JPH09101815A - Device and method for servo control and device and method using for robot control the device and the method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent generated operation from being deviated from the limit performance of a controlled target by regulating an acceleration/deceleration pattern controlling outer frame having a reference function and its peak speed arrival, time. SOLUTION: A signal generated from a command means 2 is sent to an acceleration/deceleration pattern generating means 3 and a TP value determining means 4. The means 4 determines a peak speed arrival time TP value corresponding to a position command Δθ and sends the TP value to an α value determining means 5. The means 3 generates an acceleration/deceleration pattern based upon the command signal from the means 2 and a time base extending/ shortening parameter α value. Since the means 3 has a standardizing function corresponding to the limit performance of a controlled target, the means 3 can generate scaling available pattern by scaling operation, so that motion following up the pattern is not deviated from the limit performance of the controlled target. An output from the means 3 is sent to a servo system 7 through a low pass filter 6.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a servo control for generating an acceleration / deceleration curve whose shape is defined by a characteristic limit of a controlled object, and for unifying algorithms for generating the acceleration / deceleration curve. The present invention relates to an apparatus and a servo control method, and a robot control apparatus and a robot control method using them.

[0002]

2. Description of the Related Art Generally, control is performed such that acceleration control is performed for a speed or angular velocity to be controlled and then deceleration control is performed, or deceleration control is performed after constant speed control for a predetermined period after acceleration control. When performing, it is necessary to control the shape of the acceleration / deceleration curve (or the acceleration / deceleration command curve).

A frequently used acceleration / deceleration curve is, for example, a triangular acceleration / deceleration curve.

In FIG. 27, the horizontal axis represents time t and the vertical axis represents velocity Ω = θ ⁽¹⁾ (the subscript “(n)” (n is a natural number) at the upper right of the position function θ is This means an n-th derivative with respect to time t, and this abbreviation will be used hereafter.), Which shows an example of the shape of a triangular acceleration / deceleration curve, and t = T
_The acceleration / deceleration curve a at which the speed reaches the peak value at the time of _P is 0
In the range of ≤t≤T _P , the linear acceleration part b is inclined to the upper right, and in the range of T _P <t≤2 · T _P , the linear deceleration part c is inclined to the lower right.

When the operation according to such a triangular acceleration / deceleration curve is expressed on the diagram together with the acceleration-speed characteristic or the torque-angular speed characteristic of the controlled object, the operating point is on the side of the quadrangle. A trajectory that moves along is drawn.

For example, when the object to be controlled is specified as a system including the motor and its drive circuit (hereinafter referred to as "motor system"), θ may be considered as the phase angle of the motor.
As shown in, the horizontal axis represents the torque T and the vertical axis represents the rotation speed N.
On the TN diagram (in the figure, only the first quadrant is shown).
At the time of acceleration, the operating point P moves in the direction indicated by the arrow on the side of the quadrangle d where T = T _a and N = N _a (the inclination of the acceleration part b of the acceleration / deceleration curve a is constant and constant). And the inclination of the deceleration part c is negative and constant.)

In this TN diagram, the lines corresponding to the three sides of the pentagon, which is one size larger than the quadrangle d, excluding the T axis and the N axis, indicate the limit TN line e of the motor system to be controlled. However, if the operating point P exceeds the limit T-N line e and sticks out of the outside, the performance and quality of the motor system can no longer be guaranteed.

[0008]

By the way, the above-mentioned triangular acceleration / deceleration curve a is easy and convenient for controlling the acceleration constant, but it is difficult to fully utilize the performance of the controlled object. There is a problem.

That is, in the TN diagram shown in FIG. 28, the portions f, f sandwiched between the limit TN line e and the quadrangle d.
(Indicated by diagonal lines in the figure) cannot be used effectively for control, so it is difficult to speed up the control operation, and a large current value is consumed, so a large motor with high output is required, which reduces the cost. It causes an increase.

Then, at t = T _P , the acceleration is obviously discontinuous, so that a shock may be applied to the controlled object and the performance and the like may be adversely affected.

Therefore, by adopting an acceleration / deceleration curve having another shape, for example, an S-shaped accelerating portion or decelerating portion or one characterized by the shape of a sine function, the continuity of acceleration can be obtained. Guarantee or shaded parts f, f above
However, it was difficult to systematize and unify the algorithms for the generation of these acceleration / deceleration curves. Had to do.

Therefore, an object of the present invention is to generate an acceleration / deceleration curve whose shape is defined by the limit performance of the controlled object and to unify the generation algorithm.

[0013]

In order to solve the above-mentioned problems, the present invention has the constitutions shown in (1) to (3) below.

(1) First, a limit characteristic line relating to characteristics of acceleration (or force or torque) with respect to speed (or speed, angular velocity or rotational speed) of a control target is specified, and an operating point defined by acceleration and speed is specified. Defines the shape of the acceleration / deceleration curve, the reference function thereof, and the peak velocity arrival time of the reference function under the condition that the curve is on the limit characteristic line of the controlled object. In other words, the acceleration / deceleration pattern generating means that holds the reference function and its peak speed arrival time is provided, and the outer frame of control is specified so that the operation according to the acceleration / deceleration curve generated by this does not deviate from the limit performance of the controlled object. To do.

(2) The relationship between the designated value of the servo control variable and the peak speed arrival time is defined, and the value of the peak speed arrival time corresponding to the command value is obtained, and the peak speed arrival time according to the reference function is calculated. The dimensionless time axis expansion / contraction parameter is determined by dividing by the value of. Therefore, command means for issuing a specified value of the servo control variable,
A peak velocity arrival time value determining means for determining a peak velocity arrival time corresponding to the command value, and a time axis extension / contraction parameter value determining means for determining the value of the time axis extension / contraction parameter are provided.

(3) The acceleration / deceleration curve is obtained by performing the expansion / contraction operation in the time axis direction on the reference function based on the time axis expansion / contraction parameter value and the scaling operation on the reference function value based on the designated value of the servo control variable. Is generated and sent as a control signal to the controlled object. That is,
The acceleration / deceleration pattern generation means prepared in (1) generates an acceleration / deceleration curve by performing a scaling operation on the function value or a time axis expansion / contraction operation on the reference function.

Therefore, according to the present invention, the shape of the acceleration / deceleration curve, that is, the reference function, is defined by the limit performance of the controlled object, and the acceleration / deceleration curve is changed by the scaling operation of the function value or the time axis expansion / contraction operation that limits this. Since the acceleration / deceleration curve is generated, the generation algorithm is not required for each acceleration / deceleration curve having a different shape, and the acceleration / deceleration curve can be generated according to the unified algorithm.

[0018]

DETAILED DESCRIPTION OF THE INVENTION A servo control device and a servo control method according to the present invention will be described below. In the control according to the present invention, it is necessary to understand the following two basic items.

A) The shape of the acceleration / deceleration curve is determined depending on the limit performance of the controlled object. In other words, the controlled object is always constrained to be used within the range of its limit performance, and therefore the shape of the acceleration / deceleration curve cannot be freely selected. Also,
If the shape of the acceleration / deceleration curve depends on the ability and experience of the designer, thinking and error are forced each time, and the design becomes inefficient.

Therefore, it is necessary to clarify how the shape of the acceleration / deceleration curve is defined based on the limit performance of the controlled object.

B) The acceleration / deceleration curve is generated by performing some kind of scaling operation on the reference function of the shape.

Even if the shape of the acceleration / deceleration curve is defined by the above a), it is troublesome that the generation algorithm is different for each acceleration / deceleration curve. This is because, in the case of a robot that is composed of a large number of arms and its drive system, the algorithm for generating the acceleration / deceleration curve when controlling each arm requires as many arms as the extreme example. Therefore, it is necessary for the control device of the arm drive system to be provided with a means for generating an acceleration / deceleration curve for each algorithm, which is complicated and uneconomical.

Therefore, no matter what the shape of the acceleration / deceleration curve obtained by a) is, only the reference function of the shape of the acceleration / deceleration curve is prepared and scaled in order to unify the generated algorithms. (Variation and expansion / contraction of time axis)
It is necessary to handle the acceleration / deceleration curve freely by performing an operation.

In the following description, the above two items will be always kept in mind.

First, the controlled object in the present invention is required to have its limit performance characterized by acceleration (or force, torque, etc.)-Speed (or speed, angular velocity, rotation speed, etc.) characteristics. In other words, as long as there are no limit restrictions on the controlled object, the performance of the controlled object will not be exceeded regardless of the shape of the acceleration / deceleration curve, and the criteria for determining the value of the shape of the acceleration / deceleration curve. Can not be specified, and is removed from the controlled object according to the present invention. Therefore, as long as this condition is met, the controlled object according to the present invention may have any configuration, and the system including various actuators and their drive circuits is included in the controlled object.

Next, the acceleration (or force, torque, etc.)-Speed (or speed, angular velocity, rotation speed, etc.) characteristic that defines the limit performance of the controlled object is the position or angle of time such as acceleration or torque. Taking the second derivative or the amount containing it on one axis,
When the first-order derivative of a position or angle with time, such as velocity or angular velocity, or a quantity including this is expressed on the other axis with respect to the relationship between them, the limit performance of the controlled object is specified on the diagram. Characterized by a curve (limit line).
That is, the performance and quality of the controlled object are guaranteed on the curved line and in the inner area surrounded by the curve and the two axes, but when stepping into the outer area, the performance and quality of the controlled object are guaranteed. Disappear.

For example, if the controlled object is specified to the motor system, the characteristics as shown in FIG. 1 can be exemplified.

FIG. 1 shows the torque-speed (rotational speed) characteristic of the AC (alternating current) motor system (only the first quadrant is shown). The horizontal axis represents the torque T and the vertical axis represents the speed N. This shows several examples of the limit TN line and the rated TN line.

In the example shown in FIG. 1A, the rated TN line g
1 is a line segment of N = Nm, T = T1 (T1 is rated torque)
, And a line segment descending to the right connecting both line segments, and a limit TN line g2 is a line segment of N = Nm and is inclined downward to the right in succession thereto, and T = T2. (T2 is the instantaneous maximum torque) and is formed by a line segment that intersects the T axis.

In the example shown in FIG. 1B, the rating T-
The N line h1 has the same shape as the rated T-N line g1, but the limit T-N line h2 is a line segment of N = Nm, and a line segment that slopes to the lower right and intersects the T axis at T = T2. , And a line segment descending to the right that connects both line segments.

In the example shown in FIG. 1C, the rating T
The −N line i1 has the same shape as the rated TN line g1, and the limit TN line i2 is a line segment of N = Nm, a line segment of T = T2, and a right-downward line connecting both line segments. It is composed of line segments and.

The limit T-N line and the rated T-N line composed of a plurality of line segments as described above are expressions for the sake of convenience to make the characteristics of the motor system easy to understand, and the actual limit T The −N line and the rated TN line are generally curved lines having a complicated shape.

Since the motor system has been taken up as the control target, the effective values of the mechanical power (power) and electric power (power consumption) of the motor will be described.

[0034] FIG. 2 shows a simplified equivalent circuit of the servo motor, the resistance to the variable voltage source Vccm showing a control voltage (resistance value and "R _M".) And the electromotive force (motor torque The constant is “K _t ” and the angular velocity of the motor is “Ω”
Then, −K _t · Ω. ) Is connected in series with the voltage source shown in FIG.

[0035] Incidentally, for simplification of explanation, a triangular acceleration and deceleration curve (peak rate arrival time of the shape shown in FIG. 27 is "T _P", the peak value of the speed is "Omega _p", pressurized The area enclosed between the deceleration curve and the time axis, that is, the total movement (rotation) amount of the motor is “Δθ”. However, the result obtained here will be generalized later. Be careful.

Table 1 below shows symbols and meanings of various physical quantities used in a tabular form.

[0037]

[Table 1]

First, when the motor is operating, the motor resistance R _M
It is consumed when the power to be thermally and _"I P", which is as the following equation by.

[0039]

(Equation 1)

[0040] Thus, the effective value in the period 2 · T _{P "I}
P _RMS ", this becomes as shown in the following equation.

[0041]

(Equation 2)

[0042] It should be noted, _"I P _RMS" current I _M, that is determined by only the acceleration, it is to be noted that regardless of the total amount of movement Δθ of the motor.

Next, the mechanical power supplied from the power source to the mechanical portion via the motor during the operation of the motor is expressed by " _K P
(T) ”and the effective value in the period 2 · T _P is“ _K P _RMS ”, this becomes the following formula.

[0044]

(Equation 3)

In the case of a triangular acceleration / deceleration curve, _K P
(T) is proportional to the time t as shown in the following equation [Equation 4],
Substituting the equation 4 into the equation 3 above to obtain the effective value _K P _RMS , the total movement amount Δθ of the motor can be obtained as shown in the following equation 5.
It is a function of the peak velocity Ω _p in proportion to the square of.

[0046]

(Equation 4)

[0047]

(Equation 5)

FIG. 3 shows the peak velocity Ω _p and the peak velocity arrival time T _P when the triangular acceleration / deceleration curve is continuously used.
FIG. 4 is a t-Ω diagram showing the state of one cycle when the value is changed. FIG. 4 shows the locus of the operating point corresponding to each acceleration / deceleration curve and the effective value of the mechanical power and the effective value of the electrical power. It is shown.

In FIG. 3, the acceleration / deceleration curve j _i (i = 1 to 4) has the respective peak velocity arrival times T _Pi (i = 1).
To 4) and the peak speed is Ω _pi (i = 1 to 4).

The horizontal axis of FIG. 4 is K _t · I _M, and
The vertical axis forming the quadrant is the speed Ω, and the vertical axis forming the fourth quadrant is R _M · I _M / K _t , and the acceleration / deceleration curve j
The effective value of the mechanical power corresponding to _i (i = 1 to 4) is “ _K P _RMSi ” (i = 1 to 4).

As shown in the figure, the effective value _I P _RMS of the electric power consumed by the resistor R _M is constant regardless of the acceleration / deceleration curve, and is equal to the area of the shaded portion. In addition, the effective value of mechanical power _K P _RMSi (i = 1 to 4)
Corresponds to the area of each quadrangle, as can be seen from the fact that the operating point moves on the side of the quadrangle as shown by the arrow, and differs for each operation.

Since the ability of the controlled object depends on its temperature, the algorithm for generating the acceleration / deceleration curve is _I P _RMS
It is preferable to include the value in the evaluation value.

When the controlled object is operated continuously,
The values of _I P _RMS and _K P _RMS are determined as described above, but it can be shown that the relationship between the total movement amount Δθ and the peak speed arrival time T _P is defined from the condition that these are always constant. it can.

That is, when _I P _RMS is controlled to be constant,
The following expression [expression 6] is obtained from expression [expression 2], and T _P is 2 of Δθ.
A relationship that is proportional to the 1st power is derived.

[0055]

(Equation 6)

That is, when Δθ is taken on the horizontal axis and T _P is taken on the vertical axis as shown in FIG. 5, if the relationship between the two is defined as shown by the parabola pr connecting points A and B, The temperature of the controlled object can be kept constant even when the continuous operation of the controlled object is performed at any Δθ. The point A in FIG. 5 is the minimum value T _{PMI of} T _P determined by the limit of the frequency characteristics of the controlled object.
A point corresponding to _N, and a point B indicates a point corresponding to the maximum value of the peak velocity Ω _p .

In the case of controlling _K P _RMS to be constant, the relational expression that T _P is proportional to the third power of Δθ is derived, and in the case of controlling _I P _RMS + _K P _RMS to be constant. Is Δθ of 2
Since the fourth-order equation of T _P including the power is derived, the Δθ-T _P characteristic can be defined based on these, as in the case of FIG.

Next, regarding b), the scaling operation for the reference function and the expression of the acceleration / deceleration pattern obtained by the scaling operation will be described.

The "acceleration / deceleration pattern" referred to here is
The state quantity (that is, the velocity and the angular velocity) directly described by the acceleration / deceleration curve is, of course, a broad concept including the time derivative of the state amount, the integral amount, and the like.

The acceleration / deceleration curve relating to the controlled object needs to be generated according to the unified algorithm regardless of the shape.

Therefore, by imposing as few constraints as possible on the shape of the acceleration / deceleration curve, a unified expression of the acceleration / deceleration curve including the conventional acceleration / deceleration curve having a triangular shape or a sine wave shape will be given. .

This new acceleration / deceleration pattern is characterized by the introduction of a reference function that determines the shape of the acceleration / deceleration curve and the scaling operation for the reference function.
Hereinafter, this will be referred to as a “scalable pattern”.

The standard function of the scalable pattern is called a "normalization function", which will be referred to as "θ _N (τ / T)". However, “τ” represents a time axis parameter, T represents a time constant of the control system, and is a constant necessary for making dimensionless to τ / T.

Then, the position function related to the scalable pattern is set to "Θ (t)" (where t is real time), and the time axis expansion / contraction parameter defined by the relation "t = ατ". Introduce “α”.

The amount of movement related to Θ (t) (that is, Θ ⁽¹⁾
Total amount of movement obtained by integrating (t) over the entire movement time. ) Is defined as “Δθ”, and the first-order time derivative of the normalization function θ _N θ _N ⁽¹⁾
Is referred to as "normalized speed", and the time until the peak value is reached, that is, the peak speed arrival time related to the normalized speed θ _N ⁽¹⁾ is defined as " _TPn ". If the peak velocity arrival time for velocity Θ ⁽¹⁾ , which is the first-order time derivative of the position function Θ (t), is “T _P ”, then T _P has a lower limit value (this is referred to as “T _P ”).
_PMIN ". ) Or an upper limit value (this is referred to as “T _PMAX ”), T _Pn can be selected as a certain value within the range of T _PMIN ≦ T _P ≦ T _PMAX .

The definitions of the above various quantities are summarized in Table 2 below.

[0067]

[Table 2]

FIG. 6 is a graph showing the normalized speed θ _N ⁽¹⁾ . The horizontal axis is “τ” and the vertical axis is “θ _N ⁽¹⁾ (τ /
T) ”, and“ T _all ”indicates the total travel time.

The normalized speed θ _N ⁽¹⁾ is 0 ≦ τ ≦ T _Pnu (=
T _Pn ), the acceleration part and T _{P nu} <τ ≦ T _all -T
In a constant-speed portion of the period of _pnd (which is referred to as "T _c".), Which consists of a deceleration section in the period _{T al l -T Pnd <τ ≦} T all, these partial and tau shaft It has the property that the area of the enclosed range, that is, the definite integral value is equal to 1. That is, “normalization” in the normalization function θ _N means that the total movement amount related to θ _N is adjusted to always 1. It should be noted that this may not always be included in the constant speed part (period T _c = 0).

The relational expression showing the property of the normalization function θ _N is shown below.

[0071]

(Equation 7)

[0072]

(Equation 8)

[0073]

(Equation 9)

[0074]

(Equation 10)

The expression [7] means the existence of a peak speed value of the standardized speed, and the expression [10] implicitly assumes that the deceleration portion does not include a vibration component. When the component is included, it is generally “θ _N (∞ / T) = 1”.

Next, for convenience of explanation, an assumption is made for the standardization function θ _N to the extent that generality is not impaired.

The first reason is that the acceleration time is considered in consideration of the fact that the limit characteristic line that determines the performance of the controlled object has symmetry between the portion belonging to the first quadrant and the portion belonging to the second quadrant on the characteristic diagram. T _Pnu is set equal to the deceleration time T _Pnd , and the second is to set T _Pn to the minimum value T _PMIN of T _P , and when these are expressed by a mathematical expression, the following expression is obtained. .

[0078]

[Equation 11]

[0079]

(Equation 12)

It should be noted that these assumptions are for convenience only, and the assumptions are intended to give concreteness to the description by reducing the arbitrariness of the setting of T _Pn and the shape of the acceleration / deceleration curve as much as possible. Be careful not to pass too much.

Specific examples of the standardizing function θ _N include the following functions.

I) In the case of a triangular acceleration / deceleration curve.

[0083]

(Equation 13)

Ii) In the case of an acceleration / deceleration curve obtained when the position command is given to the third-order filter as a ramp signal.

In controlling the servo system to be controlled, a system corresponding to the actual servo system is simulated by software processing to be prepared in the control device (such a servo system will be described below. By introducing a "virtual servo system") and using it as a calculation means, it can be used for generation of acceleration / deceleration curves and calculation of various estimated quantities. For example, in the control circuit of the motor system,
A circuit portion directly related to motor control and a portion related to dynamic compensation for stabilization of control are provided in the servo system. However, a virtual servo system (hereinafter, "virtual servo system" having a filter configuration corresponding to these is provided. "Motor system") is prepared in the control device as an internal model by software processing, the output when a command is given to the virtual motor system can be used as it is as a control signal to the actual servo system. it can.

For example, the virtual servo system is constituted by a third-order low-pass filter, and the output when the position command of the ramp signal as shown in FIG. 7 (this is referred to as “Δθ _REF ”) is input is used. By doing so, it becomes unnecessary to explicitly perform the calculation related to the normalization function. That is, in general, the response characteristic of the servo system is not always expressed by an analytical expression, and therefore it is useful from a practical point of view to introduce a virtual servo system that imitates the actual servo system.

In the case of the third-order filter, the function expression can be easily obtained, and the following expression is obtained.

[0088]

[Equation 14]

In the above equation, "T _i " (i = 1, 2, 3)
Is the time constant of the servo loop, " _Tf " is the break time of the ramp signal (see FIG. 7), and "exp ()" is an exponential function. Further, in deriving the above equation, the inverse Laplace transform may be performed on the output obtained from the product of the Laplace expression of the ramp signal and the loop transfer function of the filter, but the process is directly related to the gist of the present invention. Since it does not exist, the explanation is omitted.

Iii) In the case of an acceleration / deceleration curve having a shape including a sine function.

[0091]

(Equation 15)

Now, when the normalization function θ _N is defined as described above, the scalable pattern can be defined as the following equation based on this.

[0093]

(Equation 16)

First, the position function Θ (t) is obtained by multiplying the normalizing function θ _N by Δθ and multiplying the time axis τ and the time constant T by α, as shown in the following equation.

[0095]

[Equation 17]

The velocity function Θ ⁽¹⁾ (t) is obtained by multiplying Δθ by 1 / α and multiplying it by the normalized velocity θ _N ⁽¹⁾ , as shown in the following equation.

[0097]

(Equation 18)

The acceleration function Θ ⁽²⁾ (t) is obtained by dividing Δθ by the square of α as shown in the following equation, and the acceleration function θ _N ⁽²⁾ (hereinafter referred to as “normalized acceleration”). .).

[0099]

[Equation 19]

From these equations, it becomes clear that the function related to the scalable pattern can be obtained by independently performing the expansion / contraction operation in the time axis direction for the normalization function and the like and the scaling operation for the function value itself.

FIG. 8 is a diagram for explaining the scaling operation with respect to the standardized speed with τ on the horizontal axis and the speed Ω on the vertical axis. FIG. 8A shows the standardized speed with Δθ = 2. The operation of doubling the value of θ _N ⁽¹⁾ is shown in FIG.
FIG. 8B shows an operation in which α = 2 and the normalized speed θ _N ⁽¹⁾ is doubled in the time axis direction. FIG. 8C shows the operation in FIG. 8A and the operation in FIG. 8B. It shows an operation that is a combination of operations.

As described above, if the standardized speed (or the standardized function) is given in advance, it is independent in the vertical axis direction (variation direction of the function value) and the horizontal axis (time axis) direction. By adopting the rule of performing expansion / contraction operation (scaling), the normalized function θ _N or the normalized speed θ _N ⁽¹⁾
The acceleration / deceleration pattern can be freely controlled regardless of the shape of the. In other words, all the acceleration / deceleration patterns generated by such a method are included in the scalable pattern.

It should be noted that a specific example of the scalable pattern can be easily obtained by substituting the expressions of the standardization functions shown in the equations [13] to [15] into the equations [16] to [18]. Since the generation algorithm does not depend on the specific shape of the acceleration / deceleration curve with respect to the concept of the scalable pattern, the description thereof will be omitted.

As described above, it becomes clear that the position, velocity, and acceleration related to the scalable pattern can be derived based on the standardization function, and more generally, the physical quantities relating to the motion of the controlled object are calculated by the standardization function (θ _N ), the total movement amount (Δθ), and the time axis expansion / contraction parameter (α) can all be derived.

For example, in the case of the effective current value related to the controlled object, the control that keeps the effective value _I P _RMS of the electric power constant explained in FIG. 5 when the controlled object is specified to the motor system. In, it is possible to discuss the magnitude of _I P _RMS by the effective current value of the motor, and the effective current value is often more convenient when referring to the specifications of the motor.

Therefore, the effective current value of the motor is set to " _IRMS ".
Then, this can be defined by the following formula [Formula 20], and by using the formula [Formula 19] and the relation of “t = α · τ”, the following formula [Formula 21] can be obtained.

[0107]

(Equation 20)

[0108]

(Equation 21)

In the equation [21], the effective current value of the motor when the motion control is performed according to the scalable pattern is
It shows that it can be calculated from the effective value of the normalized acceleration (θ _RMS ⁽²⁾ ), Δθ and α.

In the above description, when calculating the value of the standardizing function, the amount of calculation is not considered and the value to be calculated is obtained immediately. However, when the shape of the standardizing function is complicated. Since it is expected that the calculation of the function value will take time, in such a case, it is practical to prepare a data table relating to the standardization function in advance and to make a database of the function value.

From the above description, it is clear that the item b), that is, the introduction of the scalable pattern, realizes the unification of the acceleration / deceleration pattern generation algorithm. Since the description of the relationship between the limit performance and the shape of the acceleration / deceleration curve remains, this point will be described in detail below.

The limit performance of the controlled object is defined by the limit characteristic line in the acceleration-speed characteristic diagram as described above, and it is easily understood that this serves as a design guideline for the acceleration / deceleration pattern. This is because the motion of the controlled object corresponds to the operating point or the locus of the operating point on the characteristic diagram in a one-to-one manner, and the acceleration / deceleration is such that the operating point deviates from the guaranteed operating range bounded by the limit characteristic line. This is because the generation of the pattern is not permitted or dangerous in terms of control.

Therefore, hereinafter, the behavior of the operating point corresponding to the motion according to the scalable pattern on the characteristic diagram will be described first.

FIG. 9A shows an operating point Q (θ _N on the normalized acceleration (θ _N ⁽²⁾ )-normalized velocity (θ _N ⁽¹⁾ ) diagram for the normalizing function of the scalable pattern. ⁽²⁾ , θ _N
⁽¹⁾ ) about the trajectory M, where the horizontal axis is the standardized acceleration and the vertical axis is the standardized speed. For comparison, temporal changes in the standardized speed and the standardized acceleration are shown in FIGS. 9B and 9C, and the subscript "" attached to the physical quantity in these figures is shown. "P" means the peak value of each amount.

In the locus M shown in FIG. 9A, the portion M1 shown in the first quadrant of the figure corresponds to the acceleration portion in the acceleration / deceleration curve, and the portion M2 shown in the second quadrant of the figure shows the acceleration / deceleration curve. It corresponds to the deceleration part, and the operating point Q is the locus M.
Move up in the direction indicated by the arrow.

Here, it should be noted that θ _N ⁽²⁾ is θ _N
^This means that it is a monovalent function for ⁽¹⁾ . For example, as shown in FIG. 10A, there are a plurality of acceleration values corresponding to one value of θ _N ⁽¹⁾ , and as shown in FIG. 10B, a locus including a loop is not drawn. That's what it means.

When the acceleration / deceleration curve relating to the normalization function includes the constant velocity part, the operating point Q is (0, θ
_NP ⁽¹⁾ ).

If the behavior of the trajectory related to the normalization function is clarified, the behavior of the scalable pattern obtained by performing the scaling operation on it is also clarified. At that time, the scaling operation is performed on the characteristic diagram. Figure 1 shows how it affects the trajectory of the normalization function.
1 will be described.

In FIG. 11, the horizontal axis represents the acceleration (Θ ⁽²⁾ (t)) and the vertical axis represents the velocity (Θ ⁽¹⁾ (t)).
FIG. 10 shows the locus of (Θ ⁽²⁾ , Θ ⁽¹⁾ ) and also shows the locus M related to the normalization function shown in FIG. 9.
It should be noted that, for simplification of description, only loci belonging to the first quadrant are shown in the figure.

As described with reference to FIG. 9, in the scaling operation related to the scalable pattern, the scaling operation of the function value itself and the expansion / contraction operation in the time axis direction are independent of each other.

FIG. 11A is for explaining the scaling operation (scaling by Δθ) of the function value itself, and as can be seen from the formulas [18] and [19], Θ
^{Since (2)} and Θ ^{( 1)} are proportional to Δθ, when the time axis expansion / contraction parameter is fixed to α = 1, this operation shows the trajectory M related to the normalization function as it is for both axes. It is nothing but to obtain a trajectory Ma similar to the trajectory M by multiplying by Δθ.

FIG. 11B is for explaining the expansion / contraction operation (magnification by α) in the time axis direction. If Δθ = 1 in the formulas [18] and [19], As shown in the formula [Equation 22], the speed is 1 / α of the standardized speed, and the acceleration is the standardized acceleration α.
Since it is converted to a value obtained by dividing by a square of, the operation is performed with respect to the locus M related to the normalization function as shown in the figure, where α is 1 / square for the acceleration axis and α is for the velocity axis.
It is nothing but to obtain the locus Mb by reducing the number by one.

[0123]

(Equation 22)

In the figure, 0 ≦ α ≦ 1 is set for the sake of easy understanding. However, in consideration of the arbitrariness of the selection of the peak speed arrival time T _Pn related to the normalization function, T _Pn = T _PMAX Corresponds to the case you choose.

FIG. 11C shows the operation of FIG. 11A and FIG.
1 shows a locus Ma * b obtained from a locus M relating to a normalization function by combining with the operation 1 (b), and shows a general behavior of an operating point U relating to a scalable pattern. .

As described above, also with respect to the locus of the operating point U related to the scalable pattern, the locus M of the operating point P related to the standardizing function has an important role, and both are closely related through the scaling operation. It becomes clear.

Next, regarding the standardizing function and the function relating to the scalable pattern, the physical meaning of the area surrounded by the loci of the operating points will be described.

As described with reference to FIG. 4, when the motor shaft is rotated at a constant speed with a constant torque load applied to the motor, the effective value of the mechanical power is the torque-speed characteristic diagram. It is represented by the area of a quadrangle region surrounded by the locus drawn by the operating point and the acceleration axis and the velocity axis in the following, but in the following, the area within the locus drawn by the operating point will be examined when accelerating and decelerating. .

FIG. 12 shows θ _N ⁽²⁾ and Θ ⁽²⁾ on the horizontal axis,
The vertical axis represents θ _N ⁽¹⁾ and θ ⁽¹⁾ and the locus M and the locus Ma * b are shown together.

First, regarding the normalization function θ _N , FIG.
If the area within the locus M in the θ _N ⁽²⁾ -θ _N ⁽¹⁾ characteristic diagram of (indicated by the diagonal line to the lower left of the figure) is “s”, this is given by the following equation.

[0131]

(Equation 23)

Therefore, the effective current value of the controlled object when operating according to the acceleration / deceleration curve related to the standardization function is “i _RMS ”.
Then, this can be obtained by the following equation.

[0133]

(Equation 24)

That is, it can be seen that the effective current value i _RMS is proportional to the square root of the area s in the locus M.

Further, the same calculation as described above can be performed for the locus associated with the scalable pattern.
The area within the locus Ma * b (indicated by the slanting line on the lower right of the figure ^{) in} the Θ ^{(2) −} Θ ^{( 1)} characteristic diagram of is as “S”, and the effective current value related to the controlled object is “ _IRMS ”. Then, S becomes like the following formula.

[0136]

(Equation 25)

[Equation 20], [Equation 21], [Equation 2]
4] and [Equation 25], the effective current values I _RMS and i _RMS
The following relationship is found between and.

[0138]

(Equation 26)

That is, if the effective current value i _RMS related to the standardization function, that is, θ _RMS ⁽²⁾ is known, the effective current value I _RMS in the actual operation can be known.

Based on the relationship between the function related to the scalable pattern and the normalization function in this way, various physical quantities related to the property of the acceleration / deceleration curve are set to the normalization function regardless of the shape of the acceleration / deceleration curve. It can be reduced to such an amount.

Based on the above results, a method of generating a scalable pattern considering the limit characteristics of the controlled object will be described.

FIG. 13A shows the acceleration function Θ on the horizontal axis.
⁽²⁾ Taking (t) and taking the velocity function Θ ⁽¹⁾ (t) on the vertical axis, the locus Mt of the operating point U (Θ ⁽²⁾ , Θ ⁽¹⁾ ) relating to the motion of the controlled object is shown. is intended, the peak value of the "theta _P ^(2)" the acceleration theta ⁽²⁾ in the figure, "theta _P ^(1)" indicates respectively the peak value of the speed theta ^(1), these T _P And information about Δθ is included. Now, the question is how to determine the shape of the acceleration / deceleration curve when the controlled object is moved along the locus Mt. For the sake of simplicity of explanation, it is assumed that the acceleration / deceleration curve does not include a constant velocity part.
The locus Mt has symmetry between the first quadrant and the second quadrant of the Θ ^{(2) −} Θ ^{( 1)} diagram, that is, when the part of the locus Mt belonging to the first quadrant is folded back with respect to the Θ ⁽¹⁾ axis. It can be superposed on the part of Mt belonging to the second quadrant.
Even if such an assumption is made, the generality of the following discussion will not be impaired.

Since the velocity function Θ ⁽¹⁾ (t) is a function of the acceleration function Θ ⁽²⁾ (t), the locus Mt in the figure can be expressed by the following equation.

[0144]

[Equation 27]

"Θ _F ⁽¹⁾ " in the above equation is a functional.

When the above equation is written in a more understandable form, the following equation is obtained.

[0147]

[Equation 28]

The above equation is obviously a differential equation, and by solving it, a solution like the following equation can be obtained.

[0149]

(Equation 29)

The integration constant (constant in C or Ω (t)) in the above equation is determined by considering initial conditions and boundary conditions (related to Δθ, T _P, etc.). For example, from the initial condition Ω (t = 0) = 0, C = 0, and finally, the solution shown in the following equation is obtained, and when this is conceptually illustrated in FIG.
(B).

[0151]

[Equation 30]

From the above discussion, it becomes clear that the locus Mt defines the shape of the acceleration / deceleration curve.
We will supplement that understanding by presenting some concrete examples.

FIG. 14 is a Θ ⁽²⁾ -Θ ⁽¹⁾ diagram showing an example of the characteristic line showing the limit performance of the controlled object (see the limit TN line i2 in FIG. 1C). , The limit characteristic line L is Θ ⁽¹⁾ =
Θ _MAX ⁽¹⁾ line segment Lb, Θ ⁽²⁾ = Θ _MAX ⁽²⁾ line segment La,
These line segments are connected at points A and B, and include a line segment Lab inclined downward to the right. That is, the point A is the line segment L
It is the connection point between a and the line segment Lab, and its coordinate value is (Θ _MAX
⁽²⁾ , Θ _ap ⁽¹⁾ ), and the time when the operating point passes through this point A is t = T _a . Further, the point B is a line segment Lab and a line segment Lb.
And the coordinate value is (Θ _b ⁽²⁾ , Θ _MAX ⁽¹⁾ )
And the time when the operating point passes through this point B is t = T _b .

The line segment Lab has its slope (this is "-T _m "
And ) Or the intercept on the Θ ⁽¹⁾ axis (this is referred to as “Θ _c ⁽¹⁾ ”) is determined by giving the coordinate values of points A and B, so the limit characteristic line L is expressed by the following equation. be able to.

[0155]

(Equation 31)

The above equation corresponds to the equation [27], and the following equation is obtained from the integration operation of the first equation in the equation [31].

[0157]

(Equation 32)

Further, the solution of the differential equation using the Laplace transform can be used for the second equation in the equation [31], and the following equation is obtained after the transformation.

[0159]

[Equation 33]

The Laplace expression of each quantity was specified by adding (s) to the function using the parameter s (this s is independent of the above-mentioned area s).

From the equation [32], Θ ⁽¹⁾ (s) can be obtained by the following equation.

[0162]

(Equation 34)

By applying the inverse Laplace transform to the above equation, Θ ⁽¹⁾ (t) can be obtained as the following equation.

[0164]

(Equation 35)

This equation corresponds to the above-mentioned [Equation 29], and by imposing the initial condition of the following Equation [Equation 36], the solution corresponding to the [Equation 30] is eventually obtained as [Equation 37].
You can get the formula.

[0166]

[Equation 36]

[0167]

(37)

Next, _letting T _{PMAX be the} time for the speed Θ ⁽¹⁾ to reach its maximum value Θ _MAX ⁽¹⁾ , t =
By substituting T _PMAX , the following formula [Formula 38] is obtained, and from this, T _PMAX is obtained as shown in the formula [Formula 39].

[0169]

(38)

[0170]

[Equation 39]

Summarizing the results so far, the speed function is expressed by the following equation.

[0172]

(Equation 40)

Therefore, to obtain the position function from this,
The following expression [expression 41] is obtained by integrating the first expression of the expression [40] (however, the integration constant is set to zero from the initial condition), and the second expression of the expression [40] is integrated. (However, the integration constant is set as “C”.) By the following expression [Formula 42] is obtained.

[0174]

[Equation 41]

[0175]

(Equation 42)

In determining the integration constant C, the following equation [Equation 4
3], the boundary condition at t = T _a may be used.

[0177]

[Equation 43]

By summarizing the above results, the formula [44] is obtained for the velocity and the formula [45] is obtained for the position, which is shown in FIG.

[0179]

[Equation 44]

[0180]

[Equation 45]

In this example, it is assumed that the locus L has symmetry between the first quadrant and the second quadrant, so that the acceleration part and the deceleration part of the acceleration / deceleration curve have symmetrical shapes. Of course, if the assumption is not satisfied, the solution for the speed reducer can be obtained by the same procedure as described above. It is also easy to generalize the above discussion to the case where the acceleration / deceleration curve includes the constant velocity part (since the operating point of the constant velocity part is only at the Θ ⁽¹⁾ axis).

FIG. 16 is a Θ ⁽²⁾ -Θ ^{( 1)} diagram showing another example of the characteristic line showing the limit performance of the controlled object, and the limit characteristic line L'is a straight line sloping downward to the right. It consists of line segments only. That is, in the limit characteristic line L shown in FIG. 14, a point A (Θ _MAX ⁽²⁾ , 0) is set and a point B (0, Θ _MAX ⁽¹⁾ ) is set.
Is the limit characteristic line L '.

Therefore, the condition shown in the following equation [Equation 46] is
By substituting into the equations [44] and [45], the solutions shown in the equations [47] and [48] can be obtained without resolving the differential equations.

[0184]

[Equation 46]

[0185]

[Equation 47]

[0186]

[Equation 48]

FIG. 17 is a Θ ^{(2 )} -Θ ⁽¹⁾ diagram showing another example of the characteristic line showing the limit performance of the controlled object, and the limit characteristic line L ″ is Θ ⁽²⁾ = θ. Line segment of _MAX ⁽²⁾ and Θ ⁽¹⁾
= Theta consists of a segment of _{MA X} ^(1), a trace corresponding to the triangular acceleration and deceleration curve, as described in FIG. 4 and FIG. 28,
It is expressed by the formula [49].

[0188]

[Equation 49]

Therefore, if this is solved, the solutions shown in the equations [50] and [51] can be obtained, but the intermediate calculations are omitted because they are not important (acceleration / deceleration in FIGS. 4 and 28). This supplements the explanation of the correspondence between curves and trajectories on the TN diagram.)

[0190]

[Equation 50]

[0191]

(Equation 51)

From the above, it is understood that the shape of the acceleration / deceleration curve is defined by the characteristic line (that is, the locus) that defines the performance of the controlled object.

Therefore, for the motion of the controlled object, since it is not allowed or dangerous to exceed the limit performance of the controlled object, the acceleration / deceleration curve is defined by the limit performance characteristic line of the controlled object, that is, If the standardization function related to the scalable pattern is specified from the limit performance characteristic line of the control target, it is possible to control so that the operating point or locus corresponding to the acceleration / deceleration pattern is always located within the range of the limit performance characteristic line. it can.

That is, T _PMAX (maximum peak velocity arrival time) and Δθ _MAX (= Θ (T _PMAX )) always exist for the controlled object, and these are bounded by the boundary as shown in the equation [39]. Since it is uniquely determined by the condition, the acceleration / deceleration pattern corresponding to the operation when the controlled object is drawn to the limit of its capability may be used as the reference of the scalable pattern.

Specifically, if the position function related to such an acceleration / deceleration pattern is set to "Θ _MAX (t)", the normalization function θ _N related to the scalable pattern is defined by the following equation.

[Equation 52]

Therefore, if T _P for Δθ (≦ Δθ _MAX ) is determined, the time axis expansion / contraction parameter α is determined as shown in the following formula [Equation 53], and a function related to the scalable pattern, for example, the position function Θ (T) is determined as shown in the following equation [Equation 54].

[0197]

(Equation 53)

[0198]

(Equation 54)

Now, recall the arbitrariness of the selection of the peak velocity arrival time T _Pn related to the normalization function.

FIG. 18 shows a scalable pattern in which the limit performance of the controlled object is taken into consideration. The change of the acceleration / deceleration curve and the corresponding Θ ⁽²⁾ −Θ ⁽¹⁾
It is a conceptual illustration of a correspondence relationship between a change in the locus of the operating point on the diagram.

The acceleration / deceleration curve AL shown in the figure corresponds to the limit characteristic line L of the controlled object, and the acceleration / deceleration curve al has a certain Δθ (<
_Acceleration / deceleration curves for Δθ _MAX ) and T _P (<ΔT _PMAX ), that is, obtained by performing a scaling operation on the normalization function of [Equation 52],
The characteristic line corresponding to this is l. In the drawing, the correspondence between each part of the acceleration / deceleration curve and each part of the characteristic line is indicated by the symbols.

Therefore, it is concluded that the acceleration / deceleration curve generated as the scalable pattern does not deviate from the range in which the corresponding operating point is limited by the limit characteristic line L of the controlled object.

[0203] Incidentally, as can be seen from the number 44] where the acceleration and deceleration curve t = 0, T _a, or no longer be guaranteed continuity of acceleration in T _{PMA X,} or the control target limit characteristic line shape It is necessary to take some measures to deal with the case where is not as simple as the above example, and the following two methods will be described.

The first is the "low-pass filter method",
For example, as shown in FIG. 19, the acceleration / deceleration curve corresponding to the equation [44] is passed through a low-pass filter LPF having a predetermined cutoff frequency, and its output (in FIG.
_o ⁽¹⁾ (t) ”, and the peak velocity arrival time is“ T
_P ′ ”and the maximum speed is“ Θ _MAX ⁽¹⁾ ′ ”. ) Is given to the controlled object as a control signal. That is, since a point in the acceleration / deceleration curve having a problem of continuity of acceleration or a portion near the connection point corresponds to a portion including a high frequency component, the low-pass filter LPF facilitates the connection of the components of the acceleration / deceleration curve. can do. By this, Θ ⁽²⁾ corresponding to the output of the low pass filter LPF
The locus Lo on the −Θ ⁽¹⁾ diagram is located inside the limit characteristic line L as shown by the broken line in FIG. Note that the filter configuration can be generally an n-th (n is a natural number) low-pass filter, and its order and frequency characteristics (pole arrangement) can be determined so as to match the frequency characteristics of the control target. . For example, when the virtual servo system as described above is built in the control device, the servo parameters and time constants related to the virtual servo system are fixed according to the parameters and time constants related to the actual servo system. Thus, the virtual servo system can be used for the low-pass filter, and in this case, control can be performed so that T _P ′ ≧ T _PMIN .

The second method is the "inscribed method", which is shown in FIG.
As shown in FIG. 1, when the limit characteristic line L of the controlled object is a complicated curve, it is approximated by a polygon (a side of a polygon) LL inscribed in the curve, as shown by a chain double-dashed line. Is the way to do it. That is, if the function Θ _F ^{(1) of the} equation [27] is simple, the differential equation can be solved by the above procedure, but if the function is complicated, some approximation is required. . Therefore, the accuracy of the approximation can be improved by increasing the n value of the n-sided polygon inscribed in the limit characteristic line L as an approximate function of the target function.

If there is no corner in the limit characteristic line L, it is possible to obtain a smooth acceleration / deceleration curve, the low-pass filter method is unnecessary, and the shape of the limit characteristic line L is simple. For example, even if it is a curve, it is sufficient if it can be expressed easily by an analytical expression. Since these methods are complementary, there is no practical problem due to the combination of these methods. An acceleration / deceleration curve can be obtained. That is, when the limit characteristic line L is complicated, the inscribed method is required to obtain the approximate expression, and the low-pass filter method is required to smoothly connect the approximate characteristic lines.

From the above discussion, however, the shape of the acceleration / deceleration curve is determined by the items a) and b) above, that is, the limit performance of the controlled object. It is clarified that it is generated by a unified algorithm of performing a scaling operation on the digitized function, and based on these, the configuration of the servo control device according to the present invention can be understood.

FIG. 22 shows the structure of the servo control device. The servo control device 1 includes a command means 2, an acceleration / deceleration pattern generation means 3, a T _P (peak speed arrival time) value determination means 4, and an α (time). A shaft expansion / contraction parameter) value determining means 5 and a low-pass filter 6 are included.

That is, the signal (position command, etc.) generated by the command means 2 is sent to the acceleration / deceleration pattern generation means 3 and the T _P value determination means 4
Sent to

The T _P value determining means 4 defines the relationship between the position command Δθ and the T _P value, and T _P corresponding to the Δ _P
The value is determined, and this value is sent to the α value determining means 5 in the subsequent stage.

[0211] For example, as shown in FIG. 23 (a), taking the [Delta] [theta] on the horizontal axis, taking the T _P on the vertical axis, points corresponding respectively to the lower limit and the upper limit of the T _P value MIN ([Delta] [theta] _MIN,
T _PMIN ) and the point MAX (Δθ _MAX , T _PMAX ) are determined as points unique to the controlled object.

Therefore, for example, considering a characteristic in which a point MIN and a point MIN are connected by a straight line as shown in the figure, 0 is obtained.
In the range of ≦ Δθ ≦ Δθ _MIN , T _P = T _PMIN , Δθ _MIN ≦ Δ
In the range of θ ≦ Δθ _MAX , the T _P value (where T _PMIN ≦ T _P ≦ T _PMAX ) according to the inclination of the straight line is obtained, and in the range of Δθ ≧ Δθ _MAX , T _P = T _PMAX is obtained.

The shape of the line connecting the points MIN and MAX depends on what is used as the evaluation function. For example, as shown in FIG. 5, when the effective value _I P _RMS of electric power is targeted for evaluation and control is performed to keep it constant, as shown in the formula [6], T _P is proportional to the square root of Δθ. At that time, the acceleration / deceleration curve has been described as having a triangular shape, but if the effective value _I P _RMS is associated with the effective current value, the effective current value becomes constant.
When the effective current value I _RMS is made constant in the formula [26] and the effective current value i _RMS is determined by the normalization function, a general result that α = T _P / T _Pn is proportional to the square root of Δθ Leads to.

Further, FIG. 23 (a) is generalized to the case where the acceleration / deceleration curve includes a constant velocity part, as shown in FIG. 23 (b). That is, FIG. 23B shows Δθ on the horizontal axis and T _all (= 2 · T _P + T _c ) on the vertical axis, and the characteristics are as shown by the solid line, and 0 ≦ Δθ ≦ In the range of Δθ _MIN , T _all = 2 · T _PMIN , and in the range of Δθ _MIN ≦ Δθ ≦ Δθ _MAX , the T _all value (however, 2 · T
_{In the range of PMIN} ≦ T _all ≦ 2 · T _PMAX ), Δθ ≧ Δθ _MAX , the T _P value according to the straight line having the slope of 1 / Θ _MAX ⁽¹⁾ is determined. Note that for the slope value 1 / Θ _MAX ⁽¹⁾ , the amount of movement of the constant velocity part of the acceleration / deceleration curve is Θ _MAX ⁽¹⁾ · T _c .

In FIG. 23 (b), when the dynamic range of the controlled object is large, the characteristic line is extremely close to the point MIN and the point MAX as shown by the alternate long and short dash line in the figure. In some cases, the two points coincide as shown by the chain double-dashed line (that is, T _P until this point is reached)
Is constant, but if it exceeds this, the T _P value becomes the maximum speed Θ _MAX.
^It depends on ^{( 1)} . ) It will be.

In this way, the T _P value determining means 4 sets Δ
When the T _P value corresponding to θ is determined, the α value determining means 5 determines the value of the time axis expansion / contraction parameter α based on this. However, as can be seen from “α = T _P / T _Pn ”, the division is performed only once. In addition, [Equation 52],
As shown in the formula [53], the standardizing function determined by the limit characteristic line is set to "T _Pn = T _PMAX ", but the selection of T _Pn is generally arbitrary.

The acceleration / deceleration pattern generation means 3 is the command means 2
The acceleration / deceleration pattern is generated based on the command signal from the .alpha. The operation here is nothing but the scaling operation by Δθ or α based on the normalization function, as described in FIG. That is, since the acceleration / deceleration pattern generation means 3 has a standardization function corresponding to the limit performance of the controlled object (see [Equation 52], [Equation 54], etc.), it is possible to generate a scalable pattern by a scaling operation. Yes, movements that follow the pattern never deviate from the critical performance of the controlled object.

When the low-pass filter method is used, the output of the acceleration / deceleration pattern generating means 3 is sent to the servo system 7 to be controlled after passing through the low-pass filter 6, but in some cases, it is 1 in FIG. As shown by the dotted line,
It is sent to the servo system 7 as it is. The “servo system” here means a wide system including a servo circuit and the like in addition to a drive source and / or a drive mechanism to be controlled. Further, in FIG. 22, the portion excluding the servo system 7 is a portion that can be realized by software processing in the control device.

The procedure of the servo control method according to the present invention is summarized as shown in FIG. First, step S1
The critical performance of the controlled object is specified in. For example, in the case of a motor system, the limit TN line as shown in FIG. 1 is specified from the specifications provided by the manufacturer, but in some cases, an experiment or theory for determining the limit performance of the controlled object. It can be specified by calculation.

In the next step S2, the shape of the acceleration / deceleration curve is determined based on the limit characteristic line of the controlled object. That is, as shown in the equations [27] to [30], this is nothing but solving the differential equation corresponding to the limit characteristic line.
As a result, the normalization function related to the scalable pattern is determined from the limit performance characteristic line of the controlled object.

Then, in step S3, the Δθ-T _P characteristic is defined. As a result, T corresponding to the command value Δθ
The values of _P and α can be determined (see FIG. 23).

In the next step S4, the scaling operation for Δθ and α is performed on the standardization function to generate a scalable pattern. That is, step S
In the acceleration / deceleration pattern obtained by the scaling operation (see FIG. 8) for the standardization function determined in 2, the locus of the operating point that moves on the characteristic diagram accordingly is limited by the limit performance characteristic line of the control target. It will always be located within the range without departing from it. It should be noted that, although it is a stumbling block, with respect to the scaling law for α, the time dimension such as T _P is the α multiplication rule, the velocity dimension is the α times multiplication rule, and the acceleration and current dimensions are the square of α. Obey the 1x rule of.

Then, in step S5, the control signal related to the acceleration / deceleration pattern is sent to the control target as it is or via the low-pass filter 6.

In the actual control operation of the servo controller 1, steps S1, S2 and part of S3 are already determined at the design stage.
Steps S3 to S5 according to the command from the command means 2
It should be noted that the motion control of the controlled object is performed by repeating (however, the regulation of the Δθ-T _P characteristic in step S3 is excluded).

Further, in order to generate the scalable pattern in step S4, it is necessary to prepare various amounts including the standardizing function as a database in advance. For example, taking the acceleration / deceleration patterns shown in [Equation 44] and [Equation 45] as examples, the standardization function has a shape as shown in the upper part of FIG. 15, and T _aMAX , T _PMAX or T
_PMAX - _TaMAX , θ _NP ⁽¹⁾ (peak value of normalized speed θ _N ⁽¹⁾ ) is required. Further, the calculation of the effective current value I _RMS is performed by θ _NRMS ⁽²⁾ (this is obtained by replacing T _PMIN with T _PMAX in the fourth equation of [Equation 21]), the inertia moment J, and the torque coefficient K _t. A constant related to the ability of the controlled object is required. Then, the expression of the normalization function ([Equation 4
5] See formula. Various amounts of contained) (theta _c and T _m, etc.) can of course necessary. The acceleration / deceleration pattern generation means 3 in the servo control device 1 holds data relating to these amounts.

In the above description, it is assumed that the performance of the controlled object is not affected by the influence of the surrounding environment (temperature etc.).
The performance of an actual controlled object is usually influenced by the outside air temperature and the like. In other words, the rated current value of the controlled object regulates the amount of heat generated, which causes the temperature of the controlled object to rise, and the sum of the temperature increase and the outside air temperature becomes the temperature of the controlled object and its capacity is increased. Because it affects. Therefore, in order to respond to such changes in the performance of the controlled object due to temperature, the outside air temperature and the temperature of the controlled object are constantly detected in order to generate a scalable pattern that corresponds to the limit performance of the controlled object, and It is practical to specify the limiting characteristic line and the rated characteristic line and generate an acceleration / deceleration pattern corresponding to them (see the temperature detecting means shown in FIG. 22).

[0227]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An example of application of the present invention to a robot controller will be described below with reference to FIGS.

FIG. 25 schematically shows a main part of the hardware configuration of the robot controller 8.

The robot controller 8 has a CPU (central processing unit) 9 as a core, and a storage unit 11 directly or indirectly connected to a bus 10 thereof.

The storage unit 11 stores, for example, a robot program related to the description of the operation procedure of the robot, a system program related to the description of the robot system (for example, description of coordinate conversion for determining the operation of each axis of the robot), and the like. Will be remembered.

The servo units 12, 12, ... Are independent for each drive axis that constitutes the robot 13,
It is provided to control actuators (motors, cylinders, etc.) that drive the axes of the robot 13 in accordance with commands issued from 9. The teaching pendant 14 according to the operation instruction of the robot 13 is the T. P. (Abbreviation of teaching pendant) It is connected to the bus 10 via an interface 15.

FIG. 26 shows a horizontal multi-joint type robot having four axes as an example of the configuration of the robot 13 to be controlled.

That is, the robot 13 is provided with the base portion 16, the first arm 17 and the second arm 18, and the first arm 1
7 is attached to the base shaft portion 16 in a state in which one end portion of 7 can rotate. Then, one end of the second arm 18 is rotatably attached to the other end of the first arm 17, and a tool mounting shaft 19 is provided to the other end of the second arm 18. .

A motor 20 is provided on the base shaft portion 16 for rotating the first arm 17, and the driving force of the motor 20 is transmitted via the harmonic reducer 21 to the first arm 1.
It is designed to be transmitted as the turning force of 7.

The first arm 17 is provided with a mechanism for rotating the second arm 18. That is, the motor 23 and the harmonic reducer 24 are attached to the support portion 22 fixed to one end of the first arm 17, and the pulley 25 is attached to the output shaft of the harmonic reducer 24. The pulley 26 paired with this pulley 25 is the second
The pulley 2 is fixed to the rotating shaft 18a of the arm 18.
A belt 27 is stretched between the pulley 5 and the pulley 5. That is, the driving force of the motor 23 becomes the rotational force of the pulley 25 via the harmonic speed reducer 24, and this is transmitted to the pulley 26 by the belt 27 and becomes the rotational force of the second arm 18.

The tool mounting shaft 19 is slid in the vertical direction by an actuator 28 provided on the upper surface of the rotating end of the second arm 18, and at the end of the second arm 18 near the rotating shaft 18a. The rotational force of the provided motor 29 is transmitted to the tool mounting shaft 19 by a belt or the like so that the motor 29 is rotated about the shaft. And
A tool mounting portion 30 is fixed to the lower end of the tool mounting shaft 19 so that an effector such as a hand or a tool can be attached.

[0237] As described above, the robot 13, the rotary shaft of the first arm 17 (theta ₁ axis), pivot axis of the second arm 18 (theta _2-axis), the rotation of the tool mounting portion 30 shaft (R.theta Axis and sliding axis (Rz axis).

Therefore, when applied to the robot controller 8, it is necessary to construct the configuration of the servo controller 1 shown in FIG. 22 for each drive axis of the robot. That is, FIG.
It can be considered that the servo system 7 shown in FIG. 1 corresponds to one of the drive axes of the robot, and the servo control means related to robot control is functionally connected to the CPU 9, the storage unit 11, the servo unit 12, the teaching pendant 14, and the like. Will be realized.

However, in the case of the multi-axis configuration like the robot 13, the PTP (Point Toe) for each drive axis is
In consideration of the necessity of synchronizing the start and stop of the Point operation, for example, the maximum value is selected from among the α values of the drive axis. That is, when the robot is composed of M drive axes, the i-th (i = 1, 2, ...
., M) is the α value related to the drive axis of “α _i ” (i = 1, 2,
, M), the α value common to all drive shafts (this is referred to as “α _CMN ”) may be determined by the following equation.

[0240]

[Equation 55]

The function max (X1, X2, ...
..) is a function for extracting the maximum value of the variables X1, X2, ...

In this case, the position function (eg, phase angle in the case of a motor) of the acceleration / deceleration pattern related to the i-th drive shaft is set to "Θ _i " (i = 1, 2, ..., M). , Their normalization function is “θ _Ni ” (i = 1, 2, ..., M)
Then, the relation of the following formula is obtained.

[0243]

[Equation 56]

The standardizing function is generally different for each drive axis. However, if the standardizing function for each drive axis can be unified into one function, The configuration is simple. This is because, in this case, the standardization function for each drive axis has a proportional relationship with the standardized standardization function, so that the scaling operation for one standardization function adds all drive axes. This is because the deceleration pattern can be generated.

However, since the relationship between Δθ and TP generally differs for each drive axis, such a robot has the largest value of α _i · T _PMAXi for each drive axis as shown in the following equation. Should be selected.

[0246]

[Equation 57]

Therefore, when the servo control device according to the present invention is applied to a robot control device, each drive is performed in accordance with an acceleration / deceleration pattern that does not deviate from the limit performance of the servo system related to the drive axis that constitutes the robot. Since the axis can be controlled, it is possible to stabilize robot operation control (for example, measures against runaway and reduce failures), and to speed up work by maximizing the capabilities of the servo system. Can be planned. For example, in the case of a robot that uses a pulse (stepping) motor for the servo system, by using the pulse motor to its limit, the speed of the robot operation can be increased, and further, according to the present invention. The operation according to the deceleration pattern does not force the robot to perform unreasonable work that exceeds the limit performance of the pulse motor, and, for example, eliminates the risk of causing step out of the pulse motor.

[0248]

As is apparent from what has been described above, according to the inventions of claims 1 and 2, by generating an acceleration / deceleration curve that does not deviate from the limit performance of the controlled object, The performance and quality of the controlled object can be guaranteed. In addition, it is possible to bring out the capability of the controlled object to its limit performance, and, regardless of the shape of the acceleration / deceleration curve specified by the limit performance of the controlled object, the function value of the reference function It is possible to generate the acceleration / deceleration curve to be obtained by the scaling operation or the expansion / contraction operation in the time axis direction, and it is possible to unify the algorithms for generating the acceleration / deceleration curve.

According to the third aspect of the present invention, the connectivity of the acceleration / deceleration curve can be compensated by sending the output of the acceleration / deceleration pattern generating means to the controlled object via the low-pass filter.

According to the invention of claim 4, in determining the peak speed arrival time corresponding to the designated value of the servo control variable, the condition that the effective current value related to the controlled object is made constant is imposed, whereby the operation control is performed. By keeping the effective current value at constant, the ability of the controlled object can be stabilized.

According to the inventions of claims 5 and 6,
By maximizing the capacity of the servo system related to the robot arm and tool mounting part, the work speed is increased.
Moreover, the robot operation can be stabilized without deviating from the limit performance characteristics of the servo system. Further, regardless of the structure of the robot, a unified control algorithm can be used for each servo system.

[Brief description of the drawings]

FIG. 1 is a schematic graph showing some examples of characteristic lines indicating a limit performance and a rated performance related to a motor system.

FIG. 2 is a diagram showing an example of an equivalent circuit of a servo motor.

FIG. 3 is a diagram showing several triangular acceleration / deceleration curves.

FIG. 4 is a diagram for explaining an effective value of mechanical power and an effective value of electrical power corresponding to the acceleration / deceleration curve of FIG.

FIG. 5 is a diagram showing Δθ-T _P characteristics in control in which the effective value of electric power is constant.

FIG. 6 is an explanatory diagram of a shape of a standardized speed function.

FIG. 7 is a graph showing a ramp signal.

FIG. 8 is an explanatory diagram of a scaling operation related to a normalized speed, where (a) performs only a scaling operation related to a function value, and (b) performs only a time-axis expansion operation with respect to the normalized speed. In the above case, (c) shows the case where the combined operation of (a) and (b) is performed.

FIG. 9 is a diagram for explaining a locus on a θ _N ⁽²⁾ −θ _N ⁽¹⁾ diagram related to a normalization function, and (a) is θ _N ⁽²⁾
-Θ _N ⁽¹⁾ diagram, (b) t-θ _N ⁽¹⁾ diagram, (c) t-
It is a θ _N ⁽²⁾ diagram.

FIG. 10: Regarding the trajectory on the θ _N ⁽²⁾ −θ _N ⁽¹⁾ diagram regarding the normalization function About the example where the functional relationship between θ _N ⁽²⁾ and θ _N ⁽¹⁾ is not a monovalent function FIG.

FIG. 11 is a diagram for explaining the relationship between the trajectory on the θ _N ⁽²⁾ −θ _N ⁽¹⁾ diagram regarding the normalization function and the scaling operation regarding the normalization speed, and (a) is a function value. When only the scaling operation related to is performed, (b) is a time axis extension operation with respect to the normalized speed, (c) is (a)
The respective cases where the operations of (1) and (b) are combined are shown.

[Figure 12] θ _N ⁽²⁾ relating to the normalized function - [theta] _N ⁽¹⁾ line and the area of the region surrounded by the trajectory in diagram, theta according to variable scaling pattern ^{(2) -Θ (1)} diagram It is a figure which also shows the area in the area | region enclosed by the above locus.

13A is a diagram for explaining a trajectory on a Θ ⁽²⁾ -Θ ⁽¹⁾ diagram relating to a controlled object, and FIG. 13B is an acceleration / deceleration defined by the trajectory of FIG. 13A. It is a figure for explaining a curve.

FIG. 14 is a Θ ⁽²⁾ -Θ ⁽¹⁾ diagram showing an example of a characteristic line showing the limit performance of the controlled object.

15 is a graph diagram for explaining an acceleration / deceleration pattern defined by a characteristic line of FIG.

16 is a Θ ⁽²⁾ -Θ ⁽¹⁾ diagram showing a characteristic line different from that in FIG. 14;

FIG. 17 is a Θ ⁽²⁾ -Θ ⁽¹⁾ diagram showing yet another characteristic line.

FIG. 18 is a diagram for explaining a correspondence relationship between an acceleration / deceleration curve and a locus on a Θ ⁽²⁾ −Θ ⁽¹⁾ diagram.

FIG. 19 is an explanatory diagram of a low pass filter method.

FIG. 20 shows Θ ⁽²⁾ − when the low-pass filter method is used.
FIG. 3 is a schematic diagram for explaining a trajectory on a Θ ⁽¹⁾ diagram.

FIG. 21 is an explanatory diagram of an inscribed method.

FIG. 22 is a block diagram showing a configuration of a servo control device according to the present invention.

FIG. 23 (a) shows an example of a Δθ-T _P diagram,
(B) shows an example of a Δθ-T _all diagram.

FIG. 24 is a flow chart diagram for explaining a servo control method according to the present invention.

25 shows an example of application to a robot controller together with FIG. 26, and this figure is a diagram showing an example of the hardware configuration of the robot controller.

FIG. 26 is a schematic side view showing a robot with a four-axis configuration by cutting out a part thereof.

FIG. 27 is a diagram showing a triangular acceleration / deceleration curve.

28 is a diagram showing a locus on the TN diagram corresponding to the acceleration / deceleration curve in FIG. 27 and a limit TN line.

[Explanation of symbols]

 1 Servo control device 2 Command means 3 Acceleration / deceleration pattern generation means 4 Peak velocity arrival time value determination means 6 Low pass filter 5 Time axis expansion / contraction parameter value determination means 8 Robot control device 13 Robots 17, 18 Arms 30 Tool mounting part

Claims (6)

[Claims] 1. A servo control device for performing acceleration / deceleration control corresponding to a limit performance of a controlled object, the command means issuing a command value for a servo control variable, a command value related to the servo control variable, and an acceleration For defining the relationship with the time until the speed reaches the peak value in the deceleration curve (hereinafter referred to as "peak speed arrival time"), and determining the value of the peak speed arrival time from the command value related to the servo control variable. Peak velocity arrival time value determining means, time axis extension / contraction parameter value determining means for determining the dimensionless parameter for time axis extension / contraction parameter defined by dividing the peak velocity arrival time by its reference value, and acceleration (Or force or torque) and speed (or speed, angular velocity or number of revolutions) is the speed of the acceleration (or force or torque) of the controlled object. Or speed,
The shape of the acceleration / deceleration curve is defined under the condition that the acceleration / deceleration curve is on the limit characteristic line on the characteristic diagram for the angular velocity or the number of revolutions. As a reference value of, the acceleration / deceleration curve is generated and sent to the controlled object by the expansion / contraction operation in the time axis direction for the reference function based on the time axis expansion / contraction parameter value and the scaling operation for the reference function value based on the command value from the command means. A servo control device, comprising: 2. A servo control method for performing acceleration / deceleration control corresponding to the limit performance of a controlled object, wherein the speed (or speed, angular velocity or rotation) of acceleration (or force or torque) is first controlled for the controlled object. Number) is specified, the operating point defined by acceleration (or force or torque) and speed (or speed, angular velocity or rotation speed) is placed on the limit characteristic line related to the controlled object. Under the condition that the acceleration / deceleration curve is defined, its reference function and the peak speed arrival time related to the reference function are specified, and the relationship between the specified value of the servo control variable and the peak speed arrival time is specified, and the command Obtain the value of the peak velocity arrival time corresponding to the value, and divide this by the peak velocity arrival time according to the reference function The acceleration / deceleration curve is determined by deciding the value and performing the expansion / contraction operation in the time axis direction for the reference function based on the time axis expansion / contraction parameter value and performing the scaling operation for the reference function value based on the command value of the servo control variable. A servo control method, wherein the servo control method is characterized in that it is generated and transmitted as a control signal to a control target. 3. The servo control device according to claim 1, wherein a low-pass filter is provided between the acceleration / deceleration pattern generation means and the controlled object. 4. The servo control method according to claim 2, wherein the peak speed arrival time corresponding to the command value of the servo control variable is determined by imposing a condition that the effective current value related to the controlled object is constant. A servo control method characterized in that a relationship between a command value and a peak speed arrival time is defined for a servo control variable. 5. A robot control device for controlling the operation of a robot, which comprises a plurality of arms and / or tool mounting parts, and a plurality of servo systems including drive sources for the arms or tool mounting parts and their drive circuits, The servo control means for controlling each servo system defines the relationship between the command means for issuing a command value for the servo control variable and the command value for the servo control variable and the peak speed arrival time for the acceleration / deceleration curve. Peak velocity arrival time value determining means for determining the value of the peak velocity arrival time from the command value related to the control variable, and dimensionless time axis expansion / contraction defined by dividing the peak velocity arrival time by its reference value. Time axis expansion / contraction parameter value determining means for determining the value of the parameter for acceleration, acceleration (or force or torque) and speed (or speed, angular velocity or speed) Under the condition that the operating point defined by (number) is on the limit characteristic line on the characteristic diagram for acceleration (or force or torque) speed (or speed, angular velocity or rotation speed) related to the servo system. The shape of the acceleration / deceleration curve is defined, and the peak velocity arrival time value related to the reference function of the acceleration / deceleration curve is used as the reference value of the time axis extension / contraction parameter, and the extension / contraction in the time axis direction with respect to the reference function based on the time axis extension / contraction parameter value. A robot controller, comprising: an acceleration / deceleration pattern generation means for generating an acceleration / deceleration curve by a scaling operation with respect to a reference function value based on a command value from an operation and command means and sending it to a servo system. 6. A robot control method relating to motion control of a robot, comprising: a plurality of arms and / or tool mounting parts; and a plurality of servo systems including drive sources for the arms or tool mounting parts and drive circuits therefor, When controlling each servo system, first, specify the limit characteristic line related to the speed (or speed, angular velocity or rotation speed) of acceleration (or force or torque) of the servo system, and then accelerate (or force or torque). And the speed (or speed, angular velocity, or number of revolutions), the operating point defined by the acceleration / deceleration curve shape, its reference function and the reference function The peak speed arrival time is specified, the relationship between the specified value of the servo control variable and the peak speed arrival time is specified, and the peak speed arrival corresponding to the command value is reached. The value of time is obtained, and the value of the dimensionless time-axis expansion / contraction parameter is determined by dividing this by the peak velocity arrival time value related to the above-mentioned reference function, and based on the time-axis expansion / contraction parameter value, The acceleration / deceleration curve is generated by performing expansion / contraction operation in the time axis direction and performing a scaling operation on the reference function value based on the specified value of the servo control variable, and this is sent as a control signal to the servo system. A robot control method characterized by the above.