CN111724602B - Multi-vehicle cooperative control method under urban non-signal control multi-intersection environment - Google Patents

Abstract

The invention discloses a multi-vehicle cooperative control method under an urban non-signal control multi-intersection environment, which comprises the following steps: step 1, acquiring macro traffic network operation situation prediction state information and short-term traffic network boundary control state prediction information among sub-areas of each intersection; step 2, constructing a guiding and cooperative control method of the internal and boundary traffic flow of each intersection subregion network; and 3, designing a multi-objective optimization control method for multi-intersection multi-vehicle system cooperative driving, which comprehensively considers the macroscopic traffic state and the microscopic multi-vehicle system cooperative control, on the basis of the steps 1 and 2. The invention can save computing resources, improve the traffic efficiency of multiple intersections and improve the vehicle performance.

Description

Multi-vehicle cooperative control method under urban non-signal control multi-intersection environment

Technical Field

The invention belongs to the technical field of intelligent traffic, and particularly relates to a multi-vehicle system cooperative control method in an urban non-signal control multi-intersection environment.

Background

With the rapid increase of the quantity of motor vehicles kept, the problems of traffic jam and traffic safety become one of the key factors limiting the development of cities. How to take reasonable and effective traffic control measures has important significance for relieving traffic jam, improving road driving safety and improving vehicle performance.

The control strategies of a multi-vehicle system under the traditional urban non-signal control multi-intersection environment are mainly divided into two types: centralized control and distributed control. The centralized control strategy mostly considers the whole traffic flow of multiple intersections, adopts the ways of route induction and the like to improve the traffic flow, improve the average speed and realize route selection, but rarely considers the multi-target optimization problems of improving the fuel economy and the like of a multiple vehicle system, and depends on a network controller, so that the calculation amount is huge; the distributed control strategy mostly considers the multi-objective optimization problems of improving fuel economy, following performance and the like of a multi-vehicle system, but only uses local information, only can evaluate the traffic density between adjacent road sections, possibly causes deadlock when vehicles stably run uninterruptedly, and is not suitable for road network scenes formed by large-scale intersections. Therefore, in the existing research situation, different control architectures and strategies need to be adopted for different control targets, and a universal unified architecture and strategy are lacked.

In summary, a method for cooperatively controlling a multi-vehicle system in a city non-signal control multi-intersection environment integrating "centralized type + distributed type" needs to be designed, so that cooperative scheduling of the multi-vehicle system among the non-signal control multi-intersections is realized, the method has universality, the method can be applied to a road network scene formed by a large-scale non-signal control multi-intersection, the traffic passing efficiency of the multi-intersection can be improved on a macroscopic level, the fuel consumption of the multi-vehicle system can be reduced on a microscopic level, the following performance of the multi-vehicle system can be improved, and the like.

Disclosure of Invention

Therefore, it is an object of the present invention to provide a method for cooperative control of multiple vehicle systems in an urban non-signal-controlled multi-intersection environment, which overcomes or at least alleviates at least one of the drawbacks of the prior art.

In order to achieve the aim, the invention provides a multi-vehicle cooperative control method under the environment of an urban non-signal control multi-intersection, which comprises the following steps:

step 1: constructing a control system comprising a multi-intersection region integrated regulation and control unit, an intersection subregion regulation and control unit and a vehicle-mounted dynamic decision and management control unit;

step 2: the integrated regulation and control unit of the multiple intersection sub-areas is used for providing comprehensive evaluation for the macroscopic accumulated traffic running state among the intersection sub-areas regulated and controlled by the intersection sub-area regulation and control unit according to the macroscopic traffic network running state prediction state information among the intersection sub-areas and the short-time traffic network boundary control state prediction information of the intersection sub-areas, and performing balanced control on the traffic flow states inside the intersection sub-areas and among the boundaries;

and step 3: the intersection subregion regulating and controlling unit is used for constructing a vehicle formation system according to the traffic network state information and the vehicle state information of each vehicle which is about to reach the intersection subregion control boundary, performing traffic sequence distribution, path distribution and speed distribution and sending decision information to the vehicle-mounted dynamic decision and management control unit;

and 4, step 4: the vehicle-mounted dynamic decision and management control unit is provided with a vehicle head-vehicle optimization controller and a follower vehicle-mounted controller of the vehicle formation system;

the head vehicle optimization controller receives decision information of each intersection subregion regulation and control unit, receives state information of following vehicles in the system, receives state information of tail vehicles of a front vehicle formation system, performs multi-objective collaborative optimization control, and sends optimization control decision information to all following vehicles in the system;

the follow vehicle-mounted controller receives the optimization control decision information sent by the head vehicle of the system, receives the state information sent by the front vehicle, and automatically follows the head vehicle and the front vehicle of the system to adjust the longitudinal and transverse motion state.

Further, the step 2 is implemented as follows:

step 2.1: the integrated regulation and control unit of the multi-intersection region obtains the road condition information of the current road through the real-time interactive communication between the vehicle and the road side unit;

step 2.2: the integrated regulation and control unit of the multiple intersection regions calculates and obtains the prediction state information of the operation situation of the macroscopic traffic network among the subareas of each intersection and the prediction state information of the boundary control state of the short-time traffic network of each intersection subarea;

step 2.3: the integrated regulation and control unit of the multiple intersection area quantitatively determines the weight of the two information in the step 2.2, and the optimal solution of the traffic running state of each intersection subregion is solved;

step 2.4: the integrated regulation and control unit of the multiple intersection regions establishes the comprehensive evaluation of the macroscopic accumulated traffic running state among the sub-regions of each intersection;

step 2.5: if the comprehensive evaluation of the macroscopic accumulated traffic running state does not meet the road congestion condition, returning to the step 2.2 for circular monitoring calculation, and if so, entering the step 2.6;

step 2.6: carrying out traffic flow state balance control inside each intersection subregion and between boundaries;

step 2.7: and (5) changing the real-time running path and the average running speed of each traffic flow according to the step 2.6, and reducing the traffic jam time.

Further, the step 3 is implemented as follows:

step 3.1: acquiring the distance between adjacent multi-vehicle systems and the length of a queuing queue according to the vehicle condition information of the adjacent multi-vehicle systems in the same lane at the current moment through information interaction between vehicles;

step 3.2: combining the traffic flow states regulated and controlled by the integrated regulation and control unit in the multi-intersection region through information interaction between vehicles and roadside units, and forming a plurality of vehicle formation systems by the multi-vehicle systems in the current traffic network according to the distance between the adjacent multi-vehicle systems and the queuing length acquired in the step 3.1;

step 3.3: and distributing the dynamic passing time, the passing sequence and the expected average driving speed of all the vehicle formation systems entering the sub-areas of the intersections according to the principle of passing first when entering the sub-areas of the intersections.

Further, the step 4 is implemented as follows:

step 4.1: through information interaction between each vehicle formation system and each intersection subregion regulation and control unit, a head vehicle in each vehicle formation system receives the passing sequence and speed distributed by each intersection subregion regulation and control unit;

step 4.2: designing a head vehicle optimization controller in each vehicle formation system for the purposes of safety, energy conservation, stability and comfort of a head vehicle, carrying out local expected trajectory planning, and outputting expected control quantity;

step 4.3: and sending the output expected control quantity to vehicle-mounted controllers of other following vehicles except the head vehicle in the vehicle formation system according to the communication topological structure of the multi-vehicle formation system, wherein the other following vehicles run along with the head vehicle in the vehicle formation system.

Furthermore, in step 2.2, the method for obtaining the operation situation prediction state information of the macro traffic network between the sub-areas of each intersection by calculating by the multi-intersection area integrated control unit is as follows:

step 2.2.1: assuming that each intersection subregion has a well-defined traffic flow model MFD, each time step k in subregion i is used_tTo establish a network traffic flow Q (k)_t) And total number of vehicles N_i(k_t) The relationship of (a) to (b) is as follows:

Q(k_t)＝G_i(N_i(k_t))

in the formula, G_i(. represents a function of MFD, N_i(k_t)＝n_i(k_t)+n_ij(k_t)，n_i(k_t) Indicating the number of local vehicles, n_ij(k_t) Representing the number of vehicles from sub-area i to destination sub-area j, then for k_tAt +1 there are: n is a radical of_i(k_t+1)＝n_i(k_t+1)+n_ij(k_t+1)，

In the formula, T represents a sampling interval, d_ii(k_t) And d_ij(k_t) Respectively correspond to at k_tThe traffic demand arriving at sub-area i and the traffic demand arriving at sub-area j, p at that moment_ii(k_t) Is represented at k_tProbability of one-step transition of traffic flow leaving sub-area i at any moment, p_ij(k_t) Is represented at k_tProbability of one-step transition of traffic flow from sub-area i to sub-area j at a time, Q_ji(k_t) Is at k_tTraffic flow, Q, entering sub-area i from sub-area j at a time_i,O(k_t) Is at k_tTraffic flow, Q, leaving sub-area i at any time_ij(k_t) Is at k_tThe traffic flow from the sub-area i to the sub-area j at the moment, and n represents the number of the sub-areas of the intersection;

step 2.2.2: establishing a macroscopic traffic network operation situation prediction state information objective function by minimizing the flow passing through the intersection subregion:

in the formula, N_i,c(k_t+1) denotes k_tThe critical vehicle number of the maximum spatial average traffic flow at +1 time;

step 2.2.3: establishing an objective function constraint framework:

a vehicle number constraint

In the formula, Q_i,I(k_t) Represents k_tThe total inflow of the sub-area i at the moment;

traffic flow constraint

In the formula, Q_i(k_t+1) denotes k_tSub-area i network traffic flow at +1 moment, Q_i,O(k_t+1) denotes k_tTraffic flow leaving sub-area i at +1 moment, Q_ji(k_t+1) denotes k_tThe traffic flow entering the sub-area i from the sub-area j at the moment of + 1;

link maximum traffic flow constraint

0≤Q_ij(k_t+1)≤m_ijρ_i,ij

In the formula, Q_ij(k_t+1) denotes k_t+1 flow of traffic from sub-area i into sub-area j at time m_ijRepresents the number of links, ρ, of the sub-region i and its neighbor sub-region j_i,ijIs the average saturated traffic flow on the link between sub-area i and its neighboring sub-area j;

furthermore, in step 2.2, the method for calculating and obtaining the control state prediction information of the short-time traffic network boundary of each intersection subregion by the multi-intersection region integrated control unit is as follows:

step 2.2.1: assuming a sampling cycle time c_cEqual to the sampling time step k of all intersections_tThe total number of vehicles in the link (i, j) of the intersection sub-area may beThe update is performed by the following conservation equation:

n_i,j(k_t+1)＝n_i,j(k_t)+(α_ij,I(k_t)-α_ij,O(k_t)·c_c

in the formula, n_i,j(k_t+1) denotes k_tTotal number of vehicles entering link (i, j) at time + 1; n is_i,j(k_t) Represents k_tTotal number of vehicles entering link (i, j) at time; alpha is alpha_ij,I(k_t) Represents k_tThe traffic entering the link (i, j) at a time is the sum of the traffic flowing out of its upstream link, α_ij,O(k_t) Represents k_tThe traffic leaving link (i, j) at a time is the sum of the incoming traffic of its downstream links; wherein the content of the first and second substances,

α_ij,O(k_t)＝min(β_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_c,q_ij,o(k_t)/c_c+α_ij,I(k_t),β_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_c)，β_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_cindicating capacity at the intersection, q_ij,o(k_t)/c_c+α_ij,I(k_t) Indicating the number of waiting and arriving vehicles, beta_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_cRepresenting available space of downstream road section, beta_ij,o(k_t) Represents the relative fraction, μ, of flow out of the subregion j_ijRepresents the saturated flow leaving the link (i, j), g_ij,o(k_t) Represents the feasible time length of the traffic flow to j in the link (i, j), q_ij,o(k_t) Representing the traffic density, C, of the outgoing links (i, j)_j,oRepresenting the capacity of the downstream link, n, in terms of the number of vehicles_j,o(k_t) Is the number of vehicles in link (i, j);

α_ij,I(k)＝β_ij,I·α_ij,O(k_t)，β_ij,Iis the relative fraction of flow to sub-region j;

step 2.2.2: establishing a short-term traffic network boundary control state prediction information objective function by minimizing the flow passing through the sub-area of the intersection and the difference between the total traffic flow and the optimal traffic flow:

in the formula, alpha_ij,O(k_t+1) denotes k_t(ii) traffic flow leaving link (i, j) at time +1, where α_lRepresents the lowest flow, Q_ij(k_t+1) denotes k_tThe traffic flow from sub-area i into sub-area j at time + 1.

Still further, in the step 2.3, the method for solving the optimal solution of the traffic running state of each intersection subregion is as follows:

establishing a cost function:

Q^*(k_t)＝min(w_iQ_w(k_t)+w_jQ_h(k_t))

Q^*(k_t) Represents k_tOptimal solution, w, of traffic running flow state in sub-area of each intersection at any moment_iThe method is to consider the operation situation of the sub-region i to predict the state information weight coefficient, Q_w(k_t) Is k_tForecasting state information value w of macro traffic network operation situation of sub-area i at moment_jConsidering the weight coefficient, Q, of the information predicted by the boundary control state of the short-term traffic network of the sub-area j_h(k_t) Is k_tPredicting information values of the short-time traffic network boundary control states of the sub-area j at the moment; w is a_i、w_jIs an empirical tuning value.

Further, in step 2.4, a method for making a comprehensive evaluation of the macroscopic cumulative traffic operation state between the sub-areas of each intersection is as follows:

step 2.4.1: establishing a macroscopical accumulated traffic running state comprehensive evaluation parameter index set by adopting a fuzzy comprehensive evaluation method

Q^*Represents the optimal solution of the traffic running flow state of each intersection subregion,

representing average speed, L queue length, ATTR travel-time ratio; establishing a macroscopic accumulated traffic running state comprehensive evaluation parameter evaluation set B ═ B₁,b₂,b₃,b₄]，b₁Indicating congestion, b₂Indicating light congestion, b₃Indicates substantial absence of flow, b₄Indicating unblocked;

step 2.4.2: fuzzifying the traffic running state, and performing fuzzy reasoning according to rules under different traffic states, wherein the single-factor fuzzy discrimination matrix is as follows:

wherein the content of the first and second substances,

respectively represent the optimal membership function values of the traffic operation flow state,

respectively represent the average velocity

The value of the membership degree function of (c),

respectively representing membership function values of the queuing lengths L,

respectively representing membership function values of the travel time ratio ATTR;

step 2.4.3: weighting is carried out on the comprehensive evaluation parameter indexes of each traffic running state, and a final evaluation result is obtained as follows:

in the formula, B is a fuzzy comprehensive evaluation value which falls into which evaluation range of the comprehensive evaluation parameter evaluation of the macroscopic accumulated traffic running state and belongs to which traffic condition; omega₁,ω₂,ω₃,ω₄The weight coefficients of the four traffic running state comprehensive evaluation parameter indexes are respectively,

representing a fuzzy synthesis operation.

Still further, in step 2.6, the method for performing the traffic flow state balance control inside each intersection subregion and between the boundaries includes the following steps:

step 2.6.1: establishing a cost function

In the formula, Q_i(k_t) Represents k_tInformation on the actual traffic state of sub-area i of the intersection at that moment, Q_j(k_t) Represents k_tInformation on the actual traffic state of sub-area j of the intersection at that moment, Q^*(k_t) Represents k_tThe optimal solution of the traffic running flow state of each intersection subregion at any moment;

step 2.6.2: building a constraint framework

0≤Q_i(k_t)≤Q_max

0≤Q_j(k_t)≤Q_max

In the formula, Q_maxRepresents a maximum value of the traffic state parameter;

and under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

Still further, in step 3.3, the method for distributing the expected average running speed of all the vehicle formation systems is as follows:

step 3.3.1: establishing a cost function

In the formula, V_i(k_t) Represents k_tActual average velocity, V, of sub-area i of the time intersection_i ^*(k_t) Represents k_tThe optimal reference speed of the intersection sub-area i at the moment,

and

respectively represents k_tThe average speed of the intersection sub-areas i and j under the time optimal traffic flow conditions,

is a weight coefficient;

step 3.3.2: building a constraint framework

0≤V_i(k_t)≤V_max

In the formula, V_maxIn order to set the maximum speed limit of the road,

and

respectively the minimum and maximum average acceleration of the multi-vehicle system in the intersection area;

and under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

Still further, in step 4.2, the method for designing the head-up optimization controller in each vehicle formation system is as follows:

step 4.2.1: the nonlinear dynamical equation of the vehicle is established as follows:

in the formula, S_i(k_t) And v_i(k_t) Respectively displacement and velocity, T, of the ith vehicle_q,i(k_t) Is the actual torque of the vehicle, u_i(k_t) To the desired torque, i₀Is a mechanical transmission ratio, eta_m,iFor mechanical efficiency of the transmission system, m_iAs mass of the vehicle, C_D,iIs the drag coefficient of the vehicles in the queue, rho is the air density, A_iIs the frontal area of the vehicle, g is the acceleration of gravity, r_w,iIs the rolling radius of the wheel, f is the rolling resistance coefficient, tau_iIs the longitudinal power system time lag coefficient, alpha is the road gradient, delta k_tIs a discrete time step;

discretizing the kinetic equation by an Euler method, and rewriting the nonlinear kinetic equation into:

x_i(k_t+1)＝φ_i(x_i(k_t))+ψ_iu_i(k_t)，i∈Ν

wherein x is_i(k_t) N is the state quantity of the vehicle, the number of the vehicles in the queue,

constructing an output for each vehicle in a vehicle fleetIs y_i(k_t)＝[S_i(k_t),v_i(k_t)]^T＝γx_i(k_t) Wherein

Order:

X(k_t)＝[x₁ ^T(k_t),x₂ ^T(k_t),...,x_N ^T(k_t)]^T，

U(k_t)＝[u₁(k_t),u₂(k_t),...,u_N(k_t)]^T，

Y(k_t)＝[y₁ ^T(k_t),y₂ ^T(k_t),...,y_N ^T(k_t)]^T，

Φ(X(k_t))＝[φ₁(x₁)^T,φ₂(x₂)^T,...,φ_N(x_N)^T]^T，

Ψ＝diag{ψ₁,ψ₂,...,ψ_N}，

the equation of state for the vehicle fleet as a whole is described as:

X(k_t)＝Φ(X(k_t))+Ψ·U(k_t)

Y(k_t)＝ΓX(k_t)

in the formula (I), the compound is shown in the specification,

step 4.2.2: defining a sub-prediction optimization problem on each vehicle in a vehicle formation system, carrying out optimization solution on each sub-prediction optimization problem by using information of adjacent vehicles and head vehicles, and constructing a distributed controller for each vehicle by adopting a distributed model prediction control method, wherein the method comprises the following steps:

a, establishing an objective function

J＝min(f₁(k_t)+f₂(k_t)+f₃(k_t)+f₄(k_t))

Wherein, the following cost function f of the ith vehicle and the head vehicle in the vehicle formation system₁(k_t) The following were used:

f₁(k_t)＝ω_s1(h_i(k_t)-h₁(k_t)-d_i,1)²+ω_v1(Z_i(k_t)-V₁ ^*(k_t))²

ω_s1is the distance error weight coefficient of the ith vehicle and the head vehicle in the formation system, h_i(k_t) For the predicted position of the ith vehicle in the formation system, h₁(k_t) For the predicted position of the head car, d_i,1Is the desired distance, ω, between the ith vehicle and the head vehicle in the formation system_v1Is the speed error weight coefficient, Z, of the ith vehicle and the head vehicle in the formation system_i(k_t) For the predicted speed, V, of the i-th vehicle in the formation system₁ ^*(k_t) To predict the speed of the head car in the formation system,

wherein, the following cost function f of the ith vehicle and the front vehicle j in the formation system₂(k_t) The following were used:

f₂(k_t)＝ω_s,i(h_i(k_t)-h_j(k_t)-d_i,j)²+ω_v,j(Z_i(k_t)-Z_j(k_t))²

ω_s,iis the distance error weight coefficient of the ith vehicle and the preceding vehicle j in the formation system, h_j(k_t) For the predicted position of the vehicle j ahead of the i-th vehicle in the formation system, d_i,jIs the desired distance, ω, between the ith and preceding vehicle j in the formation system_v,jIs a velocity error weight coefficient, Z, of a preceding vehicle j in a formation system_j(k_t) The predicted speed of the front vehicle j in the formation system;

wherein the stability cost function f₃(k_t) The following were used:

α_i1for the predicted output trajectory and assumed output trajectory series error weight coefficients for each vehicle in the formation system,

in order to predict the output trajectory,

is a hypothetical output trajectory;

wherein, a comfort and energy-saving cost function f₄(k_t) The following were used:

f₄(k_t)＝α_i2(u_i(k_t)-T_i(k_t))²

α_i2for comfort weight coefficient, T, of each vehicle in the formation system_i(k_t) The torque of each vehicle in the formation system is the torque when the vehicle runs at a constant speed;

b, establishing a constraint framework

v_min≤v_i(k_t)≤v_max

T_min≤u_i(k_t)≤T_max

V_i ^p(N_p|k_t)＝V_i ^*(N_p|k_t)

h_i ^p(N_p|k_t)＝h₁ ^p(N_p|k_t)-(i-1)d_i,j

T_i ^p(N_p|k_t)＝h_i(V_i ^p(N_p|k_t))

v_i(k_t) Is the actual speed, v, of the i-th vehicle in the formation system_min、v_maxMaximum and minimum speed, T, of the ith vehicle in the formation system, respectively_min、T_maxMaximum and minimum torque for the ith vehicle in the formation system, respectively，V_i ^p(N_p|k_t) The vehicle speed of the ith vehicle in the formation system is the vehicle speed when the vehicle state converges to the steady state at the time of predicting the terminal, the vehicle speed is the same as the optimal reference speed of a sub-area i of the intersection, and V is the optimal reference speed of the sub-area i of the intersection_i ^*(N_p|k_t) Is the speed h of the ith vehicle in the formation system when the ith vehicle runs at a constant speed_i ^p(N_p|k_t) For the displacement of the prediction terminal of the ith vehicle in the formation system, the displacement of the vehicle prediction terminal is designed to be consistent with the expected displacement of the ith vehicle and the head vehicle, h₁ ^p(N_p|k_t) For predicting the expected displacement of the terminal of the head car in the formation system, d_i,jFor a desired inter-vehicle distance, T, between two adjacent vehicles in a formation system_i ^p(N_p|k_t) Predicting a terminal torque, h, for the vehicle of the i-th vehicle in the formation system_i(V_i ^p(N_p|k_t) Is the vehicle equilibrium torque when the ith vehicle in the convoy system is running at a constant speed.

In summary, the beneficial effects of the present invention include:

1. the multi-vehicle system cooperative control method for the urban non-signal control multi-intersection environment can fuse a centralized and distributed multi-intersection network control framework, overcomes the defect that the distributed or centralized control framework is singly adopted, enables the framework to have universality, and can be suitable for road network scenes formed by large-scale intersections.

2. The designed traffic flow balanced dispatching control method under the non-signal control multi-intersection environment comprehensively considers the operation state prediction information of the macro traffic network among the sub-areas of each intersection and the boundary control state prediction information of the short-time traffic network of each intersection, and carries out dynamic traffic flow distribution and coordinated dispatching of a multi-vehicle system based on the multi-agent consistency control theory, thereby realizing the optimal traffic operation state among the sub-areas of the regulation and control area.

3. The designed multi-intersection multi-vehicle system multi-target collaborative optimization control method combines the current traffic running state and the multi-vehicle system running state, can improve the traffic passing efficiency of the multi-intersection on a macroscopic level, can reduce the fuel consumption of the multi-vehicle system on a microscopic level, improves the vehicle following performance of the multi-vehicle system and the like.

Drawings

In order to more clearly illustrate the technical solution in the present embodiment, the following description will be made of the drawings required for the embodiment or the prior art description:

FIG. 1(a) is a layered control architecture diagram of a multi-vehicle system cooperative control system in an urban non-signal-controlled multi-intersection environment;

FIG. 1(b) is a logic diagram of the cooperative control of a multi-vehicle system in an urban non-signal-controlled multi-intersection environment;

FIG. 2 is a schematic flow chart of a real-time online traffic running state comprehensive evaluation algorithm in the embodiment of the invention;

fig. 3 is a schematic flow chart of the traffic flow equalization optimization control in the embodiment of the present invention.

Detailed Description

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

The invention provides a multi-vehicle cooperative control system under an urban non-signal control multi-intersection environment, and a control system layered control architecture diagram is shown in figure 1 (a); fig. 1(b) is a control logic diagram.

As shown in fig. 1(a), the control architecture is divided into 3 layers: the system comprises a multi-intersection region integrated regulation and control unit, intersection subregion regulation and control units (1-n) and vehicle-mounted dynamic decision and management control units (1-n).

The vehicle-mounted dynamic decision and management control unit is provided with a formation system head vehicle optimization controller and a following vehicle on-board controller.

The multi-intersection region integrated regulation and control unit receives the traffic flow state information of each intersection subregion regulation and control unit and sends the decision information of the multi-intersection region integrated regulation and control unit to each intersection subregion regulation and control unit.

Each intersection subregion regulation and control unit receives multi-vehicle system state information sent by each formation system head vehicle (which can be designated as a first vehicle), carries out traffic sequence distribution, path distribution and speed distribution, and sends decision information of the multi-intersection region integrated regulation and control unit and each intersection subregion regulation and control unit to each formation system head vehicle controller.

And the head vehicle controller of each formation system receives the decision information of the regulation and control unit of each intersection subregion, receives the state information of other following vehicles except the head vehicle in the formation system and receives the tail vehicle information of the front formation system to perform multi-objective collaborative optimization control, and sends the optimized decision control information to all the following vehicles in the formation system.

And the follow vehicle-mounted controller receives the current state information sent by the head vehicle and receives the state information sent by the front vehicle to perform multi-target optimization control, and the head vehicle and the front vehicle of the formation system are automatically followed to adjust the longitudinal and transverse motion states.

In connection with the hierarchical control architecture of fig. 1(a), the system control logic is as shown in fig. 1 (b):

the integrated regulation and control unit of the multi-intersection region predicts state information according to the macroscopic traffic network operation situation among the sub-regions of each intersection and predicts the state information according to the short-time traffic network boundary control state of each intersection sub-region, provides comprehensive evaluation for the regulation and control unit of each intersection sub-region to realize the optimal traffic operation state among the sub-regions, and performs balance control on the traffic flow states inside the multi-intersection region and among the boundary regions by adopting a consistency optimization algorithm.

Each intersection subregion regulation and control unit obtains the traffic scheduling network state information and the vehicle control state information of each vehicle at the current moment, which is about to reach each intersection subregion control boundary, and the method comprises the following steps: the method comprises the steps of numbering and grouping a multi-vehicle system according to vehicle IDs, driving intentions, current speeds, current positions, time intervals to reach intersections and the like, planning vehicle driving paths of each grouping system and distributing optimized reference speeds for vehicles in a control network area by combining balanced control of traffic flow states inside a multi-intersection area and between boundary areas through a multi-intersection area integrated regulation and control unit, and finally achieving efficient coordinated dispatching and distribution of the multi-vehicle system inside the sub-areas of the multi-intersection and between the sub-areas.

Each vehicle-mounted dynamic decision and management control unit receives traffic scheduling management and control instruction information sent by each intersection subregion regulation and control unit, and the method comprises the following steps: the method comprises the steps that a multi-vehicle system formation instruction, a multi-vehicle system passing order instruction, optimal expected speed information and the like are adopted, distributed rolling time domain decision and dynamic optimization control are adopted as vehicle-mounted dynamic decision and management control strategies, the expected control quantity of longitudinal and transverse vehicle running is formulated with the purposes of safety, comfort, stability and energy conservation, and the expected control quantity is converted into execution driving force by means of a vehicle dynamics model; and uploading the traffic scheduling execution and control state information of the own vehicle to each intersection subregion control unit, wherein the traffic scheduling execution and control state information of the own vehicle comprises the following steps: vehicle ID, travel intent, travel path, current speed, current location, vehicle acceleration, etc.

The information sources required by the control process comprise various information from a road side unit, a vehicle-mounted controller, a multi-intersection region integrated regulation and control unit and each intersection subregion regulation and control unit. The information source is responsible for information processing and interaction of all V2V/V2I/I2I, and records running state information of all traffic systems, namely, each intersection subregion regulation and control unit and each vehicle-mounted dynamic decision and management control unit can obtain information required by the information source, and a communication network topology structure among multiple vehicle systems, information required to be transmitted and a transmission process of the information among all layers are designed in the information source module. The multi-intersection region integrated regulation and control unit receives the traffic flow state information of each intersection subregion regulation and control unit, and sends the decision information of the multi-intersection region integrated regulation and control unit to each intersection subregion regulation and control unit through I2I. And each intersection subregion regulating and controlling unit receives multi-vehicle system state information sent by each formation system head vehicle, and sends the decision information of the multi-intersection subregion integrated regulating and controlling unit and each intersection subregion regulating and controlling unit to each formation system head vehicle controller through V2I. Each on-board dynamic decision and management control unit obtains the information needed by itself from the information source through V2V.

Based on the control architecture and the control logic, the invention provides a multi-vehicle system cooperative control method facing to an urban non-signal control multi-intersection environment, which comprises the following steps:

step 1: the control system comprising the integrated regulation and control unit of the multiple intersection regions, the regulation and control units of the intersection sub regions, the vehicle-mounted dynamic decision and management control units and the information source is constructed, and the relationship is as described above.

Step 2: the method for making the control measures of the integrated regulation and control unit of the multi-intersection region specifically comprises the following steps:

step 2.1: through the real-time interactive communication between the vehicle and the road side unit, the road condition information of the current road is obtained by utilizing the integrated regulation and control unit in the multi-intersection area, wherein the road condition information comprises: traffic density, boundary average speed, queue length, travel time ratio, etc.

Step 2.2: according to the step 2.1, the prediction state information of the operation situation of the macro traffic network and the prediction information of the boundary control state of the short-time traffic network in each intersection subregion are calculated and obtained.

Step 2.3: and (3) quantitatively determining the weight of each target value in the step 2.2, and solving the optimal solution of the traffic running state of each intersection subregion.

Step 2.4: and formulating the comprehensive evaluation of the macroscopic accumulated traffic running state among the subareas of the intersection.

Step 2.5: and if the comprehensive evaluation of the macroscopic accumulated traffic running state does not meet the road congestion condition, returning to the step 2.2 for recalculation, and if so, entering the step 2.6.

Step 2.6: and performing guidance and cooperation control of the traffic flow coordination control in each intersection subregion network and the boundary traffic flow between each intersection subregion.

Step 2.7: and changing the real-time running path of each traffic flow and the average running speed of the current traffic flow according to the step 2.6, and reducing the traffic jam time and the space state of the traffic network density.

And step 3: the method comprises the following steps of making control measures of each intersection subregion regulation and control unit, wherein the control measures specifically comprise the following steps:

step 3.1: through information interaction between vehicles, the distance between adjacent multi-vehicle systems and the length of a queuing queue are obtained according to vehicle condition information of the adjacent multi-vehicle systems in the same lane at the current moment, wherein the vehicle condition information comprises current torque, current speed and real-time position information of each vehicle in the multi-vehicle systems.

Step 3.2: and (3) forming a plurality of vehicle formation systems from the multi-vehicle systems in the current traffic network by information interaction between the multi-vehicle systems and the road side unit, combining the expected running path and the expected average running speed of the multi-vehicle systems obtained in the step (2.7) and the distance between the adjacent multi-vehicle systems and the length of the queue of the adjacent multi-vehicle systems obtained in the step (3.1).

Step 3.3: all the vehicle formation systems distribute the passing time and space passing resources entering the sub-areas of the intersections according to the principle that the vehicles firstly enter the sub-areas of the intersections and pass firstly, and the centralized distribution of the passing order and the average driving speed of each vehicle formation system is realized.

And 4, step 4: the method for making the control measures of each vehicle-mounted dynamic decision and management control unit specifically comprises the following steps:

step 4.1: and 3.3, receiving the dynamic passing sequence and the expected average running speed distributed by each intersection subregion regulating and controlling unit in the step 3.3 by the head vehicles in each vehicle formation system through information interaction between each vehicle formation system and each intersection subregion regulating and controlling unit.

Step 4.2: and (4) designing a multi-objective collaborative optimization controller of the safety, the energy saving performance, the stability and the comfort of the head vehicles in each vehicle formation system according to the step 4.1, carrying out local expected trajectory planning, and outputting expected control quantity.

Step 4.3: according to the step 4.2, the output expected control quantity is sent to the vehicle-mounted controllers of other following vehicles except the head vehicle in the vehicle formation system according to the communication topological structure of the multi-vehicle formation system, and the other following vehicles drive along with the head vehicle in the vehicle formation system.

Further, the method for acquiring the "macro traffic network operation situation prediction state information" in step 2.2 is as follows:

step 2.2.1: aiming at a complex urban traffic network with uneven traffic flow distribution of each intersection subregion to be regulated, a multi-intersection subregion integrated regulation and control unit traffic flow model MFD (Macroscopic Fundamental Diagram, a known training model) is constructed on the basis of integrating the space average density balance of each intersection subregion and the conservation characteristics of the traffic running state of each intersection subregion in and out.

Assuming each intersection subregion has a well-defined traffic flow model MFD, each time step k in subregion i is used_tTo establish a network traffic flow Q (k)_t) And total number of vehicles N_i(k_t) The function is expressed as follows:

Q(k_t)＝G_i(N_i(k_t))

wherein G is_i(. cndot.) represents a function of MFD, which can be obtained by a polynomial fitting method. According to the different destinations of the traffic flow, k_tTotal number of vehicles N passing through area i at time_i(k_t) Is the sum of two components, i.e. representing the number n of local vehicles_i(k_t) And the number of vehicles n from sub-area i to destination sub-area j_ij(k_t) The sum of (a) and (b).

Then for k_tTotal number of vehicles N passing through sub-area i at +1 moment_i(k_t+1), then there are: n is a radical of_i(k_t+1)＝n_i(k_t+1)+n_ij(k_t+1), wherein the local number of vehicles n_i(k_t+1) and the number of vehicles n from sub-area i to destination sub-area j_ij(k_t+1) are respectively:

where T is the sampling interval, d_ii(k_t) And d_ij(k_t) Respectively correspond to at k_tThe traffic demand reaching the destination sub-area i and the traffic demand reaching the destination sub-area j at the moment; q_ji(k_t) Is at k_tTraffic flow, Q, entering sub-area i from sub-area j at a time_i,O(k_t) Is at k_tTraffic flow, Q, leaving sub-area i at any time_ij(k_t) Is the traffic flow leaving sub-area i to sub-area j; p is a radical of_ii(k_t) Is at k_tAt the moment, the probability of one-step transition of the traffic flow leaving the subarea i, p_ij(k_t) Is at k_tAt the moment, the one-step transition probability of the traffic flow from the subarea i to the subarea j also represents the proportion of the traffic flow to different destinations in the whole network traffic flow; n represents the number of intersection sub-regions.

Step 2.2.2: establishing a macroscopic traffic network operation situation prediction state information objective function by minimizing the flow passing through the intersection subregion:

in the formula, N_i(k_t+1) denotes k_tThe total number of vehicles passing through sub-area i at time +1, T represents the sampling interval, N_i,c(k_t+1) denotes a value corresponding to k_tThe critical number of vehicles for the maximum spatial average traffic flow at time + 1.

Step 2.2.3: establishing a constraint framework for the objective function, which comprises the following steps:

a vehicle number constraint

In the formula, Q_i,I(k_t) Represents k_tTotal inflow, Q, of the time sub-zone i_i,O(k_t) Represents k_tTraffic flow, Q, leaving sub-area i at any time_ji(k_t) Represents k_tTraffic flow, Q, entering sub-area i from sub-area j at a time_ij(k_t) Represents k_tTraffic flow, N, entering sub-area j from sub-area i at any moment_i(k_t) Represents k_tThe total number of vehicles passing through sub-area i at that moment.

Traffic flow constraint

In the formula, Q_i(k_t+1) denotes k_tSub-area i network traffic flow at +1 moment, Q_i,O(k_t+1) denotes k_tTraffic flow leaving sub-area i at +1 moment, Q_ji(k_t+1) denotes k_tThe traffic flow entering sub-area i from sub-area j at time + 1.

Link maximum traffic flow constraint

0≤Q_ij(k_t+1)≤m_ijρ_i,ij

In the formula, Q_ij(k_t+1) denotes k_t+1 flow of traffic from sub-area i into sub-area j at time m_ijRepresents the number of links, ρ, of the sub-region i and its neighbor sub-region j_i,ijIs the average saturated traffic flow on the link between sub-area i and its neighbor sub-area j.

Further, the method for acquiring the "short-term traffic network boundary control state prediction information" in step 2.2 is as follows:

step 2.2.1: constructing traffic flow models of the regulation and control units of the sub-regions of each intersection:

assuming a sampling cycle time c_cEqual to the sampling time step k of all intersections_tThen the total number of vehicles in link (i, j) of the intersection sub-area can be updated by the following conservation equation (link (i, j) refers to the link from i to j direction):

n_i,j(k_t+1)＝n_i,j(k_t)+(α_ij,I(k_t)-α_ij,O(k_t)·c_c

in the formula, n_i,j(k_t+1) denotes k_tTotal number of vehicles entering link (i, j) at time + 1; n is_i,j(k_t) Represents k_tTotal number of vehicles entering link (i, j) at time; alpha is alpha_ij,I(k_t) Represents k_tThe traffic entering the link (i, j) at a time is the sum of the traffic flowing out of its upstream link, α_ij,O(k_t) Represents k_tThe traffic leaving link (i, j) at a time is the sum of the incoming traffic of its downstream links. Wherein the content of the first and second substances,

α_ij,O(k_t)＝min(β_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_c,q_ij,o(k_t)/c_c+α_ij,I(k_t),β_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_c) In the formula beta_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_cIndicating the traffic capacity of the intersection, q_ij,o(k_t)/c_c+α_ij,I(k_t) Indicating the number of waiting and arriving vehicles, beta_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_cAnd the available space of the downstream road section is represented, and the specific value is the minimum value of the three items.

β_ij,o(k_t) Is the relative fraction, μ, of flow out of sub-region j_ijIs the saturated flow leaving the link (i, j), g_ij,o(k_t) Is the feasible time length of traffic flow to j in link (i, j), q_ij,o(k_t) Representing the traffic density, C, of the outgoing links (i, j)_j,oCapacity of downstream link, n, expressed in number of vehicles_j,o(k_t) Is the number of vehicles in link (i, j).

Wherein alpha is_ij,I(k)＝β_ij,I·α_ij,O(k_t)，β_ij,IIs the relative fraction of flow to sub-region j.

Step 2.2.2: establishing a short-time traffic network boundary control state prediction information objective function by minimizing the flow passing through the intersection subregion and the difference between the total traffic flow and the optimal traffic flow:

in the formula, n_i,j(k_t+1) denotes k_tThe total number of vehicles entering link (i, j) at time + 1; c. C_cIs the sampling cycle time, alpha_ij,O(k_t+1) denotes k_t+1 moment leaving the traffic flow in link (i, j), where α_lRepresents the lowest flow, Q_ij(k_t+1) denotes k_tThe traffic flow from sub-area i into sub-area j at time + 1.

Further, the method for solving the "optimal solution of the traffic running state of each intersection subregion" in step 2.3 is as follows:

establishing a cost function:

Q^*(k_t)＝min(w_iQ_w(k_t)+w_jQ_h(k_t))

wherein Q is^*(k_t) Represents k_tOptimal solution, w, of traffic running flow state in sub-area of each intersection at any moment_iThe method is to consider the operation situation of the sub-region i to predict the state information weight coefficient, Q_w(k_t) Is k_tForecasting state information value w of macro traffic network operation situation of sub-area i at moment_jConsidering the weight coefficient, Q, of the information predicted by the boundary control state of the short-term traffic network of the sub-area j_h(k_t) Is k_tThe short-term traffic network boundary control state prediction information value of the sub-area j at the moment. Two of the weighting coefficients are empirically derived, Q_w(k_t)、Q_h(k_t) Is calculated in step 2.2.

Further, with reference to fig. 2, in step 2.4, the specific steps of making a "macroscopic accumulated traffic running state comprehensive evaluation" between sub-areas of each intersection are as follows:

step 2.4.1: establishing a macroscopical accumulated traffic running state comprehensive evaluation parameter index set by adopting a fuzzy comprehensive evaluation method

Wherein Q^*Represents the optimal solution of the traffic running flow state of each intersection subregion,

representing average speed, L queue length, ATTR travel-time ratio; establishing a macroscopic accumulated traffic running state comprehensive evaluation parameter evaluation set B ═ B₁,b₂,b₃,b₄]Wherein b is₁Indicating congestion, b₂Indicating light congestion, b₃Indicates substantial absence of flow, b₄Indicating a clear.

Step 2.4.2: fuzzifying the traffic running state, and performing fuzzy reasoning according to rules under different traffic states, wherein the single-factor fuzzy discrimination matrix is as follows:

wherein the content of the first and second substances,

respectively represent the optimal membership function values of the traffic operation flow state,

respectively represent the average velocity

The value of the membership degree function of (c),

respectively representing membership function values of the queuing lengths L,

respectively, the membership function values of the travel time ratio ATTR.

Step 2.4.3: for each traffic running state comprehensive evaluation parameter index, weighting each evaluation parameter index according to different emphasis points and importance degrees of each basic index on the evaluation traffic state, and obtaining a final evaluation result as follows:

in the formula, B is a fuzzy comprehensive evaluation value, and the value falls into which evaluation range of the comprehensive evaluation parameter evaluation set and belongs to which traffic condition; omega₁,ω₂,ω₃,ω₄Adjusting weight coefficients for the four evaluation parameter index states respectively;

representing a fuzzy synthesis operation.

Further, with reference to fig. 3, based on a consistency algorithm, the policy method of "performing guidance and cooperative control of the internal traffic flow coordination control of each intersection subregion network and the boundary traffic flow between each intersection subregion" in step 2.6 is as follows:

step 2.6.1: establishing a cost function

In the formula, Q_i(k_t) Represents k_tInformation on the actual traffic state of sub-area i of the intersection at that moment, Q_j(k_t) Represents k_tInformation on the actual traffic state of sub-area j of the intersection at that moment, Q^*(k_t) Represents k_tAnd (4) performing optimal solution on the traffic running flow state of each intersection subregion at each moment.

Step 2.6.2: building a constraint framework

0≤Q_i(k_t)≤Q_max

0≤Q_j(k_t)≤Q_max

In the formula, Q_maxRepresenting the maximum value of the traffic status parameter.

And under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

Further, the control problem of "centralized distribution of average traveling speed" in step 3.3 is solved as follows:

step 3.3.1: establishing a cost function

In the formula, V_i(k_t) Represents k_tThe actual average speed of the intersection sub-area i at the moment,

represents k_tThe optimal reference speed of the intersection sub-area i at the moment,

wherein

And

respectively represents k_tThe average speed of the intersection sub-areas i and j under the time optimal traffic flow conditions,

are weight coefficients.

Step 3.3.2: building a constraint framework

In the formula, V_maxIs the most important roadThe speed of the motor is limited to a large value,

and

respectively the minimum and maximum average acceleration of the multi-vehicle system in the intersection area.

And under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

Further, the method for designing the multi-objective collaborative optimization controller of the head car in each vehicle formation system in the step 4.2 is as follows:

step 4.2.1: designing a nonlinear dynamic model of each vehicle formation system, which comprises the following specific steps:

in order to ensure the driving stability of the vehicle formation system and the following performance of each vehicle in the formation, the nonlinear term in the vehicle longitudinal dynamic equation needs to be considered in the control process, and the model adopts a mode of establishing a nonlinear dynamic equation to establish a nonlinear formation dynamic model.

The nonlinear dynamical equation for each vehicle is:

in the formula, S_i(k_t) And v_i(k_t) Respectively displacement and velocity, T, of the ith vehicle_q,i(k_t) Is the actual torque of the vehicle, u_i(k_t) To the desired torque, i₀Representing mechanical transmission ratio, η_m,iFor mechanical efficiency of the transmission system, m_iAs mass of the vehicle, C_D,iFor the in-queue vehicle drag coefficient, ρ represents the air density, A_iIs the frontal area of the vehicle, g is the acceleration of gravity, r_w,iIs the rolling radius of the wheel, f is the rolling resistance coefficient, tau_iIs the longitudinal power system time lag coefficient, alpha is the road gradient, delta k_tIn discrete time steps.

The above nonlinear equation can be further written in the form of the following equation by dispersing the kinetic equation by the euler method:

x_i(k_t+1)＝φ_i(x_i(k_t))+ψ_iu_i(k_t) I e Ν (N is the number of vehicles in the queue)

Wherein x is_i(k_t) Is the state quantity of the vehicle;

constructing an output of y for each vehicle in the vehicle fleet_i(k_t)＝[S_i(k_t),v_i(k_t)]^T＝γx_i(k_t) Wherein

Order:

X(k_t)＝[x₁ ^T(k_t),x₂ ^T(k_t),...,x_N ^T(k_t)]^T，

U(k_t)＝[u₁(k_t),u₂(k_t),...,u_N(k_t)]^T，

Y(k_t)＝[y₁ ^T(k_t),y₂ ^T(k_t),...,y_N ^T(k_t)]^T，

Φ(X(k_t))＝[φ₁(x₁)^T,φ₂(x₂)^T,...,φ_N(x_N)^T]^T，

Ψ＝diag{ψ₁,ψ₂,...,ψ_N}，

the equation of state for the vehicle fleet as a whole can be written as:

X(k_t)＝Φ(X(k_t))+Ψ·U(k_t)

Y(k_t)＝ΓX(k_t)

in the formula (I), the compound is shown in the specification,

step 4.2.2: according to the nonlinear queue dynamics model, a sub-prediction optimization problem is defined on each vehicle in a vehicle formation system, each sub-prediction optimization problem is optimized and solved by using information of a neighborhood vehicle and a head vehicle, and because the optimization problem only uses state information of the neighborhood vehicle and does not use global state information, the optimization problem is a Distributed optimization problem, a Distributed controller is constructed for each vehicle by adopting a DMPC (Distributed Predictive Control, Distributed model prediction Control method), and the method comprises the following steps:

a, establishing an objective function

J＝min(f₁(k_t)+f₂(k_t)+f₃(k_t)+f₄(k_t))

Wherein, the following cost function f of the ith vehicle and the first vehicle (head vehicle) in the formation system₁(k_t) The following were used:

f₁(k_t)＝ω_s1(h_i(k_t)-h₁(k_t)-d_i,1)²+ω_v1(Z_i(k_t)-V₁ ^*(k_t))²

in the above formula, ω_s1Is the distance error weight coefficient of the ith vehicle and the first vehicle in the formation system, h_i(k_t) For the predicted position of the ith vehicle in the formation system, h₁(k_t) As predicted position of the first vehicle, d_i,1Is the desired distance, ω, between the ith vehicle and the first vehicle in the formation system_v1Is the speed error weight coefficient, Z, of the ith vehicle and the first vehicle in the formation system_i(k_t) For the i-th vehicle in a formation systemPredicted speed, V₁ ^*(k_t) Is the predicted speed of the first vehicle in the convoy system, i.e. the optimal reference speed of the intersection sub-area i.

Wherein, the following cost function f of the ith vehicle and the front vehicle j in the formation system₂(k_t) The following were used:

f₂(k_t)＝ω_s,i(h_i(k_t)-h_j(k_t)-d_i,j)²+ω_v,j(Z_i(k_t)-Z_j(k_t))²

in the above formula, ω_s,iIs the distance error weight coefficient of the ith vehicle and the preceding vehicle j in the formation system, h_i(k_t) For the predicted position of the ith vehicle in the formation system, h_j(k_t) For the predicted position of the vehicle j ahead of the i-th vehicle in the formation system, d_i,jIs the desired distance, ω, between the ith and preceding vehicle j in the formation system_v,jIs a velocity error weight coefficient, Z, of a preceding vehicle j in a formation system_i(k_t) For the predicted speed, Z, of the i-th vehicle in the formation system_j(k_t) Is the predicted speed of the lead vehicle j in the formation system.

Wherein the stability cost function f₃(k_t) The following were used:

in the formula, alpha_i1For the predicted output trajectory and assumed output trajectory series error weight coefficients for each vehicle in the formation system,

in order to predict the output trajectory,

an assumed output trajectory.

Wherein, a comfort and energy-saving cost function f₄(k_t) The following were used:

f₄(k_t)＝α_i2(u_i(k_t)-T_i(k_t))²

in the formula, alpha_i2For the comfort weight coefficient, u, of each vehicle in the formation system_i(k_t) For the desired torque, T, of each vehicle in the formation system_i(k_t) The torque of each vehicle in the formation system is the torque when the vehicle runs at a constant speed.

B, establishing a constraint framework

v_min≤v_i(k_t)≤v_max

T_min≤u_i(k_t)≤T_max

V_i ^p(N_p|k_t)＝V_i ^*(N_p|k_t)

h_i ^p(N_p|k_t)＝h₁ ^p(N_p|k_t)-(i-1)d_i,j

T_i ^p(N_p|k_t)＝h_i(V_i ^p(N_p|k_t))

Wherein v is_i(k_t) Is the actual speed, u, of the ith vehicle in the formation system_i(k_t) For the desired torque, v, of each vehicle in the formation system_min、v_maxMaximum and minimum speed of the ith vehicle in the formation system, respectively, the values being selected in relation to the vehicle dynamics, T_min、T_maxThe maximum torque and the minimum torque of the ith vehicle in the formation system are respectively selected, and the values are related to the vehicle dynamic characteristics, so that the vehicle can keep running in a safe speed range and an allowable torque range; v_i ^p(N_p|k_t) The vehicle speed of the ith vehicle in the formation system is represented when the vehicle state converges to a stable state at the moment of a prediction terminal, and the vehicle speed of the prediction terminal is the same as the optimal reference speed of a sub-area i of the intersection; v_i ^*(N_p|k_t) For the speed of the ith vehicle in the formation system when the ith vehicle runs at a constant speed, namely the optimization of an intersection subregion iA reference speed; h is₁ ^p(N_p|k_t) An expected displacement of a predicted terminal for a first vehicle in the formation system; h is_i ^p(N_p|k_t) Designing the displacement of a vehicle prediction terminal to be consistent with the expected displacement of the vehicle i and the first vehicle for the displacement of the prediction terminal of the ith vehicle in the formation system; d_i,jThe expected distance between two adjacent vehicles in the formation system; t is_i ^p(N_p|k_t) Predicting a terminal torque for a vehicle of an ith vehicle in the formation system; h is_i(V_i ^p(N_p|k_t) Is the vehicle equilibrium torque when the ith vehicle in the convoy system is running at a constant speed.

Finally, it is to be noted that: the above examples are only for illustrating the technical solutions of the present invention, and are not limited thereto. It will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention; such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A multi-vehicle cooperative control method under the environment of urban non-signal control multi-intersection is characterized by comprising the following steps:

step 1: constructing a control system comprising a multi-intersection region integrated regulation and control unit, an intersection subregion regulation and control unit and a vehicle-mounted dynamic decision and management control unit;

step 2: the integrated regulation and control unit of the multiple intersection sub-areas is used for providing comprehensive evaluation for the macroscopic accumulated traffic running state among the intersection sub-areas regulated and controlled by the intersection sub-area regulation and control unit according to the macroscopic traffic network running state prediction state information among the intersection sub-areas and the short-time traffic network boundary control state prediction information of the intersection sub-areas, and performing balanced control on the traffic flow states inside the intersection sub-areas and among the boundaries;

and step 3: the intersection subregion regulating and controlling unit is used for constructing a vehicle formation system according to the traffic network state information and the vehicle state information of each vehicle which is about to reach the intersection subregion control boundary, performing traffic sequence distribution, path distribution and speed distribution and sending decision information to the vehicle-mounted dynamic decision and management control unit;

and 4, step 4: the vehicle-mounted dynamic decision and management control unit is provided with a vehicle head-vehicle optimization controller and a follower vehicle-mounted controller of the vehicle formation system;

the head vehicle optimization controller receives decision information of each intersection subregion regulation and control unit, receives state information of following vehicles in the system, receives state information of tail vehicles of a front vehicle formation system, performs multi-objective collaborative optimization control, and sends optimization control decision information to all following vehicles in the system;

the follow vehicle-mounted controller receives the optimization control decision information sent by the head vehicle of the system, receives the state information sent by the front vehicle, and automatically follows the head vehicle and the front vehicle of the system to adjust the longitudinal and transverse motion state.

2. The multi-vehicle cooperative control method in the urban non-signal-control multi-intersection environment according to claim 1, wherein the step 2 is implemented as follows:

step 2.1: the integrated regulation and control unit of the multi-intersection region obtains the road condition information of the current road through the real-time interactive communication between the vehicle and the road side unit;

step 2.2: the integrated regulation and control unit of the multiple intersection regions calculates and obtains the prediction state information of the operation situation of the macroscopic traffic network among the subareas of each intersection and the prediction state information of the boundary control state of the short-time traffic network of each intersection subarea;

step 2.3: the integrated regulation and control unit of the multiple intersection area quantitatively determines the weight of the two information in the step 2.2, and the optimal solution of the traffic running state of each intersection subregion is solved;

step 2.4: the integrated regulation and control unit of the multiple intersection regions establishes the comprehensive evaluation of the macroscopic accumulated traffic running state among the sub-regions of each intersection;

step 2.5: if the comprehensive evaluation of the macroscopic accumulated traffic running state does not meet the road congestion condition, returning to the step 2.2 for circular monitoring calculation, and if so, entering the step 2.6;

step 2.6: carrying out traffic flow state balance control inside each intersection subregion and between boundaries;

step 2.7: and (5) changing the real-time running path and the average running speed of each traffic flow according to the step 2.6, and reducing the traffic jam time.

3. The multi-vehicle cooperative control method in the urban non-signal-control multi-intersection environment according to claim 1, wherein the step 3 is implemented as follows:

step 3.1: acquiring the distance between adjacent multi-vehicle systems and the length of a queuing queue according to the vehicle condition information of the adjacent multi-vehicle systems in the same lane at the current moment through information interaction between vehicles;

step 3.2: combining the traffic flow states regulated and controlled by the integrated regulation and control unit in the multi-intersection region through information interaction between vehicles and roadside units, and forming a plurality of vehicle formation systems by the multi-vehicle systems in the current traffic network according to the distance between the adjacent multi-vehicle systems and the queuing length acquired in the step 3.1;

step 3.3: and distributing the dynamic passing time, the passing sequence and the expected average driving speed of all the vehicle formation systems entering the sub-areas of the intersections according to the principle of passing first when entering the sub-areas of the intersections.

4. The multi-vehicle cooperative control method in the urban non-signal-control multi-intersection environment according to claim 1, wherein the step 4 is implemented as follows:

step 4.1: through information interaction between each vehicle formation system and each intersection subregion regulation and control unit, a head vehicle in each vehicle formation system receives the passing sequence and speed distributed by each intersection subregion regulation and control unit;

step 4.2: designing a head vehicle optimization controller in each vehicle formation system for the purposes of safety, energy conservation, stability and comfort of a head vehicle, carrying out local expected trajectory planning, and outputting expected control quantity;

step 4.3: and sending the output expected control quantity to vehicle-mounted controllers of other following vehicles except the head vehicle in the vehicle formation system according to the communication topological structure of the multi-vehicle formation system, wherein the other following vehicles run along with the head vehicle in the vehicle formation system.

5. The method of claim 2, wherein the multi-vehicle cooperative control system is used in an urban non-signal-controlled multi-intersection environment,

1) in the step 2.2, the method for calculating and obtaining the operation situation prediction state information of the macro traffic network among the sub-areas of each intersection by the multi-intersection area integrated regulation and control unit is as follows:

step 2.2.1A: assuming that each intersection subregion has a well-defined traffic flow model MFD, each time step k in subregion i is used_tTo establish a network traffic flow Q (k)_t) And total number of vehicles N_i(k_t) The relationship of (a) to (b) is as follows:

Q(k_t)＝G_i(N_i(k_t))

in the formula, G_i(. represents a function of MFD, N_i(k_t)＝n_i(k_t)+n_ij(k_t)，n_i(k_t) Indicating the number of local vehicles, n_ij(k_t) Representing the number of vehicles from sub-area i to destination sub-area j, then for k_tAt +1 there are: n is a radical of_i(k_t+1)＝n_i(k_t+1)+n_ij(k_t+1)，

In the formula, T represents a sampling interval, d_ii(k_t) And d_ij(k_t) Respectively correspond to at k_tThe traffic demand arriving at sub-area i and the traffic demand arriving at sub-area j, p at that moment_ii(k_t) Is represented at k_tProbability of one-step transition of traffic flow leaving sub-area i at any moment, p_ij(k_t) Is represented at k_tProbability of one-step transition of traffic flow from sub-area i to sub-area j at a time, Q_ji(k_t) Is at k_tTraffic flow, Q, entering sub-area i from sub-area j at a time_i,O(k_t) Is at k_tTraffic flow, Q, leaving sub-area i at any time_ij(k_t) Is at k_tThe traffic flow from the sub-area i to the sub-area j at the moment, and n represents the number of the sub-areas of the intersection;

step 2.2.2A: establishing a macroscopic traffic network operation situation prediction state information objective function by minimizing the flow passing through the intersection subregion:

in the formula, N_i,c(k_t+1) denotes k_tThe critical vehicle number of the maximum spatial average traffic flow at +1 time;

step 2.2.3A: establishing an objective function constraint framework:

a vehicle number constraint

In the formula, Q_i,I(k_t) Represents k_tThe total inflow of the sub-area i at the moment;

traffic flow constraint

In the formula, Q_i(k_t+1) denotes k_tSub-area i network traffic flow at +1 moment, Q_i,O(k_t+1) denotes k_tTraffic flow leaving sub-area i at +1 moment, Q_ji(k_t+1) denotes k_tThe traffic flow entering the sub-area i from the sub-area j at the moment of + 1;

link maximum traffic flow constraint

0≤Q_ij(k_t+1)≤m_ijρ_i,ij

In the formula, Q_ij(k_t+1) denotes k_t+1 flow of traffic from sub-area i into sub-area j at time m_ijRepresents the number of links, ρ, of the sub-region i and its neighbor sub-region j_i,ijIs the average saturated traffic flow on the link between sub-area i and its neighboring sub-area j;

2) in the step 2.2, the method for calculating and obtaining the control state prediction information of the short-time traffic network boundary of each intersection subregion by the multi-intersection subregion integrated regulation and control unit is as follows:

step 2.2.1B: assuming a sampling cycle time c_cEqual to the sampling time step k of all intersections_tThen the total number of vehicles in the link (i, j) of the intersection sub-area can be updated by the following conservation equation:

n_i,j(k_t+1)＝n_i,j(k_t)+(α_ij,I(k_t)-α_ij,O(k_t)·c_c

in the formula, n_i,j(k_t+1) denotes k_tTotal number of vehicles entering link (i, j) at time + 1; n is_i,j(k_t) Represents k_tTotal number of vehicles entering link (i, j) at time; alpha is alpha_ij,I(k_t) Represents k_tThe traffic entering the link (i, j) at a time is the sum of the traffic flowing out of its upstream link, α_ij,O(k_t) Represents k_tThe traffic leaving link (i, j) at a time is the sum of the incoming traffic of its downstream links; wherein the content of the first and second substances,

α_ij,O(k_t)＝min(β_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_c,q_ij,o(k_t)/c_c+α_ij,I(k_t),β_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_c)，β_ij,o(k_t)·μ_ij·g_ij,o(k_t)/c_cindicating capacity at the intersection, q_ij,o(k_t)/c_c+α_ij,I(k_t) Indicating the number of waiting and arriving vehicles, beta_ij,o(k_t)(C_j,o-n_j,o(k_t))/c_cRepresenting available space of downstream road section, beta_ij,o(k_t) Represents the relative fraction, μ, of flow out of the subregion j_ijRepresents the saturated flow leaving the link (i, j), g_ij,o(k_t) Represents the feasible time length of the traffic flow to j in the link (i, j), q_ij,o(k_t) Representing the traffic density, C, of the outgoing links (i, j)_j,oRepresenting the capacity of the downstream link, n, in terms of the number of vehicles_j,o(k_t) Is the number of vehicles in link (i, j);

α_ij,I(k)＝β_ij,I·α_ij,O(k_t)，β_ij,Iis the relative fraction of flow to sub-region j;

step 2.2.2B: establishing a short-term traffic network boundary control state prediction information objective function by minimizing the flow passing through the sub-area of the intersection and the difference between the total traffic flow and the optimal traffic flow:

in the formula, alpha_ij,O(k_t+1) denotes k_t(ii) traffic flow leaving link (i, j) at time +1, where α_lRepresents the lowest flow, Q_ij(k_t+1) denotes k_tThe traffic flow from sub-area i into sub-area j at time + 1.

6. The multi-vehicle cooperative control method under the environment of urban non-signal-control multi-intersection according to claim 2 or 5, characterized in that in the step 2.3, the method for solving the optimal solution of the traffic running state of each intersection subregion is as follows:

establishing a cost function:

Q^*(k_t)＝min(w_iQ_w(k_t)+w_jQ_h(k_t))

Q^*(k_t) Represents k_tOptimal solution, w, of traffic running flow state in sub-area of each intersection at any moment_iThe method is to consider the operation situation of the sub-region i to predict the state information weight coefficient, Q_w(k_t) Is k_tForecasting state information value w of macro traffic network operation situation of sub-area i at moment_jConsidering the weight coefficient, Q, of the information predicted by the boundary control state of the short-term traffic network of the sub-area j_h(k_t) Is k_tPredicting information values of the short-time traffic network boundary control states of the sub-area j at the moment; w is a_i、w_jIs an empirical tuning value.

7. The multi-vehicle cooperative control method in the urban non-signal-control multi-intersection environment according to claim 2, wherein in step 2.4, the method for making the macroscopic accumulated traffic running state comprehensive evaluation between the sub-areas of each intersection is as follows:

step 2.4.1: establishing a macroscopical accumulated traffic running state comprehensive evaluation parameter index set by adopting a fuzzy comprehensive evaluation method

Q^*Represents the optimal solution of the traffic running flow state of each intersection subregion,

representing average speed, L queue length, ATTR travel-time ratio; establishing a macroscopic accumulated traffic running state comprehensive evaluation parameter evaluation set B ═ B₁,b₂,b₃,b₄]，b₁Indicating congestion, b₂Indicating light congestion, b₃Indicates substantial absence of flow, b₄Indicating unblocked;

step 2.4.2: fuzzifying the traffic running state, and performing fuzzy reasoning according to rules under different traffic states, wherein the single-factor fuzzy discrimination matrix is as follows:

wherein the content of the first and second substances,

respectively represent the optimal membership function values of the traffic operation flow state,

respectively represent the average velocity

The value of the membership degree function of (c),

respectively representing membership function values of the queuing lengths L,

respectively representing membership function values of the travel time ratio ATTR;

step 2.4.3: weighting is carried out on the comprehensive evaluation parameter indexes of each traffic running state, and a final evaluation result is obtained as follows:

wherein B is a fuzzy comprehensive evaluation value which falls into which evaluation parameter evaluation set of the macroscopic cumulative traffic state comprehensive evaluation parametersWithin the range, which traffic condition belongs to; omega₁,ω₂,ω₃,ω₄The weight coefficients of the four traffic running state comprehensive evaluation parameter indexes are respectively,

representing a fuzzy synthesis operation.

8. The method of claim 2, wherein in step 2.6, the method of performing the traffic flow state equalization control within each intersection sub-area and between boundaries is as follows:

step 2.6.1: establishing a cost function

In the formula, Q_i(k_t) Represents k_tInformation on the actual traffic state of sub-area i of the intersection at that moment, Q_j(k_t) Represents k_tInformation on the actual traffic state of sub-area j of the intersection at that moment, Q^*(k_t) Represents k_tThe optimal solution of the traffic running flow state of each intersection subregion at any moment;

step 2.6.2: building a constraint framework

0≤Q_i(k_t)≤Q_max

0≤Q_j(k_t)≤Q_max

In the formula, Q_maxRepresents a maximum value of the traffic state parameter;

and under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

9. The method for cooperative control of multiple vehicles at an urban non-signal controlled multi-intersection environment according to claim 3, wherein in step 3.3, the method for distributing the expected average driving speed of all vehicle formation systems is as follows:

step 3.3.1: establishing a cost function

In the formula, V_i(k_t) Represents k_tActual average velocity, V, of sub-area i of the time intersection_i ^*(k_t) Represents k_tThe optimal reference speed of the intersection sub-area i at the moment,

and

respectively represents k_tThe average speed of the intersection sub-areas i and j under the time optimal traffic flow conditions,

is a weight coefficient;

step 3.3.2: building a constraint framework

0≤V_i(k_t)≤V_max

In the formula, V_maxIn order to set the maximum speed limit of the road,

and

multiple vehicle system for intersection region respectivelyMinimum and maximum average acceleration of;

and under a constraint frame, the cost function is enabled to obtain the minimum value, namely the problem is solved.

10. The method for multi-vehicle cooperative control in the urban non-signal-control multi-intersection environment according to claim 4, wherein in the step 4.2, the method for designing the head-vehicle optimization controller in each vehicle formation system is as follows:

step 4.2.1: the nonlinear dynamical equation of the vehicle is established as follows:

in the formula, S_i(k_t) And v_i(k_t) Respectively displacement and velocity, T, of the ith vehicle_q,i(k_t) Is the actual torque of the vehicle, u_i(k_t) To the desired torque, i₀Is a mechanical transmission ratio, eta_m,iFor mechanical efficiency of the transmission system, m_iAs mass of the vehicle, C_D,iIs the drag coefficient of the vehicles in the queue, rho is the air density, A_iIs the frontal area of the vehicle, g is the acceleration of gravity, r_w,iIs the rolling radius of the wheel, f is the rolling resistance coefficient, tau_iIs the longitudinal power system time lag coefficient, alpha is the road gradient, delta k_tIs a discrete time step;

discretizing the kinetic equation by an Euler method, and rewriting the nonlinear kinetic equation into:

x_i(k_t+1)＝φ_i(x_i(k_t))+ψ_iu_i(k_t)，i∈N

wherein x is_i(k_t) N is the state quantity of the vehicle, the number of the vehicles in the queue,

G_i＝m_ig

constructing an output of y for each vehicle in the vehicle fleet_i(k_t)＝[S_i(k_t),v_i(k_t)]^T＝γx_i(k_t) Wherein

Order:

X(k_t)＝[x₁ ^T(k_t),x₂ ^T(k_t),...,x_N ^T(k_t)]^T，

U(k_t)＝[u₁(k_t),u₂(k_t),...,u_N(k_t)]^T，

Y(k_t)＝[y₁ ^T(k_t),y₂ ^T(k_t),...,y_N ^T(k_t)]^T，

Φ(X(k_t))＝[φ₁(x₁)^T,φ₂(x₂)^T,...,φ_N(x_N)^T]^T，

Ψ＝diag{ψ₁,ψ₂,...,ψ_N}，

the equation of state for the vehicle fleet as a whole is described as:

X(k_t)＝Φ(X(k_t))+Ψ·U(k_t)

Y(k_t)＝ΓX(k_t)

in the formula (I), the compound is shown in the specification,

step 4.2.2: defining a sub-prediction optimization problem on each vehicle in a vehicle formation system, carrying out optimization solution on each sub-prediction optimization problem by using information of adjacent vehicles and head vehicles, and constructing a distributed controller for each vehicle by adopting a distributed model prediction control method, wherein the method comprises the following steps:

a, establishing an objective function

J＝min(f₁(k_t)+f₂(k_t)+f₃(k_t)+f₄(k_t))

Wherein, the following cost function f of the ith vehicle and the head vehicle in the vehicle formation system₁(k_t) The following were used:

f₁(k_t)＝ω_s1(h_i(k_t)-h₁(k_t)-d_i,1)²+ω_v1(Z_i(k_t)-V₁ ^*(k_t))²

ω_s1is the distance error weight coefficient of the ith vehicle and the head vehicle in the formation system, h_i(k_t) For the predicted position of the ith vehicle in the formation system, h₁(k_t) For the predicted position of the head car, d_i,1Is the desired distance, ω, between the ith vehicle and the head vehicle in the formation system_v1Is the speed error weight coefficient, Z, of the ith vehicle and the head vehicle in the formation system_i(k_t) For the predicted speed, V, of the i-th vehicle in the formation system₁ ^*(k_t) To predict the speed of the head car in the formation system,

wherein, the following cost function f of the ith vehicle and the front vehicle j in the formation system₂(k_t) The following were used:

f₂(k_t)＝ω_s,i(h_i(k_t)-h_j(k_t)-d_i,j)²+ω_v,j(Z_i(k_t)-Z_j(k_t))²

ω_s,iis the distance error weight coefficient of the ith vehicle and the preceding vehicle j in the formation system, h_j(k_t) For the predicted position of the vehicle j ahead of the i-th vehicle in the formation system, d_i,jIs the desired distance, ω, between the ith and preceding vehicle j in the formation system_v,jIs a velocity error weight coefficient, Z, of a preceding vehicle j in a formation system_j(k_t) The predicted speed of the front vehicle j in the formation system;

wherein the stability cost function f₃(k_t) The following were used:

α_i1for the predicted output trajectory and assumed output trajectory series error weight coefficients for each vehicle in the formation system,

in order to predict the output trajectory,

is a hypothetical output trajectory;

wherein, a comfort and energy-saving cost function f₄(k_t) The following were used:

f₄(k_t)＝α_i2(u_i(k_t)-T_i(k_t))²

α_i2for comfort weight coefficient, T, of each vehicle in the formation system_i(k_t) The torque of each vehicle in the formation system is the torque when the vehicle runs at a constant speed;

b, establishing a constraint framework

v_min≤v_i(k_t)≤v_max

T_min≤u_i(k_t)≤T_max

V_i ^p(N_p|k_t)＝V_i ^*(N_p|k_t)

h_i ^p(N_p|k_t)＝h₁ ^p(N_p|k_t)-(i-1)d_i,j

T_i ^p(N_p|k_t)＝h_i(V_i ^p(N_p|k_t))

v_i(k_t) Is the actual speed, v, of the i-th vehicle in the formation system_min、v_maxMaximum and minimum speed, T, of the ith vehicle in the formation system, respectively_min、T_maxMaximum and minimum torque, V, respectively, for the ith vehicle in the formation system_i ^p(N_p|k_t) The vehicle speed of the ith vehicle in the formation system is the vehicle speed when the vehicle state converges to the steady state at the time of predicting the terminal, the vehicle speed is the same as the optimal reference speed of a sub-area i of the intersection, and V is the optimal reference speed of the sub-area i of the intersection_i ^*(N_p|k_t) Is the speed h of the ith vehicle in the formation system when the ith vehicle runs at a constant speed_i ^p(N_p|k_t) For the displacement of the prediction terminal of the ith vehicle in the formation system, the displacement of the vehicle prediction terminal is designed to be consistent with the expected displacement of the ith vehicle and the head vehicle, h₁ ^p(N_p|k_t) For predicting the expected displacement of the terminal of the head car in the formation system, d_i,jFor a desired inter-vehicle distance, T, between two adjacent vehicles in a formation system_i ^p(N_p|k_t) Predicting a terminal torque, h, for the vehicle of the i-th vehicle in the formation system_i(V_i ^p(N_p|k_t) Is the vehicle equilibrium torque when the ith vehicle in the convoy system is running at a constant speed.

Images