Multi-size deceleration control device for a rail vehicle unit, multi-size deceleration control system for a rail vehicle or train consist, and method for deceleration control of a rail vehicle unit of a rail vehicle

Description

[0001] The present invention relates to a multi-variable delay control device for a rail vehicle unit, a multi-variable delay control system for a rail vehicle or a train consist comprising at least a first rail vehicle unit and a second rail vehicle unit, each comprising a multi-variable delay control device, and a method for delay control of a rail vehicle unit of a rail vehicle.

[0002] Deceleration control systems are used to brake rail vehicles, regulating the deceleration behavior of the respective rail vehicle units. Accordingly, a rail vehicle has at least one rail vehicle unit to which a predetermined deceleration control is applied. This rail vehicle unit can correspond to the rail vehicle itself or to related deceleration units, such as individual bogies with one or more braking devices, or individual braking devices for one or more bogies. The rail vehicle can, in turn, be part of a rail vehicle consist or train consist, which comprises several rail vehicles.

[0003] For deceleration control of individual rail vehicles, controllers are used, for example, which modify the requested deceleration and thus the braking force accordingly, thereby compensating for tolerances. Deceleration control is a function that should operate uniformly throughout the entire train consist to prevent impermissible longitudinal forces between the individual rail vehicles. Due to operational coupling requirements and the necessary signal communication, deceleration control may be limited to the respective rail vehicle or to a subunit of coupled rail vehicles.

[0004] Ideally, a delay control system for a train consisting of multiple rail vehicles would be centrally located so that all vehicles in the train are controlled simultaneously. However, such a centralized delay control system for multiple rail vehicles can result in long communication times between the vehicles, which can negatively impact the control dynamics. Furthermore, communication times can vary depending on the rail vehicle configuration. This places particular demands on a robust delay control system or on the ability to switch the delay control system. In addition, existing controllers, for example, at the coupling of two rail vehicles, must be partially deactivated, which necessitates a corresponding structural switching mechanism.

[0005] A deceleration control system for multiple rail vehicles in a train consist and / or for multiple rail vehicle units can also include independent, decentralized deceleration control systems for each rail vehicle and / or each rail vehicle unit. As already mentioned, the rail vehicle unit can correspond to a rail vehicle itself or to corresponding deceleration units. Particularly in cases of tolerances in the sensor signals used for the respective deceleration control systems in the individual rail vehicle units, the decentralized deceleration control systems of the rail vehicle units may, due to the system's inherent indeterminacy within the train consist and / or with multiple rail vehicle units within a single rail vehicle, potentially work against each other, and the control variables of the deceleration control system, which are fed to the control system to actuate the corresponding brake force generators, may drift apart.The control variable of each rail vehicle unit thus influences the braking forces of the individual brake force generators. These control variables can be derived from a predetermined target deceleration compared to a measured actual deceleration. The underdetermination of the system arises in this context from the fact that the necessary sum of braking forces required for a requested deceleration of the train consist or rail vehicle can be achieved through different braking force distributions in the individual rail vehicles or rail vehicle units. This distribution of braking forces is, in turn, influenced by the individual deceleration control systems in the rail vehicles or rail vehicle units. Therefore, theoretically, an infinite number of control variable vectors can be generated to achieve a required deceleration of the train consist or rail vehicle.In other words, different output variables from all deceleration control systems can ultimately produce the same overall deceleration, making the system indeterminate. The control variable of the rail vehicle, in turn, influences the braking forces of the individual braking devices. Drifting of the control variable in the rail vehicles or rail vehicle units can be prevented, for example, by tolerance bands. However, this reduces the control accuracy. Alternatively, a control variable limit can be implemented, but this can still lead to significantly different friction between wheel and rail, which is undesirable in most cases. As a further alternative, each decentralized deceleration control system can be provided with the same actual signals to mitigate the negative effects described above.However, this can in turn lead to similar adverse effects as those discussed regarding centralized deceleration control. Furthermore, temporary disturbances in the signals used to determine the actual deceleration can generate permanent deviations in the control variables, i.e., deviations in the control variables for the duration of braking, since this too is an underdetermined system. Depending on the measurement principle, such disturbances, in the sense of a deviation from the actual deceleration, can also depend on the slip of individual wheels or the gradient of the rails. The control of the train's deceleration and the distribution of forces among the individual rail vehicles or rail vehicle units are thus closely linked. If the train and / or rail vehicle is designed for potential frictional equilibrium, this distribution of forces between the individual rail vehicles or units could be affected.Rail vehicle units, for example, may be disrupted by the deceleration control.

[0006] For the control of a braking system for a rail vehicle, DE 10 2015 110053 A1 discloses a setpoint generator for the setpoint of the total braking force, a controller for determining at least one manipulated variable value based on this setpoint, and at least one actuator that transmits this manipulated variable value to at least one braking device. The system includes means for at least approximately determining acceleration components of the rail vehicle and means for determining the actual value of the total braking force from these acceleration components, taking into account the vehicle speed and mass. The controller is designed to regulate the manipulated variable value in such a way as to reduce the control deviation between the setpoint and the determined actual value.

[0007] In view of the foregoing, the present invention is based on the objective of providing a deceleration control for a rail vehicle unit, a deceleration system for a rail vehicle or a train consist, and a method for deceleration control of a rail vehicle unit of a rail vehicle, which enables an increase in the control quality compared to previous deceleration control systems.

[0008] The problem is solved by a multi-variable delay control device for a rail vehicle unit according to claim 1, a multi-variable delay control system for a rail vehicle or a train consist according to claim 10, and a method for delay control of a rail vehicle unit of a rail vehicle according to claim 14.

[0009] According to the invention, a multi-variable delay control device for a rail vehicle unit comprises a delay controller configured to determine a delay controller input for the rail vehicle unit and a control variable controller configured to determine a control variable input for the rail vehicle unit. The multi-variable delay control device is configured to form a total control variable for the rail vehicle unit from the delay controller input and the control variable input, and to supply the total control variable to a control loop for at least one braking force generation unit and to a feedback loop for the control variable controller.

[0010] The rail vehicle unit represents a unit to which at least one brake force generation unit, or at least one predetermined braking device of the brake force generation unit, is assigned. This braking device can be controlled via the multi-variable deceleration control device assigned to the rail vehicle unit. The rail vehicle unit can therefore correspond, for example, to a rail vehicle such as a railcar or wagon, or it can be a unit of a rail vehicle that is assigned to specific axles of the rail vehicle. The rail vehicle unit thus has at least one axle with an associated braking device. If several rail vehicle units, each with its own multi-variable deceleration control device, are provided in a rail vehicle or train consist, these each represent decentralized multi-variable deceleration control devices.In particular, these decentralized multi-variable deceleration control devices are each also arranged locally on or in their respective rail vehicle units. Alternatively, the multi-variable deceleration control device can also be arranged spatially separate from the rail vehicle unit and only functionally assigned to it. Accordingly, the multi-variable deceleration control device is not located in or on the respective rail vehicle unit, but can be part of a central control unit that is connected or connectable to the rail vehicle unit, a brake force generation unit belonging to the rail vehicle unit, or at least a predetermined braking device of the brake force generation unit. In addition to at least one predetermined braking device, the brake force generation unit can also include a brake force control loop.

[0011] The multi-variable delay control device includes a delay controller to which a target delay is specified. This target delay can correspond to, or be equivalent to, the target delay of the respective train consist. Alternatively, it can also represent the target delay of the corresponding rail vehicle or rail vehicle unit. Furthermore, the delay controller is provided with a calculated actual delay, which relates equally to the actual delay of the train consist, the rail vehicle, or the rail vehicle unit. In particular, the rail vehicle unit can determine its actual delay independently. However, to reduce the number of corresponding detection and / or evaluation devices and / or for plausibility checks, an overarching provision of the actual delay can also be used, either alternatively or additionally.The deceleration controller determines the deceleration control variable for the rail vehicle unit based on a comparison between the target and actual deceleration. This deceleration control variable represents a deceleration or corrected deceleration, a braking force, or another quantity representing the braking effect.

[0012] In addition to the delay controller, the multivariable delay control device includes the actuator controller, which determines the actuator control variable. The actuator control variable and the delay controller actuator are combined to form a total actuator. This total actuator is not only fed to another control loop for determining the controlled variable of the force generation unit, but also forms the basis for determining the setpoint and actual values ​​for the actuator controller. In other words, the setpoint of the actuator controller is derived from the total actuator, and the actual value is derived from the total actuator. Thus, the total actuator is fed back to the actuator controller and, consequently, to the multivariable delay control device.The total manipulated variable is therefore determined by both the delay controller manipulated variable and the manipulated variable controller manipulated variable, whereby at least the total manipulated variable again influences the manipulated variable controller manipulated variable.

[0013] The total manipulated variable can be calculated as the sum of the delay controller input and the manipulated variable input. Alternatively or additionally, the total manipulated variable can also be calculated by weighting the delay controller input and the manipulated variable input. For this purpose, for example, the delay controller input and the manipulated variable input can each be assigned a predetermined or variable weighting factor for the sum calculation and / or each can only be considered within a specific domain. The weighting can, for example, be determined indirectly by selecting the controller parameters of the multivariable delay control device.

[0014] In one embodiment, the control section comprises a total braking force determination unit configured to determine a total braking force of the rail vehicle unit based on the total control variable, an optional braking force distribution unit configured to transfer the total braking force into a braking force distribution, and a braking force control unit configured to control the braking force generation unit according to the total braking force or the braking force distribution.

[0015] The total force determination unit thus generates a total braking force for the rail vehicle unit that corresponds to the total control variable. If only one axle or wheel is controlled via the multi-variable deceleration control device, the total braking force can be supplied to a brake force control unit configured to control the brake force generation unit. If the rail vehicle unit has several axles or wheels to be controlled, the total braking force is first supplied to a brake force distribution unit configured to transfer the total braking force into a brake force distribution system. The brake force control unit then controls the brake force generation unit according to the brake force distribution system. The brake force distribution unit is therefore to be provided depending on the requirement for brake force distribution and is thus to be understood as optional.

[0016] In one embodiment, the feedback includes a net control variable determination unit configured to determine a net control variable based on the total control variable or the total control variable taking into account a reference control variable of the rail vehicle unit.

[0017] The total manipulated variable is therefore fed back to the actuator controller as an actual and / or setpoint value via the net manipulated variable determination unit, i.e., it is fed to the net manipulated variable determination unit, which determines a net manipulated variable from the total manipulated variable. The determination of the net manipulated variable can be carried out taking into account the reference manipulated variable of the rail vehicle unit. The reference manipulated variable serves as a benchmark for evaluating the magnitude of the manipulated variable. For example, even without the deceleration control, a specific manipulated variable would be set to achieve the desired braking effect of the brake force generation unit. The reference manipulated variable can thus relate the total manipulated variable to an uncontrolled manipulated variable. Depending on the total manipulated variable and / or the specific application, the reference manipulated variable can also be omitted or...be zero or correspond to the target delay or a feedforward control of the delay controller.

[0018] According to a further training, the feedback loop includes a filter unit configured to convert the net manipulated variable into a filtered net manipulated variable. In particular, the filter parameters of the filter unit, preferably at least a time shift, are adjustable.

[0019] The net control variable can be filtered, for example, to correlate it temporally with other control variables and / or influencing variables that may need to be considered. For this purpose, the filter unit can be designed as a simple dead-time element or at least contain a dead-time element. Alternatively or additionally, the filter unit can adjust or modify the net control variable in other ways, for example, to restrict the filtered net control variable to a specific range of values. For example, the filter unit can also be designed as a low-pass filter. The parameterization of the filter unit can be a fixed value, a fixed value with a tolerance, or defined by self-learning logic. The filter parameters themselves and / or the selection of the parameterization with regard to the above variants can be adapted. For example, different braking types, such as service braking or emergency braking, can be adjusted.This can also be advantageously supported by appropriately adapted parameterization, taking into account further boundary conditions such as weather conditions, track profile and other influencing factors on the resulting or anticipated braking behavior.

[0020] In a further training course, the feedback is configured to supply the net manipulated variable or the filtered net manipulated variable as the actual value to the actuator controller.

[0021] If the feedback loop does not have a filter unit, or if the filter unit is deactivated or does not perform signal conditioning, the net manipulated variable is supplied to the actuator as the actual value. Otherwise, the filtered net manipulated variable is supplied to the actuator as the actual value. Thus, the total manipulated variable, either as a derived net manipulated variable or as a filtered net manipulated variable, is supplied back to the actuator as an actual value and, taking into account the corresponding setpoint for the actuator, is used to determine the total manipulated variable.

[0022] In one embodiment, the feedback includes a normal control variable determination unit configured to transfer the total control variable, the net control variable or the filtered net control variable, in particular taking into account at least one further influencing variable, into a normal control variable, wherein the feedback is configured to supply the normal control variable as a setpoint to the control variable controller.

[0023] The at least one influencing factor can be, for example, at least one externally supplied disturbance variable and / or at least one control variable from another rail vehicle unit, in order to adjust the total control variable, the net control variable, or the filtered net control variable in the standard control variable determination unit to the at least one control variable of the other rail vehicle unit, and / or be determined by averaging the total control variables, the net control variables, and / or filtered net control variables over a predetermined period. The respective current total control variable, net control variable, or filtered net control variable can thus be related to the corresponding average value in the aforementioned averaging process.Similar to considering the control input of another rail vehicle unit, the standard control input is calculated from the current total control input, net control input, or filtered net control input in relation to the corresponding average value. For example, the total control input, net control input, or filtered net control input can be adjusted to the corresponding average value or assume predetermined values ​​depending on the average value. Various weighting factors can be assigned to the total control input, net control input, or filtered net control input and the at least one influencing factor to determine the standard control input. The at least one influencing factor can also be set to zero, so that the standard control input corresponds to the respective total control input, net control input, or filtered net control input.The multivariable delay control device can also be configured such that the normal actuator determination unit does not receive at least one additional input variable. In such a case, the normal actuator determination unit determines the normal actuator from the total actuator, the net actuator, or the filtered net actuator, for example, as a weighted total actuator, weighted net actuator, or weighted filtered net actuator, and / or by averaging over a predefined time period. The selection of the at least one influencing variable and / or weighting factors can depend on predetermined limit values ​​and / or current operating conditions, such as the braking type (service or emergency braking), external conditions (weather or track sections), or even the load status.

[0024] In conjunction with the previously described actual value determination for the actuator controller, the respective total manipulated variable is fed back to the actuator controller for both the actual value and the setpoint. The difference between the setpoint and the actual value, for example, the difference between the standard manipulated variable and the filtered net manipulated variable, can also be fed to the actuator controller via a separate control deviation unit. The control deviation is therefore calculated outside of the actuator controller.

[0025] If the normal control variable determination unit defines the net manipulated variable or filtered net manipulated variable as the normal control variable, for example, in a case where at least one influencing variable is set to zero, the setpoint supplied to the control variable can be equal to the actual value. In such a case, there is no control deviation at the control variable. However, the normal control variable determination unit can also set the normal manipulated variable to zero, so that the setpoint supplied to the control variable can also be zero. The setpoint can be calculated from the influencing variables in conjunction with the total manipulated variable, the net manipulated variable, or the filtered net manipulated variable. If at least one influencing variable is set to zero as a factor, the component related to that factor is also set to zero.Setting to zero can also occur if an influencing variable is not a factor, but setting it to zero is defined as a condition for setting the normal control variable to zero.

[0026] The normal control variable determination unit thus transfers the total control variable, the net control variable, or the filtered net control variable into the normal control variable, taking into account at least one additional influencing variable, which can also be set to zero. In other words, the normal control variable determination unit can, for example, transfer the input variables into the normal control variable according to a weighting of the input variables. For example, the normal control variable can be the sum of the weighted input values. In such a case, if at least one additional influencing variable is set to zero, the normal control variable can then be the weighted total control variable, the weighted net control variable, or the weighted filtered net control variable. Therefore, at least the weighted total control variable, the weighted net control variable, or the weighted filtered net control variable will always be used.In one variant, the normal control variable determination unit can also be configured such that, in a case where at least one other influencing variable is set to zero, the weighting of the total control variable, the net control variable, or the filtered net control variable is set from a predetermined value to 100%, or thus adopted as the normal control variable without weighting. In other cases, the predetermined weighting of the total control variable, the net control variable, or the filtered net control variable can be maintained regardless of the at least one other influencing variable. Weighting adjustments can also be made depending on driving operation or otherwise, as will be described later in connection with a multivariable delay control system according to the invention.The weighting adjustment options described there are transferable to the multivariable delay control device, provided they are not system-specific. Conversely, corresponding aspects of the multivariable delay control device are also transferable to the multivariable delay control system.

[0027] According to further training, the feedback is configured to feed the normal control variable from the normal control variable determination unit back to the filter unit described above.

[0028] In such a case, the filter unit can be configured, for example, to iteratively determine the filtering within certain limits to minimize the deviation from the normal control variable. To achieve this, the filter unit can, for instance, first calculate a possible range for the filtered net control variable. This can be done by calculating with different filter parameter settings within a predetermined tolerance. For each parameter setting, the deviation of the filtered net control variable from the normal control variable is determined. Based on this, the filter unit parameter setting with the minimum deviation is selected. Using this parameter setting, the filtered net control variable and the normal control variable are then calculated for further processing.The feedback of the normal control variable and corresponding configuration of the filter unit described here is equally applicable to the multivariable delay control system explained below.

[0029] In one embodiment, the delay controller and the manipulated variable controller are implemented using a single controller, in particular a PI controller.

[0030] When using a PI controller that includes both the delay controller and the manipulated variable controller, the control deviation is the sum of the control deviations of the respective individual controllers. Optionally, the single controller, in this example the PI controller, can include feedforward control.

[0031] According to one embodiment, the actuator controller has a static and / or dynamic dead zone.

[0032] The static dead zone allows the actuator controller to be configured so that it only intervenes in the control process when the control deviation lies outside the dead zone. Alternatively or additionally, the dead zone can be dynamically adjustable, allowing the actuator controller to exhibit different response behaviors depending on predetermined boundary conditions.

[0033] In a further aspect, the invention relates to a multi-variable delay control system for a rail vehicle or a train consist with at least a first rail vehicle unit and a second rail vehicle unit, each of which has a multi-variable delay control device comprising the previously described normal control variable determination unit, which is configured to transfer the net control variable or the filtered net control variable into a normal control variable, taking into account at least one influencing variable.The multivariable delay control system is configured to supply at least the net manipulated variable or filtered net manipulated variable of the other railway unit as at least one influencing variable to the respective normal control variable determination unit of the multivariable delay control device of the first railway unit and / or the second railway unit, and / or to the respective filter unit of the multivariable delay control device of the first railway unit and / or the second railway unit, at least the net manipulated variable or filtered net manipulated variable of the other railway unit for adjusting the filter parameters.

[0034] The multi-variable delay control system is therefore directed at a rail vehicle or train consist comprised of several rail vehicle units. According to the previously described multi-variable delay control device, which includes at least the first and second rail vehicle units, each of these units can initially be controlled independently. However, when several such rail vehicle units are combined, completely independent delay control for each unit can, under certain circumstances, lead to the delay control systems working against each other and the manipulated variables drifting apart due to the previously mentioned underdetermination of the system.The underdetermination of the system can be reduced by the aforementioned multivariable delay control system, since the net control variable or filtered net control variable of the multivariable delay control device of at least one rail vehicle unit is fed to the multivariable delay control device of at least one other rail vehicle unit. In particular, such feeding occurs reciprocally, preferably for all rail vehicle units of the rail vehicle or train consist.

[0035] With regard to the first and second rail vehicle units, each with its respective multivariable delay control device, the normal control variable of one rail vehicle unit can be determined based on the net control variable or filtered net control variable of at least one other rail vehicle unit. For example, the net control variable or filtered net control variable of the second rail vehicle unit can be fed to the normal control variable determination unit of the multivariable delay control device of the first rail vehicle unit as at least one of the influencing variables. Alternatively or additionally, the net control variable or filtered net control variable of the second rail vehicle unit can also be fed to the filter unit of the multivariable delay control device of the first rail vehicle unit to adjust the filter parameters.

[0036] The normal manipulated variable is fed into the manipulated variable controller as the setpoint. The actual value corresponds to the net manipulated variable itself, specifically the filtered net manipulated variable. Thus, at least the manipulated variable itself is fed back into the control system.

[0037] In one embodiment, the actuator of one rail vehicle unit can also include a dynamic dead zone that can be influenced by the net actuator or filtered net actuator of the other rail vehicle unit. The net actuator or filtered net actuator of the other rail vehicle unit can also be fed to the actuator of one rail vehicle unit directly or indirectly, for example via the filter unit and / or the normal actuator determination unit of the first rail vehicle unit.

[0038] Alternatively or additionally, the net control variable or filtered net control variable of the respective other rail vehicle unit can be supplied to the respective filter unit of the multi-variable delay control device of the first and / or the second rail vehicle unit for adjusting the filter parameters. Accordingly, the parameterization of the filter unit of one rail vehicle unit, and thus the net control variable filtered by it, can be adjusted via the net control variable or filtered net control variable of the respective other rail vehicle unit. In particular, the filter parameters can be dynamically adjusted within predetermined, and possibly adjustable, limits in order to further minimize deviations from the respective net control variables or filtered net control variables of the rail vehicle units.

[0039] The same target deceleration, for example, the target deceleration of the rail vehicle or train consist, can be specified for each of the respective delay controllers of the multivariable delay control devices of the first and second rail vehicle units. Each delay controller can also be provided with a determined actual deceleration. This can, for example, be determined independently for each rail vehicle unit. In such a case, a deviation in the actual deceleration measurements of the individual rail vehicle units does not affect the control accuracy with respect to a resulting average value. Only a deviation from the standard manipulated variable occurs. The effect of dynamic disturbances in the actual deceleration measurement on the manipulated variables is thus compensated for.In a special case where only the total manipulated variable, net manipulated variable, or filtered net manipulated variable of the respective rail vehicle unit is fed back to its own actuator controller—i.e., where the total manipulated variable, net manipulated variable, or filtered net manipulated variable is set to zero as an influence variable—a compromise between the manipulated variable deviation and the control deviation can be specifically achieved by parameterizing the actuator controller, even with differing actual delays. In particular, the actual delay for control in a rail vehicle unit can be calculated as a weighted average of the actual delays of several rail vehicle units. The weightings can be dynamically adjusted.

[0040] Alternatively or additionally, the actual delay can also be determined centrally, i.e., for the entire rail vehicle or train consist, and made available to the respective delay controllers. Such a supplementary determination of the actual delay can be provided for plausibility checks, redundancy reasons, and / or to account for different actual delays at the rail vehicle unit or rail vehicle / train consist level.

[0041] The respective delay controller can also include feedforward control. Feedforward control allows for a fast response, especially with large and dynamic setpoint changes, because in this case the adjustment of the actual value to the setpoint is not performed entirely by the multivariable delay control device or system, but at least partially by the multivariable delay control device or system. The multivariable delay control device or system then compensates for the feedforward deviations from a predefined model. This allows the system to react more quickly.

[0042] If each rail vehicle unit represents a rail vehicle, the multi-parameter deceleration control system is directed at a train consist. However, the multi-parameter deceleration control system can also be directed at a rail vehicle with at least two such rail vehicle units, either alternatively or additionally. The multi-parameter deceleration control system can thus coordinate the individual multi-parameter deceleration control devices of the multiple rail vehicle units of a rail vehicle and / or train consist with comparatively low requirements for signal communication between the rail vehicle units. Accordingly, for example, the braking force distribution and, consequently, compliance with traction requirements can be improved. The control quality can be achieved or at least approximated to that of a deceleration control system for a single rail vehicle unit. The control or...Control objectives for the deceleration of the rail vehicle or train consist and the distribution of braking force can be implemented and thus designed independently.

[0043] The described multivariable delay control system thus not only regulates the delay in the individual rail vehicle units, but also compensates for the deviation of its own net control variable from a standard control variable. In this way, the remaining degrees of freedom of the underdetermined system are bound up.

[0044] In one embodiment, the normal control variable determination unit of the multi-variable deceleration control device of the first rail vehicle unit and / or the second rail vehicle unit is configured to determine the normal control variable as a weighted average of its own net control variable or its own filtered net control variable and at least one net control variable of the other rail vehicle unit. The respective weighting factors correspond, in particular, to a number of bogies or cars of the corresponding rail vehicle unit, to a contribution of the corresponding rail vehicle unit to the total braking force of the rail vehicle, to the actual or nominal masses of the corresponding rail vehicle units, and / or to the masses of the corresponding rail vehicle units and a predetermined distribution of the masses.

[0045] The aforementioned weightings can therefore be selected depending on the respective influence of the corresponding rail vehicle unit on the overall deceleration. The weighting of at least one net control variable of the other rail vehicle unit can also be set to zero, so that in such a case only the unit's own total control variable, its own net control variable, or its own filtered net control variable is fed back to the control variable controller as the setpoint. Thus, for example, control by the control variable controller can be at least partially maintained even in the event of a communication failure between the rail vehicle units. Accordingly, it can be provided that a net control variable or filtered net control variable of the other rail vehicle unit is set to zero when no corresponding signal is present.Alternatively or additionally, the normal control variable can also be set to zero if, for example, all weights are set to zero. Alternatively or additionally, the normal control variable can also be set to zero if setting at least one of the weights to zero is defined as a condition for setting the normal control variable to zero. If the normal control variable is set to zero in such a case, only the system's own total control variable, its own net control variable, or its own filtered net control variable is fed back to the control variable controller.

[0046] The standard control variable in the standard control variable determination unit is thus determined from the net control variables or the filtered net control variables of the respective rail vehicle units, in particular at least the unit itself. For example, the standard control variable can be calculated from the weighted average of the net control variables or filtered net control variables. In one embodiment, for example, the filtered own net control variable and the unfiltered net control variables of the respective other rail vehicle unit(s) are used to calculate the average and thus the standard control variable.

[0047] If the normal control variable is based, for example, on a weighted average of the net control variables or filtered net control variables, the control variable controller can ensure that the net control variables equalize, thus minimizing the deviation. In this case, the delay controller is decoupled from the control variable controller; that is, the change in the total control variable is primarily determined by the delay controller.

[0048] However, aligning the net control variables does not necessarily mean that the actual braking forces are equal. The distribution of braking forces between and within the rail vehicle units via the control loop is freely selectable and is designed independently of the deceleration controller and the control variable controller. If the deviations of the control variables are reduced to zero, this means that the deceleration controller and the control variable controller do not influence the braking force distribution, which was designed according to specific criteria. This corresponds to the decoupling described above. In other words, while the deceleration controller and the control variable controller influence the overall control variable, they do not affect the control loop for the braking force distribution itself. The control loop for the braking force distribution is therefore designed independently.

[0049] If the rail vehicle unit corresponds to a rail vehicle, the braking force distribution between the rail vehicle units is such that each rail vehicle, considered individually, would achieve the same deceleration. Other brake force distribution options are also possible, but these may affect the weightings used in calculating the normal control variable.

[0050] The brake force distribution can also be located downstream of the multi-variable deceleration control system. For example, the total control variables of the respective rail vehicle units can be processed in a central evaluation unit and transferred into a brake force distribution or corresponding control signals, which are then transmitted to the respective brake force generation devices.

[0051] According to further training, the normal control variable determination unit is configured to adjust the respective weighting proportions, in particular to set predetermined weighting proportions to zero.

[0052] Accordingly, it is not necessarily required that all net control variables or filtered net control variables of one's own or other rail vehicle units be used to determine the standard control variable, either according to predetermined weightings or individually. In other words, the weighting and / or selection of net control variables or filtered net control variables can be dynamically adjusted to changing conditions or according to specific input methods. For example, the weighting of a net control variable or filtered net control variable of a rail vehicle unit can be reduced or set to zero when that unit is unloaded, resulting in a reduction in weight and a lesser influence of weight on deceleration.Preferably, at least the net manipulated variable or filtered net manipulated variable of each rail vehicle unit is considered in the normal manipulated variable determination unit for determining the normal manipulated variable. In such a case, at least the unit's own, possibly filtered, net manipulated variable is included in the determination of the manipulated variable controller variable via the manipulated variable controller. Thus, for example, if the requirements for control accuracy are temporarily lower, an exchange of the net manipulated variables or filtered net manipulated variables can be omitted. The rail vehicle units can therefore still be operated in a controlled manner even when decoupled. In particular, even the failure of a communication link between the rail vehicle units still allows for delay control of the individual rail vehicle units.

[0053] In one embodiment, the multivariable delay control system is configured to route at least one influencing variable to the normal control variable determination unit via the filter unit or another filter unit.

[0054] Accordingly, the net control variable or filtered net control variable of another rail vehicle unit, or the net control variables or filtered net control variables of other rail vehicle units, or averages derived from them, can not only influence the parameterization of the filter unit, but can also be fed to the normal control variable determination unit as a further filtered influencing variable or variables. Thus, the filter unit of the multi-variable delay control device of the respective rail vehicle unit can filter not only its own net control variable, but also the net control variable or the filtered net control variable, and then again, the latter of the other rail vehicle unit. The normal control variable is then derived from the filtered net control variables.Furthermore, the functionalities and configurations of the filter unit previously described for the multi-variable delay control device are also applicable to the net control variables of the user's own or other rail vehicle units and / or the parameterization of the filter unit. Regarding the filtering of the user's own and other net control variables, the parameterization for the respective net control variables can be different or adaptable to different levels. At least the user's own net control variable of the respective rail vehicle unit can be filtered to establish, as far as possible, a temporal correlation between its own net control variable and the other net control variable(s). As already mentioned regarding the multi-variable delay control device itself, the filter unit can incorporate a simple dead-time element for this purpose.

[0055] In another aspect, the present invention relates to a method for deceleration control of a rail vehicle unit of a rail vehicle, comprising the steps of: Determining a delay controller input variable via a delay controller, determining a manipulated variable controller input variable via a manipulated variable controller, and forming a total manipulated variable from the delay controller input variable and the manipulated variable controller input variable.

[0056] The total manipulated variable is fed to a control section for a brake force generation unit and fed back to the manipulated variable controller via a feedback loop.

[0057] The control path allows the total control variable of the brake force generation unit to specify a quantity representing a brake force or a quantity representing a brake force distribution.

[0058] The advantages and design possibilities of the method are analogous to the explanations regarding the multivariable delay control device.

[0059] In one embodiment of the method, the total manipulated variable is transferred via feedback into a net manipulated variable that can be set to a reference manipulated variable, in particular into a filtered net manipulated variable, which is supplied to the manipulated variable controller as an actual variable and is transferred in parallel via a normal manipulated variable determination unit, in particular to take into account at least one manipulated variable of another rail vehicle unit as an influencing variable, into a normal manipulated variable, which is supplied to the manipulated variable controller as a setpoint variable.

[0060] By taking into account at least one control variable of another rail vehicle unit as an influencing variable, in particular a total control variable, net control variable or filtered net control variable of the other rail vehicle unit, the control variable controller can easily adjust the control variable controller control variable taking into account the delay behavior of the at least one other rail vehicle unit.

[0061] Further advantages and design options of the method arise with regard to the consideration of the control variables of other rail vehicle units as influencing variables, analogous to the explanations for the multi-variable delay control system. Accordingly, the method can also be directed towards feeding the net control variable or filtered net control variable of another rail vehicle unit to a filter unit, as described above.

[0062] Particularly with regard to the multivariable delay control system or analogous procedural designs, the advantage arises that each delay controller or subsequent control section can be designed independently for the respective rail vehicle unit. Adaptation or consideration of other rail vehicle units is then achieved via the actuator controller, which, in the aforementioned variants, can access actuators from the other rail vehicle units.

[0063] The deceleration control and the distribution of braking forces can be designed independently of each other, thus decoupling them. This is achieved by constraining the degrees of freedom of the respective underdetermined system accordingly. The actuator controller can then ensure that the desired braking force distribution is not influenced by the deceleration control as much as possible.

[0064] Furthermore, long communication delays between the rail vehicle units or their variance do not affect the actual deceleration control. The actuator can therefore be easily designed to be tolerant of these delays.

[0065] The invention will now be explained in more detail with reference to the accompanying figures. The figures show, in detail: Figure 1 a schematic representation of a multi-variable delay control device for a rail vehicle unit according to an exemplary embodiment; Figure 2 a schematic representation of a multi-variable delay control system for a train consist according to an exemplary embodiment; Figure 3 a schematic representation of a dynamic disturbance of the measured quantities; Figure 4a schematic representation of the target and actual acceleration as well as the manipulated variables for an ideal system with independent controllers or a multivariable delay control system according to the invention; Figure 5 a schematic representation of the target and actual acceleration as well as the manipulated variables in the case of a static measurement error with independent controllers; Figure 6 a schematic representation of the target and actual acceleration as well as the manipulated variables in the case of a dynamic measurement error with independent controllers; Figure 7 a schematic representation of the target and actual acceleration as well as the manipulated variables in the case of a static measurement error in a multivariable delay control system according to the invention; and Figure 8 a schematic representation of the target and actual acceleration as well as the manipulated variables in the case of a dynamic measurement error in a multivariable delay control system according to the invention.

[0066] Figure 1Figure 1 shows a schematic representation of a multi-variable delay control device 10 for a rail vehicle unit 1 according to an exemplary embodiment. The rail vehicle unit 1 corresponds here to a rail vehicle, for example, a railcar or a wagon. Alternatively, the rail vehicle unit 1 can also be one of several rail vehicle units of a rail vehicle. The multi-variable delay control device 10 comprises a delay controller 20 and a manipulated variable controller 30. The delay controller 20 determines a delay controller manipulated variable ua,1 from a target delay a target and an actual delay a is,1. In the embodiment shown here, the actual delay a is,1 is determined by an actual delay measuring unit (not shown) of the rail vehicle unit 1.According to an alternative embodiment, the actual deceleration aist,1 can also be determined centrally for a rail vehicle composed of several rail vehicle units 1 or a train consist of several rail vehicle units 1 or rail vehicles. The actuator 30 determines an actuator control variable uc,1 from the control deviation between a setpoint and an actual value, specifically a normal actuator uz as the setpoint and a filtered net actuator as the actual value, as these will be described later. A total actuator u1 is formed from the sum of the deceleration actuator ua,1 and the actuator control variable uc,1. The total actuator u1 is fed to a total braking force determination unit 21, which converts the total actuator u1 into a total force Fges.The total force Fges is again fed to a brake force distribution unit 22, which divides the total force Fges into a brake force distribution F1 to Fn. The brake force distribution F1 to Fn corresponds to the brake force generating devices (not shown) of a brake force generating unit 40 ( ) to be assigned to the rail vehicle unit 1. Fig. 2 The control of the brake force generation devices or the brake force generation unit 40 is carried out via a brake force control unit 23, which transfers the brake force distribution F 1 to F n into respective control commands.

[0067] The total manipulated variable u1 is also fed through a further control loop to determine the setpoint and actual value for the manipulated variable controller 30. For this purpose, a net manipulated variable un,1 is first determined in a net manipulated variable determination unit 31 from the difference between the total manipulated variable u1 and a reference manipulated variable uref,1. The reference manipulated variable uref,1 thus forms a reference point for evaluating the magnitude of the total manipulated variable u1. In the exemplary embodiment, the reference manipulated variable uref,1 corresponds to the target delay a soll. According to alternative embodiments, the reference manipulated variable uref,1 can also correspond to an optional feedforward control of the delay controller 20, or be set to zero or omitted. The respective selection of the reference manipulated variable uref,1 can correspond to a preset and / or be adjustable according to the application, for example, depending on the braking type or external conditions.The net control variable un,1 is then fed to a filter unit 32. The filter unit 32 generates a filtered net control variable u nf,1 from the net control variable un,1. For this purpose, the filter unit 32, in the illustrated embodiment, has a dead-time element to establish a temporal correlation of the net control variable un,1 with subsequently to be considered influencing variables, here net control variables un,2 to un,n of other rail vehicle units. In alternative embodiments, the filter unit 32 can also be omitted or switched on as required, or activated with respect to a transformation of the net control variable un,1 into the filtered net control variable u nf,1. In addition, the net control variable un,1 is also output via another signal branch from the multi-variable delay control device 10 to other rail vehicle units or...Corresponding multivariable delay control devices can be provided for this purpose. However, according to alternative embodiments, such a signal branch can also be omitted.

[0068] The filtered net manipulated variable u nf,1 is provided to the manipulated variable controller 30 as an actual value for determining the manipulated variable controller uc,1. According to alternative embodiments, particularly in embodiments without filter unit 32 or with corresponding deactivation of the filter unit 32, the net manipulated variable un,1 can also be directly fed to the manipulated variable controller 30 as an actual value. Two switchable signal branches can also be provided for selectively forwarding the net manipulated variable un,1 or the filtered net manipulated variable u nf,1 to the manipulated variable controller 30, either directly or as a filtered net manipulated variable u nf,1 via the filter unit 32.

[0069] To generate the normal manipulated variable uz as the setpoint for the manipulated variable controller 30, the filtered net manipulated variable u nf,1 is additionally fed to the normal manipulated variable determination unit 33. According to the optional filter unit 32, in alternative embodiments the net manipulated variable un,1 can also be fed directly to the normal manipulated variable determination unit 33. In such cases, the following description applies analogously to the net manipulated variable un,1 instead of the filtered net manipulated variable u nf,1, provided it does not concern filter-specific aspects. The normal manipulated variable determination unit 33 determines the normal manipulated variable uz from the filtered net manipulated variable u nf,1, taking into account further influencing variables, here the net manipulated variables un,2 to un,n of other rail vehicle units.According to alternative embodiments, other influencing factors can be considered as an alternative or supplement to the net control variables un,2 to un,n of other rail vehicle units, or optionally omitted. In the exemplary embodiment, the normal control variable uz is formed from a weighted average of the filtered net control variable u nf,1 and the net control variables un,2 to un,n. The weighting of the filtered net control variable u nf,1 and the net control variables un,2 to un,n corresponds here to the respective weighting of the number of bogies of the respective rail vehicle units.In alternative embodiments, the weights can alternatively or additionally correspond to a typical value of the rail vehicle unit relative to the total braking force, the respective actual or nominal mass of the respective rail vehicle unit, and / or the respective actual or nominal mass of the respective rail vehicle unit in conjunction with a predetermined distribution between the rail vehicle units. The normal control variable determination unit 33 can be configured to adjust the respective weights. In particular, individual weights can also be set to zero. For example, if all weights of the net control variables un,2 to un,n of the other rail vehicle units are set to zero, the control variable controller 30 is supplied only with its own filtered net control variable u nf,1 as the setpoint.The multivariable delay control device can therefore also be operated independently of considering the net manipulated variables un,2 to un,n. This can result in a situation where, in a case where the normal manipulated variable determination unit 33 adopts the filtered net manipulated variable u nf,1 as the normal manipulated variable uz according to influencing variables not to be considered, the filtered net manipulated variable u nf,1 represents both the setpoint and the actual value for the manipulated variable controller 30. In such a case, no control deviation occurs, and the manipulated variable controller has no further influence on the total manipulated variable u 1. The consideration of influencing variables by the normal manipulated variable determination unit is adjustable in the illustrated embodiment. Accordingly, respective influencing variables are only considered if an adjustable, predetermined limit value is exceeded.According to alternative embodiments, at least one limit value can be fixed or a limit-related consideration can be omitted.

[0070] Furthermore, in the exemplary embodiment, the multivariable delay control device is configured such that the normal control variable uz can optionally be fed back from the normal control variable determination unit 33 to the filter unit 32. This optional feedback is indicated by the dashed arrow and occurs here depending on a time shift of the net control variables un,2 to un,n relative to the filtered net control variable u nf,1. According to alternative embodiments, such feedback can also be provided entirely or omitted. The parameterization of the filter unit 32 is adjusted via the feedback of the normal control variable uz. The parameterization is determined iteratively within limits such that the deviation of the filtered net control variable u nf,1 from the normal control variable uz is minimized.In a first step, filter unit 32 calculates a possible range for the filtered net control variable u nf,1. This calculation is performed using various parameter settings of filter unit 32 within a predetermined tolerance. For each parameter setting, the deviation of the filtered net control variable u nf,1 from the normal control variable uz is calculated. The parameter setting of filter unit 32 with the minimum deviation is then selected. Finally, this parameter setting is used to calculate the filtered net control variable u nf,1 and thus, in turn, the normal control variable uz for further processing.

[0071] Figure 2Figure 1 shows a schematic representation of a multivariable delay control system 100 for a train consist according to an exemplary embodiment. The train consist is formed here from a first rail vehicle unit 1, representing a first rail vehicle, and a second rail vehicle unit 2, representing a second rail vehicle. The first and second rail vehicle units 1, 2 each have a multivariable delay control device 10 according to Figure 1. Figure 1 To simplify the representation, the respective delay controllers 20, total braking force determination units 21, braking force distribution units 22, manipulated variable controllers 30 and net manipulated variable determination units 31 are represented by the area drawn with a dashed line.

[0072] Each of the delay controllers 20 is given the same target delay atarget. To determine the respective control deviation of the actual delay atarget,1 or atarget,2 with respect to the target delay atarget, or to determine the delay controller control variable uaa,1 or uaa,2 (not shown here), the actual delay atarget,1 determined by the first rail vehicle unit 1, is supplied to the delay controller 20 of the first rail vehicle unit 1. Simultaneously, the actual delay atarget,2 determined by the second rail vehicle unit 2, is supplied to the delay controller 20 of the second rail vehicle unit 2. For each of the rail vehicle units 1, 2, the respective total control variable u 1 or u 2 (not shown here) is then formed from the sum of the respective delay control variable ua,1 or ua,2 and the respective control variable control variable uc,1 or uc,2 (not shown here).As already described for the aforementioned multi-variable deceleration control device 10, the respective total control variable u1 or u2 is supplied to the respective brake force control units 23 via their respective total brake force determination units 21 and brake force distribution units 22 as brake force distributions F1.1 to Fn.1 for the first rail vehicle unit 1 and as brake force distributions F1.2 to Fn.2 for the second rail vehicle unit 2, respectively. The brake force control units 23 of the first and second rail vehicle units 1 and 2 control their respective assigned braking devices via a brake force generation unit 40. Here, the brake generation unit 40 is designed as a single brake generation unit 40 that can be controlled by both rail vehicle units 1 and 2.In alternative embodiments, the first and second rail vehicle units 1, 2 can each have their own brake generation unit 40, which can be controlled via the respective brake force control units 23.

[0073] The multivariable delay control device 10 of the first rail vehicle unit 1 supplies its own net manipulated variable un,1 to the filter unit 32 of the first rail vehicle unit 1. Furthermore, the net manipulated variable un,1 of the first rail vehicle unit 1 is also forwarded to the normal manipulated variable determination unit 33 of the second rail vehicle unit 2. Similarly, the multivariable delay control device 10 of the second rail vehicle unit 2 supplies its own net manipulated variable un,2 to the filter unit 32 of the second rail vehicle unit 2. Likewise, the net manipulated variable un,2 of the second rail vehicle unit 2 is also forwarded to the normal manipulated variable determination unit 33 of the first rail vehicle unit 1.Accordingly, the normal control variable uz of each rail vehicle unit 1, 2 in the respective normal control variable determination unit 33 is formed by its own filtered net control variable u nf,1 , u nf,2 and the net control variable un,2 , un,1 of the other rail vehicle unit 2, 1. According to alternative embodiments, the normal control variable uz of each rail vehicle unit 1, 2 in the respective normal control variable determination unit 33 can be formed by its own filtered net control variable u nf,1 , u nf,2 and the filtered net control variable u nf,2 , u nf,1 of the other rail vehicle unit 2, 1. The net control variables un,2 , un,1 of the respective other rail vehicle unit 2, 1 can be filtered for this purpose via the respective own filter unit 32 of the other rail vehicle unit 2, 1 and / or the filter unit 32 of the rail vehicle unit 1, 2 to which they are supplied.

[0074] The standard control variable uz1 or uz2 is calculated as the weighted average of its own filtered net control variable unf,1 or unf,2 and the net control variable un,2 or un,1 of the other rail vehicle unit 1, 2. In this exemplary embodiment, the weighting corresponds to the number of bogies of the respective rail vehicle units, which is the same for both rail vehicle units 1, 2. Thus, the standard control variable uz1 of the first rail vehicle unit is uz1 = (unf,1 + uf,2) / 2, and the standard control variable uz2 of the second rail vehicle unit is uz2 = (unf,2 + uf,1) / 2.

[0075] The normal manipulated variable uz1 is transmitted as a setpoint to the manipulated variable controller 20 of the first rail vehicle unit 1, and the normal manipulated variable uz2 is transmitted as a setpoint to the manipulated variable controller 20 of the second rail vehicle unit 2. The actual value corresponds to the respective filtered net manipulated variable unf,1 or unf,2. Thus, at least the respective net manipulated variable unf,1 or unf,2, specifically the filtered net manipulated variable unf,1 or unf,2, is fed back to the delay control, specifically to the respective manipulated variable controller 30. The described procedure therefore not only controls the delay in the individual rail vehicle units 1 and 2, but also compensates for the deviation of the respective filtered net manipulated variable unf,1 or unf,2 from the normal manipulated variable uz1 or uz2. In this way, the remaining degrees of freedom of the underdetermined system are constrained.The system can therefore exhibit a defined behavior, at least in the case of quasi-static disturbances.

[0076] The advantage of the above-described multivariable delay control system 100 for two rail vehicle units is explained below in comparison to a delay control system in which only one delay controller is used per rail vehicle unit, i.e., no feedback via a manipulated variable controller takes place.

[0077] In the case of a static measurement error with respect to the determined actual decelerations a_target,1 and a_target,2 of the first and second rail vehicle units 1, 2, an offset occurs in a system without feedback. In other words, the manipulated variables generated by the respective deceleration controller drift apart. Consequently, the actual braking force distribution deviates from the desired braking force distribution, or the deceleration control distorts the braking force distribution. This can lead to a violation of the required traction conditions. In the case of a dynamic measurement error, a permanent deviation of the manipulated variables occurs in the system without feedback. Therefore, the braking force distribution is uncontrollable, and a violation of traction conditions cannot be ruled out.

[0078] Under ideal conditions, i.e., without measurement errors, the multivariable delay control system according to the invention initially shows no difference compared to a system without feedback. However, if a static measurement error occurs, the difference between the respective total manipulated variables u1 and u2 is only slight. This difference is directly dependent on the difference in the respective delay measurement. In the case of a dynamic measurement error, the disturbance can be completely corrected or at least reduced to a tolerable level. No permanent deviation of the total manipulated variables u1 and u2 occurs.

[0079] The advantages of a multivariable delay control system according to the invention, such as the multivariable delay control system 100 according to the described exemplary embodiment, are shown by the Figures 3 to 8 different progressions of acceleration signals.

[0080] Figure 3This is first shown in general terms as a schematic representation of a dynamic disturbance of the measured variables. The manipulated variables du 1 and du 2 drift apart.

[0081] Figure 4Figure 1 shows a schematic representation of the target acceleration atarget and the actual acceleration aactual, as well as the manipulated variables u1 and u2 for an ideal system with independent controllers or a multivariable delay control system according to the invention, such as the multivariable delay control system 100. In an ideal system, it is assumed that no measurement inaccuracies occur. The manipulated variables u1 and u2 behave uniformly and regulate the actual acceleration aactual to the target acceleration atarget. In practice, however, dynamic and static measurement errors repeatedly occur, which, with independent controllers, can cause a deviation of the manipulated variables u1 and u2 from each other. This makes the braking force distribution uncontrollable and can lead to a failure to comply with the frictional engagement conditions.

[0082] Figure 5Figure 1 shows a schematic representation of the target acceleration atarget and the actual acceleration aactual, as well as the manipulated variables u1 and u2, with a static measurement error for independent controllers. Due to the underdetermination of the system with independent controllers, the manipulated variables u1 and u2 drift apart for the respective brake force generation units being controlled. This also causes the frictional relationships to diverge, so that compliance with specified frictional conditions can no longer be guaranteed. Furthermore, this can also lead to excessive wear on the brake force generation units.

[0083] Additionally, it shows Figure 6A schematic representation of the target acceleration atarget and actual acceleration aactual, as well as the manipulated variables u1 and u2, is shown in the context of a dynamic measurement error with independent controllers. Here, the underdetermination of the independent controller system leads to a permanent deviation of the manipulated variables u1 and u2 of the respective brake force generation units being controlled. Correspondingly, permanent deviations also occur in the frictional relationships, which can likewise result in a violation of the specified frictional relationship conditions. Furthermore, excessive wear can also occur on the respective brake force generation unit under greater stress.

[0084] The multi-variable delay control system according to the invention, such as the multi-variable delay control system 100, can reduce the drift apart of the total control variables u 1 and u 2 for the respective brake force generation units to be controlled. Figure 7Figure 1 shows a schematic representation of the target acceleration atarget and actual acceleration aactual, as well as the total manipulated variables u1 and u2 with a static measurement error in a multi-variable deceleration control system according to the invention. This is in comparison to the behavior of the manipulated variables u1 and u2 in independent controllers for the brake force generation units, as in Figure 2. Figure 5As shown, in the multi-variable deceleration control system according to the invention, the total control variables u1 and u2 no longer drift apart in the event of a static measurement error, but instead run essentially parallel to each other with a comparatively small difference. The frictional contact conditions are therefore also essentially comparable. As a result, compliance with predefined frictional contact conditions can be ensured. Likewise, wear peaks of individual brake force generation units are reduced, and wear on the respective brake force generation units or brake components connected to them can be set at least approximately equal, unless a controlled brake force distribution dictates otherwise.

[0085] The multivariable delay control system according to the invention has an equally positive effect on dynamic measurement errors. Figure 8Figure 1 shows a schematic representation of the target acceleration atarget and actual acceleration aactual, as well as the total control variables u1 and u2, in the event of a dynamic measurement error in a multi-variable deceleration control system according to the invention. The dynamic measurement error leads only to a short-term drift between the total control variables u1 and u2, which are then realigned by the multi-variable deceleration control system according to the invention. No permanent offset of the total control variables u1 and u2 remains. Accordingly, the force transmission ratios resulting from the control variables u1 and u2 are essentially the same for the respective brake generation units after the dynamic measurement error has been corrected. In other words, the disturbance caused by a dynamic measurement error is completely corrected, and no permanent deviation of the total control variables u1 and u2 occurs.

[0086] The invention is not limited to the described embodiments.

[0087] In particular, features of the multi-variable deceleration control device and the multi-variable deceleration control system are directly transferable to corresponding process steps for the respective deceleration control and vice versa. The multi-variable deceleration control device and the multi-variable deceleration control system can, in principle, be used in any type of braking, for example, for service braking and / or emergency braking, and regardless of the type of brakes used, such as electrodynamic, electromechanical, hydraulic, or pneumatic brakes. REFERENCE MARK LIST

[0088] 1 First rail vehicle unit 2 Second rail vehicle unit 10 Multi-variable deceleration control device 20 Deceleration controller 21 Total braking force determination unit 22 Braking force distribution unit 23 Braking force control unit 30 Actuator controller 31 Net actuator determination unit 32 Filter unit 33 Normal actuator determination unit 40 Braking force generation unit 100 Multi-variable deceleration control system a target deceleration a actual,1 Actual deceleration (first rail vehicle unit) a actual,2 Actual deceleration (second rail vehicle unit) F 1 , ..., F n Braking force distribution (rail vehicle unit) F 1,1 , ..., F n,1 Braking force distribution (first rail vehicle unit) F 1,2 , ..., F n,2 Braking force distribution (second rail vehicle unit) F ges Total braking force (rail vehicle unit) u 1 Total control variable (first rail vehicle unit) ua,1 Deceleration controller control variable (first rail vehicle unit) uc,1 Control variable controller control variable (first rail vehicle unit) un,1 Net control variable (first rail vehicle unit) un,2 Net control variable (second rail vehicle unit) un,n Net control variable (nth rail vehicle unit) u nf,1 Filtered net control variable (first rail vehicle unit) u nf,2 Filtered net control variable (second rail vehicle unit) u ref,1 Reference control variable (first rail vehicle unit) uz Normal control variable.

Claims

1. A multivariable deceleration control device (10) for a rail vehicle unit (1), having: a deceleration controller (20) configured to determine a deceleration controller manipulated variable (ua,1) for the rail vehicle unit (1), and a manipulated variable controller (30) configured to determine a manipulated variable controller manipulated variable (uc,1) for the rail vehicle unit (1), wherein the multi-variable deceleration control device (10) is configured to form a overall manipulated variable (u1) for the rail vehicle unit (1) from the deceleration controller manipulated variable (ua,1) and the manipulated variable controller manipulated variable (uc,1), and supply the overall manipulated variable (u1) to a controlled system (21,22, 23) for at least one brake force generating unit (40) and a return line (31, 32, 33) to the manipulated variable controller (30).

2. The multivariable deceleration control device (10) according to claim 1, wherein the controlled system (21, 22, 23) has a overall braking force determination unit (21) configured to determine a overall braking force (Fges) of the rail vehicle unit (1) based on the overall manipulated variable (u1), an optional brake force distribution unit (22) configured to transfer the overall braking force (Fges) to a braking force distribution (F1, ..., Fn), and a brake force control unit (23) configured to control the brake force generating unit (40) according to the overall braking force (Fges) or the brake force distribution (F1, ..., Fn)3. The multivariable deceleration control device (10) according to claim 1 or 2, wherein the return line (31, 32, 33) has a net manipulated variable determination unit (31), which is configured to determine a net manipulated variable (un,1) based on the overall manipulated variable (u1) or the overall manipulated variable (u1), taking into account a reference manipulated variable (uref,1) of the rail vehicle unit (1).

4. The multivariable deceleration control device (10) according to claim 3, wherein the return line (31, 32, 33) has a filter unit (32), which is configured to transfer the net manipulated variable (u n,1) into a filtered net manipulated variable (unf,1), and wherein in particular filter parameters of the filter unit (32), preferably at least onw time shift, are adaptable.

5. The multivariable deceleration control device (10) according to claim 3 or 4, wherein the return line (31, 32, 33) is configured to supply to the manipulated variable controller (30) the net manipulated variable (un,1) or the filtered net manipulated variable (unf,1) as the actual value.

6. The multivariable deceleration control device (10) according to any one of the preceding claims, wherein the return line (33, 32, 33) has a normal manipulated variable determination unit (33), which is configured to transfer the overall manipulated variable (u1), the net manipulated variable (un,1) or the filtered net manipulated variable (unf,1), in particular under consideration of at least one further influencing variable (un,2, ... un,n), to a normal manipulated variable (uz), and the return line (31, 32, 33) is configured to supply the normal manipulated variable (uz) as the setpoint to the manipulated variable controller (30).

7. The multivariable deceleration control device (10) according to claim 6, wherein the return line (31, 32, 33) is configured to supply the normal manipulated variable (uz) from the normal manipulated variable determination unit (33) again to the filter unit (32) according to claim 4.

8. The multivariable deceleration control device (10) according to any one of the preceding claims, wherein the deceleration controller (20) and the manipulated variable controller (30) are formed by a single controller, in particular a PI controller, which takes into consideration both the deviation of the deceleration from the target deceleration and also the deviation of the manipulated variable from the normal manipulated variable.

9. The multivariable deceleration control device (10) according to any one of the preceding claims, wherein the manipulated variable controller (30) has a static and / or dynamic dead zone.

10. A multivariable deceleration control system (100) for a rail vehicle or a train set having at least one first rail vehicle unit (1) and a second rail vehicle unit (2), each having a multivariable deceleration control device (10) according to any one of claims 6 to 9, wherein the multivariable deceleration control system (100) is configured to: supply to the respective normal manipulated variable determination unit (33) of the multivariable deceleration control device (10) of the first rail vehicle unit (1) and / or of the second rail vehicle unit (2) at least the net control variable (u n,1, un,2) or the filtered net control variable (unf,1, unf,2) of the respective other rail vehicle unit (1.2) as at least one influencing variable and / or to the respective filter unit (32) of the multivariable deceleration control device (10) of the first rail vehicle unit (1) and / or of the second rail vehicle unit (2) at least the net manipulated variable (un,1, un,2) or the filtered net manipulated variable (unf,1, unf,2) of the respective other rail vehicle unit (1, 2) for adapting the filter parameters11. The multivariable deceleration control system (100) according to claim 10, wherein the normal manipulated variable determination unit (33) of the multivariable deceleration control device (10) of the first rail vehicle unit (1) and / or of the second rail vehicle unit (2) is configured to determine the normal manipulated variable (u z) as the weighted average of the ego net manipulated variable (un,1, un,2) or the ego filtered net manipulated variable (unf,1, unf,2) and the at least one net manipulated variable (un,2, un,1) of the respective other rail vehicle unit (1, 2), and wherein the respective weighting components in particular correspond to a number of trucks or carriages of the corresponding rail vehicle unit (1, 2), to a contribution of the corresponding rail vehicle unit (1, 2) to the overall brake force of the rail vehicle, to the current or nominal masses of the corresponding rail vehicle units (1, 2), and / or to the masses of the corresponding rail vehicle units (1, 2) and a predetermined distribution of the masses.

12. The multivariable deceleration control system (100) according to claim 11, wherein the normal manipulated variable determination unit (33) is configured to adapt the respective weighting components, in particular to set predetermined weighting components to zero.

13. The multivariable deceleration control system (100) according to any one of claims 10 to 12, wherein the multivariable deceleration control system (100) is configured to guide the at least one influencing variable for supplying to the normal manipulated variable determination unit (33) via the filter unit (32) or a further filter unit.

14. A method for deceleration control of a rail vehicle unit (1) of a rail vehicle, comprising the steps: - determining a deceleration controller manipulated variable (ua,1) via a deceleration controller (20), - determining a manipulated variable controller manipulated variable (uc,1) via a manipulated variable controller (30) and - forming a overall manipulated variable (u1) from the deceleration controller manipulated variable (ua,1) and the manipulated variable controller manipulated variable (uc,1), wherein the overall manipulated variable (u1) is supplied to a controlled system (21,22, 23) for a brake force generating unit (40) and supplied back to the manipulated variable controller (30) via a return line (31, 32, 33).

15. The method according to claim 14, wherein the overall manipulated variable (u1) is transferred via the return line (31, 32, 33) into a net manipulated variable (un, 1) that can be set to a reference manipulated variable (uref,1), in particular into a filtered net manipulated variable (unf,1), which is supplied to the manipulated variable controller (30) as an actual variable and is transferred in parallel via a normal manipulated variable determination unit (33), in particular to take into account at least one manipulated variable of another rail vehicle unit (2) as an influencing variable, further transferred into a normal manipulated variable (uz), which is supplied to the manipulated variable controller (30) as a setpoint variable.