Intelligent fleet following control method based on motion state estimation under communication delay

Abstract

This invention discloses an intelligent platoon following control method based on motion state estimation under communication delay. It employs a motion state estimation algorithm to estimate the state information of the preceding vehicle under delay conditions, and then transmits the estimated state information to the model predictive controller (MPC) to determine the desired acceleration of the following vehicles. This enables the platoon to move stably while improving the control effect of vehicle formation. Through this vehicle motion state estimation-based platooning control method, relatively accurate preceding vehicle state information can be obtained under communication delay conditions, thereby ensuring the accuracy of the MPC controller's decisions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent vehicle platoon longitudinal following control, and particularly relates to an intelligent vehicle platoon following control method based on motion state estimation under communication delay. BACKGROUND

[0002] With the growth of the number of cars and road traffic, a series of problems such as traffic congestion and environmental pollution have appeared. To solve this problem, a vehicle platooning scheme can be used. Vehicle platooning refers to a group of vehicles arranged longitudinally in a column and stably traveling at the same speed and ideal inter-vehicle distance. When platooning, the vehicles can maintain a small inter-vehicle distance to travel under the premise of safety, which can increase road capacity and alleviate traffic congestion. At the same time, when the vehicles platoon, the air resistance of the vehicle queue is reduced due to the interaction between the flow fields of the front and rear vehicles, fuel consumption and emissions are reduced, and environmental pollution is reduced. Therefore, vehicle platooning as an intelligent driving application with great development prospects has attracted much attention.

[0003] Currently, the stability and control performance of vehicle platooning mainly depend on scientific and reasonable control strategies. However, in the existing vehicle platoon control strategy, information is transmitted between adjacent two vehicles in the queue through wireless communication technology, and the transmission process often has communication delay, that is, when the rear vehicle receives the state information transmitted by the front vehicle, the current state information of the front vehicle has changed, so that the rear vehicle cannot accurately obtain the state information of the front vehicle, which further brings negative effects to the vehicle platoon control. Therefore, when formulating the control strategy, it is very important to consider the influence of communication delay on vehicle platooning.

[0004] For example, patent CN115951666A discloses a vehicle control method for an automatic driving vehicle equipped with a vehicle-mounted device in vehicle platooning, which aims to combine with the real-time obtained vehicle motion state information to optimize and solve the corresponding optimization control amount at each time node, so as to realize vehicle longitudinal control, but the existing scheme does not consider the influence of communication delay. For example, patent CN115951666A discloses a vehicle platoon control method based on communication delay, which adopts Kalman filtering algorithm and model predictive control method to compensate for the vehicle information deviation caused by communication delay, so as to improve the control effect of vehicle platoon; but the existing technology only makes one-step state prediction on the state information of the vehicle under communication delay. SUMMARY

[0005] Invention purposes: The purpose of the present application is to solve the problems existing in the prior art, provide a communication delay based on intelligent vehicle following control method of state estimation, according to the vehicle kinematics principle and the size of communication delay, the state information of the front vehicle in the formation is estimated, and the front vehicle state information estimation value is transmitted to the model predictive controller, the expected acceleration of the vehicle is determined, so as to improve the control effect of vehicle formation and ensure the stability of vehicle formation under the influence of communication delay.

[0006] Technical scheme: A vehicle formation longitudinal following control method under the condition of communication delay, comprising the following steps:

[0007] Step S1, based on the model predictive control principle, the vehicle kinematics equation is established

[0008] Step S1.1, the longitudinal motion process of intelligent vehicle in vehicle formation is described as follows:

[0009]

[0010]

[0011]

[0012] a des The expected acceleration of the following vehicle decided by the upper MPC controller; τ is the time constant of the engine, τ -1 Indicates s, v, a respectively represent the first derivative of vehicle position s, vehicle speed v and vehicle acceleration a;

[0013] Step S1.2, the safety distance algorithm of constant vehicle head time is used as the control research method, the inter-vehicle distance error e s And the speed error e v of the ith vehicle and the front vehicle are solved.

[0014] e s = s i-1 -s i -d i,des -l

[0015] e v = v i-1 -v i

[0016] In the above formula, s i , v i , a i respectively represent the position, speed and acceleration of the ith vehicle, l represents the length of the vehicle body, d i,des is the expected inter-vehicle distance between the ith vehicle and the front vehicle i-1, and the formula is as follows:

[0017] d i,des = d0+ h i v i ;

[0018] d0is the minimum safety distance, h i is the fixed headway of the ith vehicle;

[0019] Step S1.3, to describe the relationship between state variables and input, output variables, a state space model of the vehicle in the intelligent vehicle platoon longitudinal following control system is constructed, expressed as:

[0020]

[0021] y(t) = Dx(t)

[0022] In the above formula,

[0023]

[0024] x(t) = [e s e v a i ] T , u(t) = a des , ω(t) = a i-1

[0025] Where x(t), u(t), ω(t) and y(t) are state input, control input, disturbance input and system output respectively, A is the state coefficient matrix, B and C are input coefficient matrices, and D is the output coefficient matrix; Step S2, considering the communication delay of information transmission between vehicles in the platoon, a front vehicle motion state estimation method is constructed to estimate the motion state information of the front vehicle under the condition of communication delay, and the updated distance error and speed error are obtained. The specific method is:

[0026] Step S2.1, based on the relationship between the delay length τ of the communication delay and the prediction time step λ t , the speed information of the front vehicle i-1 is estimated and calculated

[0027] Step S2.2, according to the speed information obtained in step S2.1, the position information of vehicle i-1 is estimated:

[0028]

[0029] Further, the position and speed trajectory of vehicle i-1 are as follows:

[0030]

[0031]

[0032] Step S2.3: Based on the state information of the preceding vehicle under the delay conditions obtained in step S2.2. and use and Replace the 's' in the spacing error formula and speed difference formula in step S1.2. i-1 and v i-1 The updated spacing error and speed error As the control input for the MPC model:

[0033]

[0034]

[0035] Step S3: Calculate the desired acceleration of the following vehicle using the updated state-space model through the MPC controller. To prevent abrupt changes in acceleration, replace the controlled acceleration u with the acceleration increment Δu; Step S3.1: Perform a discrete linear transformation on the state-space model from step S1:

[0036]

[0037] y(t)=ξx(t)

[0038] Where x(t) is the system state; y(t) is the system output; u(t) is the control input; ω(t) is the disturbance input; ζ = diag(1,0,0); These are the updated forms of A, B, and C after the system state equations have been discretized;

[0039] Step S3.2: Assume that the state vector and the disturbance vector are in each sampling period t. s It is internally measurable, and the interference prediction of future times at time t is itself, i.e., ω(t+j|t)=ω(t);

[0040] Step S3.3: Use an iterative model to analyze the future N. P The predictions at each sampling time point yield the following final control prediction expression for the future at time t:

[0041]

[0042] Where Δu is the control increment, u(t|t) represents the control input predicted at time t, and N c Represents the control time domain, and N c ≤Np ;

[0043] Step S3.4, defining the target function as:

[0044]

[0045] where y ref is the reference trajectory, which is related to the output measurement value up to the current time; the symbol represents a quadratic function; P and Q are error and input weighting matrices respectively; the above formula shows that the performance index mainly considers the error size of the output and the reference and the energy size of the input;

[0046] Rewrite the above target function into the following vector form:

[0047]

[0048] Define the vector E(t) as the deviation of the system free response and the future target trajectory:

[0049] E(t) = Y ref (t) - ξ x x(t) - ξ u u(t-1) - ξ ω ω(t)

[0050] The standard form of the quadratic programming can be obtained:

[0051]

[0052] Suppose that the optimization function is constrained by the following conditions:

[0053]

[0054] The present application comprehensively considers vehicle longitudinal dynamics analysis, vehicle kinematics equation description, spacing control strategy selection and vehicle state space model construction; by establishing a motion state estimation equation, the state information of the preceding vehicle under the condition of communication delay is estimated, and an updated vehicle state space model is obtained; then the estimated state information of the preceding vehicle is input into the upper MPC controller, the expected control input of the controlled vehicle is obtained by calculation and solving, and the acceleration or braking of the controlled vehicle is realized through the lower controller.

[0055] Further, the step S1.1 first describes the longitudinal motion process of the intelligent vehicle in the vehicle platoon as the following nonlinear third-order differential equation:

[0056]

[0057] Then, the above formula is linearized according to the feedback linearization theory to obtain the following expression:

[0058]

[0059] η is the control input of the engine.

[0060] Further, the step S1.2 assumes that three vehicles equipped with the same device are in a queue on a straight road when solving the inter-vehicle distance error and the speed error according to the inter-vehicle distance strategy.

[0061] Further, the step S2 assumes that the communication delay between the vehicle i and the vehicle i-1 is τ, and the motion state information received by the vehicle i at the time t+n is actually the state information of the vehicle i-1 at the time t+n-τ. The motion state information of the vehicle i-1 at the time t+n is predicted by using the position and the speed information of the vehicle i-1 received by the vehicle i at the time t, and n is the prediction time step.

[0062] Further, the step S2.1 estimates the speed information of the preceding vehicle i-1 as follows:

[0063] If the prediction time step λ t is greater than the communication delay τ, i.e., λ t > τ, it is assumed that the state information of the preceding vehicle at the time t+n does not change and is consistent with the time t+n-τ when the state information is sent. At this time, the speed information of the preceding vehicle at the time t+n is expressed as:

[0064]

[0065] If the prediction time step λ t is less than or equal to the communication delay τ, i.e., λ t ≤ τ, the state information of the preceding vehicle is obtained according to the following formula, and the influence of the delay is eliminated:

[0066]

[0067] wherein τ' = τ - Tλ t .

[0068] Further, in the step S2.3, if an uncertain factor interferes with the communication between the vehicles, the position and speed information of the i-1th vehicle are taken as the input of the model predictive control algorithm, which can be obtained from the position and the speed trajectory of the vehicle, and P is the prediction range.

[0069] Further, the step S3.3 takes the future N P ​The iterative model of the prediction of the sampling moment is as follows:

[0070]

[0071] Then, the expression of the prediction control quantity at t moment to t+j moment is obtained, as follows:

[0072] u(t+j|t) = u(t+j-1|t) + Δu(t+j|t)

[0073] In the above formula, the control quantity u(t-1) is known.

[0074] Beneficial effects: the motion state estimation algorithm is adopted to estimate the state information of the preceding vehicle under the delay condition, the estimated state information is transmitted to the model prediction controller, and then the expected acceleration of the following vehicle is determined, so that the queue can stably travel and the control effect of the vehicle platoon is improved. Through the above vehicle motion state estimation platoon control method, the relatively accurate state information of the preceding vehicle can be obtained under the communication delay condition, and then the accuracy of the decision of the MPC controller is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 is the overall flowchart of the present application;

[0076] Figure 2 is the logic diagram of the present application;

[0077] Figure 3 is the schematic diagram of the motion state estimation of the preceding vehicle in the present application;

[0078] Figure 4 is the control effect diagram of the three control methods under the continuous acceleration and deceleration working condition in the embodiment, and the delay time length is 100 ms;

[0079] Figure 5 is the control effect diagram of the three control methods under the emergency braking working condition in the embodiment, and the delay time length is 100 ms. DETAILED DESCRIPTION

[0080] The technical scheme of the present application will be described in detail below, but the protection scope of the present application is not limited to the described embodiments.

[0081] For the problem of intelligent vehicle fleet stability and control effect under communication delay influence, the intelligent vehicle fleet following control method based on motion state estimation under communication delay of the application considers the influence of communication delay, predicts the vehicle state information under communication delay in a time domain, obtains the state trajectory in the prediction time domain at a time, when there is communication delay, the vehicle state information required by MPC control can be obtained from the state prediction trajectory, so as to obtain relatively accurate front vehicle state information, and then ensure the accuracy of the decision of the MPC controller.

[0082] As shown in Figure 1 and Figure 2 , the specific control process is as follows:

[0083] First, the vehicle kinematics equation is established based on the model predictive control principle:

[0084] First, the longitudinal motion process of the intelligent vehicle in the vehicle fleet is regarded as a nonlinear third-order model, which is described by the following differential equation:

[0085]

[0086] Where s, v and a are the position, speed and acceleration of the vehicle respectively; η is the control input of the engine.

[0087] According to the feedback linearization theory, the above expression is further linearized to obtain the following expression:

[0088]

[0089] Where a des is the expected acceleration of the following vehicle decided by the upper controller; τ is the time constant of the engine.

[0090] Next, consider three closely following intelligent vehicles of the same configuration in the vehicle fleet driving on a straight road, s i , v i , a i represent the position, speed and acceleration of the i-th vehicle respectively, and l represents the length of the vehicle body, then the inter-vehicle spacing error of the i-th vehicle and the front vehicle is:

[0091] e s = s i-1 -s i -d i,des -l (3)

[0092] Where d i,des is the expected inter-vehicle spacing of the i-th vehicle and the front vehicle, and the expected inter-vehicle spacing can be obtained through reasonable decision of the inter-vehicle spacing strategy.

[0093] This embodiment uses a constant headway safety distance algorithm as the control research method. For the i-th vehicle, the distance error between it and the vehicle in front is shown in equation (3), where the expected distance is:

[0094] d i,des =d0+h i v i (4)

[0095] Where d0 is the minimum safe distance, usually 2-5m; h i To fix the headway, the typical value range in this invention is between 0.8 and 3.6 seconds.

[0096] The speed difference between the current vehicle and the vehicle in front was calculated as follows:

[0097] e v =v i-1 -v i (5)

[0098] Finally, combining formulas (1) to (5), the state-space model of the vehicle is derived as follows:

[0099]

[0100] In the above formula,

[0101]

[0102] x(t)=[e s e v a i ] T u(t) = a des ω(t)=a i-1

[0103] Where x(t), u(t), ω(t) and y(t) are the state input, control input, disturbance input and system output, respectively; A is the state coefficient matrix; B and C are the input coefficient matrices; and D is the output coefficient matrix.

[0104] The second step is to consider the communication delay in information transmission between vehicles in the formation, and to construct a motion state estimation algorithm for the vehicle in front to predict the state information of the vehicle in front under the condition of communication delay, and update the vehicle spacing error and speed error.

[0105] like Figure 3 As shown, vehicle i receives the motion state information of vehicle i-1 at time t+n. Due to the communication delay τ, the motion state information received by vehicle i at time t+n is actually the state information of vehicle i*1 at time t+n-τ. To eliminate the effect of the delay, the state information of vehicle i-1 at time t+n is estimated using a corresponding formula.

[0106] First, let's discuss the delay duration τ and the prediction time step λ. t Based on the relationship, different formulas are used to estimate the speed information of the vehicle in front under the corresponding conditions. The specific motion state estimation methods are shown below.

[0107] Case 1: When the prediction time step λ t When the latency is greater than the communication delay τ, i.e., λ t If the speed is greater than τ, we can assume that the state information of the preceding vehicle at time t+n has not changed and is consistent with that at time t+n-τ when the state information was issued. In this case, the speed information of the preceding vehicle at time t+n can be expressed as:

[0108]

[0109] Case 2: When the prediction time step λ t When the communication delay is less than or equal to τ, i.e., λ t For a value ≤τ, based on the principles of vehicle kinematics, the following formula is used to estimate the state information of the preceding vehicle in order to eliminate the effect of delay:

[0110]

[0111] in τ′=τ-Tλ t .

[0112] Then, based on the speed information obtained from the above formula, the position information of vehicle i-1 can be estimated:

[0113]

[0114] The position and velocity trajectory of vehicle i-1 are represented as follows:

[0115]

[0116]

[0117] Next, based on the state prediction information of the preceding vehicle obtained from the above steps... and use and Replace 's' in the spacing error formula and the speed difference formula. i-1 and v i-1 The compensated spacing error and This serves as the control input for the MPC model:

[0118]

[0119]

[0120] Third step, the updated state space model is calculated by the upper controller to obtain the expected acceleration:

[0121] First, the first step state space model is discretely linearly converted:

[0122]

[0123] Wherein, x(t) is the system state; y(t) is the system output; u(t) is the control input; ω(t) is the disturbance input; ζ = diag(1, 0, 0).

[0124] It is assumed that the state vector and the disturbance vector are measurable at each sampling period t s , and the disturbance at time t is predicted for future time, i.e. ω(t+j|t) = ω(t). And the prediction for future N P sampling time is realized through the following iterative model:

[0125]

[0126] N P It is called the prediction time domain, and u(t|t) represents the predicted control input. The expression of the predicted control at time t for time t+j is as follows:

[0127] u(t+j∣t)=u(t+j-1∣t)+Δu(t+j∣t)

[0128] Wherein, Δu is the control increment.

[0129] Since the control u(t-1) is known, the future control prediction at time t can be represented by the following formula:

[0130]

[0131] Wherein, N c represents the control time domain, and N c ≤N p .

[0132] Next, the objective function is defined as:

[0133]

[0134] Wherein y ref is the reference trajectory, which is related to the output measurement value up to the current time, or is a pre-set trajectory, and the symbol represents a quadratic function, and P and Q are error and input weighting matrices, respectively.

[0135] The first term in the objective function reflects the following ability of the system, and the second term reflects the change of the control increment.

[0136] The above objective function is rewritten in the following vector form:

[0137]

[0138] Define the vector E(t) as the deviation of the system free response from the future target trajectory:

[0139] E(t) = Y ref (t) - x x x(t) - x u u(t-1) - u ω ω(t) (18)

[0140] The standard form of the quadratic programming can be obtained:

[0141]

[0142] Suppose that the optimization function is subject to the following conditions:

[0143]

[0144] Embodiment:

[0145] In this embodiment, first, a Simulink and Carsim joint simulation experiment platform is built, then two different driving conditions are selected for simulation verification under a communication delay duration of 100 ms, the first one is a continuous acceleration and deceleration condition: the initial speeds of the three vehicles are all 50 km / h, the initial positions are 30 m, 15 m and 0 m respectively, and the acceleration of the leading vehicle is a = sin(0.2t); the second one is an emergency braking condition: the initial speeds of the three vehicles are all 120 km / h, the initial positions are 100 m, 50 m and 0 m respectively, and the acceleration of the leading vehicle is -4.5 m / s 2 Finally, the control effect of the controller of the present application is compared with that of the conventional PID and MPC controllers, and the simulation results are shown in Figure 4 and Figure 5 , wherein figures a, b and c are the existing PID, the existing MPC and the technical solution of the present application respectively.

[0146] As Figure 4It can be seen that under the continuous acceleration and deceleration working condition, the dynamic adjustment fluctuation of the existing PID control method is large, the maximum speed of the vehicle 3 has exceeded 100km / h, which leads to a large fluctuation range of the spacing error, and the system control effect is poor. After a period of adjustment, the existing MPC control method reaches a stable state, and the spacing error is reduced to within 5m compared with the PID control method, and the control effect is obviously improved. Compared with the MPC control method, the control effect of the control method of the present application on the vehicle team system is further improved, the dynamic adjustment range is reduced, and the spacing error is less than 5m.

[0147] As Figure 5 It can be seen that under the emergency braking working condition, the existing PID control method has a large adjustment error, and the actual distance between the two adjacent vehicles is less than the safety distance, which is easy to cause collision phenomenon. After a period of dynamic adjustment, the existing MPC control method makes the actual distance between the two adjacent vehicles stop at 6.1m and 6.5m respectively, and the spacing error is reduced to 2.1m, which can keep the vehicle team system in a stable state. The dynamic adjustment change range of the control method of the present application is small, and the spacing error is further reduced to 1.6m, which is better than the control effect of the PID and MPC control methods.

[0148] From the above experimental results, it can be seen that the intelligent vehicle team following control method based on motion state estimation under communication delay of the present application can eliminate the influence of delay and ensure the effect of formation control.

Claims

1. An intelligent platoon following control method based on motion state estimation under communication delay, characterized in that, The method comprises the following steps: Step S1, establishing a vehicle kinematics equation based on a model predictive control principle Step S1.1, describing the longitudinal motion process of the intelligent vehicle in the vehicle fleet in the following manner: a des the desired acceleration of the following vehicle as decided by the upper MPC controller; τ is the time constant of the engine, τ -1 denotes are the first derivative of the vehicle position s, the vehicle speed v, the vehicle acceleration a, respectively; Step S1.2, the safety distance algorithm with constant time headway is used as the control research method to calculate the inter-vehicle distance error e of the ith vehicle and the front vehicle s , and the speed error e of the ith vehicle and the front vehicle v ; e s = s i-1 - s i - d i,des - l e v = v i-1 - v i In the above formula, s i , v i , a i respectively represent the position, speed and acceleration of the i-th vehicle, l represents the length of the vehicle body, d i,des is the desired inter-vehicle distance between the i-th vehicle and the preceding vehicle i-1, and the formula is as follows: d i,des = d0+ h i v i ; d0 is the minimum safety distance, h i is the fixed headway of the ith vehicle; Step S1.3, constructing a state space model of the vehicle in the intelligent vehicle fleet longitudinal following control system, expressed as: y(t) = Dx(t) In the above formula, x(t) = [e s e v a i ] T , u(t) = a des , ω(t) = a i-1 Wherein, x(t), u(t), ω(t) and y(t) are state input, control input, disturbance input and system output respectively, A is a state coefficient matrix, B and C are input coefficient matrices, and D is an output coefficient matrix; Step S2, constructing the front vehicle motion state estimation method, estimating the motion state information of the front vehicle under the communication delay condition, and updating the inter-vehicle distance error and speed error to obtain the updated distance error and speed error The specific method is: Step S2.1, estimating the speed information of the preceding vehicle i-1 based on the relationship between the delay duration τ of the communication delay and the prediction time step λ t ​ Step S2.

2. Estimating the speed information from the position information of step S2.1 Estimating the position information of vehicle i-1: Then, the position and speed trajectory of the vehicle i-1 are obtained as follows: Step S2.3, update the state information of the preceding vehicle based on the delay condition and replace s in the spacing error formula and the speed difference formula in step S1.2 with and i-1 and v i-1 the updated spacing error and speed error as the control input of the MPC model:​ Step S3, calculating the expected acceleration of the following vehicle through the MPC controller based on the updated state space model; Step S3.1, discretely linearly converting the state space model of step S1: y(t) = ζx(t) Where x(t) is the system state; y(t) is the system output; u(t) is the control input; ω(t) is the disturbance input; ζ = diag(1, 0, 0); A, B, C are the updated forms of A, B, C of the system state equation after discretization, respectively; Step S3.2, assuming that the state vector and the disturbance vector are measurable at every sampling period t s and the disturbance prediction for future time instants is its own, i.e. ω(t + j | t) = ω(t). Step S3.3, prediction of future N P sampling instants by iterating the model, the final expression of the control prediction at time t for the future is as follows: wherein, Δu is a control increment, u(t|t) represents a control input predicted at time t, N c represents a control time domain, and N c ≤N p ; Step S3.4, to ensure the following property of the vehicle in the fleet and avoid sudden changes in the controlled acceleration, the objective function is defined as: where y ref is the reference trajectory, related to the output measurements up to the current time instant; the symbol represents a quadratic function; P, Q are error and input weighting matrices, respectively, the first term in the objective function reacts to the following ability of the system, the second term reacts to the variation of the control increment; Rewrite the above objective function into the following vector form: Define the vector E(t) as the deviation of the system free response and the future target trajectory: E(t) = Y ref (t) - ξ x x(t) - ξ u u(t - 1) - ξ ω ω(t) The standard form of the quadratic programming can be obtained as follows: Suppose that the optimization function is subject to the following conditions:

2. The intelligent platoon following control method based on motion state estimation under communication delay according to claim 1, wherein, The step S1.1 first describes the longitudinal motion process of the intelligent vehicle in the vehicle fleet as the following nonlinear third-order differential equation: Then, the above formula is linearized according to the feedback linearization theory to obtain the following expression: η is the control input of the engine. 3.The intelligent vehicle platooning control method based on motion state estimation under communication delay according to claim 1, wherein, The step S1.2 assumes that three vehicles equipped with the same equipment are queue driving on a straight road when solving the inter-vehicle distance error and speed error.

4. The intelligent platoon following control method based on motion state estimation under communication delay according to claim 1, wherein, The step S2 assumes that the communication delay between the vehicle i and the vehicle i-1 is τ, and the motion state information received by the vehicle i at the time t+n is actually the state information of the vehicle i-1 at the time t+n-τ. The position and the speed of the vehicle i-1 received by the vehicle i at the time t are used to predict the motion state information of the vehicle i-1 at the time t+n, and n is the prediction time step, and P is the prediction range.

5. The intelligent platoon following control method based on motion state estimation in communication delay according to claim 1, wherein, The step S2.1 estimates the speed information of the preceding vehicle i-1 The detailed process is as follows: If the prediction time step λ t is greater than the communication delay τ, i.e. λ t > τ, it is assumed that the state information of the preceding vehicle at time t + n has not changed and is consistent with the time t + n - τ when the state information was sent. In this case, the speed information of the preceding vehicle at time t + n is expressed as: If the prediction time step λ t is less than or equal to the communication delay τ, i.e. λ t ≤ τ, then the state information of the preceding vehicle is obtained at this time according to the following formula, and the effect of the delay is eliminated. wherein τ' = τ - Tλ t .

6. The intelligent platoon following control method based on motion state estimation in communication delay according to claim 1, wherein, If there is an uncertainty factor interference in step S2.3, which causes abnormal communication between vehicles, the position and speed information of the i-1th vehicle is taken as the input of the model predictive control algorithm, and the position and speed trajectory of the above vehicle is obtained, and the speed trajectory is obtained, and P is the prediction range.

7. The intelligent platoon following control method based on motion state estimation in communication delay according to claim 1, wherein, The step S3.3 of predicting the future N P iterative model for N future sampling instants is as follows: Then, the expression of the prediction control quantity at t time on t+j time is obtained, as follows: u(t+j|t) = u(t+j-1|t) + Δu(t+j|t) In the above formula, the control quantity u(t-1) is known.

Images