CN110599772B

CN110599772B - Mixed traffic flow cooperative optimization control method based on double-layer planning

Info

Publication number: CN110599772B
Application number: CN201910887658.1A
Authority: CN
Inventors: 孙湛博
Original assignee: Southwest Jiaotong University
Current assignee: Southwest Jiaotong University
Priority date: 2019-09-19
Filing date: 2019-09-19
Publication date: 2020-08-28
Anticipated expiration: 2039-09-19
Also published as: CN110599772A

Abstract

The invention provides a double-layer planning-based hybrid traffic flow collaborative optimization control method, and belongs to the field of traffic engineering. The method adopts a double-layer optimization model based on dynamic programming recursion to carry out mixed traffic flow cooperative decision control, and is suitable for different traffic scenes in the mixed traffic flow, including expressway ramp vehicle convergence, intersection vehicle convergence and vehicle passing intersection; the double-layer optimization model comprises an upper layer model and a lower layer model, wherein the upper layer model is a vehicle sequencing problem solved by dynamic programming recursion, and the lower layer model is a single vehicle track optimization problem in different traffic scenes solved by dynamic programming recursion; and establishing an upper layer model and a lower layer model, and solving the models. The upper layer model and the lower layer model jointly ensure the optimal running of the system vehicles, so that the vehicle conflict of the vehicles in the process of converging or passing through the intersection is reduced under the mixed traffic flow environment, and the passing efficiency and the comfort of the vehicles are effectively improved.

Description

Mixed traffic flow cooperative optimization control method based on double-layer planning

Technical Field

The invention relates to a mixed traffic flow collaborative optimization control method based on double-layer planning, and belongs to the field of traffic engineering.

Background

New technologies such as Global Positioning System (GPS), wireless communication, advanced sensing, and automatic control have prompted the rapid development of autonomous vehicles (i.e., intelligent networked vehicles, which are optimally controllable vehicles, CAVs). An autonomous vehicle is defined as a vehicle that is capable of sensing and communicating with the driving environment, and the operation of the vehicle (in part or in whole) may be performed without driver action. Autonomous vehicles have better controllability and cooperativity than the difficult cooperativity of conventional human-driven vehicles (non-optimally controlled vehicles). This may therefore provide benefits such as improved fuel/energy efficiency, traffic safety and traffic stability. Mixed traffic (i.e., a state in which human-driven vehicles and autonomous vehicles are mixed) will become a major mode of road traffic before the hundreds of popularity of autonomous vehicles. Under the environment of mixed traffic flow, under the conditions of various traffic scenes such as ramp junction convergence at a highway intersection, junction intersection or junction at a T-junction and the like, traffic conflict between an automatic driving vehicle and a human driving vehicle can occur, and certain harm is caused. The scientific theoretical framework and the modeling method are used for carrying out mixed traffic flow cooperative decision control, so that traffic conflicts are reduced or eliminated, and the vehicle track is optimized to a certain degree to become a practical problem which needs to face in future traffic for a long time.

Existing research related to autonomous vehicles is mostly based on the assumption that the permeability of the autonomous vehicle is 100%, and is mostly studied from a macroscopic perspective. The model of the microscopic level is also mainly focused on the single-vehicle track optimization of the research automatic driving vehicle, and the optimization of the system level cannot be guaranteed. From a microscopic view point, a mixed traffic flow cooperative decision control optimization model aiming at system optimization is researched. The cooperative decision control under the mixed traffic flow environment is carried out aiming at various microscopic traffic scenes, such as crossroads, T-shaped intersections, expressway ramp confluence and the like, so that traffic conflicts are eliminated to the maximum extent, and the research on traffic operation efficiency and traffic capacity is basically not available.

Disclosure of Invention

The invention aims to provide a mixed traffic flow cooperative optimization control method based on double-layer planning, aiming at the defects in the prior art, and from a microscopic view, researching a mixed traffic flow cooperative decision control optimization model aiming at system optimization. The mixed traffic flow refers to the traffic flow formed by mixing a plurality of vehicles which can be optimally controlled and vehicles which can not be optimally controlled.

The technical scheme adopted by the invention for realizing the aim is as follows:

a mixed traffic flow collaborative optimization control method based on double-layer planning is characterized in that a double-layer optimization model based on dynamic planning recursion is adopted for carrying out mixed traffic flow collaborative decision control, and the method is suitable for different traffic scenes in the mixed traffic flow, wherein the different traffic scenes comprise ramp vehicle convergence on a highway, vehicle confluence at an intersection and vehicle passing through the intersection; the double-layer optimization model comprises an upper layer model and a lower layer model; the upper model is a vehicle sequencing problem solved by dynamic programming recursion; the lower layer model is a single vehicle track optimization problem in different traffic scenes solved by dynamic programming recursion;

the method comprises the following steps:

s1, establishing the upper layer model, including:

s1-1, determining the division stage, the state variable and the decision variable of the upper layer model, which are as follows:

the upper layer model is divided: assuming that an X road and a Y road are two one-way roads with intersections, n vehicles on the X road need to sequentially converge into or pass through m +1 intervals among the vehicles on the Y road, and the behavior that each single vehicle on the X road converges into or passes through the Y road is represented as a stage, and the behavior that the k-th vehicle on the X road converges into or passes through the Y road is recorded as a k-th stage, wherein k is 1,2,3, …, n;

state variables of the upper model: the number of the vehicles which can be converged into the kth vehicle on the X road or pass through the Y road in the kth stage is s_kRepresents;

decision variables of the upper model: the decision made in each stage represents that the k vehicle on the X road in the k stage can merge into or pass through the s road of the Y road_kSelecting a particular xth of individual vehicle intervals_kThe vehicles are alternately converged or passed;

s1-2, determining a state transition equation, a cost function and an objective function of the upper layer model, which are as follows:

the state transition equation of the upper model is as follows:

setting the initial condition as s₀＝m+1；

The state transition equation of the upper model indicates that s is equal to 1 when k_k＝s₁The number of the vehicle intervals for the 1 st vehicle on the X road to enter or pass through the Y road in the 1 st stage is m + 1; when k is 2,3, …, n, the (k-1) th vehicle on the X road selects the (X) th vehicle at the (k-1) th stage_k-1The interval of each vehicle is used as a state variable s after the vehicle is converged or passes through a Y road_kA change in (c); s₀M +1 represents that the number of vehicle intervals for vehicles to merge into or pass through the road Y on the road X in the initial state is m + 1;

cost function of the upper model:

cost function D of the upper model_k(s_k,x_k) Indicating a phase index function required for making a decision at the kth phase, wherein

S representing that the k-th vehicle on the X road is merged into or passes through the road where the k-th vehicle can merge into or pass through the Y road under the action of the cooperative optimization control strategy in the k stage_kAll possible cost costs arising from individual vehicle intervals;

the method is characterized in that the Y road does not directly participate in vehicle convergence or vehicle confluence or traffic flow of vehicles passing through an intersection, and the cost is consumed due to the fact that the speed of the vehicle is adjusted according to the following safety requirement of the vehicle due to the influence of vehicle convergence or passing of the vehicle on the front vehicle;

an objective function of the upper model:

setting the initial condition as f₀(s₀)＝0；

Objective function f of the upper model_k(s_k) Representing the cumulative cost consumption of the system vehicles from stage 1 to stage k, f₀(s₀) 0 denotes that the system cost is 0 in the initial state;

s2, establishing the lower layer model, including:

s2-1, determining a microscopic follow-up model, describing a follow-up state of the vehicle by using the microscopic follow-up model, and predicting an initial track of the vehicle; the following state of the vehicle comprises the speed, acceleration and position of the vehicle;

s2-2, establishing a condition constraint model for judging whether the kth vehicle on the X road at the kth stage can smoothly merge into or pass through the Y road;

s2-3, aiming at various possible situations that occur in the process that a k vehicle on a k-th stage X road converges or passes through a Y road under a mixed traffic scene, simulating a cooperative control strategy set;

s2-4, based on the vehicle initial track predicted in the step S2-1, the condition constraint model established in the step S2-2 sequentially judges that the k-th vehicle on the X-th road of the k stage is converged into or passes through the S-th road on the Y road for converging into or passing through_kThe specific situation occurring in the process of each vehicle interval is determined according to the situation that the k-th vehicle on the X-th road in the k stage can be merged into or passes through the s of the Y road_kSpecific conditions occurring in the process of each vehicle interval respectively make the cooperative control strategy corresponding to the cooperative control strategy set planned in the step S2-3;

s2-5, determining the vehicles which can be optimally controlled in the vehicles participating in the k stage as target vehicles; optimizing the running track of the target vehicle according to the cooperative control strategies made in the step S2-4 respectively, resolving the optimization problem into an optimal control problem of discrete time state constraint, and solving by using a dynamic programming idea to obtain the cooperative optimization control strategy about the target vehicle;

s2-6, calculating the k-th vehicle on the X road or the S which can be converged by the k-th vehicle or pass through the Y road under the action of the cooperative optimization control strategy in the k-th stage_kAll possible cost consumptions of one vehicle interval

(as one input in the upper model dynamic programming recursive solution process);

s3, solving the double-layer optimization model:

s3-1, solving the upper model, and determining the decision made by each stage of the upper model when the accumulated cost consumption of the system vehicle is the lowest;

s3-2, reversely deducing the decision made at each stage of the upper layer model determined in the step S3-1 to obtain the vehicle optimized track of each single vehicle on the X road of the lower layer model converging into or passing through the Y road;

and S4, solving and obtaining a mixed traffic flow cooperative decision of the double-layer optimization model aiming at system optimization by the step S3, and acting the decision on the system vehicle to control the operation of the system vehicle.

Further, the step S2-3 specifically includes:

according to the k stage X road, the k vehicle is in the s available for the k vehicle to merge or pass through the Y road_kSelecting a particular xth of individual vehicle intervals_kThe vehicles gather or pass at intervals, the k-th vehicle on the X road is marked as a vehicle k, and the X-th vehicle is positioned on the Y road_kThe front vehicle of a vehicle interval is marked as a vehicle

Rear vehicle as vehicle

Further defines a vehicle combination K, representing the combination of vehicles on the X road and the Y road directly participating in the K-th stage, and K ∈ { K }₁,K₂,K₃Vehicle combination

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k and vehicle

Vehicle combination

The vehicle participating in the kth stage comprises a vehicle k and a vehicle

At this time, the x-th road is positioned on the Y road_kThe interval of each vehicle has no front vehicle; vehicle combination

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k, now on the Xth road of Y_kNo rear vehicle participates at the interval of each vehicle;

based on the vehicle track predicted by the microcosmic following model, the vehicle is driven

Vehicle k and vehicle

The relationship between the two is that the vehicle k can smoothly merge into or pass through the vehicle

And a vehicle

In the interval between the vehicles, the vehicle k can not smoothly merge into or pass through the vehicle

And a vehicle

The spacing therebetween; the vehicle k can not smoothly merge into or pass through the vehicle

And a vehicle

The interval between them is divided into four cases: the first case is denoted as R1 and represents vehicle k and vehicle

The distance between the two adjacent layers is too close to meet the constraint condition that the two adjacent layers can be smoothly merged or passed; the second case is denoted as R2 and represents vehicle kAnd a vehicle

The distance between the two adjacent layers is too close to meet the constraint condition that the two adjacent layers can be smoothly merged or passed; the third case is denoted as R3 and represents vehicle k and vehicle

And with vehicles

The process of meeting the basic spacing requirements but merging or passing in is uncomfortable; the fourth case is denoted as R4 and represents vehicle k and vehicle

And with vehicles

The constraint conditions for smooth import or passing are not met between the two groups;

based on different vehicle combinations, different vehicle type combinations and whether the vehicle k can smoothly merge into or pass through the vehicle

And a vehicle

The planning of the cooperative control strategy set under different conditions of the interval between the two strategies specifically comprises the following steps:

the collective expression for the vehicle conditions participating in the k-th phase is as follows:

or 1 or a NaN,

α_keither the number of bits is 0 or 1,

or 1 or a NaN,

r

_k0 or 1 or 2 or 3 or 4,

wherein,

α_k、

respectively representing vehicles

Vehicle k and vehicle

A value 0 indicates that the vehicle type is a non-optimally controllable vehicle, a value 1 indicates that the vehicle type is an optimally controllable vehicle, and a symbol NaN indicates that there is no vehicle; r is_kIndicating vehicles

Vehicle k and vehicle

Relation between r _k0 means that the vehicle k can smoothly merge into or pass through the vehicle

And a vehicle

Interval between r_k1 means that the vehicle k cannot smoothly merge into or pass through the vehicle

And a vehicle

Case of spacing between R1, R _k2 means that the vehicle k is noneLaw of smooth import or passage through vehicles

And a vehicle

Case of spacing between R2, R_k3 means that the vehicle k cannot smoothly merge into or pass through the vehicle

And a vehicle

Case of spacing between R3, R _k4 means that the vehicle k cannot smoothly merge into or pass through the vehicle

And a vehicle

The case of the interval therebetween R4;

a specific list of vehicle conditions participating in phase k is as follows:

and (3) aiming at the condition of the vehicle participating in the k stage, planning a corresponding cooperative control strategy:

defining decision variables of the k-th stage lower layer model at t moment

Reflects the acceleration of the vehicle at the moment t in the k-th stage, an

Wherein

A decision variable set of the lower layer model at the kth stage at the time t;

for vehicle k:

when in use

Time t, decision variable u of vehicle k at time t_k(t) satisfies the following conditions:

when in use

when in use

Time t, decision variable u of vehicle k at time t_k(t) satisfies the following conditions: v. of_k(t)+u_k(t)τ≤v^e；

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

Wherein,

representing the vehicle condition participating in the k-th phase; τ is the reaction time of the vehicle driving; v. of_k(t)、

Respectively show a vehicle k and a vehicle

Vehicle with a steering wheel

Velocity at time t;

respectively a vehicle k and a vehicle

Vehicle with a steering wheel

The safe following acceleration of the vehicle at the time t is predicted according to the microcosmic following model;

respectively a vehicle k and a vehicle

Vehicle with a steering wheel

The safe following speed of the vehicle at the moment t + tau is predicted according to the microcosmic following model; v. of^eIs the desired speed; (in the above, L_k(t) represents the relative distance between the vehicle k and its following preceding vehicle at time t,

representing the speed of the car k before the car k follows at the time t;

indicating vehicle at time t

The relative distance between the car and the car before the car is driven,

indicating vehicles

The speed of the car before the car is followed at the time t;

indicating vehicle at time t

Relative to the car before it followsThe distance between the first and second electrodes,

indicating vehicles

The speed of the car ahead at time t. )

If the vehicle in stage k-1

The target vehicle is subjected to cooperative optimization control, and the vehicle is taken as the vehicle in the k stage

The vehicle processes that will be regarded as non-optimizable control when engaged are expressed in terms of expressions as follows:

further, the objective function of the lower model is:

setting an initial cost to

Setting an initial state variable to

Wherein, the target vehicle in the k stage is taken as the target vehiclei；

Indicating the time when the target vehicle i enters the control area in the k-th stage;

represents the time when the target vehicle i leaves the control area in the k-th stage; will be provided with

To

Is divided equally into N segments in discrete time intervals τ', i.e.

Defining a control decision time as

Objective function of the underlying model

Shows that after the target vehicle i is subjected to cooperative optimization control in the k stage, the

Cost consumption from time to time t;

representing a decision index function of the lower layer model at the kth stage at the time t;

reflecting the vehicle position and speed of the target vehicle i in the kth stage at the time t for the state variable of the lower layer model in the kth stage at the time t;

reflecting the decision variables of the lower layer model at the kth stage at the t momenti vehicle acceleration at time t;

a decision variable set of the lower layer model at the kth stage at the time t; the initial state variable is

Indicates that the target vehicle i is in the k-th stage

Vehicle position and speed at the moment.

Further, the decision index function of the k-th stage lower layer model at the time t

The method specifically comprises the following steps:

when k is equal to 1, the first step is carried out,

when k is 2,3, …, n,

wherein,

representing a decision index function of the lower layer model in the 1 st stage at the t moment;

indicating a time at which the target vehicle of the k-th stage enters the control area;

indicating the time when the target vehicle of the k-1 stage leaves the control area;

indicating the time at which the target vehicle leaves the control area at the k-th stage.

Further, the vehicle which does not directly participate in the process of vehicle confluence or vehicle passing through the intersection on the Y road in the step 1-2 is marked as a vehicle i', and the cost consumption caused by the vehicle speed adjustment of the vehicle due to the following safety requirement of the vehicle is caused by the influence of the vehicle confluence or passing of the front vehicle

The method specifically comprises the following steps:

wherein,

indicating that vehicle i' belongs to a vehicle between the rear vehicle at the k-1 th stage and the front vehicle at the k-th stage on the Y road.

Furthermore, a microscopic traffic flow simulation environment is constructed, and simulation results before and after optimization under different traffic situations are compared.

Compared with the prior art, the method has the beneficial effects that:

the invention provides a mixed traffic flow collaborative optimization control method based on double-layer planning, which is characterized in that an upper layer model of a double-layer optimization model is utilized to ensure that a system for hybrid traffic flow collaborative decision control is optimal by searching a vehicle sequence with optimal system under different traffic scenes (including ramp vehicle convergence on an expressway, vehicle convergence at an intersection and vehicle passing through the intersection), namely searching an optimal sequence with m +1 intervals existing among vehicles converged on an X road or passing through a Y road; the lower layer model of the double-layer optimization model is classified according to cooperative control strategies under different scenes, and based on a corresponding microcosmic car-following model and a condition constraint model for judging whether vehicles on an X road can smoothly merge into or pass through a Y road, vehicle tracks in the process of single vehicle convergence or vehicle confluence or vehicle passing through an intersection are subjected to optimization control, so that the vehicles on the X road can smoothly merge into or pass through the Y road, and the vehicle tracks are optimal. And the vehicle track optimization result obtained by the objective function in the lower model is used as one input in the dynamic programming recursive solving process of the upper model. The upper layer model and the lower layer model jointly ensure the optimal running of the system vehicles, so that the vehicle conflict of the vehicles in the process of converging or passing through the intersection is reduced under the mixed traffic flow environment, and the passing efficiency and the comfort of the vehicles are effectively improved.

The method can be generally applied to various microscopic traffic scenes, including highway ramp vehicle convergence, intersection vehicle convergence, vehicle passing through an intersection and the like, and can be applied only by pertinently changing the upper layer model and the lower layer model, so that the method has certain universal applicability.

A large number of simulation operation results show that the sequence of vehicle convergence or vehicle interval passing can be changed by using the double-layer optimization model, most vehicles can run at a higher speed under the action of a cooperative optimization control strategy, and an acceleration curve is smoother. The throughput of the confluence region segment increases by approximately 18% when the permeability of the autonomous vehicle (i.e. the vehicle that can be optimally controlled) is high. Under the cooperative optimization control mechanism of the mixed traffic flow, the traffic capacity of the road section can be further improved by 10-15%.

The present invention will be described in further detail with reference to the following detailed description and the accompanying drawings, which are not intended to limit the scope of the invention.

Drawings

FIG. 1 is a diagram of the basic architecture of a two-layer optimization model according to an embodiment of the present invention.

Fig. 2 is a schematic view of vehicle convergence of a highway ramp in a mixed traffic flow scene according to an embodiment of the invention.

Fig. 3 is a basic diagram of the permeability of the autonomous vehicle and the road section passing capacity under the GIPPS following model according to the embodiment of the invention.

Fig. 4 is a microscopic trace diagram of 23 vehicles on the ramp passing through the confluent region section when the permeability of the autonomous vehicle is 90% and is not controlled by the double-layer optimization model according to the embodiment of the invention.

Fig. 5 is a micro-motion trace diagram of 23 vehicles on a ramp passing through a confluence region segment when the penetration rate of the autonomous vehicle is 90% and controlled by a double-layer optimization model according to an embodiment of the invention.

Detailed Description

The following description of the embodiments of the present invention will be made with reference to the accompanying drawings.

A mixed traffic flow collaborative optimization control method based on double-layer planning is characterized in that a double-layer optimization model based on dynamic planning recursion is adopted for carrying out mixed traffic flow collaborative decision control, and the method is suitable for different traffic scenes in the mixed traffic flow, wherein the different traffic scenes comprise ramp vehicle convergence on a highway, vehicle confluence at an intersection and vehicle passing through the intersection; the double-layer optimization model comprises an upper layer model and a lower layer model; the upper model is a vehicle sequencing problem solved by dynamic programming recursion; the lower model is a single vehicle trajectory optimization problem in different traffic scenarios solved recursively with dynamic programming. Fig. 1 is a diagram showing a basic architecture of a two-layer optimization model according to an embodiment of the present invention.

Example one

The present embodiment describes the method of the present invention by taking highway ramp vehicle convergence as an example.

As shown in fig. 2, the schematic view of vehicle convergence on a ramp of an expressway under a mixed traffic flow scene is shown, where a main road (i.e., a Y road) and a ramp (i.e., an X road) are both one-way roads, a plurality of traffic flows randomly arranged between an automatically-driven vehicle a (i.e., an optimally-controlled vehicle) and a human-driven vehicle H (i.e., a non-optimally-controlled vehicle) are provided on the main road, and a plurality of traffic flows randomly arranged between the automatically-driven vehicle a and the human-driven vehicle H are provided on the ramp and need to converge between the traffic flows on the main road through a convergence region. Assume that the autonomous vehicle a and the human-driven vehicle H on the main road and the ramp both follow the microscopic follow-up model.

Description of the upper layer model: the process of converging vehicles on a ramp of an expressway, namely vehicles on the ramp are sequentially converged into a main road, starting time of an initial stage from the moment when a first ramp vehicle is ready to be converged, and then convergence of a next ramp vehicle is the next stage. The first ramp vehicle faces the selection of a plurality of spaces into which vehicles can be converged on the main road, and the number of the spaces into which vehicles can be converged on the main road is correspondingly changed along with the successful convergence of the vehicles on each ramp. The upper dynamic planning model is mainly used for finding the optimal sequence of the interval of vehicles on the ramp and merging into the main road.

The lower model describes: in the junction area of the ramps of the expressway, when vehicles normally run, the traffic flow on the same lane follows the traffic flow according to the microcosmic car following model, namely, the vehicles on the main road

And the vehicles k on the ramp follow the microscopic follow-up model to run. Suppose that when vehicle k enters the merge area on a ramp, vehicle k may observe a conflicting vehicle on the arterial road

And

at this time, according to the vehicle k and the vehicle

Whether the relative distance between the two meets the condition of conflux or not; if yes, the vehicle k successfully enters the main road; if not, judging the reason why the vehicle k cannot be converged successfully, optimizing the vehicle running track by applying a corresponding cooperative control strategy and combining with dynamic planning, and calculating corresponding consumption.

The method comprises the following steps:

s1, establishing the upper layer model, including:

the upper layer model is divided: assuming that the main road and the ramp are both one-way roads, n vehicles on the ramp need to sequentially merge into m +1 intervals existing among the vehicles on the main road, the behavior that each single vehicle on the ramp merges into the main road is represented as a stage, and the behavior that the kth vehicle on the ramp merges into the main road is recorded as the kth stage, wherein k is 1,2,3, …, n.

State variables of the upper model: the number of vehicle intervals for leading the kth vehicle on the ramp to merge into the main road in the kth stage is s_kAnd (4) showing.

Decision variables of the upper model: the decision made at each stage represents that the k-th vehicle on the ramp at the k-th stage can merge into the main road s_kSelecting a particular xth of individual vehicle intervals_kThe vehicles are alternately merged.

the state transition equation of the upper model is as follows:

setting the initial condition as s₀＝m+1；

The state transition equation of the upper model indicates that s is equal to 1 when k_k＝s₁Is shown inIn the stage 1, the number of vehicle intervals for the 1 st vehicle on the ramp to flow into the main road is m + 1; when k is 2,3, …, n, the x-th vehicle on the ramp is selected at the k-1 stage_k-1The interval of each vehicle is used as a state variable s after being converged into the main road_kA change in (c); s₀M +1 represents that the number of vehicle intervals for vehicles on the ramp to merge into the main road in the initial state is m + 1.

Cost function of the upper model:

S for showing that the k-th vehicle on the ramp is merged into the main road for the k-th vehicle to merge into the main road under the action of the cooperative optimization control strategy in the k-th stage_kAll possible cost costs arising from individual vehicle intervals;

the method is characterized in that the main road does not directly participate in the traffic flow in the vehicle converging process, and the cost is consumed due to the fact that the speed of the vehicle is adjusted according to the following safety requirement of the vehicle because the front vehicle is influenced by the vehicle converging;

as described above

The mathematical expression of (a) is as follows:

wherein,

it indicates that the vehicle i' belongs to a vehicle between the rear vehicle of the k-1 stage and the front vehicle of the k stage on the main road.

An objective function of the upper model:

setting the initial condition as f₀(s₀)＝0；

Objective function f of the upper model_k(s_k) Representing the cumulative cost consumption of the system vehicles from stage 1 to stage k, f₀(s₀) 0 indicates that the system cost is 0 in the initial state.

S2, establishing the lower layer model, including:

s2-1, determining a microscopic follow-up model, describing a follow-up state of the vehicle by using the microscopic follow-up model, and predicting an initial track of the vehicle; the following state of the vehicle includes a speed, an acceleration, and a position of the vehicle.

The microcosmic car-following model can be selected from specific microcosmic car-following models, such as a GIPPS car-following model, an IDM/EIDM car-following model and the like, and the specific model can be selected to be judged according to the simulation condition.

The GIPPS following model was selected for this example as follows:

v(t+τ)＝min(v^e,v(t)+aτ,v_safe)，

l(t+τ)＝l(t)-v(t)τ-0.5u(t)τ²，

the velocity of the vehicle at time t + τ is denoted by v (t + τ), where v^eV (t) + a τ represents the velocity obtained by acceleration at a constant acceleration a, v (t) + a τ being the desired velocity_safeIndicating a safe speed. τ is the reaction time of the vehicle driving; b is a constant deceleration; v (t) represents the speed of the vehicle at time t; u (t) represents the acceleration of the vehicle at time t; l (t) watchShowing the position of the vehicle at time t; v. of_lead(t) represents the speed of the preceding vehicle followed by the vehicle at time t; l_lead(t) represents a position of a preceding vehicle followed by the vehicle at time t; l_aIndicating the vehicle body length; l is₀Indicating a minimum safe following distance between the vehicle and a preceding vehicle to which the vehicle is following.

S2-2, establishing a condition constraint model for judging whether the kth vehicle on the kth stage ramp can smoothly merge into the main road.

Suppose that the k-th vehicle on the ramp (denoted as vehicle k) is in the merge region (i.e., z)₁<l_k(t)≤z₀) Which will merge into two vehicles of the continuous flow of the main road

And

wherein the vehicle

Is a front vehicle or a vehicle

For the rear vehicle, the following conditional constraint model (i.e., the convergence model) is used to judge whether the vehicle k can smoothly converge into the main road:

u above_k(t) represents the effect of the confluence, reflects the comfort level during the confluence, and is the distance between vehicles during the confluence, the confluent vehicle k on the ramp during the confluence and the rear vehicle of the main road

Is calibrated. Wherein,

what is shown is the utility of the bus action when unconstrained. l_aThe length of the vehicle body is long,

for a minimum safe distance of autonomous vehicle convergence,

minimum safe distance for human-driven vehicles to converge. During confluence, a confluent vehicle k on the ramp actually follows the front vehicle on the main road

Rear cars running on the main road

The actual following bus vehicles k run, and their accelerations can be calculated from the vehicle-following model.

An absolute value representing the acceleration of the vehicle k;

rear vehicle on main road

Absolute value of acceleration of (a); b_safeIndicating the maximum allowed deceleration. Phi_AFor autonomous vehicle set, phi_HSet of vehicles for human driving η₁And η₂Respectively representing a safety factor and a polite factor, the safety factor η₁Is a constant, polite coefficient η₂In a piecewise continuous form, v_thIs a given speed threshold, v^eTo desired speed, β₁And β₂Is a constant. By means of I_k(t + τ) represents a convergence decision, a value of 0 indicates that the vehicle k cannot smoothly complete convergence at time t + τ, and a value of 1 indicates that the vehicle k can smoothly complete convergence at time t + τ. When making convergence decision I_kThe value of (t + tau) is 0, namely when the vehicle k cannot smoothly complete confluence, the vehicle k on the ramp can use the end of the ramp as a stopped virtual front vehicle, continuously decelerates or even stops to wait by following a microscopic following model until a vehicle interval meeting confluence appears on the main road, and the vehicle k is converged into the main road.

S2-3, aiming at various possible situations that occur in the process that the kth vehicle on the kth stage ramp is converged into the main road under the mixed traffic scene, simulating a cooperative control strategy set. The method specifically comprises the following steps:

according to the k vehicle on the k stage ramp, the k vehicle can be converged into the main road_kSelecting a particular xth of individual vehicle intervals_kThe vehicles are converged at intervals, the kth vehicle on the ramp is marked as a vehicle k, and the xth vehicle is positioned on the main road_kThe front vehicle of a vehicle interval is marked as a vehicle

Rear vehicle as vehicle

Further defining a vehicle combination K representing a combination of vehicles on the ramp and the main road directly participating in the K-th stage, and K ∈ { K }₁,K₂,k₃Vehicle combination

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k and vehicle

Vehicle combination

The vehicle participating in the kth stage comprises a vehicle k and a vehicle

Now located on the x-th main road_kThe interval of each vehicle has no front vehicle; vehicle combination

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k, now on the x-th road_kNo rear vehicle participates at the interval of each vehicle;

Vehicle k and vehicle

The relationship between the two is that the vehicle k can smoothly merge into the vehicle

And a vehicle

The vehicle k can not smoothly merge into the vehicle

And a vehicle

The spacing therebetween; the vehicle k can not smoothly merge into the vehicle

And a vehicle

BetweenThe interval of (c) is divided into four cases: the first case is denoted as R1 and represents vehicle k and vehicle

The distance between the two adjacent beams is too close to meet the constraint condition of smooth import; the second case is denoted as R2 and represents vehicle k and vehicle

The distance between the two adjacent beams is too close to meet the constraint condition of smooth import; the third case is denoted as R3 and represents vehicle k and vehicle

And with vehicles

The process of meeting the basic spacing requirement but merging is not comfortable; the fourth case is denoted as R4 and represents vehicle k and vehicle

And with vehicles

The constraint conditions which can be smoothly merged are not met;

based on different vehicle combinations, different vehicle type combinations and whether the vehicle k can smoothly merge into the vehicle

And a vehicle

or 1 or a NaN,

α_keither the number of bits is 0 or 1,

or 1 or a NaN,

r

_k0 or 1 or 2 or 3 or 4,

wherein,

α_k、

respectively representing vehicles

Vehicle k and vehicle

A value 0 indicates that the vehicle type is a human-driven vehicle, a value 1 indicates that the vehicle type is an autonomous vehicle, and a symbol NaN indicates that there is no vehicle; r is_kIndicating vehicles

Vehicle k and vehicle

Relation between r _k0 means that the vehicle k can smoothly merge into the vehicle

And a vehicle

Interval between r_k1 means that the vehicle k cannot smoothly merge into the vehicle

And a vehicle

Case of spacing between R1, R _k2 means that the vehicle k cannot smoothly merge into the vehicle

And a vehicle

Case of spacing between R2, R_k3 means that the vehicle k cannot smoothly merge into the vehicle

And a vehicle

Case of spacing between R3, R _k4 means that the vehicle k cannot smoothly merge into the vehicle

And a vehicle

The case of the interval therebetween R4;

a specific list of vehicle conditions participating in phase k is as follows:

defining decision variables of the k-th stage lower layer model at t moment

Reflects the acceleration of the vehicle at the moment t in the k-th stage, an

Wherein

for vehicle k:

when in use

when in use

when in use

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

Wherein,

Respectively show a vehicle k and a vehicle

Vehicle with a steering wheel

Velocity at time t;

are respectivelyVehicle k and vehicle

Vehicle with a steering wheel

respectively a vehicle k and a vehicle

Vehicle with a steering wheel

The safe following speed of the vehicle at the moment t + tau is predicted according to the microcosmic following model; v. of^eIs the desired speed.

If the vehicle in stage k-1

The process of taking part in the process of being treated as human driving a vehicle is expressed by the following expression:

s2-4, based on the vehicle initial track predicted in the step S2-1, sequentially judging that the k-th vehicle on the ramp of the k-th stage merges into the main road through the condition constraint model established in the step S2-2S on which the kth vehicle can merge into the trunk road_kThe specific situation of each vehicle interval is determined according to the fact that the k-th vehicle on the kth stage ramp can be converged into the main road on the main road for the k-th vehicle to be converged into the main road_kThe specific situation occurring in the process of each vehicle interval respectively makes the cooperative control strategy corresponding to the cooperative control strategy set planned in the step S2-3.

S2-5, determining the vehicles which can be optimally controlled in the vehicles participating in the k stage as target vehicles; and optimizing the running track of the target vehicle according to the cooperative control strategies made in the step S2-4 respectively, resolving the optimization problem into an optimal control problem of discrete time state constraint, and solving by using a dynamic planning idea to obtain the cooperative optimization control strategy about the target vehicle.

S2-6, calculating S for the k-th vehicle on the ramp to merge into the main road under the action of the cooperative optimization control strategy in the k-th stage_kAll possible cost consumptions of one vehicle interval

The objective function of the lower model is:

setting an initial cost to

Setting an initial state variable to

Wherein, the target vehicle in the k stage is recorded as a target vehicle i;

To

Is divided equally into N segments in discrete time intervals τ', i.e.

Defining a control decision time as

Objective function of the underlying model

Cost consumption from time to time t;

reflecting the vehicle acceleration of the target vehicle i at the moment t in the kth stage for the decision variable of the lower layer model at the moment t in the kth stage;

Indicates that the target vehicle i is in the k-th stage

Vehicle position and speed at the moment.

Decision index function of the k-th stage lower layer model at t moment

The method specifically comprises the following steps:

when k is equal to 1, the first step is carried out,

when k is 2,3, …, n,

wherein,

S3, solving the double-layer optimization model:

s3-2, and reversely deducing the decision made at each stage of the upper layer model determined in the step S3-1 to obtain the vehicle optimized track of each single vehicle on the lower layer model ramp merging into the main road.

The method establishes a microscopic traffic flow simulation environment through programming, and the parameters and values of the simulation experiment environment are realized by computer programming as shown in the following table:

fig. 3 is a basic graph of autonomous vehicle permeability versus road segment traffic capacity for the present embodiment under the GIPPS-followed model, showing the flow-density relationship for different autonomous vehicle permeabilities observed just downstream of the confluence region segment. Wherein graph a represents a comparison of the flow-density relationship for a 100% autonomous vehicle and a 100% human-driven vehicle; the graph B, C, D reflects the flow-density relationship for autonomous vehicle permeabilities of 30%, 50%, and 70%, respectively. From fig. 3 it can be found that: when the permeability of the automatic driving vehicle is 0, namely all vehicles are human driving vehicles, the road section traffic capacity is about 2244 veh/h; when the permeability of the automatic driving vehicle is 50%, namely the ratio of the human driving vehicle to the automatic driving vehicle is 1:1, compared with the permeability of the automatic driving vehicle is 0, the road section traffic capacity is improved by about 11.8%, and reaches about 2508 veh/h; when the permeability of the automatic driving vehicle is 100 percent, namely all the automatic driving vehicles are automatic driving vehicles, compared with the permeability of the automatic driving vehicles of 0, the road section traffic capacity is improved by about 17.7 percent and reaches 2640 veh/h. Experiments using other microscopic follow models also showed similar results. It follows from this that: if the permeability of the autonomous vehicle increases, the traffic capacity of the road segment is improved.

Fig. 4 is a microscopic movement trace diagram of 23 vehicles on the ramp passing through the confluence region segment when the penetration rate of the autonomous vehicle is 90% and is not controlled by the double-layer optimization model in the embodiment. Wherein, the graph (a) is a velocity-time relation graph, and the graph (b) is an acceleration-time relation graph.

Fig. 5 is a microscopic movement trace diagram of 23 vehicles on the ramp passing through the confluence region segment when the penetration rate of the automatic driving vehicle is 90% and controlled by the double-layer optimization model. Wherein, the graph (a) is a velocity-time relation graph, and the graph (b) is an acceleration-time relation graph.

By comparing fig. 4 and fig. 5, it can be found that: by applying the double-layer optimization model, the vehicle convergence sequence can be changed, so that the global optimum is ensured. Under the action of a cooperative optimization control strategy, most vehicles can run at a higher speed, and an acceleration curve is smoother. When the permeability of the automatic driving vehicle is higher, the traffic capacity of the road section can be further improved by 10-15% under the cooperative optimization control mechanism of the mixed traffic flow.

Example two

This embodiment will explain the method of the present invention by taking intersection vehicle merging as an example.

Merging vehicles at the intersection: the X road and the Y road are two crossed one-way roads, a plurality of automatic driving vehicles (vehicles which can be optimally controlled) and human driving vehicles (vehicles which cannot be optimally controlled) are arranged on the Y road to form irregularly arranged traffic flows, and a plurality of automatic driving vehicles and human driving vehicles which form irregularly arranged traffic flows need to turn at the intersection and converge into intervals among the traffic flows on the Y road. It is assumed that the autonomous vehicle and the human-driven vehicle on both the X-road and the Y-road follow the microscopic follow-up model.

Description of the upper layer model: the intersection vehicle confluence, namely the process that vehicles to be confluent turn and converge into straight traffic flow at the intersection in sequence, is started when the first vehicle to be confluent is ready to converge into straight traffic flow, and then each vehicle to be confluent turns and converges into straight traffic flow in one stage. The first vehicle to be merged is selected to face a plurality of intervals of vehicles capable of merging, and the number of the intervals of the vehicles capable of merging is correspondingly changed along with successful merging of the vehicles to be merged. The upper dynamic planning model is mainly used for searching an optimal sequence of vehicle intervals for converging vehicles to be converged into a straight traffic flow.

The lower model describes: under the mixed traffic environment of urban intersections, when vehicles normally run, the traffic flow on the same lane follows the vehicle according to the microcosmic car following model, namely, the vehicles on the Y road

And the vehicle k on the X road runs according to the microcosmic following model. Suppose that when vehicle k on the X road enters the collision area, vehicle k can observe a colliding vehicle on the Y road

And

at this time, according to the vehicle k and the vehicle

Whether the relative distance between the two meets the condition of confluent flow or not; if the vehicle k meets the requirement, the vehicle k successfully enters the straight-going traffic flow; if not, judging the reason why the vehicle k cannot successfully converge into the straight traffic flow, optimizing the running track of the vehicle by applying a corresponding cooperative control strategy and combining with dynamic planning, and calculating corresponding consumption.

The method comprises the following steps:

s1, establishing the upper layer model, including:

the upper layer model is divided: assuming that an X road and a Y road are two one-way roads with intersections, n vehicles on the X road need to sequentially converge into m +1 intervals existing among the vehicles on the Y road, the behavior that each one-way vehicle on the X road converges into the Y road is represented as a stage, and the behavior that the k-th vehicle on the X road converges into the Y road is represented as a k-th stage, wherein k is 1,2,3, …, n.

State variables of the upper model: the number of the vehicle intervals for the k-th vehicle on the X road to merge into the Y road in the k stage is s_kAnd (4) showing.

Decision variables of the upper model: the decision made in each stage represents that the k vehicle on the X road in the k stage can be converged into the s road of the Y road_kSelecting a particular xth of individual vehicle intervals_kThe vehicles are alternately merged.

the state transition equation of the upper model is as follows:

setting the initial condition as s₀＝m+1；

The state transition equation of the upper model indicates that s is equal to 1 when k_k＝s₁The number of vehicle intervals for enabling the 1 st vehicle on the X road to merge into the Y road in the 1 st stage is m + 1; when k is 2,3, …, n, the (k-1) th vehicle on the X road selects the (X) th vehicle at the (k-1) th stage_k-1The interval of each vehicle is used as a state variable s after converging into the Y road_kA change in (c); s₀M +1 indicates that the number of vehicle intervals at which vehicles on the X road can merge into the Y road in the initial state is m + 1.

Cost function of the upper model:

S for showing that the k-th vehicle on the X road is converged into the road available for converging the k-th vehicle into the Y road under the action of the cooperative optimization control strategy in the k stage_kAll possible cost costs arising from individual vehicle intervals;

the method is characterized in that the vehicle flow which directly participates in the vehicle confluence process on the Y road is not generated, and the cost is consumed due to the fact that the speed of the vehicle is adjusted by the vehicle due to the fact that the vehicle ahead is influenced by vehicle confluence;

as described above

The mathematical expression of (a) is as follows:

wherein,

indicating vehicle i^′And the vehicles belong to the vehicles between the rear vehicle at the k-1 stage and the front vehicle at the k stage on the Y road.

An objective function of the upper model:

setting the initial condition as f₀(s₀)＝0；

S2, establishing the lower layer model, including:

And S2-2, establishing a condition constraint model for judging whether the k vehicle on the X road at the k stage can smoothly merge into the Y road.

Based on a microcosmic car-following model, acceleration constraint, distance constraint and safety constraint are added to establish a condition constraint model which accords with the vehicle confluence characteristic of the intersection by combining the road geometric characteristics of the intersection.

S2-3, aiming at various possible situations that occur in the process that the k vehicle on the X road at the k stage under the mixed traffic scene converges into the Y road, simulating a cooperative control strategy set.

S2-4, based on the vehicle initial track predicted in the step S2-1, sequentially judging that the k-th vehicle on the X road at the k stage converges to the Y road for converging the k-th vehicle on the Y road by the condition constraint model established in the step S2-2S into Y road_kThe specific situation of each vehicle interval is determined according to the situation that the k-th vehicle on the X-th road in the k stage can be supplied to the Y road on the Y road for the k-th vehicle to enter the Y road_kThe specific situation occurring in the process of each vehicle interval respectively makes the cooperative control strategy corresponding to the cooperative control strategy set planned in the step S2-3.

S2-6, calculating S for the k-th vehicle on the X road to merge into the Y road under the action of the cooperative optimization control strategy in the k stage_kAll possible cost consumptions of one vehicle interval

The objective function of the lower model is:

setting an initial cost to

Setting an initial state variable to

Wherein, the target vehicle in the k stage is recorded as a target vehicle i;

To

Is divided equally into N segments in discrete time intervals τ', i.e.

Defining a control decision time as

Objective function of the underlying model

Cost consumption from time to time t;

Indicates that the target vehicle i is in the k-th stage

Vehicle position and speed at the moment.

Decision index function of the k-th stage lower layer model at t moment

The method specifically comprises the following steps:

when k is equal to 1, the first step is carried out,

when k is 2,3, …, n,

wherein,

is shown asA decision index function of the lower layer model at the moment t in the stage k;

S3, solving the double-layer optimization model:

s3-2, and reversely deducing the decision made at each stage of the upper layer model determined in the step S3-1 to obtain the vehicle optimized track of each single vehicle on the X road of the lower layer model converging into the Y road.

EXAMPLE III

The present embodiment describes the method of the present invention by taking the vehicle passing through an intersection as an example.

Vehicle passing through the intersection: the X road and the Y road are two crossed one-way roads, a plurality of automatic driving vehicles (vehicles capable of being controlled in an optimized mode) and human driving vehicles (vehicles incapable of being controlled in an optimized mode) are arranged on the Y road to form irregularly arranged traffic flows, and the plurality of automatic driving vehicles and the human driving vehicles which form the irregularly arranged traffic flows need to pass through the intersection and sequentially pass through intervals among the traffic flows on the Y road. It is assumed that the autonomous vehicle and the human-driven vehicle on both the X-road and the Y-road follow the microscopic follow-up model.

Description of the upper layer model: the process that vehicles passing through the intersection, namely vehicles to pass through the intersection on an X road pass through vehicle intervals on a Y road in sequence at the intersection is started when a first vehicle to pass through is ready to pass through, and then each vehicle to pass through the intersection is a stage, wherein the first vehicle to pass through faces the selection of a plurality of passing vehicle intervals on the Y road, and the number of the passing vehicle intervals on the Y road is correspondingly changed along with the successful passing of each vehicle to pass through the intersection. The upper dynamic planning model is mainly used for finding the optimal sequence of the vehicle intervals of the vehicles to pass through the intersection on the Y road.

And

at this time, according to the vehicle k and the vehicle

Whether the relative distance between the two meets the condition of passing through the intersection; if the vehicle k meets the requirement, the vehicle k successfully passes through the intersection; if not, judging the reason why the vehicle k cannot successfully pass through the intersection, optimizing the running track of the vehicle by applying a corresponding cooperative control strategy and combining with dynamic planning, and calculating corresponding consumption.

The method comprises the following steps:

s1, establishing the upper layer model, including:

the upper layer model is divided: assuming that an X road and a Y road are two one-way roads with intersections, n vehicles on the X road need to sequentially pass through m +1 intervals existing between the vehicles on the Y road, and a behavior of each single vehicle on the X road passing through the Y road is represented as a stage, and a behavior of a k-th vehicle on the X road passing through the Y road is represented as a k-th stage, where k is 1,2,3, …, n.

State variables of the upper model: the number of the vehicle intervals for the k-th vehicle on the X road to pass through the Y road in the k stage is s_kAnd (4) showing.

Decision variables of the upper model: the decision made in each stage indicates that the k-th vehicle on the X-th road in the k-th stage can pass through the s-th road of the Y-th road_kSelecting a particular xth of individual vehicle intervals_kThe vehicles pass through at intervals.

the state transition equation of the upper model is as follows:

setting the initial condition as s₀＝m+1；

The state transition equation of the upper model indicates that s is equal to 1 when k_k＝s₁The number of vehicle intervals for allowing the 1 st vehicle on the X road to pass through the Y road in the 1 st stage is m + 1; when k is 2,3, …, n, the (k-1) th vehicle on the X road selects the (X) th vehicle at the (k-1) th stage_k-1Individual vehicle interval as state variable s after Y road passing_kA change in (c); s₀M +1 indicates that the number of vehicle intervals at which vehicles on the X road can pass through the Y road in the initial state is m + 1.

Cost function of the upper model:

S representing that the k-th vehicle on the X road passes through the Y road for the k-th vehicle to pass through under the action of the cooperative optimization control strategy in the k stage_kAll possible cost costs arising from individual vehicle intervals;

the method is characterized in that the Y road does not directly participate in the traffic flow of the vehicles passing through the intersection, and the cost is consumed due to the fact that the speed of the vehicles is adjusted according to the following safety requirement of the vehicles of the Y road due to the fact that the vehicles in front of the Y road are influenced by the passing of the vehicles;

as described above

The mathematical expression of (a) is as follows:

wherein,

An objective function of the upper model:

setting the initial condition as f₀(s₀)＝0；

Objective function f of the upper model_k(s_k) To representCumulative cost consumption of system vehicles from stage 1 to stage k, f₀(s₀) 0 indicates that the system cost is 0 in the initial state.

S2, establishing the lower layer model, including:

And S2-2, establishing a condition constraint model for judging whether the k vehicle on the k-th road in the k stage X can smoothly pass through the Y road.

Based on the microcosmic car-following model, the geometric characteristics of the road of the intersection are combined, and the acceleration constraint, the distance constraint and the safety constraint are added to establish a condition constraint model which accords with the characteristics of the vehicle passing the intersection.

S2-3, aiming at various possible situations that occur in the process that the k vehicle on the X road at the k stage passes through the Y road under the mixed traffic scene, a cooperative control strategy set is prepared.

S2-4, based on the vehicle initial track predicted in the step S2-1, the condition constraint model established in the step S2-2 sequentially judges that the k-th vehicle on the X-th road in the k stage passes through the S for the k-th vehicle to pass through the Y road on the Y road_kThe specific situation of each vehicle interval is determined according to the condition that the k-th vehicle on the X-th road in the k stage can pass through the s of the Y road for the k-th vehicle to pass through the Y road_kThe specific situation occurring in the process of each vehicle interval respectively makes the cooperative control strategy corresponding to the cooperative control strategy set planned in the step S2-3.

S2-6, calculating S for the k-th vehicle on the X road to pass through the road through which the k-th vehicle can pass through the Y road under the action of the cooperative optimization control strategy in the k-th stage_kAll possible cost consumptions of one vehicle interval

The objective function of the lower model is:

setting an initial cost to

Setting an initial state variable to

Wherein, the target vehicle in the k stage is recorded as a target vehicle i;

To

Is divided into N segments equally according to discrete time interval tauI.e. by

Defining a control decision time as

Objective function of the underlying model

Cost consumption from time to time t;

Indicates that the target vehicle i is in the k-th stage

Vehicle position and speed at the moment.

Decision index function of the k-th stage lower layer model at t moment

The method specifically comprises the following steps:

when k is equal to 1, the first step is carried out,

when k is 2,3, …, n,

wherein,

indicating departure of the target vehicle from the control zone in phase kThe time of day.

S3, solving the double-layer optimization model:

s3-2, and the vehicle optimized track of each single vehicle on the X road of the lower model through the Y road is obtained by reversely deducing the decision made at each stage of the upper model determined in the step S3-1.

While the present invention has been described above by way of example with reference to the accompanying drawings, it is to be understood that the invention is not limited to the specific embodiments shown herein.

Claims

1. A mixed traffic flow collaborative optimization control method based on double-layer planning is characterized in that a double-layer optimization model based on dynamic programming recursion is adopted to carry out mixed traffic flow collaborative decision control, and the method is suitable for different traffic scenes in the mixed traffic flow, wherein the different traffic scenes comprise ramp vehicle convergence on a highway, vehicle confluence at an intersection and vehicle passing through the intersection; the double-layer optimization model comprises an upper layer model and a lower layer model; the upper model is a vehicle sequencing problem solved by dynamic programming recursion; the lower layer model is a single vehicle track optimization problem in different traffic scenes solved by dynamic programming recursion;

the method comprises the following steps:

s1, establishing the upper layer model, including:

the upper layer model is divided: assuming that an X road and a Y road are two one-way roads with intersections, n vehicles on the X road need to sequentially converge into or pass through m +1 intervals among the vehicles on the Y road, the behavior that each single vehicle on the X road converges into or passes through the Y road is represented as a stage, and the behavior that a k-th vehicle on the X road converges into or passes through the Y road is recorded as a k-th stage, wherein k is 1,2,3,..., n;

the state transition equation of the upper model is as follows:

setting the initial condition as s₀＝m+1；

The state transition equation of the upper model indicates that s is equal to 1 when k_k＝s₁The number of the vehicle intervals for the 1 st vehicle on the X road to enter or pass through the Y road in the 1 st stage is m + 1; when k is 2, 3.., n, the (k-1) th vehicle on the X road selects the (X) th vehicle at the (k-1) th stage_k-1The interval of each vehicle is used as a state variable s after the vehicle is converged or passes through a Y road_kA change in (c); s₀M +1 represents that the number of vehicle intervals for vehicles to merge into or pass through the road Y on the road X in the initial state is m + 1;

cost function of the upper model:

cost function D of the upper model_k(s_k，x_k) Indicating a phase index function required for making a decision at the kth phase, wherein

the vehicle speed regulation system is characterized in that the vehicle does not directly participate in vehicle convergence or vehicle confluence or vehicle crossing intersection process on a Y road, and the cost is consumed due to the fact that the vehicle speed regulation of the vehicle is required for following safety because a front vehicle is influenced by the vehicle convergence or passing;

an objective function of the upper model:

setting the initial condition as f₀(s₀)＝0；

s2, establishing the lower layer model, including:

s2-4, based on the vehicle initial track predicted at step S2-1The condition constraint model established in the step S2-2 sequentially judges S that the k-th vehicle on the X road of the k stage is converged into or passes through the Y road and can be converged into or pass through the Y road_kThe specific situation occurring in the process of each vehicle interval is determined according to the situation that the k-th vehicle on the X-th road in the k stage can be merged into or passes through the s of the Y road_kSpecific conditions occurring in the process of each vehicle interval respectively make the cooperative control strategy corresponding to the cooperative control strategy set planned in the step S2-3;

S3, solving the double-layer optimization model:

2. The mixed traffic flow cooperative optimization control method based on double-layer planning according to claim 1, wherein the step S2-3 is specifically:

Rear vehicle as vehicle

Further defines a vehicle combination K, representing the combination of vehicles on the X road and the Y road directly participating in the K-th stage, and K ∈ { K }₁，K₂，K₃Vehicle combination

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k and vehicle

Vehicle combination

The vehicle participating in the kth stage comprises a vehicle k and a vehicle

Indicating that the vehicle participating in the k-th stage has a vehicle

Vehicle k and vehicle

And a vehicle

And a vehicle

And a vehicle

The distance between the two adjacent layers is too close to meet the constraint condition that the two adjacent layers can be smoothly merged or passed; the second case is denoted as R2 and represents vehicle k and vehicle

And with vehicles

And with vehicles

And a vehicle

or 1 or a NaN,

α_keither the number of bits is 0 or 1,

or 1 or a NaN,

r_k0 or 1 or 2 or 3 or 4,

wherein,

α_k、

respectively representing vehicles

Vehicle k and vehicle

Vehicle k and vehicle

Relation between r_k0 means that the vehicle k can smoothly merge into or pass through the vehicle

And a vehicle

And a vehicle

Case of spacing between R1, R_k2 means that the vehicle k cannot smoothly merge into or pass through the vehicle

And a vehicle

And a vehicle

Case of spacing between R3, R_k4 means that the vehicle k cannot smoothly merge into or pass through the vehicle

And a vehicle

The case of the interval therebetween R4;

a specific list of vehicle conditions participating in phase k is as follows:

defining decision variables of the k-th stage lower layer model at t moment

Reflects the acceleration of the vehicle at the moment t in the k-th stage, an

Wherein

for vehicle k:

when in use

when in use

when in use

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

For vehicles

When in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

when in use

In time, the vehicle

Decision variables at time t

The requirements are as follows:

and is

Wherein,

Respectively show a vehicle k and a vehicle

Vehicle with a steering wheel

Velocity at time t;

respectively a vehicle k and a vehicle

Vehicle with a steering wheel

respectively a vehicle k and a vehicle

Vehicle with a steering wheel

The safe following speed of the vehicle at the moment t + tau is predicted according to the microcosmic following model; v. of^eIs the desired speed;

if the vehicle in stage k-1

3. the double-layer planning-based mixed traffic flow cooperative optimization control method according to claim 1, wherein the objective function of the lower layer model is:

setting an initial cost to

Setting an initial state variable to

Wherein, the target vehicle in the k stage is recorded as a target vehicle i;

To

Is divided equally into N segments in discrete time intervals τ', i.e.

Defining a control decision time as

Objective function of the underlying model

Cost consumption from time to time t;

Indicates that the target vehicle i is in the k-th stage

Vehicle position and speed at the moment.

4. The double-layer planning-based mixed traffic flow cooperative optimization control method according to claim 3, wherein the k-th stage lower layer model is a decision index function at the time t

The method specifically comprises the following steps:

when k is equal to 1, the first step is carried out,

when k is 2,3, …, n,

wherein,

5. The double-deck planning-based mixed traffic flow cooperative optimization control method according to claim 1, wherein the vehicle on the Y road in the step S1-2, which does not directly participate in the vehicle confluence or vehicle crossing process, is marked as vehicle i', and the cost is consumed due to the vehicle speed adjustment of the own vehicle for the follow-up safety requirement because the front vehicle is influenced by the vehicle confluence or passing

The method specifically comprises the following steps:

wherein,

6. The double-layer planning-based mixed traffic flow cooperative optimization control method according to any one of claims 1 to 5, characterized in that a microscopic traffic flow simulation environment is constructed, and simulation results before and after optimization under different traffic situations are compared.