EP3722197A1 - System for motion compensation between two objects, vehicle with the system, fixed structure with the system and method using the system - Google Patents

Abstract

System for movement compensation between two objects, vehicle with the system, fixed structure with the system and method with the system It is disclosed that a relative movement between two objects is continuously measured. The movement information is used to compensate for the movement when a bridge or another element is docked from one body to the other in order to meet and maintain a target position.

Description

Field of invention

The invention relates to a system for motion compensation between two objects, at least one of the objects, such as a vehicle or watercraft, being movable. The invention also relates to a vehicle with the system and a fixed structure with the system. A procedure is also provided for the system.

Background of the invention

In the offshore industry, for example in the extraction of raw materials or wind energy generation, there are numerous operations on objects. These can occur, for example, between different ships or ships and fixed structures or ships and floating structures. For example, a wind power plant or a fixed drilling platform can be provided as the fixed structure. A floating structure is designed, for example, as a floating drilling platform or as an object attached to ropes under water. Often people and goods are transferred from one object to another during operations. For this it is necessary that the objects are connected to one another, for example by docking a ship on a fixed drilling platform, whereby it is assumed that neither people nor goods are injured or damaged. However, unwanted and difficult or unpredictable movements of the object or objects, for example due to the wave movement or the weather conditions, are problematic. If the weather conditions are not ideal, the risk of people or goods being injured or damaged during such operations increases considerably. In order to reduce or avoid the risk, the surgeries are usually performed interrupted or postponed weather conditions that are not optimal Weather-related shifts or failures lead to considerable financial losses. There is therefore a great deal of interest in systems that allow operations to be carried out even under poor conditions.

Disclosure of the invention

In contrast, the invention is based on the object of creating a system with which an object can be docked to another object in a simple, inexpensive and secure manner in terms of device technology. Furthermore, the invention is based on the object of creating a vehicle that can be docked to an object in a simple, inexpensive and secure manner in terms of device technology. In addition, it is the object of the invention to create a fixed structure to which an object can be docked in a simple, inexpensive and secure manner in terms of device technology. Furthermore, it is the object of the invention to create a method for docking an object to another object in a simple, inexpensive and secure manner in terms of device technology.

The object with regard to the system is achieved according to the features of claim 1, with regard to the vehicle according to the features of claim 10, with regard to the fixed structure according to the features of claim 11 and with regard to the method according to the features of claim 12.

Advantageous further developments of the invention are the subject of the subclaims.

According to the invention, a system is provided, in particular for motion compensation between two objects. The system is used to approach a contact area of the object provided with the system, in particular independently, to a target area of another object and then to contact the areas with one another, the contact in particular to be maintained. Alternatively, the areas can then also be positioned in a position relative to one another, in particular a position that can be determined and fixed in advance, for example also at a distance. At least one of the objects is preferably movable, the object then being a vehicle, in particular a watercraft. The system furthermore preferably has at least one sensor which can be arranged on the object provided with the system. This object can be a component, for example a ship's gangway, or one on a winch attached load, which has the contact area. The at least one sensor is preferably designed in such a way that sensor signals or data about the target area of the other object, in particular the object that does not have this sensor, such as a drilling platform, can be detected. Furthermore, at least one actuator can be provided. This can then drive and move the component and / or the object with the component, in particular in order to change the position of the contact area. This allows movement compensation between the areas. Alternatively or additionally, this can change the relative position of the areas. Furthermore, a control unit is preferably provided or the system is connected to a control unit, for example via the Internet or wirelessly or by cable. The control unit is preferably designed in such a way that it is used to process the information, in particular via a suitable algorithm, in order in particular to determine movement and / or position information of the target area relative to the contact area. The actuator can then be controlled based on the information processed by the control unit in such a way that the areas approach one another and then contact one another or then maintain the predetermined relative position, in particular a distance or relative angle, and / or that the movement compensation takes place. Furthermore, it would be possible for the relative movement not to be fully compensated, but only to be minimized to the extent that the operation can be carried out safely. For example, with existing contact, slight changes in the angle between the gangway and the other ship could be tolerated.

In other words, a relative movement between two objects, in particular continuously, can be measured. The movement information can then be used to compensate for the relative movement during docking, for example a bridge or another element from one body to the other, in order to meet and maintain the target position.

This solution has the advantage that operations, for example between two floating objects or bodies, can also be carried out in less than optimal weather conditions. This is preferably done in that the relative movement is determined, for example, between the contact points, for example between an end side of a gangway and a docking point, between the two bodies and can then be compensated for.

The system can advantageously only be installed on one of the two objects or bodies. No aids or sensors are then required on the other object, for example on the other float. This leads to a high degree of flexibility in use, since an object equipped with the system can, for example, dock onto objects without a corresponding device. Furthermore, safety is increased during operations between two objects, in particular between two floating bodies. A manual compensation of the relative movement in the prior art is practically impossible, so that no compensation currently takes place in operations of this kind and considerable risks always have to be accepted and accidents cannot be ruled out. Furthermore, a quick docking operation can be carried out with the present system. With the system according to the invention, it is not necessary, for example for operations between two floating bodies, to determine a relative movement between their contact points using a sensor system that is provided on both floating bodies. There is therefore no need to install any sensors on both bodies or objects that exchange data. In the present system according to the invention, it is also not necessary to attach aids for the sensor to the object that does not have a sensor, such as reflectors for lasers, radar or markers for cameras. In the system according to the invention, it is therefore advantageous that the relative position between the target area and the contact area can be determined. This not only determines the movement of the body on which the system is installed. If only the movement of this body were to be determined, then only absolute positioning in space would be possible, which would be problematic in particular for operations between two moving objects.

In the present system, a relative movement between two objects, for example between two floating objects or between a floating object and a fixed object, can thus be determined. The relative movement is only determined from one object, which means that no tools are required for the other object. No aids need to be attached to the other object and no information needs to be communicated from the other object to the first object.

In a further embodiment of the invention, information in the form of movement and / or position information about the contact area of the component of the object is recorded and / or established and / or the control unit at certain times or events or reported continuously. This is advantageous in order to easily determine the relative position between the contact area and the target area.

A selection algorithm can preferably be provided. This is designed, for example, in such a way that a target area can be selected via the control unit, in particular automatically, from areas detected by the at least one sensor. The target area can then be followed. If, for example, the object with the contact area approaches the further object, the object with the contact area can independently select a target area on the other object using the selection algorithm and then subsequently acquire the information required for the movement compensation and / or the To enable convergence of the areas.

In a further preferred embodiment of the invention, an input means can be provided. This is preferably designed in such a way that a target area can be selected manually, in particular via an operator, by the at least one sensor detected areas, and can then preferably be followed. As an alternative or in addition to the automatic selection, manual selection of the target area can thus also be provided. Such an input means is advantageous, for example, if an operator of the object with the contact area does not want to rely on a purely automatic target area selection. If both are possible, i.e. automatic or manual target selection, the operator can fall back on manual target area selection if necessary, for example in difficult situations.

In a further embodiment of the invention, a tracking algorithm is preferably provided. This can be configured in such a way that the, in particular the selected, target area is tracked via it on the basis of the information recorded by the at least one sensor. Thus, after the target area has been selected, an operator does not have to track this target area manually, but the tracking advantageously takes place automatically. It is conceivable that the selected target area is marked and can be shown in a display.

A position determination algorithm is advantageously provided, which can be designed in such a way that it is used to determine movement and / or position information of the, in particular selected and / or tracked, target area. The position and movement relative to the contact area can then easily be calculated.

Furthermore, a compensation algorithm is preferably provided, which is designed in such a way that the areas, i.e. the target area and the contact area, are returned to this relative position or come closer to one another after an undesired relative displacement or rotation, starting from a current relative position will. Thus, in particular, the movement compensation can be carried out via the compensation algorithm. In the case of movement compensation, the contact area and target area would preferably be in a fixed relative position to one another, although the object moves, in particular due to waves.

As already mentioned above, it can be provided that the information is ascertained or determined in the form of movement and / or position information of the contact area. A relative movement and / or a relative position can then be determined from the information about the target area and from the information about the contact area. The information of the contact area can be used, for example, in the position determination algorithm and can be recorded simultaneously or after or before the determination of the information for the target area.

In a further embodiment of the invention, one of the objects is preferably a stationary structure and the other object is movable, in particular in the form of a vehicle. It would also be conceivable that both objects can be moved. The vehicle is preferably an aircraft or a water-based vehicle or a land-based vehicle. An offshore wind turbine or a fixed drilling platform, for example, is provided as the fixed structure. A ship, a floating platform or a buoy can be provided as a water-bound vehicle. The system according to the invention is arranged either on the fixed structure or on the movable object.

In a further embodiment of the invention, the component is a load that can be moved or suspended by means of a winch and a crane arm. The contact area is preferably an underside bearing surface of the load. The target area can be a storage area for the load on the other object. Alternatively, it is conceivable that the component is a crane arm or an access bridge or staircase or a gangway of a watercraft. The contact area is then preferably a contact surface thereof, which is to be brought into contact with the other object or to be brought close to the object.

It would also be conceivable that the component is a component of the winch, such as a connecting element, such as a crane hook for the load. If the movement compensation is provided for the winch, then, for example, the lifting movement of the vehicle, in particular the watercraft, can be compensated for in the end position of the rope, where the crane hook is, for example. This means that the crane hook would essentially be in a fixed position in space, although the vehicle is moving due to waves.

In the case of a movement-compensated ship gangway, the end of the gangway could be movement-compensated via the movement compensation of the system. As a result, the end of the ship gangway can be positioned at a fixed position in space despite the ship's movement, which enables it to be guided to the target area, for example to dock on a wind turbine.

The at least one actuator or a plurality of actuators is preferably designed such that the contact area of the component can be moved in one, two or three translational direction (s) and / or in one, two or three rotary direction (s). It is therefore conceivable that some or all of the movements can be compensated for. Depending on the application, a certain number or configuration of the actuator or actuators is then used in order to move the contact area in the desired directions. It has been shown that, for example, in the case of a winch with the load on the rope, good results can be achieved if only translational relative lifting movements in the vertical direction are compensated for via the movement compensation of the system. The rotary and the remaining translational movements can then be disregarded, the corresponding movements between the contact area and the target area being possible and being tolerated. In other words, for example, with the component in the form of Load it to be movable in the direction of gravity via the actuator in the form of the winch in order to carry out movement compensation.

In a further embodiment of the invention, the component of the movable object can be movable relative to this with the at least one or a further actuator. Depending on the information detected via the at least one sensor, the component can be moved relative to its object via this at least one actuator in such a way that the contact and target areas approach one another and / or that the areas maintain a predetermined distance or contact and / or that motion compensation takes place.

In a preferred embodiment of the invention, the sensor or the sensor in combination with the selection algorithm is preferably designed in such a way that - as already mentioned above - no aids are required in the target area in order to detect it. This means that no adjustments to the object with the target area are necessary. For example, no aids or additional sensors need to be used in the target area so that the sensor can detect the target area. For example, no motion reference units (MRU) or permanently installed reference measuring points for position determination are then required, such as reflectors for radar sensors or lasers. It is therefore not necessary to measure reference measuring points and report them to the control unit in a complex manner.

For example, an imaging sensor, in particular a 3D image sensor, can be provided as the sensor. Alternatively or additionally, the sensor or another sensor can be designed as a camera. For example, a monochrome camera or a color camera or a stereo camera or an infrared camera or a time-off-flight (TOF) camera are suitable here. Alternatively or additionally, the sensor or a further sensor can be configured as a light detection and ranging (LIDAR) sensor. Furthermore, it is conceivable, alternatively or additionally, to provide a radar sensor and / or an ultrasonic sensor and / or an inertial sensor sensor, such as a motion reference unit (MRU), as the sensor. A measuring range of the sensor or sensors can preferably be adjustable via one or one adjustment actuator in each case. The adjustment actuator can be controlled, for example, by the control unit or the operator. The sensor is flexible via the Vestellaktor can be aligned to the target area. This can take place automatically, for example, and / or positioning, in particular rough positioning, takes place via the operator. It is also conceivable that, in the event of a change in position between the target area and the contact area, the sensor automatically follows the target area by activating the adjustment actuator accordingly.

The sensor is attached to the movable component, for example. For example, it can be moved together with the component and / or independently of the component via the adjustment actuator. Moving the sensor together with the component has the advantage that when the component moves with its contact area towards the target area, the sensor can, for example, also be brought closer to the target area. When using sensors on the object, it is necessary to calculate the movement / position of the contact point of the component in a complex manner using the kinematic data of the component. This calculation is omitted or simplified the closer the sensor system or the sensor is attached to the contact area.

In a preferred embodiment, the sensor in the form of a LIDAR sensor is attached to the movable component, in particular in the form of a boom. In particular, the sensor is arranged on the end face, that is to say for example at an end section or a tip of the boom, and thus for example close to the contact area or adjacent to the contact area. In addition, the sensor is thereby advantageously in an exposed position, which leads to a free measurement area.

Alternatively or additionally, at least one sensor or a further sensor and / or a sensor pair can be attached to the object with the component and thus preferably not to the movable component. A camera, in particular a pan-tilt-zoom (PTZ) surveillance camera, is provided here as the sensor. If a sensor pair is arranged, a LIDAR sensor connected to it, in particular mechanically, can be used in addition to the camera. The arrangement of the sensor or sensors on the object is preferably provided in such a way that, in this position, the target area is always provided in the measurement area despite movement or wave movement of the object. This is ensured in particular by the PTZ surveillance camera. If the LIDAR sensor is mechanically connected to the camera, it is also moved along with the camera, for example via a Adjustment actuator, which means that only one adjustment actuator is necessary. This also extends its measuring range according to the requirements. The sensor or the sensor pair are arranged, for example, on a movable platform via which the component can be moved about a vertical axis of the vehicle or the structure. The sensor or the pair of sensors can then, for example, be moved together with the platform via an actuator or adjustment actuator.

In addition to the sensor or the sensor pair, a further sensor is preferably provided on the object, which is provided in particular for automatic detection or for the automatic selection of the target area. The further sensor is attached, for example, to the platform or to the vehicle or the structure, ie preferably not to the component.

The sensor or some of the sensors or all sensors are preferably intrinsically and / or extrinsically calibrated. In intrinsic calibration, for example, a sensor is calibrated in the form of a camera for the associated optics. Furthermore, if several sensors are provided, these or some of them are preferably calibrated against one another. The calibrations can enable a coordinate transformation between sensor coordinate systems so that the control unit can easily determine the desired information. When several imaging sensors are used, in particular at the same time, the individual images (cameras) / point clouds (lidar) are merged and / or merged in the algorithm at certain points. Reasons for this can include mutual plausibility checks, creating redundancy for sensor failure or temporary occlusion or increasing the stability of the algorithms. For a meaningful merging, the relative position (3 translations, 3 rotations) between the coordinate systems of the imaging sensors should be known. This is guaranteed by the extrinsic calibration.

In a preferred solution, an input means or a further input means can be provided for the at least one actuator and / or for the at least one adjustment actuator. The input means is, for example, a joystick. The input means can be configured in such a way that the component having the contact area and / or the at least one sensor with respect to the target area, in particular in advance, is positionable. Thus, for example, rough positioning by an operator is possible in a simple manner before automatic movement compensation and / or approach of the areas takes place.

In a further embodiment of the invention it is conceivable to provide a display or a display. Possible target areas detected by the at least one sensor can then be displayed on this, for example. In particular, an image captured by the sensor by a camera can be displayed. The operator can use the display to select a target area via an input means. Furthermore, it is conceivable that a position or 3D position of the, in particular selected, target area is shown on the display based on the information recorded by the at least one sensor.

According to the invention, a vehicle with the system according to one or more of the preceding aspects is provided. The vehicle can be used as an object with the sensor.

According to the invention, a stationary structure is provided to which a vehicle, in particular a watercraft, can be docked. The structure can then have the system according to one or more of the preceding aspects. The structure can then be used as an object with the sensor. It is conceivable that the structure has the movable component.

• Erfassen von Sensorsignalen des zumindest einen Sensors,
• Auswahl des Zielbereichs anhand der erfassten Sensorsignale,
• Ermittlung der Informationen in Form von relativen Bewegungs- und/oder Positionsinformationen zwischen dem ausgewählten Zielbereich und dem Kontaktbereich des Bauteils oder des Objekts, insbesondere über den Positionsbestimmungs-Algorithmus,
• Ansteuerung des zumindest einen Aktors basierend auf den ermittelten Informationen derart, dass sich der Kontaktbereich und der Zielbereich einander annähern und/oder dass die Relativ-Bewegungskompensation oder die Bewegungskompensation, insbesondere über den Kompensations-Algorithmus, erfolgt.

According to the invention, a method with a system according to one or more of the preceding aspects is provided, which has the following steps:

• Acquisition of sensor signals from the at least one sensor,
• Selection of the target area based on the recorded sensor signals,
• Determination of the information in the form of relative movement and / or position information between the selected target area and the contact area of the component or the object, in particular via the position determination algorithm,
• Control of the at least one actuator based on the determined information in such a way that the contact area and the target area approach one another and / or that the relative movement compensation or the movement compensation takes place, in particular via the compensation algorithm.

This solution has the advantage that a movement compensation between the areas and / or an approximation of the areas is made possible in a simple manner. The compensation can take place, for example, by a corresponding control of the crane arm, the gangway or the winch, with the help of which contact with the other object is established. Operations between, for example, two floating bodies or one floating body and one solid body are thus made possible in a simple manner.

• Erfassen von Sensorsignalen des zumindest einen Sensors.
• Verfolgen oder Tracken oder Wiederfinden des gewählten Zielbereichs in den Sensorsignalen, insbesondere anhand des Verfolgungs-Algorithmus, insbesondere über die Steuereinheit. Somit muss vorteilhafter Weise nicht erneut manuell der Zielbereich ausgewählt werden.
• Ermittlung der Informationen in Form von Bewegungs- und/oder Positionsinformationen des ausgewählten Zielbereichs anhand der Sensorsignale des zumindest einen Sensors, insbesondere über den Positionsbestimmungs-Algorithmus.
• Ansteuerung des zumindest einen Aktors basierend auf den ermittelten Informationen derart, dass sich der Kontaktbereich und der Zielbereich einander annähern und/oder dass die Bewegungskompensation erfolgt. Diese Schritte können so oft wiederholt werden, bis die Bereiche kontaktiert sind oder in einem vorbestimmten Abstand zueinander angeordnet sind.

In a further refinement of the method, for example, after the actuator has been activated to initiate a relative movement between the areas, the following steps can follow:

• Detecting sensor signals from the at least one sensor.
• Tracking or tracking or retrieval of the selected target area in the sensor signals, in particular using the tracking algorithm, in particular via the control unit. The target area therefore advantageously does not have to be selected again manually.
• Determination of the information in the form of movement and / or position information of the selected target area based on the sensor signals of the at least one sensor, in particular via the position determination algorithm.
• Activation of the at least one actuator based on the ascertained information such that the contact area and the target area approach one another and / or that the movement compensation takes place. These steps can be repeated until the areas are contacted or are arranged at a predetermined distance from one another.

In other words, when the selected target area is tracked or found again, the target area is tracked with a suitable algorithm, in particular with the tracking algorithm, and the change in position is continuously tracked. For example, this can be done using one or more of the following methods: Kalman filters, particulate filters, feature-based methods, region-based methods, contour-based methods, model-based methods.

Information or position information is preferably continuously fed to the control unit for regulating the at least one actuator, in particular for the movable component, in order to implement the movement compensation, in particular with a manual assistance function or automated contact establishment.

In a further refinement of the method, the sensor signals from a plurality of sensors can be used in the determination of the information in the form of the movement and / or position information of the selected target area, in particular via the position determination algorithm. This is preferably the case in particular when the distance between the target area and the contact area exceeds a certain threshold value and / or the sensor signals of a sensor are too imprecise. For example, the sensor or the sensor pair is first used on the object and not on the component, for example the camera or the camera and the LIDAR sensor. Under certain conditions, the sensor on the component, which is preferably the LIDAR sensor, can then also be used if necessary.
In a further embodiment of the invention, if the ascertained information of the selected target area is too imprecise, it can be provided that the operator is informed of this, in particular in the display. A new target area can then be selected manually or automatically. It is conceivable that the operator initiates the automatic selection.

The distance between the contact area and the target area is preferably monitored, for example continuously or for example after a respective activation of the at least one actuator. The method can then be ended from a predetermined distance or when the areas come into contact. This means, for example, if the relative distance of the areas has fallen below a previously defined threshold value, then the contact can be regarded as established. The control can then, for example, be transferred to a general movement compensation system for the vehicle.

Preferably, the contact area and the target area can optionally be approached manually via the input means by an operator or automatically via the control unit be approximated. It is conceivable, for example, that in uncritical and simple situations the operator manually approaches the areas or intervenes in the approach if necessary.

In a further embodiment of the invention, the target area on the other object can be selected either manually via the operator or automatically via the control unit. With manual selection, the operator can first ensure that the target area is provided in the field of vision of the sensors. To this end, it can control at least one actuator and / or the at least one adjustment actuator in order to align the at least one sensor accordingly. In the manual selection, the sensor signals detected by the sensor can preferably be shown to the operator, in particular as images. This can take place, for example, in the form of a camera image and / or in the form of a LIDAR point cloud and / or in the form of a radar image. On the basis of the information presented, the operator can select the target area, for example via the input means. As already explained above, the input means is a joystick or a touch-sensitive display (touchscreen) would also be conceivable. As explained above, after the target area has been selected, based on the sensor signals, the information relating to the selected target area can then be ascertained, in particular via the position-determining algorithm. In other words, a 3D position of the target area can be determined using suitable sensor data and preferably displayed to the operator, in particular via the display. For example, a so-called sensor fusion between stereo camera data and the LIDAR point cloud can take place.

The selection algorithm can be used for the automatic selection of the target area. This is preferably configured in such a way that the target area can be determined from the sensor signals or sensor data. A surround view camera system or sensors with a large opening angle are preferably suitable as the sensor. One or more of the following methods can be used as the selection algorithm: Iterative Closest Point (ICP), RANSAC, template matching, segmentation methods such as Otsus, K-means clustering, watershed, grab-cut algorithm. Following the fully automatic selection of the target area, as already explained above, the information can then be determined in the form of the movement and / or position information of the selected target area, in particular via the position determination algorithm relative to the contact area. In other words, the 3D position of the target area can be determined using suitable sensor data and displayed to the operator.

The data acquired by the sensor or by the sensors can preferably be represented visually or graphically on the display. The selected target area can be marked here, for example. Furthermore, as already mentioned above, the current position of the 3D position of the target area can be shown in the display.

• Suche des Zielbereichs anhand eines Template-Bildes des Zielbereichs im aktuellen Kamerabild. Es kann vorgesehen sein, dass beim Positionsbestimmungs-Algorithmus, beispielsweise wenn die Entfernung zwischen dem Kontaktbereich und dem Zielbereich einen bestimmten Wert überschreitet, die Sensordaten des Sensors am Bauteil, beispielsweise einem LIDAR-Sensor, zusätzlich zur Positionsbestimmung des Zielbereichs eingesetzt werden können. Ist der Zielbereich in den Sensorsignalen, insbesondere im Kamerabild, nicht oder nur mit einer Genauigkeit ermittelbar, die unterhalb einer vorbestimmten Genauigkeitsgrenze liegt, beispielsweise nur einen geringen Kontrast aufweist, kann beim Positionsbestimmungs-Algorithmus vorgesehen sein, dass dies dem Bediener mitgeteilt wird und dieser einen neuen Zielbereich auswählt. Alternativ wäre denkbar, dass automatisch ein neuer Zielbereich ausgewählt wird, insbesondere, ohne dass dies dem Bediener mitgeteilt wird.

To determine the position of the target area, in particular using the position determination algorithm, the information from the sensor on the object, that is, not from the sensor attached to the component, is preferably used. The sensor is preferably designed in the form of the camera. A template camera image can be used as information. It can thus be provided that the sensor signals of the sensor in the form of the camera are used in the position determination algorithm. With the position determination algorithm of the camera, a template camera image can then be generated for the target area, in particular at the time of the selection of the target area. Preferably, at least the following step can be provided in the position determination algorithm:

• Search for the target area using a template image of the target area in the current camera image. It can be provided that in the position determination algorithm, for example if the distance between the contact area and the target area exceeds a certain value, the sensor data of the sensor on the component, for example a LIDAR sensor, can also be used to determine the position of the target area. If the target area in the sensor signals, in particular in the camera image, cannot be determined or can only be determined with an accuracy that is below a predetermined accuracy limit, for example only has a low contrast, the position determination algorithm can provide that this is communicated to the operator and that one selects new target area. Alternatively, it would be conceivable that a new target area is selected automatically, in particular without the operator being informed of this.

In a further embodiment of the invention it can be provided that the tracking algorithm uses the sensor data of two sensors, with which the target area can advantageously be tracked reliably. Suitable sensors are, for example, the sensor (camera) or the sensor pair on the object and not on the component and the sensor (LIDAR sensor) on the Component. With the tracking algorithm, the position of the target area in the current sensor data can preferably be tracked using a filter, in particular using a Kalman filter. The future position can preferably be calculated or estimated via the filter based on the current position of the target area, or the current position can be calculated or estimated via the filter based on the last known position.

It is disclosed that a relative movement between two objects is constantly measured. The movement information is used to compensate for the movement when a bridge or another element is docked from one body to the other in order to meet and maintain a target position.

Brief description of the drawings

• Fig. 1a bis 1d jeweils zwei Objekte, die jeweils als Fahrzeug oder Struktur ausgebildet sind, gemäß einem jeweiligen Ausführungsbeispiel, wobei jeweils eines der Objekte mit einem erfindungsgemäßen System ausgestattet ist,
• Fig. 2 in einer schematischen Darstellung ein Wasserfahrzeug mit dem System gemäß einem weiteren Ausführungsbeispiel und ein Wasserfahrzeug mit einem Zielbereich,
• Fig. 3 in einer schematischen Darstellung ein Wasserfahrzeug mit dem System gemäß einem weiteren Ausführungsbeispiel zusammen mit einem Wasserfahrzeug mit einem Zielbereich,
• Fig. 4 in einem Ablaufdiagramm ein Verfahren mit dem System gemäß einem ersten Ausführungsbeispiel und
• Fig. 5 in einem Ablaufdiagramm ein Verfahren mit dem System gemäß einem zweiten Ausführungsbeispiel.

Preferred embodiments of the invention are explained in more detail below with reference to schematic drawings. Show it:

• Figures 1a to 1d two objects, each designed as a vehicle or structure, according to a respective embodiment, one of the objects being equipped with a system according to the invention,
• Fig. 2 in a schematic representation a watercraft with the system according to a further embodiment and a watercraft with a target area,
• Fig. 3 in a schematic representation a watercraft with the system according to a further embodiment together with a watercraft with a target area,
• Fig. 4 in a flow chart a method with the system according to a first embodiment and
• Fig. 5 in a flow chart a method with the system according to a second embodiment.

Fig. 1a shows an object in the form of a floating body, which is a watercraft 1. Furthermore, a stationary structure 2 is shown as the object, which is, for example, an offshore wind turbine or a fixed drilling platform. The watercraft 1 has a component in the form of a gangway 4. This should be included dock to a docking area 6 of the structure 2. The gangway 4 can be pivoted via an actuator, not shown, so that a free-standing front end section 8 can be adjusted in height. Furthermore, the gangway 4 can be rotated about a vertical axis of the watercraft 1 via an actuator. A system according to the invention for movement compensation and for bringing the end section 8 closer to the docking area 6 is provided in the watercraft 1.

According to Figure 1b an object in the form of a watercraft 10 is also shown. Furthermore, an object in the form of a stationary structure 12 is shown. In contrast to the Fig. 1a does not have the watercraft 10, but the structure 12 has the movable gangway 4. The watercraft 10 has the docking area 6. The system according to the invention for movement compensation and for bringing the gangway 4 closer to the docking area 6 is provided in the structure 12.

According to Figure 1c a first object in the form of a watercraft 14 and a second object in the form of a watercraft 16 are shown. The watercraft 14 has a crane arm 18 which is rotatable so that its free end can be adjusted in height, and which is preferably movable about a vertical axis of the watercraft 14. The crane arm 14 interacts with a (not shown) winch for a rope 20, at the end of which a load 22 is fastened via a crane hook. The load 22 is intended to be unloaded onto the watercraft 16. For this purpose, the watercraft 14 has the system for motion compensation and for bringing the load 22 closer to the watercraft 16. At least one sensor (not shown) can be provided, for example, on the free end section of the crane arm 18, i.e. on the end section remote from the watercraft 14. This can capture a target area on the watercraft 16. A contact area is provided for the load 22, for example.

Fig. 1d shows, in addition to the watercraft 10 with the docking area 6, the watercraft 1 with the gangway 4. The watercraft 1 here has the system for combining movements and for bringing the gangway 4 closer to the docking area 6.

According to the Figures 1a to 1d it can be seen that the system can be used for motion compensation in operations between a watercraft and a stationary structure or in operations between two watercraft. It is advantageous in each case only necessary that each one of the objects from a respective Figures 1a to 1b the system has, while the corresponding other object does not have to be adjusted for the system. The system is only installed on one of the objects and then enables it to establish mechanical contact with a large number of other floating or stationary bodies. To avoid damage during or during the establishment of contact and to be able to maintain contact over a longer period of time without damage. With the system, for example, movements of the watercraft 1, 10, 14, 16 caused by wind and waves can be taken into account and compensated for by the system when the contact area is established with the target area. This is done, for example, by adjusting the position of the gangway 4 or the crane arm 18 or the length of the rope 20.

With the system from the respective Figures 1a to 1d a relative movement between the two objects, in particular between the contact area and the target area, is continuously measured. The movement information is then used to compensate for the movement when the gangway 4 or bridge or another element is docked from one object to the other in order to meet and maintain a target position.

A fully automatic and / or a semi-automatic method can be provided for the system. In Fig. 2 an arrangement for the semi-automatic process is shown and in Fig. 3 an arrangement for the fully automatic process is shown.

In Fig. 2 the watercraft 1 is shown with the gangway 4 or bridge. Furthermore, an object in the form of a watercraft 24 is shown which has a docking area 6. A sensor 26 in the form of a LIDAR sensor is arranged and fastened on the face of the gangway 4, that is to say at its free end. Furthermore, a pair of sensors 28 is provided, which is not attached to the gangway 4, but directly to the watercraft 1. For example, the pair of sensors 28 is arranged on a platform which is rotatable about a vertical axis of the watercraft 1 and on which the gangway 4 is movably mounted. The sensor pair 28 has a camera, in particular in the form of a pan-tilt-zoom (PTZ) camera, and a LIDAR sensor. The gangway has a contact area 30 at the end, which is to be brought into contact with a target area 32 for docking with the watercraft 24. The sensor pair 28 is preferably arranged on the watercraft 1 in such a way that The target area 32 remains detectable until a certain wave movement or with any wave movement. This is ensured by the PTZ camera. The wave movement or other movements would also be conceivable to at least partially compensate for the sensor pair 28 with movement compensation. The sensor pair 28 is mechanically firmly connected and can be moved together, for example via an adjustment actuator. The LIDAR sensor, for example, is connected to the camera through the mechanical connection and its field of view is thereby expanded according to the requirements.

wobei T eine Transformationsmatrix ist. Bei der üblichen Verwendung homogener Koordinaten ist die Transformationsmatrix T eine 4x4 Matrix mit einer Subrotationsmatrix R, die die Rotation zwischen den Koordinatensystemen beschreibt, und der Subtranslationsmatrix t, die die Translation zwischen den Koordinatensystemen beschreibt.
Mit Hilfe der Transformationssmatrix $${}_{K_{B1}{K}_{B2}}T$$

kann beispielsweise die Darstellung eines Vektors bezüglich des Koordinatensystems K_B1 in eine Darstellung bezüglich des Koordinatenystems K_B2 überführt werden: $$x_{K_{B1}}={}_{K_{B1}{K}_{B2}}T{x}_{K_{B2}}$$

$$x_{K_{\mathit{Bi}}}=\left(\begin{array}{c}x\\ y\\ z\\ 1\end{array}\right)$$

$${}_{K_{B1}{K}_{B2}}T=\left[\begin{array}{cc}R& t\\ 0^T& 1\end{array}\right]$$

$$R=\left[\begin{array}{ccc}{r}_{11}& r_{12}& r_{13}\\ r_{21}& r_{22}& r_{23}\\ r_{31}& r_{32}& r_{33}\end{array}\right]$$

$$t=\left[\begin{array}{c}{t}_x\\ t_y\\ t_z\end{array}\right]$$

$$0^T=\left[\begin{array}{ccc}0& 0& 0\end{array}\right]$$

The sensors 26, 28 are intrinsically, extrinsically and calibrated against one another. This enables a coordinate transformation between the sensor coordinate systems, where K _{B1 is} the coordinate system of the first sensor of the sensor pair 28, K _{B2 is} the coordinate system of the second sensor of the sensor pair 28 and K _{A is} the coordinate system of the sensor 26 is. The coordinate transformation can be done with the following common formulas: $${}_{K_{{\mathrm{B.}}_1}{K}_{\mathrm{A.}}}T,{}_{K_{{\mathrm{B.}}_2}{K}_{\mathrm{A.}}}T\mathrm{and}{}_{K_{{\mathrm{B.}}_1}{K}_{{\mathrm{B.}}_2}}T$$

where T is a transformation matrix. With the usual use of homogeneous coordinates, the transformation matrix T is a 4x4 matrix with a sub-rotation matrix R, which describes the rotation between the coordinate systems, and the sub-translation matrix t, which describes the translation between the coordinate systems.
With the help of the transformation matrix $${}_{K_{\mathrm{B.}1}{K}_{\mathrm{B.}2}}T$$

For example, the representation of a vector with respect to the coordinate system K _{B1 can be converted} into a representation with respect to the coordinate system K _B2 : $$x_{K_{\mathrm{B.}1}}={}_{K_{\mathrm{B.}1}{K}_{\mathrm{B.}2}}T{x}_{K_{\mathrm{B.}2}}$$

$$x_{K_{\mathit{Bi}}}=\left(\begin{array}{c}x\\ y\\ z\\ 1\end{array}\right)$$

$${}_{K_{\mathrm{B.}1}{K}_{\mathrm{B.}2}}T=\left[\begin{array}{cc}\mathrm{R.}& t\\ 0^{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="imgf0003.png,imgf0004.png"\; state="\{\{state\}\}">T</figure-callout>}}& 1\end{array}\right]$$

$$\mathrm{R.}=\left[\begin{array}{ccc}{r}_{11}& r_{12}& r_{13}\\ r_{\mathrm{21st}}& r_{\mathrm{22nd}}& r_{23}\\ r_{31}& r_{32}& r_{33}\end{array}\right]$$

$$t=\left[\begin{array}{c}{t}_x\\ t_y\\ t_z\end{array}\right]$$

$$0^T=\left[\begin{array}{ccc}0& 0& 0\end{array}\right]$$

With actuators 34 in Fig. 2 are shown schematically with a block, the gangway 4 can be rotated so that the height of the contact area 30 can be changed and rotated about a vertical axis 36 of the watercraft 1. The actuators 34 can be controlled via a control unit 37.

Based on Fig. 4 the procedure for the system is made Fig. 2 explained in order to bring the areas 30 and 32 closer together up to the point of contact and to carry out a movement compensation of a relative movement of these areas 30, 32. In step 36, an operator ensures that the target area 32 is off Fig. 2 is in the field of view of the sensors 26 and 28. This can take place in that the sensor pair 26 is moved together with the gangway 4 via the actuators 34, for example via the platform, and / or the watercraft 1 is moved by its drive and / or that for the sensor pair 28 and / or for the Sensor 26, an adjustment actuator or an adjustment actuator is provided, via which it is / are movable. A manual input of the change in position for the sensors 26, 28 takes place, for example, with the aid of a joystick and / or a keyboard console and / or a touch screen.

In step 38, the sensor signals or sensor data from sensors 26, 28 are recorded. The sensor data of the PTZ camera of the sensor pair 28 are used as a camera image I. ^B determined. The Sensor data of the LIDAR sensor of the sensor pair 28 are signed with a LIDAR data signature D. ^B recorded. In addition, the sensor data of the LIDAR sensor (sensor 26) is provided with a LIDAR data signature D. ^A recorded.

In step 40, the visualization takes place on the screen. The sensor signals of the scene are shown to the operator in an image. For this purpose, the image of the PTZ camera of the sensor pair 28 is shown superimposed on the LIDAR point clouds of the LIDAR sensors (sensor pair 28 and 26) on a display or monitor.

der PTZ-Kamera (Sensorpaar 28) und LIDAR-Daten-Signaturen ${\overrightarrow{D}}_0^B$

und ${\overrightarrow{D}}_0^A$

der LIDAR-Sensoren (Sensorpaar 28 und Sensor 26) erzeugt, die den Zielbereich charakterisieren. Der ausgewählte Zielbereich wird dann im Monitor mit einer Markierung, beispielsweise einer Bounding-Box, markiert.In step 42 the operator then selects based on the sensor signals displayed on the monitor I. ^B , D. ^B and D. ^{A from} a target area, the monitor being, for example, a touchscreen. After the target area has been selected, in this step 42 a template camera image is then used for the target area ${\overrightarrow{\mathrm{I.}}}_0^{\mathrm{B.}}$

the PTZ camera (sensor pair 28) and LIDAR data signatures ${\overrightarrow{\mathrm{D.}}}_0^{\mathrm{B.}}$

and ${\overrightarrow{\mathrm{D.}}}_0^{\mathrm{A.}}$

of the LIDAR sensors (sensor pair 28 and sensor 26) which characterize the target area. The selected target area is then marked in the monitor with a marking, for example a bounding box.

In the next step 44, the position determination algorithm is used to determine the information in the form of movement and position information of the target area 32 relative to the contact area 30. In other words, a 3D position of the target area is determined using the sensor signals or sensor data from the sensors 26 and 28 determined and displayed to the operator. First of all, the 3D position is determined using the PTZ camera of the sensor pair 28. So-called template matching takes place, which shows a lateral position of the target area in the captured image I. ^{B is} detected using the following formula: $$\overrightarrow{\mathrm{R.}}=\frac{{\displaystyle \sum _{\mathit{uʹvʹ}}}\left({\mathrm{I.}}_0^{\mathrm{B.}}\left(\mathit{uʹ},\mathit{vʹ}\right)\cdot {\mathrm{I.}}^{\mathrm{B.}}\left(u+\mathit{uʹ},v+\mathit{vʹ}\right)\right)}{\sqrt{{\displaystyle \sum _{\mathit{uʹvʹ}}}{\mathrm{I.}}_0^{\mathrm{<figure-callout\; id="2"\; label="B.\; uʹ\; vʹ"\; filenames="imgf0001.png"\; state="\{\{state\}\}">B.</figure-callout>}}{\left(\mathit{uʹ},\mathit{vʹ}\right)}^{}{\left(\mathit{}\right)}^2\cdot {\displaystyle \sum _{\mathit{uʹvʹ}}}{\mathrm{I.}}^{\mathrm{B.}}{\left(u+\mathit{uʹ},\mathrm{<figure-callout\; id="2"\; label="v\; +\; vʹ"\; filenames="imgf0001.png"\; state="\{\{state\}\}">v</figure-callout>}+\mathit{vʹ}\mathit{}\right)}^2}}$$

u₀ und v₀ sind also die Werte für u und v in obiger Formel für R, für die die Korrelation R maximal wird.The template image is usually smaller than the current camera image. With template matching, a moving window is placed over the camera image at every possible position and the correlation is calculated. u ' and ν' are the two pixel coordinates of the captured images. u and ν denote the a priori unknown (lateral) shifts in the two pixel coordinates between the image section and the template image. The aim is now to find values for u and v to be determined so that the closest possible correspondence between the image section and the template image results. The degree of agreement is expressed mathematically by the correlation, which is to be maximized by a suitable choice of u and v. The values of u and v that maximize the correlation are called u ₀ and v ₀ . They indicate where the template image or the target area is located in the camera image or, in other words, they describe the position r _{0 of} the target area in the camera image. The position r ₀ results as follows: $${\overrightarrow{r}}_0=\left(\begin{array}{c}{u}_0\\ v_0\end{array}\right)=\mathrm{bad}\left(\mathrm{Max}\left(\overrightarrow{\mathrm{R.}}\right)\right)$$

u ₀ and v ₀ are therefore the values for u and v in the above formula for R, for which the correlation R is maximal.

: $$\mathbb{K}=\left\{\left(\begin{array}{c}{u}_1^{B_1}\\ v_1^{B_1}\end{array}\right),\left(\begin{array}{c}{u}_2^{B_2}\\ v_2^{B_1}\end{array}\right),\dots \right\}$$

in 3D ausgedrückt werden: $${\left(\begin{array}{c}{u}^{B_1}\\ v^{B_1}\end{array}\right)}_{\in \mathbb{K}}\stackrel{{}_{K_{B_1}{K}_{B_2}}T}{\to }{\left(\begin{array}{c}{u}^{B_2}\\ v^{B_2}\end{array}\right)}_{\in \mathbb{K}}\stackrel{\mathrm{extr}\mathrm{.}\mathrm{Kalibration}}{\to }{\left(\begin{array}{c}x\\ y\\ z\end{array}\right)}_{\in \mathbb{K}}$$

Anschließend kann die 3D-Lage des Zielbereichs mit Hilfe des Random Sample Consensus (RANSAC) oder durch ein Iterative-Closest-Point (ICP) Algorithmus bestimmt werden. Falls eine relative Entfernung zwischen dem Kontaktbereich 30 und dem Zielbereich 32 über einem bestimmten Schwellenwert liegt, werden vorzugsweise die Daten D ^A des Sensors 26 zur genaueren Lagebestimmung zusätzlich verwendet. Falls der ausgewählte Zielbereich in den erfassten Sensorsignalen nur einen geringen Kontrast aufweist, so dass die Genauigkeit der Positionsbestimmung nur ungenau erfolgen kann, dann wird dem Bediener die unzuverlässige Messung signalisiert, und er kann eine neue Stelle markieren, was durch den Pfeil 46 in Fig. 4 gezeigt ist, womit im Anschluss wieder der Schritt 44 ausgeführt wird.The set of pixels assigned to the target area can then be obtained from a coordinate transformation

: $$\mathbb{K}=\left\{\left(\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="u"\; filenames="imgf0003.png,imgf0004.png"\; state="\{\{state\}\}">u</figure-callout>}}_1^{{\mathrm{B.}}_1}\\ {\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="imgf0003.png,imgf0004.png"\; state="\{\{state\}\}">v</figure-callout>}}_1^{{\mathrm{B.}}_1}\end{array}\right),\left(\begin{array}{c}{\mathrm{<figure-callout\; id="2"\; label="u"\; filenames="imgf0001.png"\; state="\{\{state\}\}">u</figure-callout>}}_2^{{\mathrm{B.}}_2}\\ {\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="imgf0001.png"\; state="\{\{state\}\}">v</figure-callout>}}_2^{{\mathrm{B.}}_1}\end{array}\right),...\right\}$$

expressed in 3D: $${\left(\begin{array}{c}{u}^{{\mathrm{B.}}_1}\\ v^{{\mathrm{B.}}_1}\end{array}\right)}_{\in \mathbb{K}}\stackrel{{}_{K_{{\mathrm{B.}}_1}{K}_{{\mathrm{B.}}_2}}T}{\to }{\left(\begin{array}{c}{u}^{{\mathrm{B.}}_2}\\ v^{{\mathrm{B.}}_2}\end{array}\right)}_{\in \mathbb{K}}\stackrel{\mathrm{extr}\mathrm{.}\mathrm{Calibration}}{\to }{\left(\begin{array}{c}x\\ y\\ z\end{array}\right)}_{\in \mathbb{K}}$$

The 3D position of the target area can then be determined using the Random Sample Consensus (RANSAC) or an Iterative Closest Point (ICP) algorithm. If a relative distance between the contact area 30 and the target area 32 is above a certain threshold, the data D. ^{A of} the sensor 26 is also used for more precise position determination. If the selected target area has only a low contrast in the detected sensor signals, so that the accuracy of the position determination can only be imprecise, then the unreliable measurement is signaled to the operator and he can mark a new point, which is indicated by the arrow 46 in Fig. 4 is shown, whereupon step 44 is carried out again.

After step 44, that is, when the movement and position information of the selected target area has been determined with sufficient accuracy via the position determination algorithm, step 48 follows. Here, the movement and / or position information is continuously fed to the control unit 37, which accordingly the actuators 34 controls. These are controlled in such a way that the contact area 30 in The direction of the target area 32 is moved. In this case, a previously defined distance is covered and then a renewed recording of the sensor signals from sensors 26 and 28 is triggered, which is shown by step 50. In step 50 it is thus provided that the sensors 26 and 28 receive the sensor signals I. ^B , D. ^B and D. ^A record.

Hierbei ist ein Messdaten-Vektor ${\overrightarrow{r}}_t^T$

aus N Messungen bis zum Zeitpunkt t gebildet und Θ _t sind Koeffizienten. Zunächst wird dann der sogenannte Kalman-Verstärkungsvektor berechnet: $${\overrightarrow{K}}_t=\frac{P_{t-1}\cdot {\overrightarrow{r}}_{t-1}}{\lambda +{\overrightarrow{r}}_{t-1}^T{P}_{t-1}{\overrightarrow{r}}_{t-1}}$$

Hierbei ist λ(0 ≤ λ < 1) der sogenannte Forgetting Factor. P_t ist die inverse Korrelationsmatrix der Messdaten. Ein A-priori-fehler ∈ zwischen dem aktuellen Messwert r_t und dem geschätzten Wert wird anschließend mit folgender Formel berechnet: $${ϵ}_t=r_t-{\overrightarrow{r}}_{t-1}^T{\overrightarrow{\theta }}_{t-1}$$

Die Koeffizienten Θ _t werden anhand der folgenden Gleichung ermittelt: $${\overrightarrow{\theta }}_t={\overrightarrow{\theta }}_{t-1}+{\overrightarrow{K}}_t{ϵ}_t$$

Des Weiteren wird eine inverse Korrelationsmatrix P_t anhand folgender Formel ermittelt: $$P_t=\frac{P_{t-1}-{\overrightarrow{K}}_t{\overrightarrow{r}}_{t-1}^T{P}_{t-1}}{\lambda }$$

Anschließend erfolgt die Schätzung der Positionsinformation des Zielbereichs zum Zeitpunkt t + 1 anhand der oben genannten Gleichung: $$r_{t+1}={\overrightarrow{r}}_t^T\cdot {\overrightarrow{\theta }}_t$$

Es muss somit keine erneute Auswahl des Zielbereichs erfolgen, sondern der Zielbereich wurde anhand der Sensorsignale und dem Verfolgungs-Algorithmus getrackt. Alternativ kann anstatt des Recurcsive Least Squares Ansatzes beispielsweise ein Kaiman-Filter verwendet werden.In the subsequent step 52, the selected target area 32 is tracked or found again in the sensor signals I. ^B , D. ^B and D. ^A using a tracking algorithm via the control unit 37. This is preferably done using a Recursive Least Square (RLS) algorithm. This is a set of mathematical equations that represent a computationally efficient, in particular recursive, estimation of the state of a process. The mean square error is minimized here. The current measured position r _t at time t is used as an input value for estimating the future position r _{t +1} . The position r _t is the last determined position information of the target area, which was initially determined in step 44 and is determined again in the following step 54, which is explained in more detail below. The estimation of the future value r _{t +1} is based on a Recursive Least Square (RLS) algorithm. This is for example in the Book: "Recursive Least Squares; Statistical Digital Processing and Modeling; by Monson H. Hayes, ISBN 0-471-59431-8 disclosed. A tracking algorithm implemented in this way exhibits extremely rapid convergence. It would also be conceivable to use a Least Mean Square (LMS) algorithm for the tracking algorithm, which has simpler mathematical operations and a lower computational resource requirement. A linear model is preferably used to estimate the value r _{t +1} : $$r_{t+1}={\overrightarrow{r}}_t^T\cdot {\overrightarrow{\theta }}_t$$

Here is a measurement data vector ${\overrightarrow{r}}_t^T$

formed from N measurements up to time t and Θ _t are coefficients. First, the so-called Kalman gain vector is calculated: $${\overrightarrow{K}}_t=\frac{P_{t-1}\cdot {\overrightarrow{r}}_{t-1}}{\lambda +{\overrightarrow{r}}_{t-1}^T{P}_{t-1}{\overrightarrow{r}}_{t-1}}$$

Here λ (0 ≤ λ <1) is the so-called forgetting factor. P _t is the inverse correlation matrix of the measurement data. An a priori error ∈ between the current measured value r _t and the estimated value is then calculated using the following formula: $${ϵ}_t=r_t-{\overrightarrow{r}}_{t-1}^T{\overrightarrow{\theta }}_{t-1}$$

The coefficients Θ _t are determined using the following equation: $${\overrightarrow{\theta }}_t={\overrightarrow{\theta }}_{t-1}+{\overrightarrow{K}}_t{ϵ}_t$$

Furthermore, an inverse correlation matrix P _{t is} determined using the following formula: $$P_t=\frac{P_{t-1}-{\overrightarrow{K}}_t{\overrightarrow{r}}_{t-1}^T{P}_{t-1}}{\lambda }$$

The position information of the target area is then estimated at time t + 1 using the above equation: $$r_{t+1}={\overrightarrow{r}}_t^T\cdot {\overrightarrow{\theta }}_t$$

There is therefore no need to select the target area again; instead, the target area was tracked using the sensor signals and the tracking algorithm. Alternatively, instead of the recurcsive least squares approach, a Kalman filter, for example, can be used.

After the target area has been tracked or found again in the sensor signals in step 52, the information in the form of movement and position information of the tracked target area on the basis of the sensor signals is generated in subsequent step 54 I. ^B , D. ^B and D. ^A determined using the sensors 26 and 28. The procedure corresponds to that in step 44.

If, for example, the position information of the target area is determined in step 54, it can be determined how large the distance between the contact area 30 and the target area 32 is. If the contact area 30 has not yet reached the target area 32, the method is carried out again from step 48, which is shown by the arrow 56. If the contact point is reached, step 58 follows after step 54. That is, if a relative distance between the contact area 30 and the target area 32 has fallen below a previously defined threshold value, then the contact between these areas is regarded as established. The control is then passed on to the general movement compensation of the object, in this case the watercraft 1.

As already stated above, a fully automatic method for motion compensation and for merging two areas is also possible with the system according to the invention, which is explained in more detail below. According to Fig. 3 is different to Fig. 2 a further sensor 60 is provided in the watercraft 1. This is a scanning sensor with a comparatively large opening angle. For example, a LIDAR sensor can be used. The sensor 60 can then be used over a large area an environment can be scanned, this preferably being done in a comparatively high resolution. A recording speed then plays a subordinate role. The sensors 26 and 60 and the sensor pair 28 are intrinsically, extrinsically and mutually calibrated, which enables a coordinate transformation between the individual sensor coordinate systems, where K _{C is} the coordinate system of the sensor 60. The coordinate transformation can be done with the following common formulas: $${}_{K_{\mathrm{C.}}{K}_{{\mathrm{B.}}_1}}T,{}_{K_{\mathrm{C.}}{K}_{{\mathrm{B.}}_2}}T,{}_{K_{\mathrm{C.}}{K}_{\mathrm{A.}}}T,{}_{K_{{\mathrm{B.}}_1{K}_{\mathrm{A.}}}}T,{}_{K_{{\mathrm{B.}}_2}{K}_{\mathrm{A.}}}T\mathrm{and}{}_{K_{{\mathrm{B.}}_1}{K}_{{\mathrm{B.}}_2}}T$$

Im folgenden Schritt 66 erfolgt dann die Ausrichtung des Sensors 26 und des Sensorpaars 28 hin zum Zielbereich 32.The following is the fully automatic procedure with the system Fig. 3 based on Fig. 5 explained in more detail. According to a step 62 is in the Fig. 5 provided that the sensor 60 picks up sensor signals. This is done in the form of the data D. ^C. In the following step 64, a potential target area is identified. This is based on a correlation of the measurement data D. ^{C carried out} with a predetermined template database with possible scenes: $$\mathbb{D.}=\left\{{\overrightarrow{\mathrm{D.}}}_1^{\mathrm{C.}},{\overrightarrow{\mathrm{D.}}}_2^{\mathrm{C.}},{\overrightarrow{\mathrm{D.}}}_3^{\mathrm{C.}},...\right\}$$

In the following step 66, the alignment of the sensor 26 and the sensor pair 28 towards the target area 32 then takes place.

According to Fig. 5 following step 68 corresponds to step 38 from Fig. 4 , whereby sensor signals are detected. The following step 70 corresponds to step 44 from FIG Fig. 4 . This means that the movement and / or position information of the selected target area is determined. If this is not possible, the method is repeated starting from step 62, which is indicated by the arrow 72. Once the desired information has been determined, steps 74, 76, 78, 80 and 82 follow after step 70, followed by steps 48, 50, 52, 54 and 58 Fig. 4 correspond.

It can thus be ascertained that, in the fully automatic method, the target area 32, that is to say the docking point, is itself identified by the method, for example with object recognition. In contrast, the target area is specified by the operator using the sensor data in the semi-automatic process.

List of reference symbols

• 1; 10; 14; 16; 24 watercraft
• 2; 12 structure
• 4 gangway
• 6 docking area
• 8 end section
• 18 crane arm
• 20 rope
• 22 load
• 26, 28; 60 sensor
• 30 Contact Area
• 32 Target area
• 34 actuators
• 36, 38, 40, 42, 44, 48, 50, 52, 54, 58, 62, 64, 66, 68, 70, 74 - 82 step
• 37 Control Unit
• 46, 56, 72 arrow

Claims (15)

System for bringing a contact area of an object (1) of the system closer to a target area (32) of another object (24) so that the areas (30, 32) subsequently contact or so that the areas (30, 32) are in a position relative to one another are positioned, wherein at least one of the objects (1, 10, 16, 14, 24) is movable, and wherein the system has at least one sensor (26, 28, 60) which can be arranged on the object (1) of the system, this object (1) having a component (4) with the contact area (30), a control unit (37) being provided which is designed in such a way that information in the form of movement and / or position information about the target area (32 ) of the other object (24) relative to the contact area (30) can be determined on the basis of sensor signals from the at least one sensor (26, 28, 60), at least one actuator (34) being provided and the component (4) and / or the object (1) is movable with the component (4) via the actuator (34) / si nd, in order to provide movement compensation between the areas (30, 32) and / or to change the relative position of the areas (30, 32) to one another, and wherein the actuator (34) based on the information processed via the control unit (37) in this way It can be controlled that the areas (30, 32) approach one another and then contact one another or then maintain the predetermined relative position and / or that the movement compensation takes place. System according to claim 1, wherein a selection algorithm is provided which is designed in such a way that a target area (32) is automatically selected from the sensor signals detected by the at least one sensor (26, 28, 60), and / or where a tracking Algorithm is provided, which is designed in such a way that the target area (32) is tracked using the sensor signals detected by the at least one sensor (26, 28, 60), and / or a position-determining algorithm is provided which is it is designed that movement and / or position information of the target area is determined via this based on the sensor signals of the sensors (26, 28, 60), and / or a compensation algorithm is provided which is designed in such a way that the contact area ( 30) and / or the target area (32) can be returned to this relative position after an undesired relative displacement, starting from a current relative position. System according to claim 1 or 2, wherein one of the objects is a fixed structure (2, 12) and the other object (1, 10, 14, 16, 24) is movable, or both objects (10, 10, 14, 16, 24 ) are movable. System according to one of Claims 1 to 3, wherein the component is a load (22) movable via a winch (20) and the contact area is a bearing surface on the underside of the load (22) or wherein the component is a crane arm or an access bridge (4). System according to one of the preceding claims, wherein the sensor (26) is attached to the component (4). System according to one of the preceding claims, wherein the sensor (26) is designed as a light detection and ranging (LIDAR) sensor. System according to one of the preceding claims, wherein the sensor (28) or a further sensor (28) or a sensor pair (28) is attached to the object (1) and not to the movable component (4). System according to claim 7, wherein the sensor (28) or the further sensor (28) is a camera or wherein the sensor pair (28) comprises a camera and a light detection and ranging (LIDAR) sensor. System according to Claim 7 or 8, in which, in addition to the sensor (28) or the sensor pair (28) on the object (1), a further sensor (60) is provided for automatic detection of the target area (32). Vehicle with the system according to one of the preceding claims, wherein this is used as an object with the sensor (26, 28, 60). Fixed structure to which a vehicle can be docked, wherein the structure (2, 12) has the system according to one of Claims 1 to 9, and wherein the structure (2, 12) is used as an object with the sensor (26, 28, 60) is. Method with a system according to one of Claims 1 to 9 with the steps: - Detecting sensor signals from the at least one sensor (26, 28, 60), - Selection of the target area (32) based on the sensor signals, - Determination of the information in the form of movement and / or position information of the selected target area, - Control of the at least one actuator (34) based on the determined information in such a way that the contact area (30) and the target area (32) approach one another and / or that the movement compensation takes place. The method according to claim 12, wherein after the activation of the at least one actuator (34) and the movement of the contact area (30), the following steps follow: - Detecting sensor signals from the at least one sensor (26, 28, 60), - tracking the selected target area (32) in the sensor signals, - Determination of the information in the form of movement and / or position information of the selected target area (32) using the sensor signals of the at least one sensor (26, 28, 60), - Control of the at least one actuator (34) based on the determined information in such a way that the contact area (30) and the target area (32) approach one another and / or that the movement compensation takes place. Method according to claim 12 or 13, wherein when determining the information in the form of movement and / or position information of the selected target area (32), if the distance between these areas exceeds a certain threshold value, the sensor signals from a plurality of sensors (26, 28, 60) can be used. Method according to one of Claims 12 to 14, the target area (32) being selected manually via an operator or automatically via the control unit (37).

Images