JP7600885B2 - Automated guided vehicle control device - Google Patents

Description

The present invention relates to a lift-up type automated guided vehicle (AGV), which lifts a cart from below for transport, and in particular to a lifter control device that enables ground transport, in which the wheels of the cart remain on the ground.

Conventional techniques for ground transportation of AGVs include the following. In Patent Document 1 and Patent Document 2, ground transportation is achieved by pressure control using a pneumatic actuator. In Patent Document 3, ground transportation is achieved by controlling the height of an electric actuator.

Patent No. 6641994 (Patent application 2015-255945) Patent No. 6848651 (Patent application 2017-87864) JP 2020-77295 A

Patent documents 1 and 2 use pneumatic actuators, which have problems such as frequent maintenance, difficulty in high-precision control, and noise. Patent document 3 uses electric actuators to control the height, but the wheels of the cart may lift off the ground on uneven, undulating ground or on slopes.

In this invention, we consider an approach using force control of an electric actuator to deal with undulations and tilts. While the first force control that can be considered is using a force sensor, a sensorless method is preferable in terms of cost. When trying to achieve highly accurate force control without a sensor, friction in the actuator becomes an issue. It is difficult to distinguish between reaction force and friction using only the reaction force estimate by an observer.

The present invention aims to solve the above problems, and its purpose is to provide a control device for an automated guided vehicle that can control thrust without a force sensor.

In order to solve the above problem, a control device for an automated guided vehicle according to claim 1 comprises:
A base mounted on a movable traveling body;
First to fourth actuators are respectively installed at the four corners of the base, and the output shafts of the actuators are configured to be movable in a direction perpendicular to the surface on the base where the actuators are installed ;
a controlled body disposed opposite to a tip end of each of the output shafts of the first to fourth actuators;
a controller that controls motors for driving the first to fourth actuators;
The controller has a thrust-to-additional torque conversion unit that converts a thrust command into an added torque using a conversion function of thrust command to added torque, and an adder unit that adds a maximum static friction torque to the converted added torque to output a motor torque reference value, and is characterized in that an output of the adder unit is used as a motor torque reference value for the motor for driving the actuator.

The control device for an automated guided vehicle according to claim 2 is the control device for an automated guided vehicle according to claim 1,
The controller is characterized by performing lifter ascent control based on a motor torque reference value using a maximum static friction torque in the positive direction, thrust control based on a motor torque reference value obtained by adding the additional torque and the maximum static friction torque, and lifter descent control based on a motor torque reference value using a maximum static friction torque in the negative direction.

The control device for an automated guided vehicle according to claim 3 is the control device for an automated guided vehicle according to claim 1 or 2,
The conversion function of the thrust command to the additional torque in the thrust-additional torque conversion unit is expressed as follows:
a first identification process in which the motor torque reference value is gradually increased from zero, the motor torque reference value when the actuator starts to move is recorded as a maximum static friction torque, and a value measured by a force sensor of a reaction force when the actuator is pressed against a wall surface with this torque is also recorded;
a second identification process is executed to determine a relationship between the maximum static friction torque recorded in the first identification process plus an additional torque as a motor torque reference value, the reaction force measured by a force sensor when the actuator is pressed against a wall surface, and the additional torque;
The relationship between the reaction force measurement value and the additional torque obtained in the second identification process is converted into a conversion function of the thrust command versus additional torque and set.

The control device for an automated guided vehicle according to claim 4 is the device according to any one of claims 1 to 3,
The controller:
a position controller that controls the position response value of each actuator output shaft, obtained by variable-transforming the angle of the actuator drive motor when the first to fourth actuators are each raised at a constant speed, so that the position response value becomes a command value for the actuator length, and outputs a force reference value for the actuator;
a disturbance observer that receives as input a force reference value of the actuator and a position response value of an output shaft of the actuator, and outputs an estimated disturbance proportional to a load of the actuator;
a load estimator that executes a pre-process for setting in advance a relational expression between a sum of outputs of a disturbance observer for an output shaft of each actuator when a controlled object is controlled to ascend at a constant speed by the first to fourth actuators and a sum of loads acting on the output shaft of each actuator, and then estimates a load by substituting the sum of outputs of the disturbance observer for an output shaft of each actuator when a controlled object, the load of which is unknown, is controlled to ascend at a constant speed by the first to fourth actuators into the relational expression;
The thrust command is determined based on the load estimated by the load estimation device.

（ただし、Ｍはアクチュエータ慣性、Ｇｒはアクチュエータの力参照値を変数変換したアクチュエータ駆動用のモータのトルク参照値から推力の換算ゲイン、Ｇｒｎはトルク参照値から推力の換算ゲインＧｒのノミナル値、Ｆ^d _ricは動摩擦、ｇは重力加速度、Ｗは荷重）
前記事前プロセスの実行により設定された関係式で表現される近似直線の傾きａを用いて、前記（５）式のトルク参照値から推力の換算ゲインＧｒをＧｒ＝Ｇｒｎ／ａに同定することを特徴としている。 The control device for an automated guided vehicle according to claim 5 is the control device for an automated guided vehicle according to claim 4,
The output of the disturbance observer is defined by equation (5):

(where M is the inertia of the actuator, Gr is the conversion gain of thrust from the torque reference value of the motor for driving the actuator obtained by variable conversion of the force reference value of the actuator, Grn is the nominal value of the conversion gain Gr from the torque reference value to thrust, ^Fdric _is the kinetic friction, g is the gravitational acceleration, and W is the load.)
The method is characterized in that the thrust conversion gain Gr is identified as Gr = Grn/a from the torque reference value in the equation (5) by using the gradient a of the approximation line expressed by the relational expression set by executing the pre-process.

(1) According to the inventions described in claims 1 to 5, force sensorless thrust control, lifter ascent and contact with the controlled object, and lifter descent can each be achieved simply by changing the thrust command value. If the controlled object is a wheeled cart, thrust control enables transportation with the wheels of the cart on the ground. Because thrust control is performed on each axis of the first to fourth actuators, travel according to the road surface is possible even if there is a slope or undulation.
(2) According to the invention described in claim 3, the relationship between the thrust force at a standstill and the motor torque reference value can be identified. By using this, force sensorless thrust force control can be realized with high accuracy. Since it becomes possible to identify the actual efficiency and the thrust force at the maximum static friction torque, more accurate force control becomes possible.
(3) According to the invention as recited in claim 4, it is possible to estimate the load without a force sensor, which can be applied to the generation of a thrust command in thrust control.
(4) According to the invention described in claim 5, the conversion gain of thrust can be identified from the torque reference value of the motor without using a force sensor. Since it is possible to identify the conversion gain (= actual efficiency), more accurate force control can be achieved.

1A and 1B show an outline of the device according to the present embodiment, where FIG. 1A is an external view and FIG. 1B is an explanatory diagram of the numbers of actuators as viewed from above the device. FIG. 2 is an explanatory diagram showing an example of connection between an actuator output shaft and a controlled object in the device according to the present embodiment. FIG. 1 is an overall configuration diagram of a first embodiment of the present invention. FIG. 2 is a configuration diagram of a controller according to the first embodiment of the present invention. FIG. 4 is an explanatory diagram of a conversion function of a thrust-additional torque conversion unit in the first embodiment of the present invention. FIG. 11 is an overall configuration diagram of a second embodiment of the present invention. FIG. 11 is an explanatory diagram of conversion variables of a thrust-additional torque conversion unit in the second embodiment of the present invention. FIG. 11 is an explanatory diagram showing the relationship between the reaction force measurement value of the actuator and the additional torque, which is an experimental result in the second embodiment of the present invention. 13 is an explanatory diagram showing the lifter average position and the sum of the force sensor output in the experimental results of thrust control in Example 2 of the present invention. FIG. FIG. 11 is an overall configuration diagram of a third embodiment of the present invention. FIG. 11 is a configuration diagram of a controller according to a third embodiment of the present invention. FIG. 11 is an explanatory diagram showing a relationship between the sum of disturbance observer outputs and the sum of loads in the third embodiment of the present invention. 5A and 5B show the response when entering a slope, where FIG. 5A is an explanatory diagram of ground contact transport based on height control, and FIG. 5B is an explanatory diagram of ground contact transport based on force control of the present invention.

The following describes embodiments of the present invention with reference to the drawings, but the present invention is not limited to the following embodiments. Example 1 of the present invention describes an example of feedforward force control that is force sensorless and takes into account only static friction. Example 2 performs pre-identification using a force sensor to achieve more accurate sensorless force control. Example 3 describes a load estimation method that uses a disturbance observer in relation to generating command values for force control.

Figure 1 shows an overview of a lifter device (hereinafter sometimes referred to as a four-axis lifter) according to an embodiment of the present invention. In Figure 1(a), a base 20 is mounted and fixed to an automated guided vehicle (AGV: mobile vehicle) 10, and the stators of first to fourth actuators 31 to 34 equipped with servo motors are fixed to the four corners of the base 20.

Each output shaft of the first to fourth actuators 31 to 34 has a degree of freedom of movement in the vertical direction (as shown in FIG. 1, in the direction perpendicular to the surface on which the first to fourth actuators 31 to 34 are installed on the base 20) , and this vertical movement is performed by the action of a servo motor that moves the output shaft of the actuator (shown as a) in the vertical direction.

A top plate 40 constituting a controlled object is disposed at the tip of each output shaft of the first to fourth actuators 31 to 34 so as to be supported by the tip. The controlled object of the present invention is not limited to the top plate 40, but also includes luggage, carts, wheeled carts, etc. that are placed on the top plate 40.

The positional relationship of each actuator on the top plate 40 is shown in Figure 1 (b). In the figure, Actuators 1 to 4 indicate the positions of the first to fourth actuators 31 to 34, and the x-axis is the travel direction of the AGV.

In this embodiment, the actuators 31 to 34 use ball screw mechanisms as an example.

Figure 2 shows an example of the connection between the output shaft 31a of the first actuator 31 and the top plate 40 among the actuators 31 to 34. In Figure 2, a spherical seat nut 51-1 is attached to the tip of the output shaft 31a.

A spherical seat support 52-1 is disposed between the spherical seat nut 51-1 and the top plate 40.

One end of a full thread (screw) 53-1 is inserted into the portion of the top plate 40 opposite the arrangement positions of the spherical seat nut 51-1, the spherical seat support 52-1, and the spherical seat support 52-1, connecting the spherical seat nut 51-1, the spherical seat support 52-1, and the top plate 40.

A spherical joint may be used instead of the spherical seat nut 51-1 and spherical seat support 52-1.

The connection configuration in FIG. 2 allows the actuator output shaft to come into contact with the top plate 40 even if the top plate 40 is tilted, and the tilt of the top plate can be mechanically tolerated.

The full thread 53-1 prevents the top plate 40 from collapsing, and the other end of the full thread 53-1 is fitted with a cap nut 54-1 (a top plate removal prevention device), which prevents the top plate 40 from coming off the full thread 53-1.

The second to fourth actuators 32 to 34 have the same connection configuration as shown in FIG. 2.

The overall system configuration of the first embodiment is shown in Figure 3. In Figure 3, the same parts as in Figure 1 are denoted by the same reference numerals. The thick lines in Figure 3 indicate mechanical connections. From the bottom, the AGV 10, the base 20, the motor 100 for driving the actuators, the actuators 31 to 34, the top plate 40, and the luggage 50 are connected.

The luggage 50 is placed on the top plate 40, and in this embodiment, it is assumed to be a wheeled cart. The motor 100 is a combination of the motor driver and the motor itself, and the actuators (31 to 34) are transmission mechanisms such as ball screws and gears.

The controller 60 provides a torque reference to the motor 100. In Example 1, the controller 60 does not require a feedback signal.

The configuration of a controller 60 (embodiment 1) is shown in Fig. 4. In the controller 60 of embodiment 1, a thrust command f ^cmd is input, a thrust-additional torque conversion unit 61 calculates an additional torque τ ^ad using a conversion function G of thrust command versus additional torque, and an adder 62 adds this to the maximum static friction torque τ ^sf to calculate a motor torque reference value τ ^ref . This is implemented for each actuator.

The thrust command f ^cmd is a constant value calculated in advance (for example, a constant multiple of the load equal to or less than 1). The maximum static friction torque τ ^sf is also a constant value. The value of the maximum static friction torque τ ^sf is the torque reference value obtained by gradually increasing the torque reference value when the actuators (31 to 34) start to move.

The contents of the thrust-additional torque conversion unit 61 are assumed to be expressed as a function of the added torque relative to the thrust command, as shown in Figure 5. This time, it is assumed to be linear. In the case of Example 1, where no prior identification is performed, the slope of the straight line in Figure 5 is calculated from the catalog value of the ball screw lead or gear ratio. For example, in the case of a ball screw, the slope is found as "slope = lead / (2π x efficiency)".

In Example 1, the thrust force at maximum static friction torque is assumed to be zero, but in reality it is not zero. A more accurate method of identifying the function that takes this into account will be described in Example 2.

The motor torque reference value τ ^ref in Fig. 4 is input to the motor driver in the block of the motor 100 in Fig. 3. The motor driver controls the current based on the motor torque reference value τ ^ref .

When the cargo 50 is a wheeled cart, the flow of ground transportation is "lifter up → constant thrust is applied to transport the cargo to the ground → lifter down," and the details of each step are assumed to be as follows.

(1) Lifting: In lift-up operations that do not require a pressing force, the maximum static friction torque (plus a small torque) is used as the torque reference value, and the lifter is slowly raised to make contact with the cart.

(2) Thrust control: After a certain time has elapsed after contact, the output of the adder 62, which is "additional torque τ ^ad corresponding to the thrust command + maximum static friction torque τ ^sf " as shown in Fig. 4, is given as a torque reference value τ ^ref to perform thrust control. After the torque reference value is changed, the wheels of the AGV 10 are driven to carry out transportation. In this way, ground transportation is performed.

(3) Lifter descent: After the ground transfer is completed and the AGV10 has stopped, the lifter begins to lower. The maximum static friction torque in the negative direction (plus a small torque) is input as the torque reference value, and the lifter is slowly lowered. In this mechanical system, the end of the range of motion is the origin, so a return to the origin is achieved.

By configuring the maximum static friction torque τ ^sf and the additional torque τ ^ad separately as shown in Figure 4, it is possible to realize thrust control that corresponds to the individual differences in friction of each actuator (31 to 34). The control system used here is a feedforward type thrust control system.

Figure 13 shows the response when entering a slope (undulation) during ground transportation using the thrust control described above in (2). Figure 13 shows the ground transportation of a cart 55 with wheels 55a, 55b mounted on the top plate 40 of the AGV 10 in Figure 1(a), where (a) shows ground transportation based on height (position) control. When one wheel 55a enters the inclined surface 91, the other wheel 55b lifts off the horizontal road surface 92, preventing normal ground transportation.

In contrast, FIG. 13(b) shows ground-contact transport using force control in Example 1. When the cart 55 tilts upon entering the inclined surface 91, the lifter naturally tilts accordingly, so that the wheels 55b do not lift up and normal ground-contact transport can be performed.

As such, this embodiment makes it possible to travel according to the road surface, even when there are slopes or undulations.

Another advantage of pre-identifying the maximum static friction torque is that the lifter can be raised and lowered by using the maximum static friction torque (plus a small torque) as the torque reference value.

Since the thrust and lifter ascent and descent are all controlled by torque control, they can be achieved simply by changing the thrust command value, making implementation easy.

In the first embodiment, the "thrust-torque conversion" was given without a force sensor using catalog values, etc., but in the second embodiment, we will describe a method of identifying the "thrust-torque conversion" using a force sensor.

The overall system configuration of the second embodiment is shown in Figure 6. In Figure 6, the same parts as in Figures 1 and 3 are denoted by the same reference numerals. In the second embodiment, the controller 70 provides the motor 100 with a torque reference value, and feeds back the motor rotation angle from the motor 100 and the reaction force from the force sensor 110. The force sensor 110 that measures the reaction force is installed at the tip of each of the actuators 31 to 34. When lifting a load 50 (e.g., a dolly 55), the entire weight is supported by the force sensor 110.

The identification process is carried out in the following order: "Identification of thrust at maximum static friction torque → Identification of the relationship between reaction force and added torque."

(1) Identification of maximum static friction torque and thrust: The torque reference value is gradually increased from zero, and the torque reference value at which the actuators (31 to 34) start to move is recorded as the maximum static friction torque, and the reaction force measurement value when this torque is pressed against the wall surface is also recorded (first identification process). The identification of the maximum static friction torque itself is the same as in Example 1. The motor rotation angle is used to detect the timing at which the motor 100 starts to move.

(2) Identification of the relationship between reaction force and additional torque: The maximum static friction torque plus the additional torque is used as a torque reference value, and the wall surface (or a sufficiently heavy car bogie) is pushed by the actuators (31 to 34). The reaction force measurement value at this time is recorded, and the relationship between the reaction force measurement value f ^res and the additional torque τ ^ad is plotted as shown in Figure 7(a) (second identification process).

The relationship in Fig. 7(a) can be converted into the relationship between the thrust command f ^cmd and the additional torque τ ^ad shown in Fig. 7(b). The relationship obtained in Fig. 7(b) is used as a conversion function of the thrust command versus the additional torque, and is substituted for Fig. 5 in the first embodiment.

That is, the controller 70 in FIG. 6 comprises a thrust-to-additional torque conversion section 61 and an adder section 62 similar to those in FIG. 4, and a conversion function of the thrust command to added torque in the thrust-to-additional torque conversion section 61 is set to a conversion function of the relationship between the thrust command f ^cmd and the added torque τ ^ad in FIG. 7(b) obtained in the first and second identification processes.

In Example 2, compared to Example 1, the thrust value at the maximum static friction torque and the actual efficiency can be identified, enabling more precise thrust control. Examples of experimental results of the identification of (1) and (2) above are shown below.

First, as a result of identifying the maximum static friction torque in (1) above, the torques of the four actuators (31 to 34) were determined to be (0.013, 0.011, 0.012, 0.018) [Nm]. Next, by adding additional torque in the identification in (2) above, the relationship between the actuator reaction force measurement value (vertical axis) and the additional torque (horizontal axis) was obtained, as shown in Figures 8 (a) to (d). Figure 8 shows individual differences between the actuators.

Based on the correspondence between the reaction force measurement value and the added torque in Figure 8, a conversion function between the thrust command and the added torque was set in the thrust-added torque conversion unit 61 in the controller 70 in Figure 6, and a thrust control experiment was performed.

In this experiment, the above steps were taken in the order of (1) lifter lift (up to 6 seconds after start), and (2) thrust control (after 6 seconds). First, a torque reference value (plus a small torque) corresponding to maximum static friction was applied, and the lift-up operation and contact with the cart were performed. The lifter was raised slowly, which helps to reduce the impact. Contact occurred about 3 seconds after the operation, and after a certain period of time had passed, after 6 seconds, the torque reference value was increased to achieve the desired thrust, as in (2).

The experimental results are shown in Figure 9. The top row of Figure 9 shows the average height of each axis of the lifter (COG height), and the bottom row shows the sum of the force sensor (110) outputs for each axis (fres All). A 5 rad/s low-pass filter is inserted in the force sensor output. It can be seen that in the period from 0 to 6 seconds in Figure 9, lifter lift is achieved from the position response, and in the period from 6 to 10 seconds in Figure 9, thrust control is achieved from the force response. The thrust command is 36 kgf, and the force sensor response also indicates approximately 36 kgf.

According to the second embodiment, the relationship between thrust and torque reference value at rest can be identified. By using this, force sensorless thrust control can be realized with high accuracy. Compared to the first embodiment, it is possible to identify the actual efficiency and the thrust at maximum static friction torque, which enables more accurate force control.

The thrust command (thrust command f ^cmd input to the thrust-additional torque converter 61 in FIG. 4) is a constant value, but is often determined by taking into consideration the load, such as a constant multiple of the load. In the third embodiment, in relation to the selection of the thrust command, estimation of the load without a force sensor is dealt with.

The overall system configuration of the third embodiment is shown in FIG. 10. In FIG. 10, the same parts as in FIG. 1 and FIG. 6 are denoted by the same reference numerals. The controller 80 in FIG. 10 provides a torque reference value to the motor 100 and feeds back the motor rotation angle from the motor 100, but does not require a force sensor.

The controller 80 includes a thrust-to-additional torque conversion unit 61 and an adder 62 shown in FIG. 4, and a load estimation device shown in FIG. 11 which will be described later, and determines a thrust command f ^cmd which is an input to the thrust-to-additional torque conversion unit 61, based on the load estimated by the load estimation device.

Load estimation consists of the following flow: (1) to (3) are the pre-processing steps, and (4) and (5) are the load estimation steps.

<Preliminary process>
(1) Implement constant speed position control on each actuator axis and lift the load. Record the output of the disturbance observer for each axis at constant speed. The disturbance observer here is assumed to be a standard disturbance observer that takes position and torque reference values as inputs and has a single inertia system as the reference model. However, the reference model may be changed to suit the plant, for example when the plant is not a single inertia system. A block diagram of the control of the load estimation device is shown in Figure 11.

In FIG. 11, reference numeral 81 denotes a variable conversion section which converts the motor angle θ ^res fed back from the motor 100 in FIG. 10 into a variable and outputs the actuator length X ^res (position response value of the actuator output shaft).

Reference numeral 82 denotes a position controller which controls the position response value ^Xres of each actuator output shaft when the first to fourth actuators are each controlled to rise at a constant speed so that it becomes equal to the actuator length command value ^Xcmd , and outputs the actuator force reference value ^Fref .

Reference numeral 83 denotes a variable conversion unit that converts the force reference value F ^ref of the actuator into a variable and outputs a torque reference value τ ^ref of the motor.

A disturbance observer 84 receives as input the actuator force reference value F ^ref and the position response value X ^res of each actuator output shaft, and outputs an estimated disturbance (disturbance estimated value) F^ ^dis proportional to the actuator load.

Reference numeral 85 denotes an adder that adds the disturbance estimate F^ ^dis of the disturbance observer 84 to the output of the position controller 82, but this adder 85 may be omitted.

86 is a load estimator that calculates a load estimated value W^ using the relationship between the sum of the output (F^ ^dis ) of the disturbance observer 84 for the output shaft of each actuator when the luggage 50 (controlled object such as a cart) is controlled to rise at a constant speed by the first to fourth actuators, and the sum of the loads acting on the output shaft of each actuator.

This block in Figure 11 estimates the load W applied to the output shaft of each actuator.

(2) The relationship between the sum of the observer outputs and the sum of the loads is plotted through experiments with multiple load patterns. Here, the sum of the observer outputs is the sum of the outputs of the disturbance observer 84 for the output shafts of each actuator, and the sum of the loads is the sum of the loads applied to the output shafts of each actuator.

(3) Derive the relationship between the sum of observer outputs and the sum of weights (e.g., least squares method).

<Load Estimation Process>
(4) The target unknown load (a controlled body with an unknown load) is lifted by constant speed position control in the same manner as in (1) above, and the sum of the outputs of the disturbance observer 84 at the constant speed is recorded.

(5) Substitute the output obtained in (4) into the relational equation between the sum of the observer outputs and the sum of the loads to derive the estimated load.

ここでＭはアクチュエータ慣性、Ｇｒはトルク参照値τ^refから推力の換算ゲイン、Ｆ^d _ricは動摩擦、ｇは重力加速度、Ｗは荷重である。このとき外乱オブザーバ８４の出力Ｆ＾^disは（２）式で表される。

ΔＧｒ＝Ｇｒ－Ｇｒｎは換算ゲインのノミナル誤差である。トルク参照値τ^refは（１）式における定速でのつり合いの式から（３）式として導出され、（３）式を（２）式に代入して整理すると外乱オブザーバ８４の出力は（４）式で表すことができる。

Ｇｒｎはトルク参照値から推力の換算ゲインＧｒのノミナル値である。 Although it is difficult to identify the friction of an actuator when it is stationary, it is relatively easy to identify dynamic friction, especially when it moves at a constant speed, because the Coulomb friction, which is speed-dependent, is constant. The load estimation in the third embodiment utilizes this property. The equation of motion of the actuator is expressed by equation (1).

Here, M is the actuator inertia, Gr is the conversion gain of thrust from the torque reference value τ ^ref , F ^d _ric is the dynamic friction, g is the gravitational acceleration, and W is the load. At this time, the output F^ ^dis of the disturbance observer 84 is expressed by equation (2).

ΔGr=Gr-Grn is the nominal error of the conversion gain. The torque reference value ^τref is derived as equation (3) from the equation for balance at a constant speed in equation (1). By substituting equation (3) into equation (2) and rearranging, the output of the disturbance observer 84 can be expressed as equation (4).

Grn is the nominal value of the conversion gain Gr from the torque reference value to thrust.

From equation (4), it can be seen that the output F^ ^dis of the disturbance observer 84 and the torque reference value τ ^ref are proportional to the load.

An example of the experimental results of the load estimation flow is shown below. Regarding the preliminary process (1) and (2), Figure 12(a) shows the time series response of the sum of the disturbance observer outputs when the load (cart) on the top plate 40 is changed from 0 to 40 kg. In Figure 12(a), it can be seen that when lifting at a constant speed (period of 6 to 10 s), the sum of the observer outputs is constant and changes according to the load on the top plate 40.

As part of the preliminary process (3), the average of the period of 8 to 10 seconds during which the observer output sum is constant and stable is taken, and the relationship with the load on the tabletop is plotted as shown in Figure 12(b). The approximate straight line obtained here becomes the content of the load estimator 86 in Figure 11.

In step (4) of the load estimation process, the unknown load is lifted using constant speed position control, and the sum of the outputs of the disturbance observer 84 is recorded.

In step (5) of the load estimation process, the recorded output sum is compared with Figure 12(b) to derive the corresponding load.

In the approximate straight line (y = ax + b) in Figure 12 (b), y is the sum of the disturbance observer's outputs, a is the slope, and x is the load. For example, if the sum of the disturbance observer's outputs is y (N), the sum of the loads can be estimated to be (y - b)/a (kg).

In this manner, the thrust command f ^cmd in FIG. 4 is determined taking into consideration the load estimated value (W^) estimated by the load estimator 86 in FIG.

As described above, according to the third embodiment, the load can be estimated without a force sensor, and this can be applied to generating thrust commands in thrust control.

を活用して、次のようにトルク参照値から推力の換算ゲインを同定する。 The system configuration and the controller configuration of the fourth embodiment are the same as those of the third embodiment, as shown in FIG. 10 and FIG. 11. In the fourth embodiment, the following equation (4) expressing the sum of the outputs of the disturbance observer is used.

Using this, the thrust conversion gain is identified from the torque reference value as follows:

(1) Perform the preliminary process (1) to (3) of Example 3 to find the slope a of the approximation line in Figure 12(b).

(2) The thrust conversion gain is changed from the torque reference value τ ^ref as Gr = Grn/a.

これは、図１２（ｂ）の近似直線ｙ＝ａｘ＋ｂに対応する（ｘは荷重、ｙは外乱オブザーバ出力）。一次の係数比較より（６）式が成り立つ。

この関係を用いて、トルク参照値から推力の換算ゲインを求めることができる。 By transforming equation (4) into equation (5), the output F^ ^dis of the disturbance observer can be expressed as a linear function of the load W.

This corresponds to the approximate straight line y=ax+b in Fig. 12(b) (where x is the load and y is the disturbance observer output). From a comparison of the first-order coefficients, formula (6) holds true.

Using this relationship, a thrust conversion gain can be determined from the torque reference value.

According to the fourth embodiment, it is possible to identify the thrust conversion gain (= actual efficiency) from the torque reference value, thereby realizing more accurate force control.

10...AGV
20: Base 31 to 34: Actuator 40: Top plate 50: Luggage 55: Cart 55a, 55b: Wheels 60, 70, 80: Controller 61: Thrust-additional torque conversion section 62: Adder 81, 83: Variable conversion section 82: Position controller 84: Disturbance observer 86: Load estimator 100: Motor 110: Force sensor

Claims (5)

A base mounted on a movable traveling body;
First to fourth actuators are respectively installed at the four corners of the base, and the output shafts of the actuators are configured to be movable in a direction perpendicular to the surface on the base where the actuators are installed ;
a controlled body disposed opposite to a tip end of each of the output shafts of the first to fourth actuators;
a controller that controls motors for driving the first to fourth actuators;
The controller has a thrust-to-additional torque conversion unit that converts a thrust command into an additional torque using a conversion function of thrust command to additional torque, and an adder unit that adds a maximum static friction torque to the converted additional torque to output a motor torque reference value, and the output of the adder unit is used as a motor torque reference value for a motor for driving the actuator. The control device for an automated guided vehicle according to claim 1, characterized in that the controller performs lifter ascent control based on a motor torque reference value using a maximum static friction torque in the positive direction, thrust control based on a motor torque reference value obtained by adding the additional torque and the maximum static friction torque, and lifter descent control based on a motor torque reference value using a maximum static friction torque in the negative direction. The conversion function of the thrust command to the additional torque in the thrust-additional torque conversion unit is expressed as follows:
a first identification process in which the motor torque reference value is gradually increased from zero, the motor torque reference value when the actuator starts to move is recorded as a maximum static friction torque, and a value measured by a force sensor when the actuator is pressed against a wall surface with this torque is also recorded;
a second identification process is executed to determine a relationship between the maximum static friction torque recorded in the first identification process plus an additional torque as a motor torque reference value, the reaction force measured by a force sensor when the actuator is pressed against a wall surface, and the additional torque;
3. The control device for an automated guided vehicle according to claim 1, wherein the relationship between the reaction force measurement value and the additional torque obtained in the second identification process is converted into a conversion function of the thrust command versus additional torque and set. The controller:
a position controller that controls the position response value of each actuator output shaft, obtained by variable-transforming the angle of the actuator drive motor when the first to fourth actuators are each raised at a constant speed, so that the position response value becomes a command value for the actuator length, and outputs a force reference value for the actuator;
a disturbance observer that receives as input a force reference value of the actuator and a position response value of an output shaft of the actuator, and outputs an estimated disturbance proportional to a load of the actuator;
a load estimator that executes a pre-process for setting in advance a relational expression between a sum of outputs of a disturbance observer for an output shaft of each actuator when a controlled object is controlled to ascend at a constant speed by the first to fourth actuators and a sum of loads acting on the output shaft of each actuator, and then estimates a load by substituting the sum of outputs of the disturbance observer for an output shaft of each actuator when a controlled object, the load of which is unknown, is controlled to ascend at a constant speed by the first to fourth actuators into the relational expression;
4. The control device for an automatic guided vehicle according to claim 1, wherein the thrust command is determined based on a load estimated by the load estimation device. The output of the disturbance observer is defined by equation (5):

(where M is the inertia of the actuator, Gr is the conversion gain of thrust from the torque reference value of the motor for driving the actuator obtained by variable conversion of the force reference value of the actuator, Grn is the nominal value of the conversion gain Gr from the torque reference value to thrust, ^Fdric _is the kinetic friction, g is the gravitational acceleration, and W is the load.)
5. The control device for an automated guided vehicle according to claim 4, wherein a gradient a of an approximation straight line expressed by the relational expression set by execution of the pre-process is used to identify a thrust conversion gain Gr from the torque reference value in the equation (5) as Gr = Grn/a.

Images