DE3513007C2 - - Google Patents

Description

The invention relates to a method for automatic Control of a crane according to the preamble of claim 1.

Such a method is from DE-Z .: hoists and Fördermittel, 1983, H. 7, pp 206-209.

In the following, the term "crane" means a crane of the type shown in FIG. 1, namely a crane in which a trolley 11 is movable on a horizontal plane and a rope 12 hangs from the trolley 11 with the load 13 attached thereto , pulling in and out the rope.

The crane shown schematically in Fig. 1 is operated automatically so that

• (1) the trolley 11 is moved to the destination within the shortest possible period of time and
• (2) the crane is controlled so that the load 13 stops swinging at the destination.

In principle, the control for stopping the pendulum movement takes place in such a way that the load 13 is pivoted forwards with respect to the trolley 11 when this is in the vicinity of the destination, whereby the trolley 11 catches up with the load 13 , which has swung approximately to the destination and is practically at rest. The trolley is then stopped.

In conventional devices for automatically controlling the crane, a speed pattern is first established which the trolley 11 is to follow. The speed of the trolley 11 is then controlled according to the speed pattern (see Mita and Kanai: "Optimum Crane Operation Method by giving attention to a Maximum Speed of the Trolley" in an article by the Association of Measurement and Automatic Control, Volume 25, No. 6, 1979). The speed curve is determined by such a calculation that the pendulum angle of the cable system at the end of the movement is equal to zero.

In this known device for automatic operation and controlling a crane is the speed curve of the Trolley for achieving the above-mentioned control operation by a geometric method or an optimal control method determined. The servo system is designed so that the The trolley follows the given speed pattern. the The greatest difficulty with this device is that the control or regulation system for the swing angle of the rope represents an open loop. Therefore it is with the above known automatic control system does not allow the pendulum angle of the rope to correct if in the control system There are error factors, such as the initial one Pendulum motion, gusts of wind or time delays.

As mentioned above, so far, the automatic operation has been used with the crane running so that the speed of the The trolley was automatically adjusted to be accurate follows a predetermined speed pattern. The speed pattern was set or determined so that the pendulum angle of the rope at the end of the run at a constant Rope length is zero. The load is at standstill Trolley up and down.

Therefore, it is the time required for one cycle of load movement equal to the sum of the time it took to move the trolley to Raising and lowering the load is required. Will this Times even only slightly reduced, this can be the case with large loads a lot of time can be saved.

Even if the speed of the trolley is exactly calculated Adjusted wisely, the rope can often result commute unexpected influences, for example as a result of Gusts of wind, the initial pendulum motion or changes different parameters. In such cases the rest of the pendulum motion becomes no longer zero when the trolley is stopped will. As a result, the one bearing the load moves Rope continues in a pendulum motion. The load can, however not be handled until the load is sufficient Dimensions has swung out, so that a lot of time is lost. In particular the containers representing the load must be precisely positioned so that the residual pendulum motion is a serious one Problem.

From the DE-Z mentioned at the beginning: Hoists and conveyors (1983), H. 7, pp. 206 to 209 is also a method for sway angle damping and positioning of the load on trolleys known with non-rigid load suspension, in which the pendulum of the rope is suppressed by regulating the speed of the trolley will. This is done using two independent feedback circuits used, one being the momentary pendulum angle on one travel controller and the other the trolley speed fed back to a speed controller, so that the load and the trolley follow a predetermined load path pattern or a predetermined trolley speed pattern follow. In other words, the position controller sees a command signal for regulation the load path and the speed controller are different Signal for regulating the trolley speed. These both command signals are combined into one command signal, with the help of which the trolley is controlled.

From DE-OS 32 28 302 it is also known to each value of Acceleration force as a solution of differential equations to obtain. The procedure for determining the control command however, it is not of the prediction type.

From EP-89 662 A1 and US-39 21 818 it is also known the rope oscillation by controlling the trolley accordingly suppress a predetermined speed pattern, however, there is no timing of the rope length he follows.

It is the object of the invention, the method of the aforementioned Kind of further training in such a way that the rope oscillation is more optimal dampened and the trolley controlled more optimally can.

The solution to this problem results from the characteristic Features of claim 1. Advantageous embodiments of these are the subject matter of subclaims 2 to 5.

According to the invention is used to control the speed of the trolley used control command from a variety of certain, mutually different control commands selected after the results of the control due to this certain control commands were estimated, if these be fed to the trolley. Besides, the decision will be made regarding the control command carried out by both the speed or the stopping position of the trolley as well the swing angle of the rope corresponding to a predetermined one Rule, namely a conclusion based on "fuzzy Values ", is determined.

In view of the above conclusion due to "fuzzy values" (which follow in the description are referred to as "fuzzy values") z. B. refer to the following literature:

• - IEEE TRANSACTION ON SYSTEMS, MAN, AND CYBERNETICS; Vol. SMC-3, No. 1 pp 28-44, January 1987. "Outline of a New approach to the Analysis of Complex Systems and Decision Process "by LOTFI A ZADEH;
• - Automatica Vol. 13 pp 235-242 Programm Press, 1977. "The Application of Fuzzy Control Systems to Industrial Process "by P.J. KING and E.H. MAMDANI;
• - EP 92 832 A2.

According to the invention, the crane can thus be more stable and at high speed controlled automatically.

The invention is explained in more detail below with reference to the drawing. Show it:

Fig. 1 is a schematic illustration of a crane,

Fig. 2 is a block diagram of a first embodiment of a device, the process is carried out with the aid,

FIGS. 3 and 4 are diagrams for explaining the calculation of weighting or evaluation values,

Fig. 5 is a flowchart for explaining the processing,

Fig. 6 is a block diagram of a second embodiment,

Fig. 7 is a diagram for explaining the procedure for calculating the evaluation values,

Fig. 8 is a phase diagram for explaining the principle for stopping the oscillating motion,

Fig. 9 in the diagram, the principle of calculation of the residual swing movement,

Fig. 10 is a block diagram of another device for automatically operating and controlling the crane,

Fig. 11 in the diagram in the form of a subdivided function,

Fig. 12 is a flowchart for illustrating the processing flow,

Fig. 13 is a schematic model diagram of a crane to which the method is suitable,

Fig. 14, the diagram of the trolley speed,

Fig. 15 is a phase diagram of the angular velocity and the swing angle of the rope,

Fig. 16 shows the diagram of a split function, which is used to determine the evaluation values,

Fig. 17 is a block diagram of another device for automatically operating and controlling a crane,

Fig. 18 is a flowchart for explaining the procedure during winding and unwinding of the rope,

Fig. 19 is a diagram for explaining the principle of suppressing the oscillating movement of the rope and

Fig. 20 is a flow chart for explaining the control of the rope length.

Fig. 2 shows the block diagram of a first embodiment of a device for the automatic control of a crane, which is used in the method according to the invention. The device contains a device 1 for measuring the pendulum angle of the cable system, a device 2 for calculating the existing pendulum angle R (t) based on a previously measured pendulum angle and a device 3 for calculating the angular velocity (t) , which is the time differential of a pendulum angle on the basis of the previously measured pendulum angle.

A tachometer generator 4 is used to measure the speed of the trolley system. A computing device 5 is used to calculate the speed V (t) of the trolley 11 from pulses received from the tachometer generator 4 , a computing device 6 is used to calculate the acceleration α (t) of the trolley 11 in the same way as by means of the computing device 5 and a Measuring device 7 is used to detect the weight m of the load to be picked up. A micro-computer 8 is used to calculate the control command u (t) on the basis of the aforementioned data, and a control device 9 is used to control the trolley 11 .

Fig. 3 is a diagram for explaining the calculation of evaluation values related to the trolley speed.

Fig. 4 shows a diagram for explaining the calculation of the pendulum motion of the cable system related evaluation values ( "fuzzy values").

The micro-computer 8 operates according to the program shown in FIG. B. 100 ms is started.

The mode of operation of the exemplary embodiment in FIG. 3 is explained with reference to FIGS. 3 to 6.

When the program starts, the values of the cable system determined by the measuring devices 2 to 6 , namely the pendulum angle R (t) , the angular speed (t) , the trolley speed V (t) , the acceleration α (t) and that of the measuring device 7 measured weight m of the load entered (step 21). A speed V _{Z is then} calculated, the current control command u (t) being maintained for Δ T seconds. Velocities V _P and V _N are calculated with the control command increased by Δ u (step 22) according to the following equations:

The running speed of the trolley is then determined as an evaluation value. It is assumed here that the trolley speed consists of two evaluation terms, namely an evaluation value (μ _{V 1} ) in a range of tolerance speeds and an evaluation term (μ _{V 2} ) which corresponds to a target speed.

A function that represents the evaluation term and is divided into sections can be defined, for example, in the manner described below. If x means a speed error and the allowable error is within ± 0.5 m / s, a subdivided function representing the evaluation term that can follow within an allowable range is defined as:

The function µ _{V 2} (x) of the evaluation term that corresponds to the target speed can be defined as follows:

The evaluation value C _V for the current or current speed can therefore be represented by the pair of the two functions µ _{V 1} (x) and µ _{V 2} (x) . If the control command is maintained or u is increased by Δ, the evaluation values C _VZ, C _VP and C _VN for the speed of the scalar quantities V _Z, V _P and V _N to the current speed as follows (step 23) are found after Δ T seconds have elapsed as calculated from equations (1), (2) and (3):

Similar to the trolley speed, the pendulum angle R _Z and the angular speed _{Z are then determined} from the pendulum angle R (t) and the angular speed (t) of the cable system if the current control command u (t) is maintained for Δ T seconds. The pendulum angles R _P , R _N and the angular _{speeds N} and _N are determined when the control command is increased by Δ u.

The movement of the cable system results from the following Equation:

Are in it the angular acceleration of the pendulum angle, G the acceleration due to gravity,l the length of the rope andm the weight of the rope system.

Are

and

then it follows from equation (7):

Here √ is a natural frequency w . R _Z and _Z result from the following equations:

In it are:

The pendulum angles R _P and R _N and the angular velocities _P and _N can also be determined in the manner described above.

From the values R and determined above, the remaining pendulum movements γ _Z , γ _P and γ _N can be determined using the following equations:

Furthermore, the current residual pendulum motion γ can be determined from the following equation (step 24):

The increase or decrease in the residual pendulum movement are determined as evaluation values. It is assumed here that the residual pendulum movement consists of two evaluation terms, namely (µ _γ ₁), which is approximately the same as the current or actual size and (µ _γ ₂), which is smaller than the actual size. The equation system representing the evaluation terms can be defined in the following way, for example. If the residual level movement is denoted by γ and the tolerance is 0.005 rad, the function µ _γ ₁ ( γ ) of the evaluation term, which is divided into sections, can be defined similar to the quantity mentioned above:

The subdivided function µ _γ ₂ ( γ ) of the evaluation term, which is smaller than the current quantity, can be defined as:

Therefore, the evaluation value C _{γ of} the residual pendulum motion can be represented by a pair of two functions µ _γ ₁ ( γ ) and µ _γ ₂ ( γ ) divided into sections. If the control command is maintained or increased by Δ u , the evaluation points C _{γ Z} , C _{γ P} and C _{γ N} for the residual pendulum movement can be calculated as follows (step 25) from the scalar variables γ _Z , γ _P and γ _N for the Determine the remaining pendulum motion after T seconds according to equations (11) to (13):

C _{γ Z} = { C _{γ Z 1} , C _{γ Z 2} }
= {µ _γ ₁ ( γ _Z ), µ _γ ₂ ( γ _Z )}
C _{γ P} = { C _{γ P 1} , C _{γ P 2} }
= {µ _γ ₁ ( γ _P ), µ _γ ₂ ( γ _P )}
C _{γ N} = { C _{γ N 1} , C _{γ N 2} }
= {µ _γ ₁ ( γ _N ), µ _γ ₂ ( γ _N )}

The tax rule is based on the valuation estimate, chosen in the following way:

• (1) maintaining the current command if the pendulum motion does not change while the trolley is running within allowable limits controlled by the current command u (t) ;
• (2) increase the control command by Δ u if the pendulum motion decreases while the trolley is running within permissible limits under a control command which has been increased by Δ u ;
• (3) Reduction of the control command by Δ u if the pendulum motion decreases while the trolley is running within permissible limits under a control command that has been reduced by Δ u,
• (4) increase the control command by Δ u if the pendulum motion does not change while the trolley is running in accordance with a target speed under a control command that has been increased by Δ u,
• (5) Decreasing the command by Δ u when the pendulum motion does not change while the trolley is running in accordance with a target speed under a command that has been decreased by Δ u.

The valuation is derived by choosing a tax rule calculated, which has a maximum value from:

(1) min (C _{VZ 1} , C _{γ Z 1} )

(2) min (C _{VP 1} , C _{γ P 2} )

(3) min (C _{VN 1} , C _{γ N 2} )

(4) min (C _{VP 2} , C _{γ P 1} )

(5) min (C _{VN 2} , C _{γ N 1} )

A control command determined according to this rule is then issued generated (step 26).

With the described embodiment it is possible to control the crane so that the Residual pendulum motion of the cable system is reduced while a constant running speed of the trolley system is retained.

In the second embodiment described below the arrangement according to the invention are the control results of the control commands depending on the values of such Parameters such as position and speed of the trolley, Pendulum angle and angular speed of the rope, determined as evaluation values. This becomes, instead of the Derivation of the rule according to the previous description, an optimal control command is determined.

Fig. 6 shows the block diagram of a second embodiment, in which with 31 a device, referred to below as a pattern generator, the target speed V _T , a target acceleration α T , a target pendulum angle R _T and a target angular speed _T from the measured position X of the crane and the expiry of the time are calculated, with 32 a computing unit which has the function of the units 5 and 6 of FIG. 2 and serves to calculate the speed and acceleration of the trolley from the measured position X of the crane, and 33 denotes a computing unit with the functions of units 2 and 3 of FIG. 2, which calculates the estimated or precalculated values R , the actual pendulum angle and the actual speed from the measured pendulum angle R of the rope.

With 41, 42 and 43 facilities for calculating the evaluation values, z. B. "Speed deviation positive (VP) ", "Speed deviation equal to zero (VZ) " and "Speed deviation negative (VN) " from a difference Δ V between the trolley speed V and a target value V _T , with 51, 52, 53 devices for calculation of evaluation values such as "acceleration deviations positive (α P)", "acceleration deviation is equal to zero (α Z)" and "negative (α N) acceleration deviation" α from a difference Δα between the trolley acceleration and the target acceleration α _T, 61, 62, 63 Devices for calculating evaluation values such as "pendulum angle deviation positive ( R P) ", "pendulum angle deviation equal to zero ( R Z) " and "pendulum angle deviation negative ( R N) " from a difference ΔR between the pendulum angle R of the rope and a target Pendulum angle R _T , and with 71, 72, 73 devices for calculating evaluation values such as "angular velocity deviation positive ( P) ", "angular velocity deviation equ I denote zero ( Z) "and" angular speed deviation negative ( N) "from the difference Δ between the angular speed of the rope and a target angular speed _T.

Further, at 34 denotes an evaluation deriving or calculating means which calculates the evaluation values for the deviation from the target values determined by the evaluation value calculating units 41 to 73 based on the evaluation control rules, and which calculates a change amount Δ u in the control command u estimates or determines, with 35 a control command integration device which adds the amount of change Δ u in the control command to the previous control command u , with 36 a device for feeding and controlling the speed of the trolley and with 37 a crane.

Fig. 7 shows a diagram for explaining the operation of the devices 41, 42, 43, which calculate the evaluation or trend values VP, VN and VN in response to the speed error.

The calculation of the evaluation or estimated values VP, VZ and VN of the speed error Δ V is described with reference to FIG. 7. If the speed error Δ V is input, the VP arithmetic unit 41 generates as shown in FIG. 7 an evaluation value of 0.0 if Δ V ≦ 0, an evaluation value which changes linearly from 0.0 to 1.0 if 0 < Δ V <0.4, and an evaluation value of 1.0 if 0.4 ≦ Δ V.

The VZ arithmetic unit 42 generates an evaluation value 0.0 when Δ V ≦ -0.5 and 0.5 ≦ Δ V , an evaluation value which changes linearly from 0.0 to 1.0 when -0.5 < Δ V <-0.1 and 0.1 < Δ V <0.5 and an evaluation value of 1.0 if -0.1 ≦ Δ V ≦ 0.1.

The VN arithmetic unit 43 generates evaluation or approximation values which are symmetrical to those generated by the VP arithmetic unit 41.

If there is a speed deviation Δ V , the values VP, VZ and VN are obtained from the arithmetic units 41, 42, 43 , which calculate the evaluation values for the speed deviation. If, for example, Δ V is equal to 0.2 m / s, evaluation values VN = 0.5, VZ = 0.75 and VP = 0.0 result.

Evaluation values α P , α Z and α N of the acceleration deviation Δα , evaluation values R P , R Z and R N of the oscillation angle deviation ΔR and evaluation values P , Z , N of the angular velocity deviation Δ are determined in the same way. On the basis of these evaluation values, the magnitude of the change Δ u in the control command is determined or estimated by the computing device 34. The determination rule is:

• (1) If the speed deviation and the acceleration deviation (VZ or α Z) are equal to zero, the control command is retained unchanged, ie Δ u = 0;
• (2) If the speed deviation and the acceleration deviation are positive (VP or α P) , the control command increases slightly, ie Δ u = +0.1 m / s;
• (3) If the speed deviation and acceleration deviation are negative (VN or α N) , the control command is slightly reduced, ie Δ u = -0.1 m / s;
• (4) If the pendulum angle deviation and the angular velocity deviation are equal to zero ( R Z or Z) , the control command is retained unchanged, ie Δ u = 0;
• (5) If the pendulum angle deviation and the angular velocity deviation are positive ( R P or P) , the control command is increased by a small amount, ie Δ u = +0.1 m / s;
• (6) If the pendulum angle deviation and the angular velocity deviation are negative ( R N or N) , the control command is reduced by a small amount, ie Δ u = -0.1 m / s.

More precisely, the estimates or projections above are calculated from the weighting values of the rules, d. H.

(1) R ₁ = min (VZ, α Z)

(2) R ₂ = min (VP, α P)

(3) R ₃ = min (VN, α N)

(4) R ₄ = min ( R Z , Z)

(5) R ₅ = min ( R P , P)

(6) R ₆ = min ( R N , N)

The rule with the highest evaluation value is selected from these rules and the change indication is generated as a pre-calculated result Δ u for the control command integration device 35 , which is determined by the rule.

This embodiment of the arrangement according to the invention enables it is also possible to steer the crane in such a way that the remaining Pendulum movement of the cable system to the same extent as with the embodiment described above is reduced.

Instead of the independent processing units described here some or all of the arithmetic units can be in In the form of a micro-computer. The program can at predetermined time intervals to generate the Control command can be started.

In the following, a third and fourth embodiment described in which a predetermined Control sequence or pattern is followed.

The dynamics of the crane can be divided into

• (1) the rope system and
• (2) the trolley system.

By analyzing each of these systems you can find the conditions to carry out the sway control according to the following Description can be found.

(1) rope system

If the acceleration u of the trolley is constant and the pendulum angle of the cable system is denoted by R and the angular velocity as, then on a phase plane / ω - R ( ω =, l as the rope length) with a point u / g (g is the acceleration due to gravity) a counterclockwise circular trajectory on the R axis as the center.

In order to stop the pendulum movement of the cable system, ie to bring the trajectory to the origin, the trolley must either be regulated in the last step
with positive acceleration u in the second quadrant - (1) of Fig. 8 (a) or
with negative acceleration u in the fourth quadrant - (2) of Fig. 8 (a).

If, however, the pendulum movement is to be suppressed while braking (u <0) of the trolley from a given speed, only the above trajectory (2) is followed. In the following, therefore, only case (2) is considered.

If the trajectory at time t lies in the fourth quadrant, as shown in Fig. 8 (b), the acceleration u = u ⁽¹⁾ , in order to bring (₀ / ω , R ₀) to the origin:

the end

Furthermore result

and as time T ⁽¹⁾ to arrive at the origin

As described, the rope system stops vibrating after the time T ⁽¹⁾ if the trolley is allowed to run at the acceleration u ⁽¹⁾ for the time T ^(1).

(2) trolley system

The position of the trolley is withx, their speed with and the target position withx _M. designated. So that results itself to the trolley at a constant acceleration stop at the desired point:

and

the end

In other words, the trolley is allowed to run for the time T ⁽²⁾ with the acceleration ^{u (2)} and is then stopped at the desired point at the time T ^(2).

From the above sections (1) and (2) it follows that the trolley can finally be stopped in a certain position and the pendulum motion can be stopped if the crane is run in such a way that u ₍₁₎ = u at the final stage of crane operation ⁽²⁾ and T ⁽¹⁾ = T ⁽²⁾ . In general, however, it cannot be guaranteed that the existence of the solutions u *, T * simultaneously satisfies the two above conditions. Therefore, errors in the solutions u *, T * of the practically set acceleration u (u ≠ u *) and the time T (T ≠ T *) serve as error factors when the final end is reached.

The following describes errors, the residual pendulum movement of the rope system and the stop position error of the trolley system at the point in time at which the final end is reached when u ≠ u * and T ≠ T *. However, it is assumed here that the trolley system comes to a standstill after a time T, ie

+uT = 0.

(1) Cable system (see Fig. 9)

Are

then

if t _f = t + T when the final end is reached.

Here is however

and the residual pendulum motion J ⁽¹⁾ is given by

(2) trolley system

The error J ⁽²⁾ in the stop position results from

Fig. 10 is a block diagram showing a device for automatically operating and controlling a crane. Therein indicate 95 probes for detecting the conditions of the crane with the lapse of time, ie for measuring the position, speed, the swing angle of the rope and the angular velocity of the oscillating motion, 96 a processor which on the basis of measuring data by the sensors 95 calculates a control command and sends it to an actuator 97 for a crane 98.

It can be said that the errors J ⁽¹⁾ and J ⁽²⁾ at the final end as shown in (1) and (2) are estimated evaluation indexes at the time when the final end is reached when the control input command u is given under the condition (x,, R ,). The control command is determined by the evaluation derivation based on the estimated evaluation values. The estimation rules are e.g. B .:

• (1) Maintaining the current value when the trolley correct at the present acceleration stopped and the pendulum motion to a standstill can be brought.
• (2) Slight increase in acceleration when the trolley properly stopped by weak acceleration and the pendulum motion properly to a standstill can be brought.
• .
  .
  .

The evaluation indices such as "holds satisfactorily" and "pendulum motion is properly stopped", which are necessary for determining (evaluation derivation) of the control command using the aforementioned control rules, are defined using functions divided into sections. Examples are shown in FIG. 11.

Fig. 11 (a) determines functions µ _GG , µ _GA, subdivided into sections, of evaluation variables,

GG (stops satisfactorily),
GA (stops correctly),

and FIG. 11 (b) determines the functions µ _SG , µ _{SA of} the evaluation variables, which are subdivided into sections,

SG (stops the pendulum motion satisfactorily),
SA (stops the pendulum movement correctly).

Using the above parameters, the estimates or forecast rules e.g. B. written down as follows will:

(1) If G = GA and S = SA , then Δ u = 0.0
(2) If G = GA and S = SA , then Δ u = +0.1
(3) If G = GA and S = SA , then Δ u = -0.1
.
.
.
(n) . . .

Now the "AND" connection of the "IF part" the first rule predetermined evaluation or changeable Group expressed as

P ₁ = GA ∧ SA where ∧ denotes a group of products.

In the following it will be considered which values are assumed by G and S when the control is corrected at time t by a quantity Δ u. The result is the following evaluation group:

P ₁ (t) = (GA ∧ G (J ⁽²⁾ , t)) ∧ (SA ∧ S (J ⁽¹⁾ , t)) (19)

In it is

r ₁ (t) = sup µ _{p 1} (t) (20)

a weighting value of the first rule.

The procedure is carried out in the same way to find the most reliable control rule from the n rules:

Fig. 12 shows the flow of the above-described process.

The above embodiment is based on the pendulum angle R of the cable system and its angular velocity. If the data for the angular velocity are not obtained, they can be determined from the following relationship:

R (t _K ) = { R (t _K ) - R (t _{K -1} )} / (t _K - t _{K -1} )

A fourth embodiment will now be described. The precalculation rule Ri is expressed as follows:

iff _i (X =A. _ix,X = _ix,R. =A. _{i R.}),
theny =G _i (X,,R.)
=P. _OK +P. _ixX +P. _ix +P. _{i ϑR}

whereinA. _ix,_ix andA. _{i R.} Represent groups of weighting values, which are the variable ranges ofX, andR. determine. f _i is a state of the forecast ruleRi expressive logical function andG _i a feature that y the endX, andR. determined and in this case a linear one Function is.

Assuming that there are a total of n estimation or precalculation rules, the truth value is determined by

where // denotes the truth values.

From the practical control data, the parameter P of the linear function g _{i is determined} in the following manner based on the least squares method.

First, in the event that the pendulum movement is correctly stopped during an operation carried out by an experienced operator or by computer simulation, observed values (X ,, R ) are stored; the data for the control command y are also stored. Assume that the data is obtained in m number and expressed as

In this case, the truth value w _i is further given by the above-mentioned prediction rule

w _i = / f _i (x _K - A _ix , _K - A _i , R _K = A _{i R} /

This results in

The above become the parameters

y = P _io + P _ixX + P _ix + P _{i ϑR}

based on the weighting method of the smallest Squares precalculated or estimated.

In this embodiment, the parameters are evaluated Tax rules to regulate the crane from favorable Get results, e.g. B. such from an operator or carried out by computer simulation Regulations. Therefore, the crane can do it automatically and correctly operated unaffected by external influences.

It should be noted here that the function g for determining the control command y from X ,, R is not restricted to the linear function mentioned in the above exemplary embodiment.

According to the first to fourth embodiments is a method for automatically controlling a crane realized in which a control command by evaluation determination is set, taking from measured data such as the position and speed of the trolley and the pendulum angle and angular speed of the rope assumed will. Therefore, the pendulum motion of the load becomes even then stopped while maintaining a high speed, when unexpected influences such as initial pendulum movements or gusts of wind are present.

Fig. 3 is a schematic model diagram showing a crane, for which the fifth embodiment is described below. The crane moves a load by raising or lowering a rope 104 , from which a load 103 is suspended, while a trolley 101 moves on a rail 102. The trolley 101 is controlled by a suitably constructed servo system according to a precalculated speed pattern (e.g. Fig. 14), so that it moves to the destination within the shortest possible time and in such a way that the rope stops vibrating, when the trolley has reached the destination.

In the arrangement of Fig. 13, the relationship between the swing angle R and the angular velocity of the pendulum motion is expressed by

R = A R + b γ

wherein

Are in it

l the pitch,
the rate of change of the rope length,
G the acceleration due to gravity,
γ the acceleration of the trolley.

If one solves the above equation analytically under the condition that l is constant, the trajectory of the solution is a counterclockwise circle on the phase plane of / ω and forms R with R equal to γ / g as the center ( ω =). Using the results, the speed pattern of Fig. 14 can be calculated and a relationship between R and / ω ( Fig. 15) can be found in each section (see "Control System for Suppressing the Swing of Crane for Use in Yards"). , Preparatory Documents in the 22nd Society of Instrument and Control Engineers Academid Lectures, pp. 533-534, 1983). According to FIG. 15, a maximum pendulum angle ( R _max = ± | γ | / g) is observed in sections (3) and (9) in the speed pattern of FIG. 14. In this condition it is stopped.

If one considers the dynamic behavior of the rope system in sections (3) and (9), the above equation shows that the system is not influenced by the rope length l or the change in rope length. In other words, in these sections the trajectory of the solution for R does not go beyond the standstill point even if the rope length l changes. This condition is called the "balanced vibrational state". If the rope is therefore caught in section (3) and released in section (9), the load can be raised or lowered while the trolley is moving to carry out the load movement. The balanced vibrational state can be determined from the observed results of R or.

I. E. in the balanced oscillation state = 0 applies. Whether the balanced state of shrinkage in the section (3) or (9) is present, depending on the positive or negative sign of the acceleration the Trolley to be determined. The basic rules of control are as follows:

• (I) Are <0 and = 0, the rope is hauled in. The retrieval speed is set so that the rope at the timet _fa a predetermined lengthl _m Has, when the end of section (3) is reached.
• (Ii) Are <0 and = 0, the rope is output. The output speed is adjusted so that the rope at the timet _fd a predetermined lengthl _M. Has, when the end of section (9) is reached. is it difficult to direct the angular velocity too measure so willΔR is used, according to the following equation is calculated: ΔR = {R. (t _k) -R. (t _{k -1})} /Δ t (22)

wherein

t = t _k - t _{k -1} .

The inhaul and output speed at certain intervals t _K repeatedly adjusted so that the estimated error Δ finally l of Seilänge is zero. The following equation is used:

(I) In section (3):

Δ l = { l (t _k ) + (t _k ) (t _fa - t _k )} - l _m (23)

(II) In section (9):

Δ l = { l (t _k ) + (t _k ) (t _fd - t _k )} - l _M (24)

wherein

l (t _k ) is the rope length at time t _k and
(t _k ) is the speed of change of the rope length at time t _k , ie the retrieval or output speed.

Regarding the basic structure of the crane or the working environment in which it is used, it is difficult to accurately measure the swing angle of the rope or its angular speed, or to precisely adjust the speed at which the rope is taken in or out. It is also difficult to measure the rope length precisely at every moment or to calculate it from the rope length previously retrieved or issued. The aforementioned values or ΔR and Δ l are therefore also inaccurate. In the following, therefore, a control method for processing these quantities as evaluation or imprecise or approximate values will be described.

Using ΔR of the equation (22) instead, the values ΔR and Δ l treated as evaluation values are expressed as Δ and Δ l . The positive or negative sign of x can be easily and clearly determined depending on the data from the system for controlling the speed of the trolley. To simplify the description, a case is considered in the following in which the rope is hauled in in section (3), it being assumed that the positive or negative sign of x has already been determined, ie the acceleration or deceleration section has already been determined.

In section (3) the evaluation and determination rules necessary for winding or unwinding the rope are as follows at each moment t _k:

In it are:

Δ(iα):Δ l (i = 0 ton), if(t _k) =i α in equation (23);
α: Minimum controllable amount of rope speed .

The symbols VS and G denote evaluation variables with the meaning "very small" or "good"; they are defined by the functions µ VS and µ G, which are subdivided into sections, for the variables shown in FIGS. 16 (a) and 16 (b). The symbols Δ and Δ l also denote weighting values which can be defined by similar functions μ Δ and μ Δ l , which are not shown and are divided into sections.

The evaluation derivation is carried out by checking the conditions of the above-mentioned rules (1) to (n) at each moment t _k and determining (t _k ) by which the value

is maximal, in which are:
∧: The calculation of a group of products,

µ₁ ( Δ ) and µ₂ ( Δ l) represent the extent to which Δ and Δ l are VS and G , respectively.

The same also applies to section ( 9 ) if the value (t _k ) is used with a negative sign and equation (24) is used instead of equation (23).

Fig. 17 shows the block diagram of a further embodiment. A processor 106 receives data from a device 105 which measures the position and speed of the trolley and calculates an operational quantity so that the trolley speed pattern of FIG. 15 is followed. It controls an actuator, for example an electric motor 107 , which accelerates or brakes the trolley. This system is constructed in the same way as the conventional one. A processor 109 receives the measurement data from a rope swing angle measuring device 108 and determines the value (t _k ) by performing evaluation derivation with reference to the trolley acceleration data supplied from the processor 106 of the trolley speed control system. Therefore, an actuator for hauling in / out the rope (e.g., a motor) 1010 is controlled accordingly. The crane is indicated schematically by block 1011.

FIG. 18 shows the flow chart of the program executed by processor 109 of FIG. 17 when the rope is being hauled in and paid out. The pendulum angle R (t _k ) of the rope is read off at the instant t _k after predetermined time intervals Δ t and the pendulum angular velocity ΔR (t _k ) of the rope is calculated according to equation (22).

Further is obtained from the or at that moment output pitch for this moment the pitch l (t _k) calculated. It will depend on the Trolley acceleration data from the processor 106 of the trolley acceleration control system the mistakeΔ l the pitch for each = ±α, ± 2α, . . ., ±n α (negative sign for the acceleration section) calculated according to equation (23) if the Trolley located in the acceleration section, or after Equation (24) when the trolley is in the braking section is located.

Then for each of the calculated errorsΔ l the sizes µ₁ (Δ) according to equation (26) and µ₂ (Δ l) according to equation (27) calculated. At this moment the function values can on the right side of the two equations by exponential function approximation or by approximation of fold lines calculated or read from a table. Further can change the group of products by choosing a smaller one Between two function values. The calculation continues with the selected smaller value according to equation (25) for all rope speeds executed. A maximum value is called the operation size(t _k) sent to the actuator hauling in the rope or issues.

A control for suppressing the oscillation of the rope will now be described. The principle of control is to repeat such an operation so that the rope length is decreased when the maximum swing angle of the pendulum motion is reached and that the cable length is increased when a swing angle equal to zero, the lowest point, is reached. Fig. 19 (a) schematically shows the mechanism and Fig. 19 (b) shows the process in which the pendulum motion is damped after the mechanism. This operation corresponds to the propagation of an oscillation, but the phase is changed by 180 °. This control is preferably carried out in the section ((6) in FIGS. 14 and 15) in which the oscillation angle should be kept at zero when the trolley is moving. However, even after the trolley has stopped, this operation must be carried out or continued in order to suppress the swaying movement resulting from the residual swaying movement of the rope.

Therefore, the basic rules of control can be summarized as follows will:

(I) If = 0 and = 0: winding the rope,
(II) If = 0 andR. = 0: unwinding the rope.

The control according to these rules can be carried out with the control arrangement shown in FIG. However, as mentioned earlier, it is difficult to accurately measure the swing angle of the rope and its angular velocity. In addition, the above-mentioned principle of damping does not require strict compliance with the condition of the two rules. Therefore, the case in which the evaluation control is carried out will also be described below.

The rules mentioned above can be sorted according to the rules express the valuation estimate or approximate estimate:

(i) If =VS and =VS, then = -α
(ii) If =VS andR. =VS, then = +α
(iii) Ifx =NVS, then = 0.

The symbols VS and NVS (not very small) represent evaluation variables, VS being determined as a function μ _VS for variables according to FIG. 16 and μ _NVS as 1 - μ _VS.

If the angular velocity cannot be measured directly, ΔR from equation (22) can be used.

The evaluation derivation fulfills the rule of the greatest Truth value under rules (i) to (iv) for which the truth values are found. I. E. it will be for rules (i) to (iv) perform the following calculations:

and it becomes one by a rule with the greatest value certain value chosen.

The above control can also be carried out according to the program of FIG. 20 with the control arrangement of FIG . The processor 109 reads R (t _k ) and (t _k ) from the cable swing angle measuring device 108 (or R ( t _k ) is calculated) and from the processor 106 x (t _k ) at predetermined times t _k in predetermined intervals Δ t are delivered. The processor 109 then performs the evaluation derivation, ie it calculates a group of products or selects a minimum value according to equation (28) and determines (t _k ), whereby the actuator 1010 is controlled, which hauls in or out the rope.

In the arrangement described, processors 106 and 109 are described and illustrated as separate units. However, they can also be replaced by a single processor.

In the fifth and sixth embodiments described becomes the rope length when the trolley is running controlled so that the load as part of the load movement is raised or lowered. Furthermore can unexpected pendulum movements of the rope are prevented, which often arise when the trolley is running. As a result the efficiency of the load movement can be seen to be considerable to enhance.

Claims (5)

1. A method for the automatic control of a crane with a trolley ( 11 ), a rope (12 ) hanging down from the trolley ( 11 ) and devices for controlling the movement of the trolley ( 11 ), wherein in a first method step at least one value of a first parameter (V (t)), which reflects the current state of the trolley (11 ), and a value of a second parameter ( R ), which reflects the current state of the rope ( 12 ), are detected, characterized in that,

• - That in a second step on the basis of the detected values of the first and second parameters values of a third parameter (V) for determining a condition of the trolley or values of a fourth parameter (r) for determining a condition of the rope are estimated, the The values of the third and fourth parameters are those values that would be obtained as a result of the control operations under a large number of different control commands if these were each fed to the trolley,
• that in a third step, on the basis of the estimated values of the third and fourth parameters, one of the different control commands is selected in accordance with a conclusion based on fuzzy or indefinite values, and
• - That in a fourth step the trolley is controlled with the aid of the selected control command.

2. The method according to claim 1, characterized in that the third step comprises:

• - Converting the estimated values of the third parameter (V) into fuzzy values using at least one first predetermined function (µ _{V 1} , µ _{V 2} ) which is subdivided into sections, which shows the relationship between the value of the third parameter (V) and the represents a fuzzy value in order to indicate the degree of fulfillment in relation to the result of the control of the trolley ( 11 ), and
• - Converting the estimated values of the fourth parameter (r) into fuzzy values using at least one second predetermined function (µ _{r 1} , µ _{r 2} ), which is divided into sections, which shows the relationship between the value of the fourth parameter (r) and the reproduces fuzzy values in order to indicate the degree of fulfillment with regard to the result of the control of the rope ( 12 ),
• the control command is selected by comparing the fuzzy values of the third and fourth parameters.

3. The method according to claim 1, characterized, that the first predetermined function relates to the optimal one Speed of the trolley and the second predetermined Function relates to the optimal swing angle of the rope. 4. The method according to claim 2, characterized, that the first predetermined function relates to the optimal one Stop the trolley and the second predetermined one Function relates to the optimal swing angle of the rope. 5. The method according to any one of claims 1 to 3, characterized, that the values of the third and fourth parameters in one Case in which the control command is in its current state is maintained, and appreciated in one case in which the control command is based on its current state by a certain amount in a certain direction for increasing or decreasing the speed the trolley is changed.