A real-time self-adaptive merging control method for autonomous vehicles

Abstract

The present application relates to a kind of automatic driving vehicle real-time self-adapting control method of merging, belong to vehicle trajectory optimization method technical field;The method includes: according to vehicle driving data, the future vehicle trajectory is predicted;The target merging gap of the minimum entrance ramp vehicle that interference to main road traffic is obtained;Design the entrance ramp vehicle non-stop merging control scheme considering the minimum fuel consumption;Using model predictive control, the future time step trajectory of entrance ramp vehicle is considered, and the control rate of entrance ramp vehicle is solved online.The present application predicts the motion state of future main road vehicle, and real-time self-adaptively selects target entry gap for entrance ramp vehicle, and completes the safe merging of entrance ramp vehicle under the condition of guaranteeing not to affect main road traffic.Meanwhile, the present application considers the motion state of entrance ramp vehicle, and smoothes the merging trajectory of entrance ramp, thereby reducing its fuel consumption.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of vehicle trajectory optimization method, and particularly relates to a real-time adaptive merging control method for an automatic driving vehicle. BACKGROUND

[0002] The merging area of the on-ramp of the urban expressway is an important conflict point and bottleneck area of urban road traffic, and the merging of the on-ramp vehicle into the main road is one of the most challenging driving scenarios, and the on-ramp vehicle must slow down or even stop at the end of the ramp to wait for a safe merging gap on the main road. Frequent deceleration and stopping will increase the fuel consumption and travel time of the on-ramp vehicle and reduce the traffic efficiency. In order to realize smooth and efficient merging, automatic driving car technology has been widely applied to many merging strategies. Most of the existing merging strategies control the deceleration of the main line vehicle to create a merging gap for the on-ramp vehicle. Although these strategies ensure the safety and efficiency of the merging of the on-ramp vehicle, the deceleration of the main line vehicle may cause the formation of serious traffic waves, which may even extend to the upstream of the main line. In addition, in the merging area, the main line vehicle must slow down to adapt to the speed difference with the ramp vehicle, which may affect the efficiency of the main line traffic. Therefore, how to control the on-ramp vehicle to realize smooth and efficient merging with the minimum disturbance to the main line traffic is a challenge. SUMMARY

[0003] The purpose of the present application is to design a real-time adaptive merging control method for an automatic driving vehicle, which realizes the smooth and efficient merging of the on-ramp vehicle without changing the motion of the main line vehicle. In which, the motion state (i.e. acceleration and deceleration) of the future main line vehicle is predicted to select the target merging gap for the on-ramp vehicle, so as to realize the merging of the on-ramp vehicle without changing the motion state of the main line vehicle, which is beneficial to minimize the disturbance to the main line traffic. At the same time, the on-ramp vehicle is controlled to adjust its speed in advance, and the merging is completed within the target merging gap, which reduces the number of frequent deceleration and stopping of the on-ramp vehicle and improves the merging efficiency.

[0004] To achieve the above-mentioned content, the technical scheme of the present application is as follows:

[0005] S1: predicting the future vehicle trajectory according to the vehicle driving data;

[0006] S2: obtaining the target merging gap of the on-ramp vehicle with the minimum disturbance to the main line traffic;

[0007] S3: designing a non-stopping merging control scheme for the ramp vehicle considering the lowest fuel consumption;

[0008] S4: using model predictive control to consider the trajectory data of the ramp vehicle within the future time step to solve the on-line control rate of the ramp vehicle.

[0009] The specific step S1 of predicting the future vehicle trajectory according to the vehicle driving data is:

[0010] S101: Collect the vehicle historical acceleration data on the trunk road, arrange the historical data, and construct an acceleration data structure based on the original time sequence;

[0011] S102: Convert the original time sequence data into a first-order cumulative generation operation sequence data according to a single variable first-order gray model.

[0012] S103: Construct a black system to obtain a first-order differential equation of the gray model.

[0013] S104: Construct a gray system, calculate its parameters by using the least square method, and obtain the acceleration prediction value corresponding to different time stamps through inverse cumulative generation operation.

[0014] The specific step S2 of taking the target merging gap of the on-ramp vehicle with the minimum traffic interference to the trunk road is:

[0015] S201: Collect the current position, speed, and acceleration information of the trunk road vehicle and the information within the future time step, and construct a gap selection set;

[0016] S202: Construct a dynamic model of the on-ramp vehicle by using a dynamic model, and determine different gap types, i.e., effective gap, invalid gap, and ineffective gap, according to the current position, speed, and acceleration information of the on-ramp vehicle.

[0017] S203: Select the effective gap set, and select the target merging gap with the shortest time and the minimum interference to the trunk road in the merging process without stopping or decelerating the trunk road vehicle to avoid the on-ramp vehicle.

[0018] The specific step S3 of designing the on-ramp vehicle non-stop merging control scheme considering the lowest fuel consumption is:

[0019] S301: Construct the on-ramp vehicle merging target in each time step and its future p f time steps, i.e., the on-ramp vehicle safely and smoothly tracks the target merging gap and merges into the trunk road without sharp acceleration and deceleration; and the fuel consumption of the on-ramp vehicle is the lowest.

[0020] S302: Set the target function as:

[0021]

[0022] Specifically, the first objective function represents the merging gap of the entrance ramp vehicle tracking target, the second objective function represents that the entrance ramp vehicle keeps a uniform speed state to the maximum extent and does not have a large acceleration and deceleration behavior, and the third objective function represents a minimum fuel consumption model, and the entrance ramp is controlled to complete the merging with the minimum fuel consumption. -δp is the weighted different time step running cost. Since the uncertainty of the running cost increases over time, this item provides a higher weight for the short-term future running cost than the long-term future, and Γ1, Γ2 and Γ3 are weight matrices.

[0023] S303: In order to realize the safe and smooth merging of the entrance ramp vehicle into the main road, position, speed, acceleration and acceleration change constraint conditions are set.

[0024] The step S4: the specific steps for solving the ramp vehicle control rate online by using the model predictive control to consider the trajectory data of the ramp vehicle in the future time step are as follows:

[0025] S401: The vehicle state equation is discretized by using the forward Euler method:

[0026] S402: The state space equation is constructed:

[0027] S403: The merging problem of the entrance ramp vehicle is reconstructed, and a quadratic term of a relaxation variable is added:

[0028] Compared with the prior art, the technical scheme of the present application has the advantages that:

[0029] The present application can realize the merging of the entrance ramp vehicle without changing the influence on the main road traffic by predicting the motion state of the main road vehicle to select the target merging gap for the entrance ramp vehicle, which helps to minimize the interference on the main road traffic; the entrance ramp vehicle adjusts the speed in advance and merges into the main line in the target merging gap, which avoids frequent deceleration and parking, greatly smooths the merging trajectory of the entrance ramp vehicle, reduces fuel consumption, and improves passenger comfort. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 The present application is a method flowchart. DETAILED DESCRIPTION

[0031] In order to further illustrate the above purposes, features and advantages of the present application, the present application will be described in detail below with reference to the accompanying drawings.

[0032] As Figure 1 shown, the implementation process of the present application includes the following steps:

[0033] S1: predicting the future vehicle trajectory according to the vehicle driving data;

[0034] Collect the vehicle historical acceleration data on the main road, organize the historical data, and construct the acceleration data structure based on the original time series, that is According to the univariate first-order gray model, the original time series data is converted into the first-order cumulative generation operation sequence data, that is

[0035]

[0036] Construct a black system Obtain the first-order differential equation of the gray model, Construct a gray system, and calculate its parameters by the least square method to make When And Apply the least square method Get the coefficient Substitute to get Get the acceleration prediction value corresponding to different timestamps by inverse cumulative generation operation, that is

[0037] S2: Take the target merging gap of the entrance ramp vehicle with the minimum traffic interference on the main road;

[0038] The minimum length of the gap that allows the entrance ramp vehicle to merge into the main road is:

[0039]

[0040] In the above minimum length calculation formula, the reaction time and braking coordination time of the driver and the maximum braking speed of the vehicle are considered.

[0041] For the entrance ramp vehicle, due to the limitation of maximum speed and acceleration, it is necessary to estimate whether the entrance ramp vehicle can reach the target merging node before entering the merging area. According to the vehicle dynamics principle, the boundary condition expression is:

[0042]

[0043] In the above formula, L is the length of the control area. L s (k f ) represents the safety distance between the entrance ramp vehicle and the main line vehicle, that is:

[0044]

[0045] Where represents the maximum speed operation.

[0046] According to the above formula, three types of gaps are distinguished:(1) Effective gap means that the gap length is greater than or equal to the minimum length, that is The on-ramp vehicle can reach the target merging node before entering the merging area. (2) Invalid gap means that the gap length meets But the on-ramp vehicle cannot reach the target merging node before entering the merging area even if it drives at the maximum speed. (3) Invalid gap means that the gap length is too small to safely merge.

[0047] The valid merging gap set for the on-ramp vehicle is:

[0048]

[0049] To minimize the travel time of the on-ramp vehicle, the nearest valid gap to the merging area is set as the target merging gap.

[0050] S3: Design a ramp vehicle non-stop merging control scheme that considers the lowest fuel consumption;

[0051] Construct the on-ramp vehicle merging target at each time step and its future p f time steps, i.e. the on-ramp vehicle safely and smoothly tracks the target merging gap and merges into the main road without sharp acceleration and deceleration. The on-ramp vehicle has the minimum fuel consumption.

[0052] Set the objective function as:

[0053]

[0054] Specifically, the first objective function indicates that the on-ramp vehicle tracks the target merging gap, the second objective function indicates that the on-ramp vehicle maximizes the uniform speed state and does not have excessive acceleration and deceleration behavior, and the third objective function indicates the minimum fuel consumption model to control the on-ramp to complete the merging with the lowest fuel consumption. Wherein, e -δp is the weighted running cost of different time steps. Since the uncertainty of the running cost increases over time, this term provides a higher weight for the short-term future running cost than the long-term future, and Γ1, Γ2 and Γ3 are weight matrices;

[0055] To achieve safe and smooth merging of the on-ramp vehicle into the main road, the constraint condition is set as follows:

[0056]

[0057] The above constraint represents the safety constraint between two adjacent on-ramp vehicles. The distance between the vehicles cannot be less than the safety distance, which is limited by the speed of the two on-ramp vehicles.

[0058]

[0059] The above constraint means that the acceleration variation value of the ramp-in vehicle needs to be constrained in order to ensure the comfort of passengers.

[0060]

[0061]

[0062]

[0063] The above constraint means that the ramp-in vehicle cannot violate its own limits, i.e., the maximum / minimum speed, the maximum / minimum acceleration, and the maximum / minimum control input, during the merging process. Among them, the minimum speed of the ramp-in vehicle is a function of its position, which ensures that the ramp-in vehicle can stop smoothly and safely at the end of the ramp when it does not find an effective gap during the merging process.

[0064]

[0065] S4: considering the trajectory data of the ramp-in vehicle in the future time steps using model predictive control, the control rate of the ramp-in vehicle is solved online;

[0066] The vehicle state equation is discretized using the forward Euler method:

[0067]

[0068]

[0069]

[0070]

[0071] The above x j 、 a j and state respectively represent the position, speed, acceleration, and acceleration variation of the vehicle;

[0072] The state space equation is constructed as follows:

[0073] I j (k+1)=A j I j (k)+B j u j (k)

[0074] Y j (k+1)=C j I j (k+1)

[0075] The above and And

[0076] Reconstruction of the entrance ramp vehicle confluence problem, add slack variable quadratic term:

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] The above are only preferred embodiments of the present application, the protection scope of the present application is not limited to the above-mentioned implementation cases, it is worth pointing out that several improvements and decorations without departing from the principles of the present application premise should be regarded as the protection scope of the present application.

Claims

1. A real-time self-adaptive merging control method for autonomous vehicles, characterized in that, The method comprises the following steps: S1: predicting a future vehicle trajectory according to vehicle driving data; S2: obtaining a target merging gap of an on-ramp vehicle with minimum interference to a main road traffic; S3: designing a non-stop merging control scheme for the on-ramp vehicle with minimum fuel consumption, and the specific steps are as follows: S301: build each time step and its future The entrance ramp vehicle converges into the target in each time step and its future, that is, the entrance ramp vehicle safely and smoothly tracks the target merging gap, converges into the main road without sharp acceleration and deceleration, and the fuel consumption of the entrance ramp vehicle is minimized. S302: setting a target function as: ； Specifically, the first objective function represents merging gap of the ramp vehicle tracking target, the second objective function represents that the ramp vehicle keeps a uniform speed state to the maximum extent and does not have a large acceleration and deceleration behavior, and the third objective function represents a minimum fuel consumption model, and the ramp is controlled to complete the merging with the minimum fuel consumption; wherein, is a weighted different time step running cost; since the uncertainty of the running cost increases over time, this item provides a higher weight for the running cost in the short term future than in the long term future, 、 and is a weight matrix; S303: in order to realize safe and smooth merging of the on-ramp vehicle into the main road, the constraint condition is set as follows: ； The above constraint represents the safety constraint between the adjacent two on-ramp vehicles, and the distance between the vehicles cannot be less than the safety distance, which is limited by the speed of the two on-ramp vehicles; ； The above constraint represents that the acceleration change value of the on-ramp vehicle needs to be constrained in order to ensure the comfort of passengers; ； ； ； The above constraint represents that the on-ramp vehicle cannot violate its own limits during the merging process, i.e., the maximum / minimum speed, the maximum / minimum acceleration, and the maximum / minimum control input; wherein the minimum speed of the on-ramp vehicle is a function of its position, which ensures that when the on-ramp vehicle does not find an effective gap during the merging process, the on-ramp vehicle can be stopped smoothly and safely at the end of the ramp; ； S4: using model predictive control to consider the future trajectory of the on-ramp vehicle within a time step, and performing online solving on the control rate of the on-ramp vehicle, and the specific steps are as follows: S401: discretizing the vehicle state equation by using the forward Euler method: ； ； ； ； The above , , and The states respectively represent the position, speed, acceleration, and acceleration change of the vehicle. S402: constructing a state space equation: ； ； The above and and ; S403: reconstructing the merging problem of the on-ramp vehicle and adding a quadratic term of the slack variable: ； ； ； ； ； 。 2. The real-time self-adaptive merge control method for an autonomous vehicle according to claim 1, wherein, The specific steps of predicting the future vehicle trajectory according to the vehicle driving data are as follows: S101: collecting historical acceleration data of the main road vehicle, organizing the historical data, and constructing an acceleration data structure based on the original time sequence; S102: converting the original time sequence data into a first-order cumulative generation operation sequence data according to a first-order gray model; S103: constructing a black system to obtain a first-order differential equation of the gray model; S104: constructing a gray system, calculating its parameters by using the least square method, and obtaining the acceleration prediction value corresponding to different time stamps through inverse cumulative generation operation.

3. The real-time self-adaptive merge control method for an autonomous vehicle according to claim 1, wherein, The specific steps of obtaining the target merging gap of the on-ramp vehicle with minimum interference to the main road traffic are as follows: S201: collecting the current position, speed, and acceleration information of the main road vehicle and the information within a future time step, and constructing a gap selection set; S202: constructing an on-ramp vehicle dynamic model by using a dynamics model, and judging different gap types, i.e., effective gap, invalid gap, and ineffective gap, according to the current position, speed, and acceleration information of the on-ramp vehicle; S203: selecting an effective gap set, in which the main road vehicle does not need to stop or decelerate to avoid the on-ramp vehicle during the merging process, and selecting a target merging gap with minimum interference to the main road and shortest time.

Images