzzy logic;
From the output current α of the second current source to the output current s of the limit difference circuit
The present invention is characterized in that it includes a subtraction circuit for subtracting n e s, and a diode operating element for cutting off a negative current of the output current of the subtraction circuit.

The predetermined grade α is generally 1. If a multi-output current mirror is used as the current mirror of the limit difference circuit, the output lines of this multi-output current mirror are connected in parallel, and at least one output line is provided with a switching element, the slope of the S function can be switched. .

An S function is such a function that a single force S shows a value of 0 until it reaches a certain value, rises at a constant slope from the above certain value, and finally becomes α (-1) at the break point Sts. It is.

According to this invention, as described above, an output current representing a value of 1-(SBθS) (the negative direction current is 0) is obtained, so this output current represents an S function. Moreover, this S
Functional circuits operate in current mode.

As will become clear in the description of the embodiments, the S-function circuit is a basic circuit for generating many fuzzy membership functions, and is an extremely useful circuit.

Examples of the present invention will be described in detail below.

In the following explanation of the embodiment, we will first clarify the various types of membership functions (Fig. 4), explain two-function circuits related to S-function circuits (Figs. 5 to 8), and then explain this The S-function circuit, which is an embodiment of the invention, will be explained (FIGS. 9 to 12), and furthermore, the developed form of the S-function circuit and its application examples will be developed.

(3) Description of Examples (a, t) Various types of membership functions and their definitions (Figure 4) Membership functions are generally used, an example of which is shown in Figure 1 (A). It is often expressed as a single curve as shown in the figure below. However, whether or not it should be expressed as a single curve is not essential for the membership function. A more important feature of the membership function is that it takes continuous values from θ to 1.

On the other hand, from the viewpoint of two-circuit design, it is easier to represent the membership function as a straight broken line as shown in Figure 1 (B), and the membership function can be expressed with a small number of parameters. can be characterized, which makes the design even easier.

Moreover, even if the membership function is represented by a broken line, the above characteristics are not lost.

Therefore, in the following description, all memberships will be expressed by straight lines or broken lines.

The membership function shown in FIG. 1(B) is only an example. There are many other types of membership. These definitions will be explained below.

Figure 4 shows 10 types of membership functions.

The first one is a function that always takes a value of 0 regardless of the value of the variable X, and this is defined as the φ function.

The second one is defined as a single function that always takes a value of 1.

The third one takes a value of 1 in the region where the variable X is small,
When it reaches a certain value zB, it decreases at a constant slope and finally reaches 0, and in the region where X is larger than that, it is a function that always takes the value of O. That is, the variable has one downward slope on the X axis. This is called the 2 function. x-2B is called the break point. The gradient can take any value. The l function can be defined by a break point ZB and a slope. Even if Z −0, ZB<0.

Include this in the l function.

The fourth one is the inverse of the two functions, and is defined as the S function. That is, there is one upward slope on the X axis. The S-function is also defined by a break point SB and a slope.

The fifth one is called the π function, which takes a value of 1 when the variable X is in a certain region, and decreases at a constant slope when Finally, it is a function that takes the value of O and is always 0 in the region where X is smaller and larger than that. It can also be called a trapezoidal function. The π function is characterized by two break points S, Z and a slope B2.

In a special case, it will be 51132''ZB2, and will have a triangular shape as shown by the chain line.

The sixth one is the inverse of the π function or is defined as a function. It can also be called a function with one valley. U
The function is defined by two breakpoints z, S and the slope BI Bl . In special cases, the shape shown by the chain line will be used (ZB□-8B
l) O The shape of the membership function becomes more complex.

The seventh one is a combination of a trapezoidal function (π function) and a function (S function) with an upward slope in a region where X is larger than the trapezoidal function (π function), and is defined as an N function. From a different perspective, this can be seen as a function with a valley (U function).
It can also be said that it is a combination of upward slope functions (S functions) in a small area of . In any case, this N
The function has three break points S, Z, S
2 B2 81 and the slope.

The eighth one is the inverse of the N function and is defined as the l function. This is also three break points Z
, S , Z and gradient BI B2
82.

The ninth one is called the W function, and it can be said to be a combination of two functions with valleys (U functions), or a function with a downward slope (L functions) with two trapezoidal functions (π functions). It can also be said to be a combination of a function) with an upward slope (S function), or a combination of an N function and an l function, or a combination of an l function and an S function. In any case, the W function has four breaks and
Points z, s, z, s and slope [
31B2 82 Bl.

The last one is the inverse of the W function and is defined as the 1M function. This also has four break points s, z
, s , z and gradient by BI B
2 82 Bl stipulated.

Furthermore, by appropriately combining two or more of the above functions,
It is easy to see that more complex membership functions can also be defined.

In FIG. 4, only the positive region of the variable X is shown, but it goes without saying that it can be extended to the negative region of X as well. In this case, the above-mentioned break points can also generally take on negative values.

Uphill slope, downward slope 2 The slopes of trapezoids, valleys, etc. can be arbitrarily set, but in the design of 2 circuits, the slope can be set to 1 (or -1).
) is the simplest.

As will be described later, even if the slope is 1, any slope can be obtained by changing the ranges of the vertical and horizontal axes when using the l1illil path. By predetermining the slope, the above-mentioned IO function can be uniquely defined by only one or an arbitrary break point.

(3, 2) l-function circuit (FIGS. 5, 6, 7, and 8) FIG. 5 shows an example of a membership function circuit that outputs the l-function. Here, the input variable is Z and the l function is fz
It is expressed as. Furthermore, this circuit operates in current mode and is a sink input and source output circuit. A sink input is a form in which an input current flows into a circuit, and a source output is a form in which an output current flows out from a circuit.

In current mode, the positive and negative values of variables and functions are represented by the direction of the current, and their absolute values are represented by the current value.

The membership l function circuit of FIG. It consists of a current source (which provides a sinking input current to the circuit) and a diode 28.The current mirror 25 consists of two N-MOS PETs.
It is made up of. A graph representing the current flowing through each portion of the circuit of FIG. 5 is shown corresponding to the arrows indicating the direction of the current. Further, a graph of the output current f2 is shown in FIG.

A current representing the value of an input variable Z (where Z≧0) flows into the input terminal 21 . Input terminal 21 and current mirror 2
A current source 23 is connected between the input side of 5 and the wire 0R24, and the value ZB
(ZB≧O) current flows out. therefore,
A current representing the difference between 2 and Z (Z-ZB) attempts to flow from the wire 0R24 toward the current mirror 25, but in reality, the current mirror 25 acts as a current blocking diode against reverse current, so the limit difference is A current of (ZOZB) will flow (see graph). Here, θ represents the calculation of the fuzzy marginal difference, and the marginal difference has the following content.

(1) The same value of sink current is output from the output side of the current mirror 25. A current source 26 is connected between the output side of the current mirror 25 and the output terminal 22 by a wire 0R27. Therefore, in wired 0R27, 1-
(ZOZB) is calculated.

A current of this value is about to be drained or sucked from the output terminal 22 (see graph). however,
Since a diode 28 is connected between the wire 0R27 and the output terminal 22 in a forward direction with respect to the output, the sink output current that is about to appear at the terminal 22 becomes zero. This is equivalent to the operation of 1θ(Z(iZB)).

The above operations can be summarized as follows.

(2) This operation is shown in a graph in FIG. The downward slope of these two functions is -1.

Note that the diode 28 can be replaced with a diode-connected MO8PET.

When input current Z is negative (however, ZI3≧0).

From the current mirror 25 to the wired 0R24 (Z
+ZB) tries to flow, but the current mirror 25 prevents this current from flowing out, so the current flowing between the current mirror 25 and the wire 0R24 is zero. Therefore, the output current of the current mirror is also 0, and the output terminal 22
The current having a value of 1 from the current source 26 is discharged as is.

If break point Zn is negative (2≧0), wired 0R24 to current mirror 24 < z
Since a current of + + z n l ) flows in, the resulting output current of the current mirror 25 also becomes (Z+lZB l). Therefore, the output can be expressed as:

(3) Equation (3) represents the operation of shifting the graph of FIG. 6 to the left so that zB is on the negative side.

When break point ZB and input terminal Z are both negative, a current force of (IZBlθIZ1) flows from wired 0R24 toward current mirror 25. Therefore, the sink output power of the current mirror 25 is also (IZB
lθIZ1), and the discharge output current is expressed by the following equation.

...(4) Equation (4) also holds true. This represents a state in which the graph in FIG. 6 is shifted to the left.

In this way, the circuit of FIG. 5 is applicable to all values of Z and values of ZB.

Figure 7 shows a bipolar transistor array (ROI
This figure shows a Z-function circuit realized using a TA78 (manufactured by IM). Circuits corresponding to the current source, current mirror, etc. in FIG. 5 are given the same reference numerals.

Further, an input circuit 21A is provided in place of the input terminal 21 in FIG. 5, and an output circuit 22A is provided in place of the output terminal 22. As the diode 28, a diode between the base and emitter of an NPN transistor (one of the TA78) is used.

FIG. 8 shows experimental results in which 71pj was determined using the circuit shown in FIG. Experiments were conducted for three different ZB (parameters). Input terminal Z.

The break point current Z, the unity value current and the output current f2 were measured as the voltage drop across the resistor in the respective circuit. fz-10μA becomes μm1, f
Z-0 μA corresponds to μm0, respectively.

As can be seen from this graph, the circuit shown in FIG. 7 has extremely excellent linearity and also has a simple nine-circuit configuration. Such excellent linearity cannot be achieved with simple voltage-mode circuits, and this is also a major reason why membership function circuits were realized with current-mode circuits. Also.

The circuit shown in FIG. 7 uses a current mirror, so it has good temperature stability, and since no resistors are used except for the current source, it is suitable for integration.

Also, as can be seen from FIGS. 7 and 8.

A highly practical two-function circuit can be realized not only by MOS PET but also by bipolar elements.

(3, 3) S function circuit (Fig. 9, Fig. 10, Fig. 11).

FIG. 12) An example of the membership S function circuit is shown in FIG. The input variable (input current) is indicated by S, and the S function output (output current) is indicated by f3. The current sB representing the break point is provided by the current source 33, and the current representing the value 1 is provided by the current source 3G.

What is the basic difference between S function circuit and Z function circuit?

It is in the direction of the current that is manually applied to wired 0R34 (corresponding to wired 0R24 in Figure 5). This wired 0R
34 is provided with the input current S as a source input and the break point current Sn as a sink input. For this reason, the direction of the sinking input current applied to the input terminal 3I is reversed by the current mirror 39. In addition, the break point current source 33 provides suction power to the circuit (fifth
(Compare with the current alX23 in the figure).

SB□s by wired 0R34 and current mirror 35
calculations are performed. Furthermore, wired OR371:
Yotte 1 (S aeS ) operation is skipped. Diode-connected MO3P acting as a diode
Since the current in the sink and output direction is blocked by ET 38, the output current is f s ”” 1θ(SB
□ A discharge output current representing S) is obtained. A graph of this output current is shown in FIG.

In this S function circuit, it is also possible to set the break point Sn to a negative value.

When SB is 0, the output current f always takes a value of 1 in the region where S≧0, so there is no particular meaning in setting SB to be negative.

It is sufficient to set S, -0.

An S-function circuit realized using bipolar transistors is shown in FIG. In this figure as well, circuits having the same functions as those shown in FIG. 9 are given the same reference numerals. Reference numeral 31A is an input circuit corresponding to the input terminal 31, and reference numeral 32A is an output circuit corresponding to the output terminal 32. The all-defined characteristics of the circuit of FIG. 11 (with SB as a parameter) are shown in FIG. It can be seen that this S-function circuit also has an excellent straight line.

(3, 4) Arbitrary setting of slope during use (Fig. 13).

Figure 14) As shown in Figure 3 for conversion circuit ■1, in discussions of membership functions, the input value of a physical quantity is normally normalized using its maximum value (or allowable value of the circuit). and the normalized value is used as the input value. For example, when dealing with height H, the maximum value (for example, 2 m) H is used to calculate the height manpower as H/H, which is the normal max.
It will be converted to a+ax.

As an example, the membership function μ811 of the set "tall people" is shown in FIG. 13(A) as an S function, and the membership function μZll of the set "short people" is shown in FIG.
) are respectively shown as Z functions.

The horizontal axis (variable) of these membership functions is S-H/
H, expressed as Z-H/H wax
It's waxed.

Therefore, on the 9 circuits, depending on how many μA the maximum value H corresponds to aX and how many μA the grade 1 of the function corresponds to, the effective slope of the membership function,
That is, it is possible to set the upward slope of the S function and the downward slope of the Z function to arbitrary values. In the above-described two-function circuit and S-function circuit using the current mirror, the slope of (output current)/(input current) is always -1 or 1.

Depending on how you use it, you can obtain any gradient.

An example of substantially changing the slope is shown in FIG. 14 using a Z function. FIG. 14(A) shows the membership function of the set "short people" when H corresponds to +++aX 100 μA and grade l corresponds to 1QpA. When we want the gradient to be 1/2 of such a membership function.

As shown in FIG. 14(B), H may be made to correspond to 50 μA as aX. Also, if you want to reduce the slope to 1/4,
As shown in FIG. 14 CC'), it is sufficient to make H correspond to 25 μA (wax).

As described above, even if the gradient of the membership function generation circuit described above is fixed to +1 or -1,
It turns out that you can set any slope depending on how you use it.

(3,5) Gradient switching control (Fig. 15, Fig. te, Fig. 1)
Figure 7.

(FIG. 18) It will be explained next that it is possible to change the slope of the membership function on the circuit configuration.

FIG. 15 shows the current source 23 in the two-function circuit shown in FIG. A configuration is shown in which the wire 0R24 and the current mirror 25 are taken out and the current mirror 25 is modified to form a current mirror 25A.

The current mirror 25A includes a current mirror 41 having two output drains having the same area, and an N-MOS for switching the parallel connection of these two output drains.
PET 42. The FET 42 is turned on and off by a control signal Vc applied to a control terminal 43.

A graph of the output signal ZθZn of this current mirror 25A is shown in FIG. Control signal V. When the current mirror 25 is set to L level, the PET 42 is off, so the current mirror 25
The slope of the output current of A is unity. In this case, current mirror 25A has the same function as current mirror 25 shown in FIG. Control signal V. When set to H level, PIE
T42 turns on and current flows to the two output drains, resulting in twice the output current, so the slope is 2.

Therefore, if such current mirror 25A is used in place of current mirror z5 in FIG. 5, a Z-function circuit whose slope can be switched depending on the level of control signal VC is realized. When the slope becomes 2, the output characteristics of the Z-function circuit are shown in broken lines in Figure 6.

It is possible to switch between any number of slopes without being limited to two types of slopes. FIG. 17 shows a part of the S-function circuit, in which the current mirror 35 of FIG. 9 is used as a current mirror 35A to select grass material. current mirror 35
At A, the current mirror 44 has three output drains, these output drains are connected in parallel, and two of them are connected to )'ET 45.46 as switching elements. FET 45
．． 4B is controlled to be turned on or off by the control signals v and 2 given to their control terminals 47 and 48, respectively.

As shown in the figure, when both PE745.4[i are off (Vcl-VO2-L), the slope of the output current is -1, and when either one is on, it is (V
-H, V -L or v -L, V C2-H
)CI C2C1 Gradient is -21 When both are turned on, (VCl""C2
``H) The slope will be -3.

(3, 6) Programmable multi-membership function circuit (Figures 19, 20, 21) 10
The 19th multi-membership function circuit allows you to freely program (or control from the outside) 9 of the 9 fuzzy membership functions, excluding the M function.
As shown in the figure. This function circuit includes a multi-fanout circuit 50, a first Z function circuit (No, l) 51.
Second Z function circuit (No, 2) 52. First S function circuit (No, l) 53° Second S function circuit (No, 2)
54. Old N (fuzzy AND) circuit 55 and MAX
(fuzzy OR) circuit 56. The variable (input) is X, and the final function (output) is f
It is given by X.

The multi-fanout circuit 50 generates multiple (four in this case) currents X having the same value and the same direction from one input current X, and an example of its specific configuration is the 20th one. As shown in the figure. This circuit is connected to a current mirror 58 for reversing the direction of the input current, and to the output side of this current mirror 58, and generates multiple (four) output currents having the same value as the input current but in the opposite direction. Multiple outputs (multi-drain)? ! It is composed of a flow mirror 59.

The four output currents S function circuit 53, 5
4 is entered. The Z-function circuits 51, 52 are each the same as shown in FIG. 5, their break points are denoted by z and Z, and the output current old B2 is denoted by f and f, respectively. 5Z
The XI ZX2 function circuits 53 and 54 are respectively the same as shown in FIG. 9, their break point is SBl'S, and the output currents are f and f, respectively, as shown in Table B2.
SXI SX2 is currently available. Therefore, the slope is now 1°-1.

The output f of the second Z function circuit 52 and the output f of the second X2S function circuit 54 are given to the old N circuit 55.
2. As shown in FIG. 21(A), the break points of these circuits 52,54 are S82≦ZB2
If the following conditions are satisfied, the old N calculation results of the outputs of these circuits 52 and 54 become a function on a trapezoid, that is, a π function. This π function (output of the old N circuit 55) is represented by f. The old N operation calculates the smallest value (smaller value) of multiple input values (π x 2 manual values)
This is because it is an operation for selecting .

The output f of the old N circuit 55, the output f of the first Z function πX number circuit 51, and the first S function circuit 53X
The output f of I is given to the MAX circuit 5G. MAX is an operation for selecting the largest value of multiple human power values. The current value corresponding to grade 1 of the function is I. shall be. 21st
Referring again to figure (A), Z8, +21°≦S, Z
2 B2 than satisfying the condition of ≦S -210
Bl If you select these breakpoints.

The output of MAX circuit 56 represents the W function.

The current mirror (fifth
Reference numeral 25 in the figure. It is possible to replace numeral 35 in FIG. 9 with a current mirror with switchable slope (such as current mirror '25A in FIG. 15). The control signals given to the control terminals in this case are shown in FIG.
v given by ZI Z2 St'
S2. By setting the levels of these control signals, it is possible to independently set any of the four gradients of the W function to a value other than 1, as shown in FIG. 21(B), for example. Fig. 21 (B) Vz t -V S 2-H, V
The state where Z 2-V s t ”” is set to L is shown. It goes without saying that gradient switching is also possible with any of the functions described below.

Next, it will be shown that the circuit of FIG. 19 can realize nine fuzzy membership functions depending on the setting of break point values. The discussion will proceed with reference to FIG. 4 and FIG. 21(A).

In addition, in the following explanation, Hl is the maximum value of the input current and the above-mentioned I
o (for example, 10 μA) plus the value ([maximum input current value] + Io).
Ha-■. This means to set it to the following value. D, C, don't care
ICare), that is, any value is acceptable.

The conditions for the circuit shown in FIG. 19 to realize each of the nine function circuits are as follows.

φ function Z-L, S-H, 5B2-H,.

Bl　　l　　BI　　I Z　B　2　””　D　-C.

or.

Z “L, S””H, ZBl””Ll
.

BI I BI I S　B2-　D,C.

1 function Z −H, others (i.e. ZS, 5) Bl l B2' BI
B2 beam, C.

(Here, it is sufficient for ZB□ to be larger than the maximum input terminal value, but in order to avoid increasing the types of control signals, we set ZBl = Hl as a sufficient condition.) Or, S-L, other beams, C .

BI I (S Bt should be less than OA, but it was set as 5Bl=Ll in order to suppress the increase in the types of control signals.) Or S -L, Z -H, other beams, C.

(Same as above, 5I32 only needs to be below OA, and z, 2 needs to be above the maximum input current.) Z functions S -H, S -H, ZBl-D, C.

Bl l 82 1 (In this case, ZBl becomes the break point.) Or S -H, Z -L, S -D, C
.

Old I 82 1 112 (ZBl is the break point in this case as well.) Or S -H, S -L, Z -LBI
I B2 1 BI
I (in this case, "82" is the break point.

Further, S82 may be less than or equal to OA. )S function Z-L, Z-L, S-D, C.

Old I B2 1 82 (in this case "" II is the break point) or Z -L, S -H, Z -D, C
.

Bl l B2 1
B2 (5I31 is the break point in this case as well.) Or Z -L, S -H, Z -HBL
I BI I 82
1 (SB2 is the break point in this case. SB3 only needs to have a value larger than the maximum input current value.) π function Z −L , S −H, S ≦2BI I
Bl l B2 82 (Break points are SB2 and Zn2.

In the case of S82'' ZI32, it becomes triangular as shown by the chain line in Fig. 4.) U function S -H, Z -D, C, ZB1+IO≦5Bl-
B2 1 B2 O (Break points are ZBl and SBI.

In the case of ZBl+l0-881-■o, the shape is shown by the chain line in FIG. ) or Z-L, S-D, C, Z
Bl+l0B2 1 B2 ≦5B1-1O N function Z-L, sB2≦ZB2≦5Bl-210l l (Break points are S, Z, S and 2 B2 Bl.) l function S -H, Z +2I ≦S ≦ ZBl l
BI O82B2 (Break points are Z, S, Z and I B2 B2.) W function ZBl+2Io≦SB2≦zB2≦5B1-2IO (as described above) Indicated by reference numeral 55 in FIG. It is easy to understand that by replacing the circuit with the MAX circuit and replacing the circuit 56 with the old N circuit, nine of the ten functions shown in FIG. 4, excluding the W function, can be realized.

(3,7) MIN circuit and WAX @ path (Fig. 22, Fig. 23, Fig. 24, Fig. 25, Fig. 26, Fig. 27, Fig. 28) Programmable multi-membership function of ffi19 Fig. Details of the old N (fuzzy AND) circuit and MAX (fuzzy OR) circuit used in the circuit can be found in the applicant's application (for example, Japanese Patent Application No. 59-598).
7121), but I will briefly explain it here.

The old N operation is defined as follows...(5) Here, μ and μY represent membership functions, respectively.

A circuit in which the MIN circuit is realized using MOS PET is shown in FIG. For convenience, the input terminals are expressed as Y by μ and μ, and the output current (old N calculation result) is given by μ2.

The direction of the input current μ is reversed by a current mirror 61. The input current μY is input to a multi-fanout circuit consisting of current mirrors 66 and 87, which produces two currents μY of equal value.

A source input current μX and a sink input current μY are given to the wired 0R62, and the wired 0RG2
is connected to the current mirror 63.

The current mirror 63 also acts as a diode, and the wired 0R 62 and the current mirror 63 constitute a fuzzy limit difference circuit. Therefore, the sink output current of the current mirror 63 is given by the following equation.

(6) Similarly, a limit difference circuit is configured by the wire 0R64 and the diode 65, and the output current of this MIN circuit is given by the following equation.

...(7) Equation (7) is the same as Equation (5).

An example in which the old N circuit is constructed using bipolar transistors is shown in FIG. From a comparison with the circuit of FIG. 22, it can be easily understood that the circuit of FIG. 23 performs the old N operation.

FIG. 24 shows 1llllJ constant results of the input/output characteristics of the circuit of FIG. 23. On the other hand, human power μ oath is used as a parameter. In the circuit shown in FIG. 23, TM01 was used as the PNP transistor, and TM78 was used as the NPN transistor.

In FIG. 19, the MAX circuit 5B has three manpowers. In general, a two-person MAX circuit can be easily configured. How to configure a 3-person MAX circuit.

As shown in Fig. 25, two-man powered MAX circuit 5
Just connect 6A and 5GB in two stages.

FIG. 26 shows an example in which a two-manpower (7) MAX circuit (56A muff', = 4; t 5BB ) is constructed using MOS r'ET.

The fuzzy MAX operation is defined by the following equation.

...(8) The input current μY is input to a two-output power mirror 71.

This generates two currents μY whose directions are opposite to those of the input terminal, one of which is applied to the wired 0R72, and the other whose direction is reversed again by the current mirror 75 and which is connected to the wired 0R.
74.

The input current μX is also manually input to the wired 0R72.

The wire 0R72 and the diode 73 constitute a limit difference circuit, and the diode 73 outputs a current given by the following equation, which flows to the wire 0R74.

(9) In the wired 0R74, the current μY is added to this current μxeμ, so the output type current μZ is as follows.

...(10) Equation (10) expresses the same content as Equation (8).

FIG. 27 shows an example in which the MAX lff1 path is constructed of bipolar transistors. In FIG.

Components corresponding to those shown in FIG. 26 are indicated by the same reference numerals with the letter A added thereto. The circuit of FIG. 27 does not completely correspond to the circuit of FIG. 26. The two current mirrors 71, 75 in FIG. 26 are replaced by three current mirrors 76, 7 in FIG.
7,78.

When a multi-output current mirror is constructed using bipolar transistors with multiple collectors.

When at least one output collector is opened, saturation occurs in that collector, causing an error in the output current of the other output collectors. In order to avoid saturation of the collector of the multi-output current mirror in any case, it is necessary to ensure a certain amount of collector-emitter voltage.

The circuit of FIG. 27 prevents collector saturation by connecting a circuit with low input resistance, such as current mirror 78, to the collector of multi-output current mirror 77. Regarding measures to avoid collector saturation in a multi-output current mirror, the applicant's patent application, Japanese Patent Application No. 59-2
6338 [i.

FIG. 28 shows an example of the measurement results of the input/output characteristics of the MAX circuit shown in FIG. 27 using μY as a parameter.

(3, 8) Simplified programmable multi-membership function circuit (Figures 29 and 30) Figure 29 shows a simplified programmable multi-membership function circuit based on the S function circuit. It shows. Here, a P-MOS FIET is used.

Therefore, the direction of the current is opposite to that of the S function circuit shown in FIG. Also, the input current is X and the output current is Z
It is shown in

Multi-output current mirror 81 generates from one input current X three currents X of the same value and opposite direction. These currents X, serve as input terminals for the three circuits described below.

The first S-function circuit includes a wired OR 84, a current mirror 85 . Wired 0R87 and diode connected MO8
It is composed of PET 88. In comparison with FIG. 9, these elements are wired OR34, current mirror 35
．． Wired 0R37 and diode connected MO9PE
738 respectively. A source input current with a value of Xt + 1 is given to the wired 0R84 as a break point. The operation of this first S-function circuit can be easily understood from the comparison with FIG. 9 and from the graph shown corresponding to the arrows indicating the direction of the current in FIG.

, the second S function circuit includes a wired OR 94, a current mirror 95 . It consists of a wired 0R97 and a current mirror 98. The current mirror 98 has the function of reversing the direction of current as well as the function of a diode.

The break point is x2 for convenience of explanation.

It is assumed that the condition of X2-1≧X1+1 is satisfied.

Furthermore, the break point x a (x a≧X
2) is provided with a circuit that generates a function having an upward slope (the slope is 1) (hereinafter referred to as an upward slope function), and this circuit is composed of a wired 0R92 and a diode-connected MO3PET 93. .

A source input current with a value of x3 is given to the wired 0R92.

The output current of this upward slope function circuit is wired 0R9
11i, it is input to the second S function circuit. In this wired 0R96, the output current of the up-slope function circuit is subtracted, and the reverse current is blocked by the current mirror 98, so the output j current of the current mirror 98 represents a π function (break point x, x
).

The current representing this π function is input to the first S function circuit at the wired 0R86 and is subtracted from the current flowing therethrough. Therefore, the output current Z is as if the π function was subtracted from the S function, and this represents an N function.

In the circuit of Figure 29, the diode-connected MO3FET
99 and 89 have been added. These PUTs work as follows. That is, between the current mirror 81 and the source/drain of the diode-connected MOS PET 93,
Current mirror 98 and diode connected MO3PET 99
The threshold voltage between the source and gate of is added.

Allow these to operate normally. In addition, the voltage between the sources and drains of the two diode-connected MO3PETs 88 and 89 (that is, the sum of these threshold values) is applied between the sources and drains of the diode-connected MO3PET 99 and the current mirror 98, and normal operation is ensured. is possible.

The circuit of FIG. 29 uses one function out of the ten functions mentioned above, excluding the l function, the W function, and the M function.

This can be achieved as follows.

φ function X IHI, X2. X3MmD, c.

(H, means setting to [maximum input current] + 1°. ■ is the current value corresponding to grade 1. φ
In the case of a function, it is sufficient if x1≧[maximum input current]. ) or x −L, x = H, xl−
D.C.

2 I 3 1 (L means set to −■. In the case of the φ function, it is sufficient that x SO is sufficient. Also, it is sufficient that x3≧[maximum input current]) 1 function x lL , x2 Adventure x3 Ruri, c.

Or x = L, x - L, ,
x2-D,C.

l 1 3 1 l function x-Ll, x3-Hl ■ (X a≧[maximum input current]).

X2-1 becomes the break point. ) S function x −H−x a −D, C− (x l+ 1 is the break point.) Or x −L , x −L 112+ (X 2≦0 is sufficient. X3+1 is the break point. ) π function wH (It is sufficient if X a≧[maximum input terminal].

x + l, X 2 1 is the break point. )■ U function X1″″L! (X and x3 are break points.) N function The above-mentioned conditions, ie, x +2≦x2≦x3+2 ■ The circuit in FIG. 29 is based on the S function circuit.

A simplified programmable multi-membership function circuit can also be realized by using an l-function circuit as the basis. In other words, 2 with values as shown in FIG. 30(A) and with Xt as the break point.
A function circuit is provided in place of the first S function circuit described above. From this Z function, π as shown in Figure 30 (B)
By subtracting the function, the A function output is obtained as shown in FIG. 30(C). However, X ≦x S x t
1 is the condition.

In such a circuit, X, X2 . It is easy to understand that by changing the conditions of X3, seven types of functions, excluding the N function, W function, and M function, can be realized among the ten functions described above.

(3,9) Expanded programmable multi-membership function circuit (FIGS. 31, 32, and 33) FIG. 31 is an expanded version of the membership function circuit shown in FIG. 29. Expansion has two meanings. One of them is that two types of grades α and β are provided. In all the circuits described above, the maximum grade is always fixed at 1, but values α and β that are variable between 1 and 0 are prepared as new grade parameters. The other point is that, as can be seen from the graph of output current Z in FIG. 31, a new membership function form, which can be called a modification of the M type, was created with the introduction of a new grade parameter.

In FIG. 31, the same elements as those shown in FIG. 29 are indicated by the same reference numerals with the letter "A" appended thereto. Hereinafter, only the points different from those shown in FIG. 29 will be explained.

The multi-output current mirror 81A generates four input currents X.

In the first S-function circuit, the wired 0R84A is provided with a source input terminal of the value x1. The wired 0R87A is provided with an input terminal for outputting the value of α.

Two wired 0R87A and 88 of the first S function circuit
A newly provided wire 0R89 is provided between A and A, into which the output current of the newly provided upward slope function circuit (first upward slope function circuit) flows. This first up-slope function circuit consists of a wired 0R82 and a diode-connected MO8FET, 83, and its break point is X4.

Therefore, the first π function (break point
．． X4. The grade α) is generated.

In the second S-function circuit, the wired 0R94A
is given a source input current of X2+β, and the wired 0R97A is given a source input current of β.

The wired 0R92A of the up slope function circuit (second up slope function circuit) attached to this S function circuit has X3-β
A source input current of is given. current mirror 99
is for inverting the output input of β to the intake input.

It goes without saying that the three input currents of value β applied to wired 0R94A, 97A and 92A can be generated by a multi-output current mirror (path shown).

By the second S-function circuit and the second upslope circuit,
A second π function of grade β is generated with break points at X2+β and X3−β.

The second π function from the first π function described above is wired 0R.
As a result of the subtraction at 86A, an M function with a maximum grade of α and a concavity of β in the center is obtained. However, α≧β,
The following conditions are required: X≦x, X2+2β≦x3≦x4.

In addition to being able to control the circuit in Figure 31 to generate 9 of the 10 functions mentioned above, excluding the W function, it is also possible to create variations thereof by setting α and β. .

Just to be sure, we will show the sufficient conditions for generating 6 functions excluding the φ function and 1 function from the 9 functions.

Z function x = x - x mL, a-1, β-D, C
.

l 2 3 1 (X 4 is the break point) or x
-L, a-1, β-1, X3-I+ X4''''HI (X 2 is the break point.) S function x -x -x -H, a-1, β-D, C.

2 3 4 I (x 1 is the break point) or x
-x-L, a-β-1,x4-H.

(X a becomes the break point.) π function α-1, β-Q, x, x-D, C.

(X, X becomes a break point.) Or,
x-xmH, α■β-1 (x, x2 are break points.)■ Or x mx-L, α-β-1(X
, X 4 becomes the breakpoint. ) U function x-L, x-H, α-β-1 (X, X3 are break points.) N function x-H, α-β-1 I (X, X, X are break points ) Then 1 function x = L, α − β = 1 (X,
X4 becomes the break point. ) M function α≦X ≦x, x+2β≦X ≦x l α−β−1 (x , x + x a , X 4 is break φ
This is the point. ) The circuit in Figure 31 is also based on the S function.

It goes without saying that an extended programmable multi-membership function circuit can also be realized by using the Z function as the basis.

FIG. 32 shows a modification of the circuit of FIG. 31 so that the slope can be switched between 1 and 2. Current mirrors 85A and 95A in FIG. 31 have been replaced by slope switchable current mirrors 85B and 95B, respectively.

These current mirrors 85B and 95B are the same as current mirror 25A in FIG. 15 and current mirror 35A in FIG. 17.

Also diode connected PET 83.93A. They are replaced by slope-switchable current mirrors 83B, 93B and are preceded by current mirrors 83C, 93C, respectively, to correct the direction of the current.

Wired 0R94A and 92A are given TJS&x and x, respectively, for simplicity.

Current mirrors 85B, 83B, 95B, 93B are P
- Since it is composed of MOSPETs, when those control voltage signals VC1 to Vc4 go to L level, the switching 1'ET is turned on, and the slope becomes 2 or -2.

The output current Z has the form shown by the broken line in FIG. Of course, the control voltage V. It goes without saying that VC1 to VC4 can be adjusted independently of each other.

(a, tO) S-function circuit applicable to crisp sets (Figures 34 and 35) The circuit in Figure 34 is an improvement of the S-function circuit (Figure 9 or Figure 32) so that it can be applied to crisp sets. This is what I did. A gradient switching circuit is also provided here. In comparison with FIG. 9 (or FIG. 32), the wired OR 104 is 34 (or 84A), and the switchable current mirror 105 is 35 (or 85B).
The wired OR107 is 37 (or 87A), and the diode 108 is diode-connected PET 3g (
or 88) respectively. Slope switching is performed by control signal vc1.

Therefore, wired OR 104 and current mirror 105
P-M as a switching element connected between
OS IcoET 10G, and wired OR10
N-MOS P[ET 1 as a switching element connected in parallel between 7 and a current source of value α (path shown)
01°P-MOS FET 102 is newly provided. FET102, 10B receives control signal V. On/off control is performed by 2. FEET 101 is controlled by the potential at node +09. This node 109 is provided between the wire 0R 104 and a current source of value x1 (the illustrated path), and its level changes to H or L level depending on the magnitude of the current flowing into or out of this node.

In a fuzzy set, whether something belongs to a fuzzy set is expressed by the degree of belonging, that is, by a continuous value from 1 to 0. Therefore, the membership function that represents this degree is.

As mentioned above, it has a sloped part. On the other hand, in the crisp set, whether something belongs to the crisp set is clearly expressed as 1 or 0. The membership function of the crisp set has a part that changes discontinuously from 1 to 0 or from O to 1 (a part with an infinite gradient).

Now, in FIG. 34, when the control voltage Vc2 is at L level, the two PETs 102.10B are on. PET 101 is connected in parallel with PET 102, so whether it is on or off, the 34th
The circuit shown works as a fuzzy set membership S function circuit. And the control voltage V. If 1 is H, the slope is 1
So, when L, the slope becomes 2. The input/output characteristics at this time are shown in FIG. 35 by solid lines and broken lines, respectively.

When the control voltage Vc2 becomes H level, PIET l0
Both L102 are turned off. Therefore, FETI
I) 6 is off, the input terminal X, is the current mirror 10
5 does not flow into node 10 from wired OR 104.
It will flow towards 9. When l'l:T 102 is off, whether the source input current α is provided to the wired OR 107 depends on the state of the FIET 101.

When X<Xl, the potential at the node 109 is at the L level, and the PET 101 is off. Therefore, the output current Z is O. When X, ≧Xt,
The node 109 becomes H level and the PET 101 is turned on. Current α is from wired 0R107 to PIET
101. Since the output current of current mirror +05 is 0, the output current Z becomes equal to α after all.

In this way, as shown by the chain line in FIG.
An output Z is obtained which inverts from 0 to 1 at . Control voltage V. 2 is at the H level, the level of the control voltage ■cl may be either H or L.

The difference between S-function circuits and Z-function circuits is that they define break points as described above? The only difference is the orientation of Ti1A. Therefore, by applying the concept of the circuit shown in FIG. 34 as is and selecting the MOSFET as a component as P type or N type as appropriate, a Z function circuit applicable to the crisp set can be constructed in the same way. .

The circuit 100 is shown by a chain line excluding the diode 108.

Since it will be used later in FIG. 40, it will be referred to here as the main part of the S-function circuit for convenience.

(3,11) Upgradient function circuit applicable to crisp sets (Figures 3G and 37) The circuit in Figure 36 is an upgradient function circuit (wired) with slope switching function shown in Figure 32. 0R82, a circuit consisting of a current mirror 83C and a current mirror 83B with a switchable slope, or a circuit consisting of a wired 0R92A, a current mirror 93C and a current mirror 93B with a switchable slope), which have been improved so that they can be applied to crisp sets. .

In comparison with Fig. 32, is the wired OR102 set to 82 (or 92A)? The S current mirror 103c corresponds to the S current mirror 83C (or 93C), and the slope-switchable current mirror 103B corresponds to the S current mirror 83B (or 93B). However, the connection order of current mirror 103c and slope-switchable current mirror 103B is as follows: current mirror 83C
(or 93C) and slope-switchable current mirror 83B (
Or the connection order of 93B) is reversed. In addition, the P type and N type of PET that constitute these current mirrors are
The type has been swapped. Then, the slope-switchable current mirror 103B is a current mirror 10g having two output drains and one of the output drains.
It consists of a PET 109 that switches one. PP and T 109 are controlled on and off by a control signal Vc3. In addition, N-MOS FET 1 for opening the gate-connected drain of the current mirror 108
07 has been newly added. This FET 107 is controlled by control signal Vc4.

The configuration of the circuit shown in FIG. 36 can be clearly seen when compared with FIG. 15. In the circuit shown in FIG. 15, PUT 107
Only a current mirror 103C and a current mirror 103C are added.

When the control signal Vc4 is at H level, this circuit
It works in the same way as the up-slope circuit for fuzzy sets in Figure 2. In other words, "If C4 is H.

PET 107 is turned on. At this time, the slope of the output current Z is equal to the control signal V. 3, and the output current Z
FIG. 37 shows input/output characteristics shown by solid lines and broken lines.

R71l When the J voltage vC4 becomes L level, PET 1
07 is off. By turning off PET 107.

PUT 108 no longer acts as a current mirror, but simply an amplifier.

In the case of X −<

If Xt≧Xt and there is a current that attempts to flow into the PET 10g even if it is a small value, this is amplified by the PE71011, and a rapidly increasing current flows on its output side. Therefore, as shown by the chain line in FIG.
A human output characteristic is obtained in which the output current Z rises almost vertically at X - X t.

Since the circuit of FIG. 36 is used in FIG. 38,
In particular, it is designated by the reference numeral 110.

(3, 12) Programmable multi-membership function circuit applicable to crisp set (Figure 38) Figure 38 shows the main parts 100 and S function circuit applicable to crisp set shown in Figure 34. Obtained by applying and improving the upslope function circuit 110 applicable to crisp sets shown in FIG. 36 to the extended programmable multi-φ membership function circuit shown in FIG. A programmable multi-membership function circuit applicable to crisp sets is shown.

In FIG. 38, the same components as those shown in FIG. 32 are given the same reference numerals. In addition, the circuit 100 in FIG.
Since two are used, this is 100A.

Similarly, the circuit 110 of FIG.
These are ll0A and ll0
It is shown by B.

From the graph shown corresponding to the arrow indicating the current flowing in the circuit, in the circuit of FIG. 38, the parameters X-X,
It is easy to understand that by changing α and β, an output current Z representing many types of fuzzy membership functions including the M function can be obtained. Also,
The slope can also be changed by changing the levels of the control voltages v-v and v-v.

It is also possible to generate many types of crisp membership functions.

[Brief explanation of drawings]

FIG. 1(A) shows a general membership function, and FIG. 1(B) shows a practical membership function approximated by a straight line. FIG. 2 shows the concept of the fuzzy control system. FIG. 3 is a block diagram showing the concept of a fuzzy system with a learning function. FIG. 4 is a graph showing various types of membership functions. FIG. 5 is a circuit diagram showing a Z function circuit constructed using MOS PET, and FIG. 6 is a graph showing its human output characteristics. FIG. 7 is a circuit diagram showing a Z function circuit constructed using bipolar transistors for determining A111 of input/output characteristics, and FIG. 8 is a graph showing measured input/output characteristics. FIG. 9 is a circuit diagram showing an S function circuit constructed using MOS FIETs, and FIG. 1O is a graph showing its input/output characteristics. FIG. 11 shows an S-function circuit configured using bipolar transistors for measuring input/output characteristics, and the 12th
The figure is a graph showing input/output characteristics with Δp1 constant. FIG. 13 is a graph showing a practical example of a membership function. FIG. 14 is a graph showing how the slope can be arbitrarily set by making the membership function and its variables correspond to the input/output current of the circuit. FIG. 15 is a circuit diagram showing a part of a Z function circuit that can switch the gradient into two, and FIG. 16 is a graph showing its input/output characteristics. FIG. 17 is a circuit diagram showing a part of an S function circuit that can switch the gradient into three, and FIG. 18 is a graph showing its human output characteristics. FIG. 19 is a block diagram showing an example of a programmable multi-membership function circuit. FIG. 20 is a circuit diagram showing an example of a multi-fanout circuit. Figure 21 (A) shows how the W function is generated by the fuzzy old N operation and fuzzy MAX operation of the Z function and the S function, and Figure 21 (B) shows the W function with the gradient switched. This is a graph showing. Figure 22 shows an MI constructed using MOS PET.
FIG. 3 is a circuit diagram showing an N circuit. FIG. 23 shows a MIN circuit constructed using bipolar transistors for measuring input/output characteristics, and FIG. 24 is a graph showing the measured input/output characteristics. FIG. 25 is a block diagram showing a three-man power MAX circuit constructed by combining two two-man power MAX circuits. Figure 26 shows an MA configured using MOS FETs.
FIG. 2 is a circuit diagram showing an X circuit. FIG. 27 shows a MAX circuit constructed using bipolar transistors for constant input/output characteristics M1, and FIG. 28 is a graph showing the measured input/output characteristics. FIG. 29 is a circuit diagram showing an example of a simplified programmable multi-membership function circuit based on an S function circuit. FIG. 30 graphically shows that a similarly simplified programmable multi-membership function circuit can be created based on two functions. FIG. 31 is a circuit diagram showing an expanded programmable multi-membership function circuit. FIG. 32 is a circuit diagram showing an extended programmable multi-membership function circuit with a slope switching function, and FIG. 33 is a graph showing its input/output characteristics. FIG. 34 is a circuit diagram showing an S-function circuit applicable to crisp sets, and FIG. 35 is a graph showing its input/output characteristics. FIG. 36 is a circuit diagram showing an upward slope function circuit applicable to crisp sets, and FIG. 37 is a graph showing its input/output characteristics. FIG. 38 is a circuit diagram illustrating a programmable multi-membership function circuit applicable to crisp sets. 31...Input terminal, 32...Output terminal 33.36.
...Current source, 34.37...Wired OR. 35...Current mirror. 38...Diode connection MO3FET. that's all

Claims (3)

[Claims] (1) A first current source that outputs a current with a value related to the break point; A limit difference circuit that calculates a limit difference that subtracts the input current from the output current of the first current source; A predetermined grade in fuzzy logic; a second current source that outputs a current representing , a subtraction circuit that subtracts the output current of the limit difference circuit from the output current of the second current source, and a diode for cutting off a negative direction current of the output current of the subtraction circuit. A fuzzy membership S-function circuit with an operating element, . (2) The fuzzy membership S-function circuit according to claim (1), wherein the predetermined grade is 1. (3) The limit difference circuit includes a multi-output current mirror, the output lines of the multi-output current mirror are connected in parallel, and at least one output line is provided with a switching element. ) The fuzzy membership S function circuit described in section 2.