JPH0291522A - Automatic balance - Google Patents

Abstract

PURPOSE:To obtain balancing at high speed with high accuracy by detecting the tilt angle of a balance beam and the angular velocity thereof, executing fuzzy control with the detection signal as an input value and obtaining balancing by means of driving a linear motor for moving the balance beam based on the inferred result. CONSTITUTION:When a weight is placed on one end side of the balance beam 4, the beam rotates centering around a horizontal axis. A detection part detects the tilt angle(rotation angle) of the beam 4 and the angular velocity thereof and gives the corresponding detection signal to a fuzzy controller 6. The fuzzy controller 6 obtains the moving distance of the balance beam 4 by applying a 1st inference rule group and a membership function to the tilt angle and the highest moving speed of the balance beam 4 and an accelerating time until the highest moving speed is attained are obtained by applying a 2nd inference rule group and the membership function to the angular velocity, then outputs the control signal of the linear motor 3. Thus, the linear motor 3 is driven to move the balance beam 4, thereby obtaining balancing.

Description

DETAILED DESCRIPTION OF THE INVENTION A. Field of Industrial Application The present invention relates to an automatic balance using fuzzy control.

B1 Summary of the Invention The present invention uses a linear motor as a movement mechanism for the balance beam in an automatic balance that changes the fulcrum position by moving the balance beam and achieves balance thereby, and also changes the inclination angle of the balance beam and its angular velocity. The balance can be balanced at high speed and with high accuracy by detecting the balance, executing fuzzy control using the detected signal as an input value, and driving a linear motor based on the inference result. It is.

C0 Prior Art In order to maintain balance when a weight is placed on one side of a balance and the balance is tilted, it is necessary to move the fulcrum of the balance.

Conventionally, the fulcrum has been automatically moved using PID control.

D1 Problems to be Solved by the Invention However, changing the fulcrum of the balance beam causes transient vibrations, making it difficult to achieve high-speed balancing operation using PID control. An object of the present invention is to provide an automatic balance that can balance at high speed and with high accuracy.

E Means for solving the problem A pedestal supported to rotate around a horizontal axis, a linear motor attached to this pedestal, a balance beam slid by this linear motor, and an inclination of this balance beam. a first inference rule group that uses the tilt angle of the balance beam as an input variable and the moving distance of the balance beam as an output variable; a fuzzy controller that stores a second inference rule group whose output variables are the maximum moving speed of the beam and the acceleration time to reach that speed and a membership function of each variable; and a linear motor driver that outputs a drive signal for the linear motor. It has.

When a weight is placed on one end of the F1 action balance beam, the beam rotates around the horizontal axis. The detection unit detects the beam inclination angle (rotation angle) and its angular velocity and provides a corresponding detection signal to the fuzzy controller.

In the fuzzy controller, the moving distance of the balance beam is determined by applying a first inference rule group and membership function to the inclination angle, and a second inference rule group and membership function is applied to the angular velocity to determine the balance beam movement distance. The maximum moving speed of the motor and the acceleration time to reach that speed are determined, and a control signal for the linear motor is output. Based on this control signal, the linear motor driver outputs a drive signal, which drives the linear motor to move the balance beam and reach equilibrium.

G. Embodiment FIG. 1 is a diagram showing the overall configuration of an embodiment of the present invention. l
is a support stand, and this support stand 1 supports a pedestal 2 so as to rotate around a horizontal axis. 3 is a linear motor which is attached to the pedestal 2, and above the linear motor 3 is a balance beam 4 which is slid by the linear motor 3.
is provided. Reference numeral 5 denotes an encoder that detects the inclination angle (80°) of the balance beam 4 from the horizontal plane as a pulse signal, and the pulse signal is taken into the fuzzy controller 6 through an input conversion section 6I in the fuzzy controller 6. The input conversion unit 61 converts the angular velocity (△θ/△
A pulse signal corresponding to t) is calculated. Note that in this embodiment, the encoder 5 and the input conversion section 61 constitute a detection section. The fuzzy controller 6 performs fuzzy inference based on these pulse signals as described later,
A control signal is created and output to the linear motor driver 7. The linear motor driver 7 converts the control signal into a drive signal and supplies it to the linear motor 3.

Regarding the fuzzy controller 7, in the fuzzy inference performed by the fuzzy controller 7, the above-mentioned inclination angle △θ (ANGLE) is used as an input variable.
and the angular velocity △θ/△t (DANGLE), and the moving distance (including the moving direction) of the balance beam 4 as the output variable (
DISTANCE), maximum movement speed of beam 4 (TOP
SPEED) and the acceleration time (ACCT IME) until the beam 4 reaches its maximum moving speed.

The fuzzy label of each variable is set as follows.

■ Regarding angle △θ (ANGLE), RL (right large), RM (right
t medium), R8 (right small)
, HO (horizontal), L S (lert
small), LM (left medium), L
L (left large)■Angular velocity △θ／△t(
DANGLE) About P B (positive
big), P M (positive medium)
), P S (positive small), Z
E (zero), N S (negative s)
mall), N M (negative medium)
m), N B (negativebig) ■About travel distance (D I 5TANCE) L M
A X (left max), L M I D (l
left medium), L S M L (left
small), Z E RO (zero), RSML
(right small), RM ID (rig
ht medium), RM AX (eight
max) ■About maximum movement speed (TOPSPEED) V HI
GH (very high 5peed), HI
G H (high speed), M I D (med
ium 5peed), S L OW (slow 5
peed), V S L OW (very slow
5peed) ■ About acceleration time (ACCTIME) V L ON G (very long time)
, LONG (long time), M I D (me
(dium time), S HORT (short
time), V S HORT (very 5hor
t time) Next, the membership functions given to the fuzzy labels of each variable are shown in FIGS. 2(A) to 2(E).

The value of a''-'e shown in the figure is appropriately determined depending on the size of the balance beam, etc., and for example, the following values are used.

a=lo, 0° b=4°/l OOm sec,
c50,Oxx,d=751vi/5ec11e=
2000m sec and these values are actually AN
It is handled as a pulse signal with a number of pulses corresponding to a numerical value, such as 500 pulses for GLE 100°.

In addition, the inference rules include a first inference rule group that uses tilt angle as an input number and travel distance as an output variable, and a second inference rule group that uses angular velocity as an input variable and maximum travel speed and acceleration time as output variables. The group is used. Specifically, each inference rule (R1-R14) is expressed as follows, and R1-R7
corresponds to the first inference rule group, and R7 to R14 correspond to the second inference rule group.

R1,lF ANGLE=RL THEN DISTA
NCB=LMAXR2:IF ANGLE=LM TH
EN DISTANCE=LMIDR3;IF ANG
LE=R5THEN DISTANCE=LSMLR4
,IF ANGLE=IIOTIIEN DISTAN
CE=ZEROR5・IF ANGLE=LS THE
N DISTANCE=RSMALR6,IF ANG
LE=LM THEN DISTANCE=RMIDR
7, IF ANGLE=LL THEN DISTAN
CE=RMAXR8, IF DANGLE=PB TH
EN TOP SPEED=VHIGHACCTIME=
SIIORTR9:IF DANGLE=PM T
HEN TOP SPEED 2 old Gll ACCTI
ME=MIDR1O, IF DANGLE=PS Tl
(EN TOP SPEED=SLOW ACCTIME
=MIDR11;IF DANGLE=ZE TIIE
N TOP SPEED=MID ACCTIME=LO
NGR12,IF DANGLE=NS Tl1EN
TOPSPEED=SLOW ACCTIMB=MID
R13・IF DANGLE=NM THEN TOP
SPEED=)11011 ACCTIME=MIDR
14,IF DANGLE=NB THEN T
IIEN TOP SPEED=VIIIGIIA
For example, MAX-MINI is a fuzzy inference method for determining the inferred value of the CCTIME=SIIORT output variable.
MAM method is used. In this case, for example, when the value △θ1 is detected as the angle △θ, the calculation is first performed on the inference rule r(1. That is, the membership value corresponding to △θ1 of the fuzzy label RL shown in FIG. 2(A) seek,
A trapezoidal function is obtained by cutting the membership function of the fuzzy label LMAX shown in FIG. 2(B) by the membership value. Similarly, calculations are performed for the inference rules R2 to R7, and the obtained function group is synthesized to obtain the center of gravity position in the horizontal axis direction, and this is used as the inferred value of the moving distance. The maximum moving speed and acceleration time are determined in a similar manner, and based on these inferred values, a control signal for the linear motor, such as a drive command pulse train, is created.

FIG. 3 is a diagram showing the logical configuration of the fuzzy controller 6, which includes a simulation section 8, a control section 9, and a support section IO.
Consisting of In the same figure, 6I is the input conversion section described above; 8. ．． 9
．． Fuzzy Reasoning Department, 8t.

9 is a storage unit that stores the first and second inference rule groups; 8;
3,9. 6. is a storage unit that stores membership functions;
is the fuzzy inference part 8. ,9. This is an output conversion unit that creates a drive command pulse train for the linear motor based on the moving distance etc. determined in . The support unit IO also has a data area 10.0 that stores the detected value of the angle Δθ from the input conversion unit 61, data upstream from the output conversion unit 6, and the like. and a man-machine interface for inputting inference rules and membership functions 10. and a monitoring unit 10 for monitoring inference rules and the like. For example, it has a CRT screen.

Next, the operation of the above embodiment will be explained. First, in the simulation unit 8, the music that enters the actual process control is stored in the storage unit 8. ．． 8. After performing a simulation in advance and debugging using the inference rule group and membership function stored in the controller 9, the determined inference rule group and membership function are respectively stored in the storage units 92 and 93 in the control unit 9. It will be stored and then put into actual operation.

In actual process control, when the weight W is placed on one end of the balance beam 4, the beam 4 begins to rotate counterclockwise.

At this time, the inclination angle △θ of the beam 4 is detected as a pulse signal by the encoder 5, and as described above, the inclination angle △θ is detected by the encoder 5 as a pulse signal.
θ and its angular velocity Δθ/ΔL are input to the input conversion section 6. is input to the fuzzy inference section 9□ of the control section 9. Then, as already detailed, fuzzy inference is performed using the inference rule group and membership function, and the moving distance of the balance beam 4 (
(including direction), maximum travel speed, and estimated acceleration time. Output converter 6. Now, based on these inferred values, a linear motor drive command pulse train is created and output to the linear motor driver 7.

The linear motor driver 7 powers up this pulse train and supplies it to the linear motor 3, which drives the linear motor and causes the balance beam 4 to slide (balanced movement).
As a result, the balance reaches an equilibrium state. After that, when the weight W is removed, the balance beam 4 rotates clockwise and is similarly balanced by fuzzy control.

■-19 Effects of the Invention The present invention uses a linear motor as a mechanism for moving the balance beam, detects the inclination angle of the balance beam and its angular velocity, executes fuzzy control using the detected signal as an input value, and calculates the inference result. The balance is maintained by driving a linear motor. Therefore, the balance can be balanced in a short time, and the balance can be balanced quickly and with high accuracy regardless of the weight of the weight placed on the beam. When changing the shape of the balance itself, this can be easily done by adjusting the membership functions of each variable.

[Brief explanation of the drawing]

Fig. 1 is a configuration diagram showing an embodiment of the present invention, Fig. 2 (A) -
(E) is a graph showing the membership function, and FIG. 3 is a block diagram showing the logical structure of the fuzzy controller. DESCRIPTION OF SYMBOLS 1... Support stand, 2... Pedestal, 3... Linear motor, 4... Balance beam, 5... Encoder, 6...
Fuzzy controller, 61... input conversion section, 7...
- Linear motor driver, 8... simulation section, 9... control section, 10... support section. Configuration of the embodiment in Figure 1 2 people not shown in the diagram 1 - Support stand 2 - pedestal 3 - Linear motor 4 - - Balance beam 5 - Encoder 6 - Fuzzy controller 6 - Input converter 7 - Linear motor driver

Claims (1)

[Claims] (1) A pedestal supported to rotate around a horizontal axis, a linear motor attached to this pedestal, a balance beam slid by this linear motor, the inclination angle of this balance beam, and its angular velocity a detection unit that detects the
A first group of inference rules in which the tilt angle of the balance beam is used as an input variable and the moving distance of the balance beam is used as an output variable, and the tilt angular velocity of the balance beam is used as an input variable, and the maximum moving speed of the balance beam and acceleration to that speed A second inference rule group that uses time as an output variable and membership functions of each variable are stored, and the first inference rule group and second inference rule are applied to the inclination angle and its angular velocity detected by the detection unit. a fuzzy controller that performs fuzzy inference by applying a set of rules and membership functions, thereby determining the moving distance, maximum moving speed, and acceleration time of the balance beam and outputting a control signal; and a fuzzy controller that outputs a control signal based on the control signal. An automatic balance comprising a linear motor driver that outputs a drive signal for a linear motor.