CONTROL METHOD FOR A VEHICLE TREATMENT PLANT

Description

[0001] The present invention relates to a control method for a vehicle treatment system having one, two, or more treatment units that can be moved by controllable actuators to perform a surface treatment of a vehicle. The invention further relates to an associated software product, a control unit, and an associated vehicle treatment system.

[0002] Vehicle treatment systems known in practice exhibit a movement behavior for the positioning movements of the treatment units relative to the vehicle that appears awkward and jerky, at least in many movement phases. The movement of a treatment unit often involves several actuators, which are often controlled completely separately. This is especially true when an actuator has a multiple drive function and contributes to the partial movement of two or more treatment units.

[0003] EP 3 936 392 A1 discloses a conventional vehicle treatment system in which previously known treatment programs are parameterized and adapted to a specific vehicle to be washed. The vehicle is recorded locally via a camera, and a request is transmitted to a cloud-based system. The more powerful processing of centralized, cloud-based image information recognition is used to improve the locally executed treatment program. Cloud-based recognition has access to image data transmitted by a large number of vehicle treatment systems and other database contents on known vehicle types, so that the cloud system provides a better database for comparative analyses than purely local processing.The adaptation and parameterization of the treatment program at the local vehicle treatment facility includes geometry information or control instructions received as the content of a control parameter message, which is created and sent by the cloud-based data processing service.

[0004] US 2008 / 300724 A1 discloses a control method for the movement of the horizontal brush of a vehicle washing system, which provides a first and a second vector speed control mode for the brush, as well as logic for switching between these control modes. The first control mode is based on measuring the input current of the rotary drive of the brush, while the second control mode is based on a geometric limitation of the brush immersion depth.

[0005] In the technical field of industrial robot control, other robot control methods for path guidance are known.

[0006] Previously known control methods and vehicle treatment systems are not optimally designed. The object of the present invention is to provide an improved control method. It is also the object of the present invention to provide a treatment system with improved movement behavior.

[0007] The vehicle treatment system comprises one, two, or more treatment units that can be moved via controllable actuators. The movement of a treatment unit is understood below to mean an adjustment movement relative to the vehicle. The adjustment movement can, in particular—relative to the vehicle—be multidimensional, i.e., it can include several translational and / or rotational components or be caused by a combination of several partial movements of rotational and / or translational actuators.

[0008] The treatment units can be of any design and perform contact or non-contact surface treatment of the vehicle. For example, they can be single- or multi-part brushes (wash brushes, polishing brushes, roof brushes, side brushes, wheel wash brushes, etc.), blowers (roof dryers, side dryers), and / or movable applicators of foams and / or fluids (high-pressure nozzles, foam nozzles). Two or more treatment units can partially or completely share a common support structure. For example, two or more treatment units can be mounted together on a movable gantry or a mobile support of the treatment system.

[0009] It is possible for a treatment unit to be moved by a group of two or more actuators. Furthermore, it is possible for at least one actuator to be involved in the movement of two or more treatment units. The treatment units are connected to the actuators via support members. The geometry of the support members is known.

[0010] The actuators are designed to move one, two, or more treatment units relative to a vehicle being treated. The actuators can therefore be referred to as positioning actuators.

[0011] A control method is described below which represents a first independent aspect of the present disclosure.

[0012] Further below, a vehicle treatment system is additionally explained as a second independent aspect of the disclosure, which can be used in combination with the first aspect or independently thereof.

[0013] The aspects can be used by different individuals or companies, at different times and / or in different locations, and are thus completely independent of one another. On the other hand, the aspects form a common concept, as they each contribute, individually or in combination, to enabling improved motion behavior for a vehicle treatment system. This includes, in particular, synchronized accelerations for a group of actuators that jointly move one, two, or more treatment units. Such motion behavior is smooth and leads to reduced energy consumption and wear. It also enables a particularly rapid execution of a treatment process.To specify the movement behavior, a dynamic plan is preferably used. This plan comprises a plurality of target values ​​for the movements of the multiple actuators in an ordered data structure. The target values ​​define at least one movement variable of the respective actuator over the duration of a treatment phase. The dynamic plan thus breaks away from a purely momentary consideration and also from a singular actuator reference.

[0014] Below, a control method is first discussed, from which the generation of the dynamic plan and the movement behavior can be understood. Further down, the treatment system with the novel movement behavior is explained. The characteristics of the treatment system can result directly or indirectly from the application of the control method. Alternatively, it is possible that the disclosed treatment system is formed independently of the execution of the control method.

[0015] The control method according to the first aspect of the present disclosure comprises at least the following steps: Creating or reading in a sequence plan for a vehicle treatment, wherein the sequence plan has setting positions for a plurality of actuators as setting values ​​(state values) in a predetermined sequence; creating a movement plan for the vehicle treatment, wherein the movement plan comprises a target path, a target setting rate and preferably a target setting rate change for the plurality of actuators, specifically as (mathematical) functions in relation to a sequence variable; creating a dynamics plan for the vehicle treatment, wherein the dynamics plan comprises target values ​​for the movements of the actuators in relation to a time variable; and wherein the dynamics plan is generated from the movement plan by a local stretching and / or compression by means of a time scaling.

[0016] The target values ​​for the actuator movements can be present in any form in the dynamic plan. Particularly preferably, the target values ​​can be present in the form of at least one of the following motion variables: As a target trajectory for the majority of the actuators with respect to the time variable, AND / OR As a target speed for the majority of the actuators with respect to the time variable, AND / OR As a target acceleration for the majority of the actuators with respect to the time variable, AND / OR As a target jerk for the majority of the actuators with respect to the time variable.

[0017] The movement variables can be represented in any form, for example, in the form of a mathematically defined function, in the form of value tuples, or in the form of a list or table. They each create a multitude of assignments of a target value to a value of the time variable.

[0018] The control method preferably further comprises controlling the actuators during the execution of the vehicle treatment with a control, in particular with a state control, on the basis of the target values ​​of the dynamic plan.

[0019] Such a control method has several advantages. It overcomes the strict connection between spatial and temporal influences when planning a treatment procedure.

[0020] The flow chart can be generated in any desired manner. Particularly preferably, it can be generated based on contour detection data of the vehicle. A separate planning method can be used to generate the flow chart. It can be executed independently of or in combination with the control method. Furthermore, a flow chart generated by the planning method can be stored on a data medium. A stored flow chart can be read in and further processed in the control method.

[0021] By creating the motion plan and dynamic plan separately, as well as separating the process variables and time variables, the control method allows for greater flexibility in planning a treatment process. In particular, physical limits and specified operating limits of the actuators, as well as interdependencies between the movements of the multiple treatment units, can be considered in separate process steps, thus preventing or resolving contradictions in the planning specifications for the movements of the individual treatment units.

[0022] Overcoming such contradictions leads to the fact that the individual movements of the actuators can be planned in a completely coordinated manner in a common sequence and implemented in a treatment process.

[0023] In the control method according to the present disclosure, (ideal or tolerated) target spatial positions of the treatment units are first planned in a specific sequence. This can be done fully automatically, semi-automatically, or manually.

[0024] For these target spatial positions, a kinematics can then be planned (still) independently of time, which supports good controllability of the actuators. This kinematics is created based on parameterizable functions, which are, however, defined with respect to a process variable. The process variable can be understood as a quasi-time that does not correspond to real time. The process variable is suitable for assigning control positions and the resulting values ​​of a target path to a common variable (auxiliary variable). A value of this process variable can represent a planned target state, and the progression of the process variable's values ​​can define the order in which these target states should occur. The process variable also defines a "distance" between these target states on a scale. However, this is not a time interval measurable in seconds, but rather a purely computational intermediate reference value.The process variable is only mapped to real time through time scaling.

[0025] Time scaling for mapping to real time leads to a significant degree of planning freedom that is not available with the direct use of a time variable: Within the framework of time scaling, the process variable and thus the sequence of the assigned target values ​​can be traversed in a positive AND, at least in sections, in a negative direction. The gradient of the time scaling ds / dt can therefore be negative in sections. In other words, certain sequences or poses can be traversed in a backward direction to resolve contradictions that have no solution option when planning purely based on a time variable. The necessary steps for this can be fully automated and, in particular, enable automated planning and control in the presence of singly or multiply chained kinematics. This will be discussed further below.

[0026] By creating a dynamic plan using time scaling, the kinematics planned with respect to the process variable are adapted to a time variable that corresponds to real time. This enables the planned movements to be executed with such dynamics that certain predefined criteria are reliably met, regardless of the complexity and number of treatment units. Such criteria can include, for example, minimizing the total treatment duration, minimizing energy consumption, or adhering to predefined maximum or minimum speeds for the movement of a treatment unit along a vehicle contour. A maximum speed can be the highest (in terms of magnitude) speed in a first direction of movement. A minimum speed can be the highest (in terms of magnitude) speed in the opposite direction.

[0027] By separating the creation of the motion plan from the creation of the dynamic plan, the planning process becomes more robust and better results are achieved.

[0028] The movements of the actuators achieved by means of the control method according to the present disclosure can be partially synchronized or fully synchronized. A partially synchronized movement of the actuators with respect to an individual treatment unit occurs when the actuators (group of involved actuators) involved in a multi-dimensional actuating movement of a treatment unit perform accelerations at the same time phases. The movement is fully synchronized with respect to the individual treatment unit when the actuators involved in its movement additionally perform accelerations proportional to one another in the corresponding time phases. A dynamic plan generated according to the above steps, in particular including the time scaling, can be stored as an intermediate result.

[0029] Particularly preferably, a dynamic plan according to the present disclosure is designed such that, on the one hand, it provides partially synchronized movements or, more preferably, fully synchronized movements for the participating actuators of each treatment unit for which planning was / is carried out according to the present control method. Furthermore, partially synchronized and, more preferably, fully synchronized movements can be present for several treatment units. This means that the dynamic plan preferably provides a multi-dimensional (adjusting) movement for a group of treatment units, so that the actuators involved in a movement of the several treatment units execute accelerations at the same time phases, and more preferably, in the corresponding time phases, execute accelerations proportional to one another.

[0030] This prevents jerky or jerky movements, which often lead to increased wear and tear and poor washing results.

[0031] Control of a treatment system according to the present method, and in particular the presence of partially or fully synchronized (adjusting) movements of the actuators, can be detected, for example, based on the following phenomenon: A vehicle is subjected to a vehicle treatment in the vehicle treatment system. The instantaneous speeds of the actuators for the movement of the roof and side brushes can be determined. Subsequently, an additional obstacle, such as a test specimen on the hood, is placed on the outer contour of the vehicle (without changing its position). The obstacle should only affect the area of ​​engagement of either the side brush or the roof brush. The same vehicle treatment is then performed again, and the instantaneous speeds are determined again.If it is determined that the gantry movement decelerates (already) before the roof brush reaches the obstacle (because it is known from contour detection that the roof brush must avoid the obstacle) and that the roof brush's lifting movement also accelerates at the same time, then the movements of these actuators are at least partially synchronized. If it turns out that the gantry movement and the roof brush movement also accelerate proportionally while avoiding the obstacle, then these actuators are fully synchronized.

[0032] If one also determines that the speed of a feed or bending movement of the side brushes also changes at the time the gantry decelerates, a partially synchronized movement exists between all the actuators involved. Finally, if one also determines a correspondingly proportional acceleration value for the changed feed or bending movement, a fully synchronized movement of all the actuators involved exists.

[0033] According to a preferred embodiment, a jerk specification (specification for changing the acceleration) can additionally be provided as a movement variable for the movement plan and / or the dynamics plan in order to achieve a smooth transition between phases with different levels or directions of acceleration. A jerk specification can achieve even smoother changing movement behavior, so that, for example, vibration excitation during highly dynamic treatment phases (adjusting movements with significant changes in the instantaneous speed of one or more actuators) is reduced or limited to a predeterminable level. This allows the damping properties of the vehicle treatment system to be designed more tightly, which in turn can contribute to an overall higher treatment speed.

[0034] As a result, the motion behavior of the treatment units is no longer perceived as awkward by applying this control method, but rather as a smooth and fluid process. Furthermore, acceleration phases can be planned so that high drive energies are only used when necessary to meet one of the specified criteria. Unnecessary abrupt accelerations (including decelerations) can thus be avoided, which positively influences the energy requirements for vehicle treatment.

[0035] Within the framework of time scaling, physical and operating limits can be considered for individual actuators or—in a coordinated manner—for multiple actuators. For each subsection of the vehicle treatment system, it can be identified which actuator actuation is the limiting factor because its operation occurs within a physical or operating limit, and the (actuating) movement of additional or all treatment units can be adapted to this actuator. The role of the "limiting factor" can alternate between the subsections of the vehicle treatment system.

[0036] Within the time scaling, physical and / or operational limits for the movement of one or more treatment units or their target points can be considered alternatively or additionally. For example, a maximum path speed for a side brush or roof brush along the vehicle contour can be specified. Alternatively or additionally, any other specified criteria can be set and taken into account within the time scaling. This will be discussed further below.

[0037] This eliminates the need to design the dynamics of the vehicle treatment system to precisely match the limits of a specific actuator. Instead, the solution space for all possible movements can be described for each section of the vehicle treatment process, and the boundaries of the solution space can be calculated. Within this solution space, the desired dynamics can be freely selected.

[0038] Phase separation, i.e., limiting the planning of the treatment process to fixed phases, can be omitted. When planning treatment processes using conventional control methods, phase separations were often necessary to avoid contradictions in the direct planning of trajectories. The most common phase separation stipulates that for a phase of the treatment process, only one monotonic movement of the gantry was permitted (first forward crossing, first backward crossing, second forward crossing, second backward crossing). Using the control method according to the present disclosure, the movements of all actuators can be combined in any desired manner. Sequential separation of movement phases is unnecessary.

[0039] Particular advantages arise when two or more actuators have interlinked kinematics. This is the case when the movement of one actuator leads to a partial movement of at least two treatment units. Interlinked kinematics often occurs with regard to the travel movement of the gantry, which simultaneously leads to a partial movement of all treatment units attached to the gantry. However, the movement of a (traveling) carriage on which a side washing brush in the form of a bent brush is arranged also leads to a partial movement for several treatment units, namely the brush segments of this side washing brush. Thus, multiply interlinked kinematics can also exist.

[0040] The control method according to the present disclosure, and in particular the separation of the reference variables "sequence," "process variable," and "time variable," allows virtually any vehicle contour to be treated contact-wise, even when using actuators with singly or multiply linked kinematics, without the need for phase separation. For example, this makes it possible to clean concave contours, such as those found on inwardly drawn body parts, under an alcove, or under a roof rack.

[0041] According to the second aspect of the present disclosure, a vehicle treatment system is provided with at least one treatment unit and a plurality of controllable actuators for executing actuating movements of the treatment unit. At least one group of actuators is provided, which are designed to jointly set the treatment unit in a multidimensional actuating movement relative to a vehicle to be treated. The vehicle treatment system comprises a controller and at least one regulator. The controller transmits target values ​​to the at least one regulator, on the basis of which the regulator generates one or more actuating signals for controlling the actuators.

[0042] According to the second aspect of the disclosure, the controller comprises at least one dynamic plan that defines, in an ordered data structure, a plurality of target values ​​for the movements of the actuators with respect to a time variable.

[0043] The dynamic plan can be generated by the control method according to the first aspect of the disclosure. However, it can also be generated in a different way or at a different location and, in particular, can be read in as a predefined plan. For example, in vehicle treatment systems that repeatedly wash identical vehicle types, particularly optimized dynamic plans can be generated and stored, which are read in for a specific treatment process and, if necessary, adapted to the current positioning of the vehicle. This approach is particularly useful for train washing systems or car and truck washing systems for cleaning fleet vehicles.

[0044] The target values ​​of the dynamic plan define at least one movement variable of the respective actuator over the duration of a treatment section.

[0045] Particularly preferably, the definition is chosen such that the target values ​​describe a current and a future movement of the respective actuator. A treatment section can, for example, comprise a front cleaning or a rear cleaning with a specific treatment unit.

[0046] A target point is defined for the treatment unit, and the target values ​​of the dynamic plan are defined for the associated group of actuators such that the target point of the treatment unit, during a multidimensional movement, follows a path that maintains a geometric reference to a contour of the vehicle to be treated during surface treatment. The contour of the vehicle to be treated can be a surface contour or a line contour. Preferably, the contour of the vehicle to be treated defines a local spatial position, in particular as a translational spatial position and / or a rotational spatial position.

[0047] The dynamic plan further defines, for each phase of the treatment section, synchronized accelerations for the actuators of the group with a multi-dimensional positioning movement of the treatment unit, so that consistent acceleration phases of the actuators result when the multi-dimensional positioning movement is executed.

[0048] The vehicle treatment system according to the second aspect of the disclosure therefore no longer exhibits phase separation between the control of actuators that jointly move a treatment unit. If the treatment unit requires two or three separate actuating movements for a treatment section according to the course of the surface contour, for example, a translation of the gantry, a translation of a carriage on which a brush is suspended, and a rotation of an articulated joint, the associated actuators will not execute these movements separately in time, but rather in a coordinated manner. The coordination is such that the associated accelerations of a guide actuator for the gantry, a push actuator for the carriage, and a tilt actuator for the articulated joint always experience accelerations at overlapping time intervals.The same can apply to all other multidimensional movements of a treatment unit caused by a group of actuators.

[0049] Coincident acceleration phases of two or more actuators essentially occur in parallel. They therefore overlap along the time axis. Furthermore, chains of acceleration phases can exist for multiple actuators, with these chains each comprising two, three, or more consecutive acceleration phases. From a macro perspective, these chains can essentially occur simultaneously for several actuators in the group.

[0050] In particular, for the duration of a multidimensional movement, there may be a chain with a plurality of acceleration phases for each of the actuators involved in the group.

[0051] Even at a micro level, overlapping acceleration phases—for example, a first acceleration phase from a first chain for a first actuator and a corresponding second acceleration phase from a second chain for a second actuator—can have exactly the same durations. Alternatively, the overlapping acceleration phases can overlap to a predetermined minimum extent, with the minimum extent preferably being greater than 60%, more preferably greater than 80%, of the duration of the longer acceleration phase.

[0052] Alternatively or additionally, the corresponding acceleration phases for each of the involved actuators can start and end at corresponding times.

[0053] The corresponding acceleration phases can further preferably provide maximum or minimum accelerations for a majority of the participating actuators at corresponding times. The geometric reference of the target point relative to the contour of the vehicle to be treated can be arbitrarily defined and, in particular, can depend on the type of treatment unit. The geometric reference can, in particular, be a fixed vertical distance from the vehicle surface AND / OR a fixed distance from the vehicle surface that is at a predetermined angle of influence relative to the vehicle surface.

[0054] The path of the target point can be determined accordingly. The path of the target point can, in particular, follow the contour of the vehicle surface and, furthermore, can be a path along the vehicle surface that is possibly smoothed at points on the vehicle surface with a curvature that exceeds a predetermined limit. Alternatively, the path of the target point can follow the contour of the vehicle surface at a parallel distance. Such a parallel path can also be smoothed at points on the vehicle surface with a curvature that exceeds a predetermined limit.

[0055] The accelerations synchronized with each other are preferably defined predominantly (for at least 60% of the time share, more preferably for at least 80% of the time share of multidimensional movements) or exclusively (for all time shares of the multidimensional movements).

[0056] The target values ​​of the dynamic plan preferably defined fully synchronized accelerations for the actuators in the group, which remain proportional to each other in corresponding time phases. This applies in particular to multidimensional positioning movements with which the treatment unit is moved from a rest position at the beginning of a treatment section to a working position and / or to multidimensional positioning movements with which the treatment unit is moved from a working position to a rest position at the end of a treatment section.

[0057] A vehicle treatment system configured according to the second aspect of the disclosure exhibits smooth movement behavior during multidimensional adjustment movements of the at least one treatment unit relative to the vehicle. This has advantages with regard to energy consumption and the stress on the mechanical structures. Furthermore, a significant reduction in the overall duration of a treatment section is achievable.

[0058] Further developments of the invention and their advantages are set out in the subclaims and the following description.

[0059] The invention is illustrated schematically and by way of example in the drawings. They show: Figure 1-3: A vehicle treatment system in three states; Figure 4: A dynamic plan for a vehicle treatment; Figure 5: Explanations for creating a flow chart for a vehicle treatment based on contour recognition data; Figure 6: A flow chart with target positions of an actuator; Figure 7: A motion plan with a single target path for the flow chart according to Figure 6; Figure 8: A motion plan with a plurality of target paths, target actuation rates and target actuation rate changes; Figure 9: A time scaling plan with curves of a section-wise constant scaling and a variable scaling; Figure 10: A state space diagram to explain the formation of a variable scaling using a first example. Figure 11: An explanatory representation for generating a dynamic plan from a motion plan using time scaling; Figures 12, 13: Explanatory representation for generating a sequence plan with a freely definable portal movement, without maintaining phase separation; Figure 14: A state space diagram to explain the formation of a variable scaling using a second example. Figure 15: An example function to explain the section-wise definition of parameterizable (mathematical) functions for a motion plan.

[0060] Figure 1shows a vehicle treatment system (1) with two or more treatment units (6, 7, 12) that can be moved by controllable actuators (21, 22, 23) in order to carry out a surface treatment of a vehicle (3). The vehicle treatment system (1) comprises a control unit (30) and a contour detection device (31) for detecting a contour of the vehicle (3) to be treated. The vehicle (3) is preferably arranged in a defined treatment zone (2) before the start of a treatment process. In this case, a one-time contour detection can be carried out, wherein the contour detection data are assumed to be constant for the entire treatment process.

[0061] Alternatively, iterative or continuous contour detection can be performed. In this case, no (fixed) arrangement of the vehicle (3) in the treatment zone (2) is required. However, for the sake of simplicity, it is assumed below that the vehicle is fixedly arranged in the treatment zone (2) for the duration of the vehicle treatment. Furthermore, for the sake of simplicity, it is assumed below that spatial coordinates (x, y, z) of the treatment system (1) are defined on the basis of a coordinate origin that has a fixed relationship to the treatment zone (2) and thus - for the duration of a treatment process - to the vehicle (3). However, the following explanations can also be applied to a different form of coordinate definition, which in particular also includes a moving coordinate system.

[0062] A relative movement takes place between the treatment units (6, 7, 12) and the treatment zone (2) to carry out the vehicle treatment. This can be configured as desired. The following illustration assumes that the relative movement is defined with respect to the aforementioned coordinate system (x, y, z) with a fixed coordinate origin.

[0063] Figures 1 to 3 show the vehicle treatment system (1) and its treatment units in different poses. A pose defines the spatial position of several (active) treatment units, and preferably all (active) treatment units (included in the planning according to the present control method) at a specific moment or state.

[0064] A planning method is explained below, which represents an independent aspect of the present disclosure. The planning method comprises the following steps: For at least one treatment unit (6, 7, 12): definition of a target point (Z1, Z2, Z3) as a geometric reference point of the treatment unit (6, 7, 12), the spatial position (R) of which can be defined in coordinates (x, y, z); determination of a sequence (i = 0, 1, 2, ...) of desired spatial positions (Ri, R0, R1, R2, ...) of the target point (Z1) to be assumed during a vehicle treatment; determination of actuating positions (Pi, P0, P1, P2, ...) of at least one actuator (21, 22, 23) involved in the movement of the target point (Z1) on the basis of a known geometry of the vehicle treatment system (1), such that the actuating positions (Pi, P0, P1, P2, ...) correlate with the desired spatial positions (Ri, R0, R1, R2, ...) of the target point (Z1).

[0065] The sequence of target spatial positions can be determined in any way. It is preferably based on contour recognition data (51) of the vehicle (3) to be treated.

[0066] A target point can be located inside or outside the physical dimensions of a treatment unit.

[0067] For a brush, it may be useful to place the target point on the rotation axis. The goal may then be to always position this target point at a specified distance perpendicular to the vehicle surface.

[0068] For a treatment unit with long-range action, such as a drying blower or a jet cleaning nozzle, it may be useful to define the target point outside the physical dimensions of the treatment unit, for example, with a specified effective distance centrally in front of a nozzle opening. The aim may then be to position a desired spatial position of this target point on the surface of the vehicle to be treated, preferably with a specified angle of impact relative to the vehicle surface. For example, with a drying nozzle, it may be desirable for its air jet to always hit the vehicle surface at a distance of 40 cm from the nozzle opening and at an angle of approximately 30°. The above angle and distance values ​​are given purely as examples. Any other distance and / or angle value can be provided.

[0069] The planning of target spatial positions with a specified distance to the vehicle surface can lead to the final result - when carrying out the planned surface treatment - of a path of the target point (e.g. a brush) that follows the contour of the vehicle surface with a parallel distance.

[0070] The planning of target spatial positions on the surface of the vehicle to be treated can lead to a path of the target point (for example of the remote treatment unit) that follows the contour of the vehicle surface when the planned surface treatment is carried out.

[0071] The aforementioned examples can be combined in any way. In both cases, the goal may be to smooth the path of the target point relative to the contour of the vehicle surface, particularly where the vehicle surface has a curvature that exceeds a predetermined limit.

[0072] The aforementioned path profiles of a target point can be generated based on the planning and control method according to the present disclosure. They also represent an independent feature of the vehicle treatment system according to the present disclosure and could also be generated based on a manual creation of a dynamic plan (70) or by reading in a dynamic plan (70) that was generated and stored outside the vehicle treatment system and its control system.

[0073] The planning process or its steps can be executed as a separate process, the result of which is available as an organized and computer-readable data structure. The data structure can be stored on a data medium and / or sent via an interface.

[0074] Preferably, a sequence plan is formed from the spatial positions of one or more treatment units or their target points, which belong to the same state. Accordingly, the positioning positions in the sequence plan preferably provide control values ​​for a plurality of actuators that, at a specific value in sequence (i), are related to the same target state for which a pose is planned (state values).

[0075] During a treatment process, the vehicle (3) preferably has a known location relative to the treatment zone (2). This can apply both to a fixed treatment zone (2), for example, in a car wash, and to a moving treatment zone (2), such as a towing area along a car wash line.

[0076] The control method according to the present invention can be used for any vehicle treatment systems (1), regardless of whether the treatment zone (2) is stationary or moving, and regardless of how the treatment units (6, 7, 12) are designed, driven and possibly coupled to one another.

[0077] In the context of the present disclosure, the terms "support member", "target point", "spatial position", "actuator" have the following meaning.

[0078] Support members are mechanically supporting components of a vehicle treatment system (1) to which a treatment unit (6,7,12) is directly or indirectly attached.

[0079] A target point (Z1, Z2, Z3) is a geometric reference point for a treatment unit whose current spatial position can be defined in coordinates. The coordinates can define a translational spatial position (position) and / or a rotational position (orientation), particularly with respect to a designated treatment zone of the vehicle treatment system.

[0080] The spatial position of the target point (Z1,Z2,Z3) of a treatment unit is physically coupled with the known geometry of the one or more support members that are relevant for the movement of the treatment unit, as well as with the current control values ​​(current control position / current control rotation, etc.) of the one or more actuators that are relevant for the movement of the treatment unit.

[0081] An actuator is a controllable motion drive. An actuator can be connected to another actuator via a support link.

[0082] A control method according to the present disclosure is explained below using an exemplary embodiment relating to a gantry car wash. The steps of the method are analogously applicable to any other embodiments.

[0083] Figures 1 to 3 first explain different spatial positions of treatment units (6, 7, 12) or their target points (Z1, Z2, Z3) for three different states during a treatment process. In Figure 5 In addition, spatial positions (R) of a target point (Z1) of a side washing brush (6) relative to the vehicle (3) or the contour detection data (51) are illustrated.

[0084] The treatment units (6, 7, 12) are moved into various spatial positions by controlled movements of actuators and the associated movements of support members. The portal (4) with the guide rails for the side washing brushes (6, 7), the roof brush (12), and the brush rods (6c, 7c) are examples of support members. The term "support member" encompasses any other rigid or elastic support means with which a treatment unit can be held or moved.

[0085] At least one actuator (21, 22, 23) is provided for the movement of each treatment unit (6, 7, 12). Furthermore, at least one support member can be provided for each treatment unit (6, 7, 12). However, two or more support members and two or more actuators can also be provided for each treatment unit (6, 7, 12).

[0086] A side washing brush (6) has in the example of Figures 1 to 3two rotatable brush sections (6a, 6b) - an upper brush section (6a) and a lower brush section (6b) - wherein these brush sections are arranged adjacent to one another in the longitudinal direction of the rotation axis. An articulated joint (17) with a articulated actuator (18) is provided between the brush sections (6a, 6b), so that the lower brush section (6b) can be pivoted through an angle relative to the upper brush section (6a).

[0087] The upper brush section (6a) is also suspended from a carriage (8) via a tilt joint (15). This tilt joint (15) allows the upper brush section (6a) and thus indirectly also the lower brush section (6b) to be pivoted outward relative to the vertical direction, for which purpose a tilt actuator (16) is provided on the tilt joint.

[0088] The carriage (8) is displaceable on a guide rail of the portal (4) in a first horizontal direction (transverse direction y of the vehicle treatment system), for which a sliding actuator (9) is provided.

[0089] The guide rail for the side brushes is attached to a gantry (4) of the vehicle treatment system, whereby the gantry (4) can be guided along a longitudinal guide in a further horizontal direction (longitudinal direction x of the vehicle treatment system). A gantry drive / guide actuator (5) is provided for this purpose.

[0090] The relevant geometrical references are known for the support, the guide rail, the carriage, the tilt joint, the upper brush section, the articulated joint and the lower brush section, e.g. their size and possibly their load-dependent deformation behavior.

[0091] Current positions can be specified for the bending actuator, the tilting actuator, the sliding actuator, and the guiding actuator. A position can be a rotational position and / or a translational position, in particular.

[0092] If a (desired) spatial position (R) of the target point (Z1, Z2, Z3) of a treatment unit (6, 7, 12) is known, at least one pose (posture) can be calculated for the parts of the treatment system that contribute to the movement of the treatment unit. The pose is defined by the geometric references of the supporting elements involved and the positioning positions of the actuators involved. From a desired target spatial position (Ri) of a target point, at least one suitable pose can be calculated through which this target spatial position (Ri) can be reached, and accordingly, an associated positioning position (Pi) can be determined for one or each actuator involved.

[0093] If all participating actuators simultaneously assume the respective target position, the treatment unit (6,7,12) will actually assume the spatial position (R) that is defined as the target spatial position for the target point (Z1).

[0094] For the following explanations, the spatial position (R) of the side washing brush (6) and in particular its target point (Z1) will be considered as an example. First, it is assumed for simplicity that the tilt joint (17) and the articulated joint (18) are exclusively in the extended position. Later, cases will be discussed (as you can see in Figure 2 and 3 shown), in which the tilt joint (18) and / or the articulated joint have a different position.

[0095] The movement of the target point (Z1) depends on the positioning positions of the gantry drive / guide actuator (5), the sliding actuator (9), and the tilt actuator (17). These are referred to below as the first actuator (21), the second actuator (22), and the third actuator (23). However, the following illustration can be applied to other actuators as desired.

[0096] Figure 4 shows the objective pursued by the control method according to the present disclosure. First, it should be noted that the dynamic plan shows the time variable (t) on the abscissa axis (40). A dynamic plan is therefore to be created that includes, with respect to the time variable (t), predetermined target values ​​for at least one movement variable for each actuator to be considered—here, the first, second, and third actuators (21, 22, 23), in particular: a target trajectory (A,B,C), a target speed (A°,B°,C°), and / or a target acceleration (A°°,B°°,C°°)

[0097] The time variable (t) should correspond to real time and be defined, for example, in milliseconds or seconds.

[0098] A target trajectory specifies which control value (e.g. in degrees or millimeters) an actuator should reach as a target value at each planned time.

[0099] The target speed specifies the current speed (e.g. in rad / s or m / s) that the actuator should reach as a target value at each planned time.

[0100] And the target acceleration specifies which instantaneous acceleration (e.g. in rad / s 2< or m / s 2< ) the actuator should reach as a target value at each planned time.

[0101] The parts (movement variables) of the dynamic plan (70) can be in any form processable by a computer, in particular as (mathematical) functions and / or (individual) values ​​in an ordered data structure such as an array, a vector, a table, etc.

[0102] The dynamic plan (70) can comprise a complete treatment process with one or more treatment cycles. Alternatively, it can be available separately for different treatment cycles and / or generated gradually for specific time periods, e.g., when contour detection is performed continuously or iteratively.

[0103] The time variable and the setpoints can be present as continuous or discrete variables. In a preferred embodiment, the dynamic plan is present as an ordered data structure that includes the corresponding setpoints for the target trajectory, the target velocity, and the target acceleration of the actuators for a plurality of points in time (values ​​of the time variable). The setpoints can be present for each point in time, for example, a set of setpoints per millisecond, or only for a subset of the possible points in time.

[0104] The setpoint values ​​from the dynamic plan can preferably be transferred to a (state) controller (32), which generates corresponding control signals directly or indirectly on the basis of the setpoint values ​​in order to control and in particular to regulate the actual movement of one or more actuators (21, 22, 23).

[0105] The following explains how such a dynamic plan is generated in several process steps using the control method.

[0106] In Figure 5 Three diagrams are shown one above the other, showing the x-coordinate of the vehicle treatment system (1) on the abscissa axis (40). On the ordinate axis (41), the diagrams show: Coordinate values ​​along a vertical axis (z) / z-coordinate, coordinate values ​​along a transverse direction (y) / y-coordinate, and values ​​of an inclination angle (a).

[0107] This is therefore a representation of spatial relationships. The diagrams contain exemplary contour detection data (51), which show the cross-sectional contour (upper diagram) and two profiles of a lateral contour (middle diagram), as well as profiles of an inclination angle at two elevations (lower diagram). At the top of Figure 5exemplary reference points (52) on the outer contour of a vehicle (3) are shown, to which the contour detection data (51) in the diagrams can belong.

[0108] The middle diagram shows an example of how, based on the contour detection data (51) of the vehicle (3), different spatial positions (R) of the side washing brush (6) can be planned in a specific sequence (i). For simplicity, it is assumed here that the side washing brush (6) is oriented exclusively vertically, so that the reference points (Z1, Z2) lie directly above one another in the plan view.

[0109] Generally speaking, for at least one treatment unit and preferably each treatment unit (6,7,12), a target point (Z1, Z2, Z3) is defined as a geometric reference point of the treatment unit (6,7,12), the (current) spatial position (R) of which can in turn be defined in coordinates (x,y,z).

[0110] A sequence (i = 0, 1, 2,...) of target spatial positions (Ri, R0, R1, R2,...) of the at least one target point (Z1) is planned, which are to be assumed during a vehicle treatment. The planning can be carried out manually, partially automated, or fully automated based on the contour recognition data (51) of the vehicle (3) to be treated. The planning can be carried out relative to a treatment zone (2) of the vehicle treatment system (1), which in Figure 5 for easy comprehension. Alternatively, it can be performed with reference to one or more characteristic points (so-called markers) on the vehicle (3) and thus independently of an external absolute coordinate system. This planning is possible for any number of treatment units, each of which can be active individually, overlapping in time, or simultaneously, and in particular, in an effective position relative to the vehicle.

[0111] In the example of Figure 5Target spatial positions (Ri, R0, R1, R2,...) are automatically planned for several steps (0, 1, 2, 3,...) in the sequence (i). For this purpose, for example, positions for the side brush (6) are determined at a predetermined grid along the x-coordinate or along the contour line, which positions are spaced from the contour line (53) by a predetermined amount in the normal direction. Alternatively, any other planning method can be used. Planning is preferably carried out in the form of a target spatial position for the target point (Z1 / Z2).

[0112] Generally speaking, a sequence (i = 0, 1, 2, ...) of target spatial positions (Ri, R0, R1, R2, ...) of the target point (Z1) is defined, which are to be assumed during a vehicle treatment, based on the contour recognition data (51) of the vehicle (3) to be treated. The aforementioned grid is not mandatory. It merely serves to facilitate the traceability of the procedure.

[0113] Based on the known geometry of the vehicle treatment system (1), it is possible to determine for each target spatial position (Ri) which actuators must assume in order for the side brush to actually assume these spatial positions. Therefore, for each target spatial position, a corresponding target position can be determined for each actuator involved in the movement of the side brush.

[0114] In the example of Figure 5 For the sake of simplicity, only a few setting positions (Pi, P0, P1, P2, ...) are entered for the first actuator (21), ie, the portal drive / guide actuator (5). Corresponding setting positions can be determined for a second, third, or further actuator (22, 23).

[0115] Generally speaking, setting positions (Pi,P0,P1,P2, ...) of at least one actuator (21,22,23) involved in the movement of the (respective) target point (Z1) are determined on the basis of a known geometry of the vehicle treatment system (1), so that the setting positions (Pi,P0,P1,P2, ...) correlate with the desired spatial positions (Ri,R0,R1,R2, ...) of the target point (Z1).

[0116] Instead of the above-mentioned procedure, any other planning types are possible. For example, manual detailed planning can be carried out for known vehicle types and their known contour data and saved with the corresponding setting positions. Determining the setting positions then involves retrieving a previously saved data set, e.g. based on a type recognition of the vehicle to be treated and an adaptation to the determined position of the vehicle (3) in the current treatment situation. Alternatively or additionally, default criteria can be defined for individual or combined poses or states for the treatment process. The default criteria can, in particular, be defined only for certain sections of the vehicle treatment and / or differently for each section. For example, a maximum path speed of a side or roof roller along a vehicle window can be set to a separate limit value.

[0117] Particularly preferably, target points are defined for a plurality of treatment units, in particular for all (active) treatment units, and for the creation of a sequence plan (50), their current spatial positions (R, L) are determined in coordinates (x, y, z), which are particularly related to the same coordinate system. This allows spatially coordinated target spatial positions (Ri, Li) of these multiple target points (Z1, Z2) to be planned as a target pose sequence (K). An explanatory example of this follows below.

[0118] A process plan can be created individually for each vehicle to be treated in the vehicle treatment system (1). Alternatively or additionally, it is possible to save a process plan and reload it. For example, process plans can be predefined for different vehicles or vehicle types and reloaded if a specific vehicle or vehicle type has been identified for an upcoming treatment process based on contour recognition.

[0119] A motion plan for the vehicle treatment is created, which includes a target path (U, V, W), a target actuation rate (U', V', W') and preferably a target actuation rate change (U'', V'', W'') for the majority of the actuators (21, 22, 23) as (mathematical) functions in relation to a process variable (s). This process is carried out in the transition from Figure 6 on Figure 7 explained.

[0120] Figure 6 shows the already in Figure 5The flow chart (50) sketched for the first actuator (21) is shown in a separate diagram. Here, the sequence (i) is shown on the abscissa axis (41), and an associated value of the setting position (Pi) is shown at a corresponding distance along the ordinate axis (x).

[0121] Figure 7 shows a target path (U) for the first actuator (21). The process variable (s) is shown on the abscissa axis (40). The ordinate axis shows target values ​​for a position of the first actuator, which correlate with the actuating positions (Pi) along the x-axis.

[0122] The sequence variable (s) can be defined in any way. For example, the values ​​(0, 1, 2, ...) of sequence (i) could be arranged on a continuous scale, with the same unit spacing between each of these values.

[0123] Alternatively, the process variable (s) can be specified manually by a planner or by another part of the program.

[0124] If a partially or fully synchronized movement of the actuators involved in the movement of a single treatment unit is desired, it is recommended to provide the same sequence variable (s) for all involved actuators and the associated target paths (U, V, W), as well as their derivatives. Even more preferably, exactly one sequence variable (s) can be used jointly for all (active) treatment units and actuators, which are taken into account when implementing the control method.

[0125] In the transition from Figure 6 to Figure 7For example, it is assumed that a path parameter along a designated movement path was determined for the target point (Z3) of the roof brush (12), which was adopted as the process variable (s) for all treatment units and actuators. The designated treatment path is shown in Figure 12 The target spatial positions for all other target points (Z1,Z1) are in this case assigned to the corresponding values ​​on the abscissa axis (40) along the process variable (s) according to the corresponding states during treatment planning. This results in a local change in the position of the points (P0 to P13) in Figure 7 opposite Figure 6 along the abscissa axis (40). This type of procedure is not limiting for the present disclosure.

[0126] The target path (U, V, W) is preferably defined in length segments of the process variable (s), i.e., for a range [s_min < s < s_max], as a function suitable for reaching the target positions (Pi) of the process plan (50). It can be specified that each target position (Pi) or each associated control value (61) is to be reached exactly. Alternatively or additionally, control values ​​(62) can be planned that are to be reached while maintaining a tolerance (T).

[0127] Particularly preferably, depending on the situation within a treatment process, a control value (61) for exact achievement can be specified for some of the setting positions (Pi), and a control value (62) for achievement within a tolerance can be specified for another part of the setting positions (Pi). The tolerance can be defined uniformly or differently for different control values ​​(62).

[0128] Parameterizable functions are selected with which the target path is defined (section by section).

[0129] Examples of suitable functions are: Third-degree polynomials (four parameters per functional section) Fifth-degree polynomials (six parameters per functional section) Seventh-degree polynomials (eight parameters per functional section) Straight line segments with parabolic ends (eight parameters per functional section) Regression splines (number of parameters can be selected) Bézier curves The selection of functions can influence the result of the planning and control method according to the present disclosure. This is because, at a later point in the planning or control method, at least one target movement variable for the movements of the actuators is generated from these parameterizable functions. Some of the aforementioned functions, in particular the polynomials, result in the movement variables of two or more actuators ultimately exhibiting congruent changes in the target values ​​in phases that are simultaneous with one another. The generation of matching acceleration phases (ki - see Figure 11) which not only run parallel to one another in time, but also meet further criteria. These criteria can include, on the one hand, that two phases (ki) for a first actuator and a second actuator begin and end at the same time (ti). A further criterion can be that the acceleration values ​​of a first actuator in a first phase are proportional to the acceleration values ​​of a second actuator in a second phase, whereby the first and second phases coincide with one another, i.e. at least run parallel to one another in time. Matching phases in the target accelerations of the dynamics plan can therefore arise simply as a result of the selection of the parameterisable function for the target paths of the motion plan. With other parameterisable functions for the motion plan, in particular the straight line sections with parabolic ends, matching phases also result, although not necessarily with the same start and end times.The start and end times of these phases may be slightly offset from each other. In these cases, however, it is also possible to subsequently adjust the phase boundaries to achieve full phase synchronization.

[0130] The following is based on the Figure 15 An example for determining a target path is explained. Here, a first mathematical function in the form of a third-degree polynomial (solid line) is selected. This function is assigned to a first function section (U_left) in the left area. s 0 ≤ s ≤ s 1 and a second functional section section (U_right) in the right area s 1 < s ≤ s 2 defined separately as: U links s = k 0 + k 1 s + k 2 s 2 + k 3 s 3 U rechts s = l 0 + l 1 s + l 2 s 2 + l 3 s 3

[0131] According to the selected function, a target actuating rate (U') and preferably a target actuating rate change (U'') are determined by derivation.

[0132] The target rate (U') and the target rate change (U'') are thus obtained by simple and double differentiation with respect to the process variable (s). They are available for both sections as: U ′ links s = dU links ds U " links s = dU ′ links ds U ′ rechts s = dU rechts ds U " rechts s = dU ′ rechts ds

[0133] The boundaries of the two sections (left, right) are at the s-values ​​of three consecutive setting positions (Qi).

[0134] To understand the mathematical transformation and parameterization, it is not necessary to consider the complex processes according to the previous figures. Instead, Figure 15 For example, three positions (Qi) can be generally defined as: A setting position Q0 at the left boundary of the left section, ie for s 0 =s(Q0), a setting position Q1 at the transition between the left and the right section, ie for s 1 =s(Q1), and a setting position Q2 at the right boundary of the right section, ie for s 2 =s(Q2) with Q0: start point, Q1: transition point, Q2: end point.

[0135] The abscissa axis (40) shows in Figure 15 the process variable (s), the ordinate axis (41) shows the amount of the setting position (Q) and thus the function value of the target path (U).

[0136] The mathematical example explained with the setting positions (Q1, Q2, Q3) can be transferred analogously to the setting positions (Pi) mentioned in the other figures. It is also the case with the second Figure 15 shown example (dashed line), where a straight line with parabolic ends is chosen as the function for both sections (left, right).

[0137] The total of eight parameters (k0, k1, k2, k3, 10, 11, 12, 13) of the two functional sections U_left, U_right for the desired path can be determined using a system of equations with eight boundary or transition conditions. These conditions can be defined appropriately depending on the application.

[0138] In the present example, it is assumed that at the starting point Q0 an actuator with a specified starting control value, a specified starting control rate and a specified starting control rate change enters the first section.

[0139] At the transition from the left section to the right section, the function should be continuous, and the actuator's control value should reach transition point Q1. Furthermore, at transition point Q1, this point should be passed through with a continuous change in the control rate.

[0140] At the end position, the actuator should return to a specified final control value with a final control rate and a final control rate change. Thus, the following boundary and transition conditions can be formulated: U links s 0 = Q 0 , Start − Stellwert U ′ links s 0 = Start − Stellrate U " links s 0 = Start − Stellraten ä nderung U links s 1 = = Q 1 U rechts s 1 = Q 1 U ′ links s 1 = U ′ rechts s 1 U " links s 1 = U " rechts s 1 U rechts s 2 = Q 2 End − Stellwert

[0141] If this system of equations is solved, the parameters (k0, k1, k2, k3, 10, 11, 12, 13) are fixed and suitable functions are defined for the two sections (left, right) as target path (U), target actuation rate (U') and target actuation rate change (U'') for the motion plan.

[0142] If required, a target jerk (U‴) can also be defined.

[0143] This method can be executed for several or all actuators (21, 22, 23) in order to define a corresponding number of target paths (U, V, W), target actuating rates (U', V', W'), and target actuating rate changes (U'', V'', W''). The method can also be executed for a chain of sections of any length. In this case, a section can be defined between two consecutive actuating positions (Pi).

[0144] Although the above example was explained for a third-order polynomial, it can be implemented in the same way for higher-order polynomials. In this case, additional boundary or transition conditions can be introduced, such as specifying the jerk (target jerk (U‴, V‴, W‴ - not shown - third derivative with respect to the process variable s) at the start point and end point and / or at individual transition points. Furthermore, two or more chained functions can be provided within a section (e.g., a straight line with parabolic ends).

[0145] Alternatively or additionally, it may be provided that the adjustment rate change and / or the adjustment jerk at one (or each) transition point should be continuous, ie "U" links ( say ) = U" right ( say ) or Unlinks ( say ) = U‴ right ( say ) , etc.

[0146] The same procedure can be performed with any other suitable function type that has a suitable number of parameters. Particularly good dynamics were observed when the target path is defined, at least in individual sections, as a straight line with parabolic ends (dashed line in Figure 15 ).

[0147] The selectable function types can be freely mixed between the target paths (U, V, W) and their sections. A function type can be selected in any way.

[0148] A preferred embodiment provides that the vehicle treatment system or its control system (in particular the control unit 30 or a software product according to the present disclosure) has a set of rules and / or a set of function templates for specific parts of the treatment process, wherein the set of rules or the function template provides certain characteristic actuating positions according to type and sequence for individual or all participating actuators and predefines suitable functions as well as associated boundary or transition conditions between these provided actuating positions. The exact location of the characteristic actuating positions can be calculated depending on a contour detection of the vehicle. Subsequently, the boundary and transition conditions can be calculated in terms of height, and the parameters of the predefined functions can be determined.

[0149] The rule set and function templates can be predefined by the manufacturer for specific vehicle types and / or special body shapes in order to achieve particularly good washing performance.

[0150] The above-mentioned steps for forming a target path (U,V,W) and the associated target actuating rate (U',V',W') and target actuating rate change (U",V",W") can be applied for two or more and in particular all actuators (21,22,23).

[0151] Figure 8illustrates a movement plan (60) in which a target path (U, V) is defined in sections by suitable functions for the first actuator (20) according to the above example, i.e. for the gantry drive / guide actuator (5), and for the second actuator (21), i.e. the carriage drive / sliding actuator (6). Furthermore, a target path (W) for a third actuator (23), i.e. the inclination actuator (17), is shown as an example. For reasons of better representation, the courses shown do not exactly correspond to the target spatial positions (Ri) and the setting positions (Pi) derived from them. What is relevant here is to get an impression of the information content of the parts of the movement plan (60).

[0152] The creation of a time scale is explained below. Examples of time scales are given in Figure 9 shown. Figure 11illustrates how the dynamic plan (70) is formed from the movement plan (60) by stretching and / or compressing it using time scaling.

[0153] Time scaling results in a mapping of the process variable (s) to the time variable (t). In other words, the motion plan created for quasi-time is adapted to real time. Physical and / or operational limits are incorporated into this adaptation.

[0154] Figure 9 explains this scaling / adaptation in the form of a time scaling plan (80). There, the time variable (t) is shown on the abscissa axis, and the process variable (41) is shown on the ordinate axis. Figuratively speaking, the processing speed of quasi-time (process variable s) is sometimes accelerated and sometimes decelerated according to the gradient ds / dt in the time scaling plan (80).

[0155] A time scaling (s=s(t)) is defined which determines the course of the process variable (s) as a function of the time variable (t). This is achieved by defining the process variable (s), at least in sections, using a parameterizable, continuous and preferably monotonically increasing function. The value of this function specifies the point in time (value of the time variable (t)) at which an assigned value of the process variable (s) is to be taken into account. The gradient of the time scaling thus specifies the step rate (ds / dt) at which the motion plan should be executed. The higher the step rate, the faster the motion plan is executed - the lower the step rate, the slower the motion plan is executed. A monotonically increasing function for the time scaling has the advantage of being easier for the system planner to understand.However, the methods according to the present disclosure are also applicable to functions that are non-monotonic and, in particular, have a negative gradient in at least one section (82'). This means that at this point, at least part of the dynamics plan is first executed backward and then forward again. This makes it possible, in particular, to provide planning for the movement of actuators with multiple, interlinked dynamics.

[0156] A first possibility of time scaling provides that (at least in sections) a constant scaling (81) is carried out with a time scaling factor (F): s = F ⋅ t

[0157] The time scaling factor (F) can theoretically be chosen as a single value for the entire execution period. Alternatively, and preferably, it is specified separately for specific time periods.

[0158] The time scaling factor (F) can be determined or specified in any way. It is preferably selected based on the physical limits or operating limits that exist for the respective actuator or the movement of the respective process element. These can be, in particular, a physically determined or manufacturer / user-defined maximum actuator speed U°max, a maximum actuator acceleration U°°max, and / or a maximum actuator jerk U°°°max.

[0159] The calculation of the time scaling factor (F) is particularly preferably carried out according to the following function (example of a time scaling factor (F_U) in relation to the first actuator and its target path (U): F U = min U ˙ max dU s ds max U ¨ max d 2 U s ds 2 max U ⃛ max d 3 U s ds 3 max 3

[0160] For the other actuators and target paths (V,W) the calculation is possible in an analogous manner.

[0161] Alternatively, an approximate calculation can be performed, which requires less computational effort.

[0162] Constant scaling (in certain sections) can be advantageous for those parts of the treatment process where relatively small changes in acceleration are expected (for the majority or all actuators), because constant scaling can increase jerk behavior, which, however, does not necessarily represent a disadvantage in the aforementioned parts of the treatment process. The advantage of constant scaling is its comparatively fast computation.

[0163] Another possibility for time scaling is to perform the time scaling (at least in sections) as a variable scaling. A preferred procedure for determining a variable time scaling is based on Figure 10 explained.

[0164] To understand the subsequent processing steps, a mathematical explanation is required. The path velocity s° is defined as the first derivative of the process variable s with respect to time (not the process variable itself!). The path acceleration s°° is defined as the second derivative of the process variable s with respect to time: s ˙ = ds dt , s ¨ = d 2 s dt 2

[0165] It should be noted here that the process variable s is not a geometric reference quantity in space, but a quasi-time. The term "orbital velocity s°" therefore also has no direct connection to a geometry. It simply refers to the gradient ds / dt with which the target values ​​defined with respect to the process variable s are passed through.

[0166] Furthermore, for mathematical transformation, the target path (U,V,W) is defined as a function of the process variable (s), which in turn is a function as a function of the time variable (t): U = U s t

[0167] The target path (U,V,W) is now derived at least twice, preferably three times according to time (not according to the process variable!) and set in relation to the path speed (s°) as well as the path acceleration (s°°) and possibly the path jerk (s°°°): U ˙ = dU s ds s ˙ U ¨ = d 2 U s ds 2 s ˙ 2 + dU s ds s ¨ U ⃛ = d 3 U s ds 3 s ˙ 3 + d 2 U s ds 2 3 s ˙ s ¨ + dU s ds s ⃛

[0168] According to the following example, the variable scaling is determined in a state space over a range of the process variable (s) by integration, based on boundary conditions at the boundaries of the range and preferably taking into account predefined dynamic criteria. The range of the process variable (s) and the predefined criteria can be chosen arbitrarily (jointly or independently of each other).

[0169] For example, the following assumes that the most time-efficient treatment possible is desired. In this case, physical limits or operating limits are considered as the default criteria, which are defined here as a maximum actuator speed (U°max) and preferably a maximum actuator acceleration (U°°max). The described procedure can be applied in any way to other default criteria, and in particular to other physical limits or operating limits. In particular, consideration of the maximum actuator acceleration (U°°max) can be omitted.

[0170] The following examples assume that the maximum actuator acceleration (U°°max) is included in the calculation. This produces a calculation result that can be translated into the actual movement of the actuators, and thus of the entire vehicle treatment system, with a particularly low expected error. If the actuator acceleration (U°°max) is omitted, a dynamic plan may result in which a somewhat larger error occurs when translated into the actual movement of the actuators. However, these errors can be compensated for by the controller to which the dynamic plan is applied.

[0171] On the other hand, in addition to the maximum actuator speed (U°max) and actuator acceleration (U°°max), a maximum actuator jerk can also be defined and considered. The following instructions can be applied accordingly to the consideration of actuator jerk. For the sake of clarity, a mathematical explanation of the two aforementioned variations (no consideration of actuator acceleration / additional consideration of actuator jerk) is omitted below.

[0172] To determine a variable time scaling, a trajectory acceleration s°° is preferably defined as a function that can be integrated at least twice and is parameterized. In the simplest case, this is a constant value or a variable with a range from 0 to s°°max. Further explained below is that the trajectory acceleration can also be calculated as a variable value from the general dynamics equation.

[0173] Alternatively, to determine a variable time scaling, it is possible to define a path velocity s° as a function that is at least once integrable and parameterizable, or to define a path jerk as a function that is at least three times integrable and parameterizable. Here, too, the instructions presented below can be applied accordingly. The following illustration assumes that a path acceleration s°° is defined.

[0174] The path acceleration s°° is integrated twice, taking into account predefined limits for the path speed (s°max, s°min) and preferably predefined limits for the path acceleration (s°°max, s°°min), in order to determine the parameterization and to calculate the process variable (s) as a (mathematical) function of the time variable (t) (result see Figure 9 ).

[0175] Figure 10shows the intermediate result after the first integration. Here, limits for the path velocity s°max, s°min, and the path acceleration (s°°max, s°°min) were determined from the maximum and minimum actuator velocity (U°max, U°min) and the maximum and minimum actuator acceleration (U°°max, U°°min). This is supported below with a mathematical explanation, which, however, represents only one of many possible calculation methods.

[0176] Particularly preferably, the maximum and minimum track speeds s°max, s°min can be calculated according to the following formulas: s ˙ max s = U ˙ max dU s ds , f ü r dU s ds > 0 U ˙ min dU s ds , f ü r dU s ds < 0 mit U ˙ min ≤ U ˙ ≤ U ˙ max s ˙ min s = U ˙ min dU s ds , f ü r dU s ds > 0 U ˙ max dU s ds , f ü r dU s ds < 0 mit U ˙ min ≤ U ˙ ≤ U ˙ max

[0177] If the value of dU(s) / ds becomes zero, the maximum / minimum orbital velocity becomes infinite.

[0178] The maximum and minimum path acceleration s°°max, s°°min can be calculated using the following formulas: s ¨ max s s ˙ = U ¨ max − d 2 U s ds 2 s ˙ 2 dU s ds , f ü r dU s ds > 0 U ¨ min − d 2 U s ds 2 s ˙ 2 dU s ds , f ü r dU s ds < 0 mit U ¨ min ≤ U ¨ ≤ U ¨ max s ¨ min s s ˙ = U ¨ min − d 2 U s ds 2 s ˙ 2 dU s ds , f ü r dU s ds > 0 U ¨ max − d 2 U s ds 2 s ˙ 2 dU s ds , f ü r dU s ds < 0 mit U ¨ min ≤ U ¨ ≤ U ¨ max

[0179] An absolute upper limit in the state space diagram (90) according to Figure 10 is specified for each value of the process variable (s) by a maximum path speed (s°max), which is calculated from the maximum actuator speed U°max according to the above formula. The maximum actuator speed U°max can, for example, be set to a specific value as an operating limit. The hatched area of ​​the state space must therefore not be crossed. In some areas, however, the absolute upper limit could reach infinity (theoretically). It is possible to additionally specify an upper limit (s°max,fix) that must not be exceeded. This is shown in Figure 10 shown as an example.

[0180] A minimum orbital velocity (s°min) is not plotted for the sake of simplicity. It would have a curve that is a mirror image of the maximum orbital velocity (s°max) relative to the abscissa axis (40).

[0181] In a preferred embodiment, the scaling (s=s(t)) is implemented monotonically increasing so that the execution of the vehicle handling does not occur in the reverse order of the motion plan. In such a case, the path velocity (s°) can only assume zero or positive values, so that consideration of the minimum path velocity (s°min) can be omitted.

[0182] However, it is not impossible to run through part of the vehicle treatment first forwards, then backwards, and then forwards again, for example in order to clean a specific part of the vehicle particularly intensively, or to have a specific treatment unit perform a movement that requires at least one actuator in a chained dynamic to perform a temporary backward movement (cf. section 82' in Figure 9 and explanations of movement sequences according to Figure 13 ). Thus, it is entirely possible, and in some applications even sensible or even necessary, to consider both a maximum path speed and a minimum path speed (s°min). The present explanations then apply analogously.

[0183] The maximum and minimum orbital speeds represent only limits that should not be exceeded. It is quite possible to perform a time scaling for all sections based on these maximum values. This can result in a course of the orbital speed s° that Figure 10 essentially follows the course of the sections s°max (variable) and possibly s°max,fix. However, it has been found that a different form of time scaling offers advantages, in particular, that the maximum path acceleration s°°, possibly the maximum path jerk, or other boundary or transition conditions are also taken into account. Various explanations of this are explained below.

[0184] To arrive at a concrete function that can define the variable time scaling (s = s(t)), a two-step approach is proposed. The order of the steps is irrelevant.

[0185] In one stage, a forward integration with two boundary conditions is carried out at the left boundary of the considered area, in Figure 10 i.e., at s=0. The boundary conditions can be chosen arbitrarily. In this example, it is assumed that the process variable (s) and the path velocity (s°) have a given initial value. In the example of Figure 10 Both starting values ​​are zero. The integration therefore begins at s=0 on the left edge. There, the orbital velocity (s°) is defined as zero (s°(0)=0). Any other known starting values ​​could be used.

[0186] In the other stage, a backward integration with two further boundary conditions is carried out at the right boundary of the considered area, in Figure 10i.e., at s=1. These boundary conditions can also be chosen arbitrarily. In this example, it is again assumed that the process variable (s) and the path velocity (s°) have a given final value. In the example of Figure 10 The right edge of the considered area is where the process variable has the value 1 (s=1). The path velocity (s°) is again defined as zero there (s°(1)=0).

[0187] Now, in the first stage (from left to right / forward integration), starting from the left edge and the known starting value located there (here: s°(0)=0), integration is carried out with a specified value of the path acceleration. In this example, this value is the determined maximum value of the path acceleration (s°°max). If the specification criteria are selected differently, a correspondingly different specified value for the acceleration can be used. In particular, a maximum path acceleration could be determined from the specification of a maximum path jerk. Further below, it is explained that a specified value for the acceleration can also vary locally and can be calculated, in particular, from the general dynamic equation for the vehicle treatment system (1).

[0188] The integration is carried out until, at a certain value of the process variable (s), the integral reaches the value of the maximum path speed (s°max, possibly s°max,fix), or until the right edge of the range (s=1) is reached.

[0189] In the second stage (from right to left / backward integration), starting from the right edge and the known starting value located there (here: s°(1)=0), integration is carried out with a further specified value for acceleration in the opposite direction, i.e. from the right edge (s=1) in the opposite direction to the abscissa axis (40) or towards the left edge. In the present example, this specified value is the minimum path acceleration (s°°min), which is calculated from the minimum actuator acceleration U°°min (braking power). Depending on the selection of the specification criteria, a different value can be used here, analogous to the above explanations.

[0190] This backward integration is also carried out until, at a certain value of the process variable (s), the backward integral (92) reaches the value of the maximum path speed (s°max, possibly s°max,fix), or until the left edge of the range (s=0) is reached.

[0191] In Figure 10 The values ​​of the orbital velocity (s°) according to the forward integral (91) and backward integral (92) up to their intersection point (93) are shown with a thick solid line. In the example of Figure 10 the chain (100) of mappings. This chain of mappings is integrated again to obtain the curve s(t).

[0192] The regions of the forward integral (91) and the backward integral (92) lying behind the intersection point (93) can be disregarded.

[0193] The resulting overall course of the variable scaling (82) is shown as an example in Figure 9 shown.

[0194] The variable scaling can comprise at least one section (82') in which a negative gradient (s°<0) exists. At such points, when mapping the process variable (s) to real time (t), a portion of the target values ​​is passed through in the negative direction of the process variable (s) (see explanations above). For ease of understanding, a monotonically increasing course of the variable scaling (82) is assumed below.

[0195] The variable time scaling (81, 82), ie the time-dependent course of the process variable (s=s(t)), is therefore preferably formed from the integral of a chain of mappings that describe the course of the orbital velocity (s°) in a state space over the process variable (s).

[0196] The chain of mappings preferably includes forward integrals (91, 96a) and backward integrals (92, 96b) of the maximum orbital acceleration (s°°max) based on boundary and transition conditions. However, it can also include a constant time scaling factor (F) in sections. This would be Figure 10 a horizontal distance, especially when the upper limit (s°max, fix) is reached.

[0197] Generally speaking, the time-dependent curve of the process variable (s=s(t)) is preferably composed in sections and is formed by an integral of a chain of mappings for the orbital velocity (s°). These mappings comprise the curves of the forward integral (91) and backward integral (92) up to their intersection point (92) or, if the value of the maximum orbital velocity (s°max, possibly s°max,fix) is reached, a replacement curve. The curve of the maximum orbital velocity (s°max, s°max,fix) for the corresponding section of the process variable (s) can often be used as a replacement curve.

[0198] Alternatively, an intermediate intersection point can be defined, allowing the closest equivalent curve to the maximum path speed (s°max) to be maintained without exceeding other physical and operating limits. The determination of such an intermediate intersection point is explained below.

[0199] In the time scaling plan (80) of Figure 9 a (section-wise) constant scaling (81) and a variable scaling (82, 82') are shown as examples.

[0200] The two scaling types (variable / constant) can be used separately or in combination. Furthermore, variable scaling can also be determined by methods other than the integration described above.

[0201] For a first group of time periods (and corresponding sections of the process variable (s)), a first scaling type, for example constant scaling, can be selected, and for a further group of time periods, a further scaling type, for example constant scaling, can be selected.

[0202] Figure 11 explains the creation of a dynamic plan (70) for vehicle treatment. The dynamic plan (70) includes a target trajectory (A,B,C), and preferably further a target speed (A°,B°,C°), and / or a target acceleration (A°°,B°°,C°°) for the majority of the actuators (21,22,23) with respect to a time variable (t).

[0203] It may be sufficient for the dynamic plan (70) to comprise only one target trajectory (A, B, C). The dynamic plan (70) may be present as an ordered data structure and stored in a physical memory. Alternatively, it may be present as a data stream that is gradually generated and processed. Particularly preferably, the dynamic plan is stored as an array, a table, or a similar data structure and covers not only a current point in time in real time, but also a future time period. This applies in particular to a dynamic plan that is stored in the controller or control unit of the treatment system. A dynamic plan can be stored in sections or in its entirety as a digital object. For example, it is possible to have the control method according to the present disclosure executed by a computing device that is temporally and / or spatially separate from the vehicle treatment system.

[0204] The target speed (A°, B°, C°) and / or the target acceleration (A°°, B°°, C°°) can be specified explicitly. Alternatively or additionally, they can be determined by derivation from the target trajectory (A, B, C). This determination can occur at any time. For example, a control system of the treatment plant can determine the target speed (A°, B°, C°) and / or the target acceleration (A°°, B°°, C°°) before transferring it to one or more controllers (32) for the actuators themselves and, in particular, calculate it by derivation (differentiation). Alternatively or additionally, the target speed (A°, B°, C°) and the target acceleration (A°°, B°°, C°°) can be determined by a controller (32) of the actuator to be addressed. A separate controller (32) can be provided for each actuator. Alternatively, one controller (32) may be provided jointly for two or more actuators.

[0205] A preferred embodiment provides for the dynamic plan to be generated completely with the target trajectory, target velocity, and target acceleration, and for the associated target values ​​to be transferred together to a controller for each control point in time. This is advantageous for utilizing the maximum accelerations of the actuators.

[0206] The dynamic plan is generated from the motion plan by locally stretching and / or compressing it using time scaling (82, 82). Stretching and / or compressing generally does not result in a change in the magnitude of the target values ​​of the target path (U, V, W), but merely in a different distribution of the target values ​​of the target path (U, V, W) relative to the time variable (t). Thus, the result of time scaling is generally different from the result that would be achieved by simply accelerating a washing program. In particular, time scaling does not cause overshoot. The dynamic plan can be generated before or during vehicle treatment. It can be generated section by section for a specific range of the process variable (s) / the time variable (t) or overall for an entire treatment process or its cycles.

[0207] In the example of Figure 11 above is the movement plan from Figure 8shown again. The included curves of the target paths, target rates and target rate changes are used as a basis for the stretching / compression according to the variable scaling curve (82) from Figure 9 used.

[0208] In the graphical representation, the scaling means that the abscissa axis (40), which shows the process variable (s) in the motion plan (60), is replaced by the time variable (t) in the dynamic plan.

[0209] The target values ​​for the movement of the actuators (21, 22, 23) are expressed in the dynamic plan based on the target trajectories (A, B, C), which are mapped as a function of the time variable (t). Additional target values ​​are defined for the target velocities (A, B, C) and preferably the target accelerations (A, B, C). Alternatively or additionally, target values ​​for a target jerk could be defined.

[0210] The graphical progression of the target trajectory (A, B, C) bears a clear resemblance to the progression of the corresponding target path (U, V, W), but is locally stretched or compressed along the abscissa axis (40). Where the gradient (ds / dt) of the scaling (81, 82) is high, the motion plan is executed comparatively faster, so that at these points the target trajectory (A, B, C) of an actuator (21, 22, 23) appears compressed compared to the corresponding target path (U, V, W). Where the gradient (ds / dt) of the scaling (81, 82) is low, the motion plan is executed comparatively slower, so that at these points the target trajectory (A, B, C) of an actuator (21, 22, 23) appears stretched compared to the corresponding target path (U, V, W).

[0211] The target accelerations (A°°, B°°, C°°) indicate coincident times (ti) at which the accelerations of the actuators (21, 22, 23) begin and end. For the sake of simplicity, these coincident times (ti) are only given for part of the time axis (t) and purely as examples. The simultaneity of the accelerations for the multiple actuators (21, 22, 23) leads to particularly smooth and supple movement behavior, especially for (all) multidimensional positioning movements of a treatment unit. This simultaneity can result as an indirect consequence of pose planning and the preferably uniform time scaling (80).

[0212] An acceleration phase (ki) lies between each two points in time (ti). It can be seen that the acceleration phases (ki) during the target acceleration (A°°) for the first actuator and the acceleration phases (ki) during the target acceleration (B°°) for the second actuator coincide. They essentially occur simultaneously.

[0213] Furthermore, several consecutive acceleration phases (ki) during the target acceleration (A°°) of the first actuator form a first chain (m). Associated and also consecutive acceleration phases (ki) during the target acceleration (B°°) of the second actuator also form a chain (n). At the macro level, the two chains form a chain of partial movements, which together lead to a fluid, multidimensional positioning movement of the treatment unit, which is moved jointly by the first actuator and the second actuator.

[0214] A particular advantage of the dynamic plan according to the present disclosure is that it can also be generated as a common dynamic plan for a plurality of actuators subject to single or multiple interlinked kinematics. In this case, too, the dynamic plan can define mutually synchronized accelerations (A, B, C) for these actuators during the duration of the treatment section for each phase with a multidimensional actuating movement of the treatment unit, so that consistent acceleration phases (ki) result for the multiple actuators (21, 22, 23) during the execution of the multidimensional actuating movement, even if they contribute to the movement of multiple treatment units.

[0215] In the treatment system (1) shown in the figures, a dynamic plan (70) can be generated which specifies target values ​​in a common ordered data structure at least for the movement of the portal drive (5), the movement of the lifting drive (14) of the roof brush (12) and the movements of the sliding actuators (9), tilting actuators (16) and bending actuators (18) of the side brushes (6, 7), by means of which partially or fully synchronized accelerations of these actuators are achieved, namely during the overlapping or simultaneous intervention of the five treatment units: Roof brush (12) Upper brush segment (6a) of the first side brush (6) Lower brush segment (6b) of the first side brush (6) Upper brush segment (7a) of the second side brush (7) Lower brush segment (7b) of the second side brush (7).

[0216] The result according to the example of Figure 11can arise even if only one of the movement variables, target trajectory, target speed, target acceleration or target jerk, is defined, since conversion between these variables can be carried out by differentiation / integration. It is therefore possible, for example, to actually define only one or two of these movement variables in the dynamics plan (70) or to transfer them to a controller (32). A particularly precise implementation of the desired multidimensional actuating movements and / or a maximum utilization of the possible power are promoted if at least two and preferably three of the movement variables are defined in the dynamics plan and transferred to a controller (32), in particular a state controller.

[0217] Separating the planning process into a motion plan and a dynamic plan offers the following additional advantages: In the motion plan, suitable spatial positions of the corresponding target points (Z1, Z2, Z3) can be defined in a specific order for all relevant states that the treatment units are to assume during vehicle treatment. From these, the corresponding actuating positions (Pi) of the actuators (21, 22, 23) can be derived. This planning step eliminates the need to consider the physical or operating limits of the actuators, enabling a targeted approach with a quickly executable and robust processing procedure. This processing procedure can be carried out manually or, preferably, semi-automatically or fully automatically.

[0218] From the positioning positions (Pi), target paths can be created in the motion plan for the majority of actuators (21, 22, 23) using simple means (function selection, parameterization, solution of simple systems of equations with boundary conditions). This also does not require consideration of the physical or operating limits of the actuators. However, any kinematic couplings between the actuators can be incorporated into this planning step, so that the motion plans can already be partially or completely synchronized with each other (with respect to a process variable s). A kinematic coupling between two actuators exists when they participate in a chained kinematics system for a treatment unit. This allows for consistent planning, at least for a simply chained kinematics system.

[0219] As part of the synchronization, additional positioning positions (Pi*) can be inserted as supplementary support points, particularly in a target path (U) for a first actuator (21), at values ​​of the process variable (s) where a specific target spatial position of another actuator (22, 23) is defined. This is particularly useful when the first actuator and the other actuator are kinematically coupled.

[0220] Kinematically coupled movements for actuators (21, 22, 23) can occur for any reason and in a wide variety of configurations. Kinematically coupled movements exist in particular for a group of actuators (21, 22, 23) when the actuators (21, 22, 23) are provided for the actuation of two or more treatment units (6, 7, 12), AND there is at least one actuator (21) in the group, the actuation of which causes at least two partial movements, namely a first partial movement for a first treatment unit (7) and at least one further partial movement for a second treatment unit (12).

[0221] In Figure 5 As an example, some additional setting positions (Pi*) of the portal drive / guide actuator (5) are shown, which are inserted as support points, for example because a target spatial position for the target point (Z3) of the roof brush (12) or for the target point (Z1) of the upper brush section (6a) would be useful there.

[0222] The target values ​​(62) derived from these additional support points (Pi*) in the target path (U) according to Figure 7 are intended to be achieved with a certain tolerance.

[0223] It is thus possible to define target spatial positions (R, L) for at least one treatment unit, and in particular for two or more treatment units (6, 7, 12) or their target points, only for specific regions in the treatment process, i.e., a minimum definition set of target spatial positions, preferably depending on the contour recognition data (50). These can, in particular, be regions in which there is a significant change in the vehicle contour or concave / convex curvatures, as well as at the beginning and end of essentially flat contour regions. The number of target spatial positions to be planned can thus be reduced to a minimum for each treatment unit.

[0224] For the other target points, corresponding additional spatial positions and associated additional target positions (Pi*) can be determined, from which corresponding additional support points in the target paths of the associated actuators can then be derived.

[0225] Particularly preferably, for each target spatial position in the minimum definition set, an associated value of the process variable (s) is determined (or set), and a pose (body posture) is determined or defined for several or all treatment units (6, 7, 12). For the pose, in particular, an overall state of all (active) actuators can be determined or defined in order to provide associated target values ​​(61, 62) in the target paths. In this way, several or preferably all target paths are synchronized with one another with respect to the process variable (s). Uniform sections then lie between the individual values ​​of the process variable (s) for which target values ​​are defined.

[0226] Alternatively, the target paths can also be planned independently of each other. This can be particularly advantageous when there is no kinematic coupling between two or more actuators. In this case, fewer target positions (Pi) tend to be available, allowing processing to take place with lower computing power requirements.

[0227] The movement plan (60) preferably comprises an exact kinematic definition of the movements of the actuators (21, 22, 23) for carrying out the vehicle treatment and is suitable for being run through at an adjustable processing speed.

[0228] Time scaling determines the processing speed so that the actuator movements can be performed according to freely selectable criteria, for example, by utilizing maximum acceleration on the one hand and minimizing energy consumption on the other. Furthermore, time scaling ensures that the actuator movements do not exceed physical limits or specified operating limits without the need to specify a phase restriction. Particularly when the kinematics are linked two or more times, time scaling can resolve inconsistencies between the individual movements of these actuators and the resulting poses.

[0229] The dynamic plan generated based on the time scaling thus allows for the control of all actuators (considered in the planning) with the ideal control parameters (current target position according to the target trajectory, current target speed, current target acceleration) according to the specified criteria for every moment of the vehicle treatment. This is possible, in particular, without the need for phase separation or separation of the control / regulation for individual actuators. Rather, the entire treatment system (1) with any number of treatment units and actuators can be operated with holistic control or regulation of all partial movements, whereby all of the boundary conditions are observed at all times. The entire treatment process can also be fully executed and visualized in a simulation environment.

[0230] Furthermore, the vehicle treatment can be fully anticipated for all parts of the vehicle (3) for which contour detection (51) has already taken place, both with regard to the current states of all actuators and treatment units, as well as with regard to the treatment duration. The states can be defined according to the dynamic plan (70), at least with regard to the target trajectory (A, B, C). Preferably, the target speed (A, B, C) and / or the target acceleration (A, B, C) are also specified.

[0231] The drawings show graphical representations of contour detection data (51), sequence plan (50), movement plan (60), dynamic plan (70) and time scaling plan (80) which are intended to explain the sequence of the control method according to the disclosure.

[0232] Alternatively or in addition to such graphical representations, these contents may be presented in other forms, particularly in an ordered data structure. These may include, in particular, mathematically defined functions, approximation curves of such functions, arrays, tables, and vectors.

[0233] A particularly preferred embodiment provides that at least the dynamic plan (70) is in such a form that for each time interval of the execution of the control method, for example for each time step of the control time, particularly preferably for each millisecond, there is a value for the target trajectory (position or angle), and preferably also for the target speed (translational or rotational) and / or the target acceleration, wherein these values ​​are transferred to a controller for the respective actuator (21, 22, 23). The dynamic plan comprises, in a common data structure, target values ​​for several actuators, in particular for a group of actuators that participate in the movement of a treatment unit. Further preferably, the dynamic plan comprises, in a common data structure, target values ​​for several actuators that are subject to singly or multiply linked kinematics for the movement of at least two treatment units.The actuator movement is controlled according to the dynamic plan and converted into an actual movement, so that the actual trajectory corresponds to or approximates the desired trajectory. The controller can have any design. For example, it can be a cascade controller. A state controller is particularly preferred.

[0234] The controller (32), in particular the cascade controller or state controller, is thus configured to control the actuator (21, 22, 23) such that the actual trajectory, and preferably also the actual speed and / or the actual acceleration, follow the target values. The controller (32), in particular the cascade controller or state controller, can compensate for any deviations between the actual values ​​and the target values ​​by appropriate control interventions.

[0235] Some more complex examples are explained below, which offer additional advantages.

[0236] Figure 14explains a possibility for generating an intersection intermediate section (96) as a section of a variable time scale (82).

[0237] In this example, the same basic assumptions are made as above for Figure 10 However, a lower maximum path speed curve (s°max) is calculated here. This can occur, for example, if the operating limits for the movement of an actuator are reduced. Such a reduction in the operating limits can have any cause, for example, the selection of a wash program with a longer exposure time or the system entering a safety mode.

[0238] In the example of Figure 14 meet (in contrast to the example of Figure 10) the curves of the forward integral (91) and the backward integral (92) do not intersect. Rather, the forward integral (91) intersects the curve of the maximum orbital velocity (s°max) at an intersection point (94). And the backward integral (92) also intersects the curve of the maximum orbital velocity (s°max) at an intersection point (95). By definition, the maximum orbital velocity (s°max) must not be exceeded.

[0239] The next step is to find the nearest equivalent curve with which the greatest possible path acceleration (s°°) and thus also the greatest possible path speed (s°) can be achieved while observing the boundary and transition conditions.

[0240] A procedure for determining the nearest replacement curve that has proven advantageous includes the following steps: Determine the tangential point (97) at which, when integrating with the maximum path acceleration (s°°max), the course of the maximum path velocity (s°max) is not exceeded but only tangentially touched, and determine its value (sT) of the process variable, as well as the maximum path velocity there (s°(sT)). From this tangential point (97): Form another forward integral (96a) and another backward integral (97b), based on the value of the maximum path velocity (s°(sT)) there at maximum path acceleration (s°°max). The forward integration and the backward integration are repeated until a new intersection point (98, 99) is found between the further forward integral / backward integral (96a, 96b) with a course of the previous forward integrals / backward integrals (91, 92) at a specific value of the process variable (s).

[0241] If, in the final step of this procedure, NO further intersection point (98, 99) with a curve of the previous forward integrals / backward integrals (91, 92) is found, but instead an intersection point with the curve of the maximum path velocity (s°max) is reached, the procedure can be iterated to find another subordinate intermediate intersection point (not shown). The procedure can be repeated until a suitable section (91, 92, 96a, 96b) is found for all values ​​of the process variable (s).

[0242] The variable time scaling method can also consider multiple actuators. For this purpose, the maximum path velocity (s°max) is determined for each actuator (to be considered) and used for the calculation. Likewise, a maximum path acceleration (s°°max) is calculated for each actuator. Of these values, the lowest calculated maximum path velocity (s°max) and the lowest calculated path acceleration (s°°max) are then used, and the time scaling of all actuators (to be considered) is determined based on these lowest calculated values.

[0243] In the above examples, it was assumed that a maximum path speed (s°°max) was calculated based on essentially static values ​​for a maximum actuator speed (U°max) and / or a maximum actuator acceleration (U°°max). However, it turns out that a maximum actuator speed (U°max) and a maximum actuator velocity (U°°max) are subject to dynamic influences and depend on the current spatial position of a treatment unit and / or can influence each other. Therefore, a static specification of these values ​​is generally very conservative, so that the true performance potential of the actuators is not utilized.

[0244] According to a preferred embodiment, a dynamic model of the treatment system or the treatment units can be included in the calculation in order to variably implement the physical limits or operating limits (at least in sections) depending on a state of the vehicle treatment system. The state can be determined, in particular, by a spatial position of one or more treatment units (angle-dependent components of gravitational forces, Coriolis forces); and / or a current speed or actuation rate of an actuator (elasticity of acceleration capability). Particularly preferably, a maximum dynamic actuator acceleration can be determined as a variable.

[0245] For this purpose, the dynamics of the treatment plant can be modeled in a system of equations, preferably in a vector definition with a mass inertia matrix (M), a vector (Vv) for the speed-dependent Coriolis and centripetal forces, and a vector (Vg) with the gravitational forces. From this, the instantaneous actuator force / torque (I) can be calculated according to the general dynamics equation for the respective values ​​of the target path (U) and its derivatives (U°, U°°). An example with a suitable vector definition is: I = M U ⋅ U ¨ + V v U U ˙ = V g U

[0246] This equation can be rearranged to U°°(U,U°). For each value of the process variable (s) / time variable (t), the dynamic maximum actuator accelerations can be determined based on the known pose and the known control values ​​(state values) for the actuators. These values ​​can be used instead of U°°max / U°°min in the above formulas to calculate the maximum path velocity (s°max) or the maximum path acceleration (s°°max). The limit U°°max / U°°min is then no longer a fixed value, but can be calculated with its true dynamic maximum value depending on the current position and orientation of the respective treatment unit as well as the state of the assigned actuators.

[0247] This variable calculation of the dynamic maximum actuator acceleration often allows the target and actual speeds to be increased significantly, especially during periods / time phases in which no or only one treatment unit is active. This minimizes unnecessary "dead times" in the treatment process, enabling rapid vehicle treatment without compromising quality. This, in turn, increases the possible number of vehicle treatments per operating phase and thus increases profitability.

[0248] Figure 12 and 13illustrate a further advantage resulting from the present control method. As already explained above, phase separation is no longer required for movement planning. Thus, for example, forward and backward movements of the gantry (4) can be combined as desired with the movements of the side brushes (6, 7) and the roof brush (12).

[0249] Figure 12 shows diagrams that are similar to the illustrations in Figure 5 are similar. In Figure 5 The x-coordinate is shown on the abscissa axis (40). Figure 12 and 13 The process variable (s) is shown on the abscissa axis (40). Figure 12 on the ordinate axes, not the y and z coordinates are plotted, but corresponding control values ​​for the spatial positions (L,R,R') which are to be assumed by the assigned actuators (21,22,23) in a desired path.

[0250] In order to still provide a spatial reference for understanding the processes, the contour detection data (51) from Figure 5 shown again. However, these are (compared to Figure 4 ) are offset from each other in the direction of the abscissa axis (40). Figure 12 and 13 the states of the treatment units (6,7,12), the target spatial positions (R,R',L) of the reference points (Z1,Z3) and the X-position of the portal (4) are related to the same process variable (s) and are thus shown synchronized with each other.

[0251] Since the target points (Z1, Z3) of the side brushes (6, 7) are offset from the roof brush (12) by a certain distance (Dx) in the x-direction, for example, when s=s*, the roof brush (12) is currently in contact with a front contour of the vehicle (3), while the side brushes (6, 7) are already in contact with a side contour of the vehicle (3). The curves of the contour detection data (51) shown in the first and second diagrams are therefore offset from one another by this distance (Dx).

[0252] The front contour of the vehicle (3) has a concave shape in the lower area and a recessed area. Such body shapes pose difficulties for cleaning with existing control methods because, when a phase restriction applies, the gantry is only allowed to move monotonously in one direction during a drive cycle—in this case, from front to back. This means that the roof brush (12) cannot be advanced into the concave area with existing control methods. Accordingly, such areas have so far been inadequately cleaned.

[0253] With the present control method, phase separation is no longer required. It is possible for a treatment unit or its target point to penetrate into concave contours of a vehicle (3) to be treated. This penetration can occur, in particular, into a concave front contour beneath an alcove and / or into a concave front contour or a concave rear contour beneath a bumper, and / or into a concave contour of a roof structure that protrudes over the windshield or rear window of the vehicle.

[0254] A preferred embodiment provides that, for the creation of a sequence plan (50) for a plurality of treatment units (6, 7, 12), target points (Z1, Z2, Z3) are provided, the current spatial positions (R, L) of which can be defined or are defined in coordinates (x, y, z), and wherein spatially coordinated target spatial positions (Ri, Li) of these target points (Z1, Z2) are planned as a target pose sequence (K). This target pose sequence illustrates Figure 12 .

[0255] It now includes areas in which the portal moves back and forth, ie in the positive X direction, then in the negative X direction and then again in the positive X direction. Accordingly, the bottom diagram in Figure 12 Sections in which the X-position of the portal and thus the setpoint for the first actuator (21) / guide actuator (5) first increases, then falls again and increases again.

[0256] In the first and second diagrams, the target spatial positions (Ri,Li) of the roof brush (6) and the side brush (12) are shown with dashed lines, at which the direction of movement of the portal (5) is reversed.

[0257] An initial back-and-forth movement of the gantry is planned, while the two side brushes (6, 7) perform a so-called over-centre wash. During this phase of the treatment, the roof brush is not yet in contact with the vehicle (3).

[0258] In previous control systems for vehicle treatment systems, an over-the-center wash is usually performed by a separate subroutine, meaning that this part of the treatment must be separated out as a separate time phase. From the beginning to the end of the over-the-center wash, the gantry in previous treatment systems usually has to be stationary. With the present control method, such a separation is no longer necessary. Instead, acceleration phases when approaching the front of a vehicle can transition directly and smoothly into a surface treatment of the vehicle front. The gantry can also be moved during the over-the-center wash, and at the end of the over-the-center wash, a direct and smooth transition to the surface treatment of the side contour can take place. This is explained below.

[0259] During the over-center wash, the gantry (4) can be moved back and forth slightly to allow the side wash brush (6) to cross the convexly curved front contour of the vehicle at an ideal washing distance. If necessary, the brush sections (6a, 6b) can also be inclined to enable the best possible adaptation to the vehicle contour.

[0260] The target spatial positions (Ri) to be assumed by the side brush (6) are planned according to their sequence (i) and planned along the process variable (s) with target positions for the guide actuator (5) and the push actuator (8) in such a way that the planned target spatial positions are gradually assumed along the concave contour.

[0261] A further back-and-forth movement of the portal is intended for cleaning the concave contour in the lower area of ​​the front apron (see upper diagram). The planning steps to be carried out are additionally described in Figure 13 explained. Figure 13 shows a sequence of states that represent the turning points of the Figure 12 illustrate the back and forth movement of the portal (4).

[0262] We assume that initially a movement of the portal (4) in the x-direction is provided so that the roof brush (12) is in a first state (see upper diagram in Figure 13 ) reaches the target spatial position (L3), i.e., comes into contact with the concave section of the front apron. Here, the gantry must reverse direction.

[0263] In the state representation shown on the right, the movement of the roof brush (12) from within the concave section of the front contour, i.e. starting from the desired spatial position (L3'), to a subsequent desired spatial position (L4) in contact with the vehicle nose is shown.

[0264] Since the diagrams are structured according to the common process variable (s), and the portal (5) moves in the opposite x-direction, the relevant section of the front contour for this movement component is shown mirrored around the vertical axis. In the lower diagram, the portal's reverse movement is evident from the falling course of the x-coordinate of the portal movement.

[0265] The middle diagram again shows which additional target spatial positions (R3, R3', R4) are to be provided for the side brush (6) for these conditions. They result from the plan specification to move the side brush along the side contour of the vehicle.

[0266] The state diagram shown further to the right shows the subsequent movement of the roof brush from the front of the vehicle (target position L4') along the height contour of the hood. From the state with the target position (L4') to the state with the target position (L5), the gantry moves again in the x-direction, in the example shown by the distance by which it was previously moved in the opposite x-direction.

[0267] The middle diagram shows which additional target spatial positions (R4', R5) are to be provided for the side brush (6) for these states. The target spatial position (R5) is identical to the target spatial position (R3), except that it is planned again for a later value of the process variable (s).

[0268] The subsequent process is analogous to the explanations above.

[0269] From the example of Figure 12 , 13It is evident that the present control method overcomes significant restrictions that previously applied to the movement of the treatment units (6, 7, 9). This makes it possible, in principle, to treat even those vehicles with an automated procedure that previously had to be rejected, such as vehicles with roof racks and vehicles with alcoves.

[0270] The control method also makes it possible to plan the movements of treatment units with a very precisely defined contour tracking, whose actuators have a multiply linked kinematics. In the example of Figures 1 to 3 the spatial position of the target point (Z2) of the lower brush section (6b) depends on the setting positions the guide actuator (5) (portal movement), the push actuator (8) (slide movement), the tilt actuator (16) (tilt of the upper brush section 6a around tilt joint 15), and the bending actuator (18) (pivoting the lower brush section 6b around bending joint 15).

[0271] Previous control methods were unable to provide fluid and smooth movement for such kinematics. However, by separating the planning into the creation of a sequence plan and a motion plan, this is possible. This is especially true if the sequence plan is based on a series of planned poses for all treatment units (6a, 6b, 7, 12) actively participating in a treatment cycle, and a corresponding derivation of positioning positions for the involved actuators (5, 8, 16, 18).

[0272] The control method disclosed herein can be implemented for any number of treatment units and any structures of support members and actuators. These include structures with one or more translational actuators, and / or one or more rotary actuators.

[0273] Where method steps are described within the scope of the present disclosure that are carried out, this also means at least an execution on a data processing device. The data processing device is designed to carry out the corresponding method steps. This can be a specific data processing architecture, such as a customized electronic circuit or an ASIC (application-specific integrated circuit). Alternatively or additionally, it can be a standardized data processing architecture, such as a programmable controller or a computer. A software product can be provided which contains instructions that, when executed on a data processing device, carry out the steps of the method.The software product may be stored on a physical data carrier and / or sent or received via an electronic interface (software streaming).

[0274] Modifications of the invention are possible in various ways. In particular, the features shown, described, or claimed for the respective embodiments can be combined, replaced, supplemented, or omitted in any way.

[0275] The planning method and the control method were described in the above explanation of the exemplary embodiments for a treatment process that extends over the entire length of the vehicle. The invention is not limited to such treatment processes over the entire length of the vehicle. In particular, it is not necessary to fully perform the steps of the control method and / or the planning method before treatment of the vehicle begins.

[0276] The steps of the control method and the sequence method can each be carried out in sections, with the sequence plan and / or the motion plan and / or the dynamics plan being generated gradually. For example, for a first execution section, only a partial data set with contour detection data may be available, which only covers a portion of the vehicle length, in particular only approximately 30% or 50% of the vehicle length. Based on this partial data set, a partial sequence plan can be generated. Alternatively, for a first execution section, only a partial data set with contour detection data can be read in and / or a partial sequence plan can be read in.

[0277] Accordingly, in a first execution section of the control method, a partial motion plan and a partial dynamic plan can be generated.

[0278] For the planning and / or execution of a complete vehicle treatment, any number of execution steps can be executed sequentially or in a temporally overlapping manner. The resulting sub-plans (sub-sequence plans, sub-motion plans, sub-dynamic plans) can be temporally and / or spatially adjacent to one another and / or overlap with one another.

[0279] A particularly preferred embodiment provides for a minimum detection range to be defined as a linear measurement along the relative movement of the vehicle and the treatment zone. The minimum detection method can be defined arbitrarily, for example, as the maximum braking distance within which the relative movement between a vehicle and a main support (gantry) of one or more treatment units can be decelerated from a maximum speed to zero. Alternatively, the minimum detection range can be defined as a maximum braking distance plus a safety margin.

[0280] The planning and / or execution of a complete vehicle treatment can be divided into several execution phases, each covering such a minimum coverage area. This allows for an even faster initiation of the treatment of a vehicle.

[0281] If the treatment system is designed as a gantry car wash, the front section of the vehicle can be detected by sensors as it enters the treatment zone, for example. The contour can be detected, for example, by sensors arranged on a movable gantry. The gantry can carry one or more of the treatment units. Based on the partial detection of the vehicle contour, a partial sequence plan, a partial movement plan and a partial dynamic plan can be generated during entry or immediately after the designated parking position has been reached, with which a first section of the vehicle treatment can be carried out immediately. During this first section of the vehicle treatment, the sensors can move relative to the stationary vehicle, whereby another section of the vehicle is detected by sensors, in particular a central section.Based on this additional contour detection data, a subsequent partial sequence plan, a partial motion plan, and a partial dynamic plan can be generated accordingly. After the front section has been treated, the vehicle treatment can continue with these subsequent partial plans. If the vehicle treatment continues with these additional partial plans, the sensors can be moved again relative to the stationary vehicle, and a further or final section of the vehicle (rear section) can now be detected by sensors. A further partial plan can then be generated for this rear section in a third execution phase, on the basis of which the treatment is completed.

[0282] During a (partial) treatment, any changes in the contour acquisition data that only become known with a delay can be translated into a change or adjustment of the target values ​​of a currently applied partial plan or a subsequent partial plan, so that overall a uniform sequence of target values ​​is created.

[0283] If the control procedure is carried out with respect to only one treatment unit, the sequence plan preferably comprises setting positions for at least a majority of those actuators involved in the movement of the one treatment unit; AND / OR the movement plan preferably comprises a target path, a target setting rate, and a target setting rate change for at least a majority of those actuators involved in the movement of the one treatment unit; AND / OR the dynamics plan preferably comprises at least one target trajectory for at least a majority of those actuators involved in the movement of the one treatment unit.

[0284] In particular: If the control procedure is carried out with respect to only one treatment unit, the sequence plan preferably comprises setting positions for all those actuators that are involved in the movement of the one treatment unit; AND / OR the movement plan preferably comprises a target path, a target setting rate and a target setting rate change for all those actuators that are involved in the movement of the one treatment unit; AND / OR the dynamic plan preferably comprises at least one target trajectory for all those actuators that are involved in the movement of the one treatment unit.

[0285] If the control method is carried out with respect to two or more treatment units, the sequence plan, the movement plan and the dynamic plan preferably comprise corresponding mappings (target values) for further actuators involved in the movement of the two or more treatment units, in particular for all those actuators involved in the movement of the two or more treatment units.

[0286] According to a preferred embodiment, the steps of the planning process that generates the schedule are executed partially or fully automatically and as part of the control process. Contour detection data about the vehicle to be treated can serve as the input for the planning process. This contour detection data can originate from any data source. For example, the vehicle treatment system can include its own contour detection sensor system or be connected to external sensors. Alternatively, contour detection data generated in advance can be read from a data storage device.

[0287] In the vehicle treatment system according to the second aspect of the disclosure, a target point is preferably defined for the at least one treatment unit, and the target values ​​of the dynamic plan (70) are defined for the associated group of actuators such that the target point of the treatment unit follows a path which, during a surface treatment, maintains a fixed geometric reference to a contour of the vehicle to be treated, in particular a fixed vertical distance from the vehicle surface.

[0288] Furthermore, the vehicle treatment system is preferably designed such that a multidimensional adjusting movement of the treatment unit takes place during a treatment section with a single or multi-travel reversing direction of movement of a portal of the vehicle treatment system.

[0289] Furthermore, the vehicle treatment system is preferably designed such that a multidimensional adjustment movement of the treatment unit during a treatment section comprises a temporary immersion of the target point into a concave front contour or rear contour of the vehicle to be treated, in particular into a concave contour below an alcove, OR into a concave contour below a bumper, OR into a concave contour below a roof structure that projects over the windscreen or rear window of the vehicle.

[0290] Below, advantageous sub-aspects of the present disclosure that can be combined with the disclosed methods and devices are briefly summarized. These sub-aspects can also be used in any desired sub-combination.

[0291] The control method according to the present disclosure preferably comprises the following step: creating the movement plan (60) with a chain of sectionally defined functions (U_left, U_right) between each two consecutive actuating positions (Pi, Qi), wherein the parameters of these functions (U_left, U_right) are determined by means of boundary and transition conditions at the actuating positions (Pi, Qi). The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0292] The control method according to the present disclosure preferably comprises the following step: A target path (U,V,W) is / is formed from a chain of sectionally defined and parameterizable functions (U_left, U_right), which are doubly or triply differentiable and in particular comprise: Third degree polynomials, AND / OR fifth degree polynomials, AND / OR seventh degree polynomials, AND / OR straight line segments with parabolic ends;

[0293] The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0294] The control method according to the present disclosure preferably comprises the following step: For defining the time scaling (s=s(t)), a time scaling plan (80) is created, which defines the process variable (s) as a function of the time variable (t). The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0295] The control method according to the present disclosure preferably comprises the following step: A time-dependent curve of the process variable (s=s(t)) is / is formed section by section based on an integration of the path speed (s°) and / or the path acceleration (s°°); and the path speed (s°) and the path acceleration (s°°) are / are calculated based on known or physical limits and / or predetermined operating limits of the associated actuator (21, 22, 23) or the treatment system (1) or the treatment units (6, 7, 12). The physical limits and / or operating limits can in particular include: a maximum acceleration capacity (U°°max) of an actuator (21,22,23), AND / OR a maximum speed (U°max) of an actuator (21,22,23);

[0296] The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0297] The control method according to the present disclosure preferably comprises the following step: For a variable scaling (82), the time-dependent course of the process variable (s=s(t)) is / is formed from the integral of a chain (100) of mappings (91, 92), wherein the mappings (91, 92) describe the course of the path velocity (s°) in a state space and wherein the chain (100) of mappings comprises forward integrals (91, 96a) and backward integrals (92, 96b) of the maximum path acceleration (s°°max) starting from boundary and transition conditions at the range boundaries; Furthermore, the following can preferably apply: the chain (100) of mappings additionally comprises, if the forward integral (91) and backward integral (92) do not directly intersect, a substitute curve, wherein the substitute curve in particular is the course of the maximum path speed (s°max) for the corresponding section of the process variable (s), OR is formed by an intersection intermediate section (96) with which the closest equivalent curve to the maximum path speed (s°max) can be maintained without exceeding the other physical limits and operating limits;

[0298] The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0299] The control method according to the present disclosure preferably comprises the following step: For variable scaling, a maximum path acceleration (s°°max) and a maximum path speed (s°max) are calculated based on known physical limits and / or predefined operating limits of the associated actuator (21, 22, 23) or the treatment system (1) or the treatment units (6, 7, 12). The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step.

[0300] The control method according to the present disclosure preferably comprises the following step: At least one rule set and / or one set of function templates is available for specific parts of the treatment process, and the rule set or set of function templates provides certain characteristic actuating positions according to type and sequence for individual or all participating actuators (21, 22, 23), and between these provided actuating positions, suitable functions and / or associated boundary or transition conditions are or will be predefined; Furthermore, it is preferred that for the creation of a movement plan (60), a rule or a function template is selected based on the contour detection (31) and parameterized for the specific treatment process.The vehicle treatment system and / or the software product and / or the control unit according to the present disclosure preferably have a corresponding design and are suitable for carrying out this step. LIST OF REFERENCE SYMBOLS

[0301] 1Vehicle treatment system 2Treatment zone 3Vehicle 4Gantry / support member 5Gantry drive / guide actuator 6Side washing brush (first) / treatment unit 6aFirst brush segment (upper) / treatment unit 6bSecond brush segment (lower) / treatment unit 6cBrush rod / support member 7Side washing brush (second) / treatment unit 7aBrush segment (upper) / treatment unit 7bBrush segment (lower) / treatment unit 7cBrush rod / support member 8First carriage 9Sliding actuator 10Second carriage 11Second sliding actuator 12Roof brush / treatment unit 13Lift guide 14Lift drive / actuator 15Tilt joint 16Tilt actuator,Second actuator 17Articulated joint 18Articulated actuator 21First actuator 22Second actuator 23Third actuator 30Control unit / controller 31Contour detection / contour detection device with sensors 32Controller / state controller 40Abscissa axis 41Ordinate axis 50Sequence plan 51Contour detection data 52Reference point 53Contour line 60Motion plan (related to sequence variable s) 61Setpoint for actuator in target path, to be achieved exactly 62Setpoint for actuator in target path, with tolerance 70Dynamic plan (related to time variable t) 80Time scaling plan (s = s(t)) 81Constant scaling (constant in sections) 82Variable scaling 82Sections of the variable scaling with negative gradient 90State space diagram 91Forward integral 92Backward integral 93Intersection of forward and backward integral 94Intersection of forward integral with maximum orbital velocity (s°max) 95Intersection of backward integral with maximum orbital velocity (s°max) 96Intersection intermediate section 96aIntersection intermediate section,First part; Forward for nearest equivalent curve 96bIntersection-intermediate section, Second part; Backward integral for nearest equivalent curve 97Tangential point 98Intersection 99Intersection 100Chain of mappings aInclination angle iSequence / counting variable kiAcceleration phase mChain of acceleration phases nChain of acceleration phases sProcess variable,s(t) s°Path speed s°°Path acceleration tTime variable tiCorresponding times for the start and end of an acceleration xx-coordinate - Longitudinal direction of the vehicle treatment system yy-coordinate - Transverse direction of the vehicle treatment system zz-coordinate - Vertical axis ATarget trajectory of the first actuator (related to time variable) A°Target speed of the first actuator A°°Target acceleration of the first actuator BTarget trajectory of the second actuator B°Target speed of the second actuator B°°Target acceleration of the second actuator CStarget trajectory of the third actuator C°Target speed of the third actuator C°°Target acceleration of the third actuator Dxx-Distance of the target points Z1, Z3 (variable if necessary) KTarget pose sequence = aggregation of coordinated target spatial positions of several treatment units based on matching states LSpatial position (position and / or orientation) of the roof brush PiPositioning positions of the first actuator (derived from the target spatial position of the corresponding target point) Pi*Additional setting position (as a support point,because target spatial position for kinematically coupled actuator is present) QiPositioning positions in example RSpatial position (position) of target point Z1 R'Spatial position (orientation) of target point Z1 RiTarget spatial position of target point Z1 USoll path first actuator (related to process variable s) U'Target actuating rate first actuator U"Target actuating rate change first actuator U_left, sectionally defined function for a U_rightTarget path in a motion plan U°maxmaximum speed of an actuator (physically determined or specified from operating limit) U°°maxmaximum acceleration capacity of an actuator (physically determined or specified from operating limit) VTarget path second actuator V'Target actuating rate second actuator V"Target actuating rate change second actuator WTarget path third actuator W'Target actuating rate third actuator W"Target actuating rate change third actuator X(s)Gantry position along x-coordinate (relative to process variable) Z1First target point Z2Second target point Z3Third target point,

Claims

1. Control method for a vehicle treatment system (1) having one, two or more treatment units (6, 7, 12) which can be moved by controllable actuators (21, 22, 23) relative to a vehicle (3) in order to carry out a surface treatment of the vehicle (3), characterized in that the control method comprises the following steps: - creating or reading in a schedule (50) for a vehicle treatment, wherein the schedule (50) has adjustment positions (Pi) for a plurality of actuators (21,22,23) as target values in a predefined sequence (i); - creating a movement plan (60) for the vehicle treatment, wherein the movement plan comprises a target path (U,V,W), a target adjustment rate (U',V',W') and preferably a target adjustment rate change (U",V",W") for the plurality of actuators (21,22,23), specifically as functions with respect to a flow variable (s); - creating a dynamic plan (70) for the vehicle treatment, wherein the dynamic plan comprises target values for the movements of the plurality of actuators in relation to a time variable (t); - and wherein the dynamic plan is generated from the movement plan by local stretching and / or compression by means of time scaling (81,82) (s=s(t)).

2. Control method according to Claim 1, wherein the target values of the dynamic plan are present in the form of at least one of the following movement variables: - target trajectory (A,B,C) for the plurality of actuators in relation to a time variable (t), AND / OR - target speed (A°,B°,C°) for the plurality of actuators in relation to the time variable (t), AND / OR - target acceleration (A°°,B°°,C°°) for the plurality of actuators in relation to the time variable (t), AND / OR - target jerk for the plurality of actuators in relation to the time variable (t).

3. Control method according to one of the preceding claims, further comprising the step of: - actuating the actuators (21, 22, 23) while carrying out the vehicle treatment with a regulating system, in particular a state regulating system, based on the target values of the dynamic plan (70).

4. Control method according to one of the preceding claims, wherein the following steps are carried out in order to create a schedule (50): - for at least one treatment unit (6,7,12): defining a target point (Z1, Z2, Z3) as a geometric reference point of the treatment unit (6,7,12) whose spatial position (R) can be defined in coordinates (x,y,z); - stipulating a sequence (i = 0,1,2,...) of target spatial positions (Ri, R0, R1, R2,...) of the target point (Z1) to be adopted during a vehicle treatment, wherein the sequence of target spatial positions is stipulated, in particular, on the basis of contour recognition data (51) relating to the vehicle (3) to be treated; - determining adjustment positions (Pi, P0, P1, P2,...) of at least one actuator (21,22,23), which is involved in the movement of the target point (Z1), based on a known geometry of the vehicle treatment system (1), such that the adjustment positions (Pi, P0, P1, P2,...) correlate with the target spatial positions (Ri,R0,R1,R2,...) of the target point (Z1).

5. Control method according to the preceding claim, wherein, in order to create a schedule (50) for a plurality of treatment units (6,7,12), target points (Z1,Z2,Z3) whose instantaneous spatial positions (R,L) can be defined in coordinates (x,y,z) are provided, and wherein spatially coordinated target spatial positions (Ri,Li) of these target points (Z1,Z2) of the plurality of treatment units (6,7,12) are planned as a target pose sequence (K).

6. Control method according to one of the preceding claims, wherein - the time scaling at least in sections provides constant scaling (81) as stretching in the case of a time scaling factor (F); AND / OR - the time scaling at least in sections provides variable scaling (82), wherein the variable scaling is formed in a state space by a chain of integrals over a range of the flow variable (s), and wherein the integrals are formed O on the basis of boundary conditions at the range limits, AND O on the basis of previously known physical limits and / or defined operating limits for the actuator movement, in particular the maximum speed (U°max) and / or the maximum acceleration (U°°max) of an actuator (21,22,23).

7. Control method according to one of the preceding claims, wherein synchronized movement paths (U,V,W) and / or synchronized trajectories (A,B,C) are created for a group of actuators (21,22,23), the movements of which are kinematically or logically coupled.

8. Control method according to one of the preceding claims, wherein kinematically coupled movements are available for a group of actuators (21,22,23) when - these actuators (21,22,23) are intended for the operation of two or more treatment units (6,7,12), AND - there is at least one actuator (21) in the group, the operation of which causes at least two partial movements, specifically a first partial movement for a first treatment unit (7) and at least one further partial movement for a second treatment unit (12).

9. Software product for a vehicle treatment system, wherein the software product is stored on a data carrier or transmitted by a data interface, characterized in that the software product comprises instructions which are designed to carry out a control method according to one of the preceding claims when executed on an electronic data processing device, in particular a control device (30) of a vehicle treatment system (1).

10. Control device for a vehicle treatment system (1), wherein the control unit has interfaces in order to connect the control device to a plurality of actuators (21,22,23) of the vehicle treatment system (1), characterized in that the control device (30) comprises instructions which are designed to carry out a control method according to one of the preceding claims.

11. Vehicle treatment system (1) having two or more treatment units (6,7,12) which can be moved by controllable actuators (21, 22, 23) to carry out a surface treatment of a vehicle (3), wherein the vehicle treatment system (1) comprises a controller (30) and a contour detection means (31) for detecting a contour of the vehicle (3) to be treated, characterized in that the controller (30) is designed to carry out a control method according to one of the preceding claims.

12. Vehicle treatment system according to the preceding claim, wherein at least one of the treatment units (6,7,12) is movable relative to the vehicle (3) by operation of two or more actuators (21,22,23), and wherein the movements of a group of actuators (21,22,23) are coupled one or more times, such that there is at least one actuator (21), the operation of which causes partial movements for two or more of the treatment units (6,7,12).

13. Vehicle treatment system (1) according to one of the preceding claims, wherein there is at least one group of actuators (21, 22, 23) designed to jointly move the treatment unit (6,7,12) relative to a vehicle (3) to be treated in a multidimensional adjustment movement, wherein the vehicle treatment system (1) comprises at least one regulator (32) and the controller transfers target values to the at least one regulator (32), on the basis of which the regulator (32) generates one or more adjustment signals for actuating the actuators, characterized in that - the controller comprises at least one dynamic plan (70) which, in an ordered data structure, defines a plurality of target values for the movements of the plurality of actuators (21, 22, 23) in relation to a time variable (t), wherein the target values define at least one movement variable (A,B,C / A°,B°,C° / A°°,B°°,C°°) of the respective actuator (21, 22, 23) over the duration of a treatment section, - and in that a target point (Z1,Z2,Z3) is defined for the treatment unit (6, 7, 12), and the target values of the dynamic plan (70) for the associated group of actuators (21,22,23) are defined in such a way that the target point (Z1,Z2,Z3) of the treatment unit (6,7,12), during a multidimensional adjustment movement, follows a path course which maintains a geometry reference with respect to a contour profile of the vehicle to be treated during a surface treatment, - and in that the dynamic plan (70), for the duration of the treatment section, defines mutually synchronized accelerations (A°°,B°°,C°°) for each phase with a multidimensional adjustment movement of the treatment unit for the actuators (21,22,23) in the group, with the result that matching acceleration phases (ki) result for the plurality of actuators (21,22,23) when carrying out the multidimensional adjustment movement.

14. Vehicle treatment system (1) according to one of the preceding claims, wherein - the target values define at least one movement variable (A,B,C / A°,B°,C° / A°°,B°°,C°°) of the respective actuator (21, 22, 23) in such a way that they describe a current and a future movement of the respective actuator (21, 22, 23); AND / OR - the target values of the dynamic plan define, for the actuators in the group, mutually fully synchronized accelerations that are proportional to each other in matching time phases; AND / OR - the matching acceleration phases (ki) of the actuators (21,22,23) in the group begin and end at matching times (ti) for each actuator (21, 22, 23); AND / OR - the target point (Z1,Z2,Z3) for the treatment unit (6, 7, 12) defines the current spatial position (L) of the treatment unit (6, 7, 12), in particular a translational spatial position and / or a rotational spatial position; AND / OR wherein - the geometry reference of the target point (Z1,Z2,Z3) relative to the contour profile of the vehicle to be treated: O is a fixed vertical distance from the vehicle surface, AND / OR O is a fixed distance from the vehicle surface at a predefined effective angle relative to the vehicle surface.

15. Vehicle treatment system (1) according to one of the preceding claims, wherein - a multidimensional adjustment movement of the treatment unit takes place during a treatment section with a direction of movement of a portal of the vehicle treatment system that reverses one or more times; AND / OR - a multidimensional adjustment movement of the treatment unit during a treatment section comprises temporary immersion of the target point (Z1,Z2,Z3) in a concave contour of the vehicle to be treated, in particular O in a concave front contour below an alcove, AND / OR O in a concave front contour or rear contour below a bumper, AND / OR O in a concave contour below a roof structure that projects beyond the windscreen or rear window of the vehicle.