TRANSFER AND POSITIONING OF GOODS BY MEANS OF CONTAINER CRANES
TECHNICAL FIELD
The present invention relates to a method of transferring goods by means of container cranes and to equipment for carrying out the method. With the aid of these cranes, containers can be transferred within a stock yard or between a ship and the quay, or inversely, by a trolley travelling on the crane which may also comprise a driver's cabin. A system of lifting ropes is associated with the trolley, a gripping device (or spreader) being suspended from the other end of this system. The spreader is provided with gripping appliances which are connected to the containers to be transferred.
BACKGROUND ART - DISCUSSION OF THE PROBLEM
Since the load, which may be a spreader or a spreader with a container, is suspended from ropes, it may start moving in relation to the trolley, both in the direction of the trolley motion and transversely of that direction. The pendulum motion which arises because of acceleration or deceleration of the trolley in the direction of travel of the trolley is normally the greatest of these motions. This motion is, of course, also influenced by the fact that, simultaneously with the trolley movement, the spreader including the container may need to be raised or lowered in relation to the trolley in order to pass crane stands and other obstacles on the quay and on board the ship.
Since the requirements for positioning in case of connection and shutdown of containers on the ship, on the quay, or on the existing vehicle are +/- 3-5 cm, it is also easy to realize that also relatively small pendulating motions may cause problems for a crane operator.
The same problems with pendulum motions and positioning also exist. in connection with so-called grab cranes.
Starting from the problems described above, it is "readily appreciated that very high demands are placed on a crane operator, both as regards accuracy and endurance. To facilitate the work of a crane operator, several different control and drive programs have been worked out over the years. The purpose of these programs have, of course, also been to reduce the effect of pendulum motions and to facilitate the positioning to the greatest possible extent.
Typical of operator-controlled crane drives is that the operator, by means of a lever, gives a reference in the form of a step or as an analogue signal. For controlled drives, the reference is given as a speed reference. To protect mechanical as well as electrical equipment and to impart to the drive a predictable behaviour for the operator, the reference of the operator is limited by a fixed ramp. The inclination of the ramp is set such that the drive, in all normal cases, manages to follow the ramp. In most cases the drives are also provided with so-called torque failure protection, which monitors that the drive manages to follow the ramp reference.
However, a method as described above means that the ramp must be set for a conceived worst case, the capacity of the drive in all other cases thus being poorly utilized. This is especially noticeable on trolley motions where the load is not only load-dependent but also greatly dependent on the combination of load and pendulum angle, that is, the angle beween a vertical line from the point of suspension of the rope in the trolley and the rope. By careless operation, pendulum motions of such a magnitude may arise that a moment more than twice as great as that which a normally dimen¬ sioned crane manages is required. On the other hand, in the case of normal operation the motor is normally utilized less than 50%.
Many existing control and operating programs for container cranes are represented by open control algorithms, that is, they" are not included in closed control systems. This means that the accuracy of the controlled parameters and the dynamic behaviours of these cannot be controlled in the desired way.
In the Swedish patent specification with publication number 429 641, there is, for example, described a method for lateral movement of a suspended load where the length of the rope between the load and the trolley may vary during the lateral movement. The method means that a speed variation is divided into two phases with an intermediate phase with an essentially constant speed, and that the respective acceleration phases are made with different lengths and are given constant values in such a way that the difference in length for the two phases is determined from a nomogram prepared in advance or from a table.
Swedish patent specification with publication number 429 748 describes a method in connection with unloading of goods during lateral movement with the aid of a trolley and where the length of the rope between the trolley and a grab may be changed. The trolley is here given a deceleration down to stationary state and thereafter immediately an acceleration in the opposite direction.
Attemts have also been made with positioning systems supplemented with mechanical systems for suppressing pendulum motions. However, this has resulted in designs which have been both expensive and bulky and in which, in addition, it has been difficult to satisfy the demands for accuracy, particularly in connection with long ropes.
The European patent specification with publication number 0 342 655 A2 describes a container crane drive which briefly operates as follows: The load movement first takes place in the form of a pure lifting movement, then the trolley with
the load is run to a position which is to be straight above the unloading station, whereafter the load is lowered. Pendulum motions during the running of the trolley is avoided by the fact that the spreader is provided with a conically shaped suppression arrangement which connects to a corresponding conical opening in the trolley. Further, there is a description of a measurement system with a radiation transmitter placed on the spreader, with the aid of which the speed of the spreader relative to a target position may be measured. However, the speed cannot be corrected until the desired target position is being approached since information for correction only becomes available when the spreader is located in such a position that the rays of the transmitter can detect the target.
In principle, the positioning problem is a three-dimensional one. Assuming that the quay plane is an x/y plane, the' coordinate axis x, for example, can be directed perpendi¬ cularly to the quay-edge and consequently the coordinate axis y along the quay-edge. The crane positioned at right angles to the the quay-edge can then suitably be placed at the y-coordinate equal to 0. For the three-dimensional positioning determination of the container, the trolley, the spreader, possibly with the container, and the desired load position on board a ship, an x/z plane is added perpen¬ dicular to the x/y plane, suitably extending through y equal to 0. During a loading operation, the coordinates for the trolley and the spreader with the container are changed.
The position of the loading and unloading stations, the trolley and the load in the x/y plane can suitably be defined as the centre of gravity of the surface projected towards the plane. For practical reasons, which will be described below, it may be necessary to have a different definition of the position.
Since a ship is secured to the quay-edge, a definite load position on board will have a fixed and given position in the x/y plane. If the crane is placed so that the y- coordinate of the trolley coincides with the y-coordinate of the load position and at the same time it is ensured, with the aid of a vehicle on the quay, that also the y-coordinate of the container coincides with the y-coordinates of the load and the trolley, the positioning problem will be reduced to an essentially two-dimensional one. This, of course, entails a considerable simplification of the degree of difficulty of the positioning since no pendulum motions transversely of the direction of movement of the ship will be initiated. However, it is self-evident that such pendulum motions may arise as a result of wind and other factors. Because it is difficult to obtain an exact alignment of the y-coordinates, in practice also a certain rotation of the container may arise when it is lifted, which in turn may lead to a lateral pendulum motion. A pendulum motion thus initiated is, however, very small compared with the pendulum motion in the direction of movement of the trolley.
As the ship's hold is successively filled, the x- and y- coordinates of the loading station will be changed, which necessitates a lateral movement of both the crane and the location of the container of the quay. The positioning problems as described above apply, of course, both to the loading and the unloading of cargo. Reducing the positioning to a two-dimensional problem is part of the state of the art.
The physical laws governing a pendulum with a movable suspension point are so well-known that a mathematical model can be produced for the whole crane system or, as it is often named, the whole process. With knowledge of the parameters included, such as the length of the pendulum, the weight of the load, and the position, velocity of movement and acceleration of the suspension point and the load, etc.,
for a certain travel path*, mathematical conditions are provided for determining, with the aid of the model, at all times' the location of a suspended load. Unfortunately, however, such a "measuring method" does not provide sufficient resolution and accuracy to be used immediately as an actual value in a closed position control.
With time, container handling has become increasingly concentrated to very specialized terminals with high demands for efficiency. The possibilities of increasing the efficiency by increased lifting and trolley speeds are small since these are already very high. Studies of the work cycle of a crane show that with present-day technique minor transfer and positioning operations constitute more than 50% of the cycle time. When purchasing new cranes and also when supplementing older cranes, it is therefore nowadays always required that these cranes should be equipped with a posi¬ tioning system as well as an active pendulum suppressing system. Since the state of the art, as described above, provides small possibilities of fulfilling these require¬ ments, there is a great neeed for a new technical approach within this field.
What constitutes the really great problem for container cranes, according to the state of the art, is the absence of an accurate position transducer which at all times may supply a measured value of the position of the load. Since it is simple to measure the position of the trolley, it would be sufficient with a transducer which shows the position of the load in relation to the trolley. If such a transducer were available, it would be possible, with present-day sophisticated closed control systems and simulation technique, to take a large step towards a process which would satisfy the demands for optimum utilization of the motor drives, a good positioning, and suppression of pendulum motions.
SUMMARY OF THE INVENTION, ADVANTAGES
The invention comprises equipment for container handling with the aid of a crane . On the crane there is arranged a travelling trolley which, via lifting ropes, supports a spreader for the containers to be transferred. The equipment comprises, inter alia, a specially produced measuring system for practically continuous position identification, relative to the trolley, of a load during its transfer. Further, the equipment comprises drive devices for both the trolley movement and the lifting movement as well as a regulator which carries out certain calculations and obtains control signals for the drive devices . Associated with the equipment are also transducers for the position of the trolley on the crane and the speed of the trolley as well as a transducer for the actual length of the lifting rope.
Outwardly, the equipment operates as a position control the task of which is to transfer a container from a given first position to a given second position. Devices for feeding to the equipment the respective positions and knowledge of possible travel paths between these are therefore also available .
To be able to stabilize a closed position control in the best way, it is desirable to have internal control loops with feedback from the speed and acceleration of the controlled quantity. In this particular case, of course, these quantities are not directly available. To obtain the best possible values of these, the equipment comprises a mathematical model of the process, as previously described. By supplying the same desired values both to the actual process and to the model and continuously allowing the difference in the result to adjust the model so that the difference is as small as possible, reliable values of the position, speed and acceleration of the controlled quantity can be obtained from the model.
Furthermore, the position system is adapted such that, during the starting cycle, the trolley is accelerated in such a way that the speed of the swinging load in the direction of movement of the trolley, that is, in the x- direction, is the same as the speed of the trolley when the x-coordinates of the load and the trolley coincide. This means that when the load has "caught up with" the trolley, the transfer of the load will take place free from pendulum motion straight below the load and with same speed as the trolley.
The equation system of the model as well as known and measured parameters also allow a possibility to calculate, at all times, the torque that has to be developed by the drive system of the trolley in order to accelerate the load to the same speed as that of the trolley. With knowledge of the moment of inertia involved, the gear change of the drive system, etc., also the acceleration for a maximum utili¬ zation of the available driving torque can be calculated. This can then be transformed into a continuously adjusted ramp reference. This means that the ramp reference can be adapted to the load and the rope angle in question, whereby the available driving torque of the system can be utilized at all times. If it is desired, from a mechanical or other point of view, to have a certain margin to the torque at rated load, the ramp reference can very simply be adapted thereto. As opposed to the state of the art, where, as previously described, a ramp reference is determined based on the most difficult travel path and thus, in case of less torque-demanding travel paths, the available torque for faster acceleration is not utilized, a maximum utilization can be attained resulting in shorter running times.
With knowledge, at all times, of the position and speed of the load and the trolley as well as the coordinates of the unloading station, the position system can determine, with the aid of the model, a deceleration to zero trolley and load speed. During the first part of the deceleration of
the trolley, the load will continue with the same speed, that is, it will carry out a swinging motion in front of the trolley in its direction of movement. With the aid of the model and with knowledge of the available deceleration torque, the time when the deceleration process is to start as well as the deceleration torque and the ramp reference of the deceleration can be determined so that, at zero trolley speed, the load is positioned in a pendulum-free manner straight above the unloading station.
To sum up, it is clear from what has been described above that the pendulum motion of the load is controlled with the aid of the movement of the trolley. This is performed by continuously controlling the speed of movement of the trolley during an acceleration/deceleration process by means of a reference computed and adapted with the aid of the model and a regulator.
In general terms, the task of the measuring system is to determine the position of a load, suspended from a rope, in relation to a movable suspension device for the rope. For this purpose, the measuring system comprises a marker device placed on the load and facing the suspension device and equipped with a number of active markers in the form of light sources and a video camera placed on the suspension device and directed towards the marker device. The shutter of the video camera is opened synchronously with the ignition of the light sources. In this way, the video camera receives a digital picture of the x/y plane with a clear view of the light sources. A video processor processes this picture and delivers x- and y-coordinates for the centre of gravity of the regions illuminated on the picture.
For the object according to the invention, the suspension device consists of the trolley of the container crane. With knowledge of where on the spreader the light sources are situated, also the centre of gravity of the load for the
surface of the load projected on the x/y-plane can be determined. Since the length of the lifting rope is available at all times, a measured value of the three- dimensional position of the load relative to the trolley can at all times be determined in a very simple manner. This measured value together with the position of the trolley is then used as actual value for the closed position control described above.
The described invention differs from the EP publication mentioned above on a plurality of points . These differences may be described as substantial and mean that a container crane according to the claims involves both novelty and inventive step. The differences reside, inter alia, in the fact that according to the invention, upon start-up of a work cycle a first given position is assumed and the target position for a second given position, that is, where the load is to be transferred, is indicated. The transfer is then performed automatically, that is, without the intervention of the crane operator, during simultaneous lifting movement and travelling movement for the trolley in order to miminize the travel time, and when the target position has been attained, the load stops without any pendulum motion arising.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows the configuration of the measuring system.
Figure 2 shows, in broad outline, the relationship between the process, a model for the process and the regulator of the equipment in an x-direction position control system.
Figure 3 shows the fundamental relationship between the internal control loops of the feedback control system, the model, the regulator and the trolley operation.
DESCRIPTION OF THE PREFERRED EMBODIMENT
By way of introduction, the measuring system which is used for the position indication, of the load in relation to the trolley will be described with reference to Figure 1. The measuring system comprises a video camera 1 with digital reproduction, placed on the trolley. It is assumed that a marker device 2 placed on the spreader 3 of the crane for a container 4 is situated in the field of view 5 of the video camera. In a preferred embodiment, the marker device is equipped with four plates packed with light emitting diodes (LEDs) and is provided with a diffusor plate 6 and a screen facing the camera. The screen has two large and two small circular openings, in the following referred to as active markers 7, 8 and 9, 10, respectively, oriented straight above the plates packed with LEDs. The centre of the active markers is aligned and the small active markers are symmetrically placed between the outer, large active markers . The reason for this doubling is that in case of short rope lengths, the outer, large active markers will be positioned outside the field of view of the video camera. Thus, during running either only the two large outer or the two small inner active markers are ignited. The selec¬ tion as to which two active markers should be ignited is made by the control system which, at all times, senses the rope length.
A video processor 11 synchronizes the shutter of the video camera and the ignition of the LEDs such that the LEDs are ignited when the shutter is open. To prevent the video camera from registering random light of various kinds, it is provided with an automatic amplification device so that only the relatively strong light emanating from the active markers is detected, that is, the video camera operates with low amplification. The digital picture of the video camera is transferred to the video processor which processes the picture. The picture is then divided into picture elements (pixels) and the picture elements which are illuminated by
the two active markers are added up into separate areas, the centre of gravity coordinates of which are determined. These values are supplied to a control device or regulator 12 and, together with the position of the trolley, form the basis of the final position determination of the load. An adjustment of the coordinate values of the video camera to a load position value, relative to the trolley, is necessary. The reason for this is that the location of the video camera on the trolley and the location of the marker device on the spreader cannot always be made with the same x/y- coordinates.
The drive systems both for the trolley and the hoisting device consist of conventional motor drives. With the aid of suitable transducers thereon, the x-position XT of the trolley and the speed x-j of the trolley can therefore at all times be available. In similar manner, with knowledge of the drum diameter of the lifting rope and the number of turns the rope has rotated, also the length Li of the rope between the trolley and the load is, at all times, available. The weight P of the load can also be determined by conventional measuring methods.
As described under the "Summary of the Invention", it is desirable from the point of view of the position control to have knowledge of the speed and acceleration of the load for stabilization. To obtain a good measure of these quanti¬ ties, a model of the swinging load during movement of the trolley is used. The model is built up in accordance with the classical laws which apply to such movement and will therefore not be described here. The fundamental rela¬ tionship between the process, that is, the crane drive as such, the model and the regulator of the equipment in an x- direction positioning system is clear from Figure 2.
A load is to be transferred from a given first position to a given second position, that is, an x-position XLR which constitutes the desired (reference) value of the positioning
system. The difference between the calculated x-position of the load, that is, XLB anc-- the reference value, is supplied both to the process, which in the figure has been designated PROCESS, 13, and to the model, designated MODEL, 14. In the imaginary basic embodiment, the process comprises the above- mentioned measuring system for the position of the load in relation to the trolley. By supplementation with the position x<- of the trolley, a measured value XLM of the position of the load in relation to the quay is obtained. The x-position X of the trolley, the speed XT of the trolley, and the length Li of the rope between the trolley and the load are continuously supplied to the model. With the aid of the equation system indicated above, the model is now able continuously to supply a calculated value XLB of the position of the load. The measured value XLM is then compared with the calculated value X^B and the difference is returned to the model for modification thereof so as to minimize the difference. This method means that the calculated value XLB is at all times a valid value of the x- position of the load and can therefore be used as actual value in the x-position control.
With the aid of the model, it is now a simple operation to obtain the first and second derived functions of the load position, that is, the speed X B of the load and the acceleration XLB of the load, which together with X B are supplied to the regulator 12.
The fundamental relationship between the feedback control loops, the model, the regulator and the trolley drive for an x-position control system is clear from Figure 3. As in Figure 2, the reference value X R is compared with the calculated postion XLB of the load. The difference is supplied to the model 14 as well as to a first amplifier 15, which in reality belongs to the regulator 12. The output of the first amplifier, in turn, constitutes the reference value for the acceleration feedback. The difference between this reference value and the actual value X B of the
acceleration, calculated with the model, is supplied to a second amplifier 16 which is also integrated with the regulator. The output of the second amplifier now constitutes the reference value for the speed return. The difference between this reference value and the actual value XLB for the speed of movement of the load, calculated with the model, is supplied to the model which in addition, and according to Figure 2, is also supplied with the position and speed of the trolley and with the current rope length. In addition to the above-mentioned equation system, the model also comprises necessary integrators 17 and 18 to obtain the desired values of the speed of movement and acceleration of the load.
The process control also includes a corresponding system for the z-position control. Since this system is designed in the same way as the x-position control, it will not be described in more detail. The two systems are integrated, both as regards the model and from the point of view of drive strategy, and together form the complete x/z position control. They also cooperate in such a way that the transfer of the load takes place within the scope of possible travel paths.