A cooperative ecological driving method for intelligent connected electric vehicle fleets based on perturbation observation

Abstract

This invention relates to a cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation. The method first establishes longitudinal dynamics and power models for the electric vehicles based on their own parameters. Then, it constructs a fleet reference trajectory optimization problem based on model predictive control (MRC) and obtains the optimal solution using an improved PSO algorithm. Finally, it utilizes active disturbance rejection control (ADRC) technology, designing a tracking differentiator, an extended state observer, and a nonlinear state error feedback law to estimate the lumped disturbance and send it to the lead vehicle. Compared to other methods, this invention considers parameter uncertainties and external disturbances, primarily utilizing model predictive control and ADRC techniques to ensure safe operation of the electric vehicle fleet while reducing the overall energy consumption of the vehicle platoon.

Description

Technical Field

[0001] This invention relates to a cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation, belonging to the field of ecological driving technology. Background Art

[0002] With the increasing severity of energy and environmental issues, electric vehicles have experienced unprecedented development. Intelligent connected electric vehicles will become a new breakthrough in solving problems such as traffic congestion, traffic safety, and traffic pollution. Intelligent connected vehicles are based on advanced vehicle equipment, combined with modern communication and internet technologies, to achieve intelligent information exchange and sharing between the vehicle and other systems. They possess functions such as complex environment perception, intelligent decision-making, and collaborative control, achieving safe, efficient, comfortable, and energy-saving driving. Their advantage lies in their ability to form effective formations, maintaining a small longitudinal distance between vehicles, thereby improving road capacity, reducing fuel consumption, and reducing exhaust emissions.

[0003] In vehicle platoon control, internal stability and chord stability are two important performance indicators. However, when there are large disturbances in front of the vehicles, even platoons that meet platoon stability standards may still experience collisions, necessitating the design of better controllers to ensure vehicle safety. Therefore, to further improve the stability and energy economy of electric vehicle fleets, this application proposes a cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation. Summary of the Invention

[0004] To overcome the shortcomings of existing research, this invention provides a cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation. Compared with other methods, it takes into account parameter uncertainty and external disturbances, and mainly utilizes model predictive control and active disturbance rejection control technology to ensure the safe driving of electric vehicle fleets while reducing the overall energy consumption of the vehicle platoon.

[0005] The specific steps of a cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation are as follows:

[0006] Step 1: Modeling

[0007] Step 1.1: Establish the longitudinal dynamics model of electric vehicle i based on the vehicle's own parameters as follows:

[0008]

[0009] Where, p i v i u i These represent the position, speed, and control input of electric vehicle i, respectively; T d (t) is the lumped disturbance of vehicle i, where u i (t) and Td (t) can be expressed as:

[0010]

[0011]

[0012] In the formula F i (t) and F i，r (t) represents the traction force and total resistance, m i F is the mass of vehicle i. i，r (t) can be expressed as follows:

[0013]

[0014] In the formula C d (d i (t)), A i，v μ i and θ(p) i (t) represents the distance d between the vehicle and the vehicle. i (t) is related to the drag coefficient, frontal area of ​​the vehicle, rolling drag coefficient, and location-dependent road gradient, where ρ and g are the air density and gravitational acceleration, respectively.

[0015] Vehicle spacing d i (t) is defined as:

[0016] d i (t)=p i-1 (t)-p i (t)-L i

[0017] Among them, L i It is the length of vehicle i.

[0018] In addition, the drag coefficient C d It can be represented as:

[0019]

[0020] Among them, C d (d i (t) is the rated drag coefficient of vehicle i, and c1 and c2 are fitting parameters.

[0021] Step 1.2, Electric Vehicle Power Model

[0022] The output power P of electric vehicle i i，e (t) can be represented as:

[0023]

[0024] Where η i，t and ηi，r These are the mechanical transmission efficiency and energy recovery efficiency of electric vehicles, respectively.

[0025] Step 2: Select the control target

[0026] Desired position of vehicle i and speed The definition is as follows:

[0027]

[0028] Where d represents the desired distance between adjacent vehicles.

[0029] The position error and speed error of following vehicle i are defined as follows:

[0030]

[0031] Step 2.1, Internal Stability

[0032] When the acceleration of the lead vehicle is 0, all following vehicles remain at the desired position and eventually reach the same speed as the lead vehicle, that is:

[0033]

[0034] Where i = 1, 2, ..., N

[0035] Step 2.2, the string stability condition is:

[0036]

[0037] Step 2.3, Energy Economy

[0038] The total energy consumption from time t to t+T is expressed as:

[0039]

[0040] Step 3: Optimization of the Fleet Reference Trajectory Based on Model Predictive Control

[0041] Step 3.1, the band-stop function with compensation factor is expressed as:

[0042]

[0043] in, day z l , z u ∈R + ...

[0044] Step 3.2: Constructing the fleet reference trajectory optimization problem based on model predictive control.

[0045] The discrete longitudinal dynamics model of the lead vehicle is expressed as:

[0046]

[0047] In the formula, Δt is the discrete time interval.

[0048] Let x 0，ref =[p 0，ref v 0，ref ] T Further acquisition:

[0049] x 0，ref (k+1)=A0x 0，ref (k)+B0u 0，ref (k)

[0050] in,

[0051]

[0052] Assuming all following vehicles remain at their desired positions within the vehicle platoon, the reference position, reference speed, and reference control input of following vehicle i are expressed as follows:

[0053]

[0054] Let N P and N c Let the prediction range and control range be the two distinct values, respectively. The cost function for vehicle i within the prediction range can be expressed as:

[0055]

[0056] Where ω1 and ω2 are weight parameters, v l and v u It is the expected speed range of the vehicle platoon [v] l v u The upper and lower bounds of the vehicle queue. Therefore, the cost function for the vehicle queue within the prediction range can be obtained as:

[0057]

[0058] Given the predictors and cost function of the vehicle platoon, the platoon reference trajectory problem can be formulated as follows:

[0059]

[0060] and satisfy

[0061]

[0062] Where F i，ref (n|k) and P i，ref，e(n|k) can be represented as:

[0063]

[0064]

[0065] Step 4: Design of Active Disturbance Rejection Controller

[0066] Specifically, the active disturbance rejection controller consists of three parts: a tracking differentiator, an extended state observer, and a nonlinear state error feedback law.

[0067] First, the tracking differentiator is specifically expressed as follows:

[0068]

[0069] In the formula, r0 and h0 are the velocity factor and the filter factor, respectively; fhan is the fastest synthesis function, and its expression is as follows:

[0070] fhan(x1,x2,r0,h0)=-r0[a / d0-sign(a)]s2-r0sign(a),

[0071] in,

[0072]

[0073] Then, the perturbation T of the electric vehicle is applied using ESO. i，d Estimate (t). Let T i，d (t) is a state variable w of electric vehicle i. i The specific expression for the extended state observer of electric vehicle i is given below: (t).

[0074]

[0075] In the formula, e i The estimation bias of the position of electric vehicle i; For system state variable p i The predicted value; For system state variable v i The predicted value; For perturbation T i，d The estimated value of (t) B i β1, β2, and β3 are compensation factors; β1, β2, and β3 are observer parameters; fal is a nonlinear function, whose specific expression is:

[0076]

[0077] Finally, the nonlinear state error feedback rate u for electric vehicle i-trajectory tracking is given. ias follows:

[0078]

[0079] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0080] Compared with other control methods, this invention takes into account the lumped resistance of the vehicle and designs an active disturbance rejection controller, which feeds forward the observed values ​​to the controller, thereby further improving the system's disturbance rejection capability.

[0081] The method of the present invention can effectively improve the fuel economy of vehicle platoons and ensure the internal stability and chordal stability of vehicle platoons. Attached Figure Description

[0082] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0083] Figure 1 This is a flowchart illustrating the implementation of the cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation, as described in this invention.

[0084] Figure 2 This defines the predictor variables within the prediction range of this invention;

[0085] Figure 3 This is a structural diagram of the active disturbance rejection control technology used in this invention;

[0086] Figure 4 This is a simulation diagram of the driving section of the present invention;

[0087] Figure 5 The figures show the simulation results of position, spacing error, speed, and control input under the control method of this invention. Detailed Implementation

[0088] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0089] A flowchart for implementing a cooperative ecological driving method for intelligent connected electric vehicle fleets based on perturbation observation. Figure 1 The specific steps are as follows:

[0090] Step 1: Modeling

[0091] Step 1.1: Establish the longitudinal dynamics model of electric vehicle i based on the vehicle's own parameters as follows:

[0092]

[0093] Where, p i v i u i These represent the position, speed, and control input of electric vehicle i, respectively; T d (t) is the lumped disturbance of vehicle i.

[0094] Where u i (t) and T d (t) can be expressed as:

[0095]

[0096]

[0097] In the formula F i (t) and F i，r (t) represents the traction force and total resistance, m i F is the mass of vehicle i. i，r (t) can be expressed as follows:

[0098]

[0099] In the above formula, C d (d i (t)), A i，v μ i and θ(p) i (t) represents the distance d between the vehicle and the vehicle. i (t) is related to the drag coefficient, frontal area, rolling drag coefficient, and location-dependent road gradient. ρ and g are air density and gravitational acceleration, respectively.

[0100] Vehicle spacing d i (t) is defined as:

[0101] d i (t)=p i-1 (t)-p i (t)-L i

[0102] Where: L i It is the length of vehicle i.

[0103] In addition, the drag coefficient C d It can be represented as:

[0104]

[0105] Among them, C d (d i (t) is the rated drag coefficient of vehicle i, and C1 and C2 are fitting parameters.

[0106] Step 1.2, Electric Vehicle Power Model

[0107] The output power P of electric vehicle i i，e (t) can be represented as:

[0108]

[0109] Where, η i，t and η i，r These are the mechanical transmission efficiency and energy recovery efficiency of electric vehicles, respectively.

[0110] Step 2: Select the control target

[0111] To improve road capacity, a constant spacing strategy will be used to adjust vehicle spacing. This requires information about the lead vehicle to ensure the chordal stability of the vehicle platoon. The desired position of vehicle i. and speed The definition is as follows:

[0112]

[0113] Where d represents the desired distance between adjacent vehicles.

[0114] The position error and speed error of following vehicle i are defined as follows:

[0115]

[0116] Step 2.1, Internal Stability

[0117] When the acceleration of the lead vehicle is 0, all following vehicles remain at the desired position and eventually reach the same speed as the lead vehicle, that is:

[0118]

[0119] Where i = 1, 2, ..., N

[0120] Step 2.2, the string stability condition is defined as:

[0121] For any ε > 0, there exists a δ > 0 that satisfies the following equation:

[0122]

[0123] Step 2.3, Energy Economy

[0124] The total energy consumption from time t to t+T is expressed as:

[0125]

[0126] Step 3: Optimization of the Fleet Reference Trajectory Based on Model Predictive Control

[0127] This step aims to optimize the platoon's reference trajectory based on Particle Swarm Optimization (MPC) to reduce the overall energy consumption of the vehicle platoon. To this end, we first assume the vehicle platoon is a rigid body, with all following vehicles positioned at the desired location within the platoon. Second, we construct a platoon reference trajectory optimization problem based on MPC, considering passenger comfort requirements, platoon driving efficiency, and the physical constraints of the vehicle platoon. Finally, we employ an improved Particle Swarm Optimization (PSO) algorithm to quickly solve the optimization problem.

[0128] Step 3.1, the band-stop function with compensation factor is expressed as:

[0129]

[0130] in, and z l , z u ∈R + Here, represents the upper and lower bounds of the interval, and c is the compensation factor. In the optimization problem, when the band-stop function... As part of the cost function, if the parameters With proper setting of c, z can be constrained to the interval [z]. l , z u ].

[0131] Step 3.2: Constructing the fleet reference trajectory optimization problem based on model predictive control.

[0132] The discrete longitudinal dynamics model of the lead vehicle is expressed as:

[0133]

[0134] In the formula, Δt is the discrete time interval.

[0135] Let x 0，ref =[p 0，ref v 0，ref ] T Further acquisition:

[0136] x 0，ref (k+1)=A o x o，ref (k)+B0u 0，ref (k)

[0137] in,

[0138]

[0139] Assuming all following vehicles remain at their desired positions within the vehicle platoon, the reference position, reference speed, and reference control input of following vehicle i are expressed as follows:

[0140]

[0141] Let N P and N c These are the prediction range and the control range, respectively. Before constructing the convoy reference trajectory optimization problem at time k, some predictor variables within the prediction range are defined as follows: Figure 2 As shown, the cost function for vehicle i within the prediction range can be expressed as:

[0142]

[0143] Where ω1 and ω2 are weight parameters, v l and v u It is the expected speed range of the vehicle platoon [v] l v u The upper and lower bounds of the vehicle queue. Therefore, the cost function for the vehicle queue within the prediction range can be obtained as:

[0144]

[0145] Given the predictors and cost function of the vehicle platoon, the platoon reference trajectory problem can be formulated as follows:

[0146]

[0147] and satisfy

[0148]

[0149] Where F i，ref (n|k) and P i，ref，e (n|k) can be represented as:

[0150]

[0151]

[0152] In the constructed vehicle reference trajectory optimization problem, u 0，ref (:|k)=[u0(0|k),u0(1|k),...,u0(N P [-1|k)] represents the unknown predictive control variable to be optimized. Let F(k) represent the optimal predictive control variable that minimizes the cost function F(k). An improved PSO algorithm is used to quickly obtain the optimal solution.

[0153] Step 4: Design of Active Disturbance Rejection Controller

[0154] Specifically, the active disturbance rejection controller consists of three parts: a tracking differentiator, an extended state observer, and a nonlinear state error feedback law. The specific structure diagram of this controller is shown below. Figure 3 As shown.

[0155] First, the tracking differentiator is specifically expressed as follows:

[0156]

[0157] In the formula, r0 and h0 are the velocity factor and the filter factor, respectively; fhan is the fastest synthesis function, and its expression is as follows:

[0158] fhan(x1,x2,r0,h0)=-r0[a / d0-sign(a)]s2-r0sign(a),

[0159] in,

[0160]

[0161] Then, the perturbation T of the electric vehicle is applied using ESO. i，d Estimate (t). Let T i，d (t) is a state variable w of electric vehicle i. i The extended state equation of electric vehicle i is as follows: (t)

[0162]

[0163] In the above formula, u i (t)=F i (t) / m i ,y i (t) represents the output of electric vehicle i.

[0164] The specific expression for the extended state observer of electric vehicle i is given below:

[0165]

[0166] In the formula, e i The estimation bias of the position of electric vehicle i; For system state variable p i The predicted value; For system state variable v i The predicted value; For perturbation T i，d The estimated value of (t) B iβ1, β2, and β3 are compensation factors; β1, β2, and β3 are observer parameters; fal is a nonlinear function, whose specific expression is:

[0167]

[0168] Finally, the nonlinear state error feedback rate u for electric vehicle i-trajectory tracking is given. i as follows:

[0169]

[0170] This embodiment considers a vehicle convoy with one lead vehicle and five following vehicles. The vehicle parameters are as follows: τ = 0.38, 0.45, 0.2, 0.3, 0.25, and 0.40; m = 7182, 7200, 7100, 7300, 7150, and 7250; L = 10, 11, 9.8, 10.5, 10.2, and 9.6; μ = ... i The values ​​are 0.003, 0.0032, 0.0031, 0.0033, 0.0032, 0.0031, and A, respectively. i，v The values ​​are 10, 10, 10, 10, 10, 10, C respectively. i，d The values ​​are 0.8, 0.83, 0.81, 0.82, 0.8, and 0.81, respectively. The gravitational acceleration and air density are set to 9.8 and 1.29, respectively; the discrete time interval Δt = 0.001 s; and the prediction range and control range in the optimization problem are set to N. P =40 and N c =0.1; the upper and lower bounds of the convoy acceleration are set to u respectively. 0，min = -2m / s 2 and u 0，max =2m / s 2 The weight parameters are set to ω2 = 1.6, ω2 = 1; the tracking controller parameters are r0 = 100000000, h0 = 0.001; the observer parameters are set to β1 = 100, β2 = 300, β3 = 1000; the nonlinear state error feedback rate parameters are set to β1 = 5.5, β2 = 0.2, α1 = 0.75, α2 = 1.5, δ = 0.02; the driving road segment used in the simulation is as follows: Figure 4 As shown.

[0171] The initial position of the rear vehicle is set to 0, and the initial positions of the other following vehicles and the lead vehicle are calculated according to the position error formula in step 1. The changes in position, spacing error, speed, and control input for different vehicles are as follows: Figure 5 As shown in the figure. It can be seen that under the control of the ecological driving method proposed in this invention, the control input gradually decreases to zero over time. Figure 5(d) As time progresses, when the acceleration of the lead vehicle reaches 0, all following vehicles will remain in the desired position, such as... Figure 5 (a), and ultimately reach with Figure 5 (c) The lead vehicle has the same speed, and the spacing error between vehicles is always 0, which means that the internal stability and chord stability of the vehicle platoon are satisfied. At the same time, the change in vehicle energy consumption does not exceed expectations.

[0172] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and these variations still fall within the protection scope of the present invention.

Claims

1. A cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation, characterized in that: Includes the following steps: Step 1: Establish the longitudinal dynamics model and power model of electric vehicle i based on the vehicle's own parameters; Step 2: Determine the control objective; Step 3: Construct a fleet reference trajectory optimization problem based on model predictive control, and use an improved PSO algorithm to obtain the optimal solution; Step 4: Using active disturbance rejection control technology, design a tracking differentiator, an extended state observer, and a nonlinear state error feedback law to estimate the lumped disturbance and send it to the leading vehicle.

2. The method for cooperative ecological driving of intelligent connected electric vehicle fleets based on disturbance observation according to claim 1, characterized in that: Step one specifically includes: 1.

1. Establish the longitudinal dynamics model of electric vehicle i: Where, p i v i u i These represent the position, speed, and control input of electric vehicle i, respectively, and T. d (t) is the lumped disturbance of vehicle i, where u i (t) and T d (t) can be expressed as: In the formula F i (t) and F i，r (t) represents the traction force and total resistance, m i It is the mass of vehicle i, F i，r (t) can be expressed as follows: In the formula C d (d i (t)), A i，v μ i and θ(p) i (t) represents the distance d between the vehicle and the vehicle. i (t) related drag coefficient, frontal area of ​​vehicle, rolling drag coefficient and location-related road gradient, ρ and g are air density and gravitational acceleration, respectively; Vehicle spacing d i (t) is defined as: d i (t)=p i-1 (t)-p i (t)-L i L i It is the length of vehicle i. Drag coefficient C d It can be represented as: Among them, C i,d is the rated drag coefficient of vehicle i, and c1 and c2 are fitting parameters; 1.

2. Establishing a power model for electric vehicles The output power P of electric vehicle i i,e (t) is represented as: Where η i，r and η i，r These are the mechanical transmission efficiency and energy recovery efficiency of electric vehicles, respectively.

3. A cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation according to claim 1 or 2, characterized in that: Step two specifically includes: Desired position of vehicle i and speed The definition is as follows: Where d represents the desired distance between adjacent vehicles. The position error and speed error of following vehicle i are defined as follows: 2.

1. Internal stability When the acceleration of the lead vehicle is 0, all following vehicles remain at the desired position and eventually reach the same speed as the lead vehicle, that is: Where i = 1, 2, ..., N 2.

2. The string stability condition is: 2.

3. Energy Economy The total energy consumption from time t to t+T is expressed as:

4. The method for cooperative ecological driving of intelligent connected electric vehicle fleets based on disturbance observation according to claim 3, characterized in that: Step three specifically includes: 3.

1. The band-stop function with compensation factor is expressed as: in, and z l , z u ∈R + , where are the upper and lower bounds of the interval, and c is the compensation factor. 3.

2. Constructing a fleet reference trajectory optimization problem based on model predictive control The discrete longitudinal dynamics model of the lead vehicle is expressed as: In the formula, Δt is the discrete time interval. Let x 0，ref =[p 0，ref v 0，ref ] T Further acquisition: x 0,ref (k+1)=A0x 0,ref (k)+B0u 0,ref (k) in, Assuming all following vehicles remain at their desired positions within the vehicle platoon, the reference position, reference speed, and reference control input of following vehicle i are expressed as follows: Let N P and N c Let the prediction range and control range be the two distinct values, respectively. The cost function for vehicle i within the prediction range can be expressed as: Where ω1 and ω2 are weight parameters, v l and v u It is the expected speed range of the vehicle platoon [v1, v] u The upper and lower bounds of the vehicle queue, and the cost function within the prediction range, can be obtained as follows: Given the predictors and cost function of the vehicle platoon, the platoon reference trajectory problem is formulated as follows: and satisfy Where F i,ref (n|k) and P i,ref,e (n|k) can be represented as:

5. A cooperative ecological driving method for intelligent connected electric vehicle fleets based on disturbance observation according to claim 1 or 4, characterized in that: Step four specifically includes: The tracking differentiator is specifically expressed as follows: In the formula, r0 and h0 are the velocity factor and the filter factor, respectively; fhan is the fastest synthesis function, and its expression is as follows: fhan(x1,x2,r0,h0)=-r0[a / d0-sign(a)]s2-r0sign(a), in, Using ESO to perturb the T of electric vehicles i，d Estimate (t) and let T i，d (t) is a state variable w of electric vehicle i. i The specific expression for the extended state observer of electric vehicle i is given below: (t). In the formula, e i The estimation bias of the position of electric vehicle i; For system state variable p i The predicted value; For system state variable v i The predicted value; For perturbation T i，d The estimated value of (t) B i β1, β2, and β3 are compensation factors; β1, β2, and β3 are observer parameters; fal is a nonlinear function, specifically expressed as: Give the nonlinear state error feedback rate u for electric vehicle trajectory tracking. i as follows:

Images