Device and computer-implemented method for moving first moving object at traffic node, and moving object comprising such device

Abstract

An apparatus and computer-implemented method of moving a first moving object at a traffic node: determining a state of the first moving object, including a location, velocity, or acceleration; receiving a trajectory of a second moving object at the traffic node including a state of location, velocity or acceleration; determining the limitation of the state of the first moving object according to the track of the state of the second moving object; determining a trajectory of the state of the first moving object starting from the state of the first moving object using a model configured to predict the trajectory of the state of the first moving object from the trajectory of the manipulation signal moving the first moving object and the state of the first moving object; determining a trajectory of the manipulation signal and a trajectory of the state of the first moving object within the limit according to a cost function defining a target of moving the first moving object with the trajectory of the manipulation signal on the trajectory of the state of the first moving object; the first moving object is moved at the traffic node using the control signal in the trajectory of the control signal.

Description

Background Art

[0001] The invention is based on a device and a computer-implemented method for moving a first mobile object at a traffic node and a mobile object comprising the device. The model predictive conditioning of the object uses an optimizer in order to determine a trajectory of the object's movement. Summary of the invention

[0002] The movement of a first mobile object at a traffic node is optimized by the apparatus and computer-implemented method for moving a first mobile object at a traffic node according to the independent claims. The method does not rely on the function of an optimizer to avoid a collision with another object.

[0003] The computer-implemented method provides for determining a state of a first mobile object, the state comprising a position, a velocity or an acceleration of the first mobile object, wherein a trajectory of a state of a second mobile object at a traffic node is received, the state comprising a position, a velocity or an acceleration of the second mobile object, wherein limits for the state of the first mobile object are determined based on the trajectory of the state of the second mobile object, wherein the trajectory of the state of the first mobile object is determined starting from the state of the first mobile object using a model, the model being constructed to predict the trajectory of the state of the first mobile object based on a trajectory of a control signal for moving the first mobile object and based on the state of the first mobile object, wherein the trajectory of the control signal and the trajectory of the state of the first mobile object are determined within the limits for the state of the first mobile object based on a cost function, wherein the cost function defines a target for moving the first mobile object using the trajectory of the control signal on the trajectory of the state of the first mobile object, wherein the first mobile object is moved at the traffic node using a control signal from the trajectory of the control signal.

[0004] In one specific embodiment, limits are predefined for the actuation signal for the first mobile object, wherein the actuation signal is determined within the limits for the actuation signal.

[0005] In one embodiment, a setpoint state at the end of a trajectory of the state of the first mobile object is predefined, wherein a trajectory of the state of the first mobile object is determined, said trajectory including the setpoint state at the end of the trajectory. In one embodiment, a permissible set of states at the end of the trajectory of the state of the first mobile object is predefined, wherein a trajectory of the state of the first mobile object is determined, in the case of which the state at the end of the optimization range is in the permissible set.

[0006] It can be provided that a disturbance variable for the state of the second mobile object is received, wherein the limitation for the state of the first mobile object is determined based on the disturbance variable for the state of the second mobile object. This means that the uncertainty of the trajectory of the state of the second mobile object occurring in the real world is also taken into account. This avoids collisions due to disturbances of the trajectory of the state of the second mobile object occurring in the real world.

[0007] It can be provided that the model includes disturbance variables for the state of the first mobile object. This means that the uncertainty relative to the ideal model of the first mobile object is also taken into account. This avoids collisions even in the case of trajectory disturbances of the state of the first mobile object that occur in the real world.

[0008] In one specific embodiment, a sequence is predefined in which the first mobile object and the second mobile object are to be moved at the traffic node, wherein a trajectory of the state of the first mobile object is determined, which trajectory complies with the sequence.

[0009] In one embodiment, the trajectory of the state of the first mobile object is determined based on the trajectory of the state of mobile objects that are located in a predetermined area around the traffic node. This means that the method jointly considers the mobile objects that are located in the predetermined area.

[0010] In one embodiment, when there is no other moving object between the first moving object and the second moving object, the first moving object and the second moving object move toward the traffic node in the same direction on one lane. Optionally, cooperative behavior of the moving objects moving sequentially on the lane can be implemented.

[0011] In one embodiment, the first mobile object and the second mobile object move toward the traffic node in the same direction on lanes adjacent to each other or on lanes intersecting at the traffic node. Optionally, cooperative behavior of the moving objects that arrive at the traffic node in sequence can be implemented.

[0012] The device for moving the first mobile object is designed to carry out the method and has advantages corresponding to those of the method.

[0013] A mobile object comprising the device has corresponding advantages.

[0014] A computer program comprising instructions executable by a computer has advantages corresponding to those of the method, which is executed when the instructions are executed by a computer. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Further advantageous embodiments can be found in the following description and the accompanying drawings. In the drawings:

[0016] Figure 1 a schematic diagram showing a mobile object having a device for moving the mobile object,

[0017] Figure 2 A schematic diagram showing an exemplary traffic node with moving objects,

[0018] Figure 3 A flow chart showing the steps of a method for moving a mobile object,

[0019] Figure 4 The merging process according to the first embodiment is shown.

[0020] Figure 5 A merging process according to the second embodiment is shown. DETAILED DESCRIPTION

[0021] Figure 1 A schematic diagram of a mobile object 100 with a device 102 for moving the mobile object 100 is shown.

[0022] Device 102 is designed to carry out a method for moving mobile object 100 .

[0023] In this example, the device 102 includes a computing device 104 designed to execute the method, in particular at least one processor and at least one memory.

[0024] The device 102 includes a receiver 106 which is designed to communicate with other mobile objects 100. The device 102 includes an actuator 108 which is designed to control the mobile object 100 in order to move the mobile object 100.

[0025] The mobile object 100 is a technical system. A technical system is a physical system, such as a mobile robot, in particular an at least partially autonomous vehicle, aircraft or ship. The vehicle may be a car or a transport vehicle of an industrial facility, such as a manufacturing facility.

[0026] exist Figure 2 A schematic diagram of an exemplary traffic node 200 with moving objects is shown in FIG.

[0027] In this example, a first vehicle 202, a second vehicle 204, a third vehicle 206, a fourth vehicle 208, and a fifth vehicle 210 are shown as moving objects. The first vehicle 202, the third vehicle 206, and the fifth vehicle 210 move in sequence on a first lane 212 toward the traffic node 200. No other vehicles are arranged between the first vehicle 202, the third vehicle 206, and the fifth vehicle 210.

[0028] Second vehicle 204 and fourth vehicle 208 move sequentially toward traffic node 200 on second lane 214 extending parallel to first lane 212. No other vehicles are arranged between second vehicle 204 and fourth vehicle 208.

[0029] In this example, the parallel running lanes taper into a common lane 216 at the traffic node 202. The vehicle is located in an area around the traffic node 200. In this example, the area starts at an entrance 218 in front of the traffic node 200 in the direction of movement of the vehicle, and ends at an exit 220 behind the traffic node 200 in the direction of movement of the vehicle.

[0030] In addition to lanes running in parallel, traffic node 200 may also include an intersection of two lanes.

[0031] Vehicles are examples of mobile objects 100. In this example, the vehicles each include a device 102.

[0032] Devices 102 are designed to communicate with one another and to execute corresponding parts of a decentralized networked control and / or regulating method for traffic node management.

[0033] The goals of traffic node management can be to improve safety, throughput, energy efficiency, and comfort.

[0034] The traffic nodes are designed without a central entity that receives information about all vehicles in order to solve control problems centrally based on this and then control the vehicles accordingly. The control technology problems are solved in a distributed manner on the individual vehicles. This saves costs or, in the first place, enables the corresponding functions to be implemented completely at remote traffic nodes without power supply. The device 100 is, for example, a control device installed in the corresponding vehicle.

[0035] The transportation node 200 may be located in a warehouse, a production workshop, an airport or a port.

[0036] The method is described using the example of controlling autonomous vehicles within the scope of a cooperative merging process at traffic node 200 , wherein the vehicles merge in a predefined sequence.

[0037] In this example, the order is defined by the time point at which the vehicle passes through the entrance 218 of the zone. The order may also be determined by other methods.

[0038] In this example, the first vehicle 202 passes through the entrance 218 first, followed by the second vehicle 204 , the third vehicle 206 , the fourth vehicle 208 , and the fifth vehicle 210 , in that order.

[0039] In this example, on the one hand, it is about realizing the merging process efficiently. This can mean, for example, that a good throughput of vehicles through the traffic node 200 should be achieved or that the merging process should be carried out in an energy-saving manner. The goal or goals are mathematically described by a cost function. In this example, the cost function is considered in the optimization problem.

[0040] In this example, it is provided that physical limits of speed or acceleration are adhered to and a collision of the vehicles is avoided.

[0041] The speed and acceleration of the vehicle are, for example, subject to restrictions.

[0042] For example, these constraints are the speed v of vehicle i for the optimization problem i or acceleration u i Constraints:

[0043] v min ≤v i ≤v max

[0044] u min ≤u i ≤u max

[0045] In this example, it is provided that a minimum distance d is maintained for vehicles in the same lane. min The position s of vehicle i at the calculated time point k on the trajectory for that vehicle i (k) and the position s of the vehicle traveling in front of the vehicle at the calculated time point k i- (k) The minimum distance d between min For example, define constraints

[0046] s i- (k)-s i (k)≥d min

[0047] The position s of vehicle i at the calculated time point k on the trajectory for that vehicle i (k) and the position s of the vehicle following the vehicle at the calculated time point k i+ (k) The minimum distance d between min For example, define constraints

[0048] s i (k)-s i+ (k)≥d min

[0049] In this example, it is provided that the minimum distance d is to be observed for vehicles in different lanes only from the traffic node 200, ie the merging point. min For example, for the merging point sm , the constraints are:

[0050]

[0051] In the case of no merging point, for example at an intersection, provision can be made that no minimum distance is set for vehicles in different lanes.

[0052] The vehicle moves on a trajectory corresponding to the state of the vehicle. The state of the vehicle is defined, for example, by its position, its velocity or its acceleration.

[0053] In this example, it is provided that the trajectories of the respective vehicles are determined such that a cost function defined for each vehicle is minimized, a physical limit defined for each vehicle or a plurality of physical limits defined for each vehicle are observed and the vehicles do not collide. In this context, for example, two vehicles do not collide if their respective positions are not identical at the same point in time.

[0054] In this example, a maneuvering signal for the respective vehicle is determined for each vehicle. In this example, the maneuvering signal is an acceleration.

[0055] In this example, a model predictive control is performed for determining the control signal.

[0056] For example, at time t k Measurements are recorded in the vehicle. These can be measurement signals of the host vehicle or measurement signals or predicted trajectories received from other vehicles.

[0057] Based on these measurements, a dynamic optimization problem is solved for each vehicle, where the state [x k , x (k+1) , x (k+2) , ...], the position of the vehicle in this example, is optimized over the prediction horizon N.

[0058] Also used for vehicle control signals [u k ,u (k+1) ,u (k+2) , ...] In this example, the setpoint acceleration is optimized in order to achieve a trajectory of states. Based on this is a dynamic model of the vehicle, which describes the relationship between actuation signals, such as acceleration, and states, such as the position of the vehicle.

[0059] It can be provided that restrictions are also taken into account within the scope of the optimization problem. Thus, the acceleration or braking capabilities of the vehicle are limited, which can be explicitly taken into account in the optimization. In merging processes, collision-freeness is important, which is taken into account, for example, by state restrictions in the cost function. In the optimization, it is thus ensured that the trajectory of the position of the vehicle on which the optimization is performed does not conflict with the received trajectory of the position of the adjacent vehicle.

[0060] In practice, uncertainties arise due to model uncertainties or disturbances, so that the actual trajectory of the vehicle's state does not necessarily coincide with the predicted trajectory of the state from the optimization problem.

[0061] In this example, only the first actuation signal u from the trajectory determined at time t is taken into account. k Applied to vehicles. At time t k+1 , record new measurements or send optimized trajectories between vehicles and solve the optimization problem again with the current information and transform the first signal u of the new trajectory of the control signal k Applicable to vehicles.

[0062] The cost function is defined, for example, by the following function, wherein one of the mentioned objectives or a plurality of the mentioned objectives is defined by q(x,u) and Q(x,u):

[0063]

[0064] The difference equations of the vehicle dynamics model are used

[0065] x(l+1|k)=Ax(l|k)+Bu(l|k)

[0066] and the measurement of the state, especially as a vector

[0067] x(k|k)=x meas (k)

[0068] In the case of one or more of the restrictions on use

[0069] u(l|h)∈U

[0070] x(l|h)∈X

[0071] and, for example, when using terminal set constraints (Terminal Set Bedingung) for model prediction regulation

[0072] x(k+N|k)∈Z f (k)

[0073] In the case of , the exemplary optimization problem is to minimize the cost function:

[0074]

[0075] In this example, the optimization problem is solved in the respective autonomous vehicle in time step k. For a state x(l|k) or an input u(l|k), k specifies the time step at which the optimization problem is solved and l specifies the time step within the forecast horizon N over which the optimization is performed.

[0076] The difference equation is a model in which the dynamic behavior of the vehicle is modeled, for example, using a double integrator, i.e. using

[0077]

[0078] Where s represents the position of the vehicle, v represents the velocity, and a represents the acceleration.

[0079] By using the sampling time t s and the state vector x = [sv] T And the definition of input u=a is discretized to determine the difference equation.

[0080] The measurements x(k|k) of the state represent constraints by means of which the dynamics model is reinitialized in time step k with the current measurements in order to then solve the optimization problem.

[0081] The limits u(l|k)∈U represent constraints, by which the braking and / or acceleration capabilities of the vehicle are taken into account. The limits x(l|k)∈X represent constraints, which on the one hand include speed limits. On the other hand, position limits can be taken into account here, where the limits of the position depend, for example, on the trajectories of other vehicles involved in the merging process.

[0082] In this example, by optimizing in time step k, for each vehicle i, the corresponding optimized trajectory of other vehicles j is used The other vehicle has already calculated the trajectory in time step k or k-1. In this example, when the vehicle merges earlier than vehicle i, the optimized trajectory of the vehicle has already been calculated in time step k. Or when the vehicle merges later than vehicle i, the optimized trajectory of the vehicle has been calculated in time step k-1 The optimized trajectory of the vehicle is used

[0083] It can be provided that a guarantee is given for collision-freeness. For this purpose, it is provided that the optimization problem can be solved in each time step k and all constraints describing collision-freeness are thereby satisfied. In order to ensure solvability in each time step k, for example, a scheme of recursive solvability (recursive feasibility) is applied. In this example, recursive solvability is guaranteed in a demonstrable manner by introducing additional final state constraints for the final state within the prediction range. The final state constraint is a condition on the nominal state at the end of the trajectory of states. The final state constraint is a condition on the allowed set of states at the end of the trajectory of states.

[0084] exist Figure 3 A flow chart with steps of a method for moving a mobile object at a traffic node is shown in FIG. The method is described by taking a first mobile object and a second mobile object as an example.

[0085] This method is applied, for example, to a first moving object and a second moving object moving in the same direction on tracks adjacent to each other.

[0086] The method is applied, for example, to a first mobile object and a second mobile object moving toward a traffic node on a lane intersecting a traffic node.

[0087] The method is applied accordingly for a plurality of moving objects.

[0088] The method uses a model that is constructed to predict a trajectory [x(k+1), ..., x(k+N)] of a state x of a first mobile object based on a trajectory [u(k), u(k+N-1)] of a control signal for moving the first mobile object and based on a state x(k) of the first mobile object at a time point k. For example, the model includes a differential equation. For example, the control signal is an acceleration of the first mobile object.

[0089] In step 302, the state x(k) of the first mobile object is determined. The state x(k) relates to a time point k. In this example, the state includes a position s(k) and a velocity v(k) of the first mobile object. The state x(k) may also include an acceleration of the first mobile object.

[0090] In step 304, the state x*(k) of the second mobile object is received until the trajectory [x*(k), x*(k+1), ..., x*(k+N)] of the prediction range N. In this example, the state x*(k) of the second mobile object includes the position and speed of the second mobile object. It can be provided that the state x*(k) includes the acceleration of the second mobile object.

[0091] Optionally, provision can be made to receive a disturbance variable for the state of the second mobile object.

[0092] In this example, provision is made to define a predefined sequence in which the first mobile object and the second mobile object are to move through the traffic node.

[0093] In step 306 , a restriction on the state of the first moving object is determined according to the trajectory of the state of the second moving object.

[0094] Optionally, provision can be made for the limitation of the state of the first mobile object to be determined as a function of the disturbance variable for the state of the second mobile object.

[0095] It may be provided that limitations are predefined for the actuation signal for the first mobile object, for example, that a limitation of the acceleration is predefined.

[0096] It may be provided that a desired state at the end of the trajectory of the state of the first mobile object or a permissible set of states at the end of the trajectory of the state of the first mobile object is predefined. For example, a final set condition is predefined.

[0097] In step 308, the model is used to determine the trajectory [x(k+1), . . . , x(k+N)] of the state of the first moving object starting from the state x(k) of the first moving object.

[0098] Within the constraints for the state of the first mobile object, the trajectory of the control signal [u(k), u(k+1), . . . , u(k+N-1)] and the trajectory of the state of the first mobile object [x(k+1), . . . , x(k+N)] are determined according to the cost function.

[0099] In this example, a trajectory of states of the first mobile object is determined, said trajectory observing this sequence.

[0100] It may be provided that the actuation signal is determined within limits for the actuation signal.

[0101] It may be provided that a trajectory of the state of the first mobile object is determined, in which case the state at the end of the optimization range is in the permissible set.

[0102] It may be provided that the model includes disturbance variables for the state of the first mobile object.

[0103] Optionally, provision may be made to determine a disturbance variable for the state of the first mobile object.

[0104] In this example, the optimization problem is solved in a manner provided with restrictions corresponding thereto.

[0105] In step 310, the trajectory [x(k+1), ..., x(k+N)] of the state of the first mobile object is sent to the second mobile object.

[0106] Optionally, provision can be made for a disturbance variable to be transmitted for the state of the first mobile object.

[0107] In step 312, the first mobile object is moved using a manipulation signal from the trajectory of manipulation signals [u(k), u(k+1), . . . , u(k+N-1)].

[0108] In this example, the first mobile object is moved using a first manipulation signal u(k) from the trajectory of manipulation signals [u(k), u(k+1), ..., u(k+N-1)].

[0109] Step 302 is then executed.

[0110] For a plurality of vehicles, the trajectory of the state of the first mobile object is determined from the trajectories of the states of the other mobile objects. In this example, mobile objects located in a predefined area around a traffic node are considered.

[0111] The method is applied, for example, to a first moving object and a second moving object that move toward a traffic node in the same direction on one lane when another moving object does not exist between the first moving object and the second moving object.

[0112] This approach is based on one or more of the following assumptions:

[0113] The participating vehicles are configured to move autonomously, the participating vehicles are equipped with vehicle-to-vehicle (V2V) communication which also allows time synchronization of the vehicles,

[0114] Assuming error-free communication,

[0115] The participating vehicles know which other vehicles are adjacent vehicles related to the merging process and then exchange optimized trajectories with the adjacent vehicles via V2V communication, presupposing that the vehicles pass through the traffic node 200, i.e., the merging point S m The order of lane merging does not change with time.

[0116] For example, this sequence is ensured by a FIFO (First In First Out) method with respect to the vehicles entering the area at the entrance 218 .

[0117] Consider a merging process with M vehicles. The first vehicle in the merging sequence has an index of i=0, and the last vehicle has an index of i=M-1. In the example of five vehicles, the first vehicle 202 in the merging sequence has an index of i=0, the second vehicle 204 has an index of i=1, the third vehicle 206 has an index of i=2, the fourth vehicle 208 has an index of i=3, and the fifth vehicle 210 has an index of i=4. More or less than five vehicles may be considered.

[0118] exist Figure 4 2 shows the flow of the merging process according to the first embodiment, in which uncertainty is not taken into account. In the merging process according to the first embodiment, no cooperative behavior between vehicles is assumed.

[0119] The steps of the merging process according to the first embodiment are described below. In this example, the method for moving a mobile object runs on the respective participating vehicles and provides a safety guarantee for collision-freeness.

[0120] Figure 4 The steps of the merging process according to the first specific embodiment shown in are respectively executed in the participating vehicles at a time k during the merging process, ie while the respective vehicle is between the entry 218 and the exit 220 of the zone.

[0121] Step 402:

[0122] In step 402, the state x of the corresponding vehicle is detected in the participating vehicles at the time point k. i (k) In this example, measurement data are detected, which are characteristic of a state, such as an acceleration, a speed or a position of the respective vehicle.

[0123] Step 404:

[0124] In step 404 , a variable i used to identify a participating vehicle is initialized to zero.

[0125] Step 406:

[0126] In step 406, for each participating vehicle i, the final set Z assigned to the vehicle is updated f,j .

[0127] Provide a final set Z for each participating vehicle f，i In this example, the final set Z for participating vehicles is determined by means of conventional optimization f，i . For example, the final set of participating vehicles is calculated offline.

[0128] In this example, an error coordinate system is used for each participating vehicle i. In the error coordinate system of the respective vehicle i, the vehicle i is arranged according to a predetermined order relative to the vehicle traveling directly in front of the vehicle i, and M subsystems E are defined. i , the subsystem describes the error dynamics with the aid of error states.

[0129] The error state z0 for the first vehicle i=0 is

[0130]

[0131] where v refrepresents the reference speed, s ref (k) = kt s v ref represents the reference position of the first vehicle at time k, v0(k) represents the speed of the first vehicle at time k, and t s represents the sampling time. In this example, the first vehicle has no vehicle ahead. In this example, it is assumed that the first vehicle is adjusted to a predetermined reference speed v ref , and the error state z0(k) of the first vehicle is calculated according to its initial position s0(k) and its initial velocity v0(k) and the time point t s to be sure.

[0132] Determine the error state z for other participating vehicles i :

[0133]

[0134] Error state z i Contains the distance error to the vehicle i-1 that is directly ahead according to the sequence and the speed relative to the reference speed v ref The speed error v i (k)-v ref In this example, a common reference speed is used for the participating vehicles so that the participating vehicles are synchronized in terms of speed. In this example, a reference distance d is defined ref =d min +t d v ref , where t d is a pre-given time interval. Based on these error state definitions, the final set Z is calculated f，i , the final set is defined as an ellipsoid around the origin of the error coordinate system. Possibilities for this are described in Christian Conte et al., “Distributed synthesis and stability of cooperative distributed model predictive control for linear systems”. In: Automatica 69 (2016).

[0135] Step 408:

[0136] In step 408 , an optimization problem is solved for each participating vehicle i.

[0137] The solution to the optimization problem includes, for each participating vehicle, the optimized trajectory x up to the prediction horizon N * i The corresponding final state at the end of the prediction horizon N is located in the final set Zf，i middle.

[0138] In this example, the following cost function is defined for the optimization problem:

[0139]

[0140] In the cost function, the deviation of the vehicle speed from the reference speed (v i (l|k)-v ref ) 2 and the setting variable cost u for the control signal of the corresponding vehicle i (l|k ) 2 is penalized as a constraint. It can be specified that the two terms are weighted by a weighting factor q or r.

[0141] In this example, the cost function includes the slip variable ρ n，i and ρ o，i , the slip variable enables the relaxation of the constraints. The slip variable is optional.

[0142] An exemplary optimization problem for vehicle i is:

[0143]

[0144] To solve the optimization problem, information about the other participating vehicles is used, which is transmitted via communication. This information is the optimized trajectory of the other participating vehicles. The trajectory is optimized in the other participating vehicles.

[0145] In this example, it is specified that the cost function is minimized under constraints. In this example, the constraints include having an initial state x from the measurement i The time-discrete system dynamics of the vehicle (k|k):

[0146] x i (l+1|k)=A i x i (l|k)+B i u i (l|k), l=k, ..., k+N-1

[0147] In this example, the constraints include the vehicle's acceleration limit u min ≤u i (l|k)≤u max , l = k, ..., k + N-1 and the speed limit v min ≤v i (l|k)≤v max , l=k,...,k+N.

[0148] The constraints include the distance n from the preceding vehicle in the same lane. i - Distance conditions

[0149]

[0150] and

[0151] ρ n，i (l)≤t d v i (l|k), l=k, ..., k+N.

[0152] These two conditions ensure that the minimum distance d min .

[0153] If the slip variable ρ n，i becomes very small, then in addition to the minimum distance, the speed-dependent distance t d v i (l|k).

[0154] This follows from the solution of the optimization problem, since ρ n，i is taken into account in the cost function and is therefore dependent on the weight p of the slip variable.

[0155] The constraints include the distance n from the following vehicle in the same lane i + Distance Condition

[0156]

[0157] The constraints include the distance from the preceding vehicle traveling in a lane parallel to the lane of the vehicle. i - Distance conditions

[0158]

[0159] ρ o,l (l)≤t d v i (l|k), l=k, ..., k+N

[0160]

[0161] Here, with the same motivation, we introduce the final variable ρ o，i , the final variable is also considered in the cost function. These distance conditions are only from the merging point S m The last of these distance conditions is introduced to ensure recursive solvability.

[0162] The constraints include the distance from the vehicle behind in a lane parallel to the lane of the vehicle. i+ Distance Condition

[0163]

[0164] Constraints include conditions on the final position and final velocity of the vehicle.

[0165]

[0166] i=1,...,M−1

[0167]

[0168] And the conditions for ensuring recursive solvability:

[0169]

[0170] Step 410:

[0171] In step 410, the optimized trajectories x are exchanged between the participating vehicles. * Ni .

[0172] Step 412:

[0173] In step 412, it is checked whether steps 408 and 410 have been performed for the M participating vehicles. For example, it is checked whether the variable is i=M-1.

[0174] If steps 408 and 410 have been performed for M participating vehicles, then step 414 is performed. Otherwise, the step is performed for the next vehicle in sequence. For example, the variable i is assigned the value i+1.

[0175] In step 414 , the participating vehicles are steered using the steer signal.

[0176] In this example, a first value of the optimized trajectory of the control signal u(k), in this example the acceleration of the vehicle, is applied for each vehicle. This means that the required acceleration is set in each vehicle involved in the merging process by a subordinate acceleration controller.

[0177] exist Figure 5 FIG. 4 shows the flow of the merging process according to the second embodiment, in which uncertainty is taken into account.

[0178] To take uncertainty into account, the model with optimal deterministic input signal is extended with disturbance parameter input, i.e.

[0179] x i (k+1)=A i x i (k)+B i u i (k)+Di w i (k)

[0180] Interference parameter w i is usually unknown, and therefore not measurable, but it is assumed that the limits of the interference are known, for example in the form:

[0181]

[0182] Due to this disturbance, errors occur in the prediction of the trajectory. This may cause the true trajectory s i (l) Deviation from the trajectory predicted in the optimization problem Therefore, we now consider the possible set of disturbance parameters W in the trajectory prediction i , which results in a set of possible trajectories.

[0183] Provision may be made to model the uncertainty of the communication in the presence of interference.

[0184] In this example, according to the second implementation of interference uncertainty, in step 410, instead of communicating trajectories between vehicles, contracts are communicated. The contract describes a set of possible trajectories based on optimal deterministic input signals.

[0185] In step 408, the optimization is determined using a model in which the optimal deterministic input signal is extended by the disturbance variable input. In this example, the same cost function is used as in the first specific embodiment.

[0186] In this example, instead of the difference equation, the following regulation law is used:

[0187] x l (l+1|k)=(A i +B i K w，i )x i (l|k)+B i c i (l|k), l=k, ..., k+N-1

[0188] u i (l|k)=c i (l|k)+K w，l x i (l|k), l=k...,k+N-1

[0189] x i (k|k)=x i (k)

[0190]

[0191] zi (k+N|k)∈Z f，i (k)

[0192] z i+1 (k+N|k)∈Z f，i+1 (k)

[0193] z i+2 (k+N|k)∈Z f，i+2 (k)

[0194] This means that the controller is supplemented with a state controller in order to compensate for disturbances. The final set of optimization problems for the model predictive control is determined, for example, taking into account the uncertainties. The constraints for the control signal u and the state x are also considered in the contract. An example of this is described in Christian Conte et al. "Robust distributed model predictive control of linear systems". In: 2013 European Control Conference (ECC). The other conditions are constraints on the final set.

[0195] In order to take uncertainty into account in the optimization, contracts are exchanged between vehicles. These contracts contain, for example, a set of predicted possible trajectories of the vehicle. In this example, the vehicle takes into account the worst-case trajectories of neighboring vehicles in its own optimization from the obtained contracts.

[0196] The remaining steps of the merging process according to the second specific embodiment are described as for the merging process according to the first specific embodiment, wherein in this example, the state x of the respective vehicle is determined. i A step 406 for updating the final set is performed between the step 402 of (k) and the step 404 for initializing the variable i for identifying the participating vehicles to zero.

[0197] In the case of the merging process according to both embodiments, a cooperative behavior of the vehicles with one another can optionally be sought. For example, this can allow one of the participating vehicles to accelerate for the vehicle merging behind it, so that the traffic flow is improved overall.

[0198] For example, cost functions and constraints are adapted in the optimization problem for the cooperation of vehicles. For this purpose, a slip variable is introduced for each vehicle traveling behind a vehicle. and

[0199] In the optimization problem mentioned above, the condition

[0200]

[0201] Replacement conditions

[0202]

[0203] In the optimization problem mentioned above, the condition

[0204]

[0205] Conditioned

[0206]

[0207] replace.

[0208] The cost function is extended with the slip variable:

[0209]

[0210] Then the weight w can be i 、w n and w o To influence the collaborative behavior, which can be achieved through w n = 0 and w o = 0 to stop the cooperation for the two following vehicles. As explained here for the nominal case, the cooperative behavior can also be integrated into the robust case.

[0211] The distributed cooperative control of the objects moved at the nodes is formulated in this example as model predictive control, in which the optimization problem is repeatedly solved at runtime taking into account the system dynamics of the vehicles. Collision-freeness is formulated and guaranteed in the form of constraints, such as safety distances between vehicles.

[0212] The participating vehicles are equipped with V2X technology, for example, for communication. Communication takes place either directly or via a cloud / edge connection. It is possible that a non-communicating vehicle is considered a participating vehicle by having its trajectory detected by another participating vehicle. However, the non-communicating vehicle is not controlled using control signals.

[0213] In general, the following advantages are obtained:

[0214] Model predictive control is an advantageous implementation in order to react to behaviors that were not anticipated in advance. Collision avoidance can be achieved directly via the limits and can be ensured at every point in time.

[0215] Explicit consideration of disturbances enables extended formal guarantees to be given to the nominal model.

[0216] The solvability of the optimization problem as a function of the disturbance can also be used as a requirement or contract for the subordinate control functions. In order to avoid collisions, under the described assumptions, there are formal guarantees for compliance with the constraints.

[0217] The cost function can take into account very different criteria, such as throughput, energy efficiency, comfort.

[0218] By exchanging the entire trajectory between participating vehicles, the method is robust against typical communication problems such as packet failures and delays. For example, in doubtful cases, the last transmitted trajectory of one or more other participants is used until a new trajectory is transmitted.

[0219] Taking uncertainties and disturbances into account in the design of the control function enables the use of the function within the scope of a safe concept. The sum of all uncertainties that can be tolerated by the function can be represented by the disturbance influence. Depending on the disturbance actually present, the control function can be easily adapted (more conservative / higher performance).

[0220] The optional consideration of other participants by exchanging certain additional variables makes it possible to integrate cooperative behavior into the functionality without losing formal guarantees, such as collision-freeness.

[0221] Since the optimization problem for model predictive control is formulated in such a way that a safety guarantee can be given, the safety guarantee applies. The constraints in the optimization problem are optionally predefined in such a way that collision-freeness can be guaranteed even during merging processes, for example, while taking into account uncertainties. In addition, the recursive solvability of the optimization problem for model predictive control is ensured, i.e. if the optimization problem is solvable in one step, it is also solvable in all other steps. This is particularly important when taking into account constraints that are important for collision-freeness.

Claims

1. A computer-implemented method for moving a first mobile object (202) at a traffic node (200), characterized in that Determining (302) a state of a first mobile object (202), the state comprising a position, a velocity or an acceleration of the first mobile object, wherein a trajectory of a state of a second mobile object (204) at the traffic node (200) is received (304), the state comprising a position, a velocity or an acceleration of the second mobile object (204), wherein constraints on the state of the first mobile object (202) are determined (306) based on the trajectory of the state of the second mobile object (204), wherein the trajectory of the state of the first mobile object (202) is determined (308) based on the state of the first mobile object (202) using a model, the model being configured to determine the trajectory of the state of the first mobile object (202) based on a constraint for moving the first mobile object (202) A trajectory of a control signal of a first mobile object (202) and a trajectory of a state of the first mobile object (202) predicted based on the state of the first mobile object, wherein the trajectory of the control signal and the trajectory of the state of the first mobile object (202) are determined (308) within constraints for the state of the first mobile object (202) based on a cost function, wherein the cost function defines a target for moving the first mobile object (202) using the trajectory of the control signal on the trajectory of the state of the first mobile object (202), and wherein the first mobile object (202) is moved (312) at the traffic node using a control signal from the trajectory of the control signal.

2. The method according to claim 1, characterized in that Limitations for an actuation signal for the first mobile object (202) are predefined (306), wherein the actuation signal (202) is determined (308) within the limitations for the actuation signal (202).

3. The method according to any one of the preceding claims, characterized in that A rated state at the end of a trajectory of the state of the first mobile object is pre-given (306), wherein a trajectory of the state of the first mobile object (202) is determined (308), the trajectory including the rated state at the end of the trajectory, or an allowed set of states at the end of the trajectory of the state of the first mobile object is pre-given (306), wherein a trajectory of the state of the first mobile object (202) is determined, in the case of which the state at the end of the optimization range is within the allowed set.

4. The method according to any one of the preceding claims, characterized in that A disturbance variable for the state of the second mobile object (204) is received (304), wherein a limitation for the state of the first mobile object (202) is determined (306) based on the disturbance variable for the state of the second mobile object (204).

5. The method according to any one of the preceding claims, characterized in that The model includes (308) disturbance variables for the state of the first mobile object.

6. The method according to any one of the preceding claims, characterized in that A sequence is predefined (304) in which the first mobile object (202) and the second mobile object (204) are to move toward the traffic node (200), wherein a trajectory of the state of the first mobile object (202) is determined (308), the trajectory adhering to the sequence.

7. The method according to any one of the preceding claims, characterized in that A trajectory of the state of the first mobile object (202) is determined (308) based on (304) the trajectory of the state of mobile objects (202, ..., 210), which are located in a predefined area around the traffic node (204).

8. The method according to any one of the preceding claims, characterized in that When there is no other moving object (206, ..., 210) between the first moving object (202) and the second moving object (204), the first moving object (202) and the second moving object (204) move toward a traffic node in the same direction on one lane (304).

9. The method according to any one of the preceding claims, characterized in that The first moving object (202) and the second moving object (204) move (304) toward a traffic node (200) in the same direction on adjacent lanes (212, 214) or on lanes intersecting at a traffic node (200).

10. A device (102) for moving a first mobile object (100) at a traffic node (200), characterized in that: The device (102) is designed to carry out the method according to any of the preceding claims.

11. A mobile object (100; 202, ..., 210), characterized in that: The mobile object (100; 202, ..., 210) comprises a device (102) according to claim 10.

12. A computer program, characterized in that The computer program includes instructions executable by a computer, and when the instructions are executed by the computer, the method according to any one of claims 1 to 9 is executed.

Images