EP0394147A1 - Device for controlling automatically the position and the oscillations of a suspended load during its transportation by a lifting device - Google Patents

Abstract

Device for controlling automatically the position and the oscillations of a suspended load during its transportation by a lifting device, comprising: - at least one guide means on which moves a translationally moveable carriage, actuated by a motor, to which is hooked a cable (or a rod) from the other end of which is suspended the load to be transported; - at least one position sensor, - an angle sensor, arranged on the carriage, characterised in that it comprises: - a first control loop comprising: . a first comparator (22> . a proportional corrector (23> . a second so-called position control loop, comprising: . a second comparator (20> . a proportional and integral corrector (21> - an adder (24). Application to any transportation of a suspended load requiring special conditions. <IMAGE>

Description

The present invention relates to an automatic device for controlling the position and oscillations of a suspended load during its transfer, in particular by means of a lifting device.

The invention can be adapted to lifting devices of various types, such as for example gantries, overhead cranes or cranes. The handling of a load by a lifting device causes the load to swing, which it is desirable to absorb as completely as possible so that the load can be deposited in a specific location without difficulty. In addition, in the context of handling environmentally hazardous loads, it is desirable to transfer said loads in a relatively short time and above all, taking care that they remain in a perfectly stable and vertical plane.

Currently, only crane operators with long experience manage to achieve quite relative stabilization of load oscillations. However, these stabilizations are not satisfactory since they generally require time to obtain them.

Rigid masts or telescopic masts integral with a carriage movable in translation have also been used. However, this type of installation remains expensive and cumbersome, because it involves the arrangement of a volume located above the carriage in order to allow the passage of the mast. So, if this achievement is satisfactory on the technical level, on the other hand, it is not compatible with mass industrialization adapted to any situation.

It has also been proposed to guy the load by means of symmetrical inclined cables. This guy wire turns out to be long and tedious to set up, and is therefore not suitable for repetitive load transfer.

In French patent FR-A-2 399 378, a load anti-sway device of the type in question has been proposed, in which the transfer member undergoes determined sinusoidal acceleration laws. However, if this law is valid for a given period, that is to say for a given cable length, on the other hand, it is not suitable for a variation of said cable length.

In French patent application FR-A-2 598 141, a system controlled in minimum time has been described. In this system, the load is accelerated in both directions to compensate for the swaying. In fact, we do not deal with oscillations during the transfer, but only at the start and at the finish. However, as already said, these oscillations are prohibitive for certain applications, in particular in the context of the environment and the safety of personnel. Such a system has also been described in German patent DE-A-3,513,007, which as in the previous case, operates in open loop, in other words, does not have a servo device, both on the position and on the oscillation angle of the pendulum load, and therefore does not allow continuous control and damping oscillations during transfer.

British patent GB-A-2,030,727 describes a system of the type in question, in which one is concerned only with the absence of residual oscillations at the place of final transfer of the pendulum load. To do this, he suggests adding an analog simulator to the installation, designed to calculate repeatedly tive the speed to be imposed on the transfer member, and this as a function of the data entered initially, namely the coordinates of the initial and final positions of the load to be transferred. As in the previous cases, there is no control on the position and on the oscillation of the load, so that we find the same drawbacks.

The present invention aims to overcome these drawbacks. It offers a servo device for quickly transferring a suspended load with small oscillations, comprising:
- At least one guide means on which moves a mobile carriage in translation actuated by a motor, to which is hung a cable (or a rod) at the other end of which (from which) is suspended said load to be transferred;
- At least one position sensor intended to permanently locate the position of the carriage relative to a determined point, and to provide a so-called position signal corresponding to this location;
- An angular sensor arranged on the carriage, intended to measure the value of the angle formed by the cable or the rod relative to the vertical, and to supply a so-called angular signal corresponding to this measurement.

The present invention is characterized in that said device comprises:
- a first control loop comprising:
. a first comparator intended to provide the angular difference between the angular signal thus determined and a predetermined reference angular value, by means of a reference model, as a function of the destination of the load and of the desired translation speed;
. a proportional corrector intended to multiply the difference thus determined by a constant fixed in advance;
- a second so-called position control loop, comprising:
. a second comparator intended to provide the difference between the position signal and a theoretical reference position predetermined by the same reference model, according to the same data;
. a proportional and integral corrector intended on the one hand, to multiply the said deviation by a predetermined constant and on the other hand, to integrate the said deviation with respect to time over an equally predetermined interval;
an adder, connected respectively to the output of the proportional corrector, of the proportional and integral corrector and of an amplifier intended to amplify the value of the reference speed, also determined by means of said reference model, as a function of the same data, the said adder being intended to add the values from these outputs to deliver a signal capable of actuating the carriage control motor.

In other words, the present invention is characterized in that the load transfer device is provided with two control loops connected on the one hand to sensors measuring respectively the angle formed by the cable or the rod relative to the vertical and the position of the carriage, and on the other hand to the motor control member actuating the carriage, the control being carried out as a function of the values determined by a reference model taking account of the desired speed and from the load transfer point.

According to the present invention, the reference values determined by the so-called reference model, namely the angular position, the speed and the position, come from the following law:
Γ = ϑ˝ (l₀ + vt) + 2vϑ ′ + gϑ
where ϑ, ϑ ′ and ϑ˝ denote respectively the angle, the angular speed and the angular acceleration, g denotes the acceleration of gravity, l₀ the initial length of the cable (or rod), v the constant speed of variation in the length of the cable (or rod), and Γ the acceleration of the carriage.

In a particular case of the present invention, for which the length of the pendulum defined by the load does not vary, the reference values determined by the reference model come from the following law:
Γ = ϑ˝l₀ + gϑ

Advantageously, in practice:
- The carriage is movable in direction on a first rectilinear guide means, itself movable in translation on a second rectilinear guide means, coplanar but in a direction perpendicular to that of said first guide means, said second guide means being actuated by means of a motor also controlled by means of two control loops (in position and in angle) of the same kind as those previously described for the directional movement, the control being dependent on a reference model which is specific to said engine;
- the carriage is mobile on an overhead crane;
- The carriage is movable in a radial movement on a first guide means, itself movable in rotation by means of a rotation member, actuated by means of a motor also controlled by a loop of enslavement according to a reference model which is also specific to it;
- the carriage is movable along the horizontal boom of a crane;
- the servo calculations are carried out approximately twenty times per second by means of a digital calculation unit.

• La figure 1 représente un chariot de transfert guidé par un rail unidirectionnel.
• La figure 2 représente le même chariot susceptible de se déplacer selon deux directions différentes et coplanaires, tel qu'un pont roulant.
• La figure 3 est un schéma représentant le déplace­ment plan de la charge au moyen d'un pont roulant, en accord avec la présente invention.
• La figure 4 représente une grue munie de sa charge susceptible de se déplacer également selon au moins deux directions différentes et coplanaires.
• La figure 5 est un graphique représentant en regard l'une de l'autre les courbes caractéristiques respec­tivement de l'accélération, de la vitesse, de l'angle et de la position communiquées au chariot selon le modèle de référence conforme à l'invention pour une longueur fixe du pendule défini par la charge, correspondant à la longueur initiale l₀.
• La figure 6 est un schéma du principe de fonction­nement et de l'asservissement du dispositif conforme à l'invention.
• La figure 7 est un schéma de principe de la déter­mination des valeurs de référence au moyen du modèle de référence conformément à la présente invention.
• La figure 8 est un graphique représentant deux courbes en regard l'une de l'autre correspondant respec­tivement à la variation angulaire et à la position de la charge en fonction du temps, dont la longueur varie au cours du transfert, l'appareil de levage auquel est suspendu la charge étant asservi conformément à l'inven­tion.
• Les figures 9 et 10 représentent une courbe corres­pondant au signal d'accélération fourni par le modèle de référence à l'appareil de levage auquel est suspendue la charge, dans lesquelles respectivement, la longueur du pendule constitué par la charge varie jusqu'à la fin du transfert, ou est stoppée avant la fin de celui-ci.

The manner in which the invention can be implemented and the advantages which result therefrom will emerge more clearly from the following exemplary embodiments, given by way of indication but not limitation, in support of the appended figures.

• Figure 1 shows a transfer carriage guided by a unidirectional rail.
• FIG. 2 represents the same carriage capable of moving in two different and coplanar directions, such as an overhead crane.
• Figure 3 is a diagram showing the planar displacement of the load by means of an overhead crane, in accordance with the present invention.
• FIG. 4 represents a crane provided with its load capable of also moving in at least two different and coplanar directions.
• FIG. 5 is a graph showing opposite each other the characteristic curves of the acceleration, the speed, the angle and the position respectively communicated to the carriage according to the reference model according to the invention for a fixed length of the pendulum defined by the load, corresponding to the initial length l₀.
• Figure 6 is a diagram of the operating principle and the control of the device according to the invention.
• Figure 7 is a block diagram of determining reference values using the reference model in accordance with the present invention.
• FIG. 8 is a graph showing two curves facing each other corresponding respectively to the angular variation and to the position of the load as a function of time, the length of which varies during the transfer, the lifting device to which is suspended the load being controlled according to the invention.
• Figures 9 and 10 show a curve corresponding to the acceleration signal provided by the reference model to the lifting device to which the load is suspended, in which respectively, the length of the pendulum constituted by the load varies until the end of the transfer, or is stopped before the end of it.

The device according to the invention will first of all be described in the case of a rectilinear unidirectional movement of a load (5) (see FIG. 1), suspended by means of a cable (4) from a carriage ( 1). The carriage (1) moves on non-deformable rails (2), by means of a motor, not shown. The carriage (1) further comprises a winding cylinder (3) capable of varying the height of the load (5).

The motor causing the carriage (1) to move is a variable speed motor, the speed of which is controlled by a device commonly known as an "electronic speed variator", to which a speed setpoint is applied defined by a voltage applied between two points , said voltage being proportional to this speed.

où g est l'accélération de la pesanteur et l la longueur du pendule.During movement, the suspended load (5) has an angular deviation from the vertical. As is quite known, the theory gives as value of the period of the oscillations:
T = 2π√ $\frac{\text{l}}{\text{g}}$

where g is the acceleration of gravity and l the length of the pendulum.

The carriage (1) has an angular sensor, not shown, located in the vertical plane defined by the movement, and intended to measure the angle ϑ. This potentiometer type angular sensor is well known for this application.

The device according to the invention also has a sensor for the actual position of the carriage on the rails (2) relative to a predetermined original position. This sensor, not shown, may be constituted by a rack fixed to the rails (2), on which a digital encoder meshes fixed to the carriage (1). In this way, the actual position of the carriage is precisely defined and instantly known.

(pos.réf.- pos. réelle)dt
où K₁ et K₂ désignent les dites constantes prédetermi­nées.In order to limit the oscillations of the suspended load (5) during transfer while carrying out the latter at a bearing speed predetermined by the operator, the motor for actuating the carriage (1) is doubly controlled. The actual position measured as previously described, is compared with a reference position determined by means of a reference model which will be described later. The difference obtained at the level of a comparator (20) is then subjected to a processing at the level of a proportional and integral corrector (21). This processing consists, in a known manner, on the one hand of multiplying the difference thus obtained by a predetermined constant as a function of the characteristics of the system, and on the other hand, of adding to this value the integral with respect to time of said difference of position, the output signal of said corrector (21) being of the form:
K₁ (pos.ref-pos.eal) + K₂ ∫ $\genfrac{}{}{0ex}{}{\text{t}}{\text{o}}$

(ref pos. - actual pos.) dt
where K₁ and K₂ denote the said predetermined constants.

Likewise, the actual measured value of the angle ϑ formed by the suspended load with the vertical, is also compared by means of a comparator (22) with a reference angle also determined by means of the reference model. The angular difference thus determined undergoes processing in a proportional corrector (23). The processing followed in the proportional corrector consists of a simple multiplication by a predetermined constant also a function of the characteristics of the system of the value of said difference.

The two values thus determined, from the correctors (23) and (21) are added in a summator (24) to a third value corresponding to the reference speed from the reference model, which is previously multiplied by a constant in the amplifier (25).

The signal from the adder (24) is then transformed into an analog signal by means of a digital analog converter not shown, then transferred to the actuation motor of the carriage (1). The two control loops are thus completely described.

Advantageously, the comparators (20,22) as well as the two correctors (21,23) and the amplifier (25), as well as the summator (24) and the digital to analog converter, are concentrated in one and the same card. electronic processing. Said card is managed by means of a microprocessor entering the angle and actual position measurements as well as the servo commands some twenty times per second. In this way, the possible oscillations of the load are rigorously controlled and immediately compensated for by said control loops.

The same device is applied in the context of a plan transfer. As can be seen in Figure 2, the carriage (1) moves in direction (Y) on the rails (2), the latter themselves moving in translation (X) coplanar and perpendicular to the direction (Y) of the carriage (1). The rails (2) integral with each other are guided on rectilinear rails (6), also undeformable.

The device according to the invention involves a displacement of the load in a vertical plane, in particular shown in FIG. 3. In this way, the movement of the load results from the component of two movements in translation on the one hand, in the direction on the other hand. The operator himself defines the level speeds for each of the movements V _X and V _Y. Depending on the spaces to be traversed P _X and P _Y , the respective times P _X / V _X and P _X / V _Y are not necessarily equal. The slowest movement imposes its law on the resulting movement and in fact, the speed of the second is adjusted to achieve the movement at the same time and thus have a synchronization of the two movements which results in a movement located in a vertical plane passing through the starting and ending points of the charge.

The two movements according to (X) and (Y) are independent and virtually decoupled, in particular for the control, apart from the fact that the speeds are calculated beforehand in order to obtain a displacement according to a fixed vertical plane containing the two starting points. and destination of the load.

This therefore results in two laws of simultaneous movements, one along the X axis, the other along the Y axis. Each corresponds to a reference model and a servo. The microprocessor performs the calculations for both at each time period, about twenty times per second. In this way, there are two servo loops for each of the two components of the movement, respectively a servo loop in position in association with a position sensor, and an angular servo loop in association with an angular sensor. So there are finally four control loops.

In this embodiment, the angular sensors are located in two perpendicular planes, namely one defined by the vertical plane of translation, and the other by the vertical plane of direction.

In another embodiment, the device is suitable for transferring a load suspended from a crane (see FIG. 4). In known manner, the movements of a crane are respectively radial along the carrying beam (10) and circular around its pivot axis (11). As in the two previous cases, the displacement of the load is carried out along a vertical plane, from the position P₁ located at the distance R₁ from the pivot axis, and forming an angle α₁ with respect to the reference axis OX, at a position P₂, distant from the pivot axis (11) by a distance R₂, forming with the reference axis OX an angle α₂. In this way, this displacement in a straight line requires beforehand to calculate in the Cartesian space OX-OY the equation of the said straight line, to determine in projection on the axes OX-OY the laws of motion with the reference models and enslavements as before, and then to transform the Cartesian equations into polar equations corresponding respectively to the circular movement and to the radial movement. In fact, the microprocessor therefore simply performs an additional step of transforming Cartesian coordinates into polar coordinates, and this by means of an appropriate converter.

The reference model according to the present invention will now be described in more detail. In the different calculation steps which follow, we will assimilate the suspended load to a simple pendulum of mass m.
g represents the acceleration of gravity;
l₀ the initial length of the pendulum;
v the constant speed of variation of the length of the load support;
ϑ the angle of the cable with the vertical;
Γ the acceleration of point P representing the carriage.

The pendulum theory sufficiently known not to be developed further here, allows to arrive with the theory of LAGRANGE to the following expression:
2mll′ϑ ′ + ml²ϑ˝ = mΓlcosϑ - mglsinϑ
where the symbols ′ and ˝ respectively represent the first derivative and the second derivative of the corresponding variable with respect to time.

Since the angle ϑ with respect to the vertical remains sufficiently small, it is reasonable to linearize this equation with respect to ϑ:
2l′ϑ ′ + lϑ˝ + gϑ = Γ

The variation of l is assumed to be linear from l₀. We can therefore write:
l = l₀ + vt
hence the following expression:
ϑ˝ (l₀ + vt) + 2vϑ ′ + gϑ = Γ (Ω)
which corresponds to a variable linear differential equation. In the simplest case where the length of the pendulum remains constant, and using the LAPLACE transform to solve the previous equation, we find that for a constant acceleration Γ, the period T of oscillation of the pendulum is equal to:
T = 2π√ $\frac{\text{l}}{\text{g}}$

These different results are listed in FIG. 5 in which the curve I represents the constant acceleration Γ, the curve II represents the variation of the speed, the curve III represents the variation of the angle and finally, the curve IV that of the position and this, for different phases, respectively phase A of speed increase, phase B which corresponds to a plateau corresponding to the maximum speed, and phase C corresponding to the descent in speed. The acceleration Γ is constant for a time T equal to the period of the pendulum and the deceleration of the same value for the same duration. In fact, the other three curves II, III and IV are deduced therefrom. As we can see, for a constant acceleration, the increase in speed is linear. On the other hand, when the acceleration stops, a speed step is obtained. Likewise, the deceleration being constant, there is a symmetrical speed descent with respect to the ascent.

The position curve is obtained by simple integration of that of speed with respect to time.

The value of the angle is a sinusoidal function of time, which is canceled during the stop and becomes symmetrical during the deceleration compared to its value during the acceleration phase.

For each movement, the operator fixes the system data which are:
- the predetermined value of the level speed of the carriage;
- the length of the pendulum;
- the coordinates of the point of arrival.

These three data allow the three references to be calculated as a function of time, respectively the speed of the carriage v, the angle of deflection ϑ and the position of the carriage P. In fact, the law is thus perfectly defined by this reference model. If the pendulum is perfect, correctly identified, if the motor actuating the carriage has no dynamics and its gain is equal to 1, and in the absence of disturbance, the real behavior of the pendulum is that of the reference model.

However, due to the observed and measured differences between reality and theory, the model is associated with a control system, the operating principle of which is shown in FIG. 6.

The servo-control applied to the motor makes it possible to follow as closely as possible the values defined by the reference model to arrive at a zero angular deviation at the point of arrival of the load by correcting the differences between theory and reality. In addition, the effects of possible disturbances are also attenuated, for example the oscillations of the start when the load is taken and the unexpected shocks during transfer.

As already said, the microprocessor performs the various calculations approximately 20 times per second in order to readjust each time the speed reference value of the carriage.

The servo-control at the end of movement is maintained until the position is reached with the required tolerance and the oscillation is damped to reach a maximum value within the required tolerance.

In the general case where the length of the pendulum is variable, for example to avoid an obstacle, the control remains the same, but the reference model is slightly more complex. The starting point becomes again the equation seen above referenced Ω:
ϑ˝ (l₀ + vt) + 2vϑ ′ + gϑ = Γ (Ω)

The principle of the command is to generate an acceleration law Γ such that the behavior of the pendulum constituted by the load is that of the pendulum of constant length l₀. It then becomes possible to use the device described in the case of controlling a pendulum of constant length. The variable linear terms are then compensated by generating a command of the type:
Γ = τ _P + compensation terms
the term τ _P will be defined later.

The compensation terms are functions of the variation of the length l of the pendulum and make it possible to transform the pendulum of variable length into a pendulum of constant length l₀ whose period is:
T = 2π√ $\frac{\text{there}}{\text{g}}$

The compensation terms are taken from the reference model already used in the case of the pendulum of constant length, this model thus giving ϑ ′ _ref , ϑ˝ _ref by deriving ϑ _ref .

The rise of the load begins at the same time as the movement of the carriage, but the fact that this rise can end either during the rise in speed of the carriage, or during the leveling, or even during deceleration, five different cases can present themselves for which the previous equation is written differently.

• 1/ Pendant la mise en vitesse du chariot et alors que la longueur du pendule varie avec une vitesse v, l'équation de commande devient :
  Γ = ϑ˝_ref.vt + 2vϑ′_ref + τ_P
  pour t ≦ T ou t ≦ t_e, t_e représentant le temps de variation de la longueur du pendule.
  L'équation Ω devient alors :
  ϑ˝l₀ + gϑ = τ_P
  dans le cas où ϑ′ = ϑ′_ref et ϑ˝ = ϑ˝_ref
• 2/ Pendant la mise en vitesse du chariot et après la variation de longueur du pendule (si celle-ci est différente de l₀), la commande devient :
  Γ = ϑ˝_ref(l_e - l₀) + τ_P
  pour t_e ≦ t ≦ T et l_e longueur finale du pendule en t_e.
  L'équation Ω devient alors :
  ϑ˝l₀ + gϑ = τ_P si ϑ˝ = ϑ˝_ref
• 3/ Pendant le palier et quelle que soit la varia­tion de longueur du pendule, la commande a pour valeur :
  Γ = O
• 4/ Pendant la décélération du chariot et alors que la longueur du pendule varie avec une vitesse v′, la commande est donnée par :
  Γ = ϑ˝_ref.v′t + 2v′ϑ′_ref + ϑ˝_ref(l_ε - l₀) + τ_P
  après translation de l'origine des temps en t = t_d, instant de début de décélération.
  L'équation Ω qui vaut dans ce cas :
  ϑ˝(l_e + v′t) + 2v′ϑ′ + gϑ = Γ
  devient alors :
  ϑ˝l₀ + gϑ = τ_P si ϑ˝ = ϑ˝_ref et ϑ′ = ϑ′_ref
• 5/ Pendant la décélération du chariot et alors que la variation de la longueur du pendule est arrêtée en :
  t = t_d + t′_e
  et qu'elle est égale à l₀ + vt_e + v′t′_e, la commande est donnée par :
  Γ = ϑ˝_ref(vt_e + v′t′_e) + τ_P
  l'équation Ω devient alors :
  ϑ˝l₀ + gϑ = τ_P si ϑ˝_ref = ϑ˝

The example which follows illustrates a particular case of variation of length of the pendulum, but it is understood that one can generalize this type of calculation to any type of variation of length.

• 1 / During the setting in speed of the carriage and while the pendulum length varies with a speed v, the control equation becomes:
  Γ = ϑ˝ _ref .vt + 2vϑ ′ _ref + τ _P
  for t ≦ T or t ≦ t _e , t _e representing the time of variation of the length of the pendulum.
  The equation Ω then becomes:
  ϑ˝l₀ + gϑ = τ _P
  in the case where ϑ ′ = ϑ ′ _ref and ϑ˝ = ϑ˝ _ref
• 2 / During the setting in speed of the carriage and after the variation of the pendulum length (if this is different from l₀), the command becomes:
  Γ = ϑ˝ _ref (l _e - l₀) + τ _P
  for t _e ≦ t ≦ T and _e final length of the pendulum at t _e.
  The equation Ω then becomes:
  ϑ˝l₀ + gϑ = τ _P if ϑ˝ = ϑ˝ _ref
• 3 / During the landing and whatever the length of the pendulum, the value of the command is:
  Γ = O
• 4 / During the deceleration of the carriage and while the length of the pendulum varies with a speed v ′, the command is given by:
  Γ = ϑ˝ _ref .v′t + 2v′ϑ ′ _ref + ϑ˝ _ref (l _ε - l₀) + τ _P
  after translation of the origin of the times in t = t _d , time of start of deceleration.
  The equation Ω which is valid in this case:
  ϑ˝ (l _e + v′t) + 2v′ϑ ′ + gϑ = Γ
  then becomes:
  ϑ˝l₀ + gϑ = τ _P if ϑ˝ = ϑ˝ _ref and ϑ ′ = ϑ ′ _ref
• 5 / During the deceleration of the carriage and while the variation of the pendulum length is stopped by:
  t = t _d + t ′ _e
  and that it is equal to l₀ + vt _e + v′t ′ _e , the command is given by:
  Γ = ϑ˝ _ref (vt _e + v′t ′ _e ) + τ _P
  the equation Ω then becomes:
  ϑ˝l₀ + gϑ = τ _P if ϑ˝ _ref = ϑ˝

The position reached at the end of the trajectory is given by the double integral of Γ. From the result of this calculation, it suffices to choose the constant τ _P so that the position reached at the end of the trajectory is that desired. For example :
if v = V from 0 to t _e with t _e = T _e
and v ′ = -V from t _d to t ′ _e + t _d with t ′ _e = T _e
we show that:

For T _e = T or T _e > T, there is no change compared to the pendulum with constant length:

The reference model corresponds to that shown in Figure 7, which summarizes the five cases listed above, and for which the functions f and h depend on the situation considered as shown in the table below, in which v denotes the speed of rise of the pendulum, that is to say the speed of decrease of the length of the pendulum, and v ′ its speed of descent. interval t function f function h O ≦ t ≦ t _e vt 2v t _e ≦ t ≦ T l _e - l _o 0 T ≦ t ≦ t _d 0 0 t _d ≦ t ≦ t _d + t ′ _e l _e + v ′ (t + t _d ) 2v ′ t _d + t ′ _e ≦ t ≦ t _d + T vt _e + v′t ′ _e 0

The functions f and h are respectively homogeneous at a length and at a speed, which multiplied respectively by an angular acceleration ϑ˝ and an angular speed ϑ ′, then become homogeneous at an acceleration. The following expression, shown diagrammatically in FIG. 7, then constitutes the compensation terms mentioned above, associated with the value of τ _P :
h.ϑ ′ + f.ϑ˝: compensation terms

Thus, according to FIG. 7, the reference model for a linear pendulum of length l₀ (30) determines the values ϑ _ref , ϑ ′ _ref and ϑ˝ _ref . The modules (31) and (32) respectively determine the values of the functions h and f, according to the different phases of the load transfer. The functions f and h thus determined are then multiplied respectively at ϑ˝ _ref and ϑ ′ _ref , the expressions thus obtained are then summed and added to the value of τ _P by means of a summator (33), the said value of τ _P being calculated beforehand as a function of the load destination position.

In this way, the output signal from the adder (33) corresponds to the value of Γ, that is to say to the acceleration signal communicated to the load, by the intermediary diary of the lifting device. The value of Γ is corrected on average 20 times per second, in order to take into account on the one hand any disturbances likely to occur during the transfer, and on the other hand, according to the different transfer phases.

The signal Γ is integrated successively twice, by means of the integrators (34) and (35), in order to give the values of v _ref and of the reference position.

Advantageously, all the integrators and summers are integrated within the same microprocessor ensuring the calculation of the reference model.

FIG. 8 shows an example of the consequence of the servo-control on the angular oscillations as a function of time in the form of a curve. In order to better reflect the different phases of the transfer, the curve representing the position of the load during its transfer as a function of time has also been shown opposite this curve.

There are clearly three successive phases, corresponding respectively to the acceleration at the start of the transfer, generating an oscillation of relatively small amplitude (the ordinate scale is graduated in radians), the said amplitude being rapidly damped by means of l enslavement. We then pass to a phase of almost negligible oscillations, corresponding to the phase at level speed. Then a new phase of greater inverse amplitude, corresponding to the deceleration appears. The load destination point is then reached. In fact, there is an additional phase, following the passage of the load through the final position, during which it is subjected to the control described above, in order to precisely dampen the residual oscillations.

In Figures 9 and 10, there are shown in the form of curves two acceleration signals et supplied to the lifting device and resulting from the processing described above, on the one hand, in the case where the duration of variation of the length of the pendulum defined by the load corresponds to that of the transfer, and on the other hand, in the case where this duration is less than the duration of transfer.

As can be seen, in the first case, during the acceleration and deceleration phases the corresponding signals are generally symmetrical, the phase at constant speed of course generating a zero signal.

On the other hand, in the second case, for which the variation of the length of the pendulum is interrupted during the deceleration phase, we observe the appearance of a plateau in the signal Γ, corresponding to the signal applied in the case of a pendulum of constant length, as shown in Figure 5.

Thus, the device according to the invention makes it possible to obtain systematic and rapid stabilization of the oscillations and other disturbances liable to cause oscillations during the transfer of a suspended load. The double servo-control both in angular deviation and in position is therefore particularly suitable for any transfer of suspended load, in particular in the context of activities requiring special precautions, such as for example the nuclear industry, etc.

Claims (8)

1 / Device for automatic control of the position and oscillations of a suspended load during its transfer by a lifting device comprising:
- a guide means (2,6) on which moves a carriage (1) movable in translation actuated by a motor, to which is hung a cable (4) (or a rod) at the other end of which (from which) the load to be transferred is suspended (5);
a position sensor, intended to permanently locate the position of the carriage (1) relative to a determined point and to provide a so-called position signal corresponding to this location,
- an angular sensor, arranged on the carriage (1), intended to measure the value of the angle ϑ formed by the cable (or the rod) relative to the vertical and to supply a so-called angular signal corresponding to this measurement;
characterized in that it comprises:
- a first control loop comprising:
. a first comparator (22) intended to provide the angular difference between the angular signal thus determined and a predetermined reference angular value by means of a reference model, as a function of the destination of the load (5) and of the speed desired translation;
. a proportional corrector (23) intended to multiply the difference thus measured by a predetermined constant;
- a second so-called position control loop, comprising:
. a second comparator (20) for providing the difference between the position signal and a theoretical reference position predetermined by the same reference model, according to the same data;
. a proportional and integral corrector (21) intended on the one hand, to multiply said deviation by a predetermined constant and on the other hand, to integrate said deviation with respect to time over an equally predetermined interval;
- an adder (24), connected respectively to the output of the proportional corrector (23) proportional and integral corrector (21) and of an amplifier (25) intended to amplify the value of the reference speed, also determined by means of said reference model, according to the same data, said summing device (24) being intended to sum the values coming from these outputs to deliver a signal capable of actuating the motor (26) for controlling the carriage (1). 2 / servo device according to claim 1, characterized in that the reference values determined by the reference model, namely, the angular position, the speed and the position, come from the following law:
Γ = ϑ˝ (l₀ + vt) + 2vϑ ′ + gϑ
by successive integration, where ϑ, ϑ ′ and ϑ˝ denote respectively the angle, the angular velocity and the angular acceleration, g denotes the acceleration of gravity, l₀ the initial length of the cable (or rod), v the speed of variation of the length of the cable (or of the rod), and Γ the acceleration of the carriage. 3 / servo device according to claim 2, characterized in that the length of the pendulum defined by the load is constant, the reference values determined by the reference model being derived from the following law:
Γ = ϑ˝l₀ + gϑ 4 / servo device according to one of claims 1 and 2, characterized in that the carriage (1) is movable in direction on a first straight guide means (2), itself movable in translation on a second means guide guide (6), coplanar but in a direction perpendicular to that of said first guide means, said first guide means being actuated by means of a motor also servo-controlled by means of two servo loops respectively in position and at an angle, according to its own reference model. 5 / servo device according to claim 4, characterized in that the carriage (1) is movable on an overhead crane. 6 / servo device according to claim 1, characterized in that the carriage (1) is movable in translation on a first guide means, itself movable in rotation by means of a rotation member, actuated by means of a motor also controlled by two control loops according to a reference model which is also specific to it. 7 / servo device according to claim 6, characterized in that the carriage (1) is movable along the horizontal boom (10) of a crane. 8 / servo device according to one of the preceding claims, characterized in that the servo calculations are carried out approximately twenty times per second, by means of a digital calculation member.

Images