CN112477846B - Intelligent networking electric automobile queue control method giving consideration to stability and energy conservation - Google Patents

Abstract

The invention provides an intelligent networking electric automobile queue control method giving consideration to stability and energy conservation, which comprises the following steps: constructing a vehicle motor model and a vehicle distance strategy of an electric vehicle queue; constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function; constructing a linear state feedback controller for the vehicles by using a feedback control strategy, and simultaneously analyzing the control gain range of each vehicle according to the queue stability; and optimizing control gain by applying a selected genetic algorithm, solving by taking the target function as a fitness function to obtain the optimal control gain, and realizing the goals of queue stability and energy conservation. The invention provides an intelligent networking electric automobile queue control method considering stability and energy conservation, aiming at intelligent networking electric automobiles, and by optimizing the working point of a motor, the energy consumption efficiency of each automobile is obviously improved, and the integral energy efficiency of the automobile queue is improved.

Description

Intelligent networking electric automobile queue control method giving consideration to stability and energy conservation

Technical Field

An intelligent networking electric automobile queue control method giving consideration to both stability and energy conservation optimizes energy consumption of a vehicle queue on the premise of ensuring the queue stability of the vehicle queue, and belongs to the field of economic control of intelligent networking electric automobile queues.

Background

The intelligent vehicle queue control technology is that intelligent vehicles running on a road form a queue, the vehicles in the queue can obtain information of surrounding environment and the road through an environment sensing technology, vehicle state information sharing among vehicle queues is realized through a vehicle-to-outside information exchange (V2X) wireless communication technology, single vehicle node control in the queue is completed on the basis, and a cooperative control technology for stable running of the vehicle queues is achieved on the whole.

For an intelligent vehicle queue, the queue stability and the queue energy-saving effect are two crucial performance indexes, and the research enthusiasm of a large number of scholars at home and abroad is attracted. Most of the available literature and patents focus on the cohort stability of the study cohort. However, few have explored vehicle fleet stability while considering energy efficiency. When the queue encounters interference, the energy efficiency of the queue can be influenced by the changed speed, and the common vehicle queue control method is mostly based on that vehicles in the queue run along with a pilot vehicle to track the pilot vehicle speed in real time, so that the method can cause the phenomenon that the number of unnecessary acceleration and deceleration times of the following vehicles in the vehicle queue is increased, and the energy consumption of the whole queue is increased. In addition, most of the current domestic and foreign researches on vehicle queues are directed at fuel vehicles and hybrid electric vehicles, few researches are carried out on applying the energy-saving control of the vehicle queues to pure electric vehicles, and the electric vehicles have the natural advantages of high motor response speed and easiness in torque accurate regulation and control of a drive-by-wire system, and the motors feed back braking to recover energy, so that the energy-saving optimization effect of the electric vehicle queues is expected to be superior to that of the fuel vehicles and the hybrid electric vehicles.

Disclosure of Invention

The purpose of the invention is as follows: the invention aims to solve the problem of road traffic jam, improve the traffic efficiency, improve the driving safety and improve the vehicle economy.

The technical scheme is as follows: in order to solve the above problems, the present invention provides the following technical solutions:

an intelligent networking electric automobile queue control method giving consideration to both stability and energy conservation is characterized in that: the method comprises the following steps:

step 1: constructing a vehicle motor model and a vehicle distance strategy of an electric vehicle queue;

step 2: constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function;

and step 3: constructing a linear state feedback controller for the vehicles by using a feedback control strategy, and simultaneously analyzing the control gain range of each vehicle according to the queue stability;

and 4, step 4: and optimizing control gain by applying a selected genetic algorithm, solving by taking the target function as a fitness function to obtain the optimal control gain, and realizing the goals of queue stability and energy conservation.

In step 1, the vehicle motor model is as follows:

wherein, P_mIs the motor power, w_mIs the motor speed, T_mIn order to be the torque of the motor,

in order to achieve the driving efficiency of the motor,

generating efficiency for the motor;

in step 1, the inter-vehicle distance strategy is: during the vehicle queue running, the queue control aims at keeping the distance between adjacent vehicles at a desired inter-vehicle distance and keeping the speed of the following vehicle consistent with that of the pilot vehicle, namely

Wherein v is_i(t)The speed of the ith following vehicle at time t; v. of_0(t)The speed of the piloted vehicle at time t; p is a radical of_i-1(t)The motor power of the i-1 th following vehicle at the time t; p is a radical of_i(t)The motor power of the ith following vehicle at the time t; n is the number of following vehicles; d_i-1,iRepresenting the desired inter-vehicle distance between node i-1 and node i, d_i-1,i>0；

The definition of the expected vehicle distance satisfies the following operational expression:

the geometry of the train being determined by the desired vehicle spacing d_i-1,iThe specific selection decision of (2);

for constant distance type of vehicle spacing, d_i-1,iIs a fixed constant, i.e.

d_i-1,i＝d₀,i∈n (14)

Wherein d is₀Is a constant greater than zero;

for constant-headway type vehicle spacing, d_i-1,iIs a linear function related to the vehicle speed, i.e.

d_i-1,i＝t_hv_i+d₀,i∈n (15)

Wherein, t_hThe time interval of following the vehicle;

for non-linear distance type inter-vehicle distance, d_i-1,iIs a non-linear function related to the vehicle speed, i.e.

d_i-1,i＝g(v_i),i∈n (16)

The goals of vehicle fleet control are:

d_min≤d_i-1,i≤d_max (17)

wherein d is_minRepresenting the minimum allowable inter-vehicle distance, d_maxRepresenting the maximum allowable car-to-car distance.

In the step 1, a vehicle motor model and an inter-vehicle distance strategy of an electric vehicle queue are constructed according to a selected workshop information flow topological structure, wherein the selected workshop information flow topological structure is in a front vehicle-navigator following mode, namely a set of one navigator and n following vehicles is defined as the queue, and each vehicle i can receive information of the navigator and adjacent vehicles.

The plant information flow topology is described by a directed graph g, wherein the directed graph g represents the information flow topology between the following vehicles in the queue_n+1Representing the information flow between the following vehicle and the pilot vehicle:

directed graph g_n＝{v_n,ε_n,A_nIn which the vertices are collectedv_n1,2, …, n, and an edge set epsilon_n＝v_n×v_nAdjacent matrix A_n＝[a_ij]∈R^n×nWherein a is_ijIs a non-negative contiguous element;

directed graph g_n+1＝{v_n+1,ε_n+1,A_n+1V, node collection v_n+1＝{v₀,v₁,…,v_n}, edge union ε_n+1＝v_n+1×v_n+1Adjacent matrix A_n＝[a_ij]∈R^(n+1)×(n+1)；

The topology of the information flow in the queue is represented by g_nAnd g_n+1Characterized as the following three matrices:

(1) adjacency matrix A_n；

(2) Laplace matrix

(3) Traction matrix

Wherein A is_n＝[a_ij]∈R^n×nIs defined as follows

Wherein, a_ij1, namely representing that the ith vehicle can receive the state information from the jth vehicle; suppose this g_nIn which no self-loop is present, i.e. all of a_ij0, i belongs to n; the information set of all vehicle state information that the vehicle i can receive is defined as:

N_i＝{j|a_ij＝1}

the set represents a collection of all vehicles that can acquire information, in the following vehicles, by vehicle i through V2X communication or radar detection; the degree of entry of vehicle i is defined as

At the same time, the in-degree matrix is defined as

Corresponds to g_nLaplacian matrix of

Is defined as

And corresponds to g_n+1Traction matrix P ∈ R_n×nFor describing the case where the following vehicle acquires the relevant pilot vehicle state information, the following is defined as follows:

wherein if {0, i }. epsilon. [ epsilon ], [ epsilon ]_n+1Then p is_i0, otherwise p_iWhen the ith vehicle can acquire the relevant state information of the pilot vehicle, the value p is represented as 0_i1 is ═ 1; at this time, the ith vehicle is also called as being directly towed by the pilot vehicle.

Under the workshop information flow topological structure of the front vehicle-navigator following mode, the information set of the vehicle i is

The adjacency matrix and the Laplace matrix of the topological structure of the workshop information flow are

Piloting vehicle reachable set of vehicle i

(P_i)_PLF＝{0}

Therefore, the traction matrix representation of the plant information flow topology is as follows

And 2, constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function:

dynamic model of single vehicle

The vehicle queue consists of n +1 vehicle groups, wherein the pilot vehicle is manually controlled and is numbered as 0, and the number of the whole vehicle queue is from 0 to n;

the longitudinal dynamics model of the ith vehicle is:

wherein, V_iIs the speed of the ith vehicle in the queue; δ is a mass transfer factor; u. of_iIs an input to a vehicle system; g is the acceleration of gravity; f is the rolling resistance coefficient; theta_iIs a road slope angle; ρ is the air density; a. the_iIs the cross-sectional area of the vehicle; c_d,iIs the air resistance coefficient; m is_iIs the mass of the ith vehicle;

vehicle queue model

The intelligent vehicle fleet is represented as:

P_p＝[P₀,P₁,…P_n]^T,V_p＝[V₀,V₁,…V_n]^T,F_p＝[F₀,F₁,…F_n]^T,F_f,p＝[F_f,0,F_f,1,…F_f,n]^T,F_g,p＝[F_g,0,F_g,1,…F_g,n]^T,F_A,p＝[F_A,0,F_A,1,…F_A,n]^T；

the dynamics of the cohort are as follows:

wherein, P_p、V_p、F_p、F_f,p、F_g,pAnd F_A,pIs the set of position, speed, resultant longitudinal tire force, rolling resistance, ramp inclination resistance and air resistance of each vehicle; m is_pThe mass of the vehicle, being a homogenous vehicle train, has m_p＝m_iThe mass of each vehicle is the same;

the mathematical structure of the queue is described as follows:

wherein D is_dIs the desired vehicle spacing; l is_iIs the length of the vehicle;

is the error between the desired pitch and the actual pitch;

is a pair of

Derivation is carried out; p_iThe position of the ith vehicle in the queue;

vehicle energy consumption model

The ith vehicle in the queue from the initial time t₀The energy consumption of the destination reached for time t is:

wherein f is_iFor the energy consumption of the ith vehicle:

wherein the content of the first and second substances,

is the efficiency of the motor; p_m,iThe motor power required by the ith vehicle;

motor power P required by ith vehicle_m,iComprises the following steps:

wherein, P_f,iIs the power consumed by the vehicle to overcome the grade resistance, P_f,i＝m_igfcos(θ_i)V_i；P_g,iIs the power consumed by the vehicle to overcome the grade resistance, P_g,i＝m_igfcos(θ_i)V_i；P_A,iIs the power consumed by the vehicle to overcome the air resistance;

P_a,iis the power required for the acceleration of the vehicle,

η_iis the mechanical transmission efficiency.

In step 3, the transfer function of the linear state feedback controller is:

controlling gain k₁,k₂,k₃,k₄The ranges of (A) are as follows:

step 4, in the control gain range obtained in the step 3, optimizing the control gain by using a genetic algorithm, and solving the optimal control gain, wherein the steps are as follows:

the optimization objective is to minimize the vehicle energy consumption J, which is expressed as follows:

to control the gain k_jIs an independent variable, j is 1,2,3, 4; taking the energy consumption J of the vehicle queue as a fitness function:

s.t.

M_min≤M_i≤M_max,

k₃ ²+2k₁k₃≥0,

k₄ ²+2k₁k₄-2k₁-2k₃≥0,

k₁＞0,k₂＞0,k₃＞0,k₄＞0,

k₁＜20,k₂＜20,k₃＜20,k₄＜20

wherein M is_minAnd M_maxRepresents a minimum value and a maximum value of the motor torque; j. the design is a square_i(k_j) Is the ith vehicle at control gain k_jEnergy consumption under control.

The basic operation process of the genetic algorithm is as follows:

a) initializing, namely setting an evolution algebra counter t to be 0, setting a maximum evolution algebra G, and randomly generating N individuals as an initial population P (0);

b) evaluating individuals, namely calculating the fitness of each individual in the population P (t);

c) selecting operation, namely acting the selecting operator on the group;

d) performing cross operation, namely acting a cross operator on the group;

e) mutation operation, namely acting mutation operators on the population; the group P (t) is subjected to selection, intersection and mutation operation to obtain a next generation group P (t + 1);

f) and (5) judging a termination condition, namely if t is G, outputting the individual with the maximum fitness obtained in the evolution process as an optimal solution, and terminating the calculation.

Has the advantages that: compared with the prior art, the invention has the advantages that:

1. innovatively proposes to improve vehicle economy by optimizing control gain;

2. gain is optimized and controlled through a genetic algorithm, so that the calculation efficiency is higher, and the method is suitable for real vehicle scenes;

3. the intelligent networking electric automobile queue control with both stability and energy conservation is realized;

4. road gradient conditions are considered in the queue economy, and the queue economy better accords with actual road situations.

Drawings

FIG. 1 is a schematic diagram of a workshop information flow topological structure selecting a front vehicle-navigator following mode according to an embodiment of the invention;

FIG. 2 is a flow chart of a genetic algorithm according to an embodiment of the present invention;

fig. 3 is a control method for an intelligent networked electric vehicle queue with both stability and energy saving in accordance with an embodiment of the present invention.

Detailed Description

The invention will be further described with reference to the accompanying drawings in which:

fig. 1 is a flow chart of the intelligent networked electric vehicle queue as an object for selecting a front vehicle-navigator following mode by a vehicle queue workshop information flow topological structure according to fig. 3.

Firstly, establishing a basic model of a vehicle queue: and selecting a workshop information flow topological structure, and constructing a vehicle motor model and a vehicle distance strategy according to the selected workshop information flow topological structure.

The workshop information flow topological structure selects a front vehicle-navigator following mode, namely, a set of one navigator vehicle and n following vehicles is defined as a queue, and each vehicle i can receive information of the navigator vehicle and adjacent vehicles. The workshop information flow topological structure is described by algebraic graph theory, a set comprising a pilot vehicle and n following vehicles is defined as a queue, and the information flow topological structure between the following vehicles in the queue is represented by a directed graph g. Directed graph g_n＝{v_n,ε_n,A_nV, in which the vertex sets v _n1,2, …, n, and an edge set epsilon_n＝v_n×v_nAdjacent matrix A_n＝[a_ij]∈R^n×nWherein a is_ijAre non-negative contiguous elements. In the directed graph, each vertex refers to a following vehicle, if the vertex i refers to the ith following vehicle, j, i represents a directed edge from the following vehicle j to the vehicle i, which indicates that the following vehicle i can obtain state information from the vehicle j, and the information flow between the following vehicle and the pilot vehicle can be used as an augmented directed graph g_n+1＝{v_n+1,ε_n+1,A_n+1Denotes, where the node collection v_n+1＝{v₀,v₁,…,v_n}, edge union ε_n+1＝v_n+1×v_n+1Adjacent matrix A_n＝[a_ij]∈R^(n+1)×(n+1). The topology of the information flow in the queue can be represented by g_nAnd g_n+1Characterized as the following three matrices:

(1) adjacency matrix A_n；

(2) Laplace matrix

(3) Traction matrix

Wherein A is_n＝[a_ij]∈R^n×nIs defined as follows

Wherein, a_ijThat is, 1 represents that the ith vehicle can receive the status information from the jth vehicle. Suppose this g_nIn which no self-loop is present, i.e. all of a_ij0, i ∈ n. The information set of all vehicle state information that the vehicle i can receive is defined as:

N_i＝{j|a_ij＝1} (2)

the set represents a collection of all vehicles that can acquire information by vehicle i through V2X communication or radar detection in the following vehicle. The degree of entry of vehicle i is defined as

At the same time, the in-degree matrix is defined as

Corresponds to g_nLaplacian matrix of

Can be defined as

And corresponds to g_n+1Traction matrix of

The following vehicle state information acquisition method can be used for describing the situation that the following vehicle acquires the related pilot vehicle state information, and is defined as follows:

wherein if {0, i }. epsilon. [ epsilon ], [ epsilon ]_n+1Then p is_i0, otherwise p_iWhen the ith vehicle can acquire the relevant state information of the pilot vehicle, the value p is represented as 0_i1 is ═ 1; at this time, the ith vehicle is also called as being directly towed by the pilot vehicle. The purpose of using the traction matrix is to represent the information interaction between the vehicles in the queue and the pilot vehicle in the form of a matrix.

Two related underlying concepts are given below:

(1) directed path: if there is a set of directed edge sequence sets (i)₁,i₂),(i₂,i₃),…(i_k-1,i_k) Wherein

Then the directed edge sequence set is called as a slave node i₁To i_kA directed path.

(2) And (3) directed spanning tree, if the tree consisting of the edge sets in the graph can connect all nodes in the graph, the tree consisting of the edge sets is called a directed spanning tree.

If the edges are combined_n+1When there is a case that a subset constitutes a directed spanning tree, the directed graph g_n+1Including a directed spanning tree. g_n+1The minimum of the directional spanning trees which take the pilot vehicle as the starting point is the basis for meeting the system control. In short, there should be at least one directed path from the lead vehicle to any one of the following vehicles in the queue. This means that any one following vehicle can directly or indirectly obtain relevant state information from the lead vehicle. Under a preceding vehicle-navigator following mode (PLF) vehicle information flow topology, the information set of vehicle i is

Therefore, the adjacency matrix and the Laplace matrix of the topology of the plant information flow are

Piloting vehicle reachable set of vehicle i

(P_i)_PLF＝{0},i＝n (9)

Therefore, the traction matrix representation of the plant information flow topology is as follows

Because the vehicle queue is a pure electric vehicle, a motor model of the vehicle is constructed. The invention adopts a permanent magnet synchronous motor as a driving/braking motor of a vehicle, and the output power model is as follows:

wherein P is_mIs the motor power, w_mIs the motor speed, T_mIn order to be the torque of the motor,

in order to achieve the driving efficiency of the motor,

the generating efficiency of the motor is obtained.

The inter-vehicle distance strategy is used for describing the geographical position relationship between adjacent vehicles in the queue, and generally, in the running process of the vehicle queue, the queue control aims to maintain the distance between the adjacent vehicles to be the expected inter-vehicle distance and keep the speed of the following vehicle consistent with that of the pilot vehicle, namely, the inter-vehicle distance strategy is used for describing the geographical position relationship between the adjacent vehicles in the queue, namely, the distance between the following vehicle and the pilot vehicle is maintained to be the expected inter-vehicle distance

Wherein d is_i-1,i>0 represents the desired inter-vehicle distance between node i-1 and node i. The definition of the expected vehicle distance satisfies the following formula

The geometry of the train being determined by the desired vehicle spacing d_i-1,iIs determined by the specific selection of (a). For constant distance type of vehicle spacing, d_i-1,iIs a fixed constant, i.e.

d_i-1,i＝d₀,i∈n (14)

Wherein d is₀Is a constant greater than zero. For constant-headway type vehicle spacing, d_i-1,iIs a linear function related to the vehicle speed, i.e.

d_i-1,i＝t_hv_i+d₀,i∈n (15)

Wherein, t_hThe time interval for following the car. Since the platoon driving safety is closely related to the speed of the vehicles in the platoon, the higher the vehicle speed, the larger the minimum safety inter-vehicle distance required, and for non-linear distance inter-vehicle distances, d_i-1,iIs a non-linear function related to the vehicle speed, i.e.

d_i-1,i＝g(v_i),i∈n (16)

The goal of vehicle fleet control is not to require the following vehicle to be consistent with the lead vehicle speed, but rather to allow the distance between adjacent vehicles to fluctuate within a safe distance for better vehicle fleet economy, i.e., vehicle fleet economy

d_min≤d_i-1,i≤d_max (17)

Wherein d is_minRepresenting the minimum allowable inter-vehicle distance, d_maxRepresenting the maximum allowable car-to-car distance.

Secondly, constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function:

(1) dynamic model of single vehicle

The vehicle train consists of n +1 cars, where the lead car is manually controlled, numbered 0, and the entire vehicle train is numbered from 0 to n. Because all vehicles are pure electric vehicles and the motor has the characteristic of quick response of torque, the influence of inertial lag on subsequent vehicles is not considered.

The complete vehicle dynamics system is a nonlinear system, and the running state of the vehicle can be described more accurately by adopting an accurate vehicle model, however, the accurate model also increases the computational analysis burden of the vehicle system, and the required analysis result is difficult to quickly feed back, so the invention ignores the influence caused by the sliding of the vehicle tires and the vertical and transverse motion of the vehicle, and mainly studies the longitudinal motion of the vehicle. In which the vehicle longitudinal dynamics take into account air resistance, tire force, rolling resistance and tilting resistance. Will P_iAnd V_iIndicated as the position and speed of the ith vehicle in the queue. The longitudinal dynamics model of the ith vehicle can then be written as

Wherein m is_iRepresenting vehicle mass, F_i,F_f,i,F_g,iAnd F_A,iRespectively the resultant longitudinal tire force, rolling resistance, ramp inclination resistance and air resistance, F_f,i,F_g,iAnd F_A,iCan be represented as:

wherein g is the acceleration of gravity; f is the rolling resistance coefficient; theta_iIs the road grade angle, and in practical cases, the road grade angle theta can be estimated by using some advanced positioning technology, such as Global Positioning System (GPS) and Geographic Information System (GIS)_i(ii) a ρ is the air density; a. the_iIs the cross-sectional area of the vehicle; c_d,iIs the coefficient of air resistance, the magnitude of which depends on the inter-vehicle distance, C_d,iDistance d from vehicle_iThe relationship of (a) to (b) is as follows:

wherein, C_dIs the air resistance coefficient of the pilot vehicle,

is gamma₁And gamma₂Is the air resistance reduction of the parameter. Can be combined with F_iExpressed as:

F_i＝δm_iu_i (21)

where δ is the quality conversion factor, u_iIs an input to the vehicle system. The aforementioned longitudinal dynamics model can then be converted into:

(2) vehicle queue model

In this embodiment, the intelligent vehicle queue is composed of n +1 vehicles traveling from left to right, and is denoted by P_p＝[P₀,P₁,…P_n]^T,V_p＝[V₀,V₁,…V_n]^T,F_p＝[F₀,F₁,…F_n]^T,F_f,p＝[F_f,0,F_f,1,…F_f,n]^T,F_g,p＝[F_g,0,F_g,1,…F_g,n]^T,F_A,p＝[F_A,0,F_A,1,…F_A,n]^TThe kinetics of the cohort are as follows:

wherein m is_pIs the mass of the vehicle, and is a uniform vehicle queue, so that m is_p＝m_iThe mass of each vehicle is the same. The mathematical structure of the queue can be described as follows:

wherein D_dIs the desired vehicle spacing, L_iIs the length of the vehicle and is,

is the error between the desired pitch and the actual pitch.

(3) Vehicle energy consumption model

The energy consumption of a vehicle depends on a number of factors, including motor torque, speed, efficiency, and gear ratio. The driving force equation of the electric vehicle can be calculated by the following equation.

Wherein M is_t,iIs the combined torque of the motor torques transmitted to the tyre, M_iIs the motor output torque, j_iIs the speed ratio of the main reducer, eta_iIs the mechanical transmission efficiency, r_iIs the wheel radius.

Motor power P required by ith vehicle_m,iCan be expressed as:

wherein P is_f,i＝m_igfcos(θ_i)V_iIs the power consumed by the vehicle to overcome the rolling resistance, P_g,i＝m_igfcos(θ_i)V_iIs the power consumed by the vehicle to overcome the grade resistance,

is the power consumed by the vehicle to overcome the air resistance,

is the power required for vehicle acceleration.

Then, the energy consumption of the ith car can be expressed as:

wherein

It is the efficiency of the motor, which is determined by the rotational speed and output torque of the motor. From the above, it can be seen that the ith vehicle in the queue has started from the initial time t₀The energy consumption of the destination reached for time t is:

and thirdly, constructing a linear state feedback controller for the vehicles by using a feedback control strategy, and analyzing to obtain a control gain range of each vehicle according to the queue stability. This step proposes a distributed controller for vehicle queues that ensures the stability of its queue, i.e. that the transient errors do not spread gradually backwards along the queue. In addition, the value range of the control gain can be obtained according to calculation, and the control gain is optimized by taking improvement of the energy-saving efficiency of the vehicle queue as a guide.

The controller acting on the ith vehicle in the queue needs to apply all other vehicle information that it can receive to realize the control of the vehicle i and thus accomplish the macro driving goal of the whole queue. Wherein, all other vehicle information that the vehicle i can receive is defined as:

I_i＝N_i∪P_i (27)

the design idea of the distributed controller is as follows:

wherein k is_ij,pAnd k is_ij,vReferring to the controller gain, the above equation is a general form of a linear controller.

The goal of the feedback control fleet control is to track the speed of the leading vehicle following the vehicle to maintain state consistency while satisfying the conditions in the following equation, which means that the vehicle spacing error and speed error in the fleet can be reduced to zero.

In order to achieve energy saving and simplify the queue control, the linear state feedback controller described above is used for each following vehicle.

Wherein k is_j(j ═ 1,2,3,4) is the control gain;

is the desired distance between the lead vehicle and the ith following vehicle;

is the sum of all vehicle lengths between the lead vehicle and the ith following vehicle.

Here, a new tracking error ζ is defined for the ith following vehicle_iComprises the following steps:

therefore, equation (30) can be rewritten as:

by combining equations (22) and (31), the following equation can be obtained:

wherein a is₀Is the acceleration of the leading vehicle, defines x_i＝[ζ_iζ_i]^T,ψ_i＝[a₀ cos(θ_i) sin(θ_i) V_i ²]^TAs status and other information of the queue system. Therefore, equation (15) can be rewritten as

By combining equations (32) and (34), the model of the vehicle fleet may be represented by equation (35):

wherein x (t) ═ x₁(t),x₂(t),…,x_n(t)]^TRepresenting the state vector of each vehicle in the queue, [ psi (t) ]₁(t),ψ₂(t),…,ψ_n(t)]^TRepresenting other information vectors, y (t) being control outputs

Q₁，Q₂Is a coefficient matrix represented by (36):

wherein I_s(s ═ 2 or n) is an identity matrix of order s, the symbols

Represents the kronecker product, and K_aAnd K_bThen is a matrix representing the in-queue vehicle control gains:

in the above formula

The erroneous transfer of transient errors may lead to a breakdown of the vehicle train or to a collision accident, and therefore the train stability of the vehicle must be satisfied. The PLF workshop information flow topological structure is adopted, namely the inter-vehicle distance error of adjacent vehicles in a queue is measured by a radar, the state information of a pilot vehicle is transmitted by a V2X wireless communication technology, a vehicle controller is further designed according to the error information, and the acceleration of the vehicle is used as control input.

From equation (23) we can derive:

by combining equations (29) and (38), the following can be obtained:

by using the laplace transform, equation (39) can be transformed into:

based on the transfer function, for any range of values for which w >0 the control gain can be obtained, the following formula must be satisfied:

wherein

Because γ ≧ 0, if σ ≧ 0, then there is

I.e. the queue stability of the vehicle queue can be guaranteed and from this the control gain (k) can be derived₁,k₂,k₃,k₄) The ranges of (A) are as follows:

fourthly, optimizing control gain by applying a selected genetic algorithm, solving to obtain optimal control gain, and realizing the goals of queue stability and energy conservation;

genetic algorithm is a search algorithm used for solving the optimal result in computational mathematics, and for an optimization problem, the genetic algorithm is generally realized by providing a large number of candidate solutions through a computer and enabling the target to be optimized to better evolve so as to obtain the optimal solution. The genetic algorithm mainly uses the phenomena of heredity, mutation, natural selection, hybridization and the like contained in the biological population in the evolution process.

The basic operation process of the genetic algorithm is as follows, and the specific flow is shown in fig. 2:

a) and (3) initializing, namely setting an evolution algebra counter t to be 0, setting a maximum evolution algebra G, and randomly generating N individuals as an initial population P (0).

b) And (4) evaluating individuals, namely calculating the fitness of each individual in the population P (t).

c) And (4) selecting operation, namely acting a selection operator on the group. The purpose of selection is to inherit optimized individuals directly to the next generation or to generate new individuals by pairwise crossing and then to inherit them to the next generation. The selection operation is based on fitness evaluation of individuals in the population.

d) And (4) performing cross operation, namely applying a cross operator to the group. What plays a core role in genetic algorithms is the crossover operator.

e) And (4) mutation operation, namely acting mutation operators on the population. I.e., to vary the gene values at certain loci of the individual strings in the population. And (t) obtaining a next generation group P (t +1) after selection, crossing and mutation operations of the group P (t).

f) And (5) judging a termination condition, namely if t is G, outputting the individual with the maximum fitness obtained in the evolution process as an optimal solution, and terminating the calculation.

Since genetic algorithms start from a collection of problem solutions, rather than from a single solution, it is a great difference between genetic algorithms and traditional optimization algorithms. The traditional optimization algorithm is used for solving the optimal solution from a single initial value in an iteration mode, and the local optimal solution is easy to enter in a wrong mode. The genetic algorithm starts to search from an initial population, has large coverage and is beneficial to global preference.

Whether traveling on a flat road or on a slope, the vehicles in the fleet must track the speed of the lead vehicle. On a grade-changing highway, the speed of the lead vehicle may change with the grade of the downhill slope of the road, which may lead to unnecessary acceleration and deceleration of the following vehicle, and the unnecessary deceleration behavior may increase the energy consumption of the vehicle. Different control gains may cause the effect of following vehicle tracking in the queue to be different. Therefore, in this step, we will reduce the energy consumption of the vehicle by optimizing the control gain in the above-mentioned feedback controller.

The optimization goal of this step is to minimize vehicle energy consumption, which can be expressed as follows:

in the present application, Genetic Algorithms (GA) are selected for optimizing control gains, the implementation of GA is generally by computer simulation, and for an optimization problem, the objective to be optimized can be better evolved by abstract representation of a large number of candidate solutions. Here to control the gain k_j(J ═ 1,2,3,4) as an argument, and energy consumption J of the vehicle fleet as a fitness function:

wherein M is_minAnd M_maxRepresenting the minimum and maximum values of the motor torque. J. the design is a square_i(k_j) Is the ith vehicle at control gain k_j(j-1, 2,3,4) under control. In addition, since the controller is designed based on the distance error and the speed error between the vehicle and the adjacent front vehicle and the pilot vehicle in the queue, the control gain must be set to satisfy k_j>0 (j-1, 2,3,4), and to ensure that the control input does not exceed the vehicle acceleration limit, the control gain should also be less than 20.

Claims (7)

1. An intelligent networking electric automobile queue control method giving consideration to both stability and energy conservation is characterized in that: the method comprises the following steps:

step 1: constructing a vehicle motor model and a vehicle distance strategy of an electric vehicle queue;

step 2: constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function;

and step 3: constructing a linear state feedback controller for the vehicles by using a feedback control strategy, and simultaneously analyzing the control gain range of each vehicle according to the queue stability;

and 4, step 4: optimizing control gain by applying a selected genetic algorithm, taking a target function as a fitness function, solving to obtain optimal control gain, and realizing the goals of queue stability and energy conservation;

in step 1, the vehicle motor model is as follows:

wherein, P_mIs the motor power, w_mIs the motor speed, T_mIn order to be the torque of the motor,

in order to achieve the driving efficiency of the motor,

generating efficiency for the motor;

in step 1, the inter-vehicle distance strategy is: during the vehicle queue running, the queue control aims at keeping the distance between adjacent vehicles at a desired inter-vehicle distance and keeping the speed of the following vehicle consistent with that of the pilot vehicle, namely

Wherein v is_i(t)The speed of the ith following vehicle at time t; v. of_0(t)The speed of the piloted vehicle at time t; p is a radical of_i-1(t)The motor power of the i-1 th following vehicle at the time t; p is a radical of_i(t)The motor power of the ith following vehicle at the time t; n is the number of following vehicles; d_i-1,iRepresenting the desired inter-vehicle distance between node i-1 and node i, d_i-1,i>0；

The definition of the expected vehicle distance satisfies the following operational expression:

the geometry of the train being determined by the desired vehicle spacing d_i-1,iThe specific selection decision of (2);

for constant distance type of vehicle spacing, d_i-1,iIs a fixed constant, i.e.

d_i-1,i＝d₀,i∈n

Wherein d is₀Is a constant greater than zero;

for constant-headway type vehicle spacing, d_i-1,iIs a linear function related to the vehicle speed, i.e.

d_i-1,i＝t_hv_i+d₀,i∈n

Wherein, t_hWhen the car is followingDistance; v. of_iThe speed of the ith vehicle is the speed of the ith vehicle;

for non-linear distance type inter-vehicle distance, d_i-1,iIs a non-linear function related to the vehicle speed, i.e.

d_i-1,i＝g(v_i),i∈n

The goals of vehicle fleet control are:

d_min≤d_i-1,i≤d_max

wherein d is_minRepresenting the minimum allowable inter-vehicle distance, d_maxRepresenting the maximum allowable car-to-car distance.

2. The intelligent networked electric automobile queue control method with both stability and energy conservation as claimed in claim 1, wherein: in the step 1, a vehicle motor model and an inter-vehicle distance strategy of an electric vehicle queue are constructed according to a selected workshop information flow topological structure, wherein the selected workshop information flow topological structure is in a front vehicle-navigator following mode, namely a set of one navigator and n following vehicles is defined as the queue, and each vehicle i can receive information of the navigator and adjacent vehicles.

3. The intelligent networked electric automobile queue control method with both stability and energy conservation as claimed in claim 2, characterized in that: the plant information flow topology is described by a directed graph g, wherein the directed graph g represents the information flow topology between the following vehicles in the queue_n+1Representing the information flow between the following vehicle and the pilot vehicle:

directed graph g_n＝{v_n,ε_n,A_nV, in which the vertex sets v_n1,2, …, n, and an edge set epsilon_n＝v_n×v_nAdjacent matrix A_n＝[a_ij]∈R^n×nWherein a is_ijIs a non-negative contiguous element;

directed graph g_n+1＝{v_n+1，ε_n+1，A_n+1V, node collection v_n+1＝{v₀,v₁,…,v_n}, edge union ε_n+1＝v_n+1×v_n+1Adjacent matrix A_n＝[a_ij]∈R^(n+1)×(n+1)；

The topology of the information flow in the queue is represented by g_nAnd g_n+1Characterized as the following three matrices:

(1) adjacency matrix A_n；

(2) Laplace matrix

(3) Traction matrix

Wherein A is_n＝[a_ij]∈R^n×nIs defined as follows

Wherein, a_ij1, namely representing that the ith vehicle can receive the state information from the jth vehicle; suppose this g_nIn which no self-loop is present, i.e. all of a_ij0, i belongs to n; the information set of all vehicle state information that the vehicle i can receive is defined as:

N_i＝{j|a_ij＝1}

the set represents a collection of all vehicles that can acquire information, in the following vehicles, by vehicle i through V2X communication or radar detection; the degree of entry of vehicle i is defined as

At the same time, the in-degree matrix is defined as

Corresponds to g_nLaplacian matrix of

Is defined as

L＝D_n-A_n

And corresponds to g_n+1Traction matrix P ∈ R_n×nFor describing the case where the following vehicle acquires the relevant pilot vehicle state information, the following is defined as follows:

wherein if {0, i }. epsilon. [ epsilon ], [ epsilon ]_n+1Then p is_i0, otherwise p_iWhen the ith vehicle can acquire the relevant state information of the pilot vehicle, the value p is represented as 0_i1 is ═ 1; at the moment, the ith vehicle is directly pulled by the pilot vehicle;

under the workshop information flow topological structure of the front vehicle-navigator following mode, the information set of the vehicle i is

The adjacency matrix and the Laplace matrix of the workshop information flow topological structure are as follows:

the piloted vehicle reachable set for vehicle i is:

(P_i)_PLF＝{0}

therefore, the traction matrix of the plant information flow topology is represented as follows:

4. the intelligent networked electric automobile queue control method with both stability and energy conservation as claimed in claim 1, wherein: and 2, constructing a single vehicle dynamics model and a vehicle queue model to complete the establishment of a vehicle energy consumption objective function:

dynamic model of single vehicle

The vehicle queue consists of n +1 vehicle groups, wherein the pilot vehicle is manually controlled and is numbered as 0, and the number of the whole vehicle queue is from 0 to n;

the longitudinal dynamics model of the ith vehicle is:

wherein, V_iIs the speed of the ith vehicle in the queue; δ is a mass transfer factor; u. of_iIs an input to a vehicle system; g is the acceleration of gravity; f is the rolling resistance coefficient; theta_iIs a road slope angle; ρ is the air density; a. the_iIs the cross-sectional area of the vehicle; c_d,iIs the air resistance coefficient; m is_iIs the mass of the ith vehicle;

vehicle queue model

The intelligent vehicle fleet is represented as:

P_p＝[P₀,P₁,…P_n]^T,V_p＝[V₀,V₁,…V_n]^T,F_p＝[F₀,F₁,…F_n]^T,F_f,p＝[F_f,0,F_f,1,…F_f,n]^T,F_g,p＝[F_g,0,F_g,1,…F_g,n]^T,F_A,p＝[F_A,0,F_A,1,…F_A,n]^T；

the dynamics of the cohort are as follows:

wherein, P_p、V_p、F_p、F_f,p、F_g,pAnd F_A,pIs a collection of position, speed, resultant longitudinal tire force, rolling resistance, ramp inclination resistance, and air resistance of each vehicle; m is_pThe mass of the vehicle, being a homogenous vehicle train, has m_p＝m_iThe mass of each vehicle is the same;

the mathematical structure of the queue is described as follows:

wherein D is_dIs the desired vehicle spacing; l is_iIs the length of the vehicle;

is the error between the desired pitch and the actual pitch;

is a pair of

Derivation is carried out; p_iThe position of the ith vehicle in the queue;

vehicle energy consumption model

The ith vehicle in the queue from the initial time t₀The energy consumption of the destination reached for time t is:

wherein f is_iFor the energy consumption of the ith vehicle:

wherein the content of the first and second substances,

is the efficiency of the motor; p_m,iThe motor power required by the ith vehicle;

motor power P required by ith vehicle_m,iComprises the following steps:

wherein, P_f,iIs the power consumed by the vehicle to overcome the grade resistance, P_f,i＝m_igfcos(θ_i)V_i；P_g,iIs the power consumed by the vehicle to overcome the grade resistance, P_g,i＝m_igfcos(θ_i)V_i；P_A,iIs the power consumed by the vehicle to overcome the air resistance;

P_a,iis the power required for the acceleration of the vehicle,

η_iis the mechanical transmission efficiency.

5. The intelligent networked electric automobile queue control method considering stability and energy conservation as claimed in claim 4, wherein: in step 3, the transfer function of the linear state feedback controller is:

controlling gain k₁,k₂,k₃,k₄The ranges of (A) are as follows:

6. the intelligent networked electric automobile queue control method with both stability and energy conservation as claimed in claim 5, wherein: step 4, in the control gain range obtained in the step 3, optimizing the control gain by using a genetic algorithm, and solving the optimal control gain, wherein the steps are as follows:

the optimization objective is to minimize the vehicle energy consumption J, which is expressed as follows:

to control the gain k_jIs an independent variable, j is 1,2,3, 4; taking the energy consumption J of the vehicle queue as a fitness function:

min.

s.t.

M_min≤M_i≤M_max,

k₃ ²+2k₁k₃≥0,

k₄ ²+2k₁k₄-2k₁-2k₃≥0,

k₁＞0,k₂＞0,k₃＞0,k₄＞0,

k₁＜20,k₂＜20,k₃＜20,k₄＜20

wherein M is_minAnd M_maxRepresents a minimum value and a maximum value of the motor torque; j. the design is a square_i(k_j) Is the ith vehicle at control gain k_jEnergy consumption under control.

7. The intelligent networked electric automobile queue control method considering stability and energy conservation as claimed in claim 6, wherein: the basic operation process of the genetic algorithm is as follows:

a) initializing, namely setting an evolution algebra counter t to be 0, setting a maximum evolution algebra G, and randomly generating N individuals as an initial population P (0);

b) evaluating individuals, namely calculating the fitness of each individual in the population P (t);

c) selecting operation, namely acting the selecting operator on the group;

d) performing cross operation, namely acting a cross operator on the group;

e) mutation operation, namely acting mutation operators on the population; the group P (t) is subjected to selection, intersection and mutation operation to obtain a next generation group P (t + 1);

f) and (5) judging a termination condition, namely if t is G, outputting the individual with the maximum fitness obtained in the evolution process as an optimal solution, and terminating the calculation.

Images