JPH07144881A - Method and device for preventing swing of load of crane - Google Patents

Abstract

PURPOSE:To converge the residual swing due to an error of a model to the minimum by increasing and decreasing the acceleration/deceleration control time, which is obtained by fuzzy inference, so that the swinging angle of load is reduced at the next automatic operation time on the basis of the swinging angle of load at the time when the two-stage acceleration/deceleration is fished. CONSTITUTION:In a fuzzy control unit 4, an operation speed pattern, which includes the two-stage acceleration/deceleration for preventing swinging of load, is inferred by a first fuzzy inference block 1, and the acceleration/ deceleration control time for deciding the acceleration/deceleration timing of the later stage of this pattern is inferred by a second fuzzy inference block 3. A swinging angle measuring device 20 measures the swinging angle of a hoisted load. A learning control unit 9 sets the increase and decrease of the acceleration/deceleration control time to be output from the fuzzy control unit 4 so that the swinging angle of a hoisted load is reduced at the next automatic operation time on the basis of the swinging angle of load computed at the time when the two-stage acceleration/deceleration is finished. An output control unit 8 controls a crane 6 through a driving unit 6 on the basis of the speed pattern, which is corrected through learning.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a crane steady-state driving method and apparatus capable of reducing the swing angle of a suspended load during automatic crane operation by using a learning function.

[0002]

2. Description of the Related Art Generally, when a jib crane is operated, a swing of the jib causes a shake of the cargo due to acceleration and deceleration, which impairs the safety of work and stops the shake. There is a problem that it takes extra time.

[0003] Conventionally, in the case of manual driving, the driver has empirically learned the acceleration and deceleration driving techniques to dampen the vibration generated by the driving technique. That is, a two-stage acceleration / deceleration operation method is performed in which the shake generated in the former stage is canceled in the latter stage by appropriately interrupting the acceleration / deceleration. However, in automatic driving, there is no such thing that appropriately stops the shake.

Therefore, the applicant of the present invention creates a turning operation speed pattern and an up-and-down operation speed pattern including two-step acceleration / deceleration for steadying the load by the first fuzzy inference based on the cargo handling data and the rule map. , The crane posture and suspended load height at the time of two-step acceleration / deceleration in the speed pattern are calculated, the optimum command speed pattern is determined by the second fuzzy inference, and the jib crane is automatically controlled and operated according to the command speed pattern. (Japanese Patent Application No. 3-277989).

[0005]

However, in the case of the conventional steady rest operation method, the steady rest of the load is carried out by the two-stage acceleration method, that is, the two-stage acceleration / deceleration. Is only determined and output by fuzzy reasoning. That is, since the open-loop control is used, in the conventional automatic crane operation device, once the steadying operation of the suspended load is determined, the steadying performance further cannot be obtained. In addition, even if preset parameters for steady rest are set by simulation,
It is customary for residual shake to remain due to model errors. Therefore, in reality, on-site adjustment work for the two-step acceleration / deceleration timing is required, which is actually a rather complicated work.

Therefore, an object of the present invention is to solve the above-mentioned problems and eliminate the on-site adjustment work of steady rest control parameters.
An object of the present invention is to provide a more accurate crane steadying operation method and device.

[0007]

In order to achieve the above object, the load steadying operation method for a crane according to the present invention creates an operating speed pattern including two-step acceleration / deceleration for stopping load steady by a first fuzzy inference. The acceleration / deceleration timing of the subsequent stage in the speed pattern
The deceleration control time is determined by the second fuzzy inference,
While automatically controlling the jib crane according to this command speed pattern, the hanging load shake sensor provided on the crane measures the load shake angle at the time when the two-stage acceleration / deceleration is completed, and based on the load shake angle, the next time The learning function is used to adjust the acceleration / deceleration control time so that the swing angle of the suspended load becomes small during the automatic operation of (1).

Further, the load steadying operation device for a crane according to the present invention creates an operating speed pattern including a two-step acceleration / deceleration for stopping the load by the first fuzzy inference, and the subsequent acceleration / deceleration timing in the speed pattern. The acceleration / deceleration control time for determining the acceleration / deceleration control time by the second fuzzy inference, the deflection angle measuring device mounted on the crane for measuring the deflection angle of the suspended load, and the deflection angle signal from the deflection angle measuring device. Based on the deflection angle processing device that calculates the deflection angle of the load at the time when the two-step acceleration / deceleration is completed, and the deflection angle obtained from the deflection angle processing device, the load deflection angle of the suspended load at the next automatic operation is The learning control unit that adjusts the acceleration / deceleration control time of the output of the fuzzy control unit so as to reduce the size, and the clay pattern according to the speed pattern corrected by the learning control unit. Is of the configuration in which an output control unit for automatic operation (the second aspect).

[0009]

In the load steadying operation method of the first aspect, the learning function of changing the crane operation timing in the next automatic operation works based on the deflection angle data obtained as a result of the automatic operation. Therefore, even if the residual shake remains due to the error of the simulation model, the residual shake automatically converges to the minimum value.

In the case of the load steadying operation device according to claim 2, the fuzzy control unit creates an operating speed pattern including two-stage acceleration / deceleration of the load steadying by the first fuzzy inference,
The second fuzzy inference determines the acceleration / deceleration control time that determines the subsequent acceleration / deceleration timing in the speed pattern.

On the other hand, the deflection angle measuring device measures the deflection angle during the operation of the crane, and the deflection angle processing device determines the deflection angle of the load when the two-stage acceleration / deceleration is completed, that is, the deflection when the crane is stopped or the crane speed changes. Calculate the angle. The learning control unit determines the acceleration / deceleration control time of the output of the fuzzy control unit based on the load deflection angle obtained from this deflection angle processing device so that the load deflection angle of the suspended load becomes smaller at the next automatic operation. Adjust the setting.

In short, the steady rest operation control device learns the acceleration / deceleration timing of the crane such that the deflection angle when the crane is stopped becomes "0", and from the next operation of the crane, the learned acceleration / deceleration timing is used. Act to control.

In the load steadying operation control device of the open control system having only the fuzzy control unit, once the steadying motion operation of the suspended load is determined, no further steady rest performance can be obtained. Since the feedback control system from the shake angle measuring device is provided, the steady rest accuracy can be improved until the measurement accuracy of the shake angle measuring device is limited.

[0014]

An embodiment of the present invention will be described in detail below with reference to the accompanying drawings. The load steadying operation control device for a jib crane shown in FIG. 3 includes a first fuzzy inference block 1 for first creating a turning operation speed pattern and an undulating operation speed pattern, and one of those speed patterns that has a longer operation time first. The fuzzy control section 4 is composed of a priority determination block 2 for starting the operation and a second fuzzy inference block 3 for determining an optimum command speed pattern based on the speed pattern. In addition, the operation control device includes an input unit (setting device) 5 for appropriately inputting cargo handling data to a fuzzy control unit 4 in which the blocks 1, 2, and 3 are incorporated, and a determined command speed pattern. The output control unit 8 outputs an operation signal to the drive unit 7 of the jib crane 6.

Furthermore, when the load shake sensor 19 (described later) actually measures the load shake angle θ ° (see FIG. 7) as the residual shake, this operation control device further determines the load shake angle θ ° during the next automatic operation. The learning control unit 9 that optimally adjusts the operation timing of the crane such as turning, undulating, traveling, that is, the two-stage acceleration / deceleration timing by using the learning function based on the deflection angle so that There is. The adjustment of the two-stage acceleration / deceleration timing of this crane is performed by setting the time length of the constant speed operation section (D1, D2 in FIG. 2) in the command speed pattern determined by the second fuzzy reasoning block 3 to the load deflection angle. It is performed by adjusting the adjustment time Δt (s) according to θ °. However, in the present embodiment, by controlling the fuzzy output t (s) from the fuzzy control unit 4, that is, the acceleration / deceleration control times T1 and T2 up to the acceleration / deceleration start times t2 and t4 in the latter stage of FIG. The constant speed operation section lengths D1 and D2 are indirectly adjusted.

The jib crane 6 is constructed by equipping a tower 10 with a swivel type hoisting jib 11, and a swivel motor 12 for swiveling drive and a hoisting motor for hoisting motion are provided at the base end of the jib 11. 13 and a hoisting machine 15 for hoisting or lowering (unwinding) a rope (wire) 14 for hanging the luggage A. Then, when a predetermined operation signal is input to the drive unit 7 of the motors 12 and 13, the jib 11 swings and moves by a predetermined angle α and moves up and down by a predetermined angle β to move the suspension rope 14 up and down. Together, the luggage A is transported from a predetermined carry-out point B to a carry-in point C.

The input unit 5 is constructed so that the ground heights of the carry-out point and the carry-in point C, planar positions, and the hanging load amount can be set and input by operating buttons or the like. . Then, by converting these data, the initial hoisting angle β0 and the initial turning angle α0 of the jib 11, and the initial suspended load ground height H at which the jib 11 should start the turning and hoisting motions.
0, the final hoisting angle βE and the final turning angle αE, which are the target values, and the final suspended ground height HE at which the jib 11 should end the turning and hoisting motions are input to the first fuzzy inference block 1.

The first fuzzy inference block 1 calculates the initial and final suspension rope lengths L0 and LE from the input data, and writes them on the rule map 16 in consideration of the suspended load cycle at these rope lengths. From the acceleration / deceleration pattern (timing), the one corresponding to the suspension rope length close to the calculated values L0 and LE is extracted, and the fitness is calculated by the membership function to obtain a definite value. Then, together with the steady speed (maximum speed) derived from other data and its operation time, a basic turning operation speed pattern and undulation operation speed pattern from start to stop are created (FIG. 8). . The rule map 16 is prepared in advance by an empirical rule and computer simulation, and in the present embodiment, the acceleration / deceleration of the front stage and the rear stage and its length are omitted as the two-stage acceleration / deceleration pattern. The lengths D1 and D2 (FIG. 2) of the constant velocity section between the front stage and the rear stage are set to be different for each suspension rope length.

The priority determination block 2 compares the total operating times of the speed patterns created by the first fuzzy inference block 1 and, as shown in FIG. 9, sets the acceleration start time T0 of the shorter speed pattern to the longer one. It is set as the acceleration end time of the speed pattern. At this time, if the total operating time length of the shorter speed pattern is sufficiently shorter than the other and the operation ends within the other operating time, the start time is set appropriately after the longer acceleration end time. It doesn't matter.

The second fuzzy inference block 3 calculates the crane attitude (α, β) and the suspended load height (H) at the two-stage acceleration / deceleration in the speed pattern set by the priority determination block 2, and this calculation The suspension rope length (L) at that time is calculated based on the value. This is because the suspension rope length L changes due to the horizontal pulling action due to the hoisting motion of the jib 11, and further changes due to the hoisting performed while the load A turns and hoists and moves, so the timing of two-stage acceleration / deceleration. It is also necessary to change. This suspension rope length L is related to the transportation route, and for example, if there is an obstacle or the like on the way, it changes independently of the loading / unloading points B and C. Therefore, the cargo handling information is input from the input unit 5 to the second fuzzy inference block. 3 is transmitted. Then, by using the calculated suspension rope length L, the acceleration / deceleration timing initially created based on the correction rule map 17 is corrected using fuzzy inference,
The optimum command speed pattern is determined.

The speed pattern, that is, the operation control data output to the crane output control unit 8 is, as shown in FIG. 2, an acceleration trigger t1 (acceleration start condition), an acceleration control time T1,
Deceleration trigger t3 (deceleration start condition), deceleration control time T2
Composed by. The acceleration control time T1 and the deceleration control time T2 are time data obtained by the load shake control calculation, and the acceleration / deceleration triggers t1 and t3 are position data obtained by setting the operation path to the target position of the crane. Is. Acceleration control time T1, deceleration control time T2
Is to determine the second (post-stage) acceleration or deceleration timing t2, t4 in the two-stage acceleration control and the two-stage deceleration control, and the hanging motion sensor 19 detects these timings based on the learning function of the learning control unit 9. By performing variable control so that the load shake angle θ is minimized, it is possible to automatically operate to the target position while suppressing the load shake.

If the suspension rope length L changes between the first acceleration and the start of the second-stage acceleration, the cycle of the load swing also changes. For this reason, in the present embodiment, as shown in FIG. 10, the acceleration timing based on the suspension rope length L0 at the start point (time 0) of the acceleration of the turning motion speed pattern and the suspension rope at the start point of acceleration in the subsequent stage. Long L
The acceleration timing based on S1 is obtained by fuzzy inference, and then the average value of these is determined as the final value TS2. Similarly, the deceleration timing TS4 and the subsequent acceleration / deceleration start times TK2 and TK4, which are the acceleration / deceleration timings of the undulation speed pattern, are also determined.

The output control unit 8 is based on the command speed pattern determined by the second fuzzy inference block 3, and more accurately, the two-stage acceleration / correction is further corrected by the learning control unit 9.
An operation signal is output to the drive unit 7 based on the deceleration timing to control the swing motor 12 and the hoisting motor 13. By this control, the jib 11 operates appropriately and cooperates with the hanging rope 14 to carry the load A from the carry-out point B to the carry-in point C.

By the way, in order to apply the above-mentioned load steadying control during the automatic operation of the jib crane, a sensor capable of detecting the load swinging angle θ in real time is required. Therefore, the crane 6 is provided with the load shake sensor 19 including the shake angle measuring device 20 and the shake angle processing device 24.

As shown in FIGS. 6 and 7, the deflection angle measuring device 20 is provided in the tracking device main body and a tracking device 22 provided at the tip of the jib of the crane 6 and comprising a light transmitting portion, a light receiving portion and an angle control portion. Lightwave range finder 21 for measuring the distance to be lifted and crane hook 1 at the tip of rope 14 of the crane
8 and comprises a light wave range finder 21 and a reflecting prism (corner cube) 23 directed toward the tracking device 22. However, the deflection angle measuring device 20 is basically mounted on a crane as long as it can measure the deflection angle of a suspended load, and the suspension rope length can be detected by other means.
The optical distance meter 21 is not always essential to the present invention.

The tracking device 22 emits a beam of light, which uses a light emitting diode as a light source, from the light transmitting unit, irradiates the beam toward the reflecting prism 23, and the reflected light is imaged on the light receiving unit. The tracking device 22 sets the X-axis according to the amount of deviation from the optical axis center.
-The pulse motor of the Y biaxial mount is driven to constantly track the reflection prism 23 attached to the crane hook 18.

The deflection angle processing device 24 is the deflection angle measuring device 2
In response to the deflection angle signal from 0, that is, the tracking device 22
By detecting the rotation angles of the X-axis and the Y-axis, the data of the load deflection angle θ at the time when the two-step acceleration / deceleration is completed is taken out. In this manner, the swing angle (hook angle of the suspended load) θ of the hook 18 that suspends the load A is measured. At the same time, the lightwave distance meter 21 coaxially arranged with the tracking device 22 more accurately measures the suspension rope length L, which is the distance between the tip of the jib 11 and the hook 18. After all, the suspended load X,
Y and Z coordinate values are calculated. Therefore, the swing angle measuring device 20 can measure the hanging load swing angle θ in all directions.

The load shake sensor 19 can be constructed by using a commercially available reflection prism tracking type optical wave distance measuring device.

As shown in FIGS. 3 and 4, the learning control section 9 receives the load deflection angle θ obtained from the deflection angle processing device 24 of the load deflection sensor 19, and the turning or undulating operation speed pattern is accelerated by two steps.・ Learning is performed by comparing the load deflection angle θ at the time points TE and TE at which the deceleration is completed with a target value “0 °” and outputting an error e °, and a neural network using the e ° as a learning signal. Two-step acceleration based on the results
It has a two-stage timing adjuster 26 that determines the adjustment time Δt (s) of the deceleration timing, and an adder 27 that adds the adjustment time Δt (s) to the fuzzy output t (s) from the fuzzy controller 4. It is configured.

The two-stage timing adjuster 26 has an input calculation circuit 28, an intermediate calculation circuit 29 and an output calculation circuit 30 as nodes of the neural network. The input calculation circuit 28 has a suspension rope length L and a fuzzy output t (s). ) Has been entered. Based on this calculation condition, the neural network determines a target value of which node in the output calculation circuit 30 should be the maximum value, that is, a target value related to the weighting parameter value of each node of the intermediate calculation circuit 29. Is calculated and set. The specific calculation content or function of each node is determined by the value of the weighting parameter stored by the node at that time. Therefore,
As a temporary form, one of the intermediate calculation circuits 29 first takes the maximum value, and the judgment as a result is obtained by the output calculation circuit 30 as the adjustment time Δt (s).

Each node of the intermediate calculation circuit 29 adjusts the internal weighting parameter in a certain direction based on the learning signal (correction command) at the two-stage acceleration / deceleration completion time points TE, TE.
As a result, the determination Δt (s) is determined again by the output calculation circuit 3
0, and the weighting parameter inside each node is further adjusted based on the correction command. By such an iterative or learning function, the same judgment result as the target value “0 °” is finally obtained in the output calculation circuit 30, and the adjustment of the weighting parameter inside each node is completed. Since the adjustment here is based on the learning function, it does not matter whether the adjustment direction starts in the correct direction or is performed by trial and error, and in any case, the correct correction is performed as a result.

Next, a specific example of the method for controlling the load steadying operation of the jib crane will be described as the operation of the above configuration.

When carrying the cargo A automatically by the jib crane 6, the cargo handling data is first inputted to the input section 5. This cargo handling data is the crane posture (α0, αE
, Β0, βE) and suspended load ground height (H0, HE) are input to the first fuzzy inference block 1. In the first fuzzy inference block 1, the suspension rope length (L0, LE) is calculated from these input data, and the two-stage acceleration / deceleration can be performed by the fuzzy inference to stabilize the load according to the suspension rope length (L0, LE). The pattern is determined, and the turning operation speed pattern and the undulating operation speed pattern shown in FIG. 8 are created.

The created turning operation pattern and undulating operation speed pattern are input to the priority order determination block 2,
The operating hours are compared. If the turning operation speed pattern is longer, as shown in FIG. 9, the start time T of the undulating operation speed pattern is set with the acceleration start time as the origin (0).
0 is delayed until time TE at which the turning speed pattern ends the two-step acceleration.

The turning operation speed pattern and the undulating operation speed pattern after the priorities are determined are input to the second fuzzy inference block 3. Then, as shown in FIG. 10, since the suspension rope length L changes with time (see the transportation route indicated by the two-dot chain line), the crane posture (α, β) and the suspended load height (in the two-stage acceleration / deceleration section) ( By calculating H), the suspension rope length (L) at that time is estimated. Then, the fuzzy inference is used to determine the acceleration start time value (TS2, TK2) and the deceleration start time value (TS4, TK4) of the subsequent stage in each speed pattern, and the optimum command speed pattern is determined.

The determined speed pattern for command is input to the output control 8, and operation signals according to this speed pattern for command are sequentially transmitted to the drive unit 7, and the jib 11 swings and undulates while carrying out steadying of the load, so that the load A is automatically transported in the shortest time.

On the other hand, it is assumed that the fuzzy output timing is t (s) at a certain rope length, and when the crane shake is stopped at this value, the residual shake is measured as the load shake angle α °. Originally, when the two-step acceleration is completed (when the steady rest acceleration is completed), TE and TE are desired to have a residual shake of 0 °, so the target value in the comparison unit 25 is 0 °. That is, this means that the two-step acceleration timing t (s) deviates from the true value.

Therefore, the two-stage timing adjuster 26
The time Δt (s) corresponding to the deviation is output.
The input of the two-stage timing adjuster 26 is the rope length L for two-stage acceleration and the fuzzy output (two-stage timing t (s)) at that time, and the output is the adjustment time Δt.
(S). However, Δt takes positive and negative values.

The learning of this neural network is t +
When the two-stage acceleration is performed in Δt (s) time, the residual shake angle α ° at the completion of acceleration and the target value of the residual shake angle 0 °
The learning of the neural network is learned using the error e ° between and as the learning signal. And by stacking the number of floors for automatic operation of cranes, this learning loop goes around,
As the learning progresses, the residual shake angle α ° gradually becomes smaller. A summary of the above learning flow is shown in FIG.

In this way, since the steady rest performance is automatically improved by the learning function, it is possible to completely eliminate the shake of goods during transportation up to the limit of the measurement accuracy of the shake angle, and the safety of automatic driving. It is possible to achieve the improvement of and the reduction of working time.

Although the jib crane has been described in this embodiment, the present invention is not limited to the jib crane but can be widely applied to various other types of cranes.

[0042]

In summary, according to the present invention, the following excellent effects can be obtained.

1) According to the method described in claim 1, the learning function can be used to optimally adjust the operation timing in the automatic operation of the crane so that the remaining deflection angle becomes zero. Therefore, even if the residual shake may remain due to a model error with the steady-state control parameter preset by simulation or the like, the residual shake is automatically shifted to the minimum value.

Further, it is possible to reduce the residual runout caused by the individual difference (machine difference) of the cranes by operating the cranes, and the complicated adjustment of the control parameters at the site becomes unnecessary.

2) According to the second aspect of the invention, the sway angle measuring device is attached to the crane, the sway angle processing device is mounted, and the crane operation timing is determined in the next automatic operation based on the sway angle data obtained therefrom. Since the learning function is provided so that the value can be changed, it is possible to optimally adjust the operation timing in the automatic operation of the crane and converge so that the remaining deflection angle becomes zero. Therefore, the residual runout caused by the individual difference (machine difference) of the crane can be made zero by simply operating the crane.

[Brief description of drawings]

FIG. 1 is a schematic block diagram showing a load steadying operation device for a crane according to the present invention.

FIG. 2 is a diagram exemplifying an operating speed pattern including two-stage acceleration / deceleration.

FIG. 3 is a block diagram showing an embodiment of a load steadying operation device for a crane according to the present invention.

FIG. 4 is a diagram showing a configuration of a learning control unit in FIG.

5 is a diagram showing a learning loop of the neural network of FIG.

FIG. 6 is a side view showing attachment of a load shake sensor to a jib.

FIG. 7 is a front view showing attachment of a load shake sensor to a jib.

FIG. 8 is a diagram showing a velocity pattern created by a first inference block.

FIG. 9 is a diagram showing a speed pattern created by a priority order determination block.

FIG. 10 is a diagram showing a velocity pattern created by a second inference block.

[Explanation of symbols]

 1 1st fuzzy inference block 2 2nd priority determination block 3 2nd fuzzy inference block 4 Fuzzy control part 5 Input part 6 Jib crane 7 Drive part 8 Output control part 9 Learning control part 10 Tower 11 Jib 12 Swing motor 13 Relief motor 14 Rope 15 Winding machine 16 Rule map 17 Correction rule map 18 Crane hook 19 Load deflection sensor 20 Deflection angle measuring device 21 Lightwave distance meter 22 Tracking device 23 Reflective prism 24 Deflection angle processing device 25 Comparing unit 26 Two-step timing adjuster 27 Addition unit 28 Input calculation circuit 29 Intermediate calculation circuit 30 Output calculation circuit A Baggage B, C Loading / unloading points D1, D2 Constant velocity section H0 Initial load ground height HE Final load ground height L0 Initial lifting rope length LE Final lifting rope length T0 Acceleration start Time T1 Acceleration control time T2 Deceleration Control time TE Two-step acceleration / deceleration end point TS2, TK2 Post-stage acceleration start time value TS4, TK4 Post-stage fast deceleration start time t1 Acceleration trigger (acceleration start condition) t2 Post-stage acceleration timing t3 Deceleration trigger (deceleration start condition) t4 Later deceleration timing t (s) Fuzzy output Δt (s) Adjustment time α Swing angle α0 Initial swivel angle αE Final swivel angle β Undulating angle β0 Initial hoisting angle βE Final hoisting angle θ Load deflection angle

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kira Kubo, 1-2-7 Moto-Akasaka, Minato-ku, Tokyo Kashima Construction Co., Ltd. (72) Inventor Tsutomu Nemoto 1-2-7, Moto-Akasaka, Minato-ku, Tokyo Kashima Construction Co., Ltd. (72) Inventor Tatsuro Sato 1-2-7 Moto-Akasaka, Minato-ku, Tokyo Kashima Construction Co., Ltd. (72) Inventor Shigeki Murayama 3-1-15-1 Toyosu, Koto-ku, Tokyo Ishikawajima Harima Heavy Industries Co., Ltd. Toji Technical Center (72) Inventor Toshiaki Saito 3-15-15 Toyosu Koto-ku, Tokyo Ishikawajima Harima Heavy Industries Co., Ltd. Toni Technical Center (72) Inventor Shigeru Okano Chiyoda, Tokyo 1-chome, Ogawamachi, Kanda-ku Ishikawajima Transporter Co., Ltd. (72) Inventor Yoshihiro Muta 1-1-chome, Kandaogawamachi, Chiyoda-ku, Tokyo Ishikawajima Within Transport Machine Co., Ltd.

Claims (2)

[Claims] 1. An operating speed pattern including two-stage acceleration / deceleration for stabilizing a load is created by a first fuzzy inference, and an acceleration / deceleration control time for determining an acceleration / deceleration timing of a subsequent stage in the speed pattern is defined by a second fuzzy reasoning. Determined by fuzzy reasoning, the jib crane is automatically controlled and operated according to this command speed pattern, while the suspended load shake sensor provided on the crane measures the load swing angle at the time when two-stage acceleration / deceleration is completed, and the load A load steadying operation method for a crane, characterized by using a learning function to adjust the acceleration / deceleration control time so that the swing angle of a suspended load becomes smaller at the next automatic operation based on the swing angle. 2. An operating speed pattern including two-step acceleration / deceleration for stabilizing the load is created by a first fuzzy inference, and an acceleration / deceleration control time for determining an acceleration / deceleration timing of a subsequent step in the speed pattern is set as a second A fuzzy control unit that is determined by fuzzy inference, a deflection angle measurement device that is mounted on a crane and that measures the deflection angle of a suspended load, and a two-stage acceleration / deceleration completion point when the deflection angle signal from the deflection angle measurement device is received. A deflection angle processing device for calculating a load deflection angle, and based on the load deflection angle obtained from the deflection angle processing device, so that the load deflection angle of the suspended load is reduced at the next automatic operation,
A crane comprising: a learning control unit that adjusts the acceleration / deceleration control time of the output of the fuzzy control unit, and an output control unit that automatically operates the crane according to the speed pattern corrected by the learning control unit. Load steady operation device.