Fusion predictive lane change cruise control method and device

Abstract

The application provides a fusion type predictive lane-changing cruise control method and device, and relates to the technical field of vehicles.The method comprises the following steps: planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information; determining a predicted state of an obstacle vehicle in a preset distance range according to a current state of the obstacle vehicle; constructing a discretized behavior state space of the self-driving vehicle in a road section according to the energy-saving vehicle speed sequence; determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; and obtaining a future optimal recommended state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so that the fusion type predictive lane-changing cruise control can be performed on the self-driving vehicle, and the energy-saving effect of the vehicle can be enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vehicle technology, in particular to a fusion type predictive lane-changing cruise control method and device. BACKGROUND

[0002] With the rapid development of China's economy, the volume of transportation is increasing year by year. In order to achieve the goal of low-carbon transportation, energy-saving driving technologies such as predictive cruise control are currently used to reduce the energy consumption of self-driving vehicles and achieve low-carbon transportation energy saving and emission reduction.

[0003] In a complex and variable traffic environment, the vehicle speed may change frequently due to the influence of the front obstacle vehicle in the current lane, increasing the energy consumption of the vehicle and reducing the driving efficiency of the vehicle. At present, the lane-changing decision and control technology mainly controls the vehicle to change lanes by calculating the acceleration benefit of the vehicle before and after changing lanes to reduce the negative impact of the front obstacle vehicle. This lane-changing decision and control technology mainly controls the vehicle to change lanes based on the dynamic information of the surrounding vehicles from the vehicle perspective, without considering the road slope characteristics and the state information of the surrounding vehicles in the super-visual range, resulting in a large gap between the actual vehicle state of the lane-changing and the ideal vehicle state under the current road characteristics, and thus leading to poor energy-saving effect of the vehicle. SUMMARY

[0004] In view of the problems in the prior art, the present application provides a fusion type predictive lane-changing cruise control method and device.

[0005] The present application provides a fusion type predictive lane-changing cruise control method, comprising:

[0006] Based on the static road information, an energy-saving vehicle speed sequence of the self-driving vehicle is planned;

[0007] The current state of the obstacle vehicle in the preset distance range is determined to determine the predicted state of the obstacle vehicle; and a discrete behavior state space of the self-driving vehicle in the road section is constructed according to the energy-saving vehicle speed sequence;

[0008] Based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle, a lane-changing state point of the self-driving vehicle is determined from the discrete behavior state space;

[0009] According to the energy-saving vehicle speed sequence and the lane-changing state point, a state sequence of the self-driving vehicle is obtained to perform fusion type predictive lane-changing cruise control on the self-driving vehicle.

[0010] According to the fusion type predictive lane-changing cruise control method provided by the present application, the lane-changing state point of the self-driving vehicle is determined from the discrete behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle, and specifically comprises:

[0011] inflating the energy-saving vehicle speed sequence to obtain a set of simulated vehicle speed space points centered on the energy-saving vehicle speed;

[0012] determining a lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the simulated vehicle speed space points and the predicted state of the obstacle vehicle.

[0013] According to the present application, a fusion type predictive lane-changing cruise control method is provided, which obtains a state sequence of the autonomous vehicle based on the energy-saving vehicle speed sequence and the lane-changing state point, specifically including:

[0014] determining a travel cost of each lane-changing state point and the corresponding simulated vehicle speed;

[0015] forward simulating a plurality of simulated state sequences by simulating different lane-changing state points and simulated vehicle speeds, and determining a travel cost of each simulated state sequence;

[0016] comparing the travel costs of the simulated state sequences to obtain a state sequence of the autonomous vehicle with the minimum cost; the state sequence of the autonomous vehicle includes a vehicle speed sequence and a behavior sequence.

[0017] According to the fusion type predictive lane-changing cruise control method provided by the present application, before determining a travel cost of each lane-changing state point and the corresponding simulated vehicle speed, the method further includes:

[0018] establishing a cost model based on the fuel consumption, travel efficiency of the autonomous vehicle at the lane-changing state point, and the speed deviation between the simulated vehicle speed and the energy-saving vehicle speed;

[0019] determining a travel cost of each lane-changing state point and the corresponding simulated vehicle speed, specifically including:

[0020] determining a travel cost of each lane-changing state point and the corresponding simulated vehicle speed based on the cost model.

[0021] According to the fusion type predictive lane-changing cruise control method provided by the present application, determining a lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the simulated vehicle speed and the predicted state of the obstacle vehicle, specifically including:

[0022] determining a behavior state point that meets a preset lane-changing safety standard as the lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the simulated vehicle speed and the predicted state of the obstacle vehicle.

[0023] According to the fusion type predictive lane-changing cruise control method provided by the present application, planning an energy-saving vehicle speed sequence of the autonomous vehicle based on static road information, specifically including:

[0024] Determine the planning distance domain of the energy-saving vehicle speed sequence of the self-driving vehicle, divide the planning distance domain into at least two stages based on road information, and plan the energy-saving vehicle speed sequence of the self-driving vehicle for each stage;

[0025] When the self-driving vehicle reaches the second stage, the step of planning the energy-saving vehicle speed sequence of the self-driving vehicle is repeated, and the energy-saving vehicle speed sequence of the self-driving vehicle is dynamically planned;

[0026] According to the lane, the energy-saving vehicle speed sequence corresponds to a discretized behavior state space, which specifically includes:

[0027] According to the lane and the discrete step number, the energy-saving vehicle speed sequence in the first stage of the distance domain corresponds to a discretized behavior state space.

[0028] According to the present application, a kind of fusion predictive lane-changing cruise control method is provided, the current state includes state quantity set and control quantity set, the current state of the obstacle vehicle in the preset distance range is determined to determine the predicted state of the obstacle vehicle, specifically includes:

[0029] The state quantity set and control quantity set of the obstacle vehicle in the preset distance range are obtained to obtain the discrete state space of the obstacle vehicle;

[0030] The predicted state of the obstacle vehicle is determined based on the preset microscopic traffic model and the discrete state space.

[0031] According to the present application, a kind of fusion predictive lane-changing cruise control method is provided, the current state of the obstacle vehicle in the preset distance range is determined to determine the predicted state of the obstacle vehicle;After the lane is planned to the discretized behavior state space of the energy-saving vehicle speed sequence corresponding to the section of self-driving vehicle, the method further includes:

[0032] In the case that the state sequence of the self-driving vehicle is not obtained, prompt information that cannot change lanes is sent to the vehicle end, to carry out preset cruise control of single lane to the self-driving vehicle.

[0033] The present application also provides a kind of fusion predictive lane-changing cruise control device, including: energy-saving vehicle speed planning module, for based on static road information, planning energy-saving vehicle speed sequence of self-driving vehicle;

[0034] Behavior state discretization module, for determining the predicted state of the obstacle vehicle according to the current state of the obstacle vehicle in the preset distance range;The discretized behavior state space of the self-driving vehicle in the section is constructed according to the energy-saving vehicle speed sequence;

[0035] Lane-changing state determination module, for determining the lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle;

[0036] a state sequence determination module configured to obtain a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so as to perform the fusion predictive lane-changing cruise control on the self-driving vehicle.

[0037] The application further provides a vehicle comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the fusion predictive lane-changing cruise control method according to any one of the above when executing the computer program.

[0038] The fusion predictive lane-changing cruise control method and device provided by the application can determine the lane-changing state point of the self-driving vehicle in the discrete behavior state space of the self-driving vehicle in the energy-saving vehicle speed corresponding road section by using the static road information and combining the predicted state of the obstacle vehicle in the preset distance range, obtain the state sequence of the self-driving vehicle according to the energy-saving vehicle speed and the lane-changing state point, so as to control the self-driving vehicle to change lanes at the energy-saving vehicle speed, reduce the gap between the actual self-driving vehicle state and the ideal self-driving vehicle state under the current road characteristics, and increase the energy-saving effect of the self-driving vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0040] Figure 1 is a flowchart of the fusion predictive lane-changing cruise control method provided by the application.

[0041] Figure 2 is a schematic diagram of the preset distance range in the fusion predictive lane-changing cruise control method provided by the application.

[0042] Figure 3 is a schematic diagram of the discrete behavior state space in the fusion predictive lane-changing cruise control method provided by the application.

[0043] Figure 4 is a schematic diagram of the forward simulation in the fusion predictive lane-changing cruise control method provided by the application.

[0044] Figure 5 is a schematic diagram of the lane-changing judgment in the fusion predictive lane-changing cruise control method provided by the application.

[0045] Figure 6 is one of the framework schematic diagrams of the fusion predictive lane-changing cruise control method provided by the application.

[0046] Figure 7 Figure 2 is a schematic diagram of a framework of the fusion type predictive lane-changing cruise control method provided by the present application.

[0047] Figure 8 Figure 3 is a schematic diagram of a framework of the fusion type predictive lane-changing cruise control method provided by the present application.

[0048] Figure 9 Figure 4 is a schematic diagram of engine fuel consumption of the fusion type predictive lane-changing cruise control method provided by the present application.

[0049] Figure 10 Figure 5 is a schematic diagram of a framework of the fusion type predictive lane-changing cruise control method provided by the present application.

[0050] Figure 11 Figure 6 is a schematic diagram of a structure of the fusion type predictive lane-changing cruise control device provided by the present application.

[0051] Figure 12 Figure 7 is a schematic diagram of a structure of the vehicle provided by the present application. DETAILED DESCRIPTION

[0052] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0053] The fusion type predictive lane-changing cruise control method and device of the present application will be described below. Figures 1-11

[0054] Figure 1 Figure 1 is a schematic diagram of a flow of the fusion type predictive lane-changing cruise control method provided by the present application, as shown in the figure, the method comprises the following steps. Figure 1

[0055] Step 101, planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information.

[0056] The road information refers to information describing road states, such as slope and speed limit. The road states can include road shapes such as road surface width, curve, road slope, connection relationship between roads such as intersection and tunnel, and road elevation, etc. Among them, the road information can be obtained from a remote end, or can be preset in the main body of the method, etc. Preferably, the road information in the present embodiment can be three-dimensional road information.

[0057] ​​Exemplarily, a preset power system and vehicle speed cooperative optimization strategy can be adopted based on static road information to plan a gear sequence with which the self-driving vehicle travels at a lowest cost in a target road section, and a vehicle speed under the gear sequence.

[0058] In step 102, a predicted state of the obstacle vehicle is determined according to a current state of the obstacle vehicle within a preset distance range, and a discretized behavior state space of the self-driving vehicle in the road section is constructed according to the energy-saving vehicle speed sequence.

[0059] The obstacle vehicle refers to a vehicle surrounding the ego vehicle within the preset distance range, including a front vehicle, a rear vehicle and a lateral vehicle. The preset distance range can be set according to actual needs, which is not limited in the embodiment. Exemplarily, as shown in Figure 2 the preset distance range can be a range of 1000 meters in front of the ego vehicle and 200 meters behind the ego vehicle, so as to cooperate with the existing traffic convention, obtain the sudden traffic condition in front of the ego vehicle, and adjust the cruise control.

[0060] As shown in Figure 3 exemplarily, the number of discrete steps can be determined based on a time for the self-driving vehicle to travel from a current stage to a first stage in the planned path according to the energy-saving vehicle speed and a discrete step length, and the behavior state space of the road section corresponding to the energy-saving vehicle speed is discretized according to the number of discrete steps and the lane. The discrete step length can be 0.5s, and the number of discrete steps can be , , is the time for the energy-saving vehicle speed to travel from the current stage to the first stage in the planned path, and s1, s2, …, sk, …, LC=0 represents a left lane, LC=1 represents a middle lane, and LC=2 represents a right lane. The end point can represent an end point of the road section corresponding to the energy-saving vehicle speed.

[0061] The state of the obstacle vehicle refers to a state of the obstacle vehicle related to lane changing with the self-driving vehicle. The current state can be a state of the obstacle vehicle under a current behavior state point in the discretized behavior state space, and the predicted state can be a state of the obstacle vehicle under a subsequent behavior state point in the discretized behavior state space.

[0062] The lane refers to a specific area guiding the vehicle to travel in a fixed direction. The number of discrete steps refers to the number of discrete behavior state points into which the behavior state is divided, which can be used to describe the number of times the behavior state is updated or calculated within a specified range.

[0063] The discretized behavior state space includes a plurality of behavior state points. Exemplarily, the number of behavior state points under each discrete step can correspond to the number of lanes.

[0064] Step 103, determining a lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle.

[0065] The lane-changing state point refers to a behavior state point of the autonomous vehicle that can safely change lanes at the energy-saving vehicle speed under the corresponding predicted state of the obstacle vehicle. The lane-changing state point includes a discrete step and a lane-changing behavior, and the lane-changing behavior includes left lane-changing or right lane-changing.

[0066] Step 104, obtaining a state sequence of the autonomous vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so as to perform the fusion predictive lane-changing cruise control on the autonomous vehicle.

[0067] The state sequence includes the energy-saving vehicle speed sequence and a behavior state at each discrete step, and the behavior state includes lane-changing behavior and non-lane-changing. It can be understood that the behavior state point corresponding to the discrete step other than the discrete step corresponding to the lane-changing state point is non-lane-changing, so that the state sequence completed by the autonomous vehicle can be obtained.

[0068] The fusion predictive lane-changing cruise control method provided by the embodiment of the present application plans an energy-saving vehicle speed sequence of the autonomous vehicle based on road information, determines a lane-changing state point of the autonomous vehicle in the discretized behavior state space of the autonomous vehicle in the energy-saving vehicle speed sequence corresponding to the road segment in combination with the predicted state of the obstacle vehicle within the preset distance range, and obtains a state sequence of the autonomous vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so as to control the autonomous vehicle to change lanes at the energy-saving vehicle speed, reduce the gap between the actual autonomous vehicle state of the lane-changing and the ideal autonomous vehicle state under the current road characteristics, and increase the energy-saving effect of the autonomous vehicle.

[0069] It can be understood that the main purpose of the energy-saving vehicle speed sequence of the autonomous vehicle planned based on the static road information is to reduce energy consumption such as power consumption and fuel consumption. In the actual traffic, under the complex situation of vehicle interaction in the multi-lane traffic flow, directly changing lanes based on the energy-saving vehicle speed sequence may cause the autonomous vehicle to miss the lane-changing opportunity, thereby increasing the energy consumption of the vehicle as a whole and reducing the energy-saving effect of the autonomous vehicle.

[0070] To solve this technical problem, based on the above embodiment, the lane-changing state point of the autonomous vehicle is determined from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle, and specifically includes:

[0071] Performing inflation processing on the energy-saving vehicle speed sequence to obtain a group of simulated vehicle speed space points centered on the energy-saving vehicle speed;

[0072] Determining the lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the simulated vehicle speed space points and the predicted state of the obstacle vehicle.

[0073] Exemplarily, the speed state space of the simulation speed space point can be further discretized by the following formula to obtain a speed state space of the simulation speed space point in order to improve the feasible solution space of the lane changing problem of the self-driving vehicle under the condition of the energy-saving speed as close as possible.

[0074]

[0075] wherein, the speed state space of the simulation speed space point, the energy-saving speed sequence of the self-driving vehicle.

[0076] The speed threshold between adjacent simulation speeds in the speed state space of the simulation speed space point can be set according to the actual working condition, which is not limited in the embodiment.

[0077] The working principle and technical effect of determining the lane changing state point of the self-driving vehicle from the discretized behavior state space based on the simulation speed and the predicted state of the obstacle vehicle are basically the same as those of determining the lane changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving speed sequence and the predicted state of the obstacle vehicle, which will not be repeated here. The difference is that a plurality of simulation speeds are included under the same discretization step, and it is necessary to judge whether the behavior state point is a lane changing state point based on the predicted state of the obstacle vehicle and each simulation speed of each behavior state point.

[0078] In the embodiment, the energy-saving speed sequence is expanded to make the speed of each behavior state point be one of at least two simulation speeds, so that the self-driving vehicle can select the best lane changing opportunity in a wider speed range and better adapt to the complex changes in the dynamic traffic environment.

[0079] Frequent lane changing can easily lead to unstable vehicle body, increase the risk of rollover or loss of control, and frequent lane changing can make it difficult for surrounding vehicles to predict the driving intention of the vehicle, increasing the risk of collision. In order to reduce traffic risk, in an embodiment, before determining the lane changing state point from the discretized behavior state space according to the simulation speed and the state of the obstacle vehicle within the preset distance range, the method further comprises: determining that the state point interval between the last lane changing state point and the target behavior state point is within the allowed lane changing interval range.

[0080] The target behavior state point refers to the behavior state point to be judged whether it is a lane changing state point. Exemplarily, the lane changing interval range can be greater than 6 behavior state points. In the case where the lane changing interval between the target behavior state point and the last lane changing state point is 5 behavior state points, it is directly determined that the behavior state point is a non-lane changing state point.

[0081] It needs greater steering force and longer distance to cross two lanes at one time, which is easy to cause the vehicle to lose control or roll over, and crossing two lanes at one time increases the blind area of the driver and increases the risk of collision with other vehicles. In order to reduce the risk of traffic, in an embodiment, the lane changing behavior in the lane changing state point is that the autonomous vehicle turns from the lane to the adjacent lane.

[0082] As shown in Figure 4 Based on any of the above embodiments, the autonomous vehicle state sequence is obtained according to the energy-saving vehicle speed sequence and the lane changing state point, specifically including:

[0083] Determining the driving cost of each lane changing state point and the corresponding simulation vehicle speed;

[0084] Forward simulation of different lane changing state points and simulation vehicle speeds obtains a plurality of simulation state sequences, and determines the driving cost of each simulation state sequence;

[0085] By comparing the driving costs of each simulation state sequence, the autonomous vehicle state sequence with the minimum cost is obtained; the autonomous vehicle state sequence includes a vehicle speed sequence and a behavior sequence.

[0086] Forward simulation means that the autonomous vehicle is simulated in the possible driving state from the current state. Tree search can be performed in the driving space to realize the cycle forward simulation of the driving cost of the autonomous vehicle under different behaviors and vehicle speeds. For example, the autonomous vehicle starts from the current state and drives at the energy-saving vehicle speed for 3 behavior state points, and then changes lanes to the right at the first simulation vehicle speed at the 4th behavior state point, obtaining the vehicle speed sequence in the first simulation state sequence: energy-saving vehicle speed, energy-saving vehicle speed, energy-saving vehicle speed, first simulation vehicle speed; behavior sequence: no lane change, no lane change, no lane change, lane change to the right.

[0087] In this embodiment, by simulating the driving cost of the host vehicle under different behaviors and vehicle speeds, the driving strategy can be adjusted in real time according to the dynamic traffic environment, and the risk of frequent and large changes in vehicle speed under complex traffic conditions and increased energy consumption is reduced.

[0088] Based on any of the above embodiments, before determining the driving cost of each lane changing state point and the corresponding simulation vehicle speed, the method further includes:

[0089] Establishing a cost model based on the fuel consumption, driving efficiency of the autonomous vehicle at the lane changing state point, and the speed deviation between the simulation vehicle speed and the energy-saving vehicle speed;

[0090] Determining the driving cost of each lane changing state point and the corresponding simulation vehicle speed, specifically including:

[0091] Determining the driving cost of each lane changing state point and the corresponding simulation vehicle speed based on the cost model.

[0092] wherein the driving efficiency refers to a parameter for measuring the consumption of resources such as electric quantity, fuel quantity, etc. of the vehicle in completing a driving task. The driving efficiency of the vehicle can be measured by the driving distance of the vehicle in the same time, or the fluctuation rate of the vehicle speed, etc.

[0093] Exemplarily, the fuel consumption of the self-driving vehicle can be calculated by the following formula:

[0094]

[0095] wherein, is the fuel consumption of the self-driving vehicle, is the instantaneous fuel consumption of the engine, is the engine speed, is the engine torque.

[0096] The driving efficiency can be calculated by the following formula:

[0097]

[0098] wherein, is the driving efficiency, is the, is

[0099] The speed deviation between the simulated speed and the energy-saving speed can be calculated by the following formula:

[0100]

[0101] wherein, is the speed deviation, is the energy-saving speed, is the simulated speed.

[0102] In an embodiment, the fuel consumption of the self-driving vehicle, the driving efficiency and the speed deviation can be weighted to establish a cost model. The specific weighting method can be determined according to actual needs, which is not limited in the present embodiment.

[0103] Exemplarily, the fuel consumption of the self-driving vehicle, the driving efficiency and the speed deviation can be weighted to establish a cost model based on the following formula:

[0104]

[0105] wherein, is a weight factor, i=1, 2, 3, 4, and is a standardization constant for standardizing the three utility functions;

[0106]

[0107]

[0108]

[0109] and are the upper and lower bounds of the driving speed of the autonomous vehicle at the discrete speed, respectively, and the fuel consumption of the vehicle when driving on the corresponding slope.

[0110] In this embodiment, by combining the fuel consumption, driving efficiency and speed deviation of the autonomous vehicle, the comprehensive optimization of the vehicle driving cost is realized, so as to balance the driving performance of the vehicle in different dimensions, improve the energy saving effect of the vehicle while ensuring the driving efficiency, and improve the user experience.

[0111] On the basis of the foregoing embodiments, the predictive optimal passage problem can be constructed based on the state matrix, the constraint condition and the driving cost function, and then the state sequence of the autonomous vehicle with the minimum cost is found through forward behavior tree search, as shown in the following formula

[0112]

[0113] As shown in Figure 5 , in order to reduce the collision risk when the vehicle changes lanes, based on any one of the foregoing embodiments, the lane changing state point of the autonomous vehicle is determined from the discretized behavior state space based on the simulated speed and the predicted state of the obstacle vehicle, specifically including:

[0114] Based on the simulated speed and the predicted state of the obstacle vehicle, the behavior state point meeting the preset lane changing safety standard is determined from the discretized behavior state space as the lane changing state point of the autonomous vehicle.

[0115] The specific lane changing safety standard can be set as required, and this embodiment does not make further limitation thereon. Exemplarily, the relative distance between the autonomous vehicle and the vehicle in the adjacent lane and the safe lane changing distance can be determined, if , the vehicle cannot change lanes, otherwise the vehicle can change lanes. The following formula can be used for determination:

[0116]

[0117] wherein, is the simulated speed of the ego vehicle selected in the fusion stage in the state space, is the set following distance.

[0118] may be taken as , wherein, taking the current lane as a 1-lane, the lane on one side of the current lane as a 0-lane, and the lane on the other side of the current lane as a 2-lane as an example, is the distance between the self-driving vehicle and the front vehicle in the 0-lane, is the distance between the self-driving vehicle and the rear vehicle in the 0-lane, is the distance between the self-driving vehicle and the front vehicle in the current lane, is the distance between the self-driving vehicle and the front vehicle in the 2-lane, is the distance between the self-driving vehicle and the rear vehicle in the 2-lane. Figure 5 In the above formula, the 0-lane is the lane with i = 0, the 1-lane is the lane with i = 1, and the 2-lane is the lane with i = 2.

[0119] Based on any of the above embodiments, based on the static road information, a sequence of energy-saving speeds of the self-driving vehicle is planned, specifically including:

[0120] determining a planning distance domain of the sequence of energy-saving speeds of the self-driving vehicle, dividing the planning distance domain into at least two stages based on the road information, and planning the sequence of energy-saving speeds of the self-driving vehicle for each stage;

[0121] when the self-driving vehicle reaches the second stage, repeating the step of planning the energy-saving speed of the self-driving vehicle to dynamically plan the sequence of energy-saving speeds of the self-driving vehicle.

[0122] The planning distance domain refers to the distance in front of the self-driving vehicle for which the energy-saving speed is to be planned. For example, the planning distance domain can be 1 kilometer in front of the vehicle, or 1.5 kilometers in front of the vehicle, etc.

[0123] For example, it can be determined that the distance of 1 kilometer in front of the vehicle is the planning distance domain of the energy-saving speed, and the planning distance domain is divided into at least two stages according to the road information such as tunnels, speed limits, and slopes within the distance of 1 kilometer in front of the vehicle.

[0124] According to the lane, the sequence of energy-saving speeds corresponds to a discretized behavior state space, specifically including:

[0125] According to the lane and the number of discrete steps, the sequence of energy-saving speeds of the first stage in the planning distance domain corresponds to a discretized behavior state space.

[0126] For example, the number of discrete steps can be obtained according to the discrete step length 0.5s, , , , wherein The discrete behavior state space can be obtained based on the lane and the discrete step number, and the behavior state of the first stage in the planning distance domain of the self-driving vehicle. When the self-driving vehicle reaches the second stage, the first stage in the new planning distance domain is planned according to the new planning distance domain obtained by repeatedly planning the energy-saving speed of the self-driving vehicle, and the discrete behavior state space corresponding to the behavior state of the first stage in the new planning distance domain is obtained.

[0127] In the embodiment, the discrete behavior state space for dynamically controlling the lane changing behavior is corresponded to the first stage of the static planning energy-saving speed, so that the influence of the energy-saving speed can be fully considered in the lane changing process, so that the energy-saving behavior state can be selected in the lane changing process of the self-driving vehicle based on the energy-saving speed, and the energy consumption of the self-driving vehicle is reduced.

[0128] Based on any of the above embodiments, the current state includes a set of state variables and a set of control variables, and the predicted state of the obstacle vehicle is determined according to the current state of the obstacle vehicle within the preset distance range, specifically including:

[0129] The discrete state space of the obstacle vehicle is obtained according to the set of state variables and the set of control variables of the obstacle vehicle within the preset distance range;

[0130] The predicted state of the obstacle vehicle is determined based on the preset microscopic traffic model and the discrete state space.

[0131] The set of state variables refers to a set of variables representing the state of the vehicle at a moment. For example, the set of state variables can be represented by the following formula:

[0132]

[0133] wherein, represents the longitudinal position of the vehicle at the moment, represents the vehicle speed at the moment, represents the current lane number of the vehicle at the moment, , when the vehicle is in the middle lane , in the right lane , in the left lane , the first vehicle in the middle lane , the second vehicle in the middle lane , the nth vehicle in the middle lane , the first vehicle in the left lane , the second vehicle in the left lane , the pth vehicle in the left lane , the self-driving vehicle , the first vehicle in the middle lane , the second vehicle in the middle lane , the nth vehicle in the middle lane , the first vehicle in the left lane Indicates the first car in the right lane. Indicates the second car in the right lane. represents the qth vehicle in the right lane.

[0134] The control variable set refers to the set of variables that characterize the change in vehicle state at a given moment. For example, the control variable set can be expressed as follows:

[0135]

[0136] in, express The vehicle's acceleration at that moment, express The vehicle's left lane change decision at time , 0 means no lane change, 1 means changing lane to the left. express The vehicle's lane change decision to the right at time t is determined in the same way as the lane change to the left, and SV is the self-driving vehicle.

[0137] For example, the discrete state space of the obstacle vehicle can be expressed by the following formula:

[0138]

[0139] Where the discrete time step is ΔT, and A and B are the coefficient matrices of the state space equation, as shown below:

[0140]

[0141] In the state quantity set The longitudinal position of the vehicle at the time instant can be calculated using a vehicle following model. For example, the acceleration of a vehicle following a traffic vehicle can be calculated using an Intelligent Driver Model (IDM) as shown below:

[0142]

[0143] in, Indicates the maximum acceleration of the vehicle, represents the desired speed of the vehicle, express The speed of any vehicle at time, express The longitudinal relative speed of any vehicle and its preceding vehicle at time , Indicates the acceleration index, usually taken as , express The actual following distance of any vehicle at time, express Any vehicle desired following distance at time t, Minimum vehicle distance at static state, Desired vehicle headway, Comfortable deceleration set by vehicle.

[0144] In order to avoid lane changing operation in unknown state and reduce potential risk, based on any of the above embodiments, the predicted state of the obstacle vehicle is determined according to the current state of the obstacle vehicle within the preset distance range; after the discrete behavior state space of the autonomous vehicle in the energy-saving vehicle speed corresponding section is planned according to the lane, the method further comprises: under the condition that the state sequence of the autonomous vehicle is not obtained, sending a lane changing impossible prompt information to the vehicle end to perform single-lane preset cruise control on the autonomous vehicle.

[0145] As Figure 6 shown, exemplarily, based on the power system constraint, speed limit constraint, static slope information, road speed limit, energy consumption model and vehicle dynamics model, the upper layer of the cloud removes the infeasible state transition point under the power system constraint to perform dynamic programming solution, after the vehicle gear and speed coordination planning is completed, the energy-saving vehicle speed can be sent to the lower layer to construct the forward traffic drivable space based on IDM prediction and search the forward behavior tree based on the driving space to perform lane changing fusion processing, and by judging whether the lane changing fusion processing has a solution, different prompt information can be sent to the vehicle end. In the case of having a solution, the first prompt information based on predictive adaptive cruise control can be sent to the vehicle end to control the vehicle through mode 1, and in the case of having no solution, the second prompt information based on lane changing safety judgment and lane changing path planning and tracking control can be sent to the vehicle end to control the vehicle through mode 2.

[0146] In an embodiment, in the case of having no solution, the energy-saving vehicle speed of the upper layer can be sent to the vehicle end to control the vehicle based on the energy-saving vehicle speed.

[0147] In order to specifically explain the function of the fusion predictive lane changing cruise control method provided by the embodiment, a specific example is provided below.

[0148] The cloud control system can also be called a cloud control digital twin system, has the typical characteristics of cyber-physical systems (Cyber-Physical Systems, CPS), and is a new generation of traffic system for real-time collaborative computing of large-scale networked applications.

[0149] As Figure 7As shown in Figure 1, the cloud control system consists of a physical space and a cyberspace. The physical space can include the physical layer, which can also be referred to as the physical system layer or the traffic physics layer, while the cyberspace can be the edge cloud. The physical space encompasses participants in mixed road traffic scenarios, including intelligent connected vehicles (ICVs), human drivers, pedestrians, roadside equipment, communication facilities, and basic traffic elements (such as speed limits, grade curvature, and traffic lights). Roadside sensing devices use cross-domain collaborative perception algorithms to collect real-time status information about vehicles on the road and upload this data to the edge cloud.

[0150] The information space includes the Cloud Control Base Platform (CCBP) and the Cloud Control Application Platform (CCAP). The CCBP can also be called the cloud control foundation layer or the twin state layer, while the CCAP can also be called the cloud control application platform layer or the twin application layer.

[0151] The physical system layer can send information such as physical space elements to the cloud control basic platform layer through real-time data collection. The cloud control basic platform layer can integrate and perceive this information through standardized communications to obtain twin data, perform real-time digital mapping of the physical space, and further perform future inference on the status information of traffic participants, such as behavior recognition and intention estimation, vehicle speed prediction, and trajectory prediction, to obtain basic data and send this basic data to the cloud control application platform layer. The cloud control application platform layer can deeply mine and utilize static road information and dynamic traffic reasoning information, perform control calculations through vehicle safety collaborative application 1, energy-saving collaborative application 2, and efficient collaborative application n, and send the optimization results as application outputs to the cloud control basic platform layer. The cloud control basic platform layer then performs integrated control to generate instructions, which are then issued to the physical space, thus realizing the integrated application of digital twin technology in the cloud control system.

[0152] like Figure 8 As shown, in this embodiment, the cloud control system's cloud control application platform can implement a fused predictive lane change cruise control method. The upper cloud layer combines information such as speed limits, slopes, and weather conditions provided by static road maps, removes infeasible state transition points under powertrain constraints, and performs energy-saving speed planning for the self-driving vehicle. During the fusion process, only the first phase of the planned energy-saving speed range domain is used, which can be referred to as the first-phase fusion. Roadside perception and vehicle-road perception fusion can be performed through multiple roadside collaborative base stations, and road information can be sent to the edge cloud. The edge cloud's twin state layer then performs twin mapping of road objects (such as location, size, and type) and environmental factors such as the road surface and weather. After digital inference is performed on traffic participant behavior, speed, and trajectory predictions to generate twin data, dynamic traffic participant prediction information is sent to the twin application layer for layered fusion of dynamic and static information, and the results are then distributed to the traffic physical layer.

[0153] Specifically, the optimal energy-saving speed and gear sequence of the self-driving vehicle can be planned based on road information. The lane-changing behavior and speed can be further integrated based on the energy-saving speed of the upper cloud layer at the lower cloud layer. The speed of the forward traffic vehicle is predicted by using a microscopic traffic model to construct the drivable space of the self-driving vehicle. Then, with the goal of minimizing driving costs, forward simulation is performed using tree search and other methods to obtain the speed sequence and behavior sequence of the self-driving vehicle with the lowest cost, which is then sent to the vehicle end to perform integrated predictive lane change cruise control on the self-driving vehicle.

[0154] Moreover, if the self-driving vehicle state sequence is not obtained, a prompt message indicating that lane change is impossible can be sent to the vehicle end, that is, there is no optimal lane change speed fusion sequence, and the system can enter a preset control mode such as Predictive Adaptive Cruise Control (PACC) / Predictive Cruise Control (PCC) to perform single-lane energy-saving driving control.

[0155] Specifically, based on the vehicle dynamics model and the engine's high-efficiency range, static slope information plays a crucial role in controlling the vehicle's powertrain. To improve the vehicle's energy efficiency, the cloud can acquire road slope information ahead of the vehicle, enabling predictive planning of the vehicle's speed and powertrain.

[0156] Exemplarily, the vehicle driving equation is discretized in the distance domain by the forward Euler method to obtain the vehicle state transfer equation shown below.

[0157]

[0158] in, for k The speed of the stage, for k+ The speed of stage 1, is the distance between two adjacent stages; For the stage k The total transmission ratio of the driveline; For the stage k of engine torque; For the stage k The road slope angle; is the rotation mass conversion factor; is the mass of the car (kg); Acceleration (m / s 2 ); is the transmission gear ratio; is the mechanical efficiency of the transmission system; is the tire radius; is the main reducer transmission ratio; is the air resistance coefficient; is the wind area; is the gravity acceleration; is the rolling resistance coefficient.

[0159] In the vehicle state transition equation, the vehicle speed v is the state quantity, the gear ge and the torque e are the control quantities, and through reasonable planning of these control quantities, effective control of the vehicle state can be achieved to achieve the goal of energy-saving driving.

[0160] The target function for calculating the driving cost can be obtained based on the engine instantaneous fuel consumption and the driving efficiency of the vehicle as follows:

[0161]

[0162] wherein, is the engine instantaneous fuel consumption model in vehicle driving, and are the weighting coefficients;

[0163] The corresponding engine fuel consumption model is shown in Figure 9 , which can be represented by the following formula:

[0164]

[0165] is the engine instantaneous fuel consumption, with the unit of kg / s; is the coefficient of each term of the polynomial; is the engine speed, with the unit of rpm, is the engine torque, with the unit of N·m.

[0166] Thus, on the basis of minimizing the driving cost, the energy-saving vehicle speed model of the self-driving vehicle can be obtained in combination with the constraints of the vehicle power system and the traffic speed limit as follows:

[0167]

[0168] wherein, , are the minimum and maximum torque boundaries; , are the minimum and maximum vehicle speed boundaries; , are the minimum and maximum acceleration boundaries; , are minimum and maximum speed boundaries; , are minimum and maximum gear boundaries.

[0169] As Figure 10 shown, according to the calculation principle of the dynamic programming algorithm, a phase, gear ge and vehicle speed v three-dimensional state space can be constructed, and then the infeasible state points under the constraints of road speed limit and state transition equation are removed. Further, according to the reverse calculation and forward solving process, the optimal vehicle speed and gear sequence state points are calculated to plan the energy-saving speed of the autonomous vehicle. The phase can be a future distance.

[0170] After obtaining the energy-saving speed of the autonomous vehicle, the speed of the vehicle in front of the road can be predicted based on the IDM model to obtain the drivable space of the autonomous vehicle at each time in the future without collision with other vehicles; the energy-saving speed is used to obtain the discretized behavior state space of the autonomous vehicle in the corresponding section of the lane planning; the energy-saving speed is expanded to obtain at least two simulation speeds; based on the simulation speed and the predicted state of the obstacle vehicle, the behavior state point that meets the preset lane-changing safety standard is determined from the discretized behavior state space as the lane-changing state point of the autonomous vehicle; a cost model is established based on the fuel consumption, driving efficiency of the autonomous vehicle based on the lane-changing state point, and the speed deviation between the simulation speed and the energy-saving speed; the driving cost of each lane-changing state point and the corresponding simulation speed is determined based on the cost model; the driving cost of each lane-changing state point and the corresponding simulation speed is determined; forward simulation of different lane-changing state points and simulation speeds obtains a plurality of simulation state sequences, and the driving cost of each simulation state sequence is determined; the driving costs of each simulation state sequence are compared to obtain the speed sequence and behavior sequence of the autonomous vehicle with the minimum cost, so as to perform fusion predictive lane-changing cruise control on the autonomous vehicle.

[0171] The fusion predictive lane-changing cruise control device provided by the present application is described below, and the fusion predictive lane-changing cruise control device described below can be referred to in conjunction with the fusion predictive lane-changing cruise control method described above.

[0172] Figure 11 An example of a structural schematic diagram of a fusion predictive lane-changing cruise control device is shown in Figure 11 , which comprises:

[0173] An energy-saving speed planning module 1101 is configured to plan an energy-saving speed sequence of an autonomous vehicle based on static road information;

[0174] The behavior state discretization module 1102 is configured to determine a predicted state of the obstacle vehicle according to a current state of the obstacle vehicle within a preset distance range; and construct a discretized behavior state space of the autonomous vehicle in the road section according to the energy-saving vehicle speed sequence.

[0175] The lane-changing state determination module 1103 is configured to determine a lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle.

[0176] The state sequence determination module 1104 is configured to obtain a state sequence of the autonomous vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so as to perform the fusion-type predictive lane-changing cruise control on the autonomous vehicle.

[0177] According to any one of the above embodiments, the lane-changing state determination module 1103 is specifically configured to:

[0178] perform inflation processing on the energy-saving vehicle speed sequence to obtain a group of simulated vehicle speed space points centered on the energy-saving vehicle speed points;

[0179] determine the lane-changing state point of the autonomous vehicle from the discretized behavior state space based on the simulated vehicle speed space points and the predicted state of the obstacle vehicle.

[0180] According to any one of the above embodiments, the state sequence determination module 1104 is specifically configured to:

[0181] determine a driving cost corresponding to each lane-changing state point and the simulated vehicle speed;

[0182] forward simulate a plurality of simulated state sequences by using different lane-changing state points and simulated vehicle speeds, and determine driving costs of the simulated state sequences;

[0183] compare the driving costs of the simulated state sequences to obtain a state sequence of the autonomous vehicle with a minimum cost; the state sequence of the autonomous vehicle includes a vehicle speed sequence and a behavior sequence.

[0184] According to any one of the above embodiments, the fusion-type predictive lane-changing cruise control device further includes a driving cost determination module configured to:

[0185] establish a cost model based on fuel consumption, driving efficiency of the autonomous vehicle at the lane-changing state point, and a speed deviation between the simulated vehicle speed and the energy-saving vehicle speed;

[0186] determine the driving cost corresponding to each lane-changing state point and the simulated vehicle speed, specifically including:

[0187] determine the driving cost corresponding to each lane-changing state point and the simulated vehicle speed based on the cost model.

[0188] Based on any of the above embodiments, the lane-changing state determination module 1103 is specifically configured to determine, from the discretized behavior state space, a behavior state point that meets a preset lane-changing safety standard as the lane-changing state point of the self-driving vehicle based on the simulated vehicle speed and the predicted state of the obstacle vehicle.

[0189] Based on any of the above embodiments, the energy-saving vehicle speed planning module 1101 is specifically configured to determine a planning distance domain of the energy-saving vehicle speed sequence of the self-driving vehicle, divide the planning distance domain into at least two stages based on road information, and plan the energy-saving vehicle speed sequence of the self-driving vehicle for each stage.

[0190] When the self-driving vehicle reaches the second stage, the step of planning the energy-saving vehicle speed of the self-driving vehicle is repeated to dynamically plan the energy-saving vehicle speed of the self-driving vehicle.

[0191] The discretized behavior state space corresponding to the energy-saving vehicle speed sequence planned according to the lane specifically includes:

[0192] The discretized behavior state space corresponding to the energy-saving vehicle speed sequence of the first stage in the distance domain planned according to the lane and the discrete step number.

[0193] Based on any of the above embodiments, the current state includes a state quantity set and a control quantity set; and the behavior state discretization module 1102 is specifically configured to:

[0194] The discretized state space of the obstacle vehicle is obtained according to the state quantity set and the control quantity set of the obstacle vehicle within a preset distance range;

[0195] The predicted state of the obstacle vehicle is determined based on a preset microscopic traffic model and the discretized state space.

[0196] Based on any of the above embodiments, the fusion predictive lane-changing cruise control device further includes a prompt information sending module configured to, in the case where the self-driving vehicle state sequence is not obtained, send prompt information of being unable to change lanes to a vehicle end to perform preset cruise control on the self-driving vehicle in a single lane.

[0197] Figure 12 An example of a structural schematic diagram of a vehicle is shown as Figure 12As shown, the vehicle can include a processor 1210, a communications interface 1220, a memory 1230, and a communications bus 1240, wherein the processor 1210, the communications interface 1220, and the memory 1230 communicate with each other through the communications bus 1240. The processor 1210 can invoke the logic instructions in the memory 1230 to execute the fusion predictive lane-changing cruise control method, which includes: planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information; determining a predicted state of an obstacle vehicle in a preset distance range according to a current state of the obstacle vehicle; constructing a discretized behavior state space of the self-driving vehicle in a road section according to the energy-saving vehicle speed sequence; determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; obtaining a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, to perform fusion predictive lane-changing cruise control on the self-driving vehicle.

[0198] In addition, the logic instructions in the memory 1230 described above can be implemented in the form of a software functional unit and sold or used as an independent product, which can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or the part that contributes to the prior art, or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0199] In another aspect, the present application also provides a computer program product comprising a computer program, which can be stored on a non-transitory computer readable storage medium, and the computer program is executable by a processor to enable a computer to perform the fusion predictive lane change cruise control method provided by the above-mentioned methods, which comprises: planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information; determining a predicted state of an obstacle vehicle according to a current state of the obstacle vehicle within a preset distance range; constructing a discretized behavior state space of the self-driving vehicle in a road section according to the energy-saving vehicle speed sequence; determining a lane change state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; and obtaining a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane change state point to perform fusion predictive lane change cruise control on the self-driving vehicle.

[0200] In yet another aspect, the present application also provides a non-transitory computer readable storage medium having a computer program stored thereon, and the computer program is executable by a processor to implement the fusion predictive lane change cruise control method provided by the above-mentioned methods, which comprises: planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information; determining a predicted state of an obstacle vehicle according to a current state of the obstacle vehicle within a preset distance range; constructing a discretized behavior state space of the self-driving vehicle in a road section according to the energy-saving vehicle speed sequence; determining a lane change state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; and obtaining a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane change state point to perform fusion predictive lane change cruise control on the self-driving vehicle.

[0201] The device embodiments described above are only schematic, wherein the units illustrated as separate components can or can not be physically separate, and the components illustrated as units can or can not be physical units, i.e., can be located in one place or distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purposes of the present embodiment scheme. Those skilled in the art can understand and implement without creative labor.

[0202] Those skilled in the art can clearly understand the technical solutions of the various embodiments from the above description of the embodiments, and the various embodiments can be implemented by means of software with the necessary general hardware platforms, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that makes a contribution, can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0203] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some technical features therein; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A fusion predictive lane change cruise control method, characterized by, The method comprises the following steps: planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information; determining a predicted state of an obstacle vehicle according to a current state of the obstacle vehicle within a preset distance range, and constructing a discretized behavior state space of the self-driving vehicle in a road section according to the energy-saving vehicle speed sequence; determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; obtaining a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, so as to perform a fusion type predictive lane-changing cruise control on the self-driving vehicle; determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle, specifically comprising the following steps: performing inflation processing on the energy-saving vehicle speed sequence to obtain a group of simulated vehicle speed space points centered on the energy-saving vehicle speed points; determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the simulated vehicle speed space points and the predicted state of the obstacle vehicle; planning an energy-saving vehicle speed sequence of a self-driving vehicle based on static road information, specifically comprising the following steps: determining a planning distance domain of the energy-saving vehicle speed sequence of the self-driving vehicle, dividing the planning distance domain into at least two stages based on road information, and planning the energy-saving vehicle speed sequence of the self-driving vehicle for each stage; when the self-driving vehicle reaches the second stage, repeating the step of planning the energy-saving vehicle speed sequence of the self-driving vehicle to dynamically plan the energy-saving vehicle speed sequence of the self-driving vehicle; planning a discretized behavior state space corresponding to the energy-saving vehicle speed sequence according to a lane, specifically comprising the following steps: planning a discretized behavior state space corresponding to the energy-saving vehicle speed sequence of the first stage in the distance domain according to a lane and a discrete step number.

2. The fused predictive lane change cruise control method of claim 1, wherein, obtaining a state sequence of the self-driving vehicle according to the energy-saving vehicle speed sequence and the lane-changing state point, specifically comprising the following steps: determining a driving cost of each lane-changing state point and a corresponding simulated vehicle speed; forward simulating a plurality of simulated state sequences by using different lane-changing state points and simulated vehicle speeds, and determining driving costs of the simulated state sequences; comparing the driving costs of the simulated state sequences to obtain a state sequence of the self-driving vehicle with the minimum cost; the state sequence of the self-driving vehicle comprises a vehicle speed sequence and a behavior sequence.

3. The fused predictive lane change cruise control method of claim 2, wherein, Before determining the driving cost of each lane-changing state point and the corresponding simulated vehicle speed, the method further comprises the following steps: establishing a cost model based on the fuel consumption, driving efficiency of the self-driving vehicle at the lane-changing state point, and a speed deviation between the simulated vehicle speed and the energy-saving vehicle speed; determining the driving cost of each lane-changing state point and the corresponding simulated vehicle speed, specifically comprising the following steps: determining the driving cost of each lane-changing state point and the corresponding simulated vehicle speed based on the cost model.

4. The fused predictive lane change cruise control method of claim 1, wherein, determining a lane-changing state point of the self-driving vehicle from the discretized behavior state space based on the simulated vehicle speed and the predicted state of the obstacle vehicle, specifically comprising the following steps: determining, from the discretized behavior state space based on the simulated vehicle speed and the predicted state of the obstacle vehicle, a behavior state point meeting a preset lane-changing safety standard as the lane-changing state point of the self-driving vehicle.

5. The fused predictive lane change cruise control method of claim 1, wherein, The current state includes a set of state variables and a set of control variables, and the predicted state of the obstacle vehicle is determined according to the current state of the obstacle vehicle within a preset distance range, and specifically includes: The set of state variables and the set of control variables of the obstacle vehicle within the preset distance range are obtained to obtain a discrete state space of the obstacle vehicle; The predicted state of the obstacle vehicle is determined based on a preset microscopic traffic model and the discrete state space.

6. The fused predictive lane-change cruise control method of claim 1, wherein The predicted state of the obstacle vehicle is determined according to the current state of the obstacle vehicle within a preset distance range; After the discrete behavior state space of the autonomous vehicle in the corresponding road section of the energy-saving vehicle speed sequence is planned according to the lane, the method further includes: In the case that the state sequence of the autonomous vehicle is not obtained, a lane changing prompt information is sent to the vehicle end to perform a single-lane preset cruise control on the autonomous vehicle.

7. A fusion predictive lane change cruise control device characterized by, It includes: An energy-saving vehicle speed planning module is configured to plan an energy-saving vehicle speed sequence of an autonomous vehicle based on static road information; A behavior state discretization module is configured to determine a predicted state of an obstacle vehicle according to a current state of the obstacle vehicle within a preset distance range, and to construct a discrete behavior state space of the autonomous vehicle in a road section according to the energy-saving vehicle speed sequence; A lane changing state determination module is configured to determine a lane changing state point of the autonomous vehicle from the discrete behavior state space based on the energy-saving vehicle speed sequence and the predicted state of the obstacle vehicle; A state sequence determination module is configured to obtain a state sequence of the autonomous vehicle according to the energy-saving vehicle speed sequence and the lane changing state point, so as to perform a fusion predictive lane changing cruise control on the autonomous vehicle; The lane changing state determination module is specifically configured to: Perform inflation processing on the energy-saving vehicle speed sequence to obtain a group of simulated vehicle speed space points centered on the energy-saving vehicle speed points; Determine a lane changing state point of the autonomous vehicle from the discrete behavior state space based on the simulated vehicle speed space points and the predicted state of the obstacle vehicle; The energy-saving vehicle speed planning module is specifically configured to: Determine a planning distance domain of the energy-saving vehicle speed sequence of the autonomous vehicle, divide the planning distance domain into at least two stages based on road information, and plan the energy-saving vehicle speed sequence of the autonomous vehicle for each stage; When the autonomous vehicle reaches the second stage, the step of planning the energy-saving vehicle speed sequence of the autonomous vehicle is repeated to dynamically plan the energy-saving vehicle speed sequence of the autonomous vehicle; According to the lane, the discrete behavior state space corresponding to the energy-saving vehicle speed sequence is planned, and specifically includes: According to the lane and the discrete step number, the discrete behavior state space corresponding to the energy-saving vehicle speed sequence of the first stage in the distance domain is planned.

8. A self-driving vehicle comprising a memory, a processor, and a computer program stored on the memory and running on the processor, wherein, The processor executes the computer program to realize the fusion predictive lane changing cruise control method according to any one of claims 1 to 6.

Images