Control method for operating a beam brake

Abstract

A control method for operating a beam brake (BLK) that acts on the wheels (RD) of a shunting system (AAL), where a manipulated variable is determined for the shunt to generate an actuating force for the beam brake. In a static analysis, the lateral forces acting on the wheels in the brake beams of the beam brake are calculated using a finite element method as a function of an actuating force acting on the brake beams by at least one actuator. In a dynamic analysis, a state-space model is calculated. An optimum is calculated for the time-dependent behavior of the manipulated variable, taking into account the model for the beam brake and the state-space model, and using the target run-out speed of the shunt from the beam brake as a first boundary condition.Taking into account the found optimum for the time course of the manipulated variable, control signals are output for the beam brake.

Description

Technical area

[0001] The invention encompasses the following subject matter: a control method for operating a beam brake. The invention further encompasses the following subject matter: a control device for a beam brake. The invention further encompasses the following subject matter: a computer program comprising program instructions. Technical background

[0002] According to the state of the art, it is known to calculate the required braking work of beam brakes before a process to be braked reaches the beam brake. Before the braking process begins, for example, the initial braking level (which, for example, is associated with a fixed piston stroke in hydraulic beam brakes) is determined from an energetic consideration of at least the entry speed, the target exit speed, and the mass of the process.

[0003] Common control concepts are based on a PID controller, which is not optimally suited for this application due to the nonlinearity of the controlled system. Due to good precontrol of the initial braking level and low controller dynamics, achieved through appropriate controller parameterization, the closed control loop usually delivers acceptable results regarding the stopping speed. Due to the nonlinear control system and the linear controller, the manipulated variable curve is subject to a certain oscillation during the braking process. This oscillation only remains reliably stable due to the controller's low dynamics and the lack of the possibility of positive acceleration of the process by means of the beam track brake. Increasing the speed of the process by means of the beam track brake is not possible.

[0004] To date, for example, according to DE 1605342 B2, the target speed trajectory during the braking process is iteratively weighted based on the number of axles in the effective braking range and the braking distance traveled by the beam-type brake, with the goal of achieving the target run-out speed when the last axle leaves the beam-type brake. In addition to the weighting based on the number of axles in the beam-type brake, heuristically determined parameters for the various bogie types are used to account for the influence of closely spaced axles. The actual speed is measured using a Doppler radar, for example. The speed error is fed to the PID controller, and the beam-type brake is activated depending on the newly calculated braking level.

[0005] In addition, DE 102018200867 A1 describes that by plotting the speed square over the location, a statement about the friction coefficient is made possible and this information is either passed on to the subsequent braking squad, if applicable, or the target run-out speed in the current beam track brake is reduced in order to reduce the necessary braking work in the subsequent braking squad.

[0006] The brake control system must also take into account the problem of wheels climbing up the rail in the beam brake. This means that the wheel is lifted off the rail due to an excessive lateral force exerted by the beam brake, which in the worst case scenario can lead to the complete derailment of the relevant train. The previous method according to EP 95109164 A1 for preventing climbing up is based on limiting the lateral force depending on the axle with the lowest weight, regardless of where this axle is located in the beam track brake and how many other axles are located in the beam track brake. This is the only, albeit restrictive, way to reliably prevent climbing up. The determination of this maximum permissible lateral force is usually carried out with the help of heuristic tests to determine when climbing up begins.The weight of the lightest axle is varied and the test is repeated. The recorded data points are then transferred to the brake control system and linearly interpolated during operation to calculate the maximum permissible braking level. If problems due to climbing occur during operation, these limits can be adjusted. Based on this approach, climbing can be largely prevented, but a potentially usable portion of the braking energy is not utilized.

[0007] DE 102016218839 A1 describes another solution to address the problem of climbing, but it is based on detecting the climbing rather than preventing it. This means that the climbing is counteracted as soon as it is detected by a suitable sensor. At this point, however, active countermeasures are already required, which represents a relatively greater intervention in the process. Furthermore, there is a risk that the countermeasure is not suitable for preventing further climbing.

[0008] The current state of the art presents the problem that the heuristics described above only imperfectly reflect reality. Therefore, reserves must be maintained by the beam brakes, and deviations between actual and target runout speeds must be accepted. These limitations can negatively impact the performance of the discharge system and increase hardware procurement costs. Summary of the invention

[0009] The object of the invention is to resolve the described problems in the prior art. In particular, it is the object to provide a control method for beam brakes used in a processing system, with which the beam brakes can be operated individually for each process while simultaneously optimally utilizing the potentially available braking power of the beam brake. Furthermore, the object of the invention is to provide a control device or a computer program that can control one or more beam brakes in such a way that the aforementioned problem related to the method is solved.

[0010] According to a first aspect of the invention, a control method is described for operating a beam brake which acts on wheels of a discharge system in a discharge system and in which a control variable for generating an actuating force for the beam brake is determined for the discharge (with the aim of achieving a desired discharge speed).

[0011] The manipulated variable can, for example, be the piston stroke in a hydraulic actuation device for the beam brake. The beam brake can also be electrodynamically driven, in which case the manipulated variable is determined by the electrical power (i.e., voltage and current, with the current preferably being changed as the manipulated variable).

[0012] Run-off systems are defined as track systems where non-powered rail vehicles can run off due to a gradient in the track. Beam brakes are used to reduce speed.

[0013] According to the invention, the real processes in beam brakes are modeled in such a way that a) (in a static view) a model of the transverse forces acting on the wheels in the brake beams of the beam brake is calculated using a finite element method as a function of an actuating force acting on the brake beams by at least one actuator (for example a hydraulic cylinder and a suitable mechanical deflection device), b) (in a dynamic view) a state space model is calculated, describing a process dynamics of the process by the temporal course (continuous or preferably at discrete time intervals) of a process position as the first state variable and the temporal course (continuous or preferably at discrete time intervals) of a process speed as the second state variable, as well as describing an actuation dynamics of the beam brake as at least one further state variable,where the change in the respective time derivative of all state variables (continuously or preferably at discrete time intervals) is also taken into account (i.e., calculated or measured). c) an optimum is calculated for the temporal progression of the manipulated variable, taking into account the model for the beam brake and the state-space model and based on the target run-out speed of the run-off from the beam brake as a first boundary condition, d) control signals for the beam brake are output, taking into account the optimum found for the temporal progression of the manipulated variable.

[0014] Due to the longitudinal extension of the brake beam, the actuating force acting on the brake beam is usually exerted through several positions. These positions include, for example, spring assemblies that act on the brake beam(s) at several points and can be subjected to a variable preload by a hydraulically driven linkage. The linkage can be driven by one or more hydraulic cylinders.

[0015] One advantage of the invention is that, even before a beam brake's rundown is reached, a possible optimal trajectory of the speed and the manipulated variable curve can be calculated using the optimization algorithm, taking into account assumptions. Depending on the optimization problem, oscillations in the control behavior can be minimized. These oscillations can be minimized by determining the initial braking level more precisely based on the models and by taking existing nonlinearities into account, so that they do not have to be compensated for by the control behavior alone. This makes it possible to better utilize the available maximum braking work. This, in turn, means that fewer safety reserves need to be considered in the design.

[0016] In the future, due to the introduction of digital automatic coupling (DAK), the requirements for target braking will also increase with regard to the permissible approach speed (to sequences already completed) on the directional track, and thus with regard to the precision of the last brake sequence before the sequences come to a standstill (usually the directional track brakes). This is also where the present invention comes in, as more precise control of the affected beam brakes is possible for the reasons mentioned above.

[0017] According to a further aspect of the invention, a control device for a beam brake is described, comprising a processor and a memory unit.

[0018] The control device can be composed of several brake controllers. Preferably, a central control device can be provided that communicates with brake controllers for entire brake squads or individual brakes and controls their interaction. The central control device can also be configured as a computer for performing calculations using the finite element method.

[0019] According to this aspect, the aspects of the invention explained above are determined in that the control device is configured to carry out a method according to any one of claims 1-15 with the aid of the processor by calling up data in the memory unit.

[0020] The advantages mentioned above for the process therefore also apply to the control device.

[0021] According to a further aspect of the invention, a computer program is described, comprising program instructions which, when the program is executed by a computer, cause the computer to carry out steps a) and b) and c) of the method according to one of claims 1 to 15.

[0022] Furthermore, according to a further aspect of the invention, a computer program containing program modules with program instructions for implementing the aforementioned method according to the invention and / or its exemplary embodiments is described, wherein the method according to the invention and / or its exemplary embodiments can be implemented by means of the computer program. Thus, the above-mentioned advantages for the method also apply correspondingly to the control device.

[0023] Furthermore, according to a further aspect of the invention, a provision device for storing and / or providing the computer program is described. The provision device is, for example, a storage unit that stores and / or provides the computer program. Alternatively and / or additionally, the provision device is, for example, a network service, a computer system, a server system, in particular a distributed, for example, cloud-based computer system and / or virtual computer system, which stores and / or provides the computer program preferably in the form of a data stream.

[0024] The provision takes place in the form of program data sets as a file, in particular as a download file, or as a data stream, in particular as a download data stream, of the computer program. This provision can also take place, for example, as a partial download consisting of multiple parts. Such a computer program is transferred, for example, using the provision device into a computing environment, so that the method according to the invention can be executed in a computing instance.

[0025] A device is computer-aided or computer-implemented if it has at least one computer or processor, or a method if at least one computer or processor carries out at least one method step of the method.

[0026] A computing environment is an IT infrastructure consisting of components such as computers, storage units, programs, and data to be processed by the programs, which are used to execute at least one application that has to perform a specific task. The IT infrastructure can also consist of a network of these components.

[0027] Computing instances (or instances for short) form functional units within a computing environment that can be assigned to applications (given, for example, by a number of program modules) and can execute them. These functional units form self-contained systems, physically (e.g., a computer, processor) and / or virtually (e.g., a program module), when the application is executed.

[0028] Computers are electronic devices with data processing capabilities. Computers can be clients, servers, handheld computers, communication devices, and other electronic data processing devices that may contain processors and memory units and may also be connected to a network via interfaces.

[0029] Processors can be, for example, converters, sensors for generating measurement signals, or electronic circuits. A processor can be a central processing unit (CPU), a microprocessor, a microcontroller, or a digital signal processor, possibly combined with a memory unit for storing program instructions and data. A processor can also be a virtualized processor or a soft CPU.

[0030] Storage units can be implemented as computer-readable memory in the form of random-access memory (RAM) or data storage (hard disk or data carrier).

[0031] Program modules are individual software functional units that enable a program sequence of method steps according to the invention. These software functional units can be implemented in a single computer program or in several communicating computer programs. The interfaces implemented in this way can be implemented in software within a single processor or in hardware if multiple processors are used.

[0032] Interfaces can be implemented in hardware, for example wired or as a radio connection, or in software, for example as interaction between individual program modules of one or more computer programs. General embodiments of the invention

[0033] Variants describing further developments of the invention are explained below without limiting the basic idea of ​​the invention.

[0034] According to a variant, the aspects of the invention explained above are determined by the fact that a model predictive control of the beam brake is carried out with the control signals.

[0035] The application of a control system combined with a closed-loop control system can also be referred to as model-predictive control. The open-loop control component determines the model-predictive component of the closed-loop control system. The model makes it possible to determine the parameters of the braking process before braking occurs to control the braking process. The closed-loop control system allows for intervention if the model deviates from reality, i.e., if the braking process parameters used for control do not lead to the predicted result. Advantageously, comparatively small corrections are required, which means that fewer manipulated variable changes are necessary, thus dampening the oscillation behavior of the control loop.

[0036] The principles of open-loop and closed-loop control must be distinguished from one another. Open-loop control is based on a heuristic that influences the behavior of a system in a targeted manner, with the goal of ensuring that the system responds to changes in a desired manner. Within this framework, the system's control system operates stably. However, open-loop control alone cannot prevent additional changes in the system's state due to unforeseen disturbances. Control, on the other hand, means counteracting changes in the system's state caused by disturbances by changing a controlled variable depending on the disturbance that occurs. However, this can lead to control oscillations.

[0037] A key advantage of this invention is the improved control during braking processes and, as a result, statistically better adherence to the target run-out speed from the individual beam brakes, thereby increasing process capability. Due to the continuous identification of the friction coefficient and the roll factor, the processes can be easily categorized.

[0038] According to a variant, the aspects of the invention explained above are determined by calculating the actuation dynamics for at least one actuation cylinder, each with a piston, a first chamber and a second chamber and actuation valves, taking into account at least one of the following further state variables, e) describing a piston dynamics of the piston by a piston position as the third state variable and by a piston speed as the fourth state variable, f) describing a pressure build-up dynamics in a first chamber by a first chamber pressure in the first chamber as the fifth state variable and by a second chamber pressure in the second chamber as the sixth state variable, g) describing a valve dynamics, wherein each chamber (ie the first chamber and also the second chamber) has at least one inlet valve and one outlet valve, by a respective valve position as the seventh state variable and ninth state variable and by a respective valve speed as the eighth state variable and tenth state variable.

[0039] This variant makes sense if the beam brake is adjusted hydraulically by an actuating cylinder and a connected mechanical deflection device, which may also include spring assemblies. Alternatives to this are electrodynamic drives for the deflection device of the beam brake.

[0040] The advantage of this variant arises from the fact that the actuation dynamics for the beam brakes contribute to a delayed implementation of control signals. If this dynamic is already taken into account in the control system, the dynamics are reduced as a disturbance in the beam brake technical system, or at least their effect is lessened. This advantageously means that a control system running in parallel with the control system requires fewer corrections, thus reducing the likelihood of oscillations occurring in the manipulated variable curve.

[0041] According to a variant, the aspects of the invention explained above are determined in that, on the basis of a deviation of the acceleration of the process in the beam brake from a target acceleration in the beam brake calculated on the basis of the temporal course of the manipulated variables, an additional control of the manipulated variable takes place.

[0042] The actual temporal behavior of a specific piston position is not exactly known despite modeling and is preferably additionally measured using a slot position sensor or similar. This also makes it possible to implement a position control loop; there are at least one, preferably two different inlet valves and one outlet valve for each chamber (i.e. the first chamber A and the second chamber B), each of which could also be referred to as a coarse and fine valve depending on their function. The respective valves are switched off via the control system using heuristically determined lead times that take into account the difference between the actual and target position. This advantageously uses a technically easy-to-implement option to control the process deviating from the control parameters originally specified by calculation from the model.The focus is on achieving the target position within a certain tolerance and on avoiding overshoots.

[0043] According to a variant, the aspects of the invention explained above are determined by additionally taking into account a second boundary condition aimed at preventing the wheels of the drain from climbing.

[0044] Climbing of the wheels must be prevented at all costs, as this could, in the worst case, lead to a complete derailment of a discharge line in the discharge system. Therefore, a safety margin must be maintained with respect to the braking force that would cause climbing. However, this reduces the effective braking energy that can be generated by the beam brake.

[0045] However, the reliable prevention of climbing through the less restrictive approach according to the invention can increase the usable braking energy for corresponding processes, while maintaining comparable safety regarding climbing prevention. This, in turn, enables the use of smaller-sized beam brakes or the utilization of a higher potential braking energy of the beam brakes of the same size. This advantageously allows either the investment in the process system to be reduced or the performance of the beam brake to be increased, thus increasing the profit margin in both cases.

[0046] According to a variant, the aspects of the invention explained above are determined by checking as a second boundary condition whether the difference between h) a torque resulting from a first braking force component generated by a front contact area of ​​a wheel band on the brake beam, seen in the direction of travel, and a height offset between the top of the rail on which the wheel stands and the front contact area as the minuend, and i) a torque resulting from a second braking force component generated by a rear contact area of ​​the wheel band on the brake beam, seen in the direction of travel, and a height offset between the top of the rail and the rear contact area, as the subtrahend, j) as a torque resulting from half the weight of an axle carrying the wheel and a horizontal offset between the axle and the front contact area.

[0047] This boundary condition is advantageously a model that can be calculated with little computational effort and at the same time represents a very good approximation to the real-life conditions for the wheel to climb during braking in the beam brake.

[0048] According to a variant, the aspects of the invention explained above are determined in that, when calculating the first braking force component and the second braking force component, a friction coefficient (e.g. an available literature value or the result of a calculation explained below) for the pairing between the wheel bandage and the brake beam and a roll factor (e.g. an available literature value or the result of a calculation explained below) for the braking effect of the process are taken into account, wherein the product of the lateral force and the friction coefficient and the roll factor each results in the braking force component.

[0049] As long as no calculated values ​​are available using the finite element method, literature values ​​or empirical values ​​for the friction coefficient and roll factor are used. However, it is advantageous to calculate values ​​identified individually for the braking process through calculation steps, which are then iteratively approximated to the actual conditions during the ongoing braking process.

[0050] According to one variant, the above-explained aspects of the invention are determined by additionally taking into account a third boundary condition aimed at ensuring uniform acceleration of the process while passing through the beam brake. The aforementioned uniform acceleration by the beam brake represents a negative acceleration.

[0051] According to a variant, the aspects of the invention explained above are determined by additionally taking into account a fourth boundary condition aimed at avoiding a jerk in the process.

[0052] Jerk is the time derivative of acceleration; in other words, jerk is caused by changing acceleration. Reduced jerk acceleration reduces the likelihood of cargo shifting during braking. Especially in tank cars filled with an inertial liquid, this can otherwise lead to sloshing of the cargo (so-called sloshing). This can be mitigated by reduced jerk acceleration (more on this below).

[0053] According to a variant, the aspects of the invention explained above are determined by additionally taking into account a fifth boundary condition aimed at avoiding changes in the manipulated variable.

[0054] This advantageously reduces the stresses on the brake mechanism, extending the service life of the beam brake and extending maintenance intervals. This contributes significantly to the cost-effective operation of the discharge system.

[0055] According to a variant, the aspects of the invention explained above are determined by taking into account at least one further boundary condition in addition to the first boundary condition, the boundary conditions taken into account being weighted.

[0056] Depending on the definition of the constraints (boundary conditions), the desired operational aspects can be advantageously taken into account. These include, for example, energy- and / or wear-optimized trajectories during the braking process, in which as few changes in the control variables as possible occur. Likewise, a deceleration that is as uniform as possible or jerk-optimized can be calculated, which would be particularly advantageous for the braking process and the subsequent free running of the aforementioned sloshing cars with regard to prediction. It is also possible that further constraints will be found in the future that could potentially reduce noise emissions.

[0057] Unlike previous control concepts, which can only be parameterized using the standard parameters P, I, D and a process-specific controller gain, the invention allows for the dynamics to be taken into account, e.g., for the opening of the brake, in a state-based manner and no longer requires hard-coded times.

[0058] For the operation of high-performance stations, it is also important that the constraints of the model predictive controller allow for delayed braking, which is characterized by the fact that boundary signs of the dispatch system can be cleared earlier and thus the dispatch sequence times can be reduced (this means an increase in the dispatch performance of the dispatch system).

[0059] According to a variant, the aspects of the invention explained above are determined by calculating the product of the process parameters of the friction coefficient and the roll factor by subtracting from the product of the mass of the run and the acceleration (corresponding to the time derivative of the speed) a running resistance, taking into account at least one of the following running resistance components: slope resistance, curve resistance, switch resistance, rolling resistance and air resistance, and dividing the difference obtained by the lateral force generated by the brake beam for braking.

[0060] If the product of the process parameters of the friction coefficient and the roll factor is known for a process, a comparatively accurate prediction of the braking force curve in the beam brake can be made on this basis (more on this below). The aforementioned properties of the process are incorporated into the parameterization and thus advantageously improve the agreement with the real conditions. A further operational advantage is the model-based optimization of the speed signal, which helps to bridge both the failure of a component and, for example, the short-term shadowing of the Doppler radar. In addition, further measurements of process parameters, such as the wheel tire width or the running circle diameter, can be used directly to improve the prediction and take this into account quantitatively.

[0061] According to a variant, the aspects of the invention explained above are determined by calculating the process parameter of the friction coefficient by subtracting from the product of the mass of the run and the acceleration (corresponding to the time derivative of the speed) a driving resistance, taking into account at least one of the following driving resistance components: inclination resistance, curve resistance, switch resistance, rolling resistance and air resistance, and dividing the difference obtained by the lateral force generated by the brake beam for braking and by the roll factor determined for the wheel.

[0062] This embodiment of the invention requires that the roll factor of the relevant process be determined by measurement in a conventional manner. This has the advantage that, by applying the method already described above, a comparatively accurate statement about the friction coefficient is possible, since this is the only variable in the associated calculation equation. This advantageously enables further improvement in the parameterization of the model, which is thus closer to reality.

[0063] According to a variant, the aspects of the invention explained above are determined in that an optimized value is calculated using a signal filter taking into account the stochastic distribution of the measured values ​​of a speed measurement for the process and a stochastic distribution of the predicted values ​​for a speed profile of the process.

[0064] Such a calculation can be performed in a conventional manner, for example, using a Kalman filter. By comparing the predicted values ​​with the actually measured values, the quality of the calculation method can be assessed and, taking the insights gained into account, further improved for subsequent processes.

[0065] According to a variant, the aspects of the invention explained above are determined by k) the time course of the manipulated variable (this can be the calculated course or the actual course of the manipulated variable influenced by control) of a braking process l) and either the friction coefficient determined during the braking process in question and the roll factor determined during the braking process in question or the product of the friction coefficient and the roll factor determined during the braking process in question are stored and used in a subsequent braking process with the beam brake in question (to take account of the braking characteristics) or with another beam brake (to take account of the sequence characteristics of the sequence in question).

[0066] Further application of the knowledge gained from a specific braking process allows for more precise selection of the initial braking level and better prediction of the speed trajectory, thus enabling more stable process control. Overall, the method according to the invention contributes to a more efficient operation of a process plant during ongoing operation. Changes in state variables over the operating period, which may occur, for example, due to wear, can also be compensated for by adaptively adjusting the process. Exemplary embodiments of the drawing

[0067] Further details of the invention are described below with reference to the drawings. Identical or corresponding drawing elements are provided with the same reference numerals in the individual figures and are explained repeatedly only to the extent that differences arise between the individual figures.

[0068] The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiments each represent individual, independently considered variants of the invention, which also further develop the invention independently of one another and are thus also considered components of the invention, either individually or in a combination other than that shown. Furthermore, the described components can also be combined with the variants of the invention described above. Figure 1shows, in a schematic sketch, an embodiment of an AAL processing system with its functional relationships, specifically with an embodiment of the control device according to the invention, in which a computer program 1 for executing an embodiment of the method according to the invention is installed. Also shown is an embodiment of a computing environment for the AAL processing system as a block diagram, wherein the individual computing instances execute program modules, each of which can run in one or more of the computers shown as examples and implemented as controllers, and wherein the interfaces shown can accordingly be implemented in software in one computer or in hardware between different computers. Figure 2shows the schematic structure of the actuation of a beam brake BLK, which is driven by a hydraulic cylinder, suitable for control by an embodiment of the method according to the invention. Figure 3 shows a simplified top view of a real geometry of the beam brake BLK according to Figure 2 , represented with realistic proportions, and a transformation to a model for the application of a finite element method according to an embodiment of the method according to the invention. Figure 4 shows the result of the FEM calculation according to an embodiment of the inventive method for three different wagons that differ in the number of axles and / or axle spacing. The abscissa represents the position of the first axle s in meters. Figure 5 shows schematically the forces acting on a drain to illustrate the parameters to be taken into account in the method according to the invention. Figure 6shows schematically the forces acting on a wheel RD of the process in order to clarify the understanding of the phenomenon of climbing wheels RD. Figure 7 shows the transverse force F in Newton N of the beams on the wheel drum RB, plotted against the position of the first axis s in meters. Figure 8 shows a summary of the FEM calculation, where the torques caused by the braking force and the weight force are shown in Newton meters Nm over the wheel travel s in meters. Figure 9 shows exemplary probability distributions of a state variable based on the state space model, the measurements during the execution of a process and an estimated optimum as a diagram. Figure 10 shows an embodiment of the method according to the invention as a flow chart, wherein the individual method steps can be implemented individually or in groups by program modules. Detailed description of the drawing

[0069] Figure 1 shows a schematic sketch of an embodiment of a discharge system AAL with an embodiment of the control device according to the invention, in which a computer program 1 for carrying out the method according to the invention is installed. The upper part of the Figure 1 the track layout of the AAL discharge system and the lower part of the Figure 1 the gradient profile or a longitudinal section of the AAL drainage system.

[0070] In a computing environment created by the controllers, the computing instances used and other connected devices are connected by a first interface S1, a second interface S2, a third interface S3, a fourth interface S4, a fifth interface S5, a sixth interface S6 and a seventh interface as well as S7, in a central control device ZST a first processor PR1 is connected to a first memory unit SE1 by an eleventh interface S11, in a hill brake control BBS a second processor PR2 is connected to a second memory unit SE2 by a twelfth interface S12, in a valley brake control TBS a third processor PR3 is connected to a third memory unit SE3 by a 13th interface S13 and in a directional track brake control RBS a fourth processor PR4 is connected to a fourth memory unit SE4 by a 14th interface S14.

[0071] According to the presentation of the Figure 1 The AAL discharge facility, which is part of a shunting facility for rail traffic, has a discharge ramp ABR starting from a mountain peak BG, which is connected to an intermediate incline ZNG, a distribution zone VTZ with distribution switches VW1 ... VW7 and directional tracks RG1 ... RG8. In addition, Figure 1 Track brakes in the form of a mountain brake squad BB with a first mountain brake BB2 and a second mountain brake BB1, a valley brake squad TB with a first valley brake TB1 and a second valley brake TB2 and a directional track brake squad RGB with a first directional track brake RB1, a second directional track brake RB2, a third directional track brake RB3, a fourth directional track brake RB4, a fifth directional track brake RB5, a sixth directional track brake RB6, a seventh directional track brake RB7 and an eighth directional track brake RB8 can be seen.

[0072] In addition to the components of the AAL drainage system mentioned above, Figure 1 By way of example, a first process AB1, a second process AB2 and a third process AB3 are shown, which are pushed over the hump by a push-off locomotive ABL or have been pushed off at a push-off point AP (which does not necessarily have to be on the hilltop BG and is shown as an example for the third process AB3) and subsequently move along the hump system AAL, driven by the acting force of gravity.

[0073] To control the valley brake relay TB, containing the valley brakes TB1, TB2, Figure 1The valley brake control TBS is indicated, which is connected to the valley brake relay TB via a first interface S1, which can be wired or wireless. To control the uphill brake relay BB, containing the uphill brakes BB1, BB2, the uphill brake control BBS is also indicated, which is connected to the uphill brake relay BB via the second interface S2, which can be wired or wireless. Similarly, the directional track brake relay RGB, containing the directional track brakes RB1 ... RB8, is connected to the directional track brake control RBS via the third interface S3. For reasons of clarity, Figure 1Only one interface between the respective brake squad and the respective track brake control is shown as an example. Of course, each track brake can be controlled. It is also possible to provide a separate control for each track brake rather than a common control for the entire brake squad (not shown).

[0074] The downhill brake control TBS is connected to a central control device ZST of the AAL processing system via the fourth interface S4, the uphill brake control BBS via the fifth interface S5, and the downhill track brake control RBS via the sixth interface S6. This means that these control components together form a control device for controlling the track brakes, i.e., the uphill brakes BB1, BB2, the downhill brakes TB1, TB2, and the downhill track brakes RB1 ... RB8, in the form of a distributed control system. Alternatively, it would of course also be possible, for example, for the uphill brakes BB1, BB2, the downhill brakes TB1, TB2, and the downhill track brakes RB1 ... RB8 to be directly connected to and controlled by the central control device ZST (not shown). The control devices also perform the necessary control steps according to the invention.Overall, the process carried out by the control devices is also referred to as model predictive control.

[0075] The determination of control parameters for the track brakes in the form of the uphill brakes BB1, BB2, the valley brakes TB1, TB2 and the directional track brakes RB1 ... RB8 of the AAL process system is carried out in a distributed manner in the controls in accordance with the invention in such a way that a cross-brake analysis or optimization of the respective speeds of the processes is carried out, whereby the processes that take place in the brakes are simulated using a finite element method based on a model and a state model is calculated that takes into account the process dynamics of the process and the actuation dynamics of the beam brake BLK.

[0076] When controlling the braking sequences, the results of a measurement of the process characteristics of a particular process can be taken into account. These measurements are created by a measuring station MST at the earliest possible point in the process. The measuring station MST can contain processors and memory units (not shown) that interact with sensors (not shown). The metrological recording of processes is known per se and therefore not the subject of a detailed description. The measuring station MST is connected to the central control system via a seventh interface S7, whereby the first processor can, for example, perform the calculation for the beam brakes using the finite element method. Alternatively (not shown), an external computer 1 connected to the central control system can also be used for the calculation.

[0077] In the example, all brakes, i.e. the uphill brakes BB1, BB2, the downhill brakes TB1, TB2 and the directional track brakes RB1 ... RB8 are designed as beam brakes BLK, so that the following figures for all brakes of the discharge system AAL according to Figure 1 are valid. An example of the design of these BLK beam brakes will be explained below.

[0078] The hydromechatronic system of the beam track brake according to Figure 2(in the case of hydraulic actuation) essentially consists of a hydraulic control mechanism that controls the volume flow of oil into a first chamber KM1 via a first inlet valve EL1 and a first outlet valve AL1, and the volume flow of oil into a second chamber KM2 of an actuating cylinder BZL via a second inlet valve EL2 and a second outlet valve AL2. The control is carried out electrically by the control device, consisting of the aforementioned brake controls ZST, BBS, TBS, RBS. For a better understanding of the functioning of the beam brake BLK system, Figure 2 shown as a top view of the single-rail beam brake BLK with three spring assemblies FP.

[0079] The double-acting actuating cylinder BZL controls the preload of each spring assembly FP via its piston KB, a rod assembly GST, and deflections ULK. Due to mechanical stops ASL, movement of the beams occurs above a certain spring force. A distinction is made between the outer beam ABK and the inner beam IBK; both are mounted and can move orthogonally to a rail SN. The distance between the outer beam ABK and the inner beam IBK can be changed via a mechanical deflection ULK below the rail SN (not shown in detail); this is referred to as the brake groove width. The brake groove width is reduced for braking operations to a width smaller than the standard wheel tire width, which causes the wheel tires RD to cut open the brake groove and then increase the forces of the additionally preloaded spring assemblies FP.The wheel bandage RB of a track-guided wheel RD is formed by a widening of the wheel cross-section on the outer circumference of the wheel RD, which forms the tread of the wheel RD. This widening is thus detected by the beam brake BLK during braking.

[0080] For each change in piston stroke, power is required in the form of a volume flow of hydraulic oil at a nominal pressure of 180 bar. A trigonometric portion of the forces from the spring assemblies FP acts on the piston KB via the deflection ULK, thus additionally influencing the movement of the piston KB.

[0081] The Figure 3 The geometry of an example of the directional track brake RB1 ... RB8 with three spring assemblies FP can be seen. As in Figure 2 As described above, the outer beam ABK and the inner beam IBK of the beam brake BLK can be seen, which encompass the rail SN. The stops ASL and the deflection ULK are housed in housings and Figure 3cannot be seen in more detail, as can the piston KB and the associated valves.

[0082] One embodiment of the invention establishes a causal chain, which is briefly outlined and then specified in more detail below. The causal chain consists of the following steps: 1. FEM (finite element method) for calculating the acting transverse forces in the beam track brake 2. Improvement of the identification of a priori unknown process parameters 3. State space model for calculating the dynamics of the beam track brake system 4. Model predictive control concept taking constraints into account 5. Use of model-based signal filters

[0083] For better comprehensibility, the further course of the description is also divided into these five steps. 1. FEM for calculating the acting transverse forces in the beam track brake

[0084] The invention is based, as already described, on a FEM that has its origins in structural analysis. The beam bending of the brake beams is calculated as a function of bearings and external forces (in Figure 3 (Shown as an example for the inner beam IBK, but equally valid for the outer beam ABK). In the application of the beam brake BLK, the bearings correspond to the contact points between the wheel drum RB of wheels RD of the unloading track (not shown) and the brake beams, whose unloading movement is simulated quasi-continuously by the beam track brake. The external forces correspond to the forces F1, F2, and F3 from the spring assemblies FP, whose preload is ensured by the hydraulic system described above.

[0085] The abstraction of reality is in the Figure 3This is the version with three spring assemblies FP, but the abstraction principle can be adapted for any other brake type. Based on all constraints, the beam can be divided into n nodes K1 ... K9 (here 9) and n-1 bars A1 ... A8 (here 8), each with a calculable length L, corresponding to the sections mentioned above. These bars, along with the area moment of inertia I and the elastic modulus E of the beam, represent the input variables of the FEM. For each bar, a stiffness matrix Ki is created in a conventional manner, structured as follows: K i = K i i , i K i i , i + 1 K i i + 1 , i K i i + 1 , i + 1 = 12 ⋅ E ⋅ I L 3 − 6 ⋅ E ⋅ I L 2 − 12 ⋅ E ⋅ I L 3 − 6 ⋅ E ⋅ I L 2 − 6 ⋅ E ⋅ I L 2 4 ⋅ E ⋅ I L 6 ⋅ E ⋅ I L 2 2 ⋅ E ⋅ I L − 12 ⋅ E ⋅ I L 3 6 ⋅ E ⋅ I L 2 12 ⋅ E ⋅ I L 3 6 ⋅ E ⋅ I L 2 − 6 ⋅ E ⋅ I L 2 2 ⋅ E ⋅ I L 6 ⋅ E ⋅ I L 2 4 ⋅ E ⋅ I L

[0086] To couple the adjacent bars, a total stiffness matrix KG< is created, in which the elements K ii,i with the previous staff and K ii + 1,i + 1superimposed on the subsequent one. The forces of the spring assemblies FP are considered in a vector p depending on their assigned node at a specific index. To comply with the deformation conditions, the nodal displacement variables vj, which consist of deflection and rotation, must match the bar end deformations vji at each node. Accordingly, known nodal displacement variables, such as the wheel tire width RB of each axle, can also be considered in the vector v. K G ⋅ v + p = 0 .

[0087] From this information, matrices and vectors are then generated that take these aspects and the coupling of the members into account. This results in the following matrix equation: K 1 1,1 K 1 1,2 0 0 K 1 2,1 K 1 2,2 + K 2 2,2 K 2 2,3 0 0 K 2 3,2 K 2 3,3 + K 3 3,3 K 3 3,4 0 0 K 3 4,3 K 3 4,4 ⋅ v 1 v 2 v 3 v 4 + p 1 p 2 p 3 p 4 = 0 .

[0088] Naturally, the wheel rims RB can only absorb compressive forces and not tensile forces. Therefore, iterations may be necessary. Depending on the sign of the acting shear force, certain nodes are deleted from the matrix equation and the adjusted system of equations is solved again until no more tensile forces occur on the wheel rims RD. The matrix KG< remains symmetric, has a band structure, and is positive definite. This means that special methods can be used to solve the system of equations that require less computation time. For example, the Cholesky decomposition is used. In principle, however, the solution method used is irrelevant.These special procedures have the advantage that a simulation can also be carried out during the execution of the processes at the AAL process plant in order to individually optimize the braking process at the involved beam brakes BLK for each process, because the simulation procedure is fast enough for this.

[0089] The result is the beam bending, which has been confirmed to be consistent in several static tests. The calculation is based on first-order engineering bending theory.

[0090] The final step deals with the node-wise determination of the internal forces, which consist of the shear force Q and the bending moment M. The indices of the displacement variables vi refer to the position of the element in the vector v: Q i i M i i Q i + 1 i M i + 1 i = K i ⋅ v 2 ⋅ i − 1,1 v 2 ⋅ i , 1 v 2 ⋅ i + 1,1 v 2 ⋅ i + 2,1

[0091] This shear force can be calculated for each contact point between the beam and the wheel tire RB, thus enabling a location-dependent description of the acting force and can be compared with the creep limit in the prediction (more on this below). Properties of the process, e.g., with regard to the geometry of the wheelset RD, can be fully considered if the measured values ​​are available; otherwise, default or standard values ​​are used for calculations.

[0092] In Figure 4As a result of the FEM calculation, the curve of the braking force FB in Newtons N over the distance s of the first axle through the beam brake BLK in meters m is shown as an example for 3 different processes. Graph G1 represents the braking force curve of a 4-axle wagon with bogies, where the distance between the axles in the bogie is 1.8 m and the distance between the two facing axles of the bogies is 7 m. Graph G2 represents the braking force curve of a 2-axle wagon with an axle distance of 9 m. Graph G3 differs from graph G2 only in that the axle distance is 10.6 m. For this reason, graphs G2 and G3 mainly only differ in an offset in the direction s in the second half of their curve. 2. Improving the identification of a priori unknown process parameters

[0093] The braking force FB of a track in a beam brake is calculated from the product of the acting transverse force of the beams FQ , the friction coefficient µ, and a so-called roll factor ς, which takes into account the height of the brake beams in relation to the running circle diameter. This results in the following formula: F B = F Q ∗ μ ∗ ς

[0094] The braking force of a freight car in a beam-type track brake has a non-linear curve plotted against location, as the beams' lateral forces acting on the track depend on the position of the axles in the brake during the track run and the number and position of the spring assemblies (FP). Currently, the control system does not know the actual lateral forces exerted by the beams. Therefore, the product of the friction coefficient and the roll factor cannot be determined from the above formula (at least without using additional measured values; more on this below).

[0095] By discretely deriving the actual speed, taking into account the running mass and the driving resistances (inclination, curve, switch, rolling and air resistance), the product of the friction coefficient and the roll factor can be identified in each calculation cycle: μ ∗ ς = m ∗ a − F N + F Bo + F W + F R + F L F Q with µ friction coefficient ς roll factor m run-off mass a acceleration of the run-off FN inclination resistance force F Bo curve resistance force FW switch resistance force FL air resistance force FQ lateral force on the brake beam FB braking force

[0096] If a measuring device is installed to determine the running circle diameter of each axle, the roll factor can be calculated from this and can therefore further improve the identification of the friction coefficient, since an unknown is eliminated.

[0097] By means of a suitable averaging of these identifications, the future behavior of the current process can be predicted and, furthermore, the identification can also be used to improve the determination of the initial braking level of subsequent processes, which is possible if communication between the individual brake controls is ensured, for example via the central control according to Figure 1 . The division of this identification is due to the fact that one part of the friction pairing is process-specific and the other is brake-specific. 3. State space model for calculating the dynamics of the beam track brake system

[0098] A state-space model is a form of system description for dynamic transmission systems in the time domain. All relationships between input, output, and state variables are represented as matrices and vectors. All physical quantities in which energy is stored are assigned a state variable xi.

[0099] In addition, a vector xi is created with the temporal change of all states. The state-space model then comprises four interdependent dynamics: the process dynamics (states: process position, process velocity), the piston dynamics (position of piston KB, piston velocity), the pressure build-up dynamics (pressure in the first chamber KM1 and pressure in the second chamber KM2), and the valve dynamics of the four valves that control the volume flow in the piston KB. Furthermore, the four input signals for controlling the valves are taken into account in the input vector u.

[0100] The resulting vector equation exhibits various nonlinearities and can be seen in the following diagram. For better understanding, symbols not yet introduced are listed here: ΣF W sum of all other considered driving resistances, m A mass of the outlet, m K mass of the piston and the rod, α ratio of the piston areas of the first chamber and second chamber, AK piston area, F r frictional force of the piston movement, F ext force acting on the piston by the spring assemblies, EA , EB compression modulus of chamber A, B, VA , VB volume of chamber A, B, QA , QB volume flow in chamber A, B, Q li internal leakage flow, c valve constant, T valve time constant, d valve damping, u input variables (opening degree of the four valves)

[0101] Process dynamics, states: x 1 Position of the drain x 2 Speed ​​of the drain x˙1=x2x˙2=1mAFNx1−FB+∑FW

[0102] Piston dynamics, states: x 3 Position of the piston x 4 Speed ​​of the piston x˙3=x4x˙4=1mKx5−αx6∗AK−Frx4−Fextx3

[0103] Pressure build-up dynamics, states: x 5 pressure in first chamber x 6 pressure in second chamber x˙5=EAx5VAx3QAx5x7,9,11,13−AK∗x4+Qlix5x6x˙6=EAx6VBx3QBx6x7,9,11,13−αAK∗x4+Qlix5x6

[0104] Valve dynamics, states: x 7,9,11,13 Position of the four valves x 8,10,12,14 Speed ​​of the four valves x7,9,11,13.=x8,10,12,14x˙8,10,12,14=c7,9,11,13T2u8,10,12,14−x7,9,11,13−2∗d∗T∗x8,10,12,14

[0105] The temporal behavior of a specific piston positioning is not known in advance and is measured using a slotted position sensor. This position control loop can essentially be viewed as a five-point controller; there are two different logic valves for each direction, each referred to as a coarse and a fine valve. The respective valves are switched off via the control system using lead times that take into account the difference between the actual and target positions. The focus is on achieving the target position within a certain tolerance and on avoiding overshoots. For each change in piston stroke, energy in the form of volume flow at a nominal pressure of 180 bar is required for a certain time. A trigonometric component of the forces from the spring assemblies acts on the piston via the deflection and thus additionally influences the movement of the piston.

[0106] The valve constants can also be determined model-based from the results of the reaction time measurements. During these measurements, the entire piston stroke used is traversed in both directions, and the time in each direction is recorded. This process is performed once with the fine valve and then repeated with both the fine and coarse valves (provided that one fine valve and one coarse valve are provided for each of the two intake valves and the two exhaust valves). This result is then compared and aligned with the modeling results.

[0107] Figure 5 shows the forces acting on a process as an example. - FB braking force m A see Inertial force with the second derivative of the distance s of the process and the mass m A of the process m A gWeight force of the process, where g is the acceleration due to gravity

[0108] Figure 6shows the forces acting on a wheel of the drain. This schematic representation can be used to explain the climbing process in particular. RBWheel bandage K1, K2Contact points between brake beam and wheel bandage SOTop edge of rail F g Weight force FB Braking force FActuating force aHorizontal distance between car axle and contact point hVertical distance between car axle and contact point h'Vertical distance between contact point and top edge of rail (attack height)

[0109] The braking force is transmitted across a contact surface between the brake beams and the wheel drum, which can be ideally considered as contact points K1 and K2. Climbing in a beam-type track brake occurs when the following torque inequality (10) is true. The braking force transmitted at contact point K1 is considered with a positive sign, as it promotes climbing, whereas the braking force at contact point K2 inhibits climbing and is therefore considered negative. If the beam is not in contact with one of the contact points, the force is set to zero. F B K1 ∗ h − F B K2 ∗ h > F g 2 ∗ a 4. Model predictive control concept taking into account constraints, i.e. boundary conditions

[0110] The prerequisites for using a model-predictive controller were described above. In other words, there is an FEM model for determining the braking forces, and the time behavior of the brakes can be predicted using the state-space model. The key difference compared to conventional control technology is that the future system behavior is taken into account when calculating the current manipulated variable, allowing the system to be controlled before control begins, taking the system behavior into account. When determining the future manipulated variables, a quality functional J and any input, output, and state restrictions can be considered as constraints. This prediction and optimization takes place over a time horizon N to be defined. The calculated manipulated variable curve is then optimal in terms of the quality functional. The constraints can also be taken into account in this quality functional. u = argmin J x 0 , u with J x 0 , u = ∑ k = 0 N 〚 x k 〛 Q + 〚 u k 〛 R

[0111] The indices Q and R represent weighting factors that allow manual adjustment of the influence on optimization. These weighting factors can be based on empirical values ​​and adjusted as appropriate based on insights gained from operating a wastewater treatment plant.

[0112] The design of this quality functional offers a wide range of possibilities; here are just a few examples of possible constraints: uniform acceleration throughout the entire braking process, jerk optimization, energy consumption and wear optimization, prevention of axle climbing

[0113] When accelerating a process evenly over the entire braking process, it is necessary to keep the effective braking force constant. Due to the non-linearity (compare Figure 4) an almost continuous adjustment of the braking level and thus of the piston stroke during the braking process.

[0114] When using jerk optimization, the optimization is based on the continuous change in acceleration. This method has the advantage, especially for tank cars where a sloshing effect can occur, of significantly reducing the sloshing effect. This also results in the car's free running behavior after the beam track brake being more uniform and easier to predict, and, for example, the required distances to trailing or leading processes can be reduced.

[0115] When optimizing energy consumption and wear, the focus is primarily on reducing changes in the control variables, as these are associated with the use of pressurized hydraulic oil and mechanical movements of the rods of the beam track brake due to the cylinder movement.

[0116] A particularly important constraint is the avoidance of climbing to prevent derailments. Depending on the braking level and axle spacing, potentially critical situations can be identified even before the train enters the beam-type retarder and taken into account in the trajectory planning of the braking process. A key component in calculating whether climbing will occur is the friction coefficient, which can be initialized using literature values ​​but cannot otherwise be determined a priori.

[0117] However, reference can be made to the procedure described in step 2 (see above). By calculating the future transverse forces acting on a wheel tire and identifying additional process parameters, the local areas where climbing could occur can be determined. In these areas, the braking level is reduced based on the quality function J. Care must be taken to ensure that the time horizon is sufficiently long to account for the dynamics of the beam track brake with regard to the adjustment of the braking level.

[0118] To illustrate the principle of braking an axle of a two-axle wagon in a beam track brake with five spring assemblies is shown in Figure 7The friction coefficient for the example is 0.3; the roll factor is 0.484 at a constant maximum braking level. For this purpose, the transverse force FQ is plotted against the position of the first axle s as it travels through the beam brake: If the acting transverse forces are assigned to the two possible contact points, the torque generated by the braking force is shown in the upper part of Figure 8 . In addition, the climbing limit KG, which results from the torque caused by the weight force, is shown.

[0119] If the sign of the torques at the contact points K1 and K2 (see Figure 6 ) is taken into account, the lower illustration results according to Figure 8In this example, there are areas where the torque caused by the braking force is greater than the torque caused by the weight. In these areas, climbing is likely and must be prevented by control technology.

[0120] The described method provides for a reduction in the braking level in advance to reduce the risk of climbing in these areas. Since the dynamics of the beam track brake are known due to the state-space model, the reduction in the braking level can advantageously be implemented sufficiently early and at the same time as late as possible.

[0121] This control concept requires significantly more computing power than the state-of-the-art and relies on the available information on the characteristics of the process (axle masses, axle spacing, possibly running circle diameter, possibly edge drum width) and the beam track brake (number of power units, span width, spring characteristic curve) to enable a robust prediction. Should some of this information not be available for a particular process, empirical values ​​for the calculation or existing state-of-the-art methods can be used. This ensures the functionality of the process system at all times, albeit not with the performance of the method according to the invention. 5. Use of model-based signal filters

[0122] From the methods already introduced in steps 1 to 4 above, a further possibility arises, which concerns the plausibility and optimization of the speed measurement of the Doppler radar device, if one is used. This is shown schematically in Figure 9shown. Using the calculated lateral forces from the FEM and the identification of the friction coefficient and roll factor, and with the help of the mass of the drain and the topology and topography of the drain system, an interval can be calculated in which the next measured value should lie. A Kalman filter, for example, can be used for this. This is a two-stage process: first, the system state is predicted and then updated with the measurement result and the associated measurement noise. The standard deviation is also updated in each case and can be weighted using the relevant matrices. It is assumed that both the prediction PRE and the measurement MSN are subject to random scatter and that their probability density corresponds to a normal distribution, resulting in two normal distributions.From these two normal distributions, an optimal state estimation EST is then performed, which takes into account the data from the modeling, the measurement and their standard deviations.

[0123] This procedure can be used in addition to already implemented plausibility checks to further increase the quality of speed measurement based on models.

[0124] Finally, an example of the process flow is shown using Figure 10 A distinction is made between process steps of the controls according to Figure 1 , characterized by a region CMP, which is separated by a dot-dash line from a measuring region MSN, in which various measuring devices according to Figure 1 be used.

[0125] After starting the procedure, data is first entered in an input step for data 2 (short DAT_IN), in which stored default values ​​for the control of the beam brakes BLK can be used, but also data on the currently running process are taken into account, which at least the wheel tire width and the running circle diameter are determined in a process-related measuring step 3 (short MSN_ABL).

[0126] This is followed by a calculation step using FEM 4 (CLC-FEM for short) to calculate the transverse forces acting in the beam track brake. This is followed by a calculation step for the friction coefficient and, if applicable, the roll factor of the identification 5 (OPT_ID for short) of a priori unknown process parameters. This is followed by a calculation step for the state-space model 6 (CLC_LRM for short) to account for the dynamics of the beam track brake system. Taking the previous calculation steps into account, a creation step for the model-predictive control 7 (CLC_CRL for short) is carried out to create a control concept taking into account the constraints described in more detail above.

[0127] With the model-predictive control concept, model-predictive control of the beam brakes BLK involved in the process is carried out in a control step for process 8 (CRL_ABL for short). This process is accompanied by a measurement step for process 9 (MSN_PRC for short). During this step, measured values ​​are created that describe the actual process. If the reported measured values ​​indicate a deviation from the calculated and thus expected process, the respective beam brake BLK is controlled by adjusting the process parameters, particularly the piston stroke. This is indicated by arrows in both directions between the measurement step for process 9 and the control step for the process.

[0128] Once the process has passed any beam brake, an optimized estimation step 10 (OPT_EST for short) follows. In this step, the calculated parameters are compared with the process parameters generated by the measurement, and an estimate for better input parameters is made for subsequent processes. If this estimation leads to a modification, the determined input parameters are stored in a setting step for new default values ​​11 (SET_DFT for short) for the input parameters.

[0129] This is followed by a query step for end of procedure 12 (STP for short). If the result is positive, i.e., when the last procedure has been completed, the procedure is stopped. Otherwise, the procedure begins again with a new input step for data 2 concerning the subsequent procedure. List of most important reference symbols

[0130] AALAblaufanlage CMPComputer BLKBemmerbremsen RDWheels S1 ... S7Interface ZSTCentral control device BBSUphill brake control TBSValley brake control RBSDirectional track brake control SNRail BGMountain peak ABRARun-off ramp ZNGIntermediate gradient VW1...VW6Distribution switches VTZDistribution zone RG1...RG8Directional tracks BBUphill brake relay BB2First uphill brake BB1Second uphill brake TBTValley brake relay TB1First valley brake TB2Second valley brake RGBDirectional track brake relay RB1First directional track brake RB2Second directional track brake RB3Third directional track brake RB4Fourth directional track brake RB5Fifth directional track brake RB6Sixth directional track brake RB7Seventh directional track brake RB8Eighth directional track brake AB1First Process AB2, second process AB3, third process ABL, push-off locomotive AP, push-off point TB1, TB2, valley brakes BB1, BB2, uphill brakes RB1...RB8 directional track brakes MST measuring station EL1 first inlet valve AL1 first outlet valve KM1 first chamber EL2 second inlet valve AL2 second outlet valve KM2 second chamber BZL actuating cylinder FP spring package KB piston GST rod ULK deflection ASL stop ABK outer beam IBK inner beam EL1...EL2 inlet valve AL1...AL2 outlet valve 2 input step for data 3 process-related measurement step 4 calculation step with application of FEM 5 optimization step of identification 6 calculation step for the state space model 7 creation step for model predictive control 8 control step for the process 9 measurement step for the process 10 step of an optimized estimate 11 definition step for new default values ​​12 query step for end of process.

Claims

1. Control method for operating a beam brake (BLK) which acts on wheels (RD) of a pay-off system (AAL) and in which a control variable for generating an actuating force for the beam brake (BLK) is determined for the pay-off system, characterized in thata) in a static analysis, a model of the transverse forces acting on the wheels (RD) in the brake beams of the beam brake (BLK) is calculated using a finite element method as a function of an actuating force acting on the brake beams by at least one actuator, b) in a dynamic analysis, a state space model is calculated, describing a process dynamics of the process by the temporal course of a process position as the first state variable and the temporal course of a process speed as the second state variable, as well as describing an actuation dynamics of the beam brake (BLK) as at least one further state variable, whereby the change in their respective time derivatives of all state variables is also taken into account,c) an optimum is calculated for the temporal course of the manipulated variable, taking into account the model for the beam brake (BLK) and the state-space model and based on the target run-out speed of the discharge from the beam brake (BLK) as a first boundary condition, d) control signals for the beam brake (BLK) are output, taking into account the optimum found for the temporal course of the manipulated variable.

2. Control method according to claim 1, characterized in that the control signals are used to perform model predictive control of the beam brake (BLK).

3. Control method according to claim 1 or 2, characterized in thatthe actuation dynamics for at least one actuation cylinder (BZL), each with a piston (KB), a first chamber (KM1) and a second chamber (KM2) and actuation valves is calculated taking into account at least one of the following further state variables, e) describing a piston dynamics of the piston (KB) by a piston position as the third state variable and by a piston speed as the fourth state variable, f) describing a pressure build-up dynamics in a first chamber (KM1) by a first chamber pressure in the first chamber (KM1) as the fifth state variable and by a second chamber pressure in the second chamber (KM2) as the sixth state variable, g) describing a valve dynamics, wherein each chamber has at least one inlet valve (EL1 ... EL2) and one outlet valve (AL1 ...AL2), by a respective valve position as the seventh state variable and ninth state variable and by a respective valve speed as the eighth state variable and tenth state variable.

4. Control method according to one of the preceding claims, characterized in that Based on a deviation of an actual acceleration of the process in the beam brake (BLK) from a target acceleration in the beam brake (BLK) calculated on the basis of the temporal course of the control variables, a predictive control of the control variable is additionally carried out.

5. Control method according to one of the preceding claims, characterized in that In addition, a second boundary condition aimed at preventing the wheels (RD) of the drain from climbing is taken into account.

6. Control method according to claim 5, characterized in thatas a second boundary condition, it is checked whether the difference between h) a torque resulting from a first braking force component which is generated by a front contact area of ​​a wheel bandage (RD) on the brake beam, seen in the direction of travel, and a height offset between a rail top side (SN) on which the wheel (RD) stands and the front contact area as minuends, and i) a torque resulting from a second braking force component which is generated by a rear contact area of ​​the wheel bandage (RD) on the brake beam, seen in the direction of travel, and a height offset between the rail top side (SN) and the rear contact area, as subtrahends, is greater, j) than a torque resulting from half the weight of an axle carrying the wheel (RD) and a horizontal offset between the axle and the front contact area.

7. Control method according to claim 6, characterized in thatwhen calculating the first braking force component and the second braking force component, a friction coefficient for the pairing between the wheel band (RD) and the brake beam and a roll factor to take into account the attack height in relation to the running circle diameter, whereby the product of the lateral force and the friction coefficient and the roll factor each results in the braking force component.

8. Control method according to one of the preceding claims, characterized in that In addition, a third boundary condition aimed at a uniform acceleration of the process while passing through the beam brake (BLK) is taken into account.

9. Control method according to one of the preceding claims, characterized in that In addition, a fourth boundary condition aimed at avoiding a jerk in the process is taken into account.

10. Control method according to one of the preceding claims, characterized in thatIn addition, a fifth boundary condition aimed at avoiding changes in the manipulated variable is taken into account.

11. Control method according to one of claims 5 to 10, characterized in that in addition to the first boundary condition, at least one further boundary condition is taken into account, whereby the boundary conditions taken into account are weighted.

12. Control method according to one of the preceding claims, characterized in that the product of the process parameters of the friction coefficient and the roll factor is calculated by subtracting a running resistance from the product of the mass of the run and the acceleration, taking into account at least one of the following running resistance components: inclination resistance, curve resistance, switch resistance, rolling resistance and air resistance, and dividing the difference obtained by the lateral force generated by the brake beam for braking.

13. Control method according to one of the preceding claims 1 to 11, characterized in that the process parameter of the friction coefficient is calculated by subtracting a driving resistance from the product of the mass of the run and the acceleration, taking into account at least one of the following driving resistance components: inclination resistance, curve resistance, switch resistance, rolling resistance and air resistance, and dividing the difference obtained by the lateral force generated by the brake beam for braking and by the roll factor determined for the wheel (RD).

14. Control method according to one of the preceding claims, characterized in that An optimized value is calculated using a signal filter taking into account the stochastic distribution of the measured values ​​of a speed measurement for the process and a stochastic distribution of the predicted values ​​for a speed curve of the process.

15. Control method according to one of the preceding claims, characterized in thatk) the time course of a braking operation l) as well as either the friction coefficient determined during the braking operation in question and the roll factor determined during the braking operation in question or the product of the friction coefficient and the roll factor determined during the braking operation in question are stored and used in a subsequent braking operation with the relevant beam brake (BLK) to take into account the braking characteristics or with another beam brake (BLK) to take into account the sequence characteristics of the sequence in question.

16. Control device for a beam brake (BLK), comprising a processor and a memory unit, characterized in that the control device is configured to carry out a method according to one of claims 1-15 with the aid of the processor by calling up data in the memory unit.

17. A computer program comprising program instructions which, when the program is executed by a computer (1), cause the computer (1) to carry out steps a) and b) and c) of the method according to any one of claims 1 to 15.

Images