A method and device for multi-objective optimization control of heterogeneous commercial vehicle fleet under time delay conditions

Abstract

The application discloses a kind of time delay conditions under heterogeneous commercial vehicle fleet multi-objective optimization control method and device, comprising: S1 establishes the dynamics model of each car in commercial vehicle queue and queue system model;S2 based on distributed model predictive control method designs queue controller, establishes cost function and constraint condition, defines sub-predictive optimization problem for each vehicle, designs follow car error, economy and comfort cost function, reduces follow car error in heterogeneous vehicle queue driving, reduces energy consumption, improves comfort, improves queue comprehensive performance;S3 by integrating buffer and delay compensator on controller, add delay compensation link for control system, improve the control effect of heterogeneous vehicle fleet under delay condition, realize the stable control of heterogeneous queue under non-ideal communication condition;S4 solves target function using simulated annealing and particle swarm optimization algorithm, solves the problem of nonlinear discontinuous target value in target function according to probability method, reduces operation time, improves solving efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent transportation, in particular to a heterogeneous commercial vehicle fleet multi-objective optimization control method and device under time delay conditions. BACKGROUND

[0002] Highway freight transportation has always been the foundation of the world's economic development, and freight demand shows a trend of increasing year by year. However, with the substantial increase in highway traffic flow, congestion is becoming increasingly serious, traffic efficiency is low, and single vehicle energy consumption is large. Multi-vehicle platoon control is a typical cooperative motion control, and the platoon vehicles can automatically adjust the motion state according to the surrounding vehicle state information, and achieve the expected platoon geometric configuration through speed control. With the evolution of advanced control theory and the increasing use of platoon in the logistics industry, the multi-objective control problem of platoon operation is increasingly prominent, and the control algorithm should take into account the following performance such as followability, economy, comfort and the like. In addition, when the number of commercial vehicles increases, the time delay increases, and the stability margin of closed-loop dynamics decreases. Non-ideal communication with time delay has a significant impact on the performance of vehicle platoon control, especially in large-scale platoon systems.

[0003] Currently, there are several methods for studying commercial vehicle platoon control, such as PID control, LQR control, and distributed model predictive control (DMPC). The advantage of PID control is that the algorithm is simple and relatively mature, but when facing relatively complex changing systems, the application of PID controller will lead to the problem of unsatisfactory control effect. Although the LQR control algorithm is simple to implement and can obtain the optimal control strategy, it is mainly designed for linear systems and has poor control effect for nonlinear systems. Distributed model predictive control can meet the needs of nonlinear complex systems, and for vehicle platoon control, using DMPC control algorithm has the advantages of high safety, stable control, and good adaptability to complex working conditions. Therefore, it has great research value to use distributed model predictive control for commercial vehicle platoon driving.

[0004] For the control of heterogeneous platoon, there are currently studies on the heterogeneous dynamic characteristics of the platoon system, fixed or variable inter-vehicle distance strategy, and inter-vehicle information flow topology. However, most of the researches only improve the single-sided control effect of heterogeneous platoon stability or economy, and do not take into account the followability, economy, and comfort of multiple platoon performances.

[0005] Communication time delay has a non-negligible impact on vehicle dynamics stability and platoon string stability. The common delay processing methods are divided into delay compensation methods based on compensation controller and information reconstruction. Among them, the method based on compensation controller mainly includes model predictive controller, H ∞Robust controller and neural network controller, while the delay compensation method based on information reconstruction mainly solves the problem of communication delay. However, most of the researches are for homogeneous vehicle platoon with the same system dynamics parameters, and most of them use linear controller, which has limited application in complex nonlinear systems. SUMMARY

[0006] To solve the problems raised in the background art, the purpose of the present application is to provide a heterogeneous commercial vehicle platoon multi-objective optimization control method under time delay conditions, a distributed model predictive multi-objective controller is designed according to the commercial vehicle platoon model, a delay compensation link is added to the control system for the time delay environment, and finally the simulated annealing and particle swarm optimization algorithm are used to solve the objective function, so as to realize the stable control of the following performance, economy and comfort of the heterogeneous platoon under non-ideal communication conditions.

[0007] The application adopts the following technical scheme:

[0008] The present application provides a heterogeneous commercial vehicle platoon multi-objective optimization control method under time delay conditions, which comprises the following steps:

[0009] S1: Establishing the dynamics model of each vehicle in the commercial vehicle platoon and the platoon system model;

[0010] S2: Designing the controller of the heterogeneous commercial vehicle platoon based on the distributed model predictive control method, establishing the cost function and the constraint condition;

[0011] S3: Adding a delay compensation link to the control system;

[0012] S4: Using the simulated annealing and particle swarm optimization algorithm to solve the objective function.

[0013] Further, the dynamics model of each vehicle in S1 is as follows:

[0014] The acceleration is written as the equation of driving power, braking force and resistance, according to Newton's second law, the relationship between acceleration a i (t), driving force F i,T (t) and resistance F i,r (t) can be expressed as

[0015]

[0016] Since the driving force Equation (1) is rewritten as

[0017]

[0018] In the formula, m i is the mass of the i-th vehicle, r w,i is the wheel rolling radius of the i-th vehicle, T i(t) is the motor output torque of the ith vehicle, η m,i is the mechanical efficiency of the powertrain of the ith vehicle, the resistance F i,r (t) is composed of rolling resistance, wind resistance and slope resistance, which can be expressed as

[0019]

[0020] where g and f are the gravity acceleration and the rolling coefficient respectively, C d,i is the air resistance coefficient of the ith vehicle, r is the air density, A i is the frontal area of the ith vehicle.

[0021] Further, the queue system model in S1 is as follows:

[0022] Let the system state variable be the vehicle displacement, vehicle speed and motor output torque, i.e. the state variable x i (t) = [S i (t), v i (t), T q,i (t)] T , the control variable is the vehicle motor torque, i.e. u i (t) T q,i (t),

[0023] Then the state equation is

[0024]

[0025] Rewrite equation (4) as

[0026]

[0027] where: i is the corresponding number of each vehicle; S i (t) is the displacement of the ith vehicle; v i (t) is the speed of the ith vehicle; T i (t) is the actual torque of the motor of the ith vehicle; e i (t) is the expected torque of the motor of the ith vehicle; t i is the time delay parameter of the power system of the ith vehicle;

[0028] The discrete dynamics equation of the system can be obtained, where△t is the time step for each calculation:

[0029]

[0030] Further, the implementation of S2 includes defining 3 types of variables, as follows:

[0031] (1) System prediction variables:

[0032] Predicted position of the ith vehicle at time t + n;

[0033] Predicted velocity of the ith vehicle at time t + n;

[0034] Predicted acceleration of the ith vehicle at time t + n;

[0035] Predicted control input of the ith vehicle at time t + n;

[0036] Predicted control output of the ith vehicle at time t + n;

[0037] (2) Optimal variables:

[0038] Optimal control input of the ith vehicle at time t + n;

[0039] (3) Hypothetical variables:

[0040] Hypothetical position of the ith vehicle at time t + n;

[0041] Hypothetical velocity of the ith vehicle at time t + n;

[0042] Hypothetical acceleration of the ith vehicle at time t + n;

[0043] Hypothetical control input of the ith vehicle at time t + n;

[0044] Hypothetical control output of the ith vehicle at time t + n.

[0045] Further, the cost function in S2 includes the following:

[0046] (1) Car-following error cost function

[0047] First, the car-following error cost function of the ith vehicle (i = 2, 3, 4, …, N) and the leading vehicle is established; let the system output be and the expected state of the leading vehicle be then the car-following error cost function of the ith vehicle and the leading vehicle is:

[0048]

[0049] In the formula: A i is the error weight coefficient matrix of the ith vehicle and the leading vehicle;

[0050] Secondly, the following error cost function of the ith vehicle (i = 2, 3, 4, …, N) and the preceding vehicle is established: define the expected state of the preceding vehicle as Then the following error cost function of the ith vehicle and the preceding vehicle is:

[0051]

[0052] In the formula: B i is the error weight coefficient matrix of the ith vehicle and the preceding vehicle; is the position of the preceding vehicle at time t; is the speed of the preceding vehicle at time t;

[0053] (2) Economic cost function

[0054] J 3,i (k|t) = ||C i P i (k|t)·△t||2 (9)

[0056] In the formula: C i is the energy consumption weight coefficient matrix of the ith vehicle; P i (k|t) is the motor power of the ith vehicle;

[0057] (3) Comfort cost function

[0058] By controlling the torque change rate at a small level, the comfort of each vehicle in the queue during driving is ensured, so the vehicle comfort cost function is:

[0059]

[0060] In the formula: D i is the comfort weight coefficient matrix of the ith vehicle; is the vehicle torque of the vehicle driving at a constant speed at a speed of is shown in formula (11):

[0061]

[0062] Further, the controller of the heterogeneous commercial vehicle queue in S2 is as follows:

[0063]

[0064] In the formula: v min is the minimum speed of the commercial vehicle driving on the highway; v max is the maximum speed of the commercial vehicle driving on the highway; T min is the minimum torque that the on-board motor can reach; T max is the maximum torque that the on-board motor can reach.

[0065] Further, the constraint condition of S2 is a penalty function as follows:

[0066]

[0067] Further, the delay compensation link of S3 includes a buffer, a compensator and a DMPC controller module; the buffer receives data from other vehicles according to information topology, and selects the latest data by comparing with cached data; the compensator predicts the assumed state with delay, and obtains the estimated value of the assumed state within the prediction horizon; in the DMPC module, a nonlinear longitudinal dynamics model of the vehicle is established, the control input is calculated, and the assumed state of the ego vehicle is sent to adjacent vehicles; the details are as follows:

[0068] The data transmitted between vehicles include time stamp, vehicle code and assumed state trajectory, which are generated by the DMPC controller and represent the possible trajectory in the future according to the optimal solution; the prediction horizon of the assumed state is N p , and there is a certain time delay t ij (t) when vehicle j transmits information to vehicle i through the communication network, and there is a certain time delay between the assumed state trajectory sent by vehicle j at time t and received by vehicle i and the true value, which is defined as:

[0069]

[0070] The buffer and the compensator select and predict the state to obtain the estimated value available for the DMPC controller to calculate; in the buffer, at time t, the newly arrived data is compared with the cached data j0 arrived at t , the new data is stored and transmitted to the compensator according to the time stamp of the data, and the old data is deleted; when no data arrives at t, the cached data

[0071] In the compensator, the error caused by time delay is compensated to obtain the assumed state estimation of the leading vehicle and the adjacent vehicle in the prediction horizon [t, t+N p -1], i.e. the assumed state estimation; the assumed state in the prediction range [t, t j -t ij (t j )+N p -1] is retained, and the part exceeding the prediction range is compensated by the prediction algorithm, the constant speed method is used to predict the future state of the vehicle, and the basic trajectory prediction equation based on the CV method is as follows:

[0072]

[0073] There is a certain error between the estimated state and the true state of other vehicles, and the estimated error is controlled within a certain range considering the speed, acceleration, torque and prediction time domain limit, so the estimated error has an upper limit, and the estimated state at time t+k can be expressed as:

[0074]

[0075] wherein

[0076]

[0077] Subsequently, the vehicle j obtains the estimated hypothetical state trajectory as follows:

[0078]

[0079] The estimated value is used for the DMPC controller to obtain a control input acceptable error.

[0080] Further, the implementation of the S4 comprises:

[0081] S4.1 initializes each particle parameter; the acceleration value of each vehicle in the vehicle platoon forms a particle group p1={a1, a2, …, ai}, the particle group size N is equal to the number of vehicle members, and a random update factor c1, an annealing rate λ, a maximum iteration number i and a minimum fitness value d are set during initialization; N ; max stop ;

[0082] S4.2 sets the fitness function; the objective function is taken as part of the fitness function, and the other part is the constraint set as the penalty function C(k); the fitness value of the initialized particle under the fitness function is calculated and recorded, d stop is equal to the fitness value, and the corresponding particle group is recorded as p g , and the acceleration value of each vehicle at this time forms;

[0083] S4.3 sets the initial temperature t0 as the minimum fitness value, i.e. t0=d stop ;

[0084] S4.4 uses the random update factor c1, the acceleration range and the particle group p g (i) to obtain a new particle group p g (i+1), and then i is increased by 1 until the maximum iteration number i max is reached, as shown in equation (19);

[0085] p g (i+1)=p g (i)+c1[a min ,a​max ] N (19)

[0086] S4.5 Records the fitness function value d at the current temperature t. temp If d temp ≤d stop Then the acceleration particle population p f Accept and equal to d temp The corresponding acceleration particle population p g If d temp >d stop Then p f by The probability of accepting p g , making p f =p g (i+1), or with The probability of accepting p g The same applies, as shown in equations (20)-(21);

[0087] if d temp ≤d stop :p f =p g (i+1)(20)

[0088] if d temp >d stop :

[0089]

[0090] S4.6 Returns to step S4.4 until i reaches i max Then let i = 1, p g (i)=p f ;

[0091] S4.7 performs cooling operation t k+1 =λt k ;

[0092] After S4.8 cools down, return to step S4.4 until the lowest temperature is reached, at which point the acceleration particles and the global optimum are output.

[0093] The present invention also proposes a multi-objective optimization control device for heterogeneous commercial vehicle fleets, which is capable of performing the control method described above.

[0094] The beneficial effects of this invention are:

[0095] 1. This invention defines a sub-prediction optimization problem for each vehicle in a heterogeneous fleet, and designs following error, economy and comfort cost functions, thereby reducing following error, reducing energy consumption, improving comfort and improving the overall performance of the fleet during the operation of heterogeneous commercial vehicle platoons.

[0096] 2. This invention uses simulated annealing and particle swarm optimization algorithms to solve the objective function, and uses a probabilistic method to solve the problem of nonlinear and discontinuous objective values ​​in the objective function, thereby reducing computation time and improving solution efficiency.

[0097] 3. To address the negative impact of time delay conditions, this invention integrates a buffer and a delay compensator on the controller, improving the control performance of heterogeneous commercial vehicle fleets under delay conditions, thereby achieving stable control of heterogeneous queues under non-ideal communication conditions. Attached Figure Description

[0098] Figure 1 This is a flowchart of the multi-objective optimization control method for heterogeneous commercial vehicle fleets based on distributed model predictive control in this invention.

[0099] Figure 2 This is a force diagram of the longitudinal motion of a commercial vehicle in this invention.

[0100] Figure 3 This is a flowchart of the simulated annealing and particle swarm optimization algorithms used in this invention to solve the objective function. Detailed Implementation

[0101] This invention proposes a multi-objective optimization control method for heterogeneous commercial vehicle fleets based on distributed model predictive control, combined with... Figures 1-3 Further explanation of the invention:

[0102] Figure 1 This is a flowchart of a multi-objective optimization control method for heterogeneous commercial vehicle fleets based on distributed model predictive control. It includes the following steps:

[0103] S1: Establish the dynamic model of each vehicle in the commercial vehicle queue and the queue system model;

[0104] S2: Design a controller for a heterogeneous commercial vehicle queue based on a distributed model predictive control method, and establish the cost function and constraints;

[0105] S3: Add a delay compensation component to the control system under non-ideal communication conditions;

[0106] S4: The objective function is solved using simulated annealing and particle swarm optimization algorithms.

[0107] Step 1: Establish the dynamic model of each vehicle in the commercial vehicle platoon and the platoon system model. Details are as follows:

[0108] The forces acting on a commercial vehicle during forward motion are as follows: Figure 2 As shown, acceleration can be expressed as an equation representing driving power, braking force, and drag. According to Newton's second law, acceleration a... i (t), driving force F i,T (t) and resistance F i,r The relationship between (t) can be expressed as:

[0109]

[0110] Due to driving force Equation (1) can be rewritten as follows

[0111]

[0112] In the formula, m i Let i be the mass of the i-th vehicle. g For the speed ratio of the reducer, r w,i Let T be the wheel rolling radius of the i-th vehicle. i (t) represents the output torque of the motor of the i-th vehicle, η m,i Let F be the mechanical efficiency of the transmission system of the i-th vehicle, and let F be the resistance. i,r (t) consists of rolling resistance, wind resistance, and ramp resistance, and can be expressed as:

[0113]

[0114] In the formula, g and f are the acceleration due to gravity and the rolling coefficient, respectively, and α is the slope on which the vehicle is located. d,i Z is the drag coefficient of the i-th vehicle, r is the air density, and Z is the air resistance coefficient. i The frontal area of ​​the i-th vehicle.

[0115] Let the system state variable x i (t) represents the vehicle displacement S i (t), vehicle speed v i (t) and motor output torque T i (t), i.e., x i (t)=[S i (t),v i (t),T i (t)] T Control quantity u i (t) represents the vehicle motor torque, i.e., u i (t)=T i (t), the state equation is

[0116]

[0117] In the formula, N represents the number of vehicles.

[0118] Equation (4) can be rewritten as

[0119]

[0120] where i is the corresponding number of each vehicle; S i (t) is the displacement of the ith vehicle; v i (t) is the speed of the ith vehicle; T i (t) is the motor output torque of the ith vehicle; e i (t) is the motor desired torque of the ith vehicle; t i is the time delay parameter of the power system of the ith vehicle.

[0121] Thus, the discretized dynamic equation is obtained, where At is the time step of each step calculation.

[0122]

[0123] Step two, design the controller of the heterogeneous commercial vehicle platoon based on the distributed model predictive control method, establish the cost function and the constraint condition. Specifically as follows:

[0124] According to the longitudinal dynamics modeling of the commercial vehicle platoon, a sub-predictive optimization problem is defined for each vehicle in the heterogeneous vehicle platoon, and the information transmitted by the communication topology structure is used to optimize the solution to obtain the control input of the vehicle.

[0125] Let N p be the prediction horizon length of DMPC, and define the following 3 types of variables.

[0126] (1) System prediction variables:

[0127] the predicted position of the ith vehicle at t+n;

[0128] the predicted speed of the ith vehicle at t+n;

[0129] the predicted acceleration of the ith vehicle at t+n;

[0130] the predicted control input of the ith vehicle at t+n;

[0131] the predicted control output of the ith vehicle at t+n.

[0132] (2) Optimal variables:

[0133] the optimal control input of the ith vehicle at t+n.

[0134] (3) Assumed variables:

[0135] the assumed position of the ith vehicle at time t+n;

[0136] the assumed speed of the ith vehicle at time t+n;

[0137] the assumed acceleration of the ith vehicle at time t+n;

[0138] the assumed control input of the ith vehicle at time t+n;

[0139] the assumed control output of the ith vehicle at time t+n.

[0140] The cost function of vehicle i in the platoon is established as follows.

[0141] (1) Following error cost function

[0142] First, the following error cost function of the ith vehicle (i = 2, 3, 4, …, N) and the lead vehicle is established. Let the system output be where is the predicted position of the ith vehicle at time t+k, is the predicted speed of the ith vehicle at time t+k, is the predicted motor torque of the ith vehicle at time t+k, and the expected state of the lead vehicle is where is the assumed position of the lead vehicle at time t+k, d is the vehicle spacing, is the assumed speed of the lead vehicle at time t+k, a is the assumed motor torque of the lead vehicle at time t+k, then the following error cost function is:

[0143]

[0144] where: A i is the error weight coefficient matrix of the ith vehicle and the lead vehicle, take A i = diag(10, …, 10).

[0145] Secondly, the following error cost function of the ith vehicle (i = 2, 3, 4, …, N) and the preceding vehicle is established. Similarly, the expected state of the preceding vehicle is defined as where is the assumed position of the preceding vehicle at time t+k, d is the vehicle spacing, is the assumed speed of the preceding vehicle at time t+k, is the assumed motor torque of the lead vehicle at time t+k, then the following error cost function is:

[0146]

[0147] where B i is the error weight coefficient matrix of the ith vehicle and the front vehicle, and B i = diag(5,...,5).

[0148] (2) Economic cost function

[0149] J 3,i (k|t) = ||C i P i (k|t) ·△t||2 (9)

[0151] where C i is the energy consumption weight coefficient matrix of the ith vehicle, and C i = diag(10,...,10); P i (k|t) is the motor power of the ith vehicle.

[0152] (3) Comfort cost function

[0153] The comfort of vehicles in the queue is ensured by controlling the torque change rate to be at a small level. Therefore, the vehicle comfort cost function is:

[0154]

[0155] where D i is the comfort weight coefficient matrix of the ith vehicle, and D i = diag(8,...,8); is the vehicle torque of the vehicle running at a constant speed at a vehicle speed v , as shown in equation (11).

[0156]

[0157] The formation cooperative DMPC controller can be obtained from equations (7)-(10) as follows:

[0158]

[0159] where s.t. represents the constraint condition, J i (t) is the objective function of the ith vehicle (i = 1, 2, 3,..., N); v min is the minimum vehicle speed specified for commercial vehicles running on the highway; v max is the maximum vehicle speed specified for commercial vehicles running on the highway; T min is the minimum torque that the on-board motor can reach; T max is the maximum torque that the on-board motor can reach; Np Predicted horizon length for DMPC.

[0160] The constraint is written as a penalty function C(k):

[0161]

[0162] where C i (k) is the penalty function for the ith vehicle (i = 1, 2, 3, …, N).

[0163] Step three, add a delay compensation link to the control system under non-ideal communication conditions.

[0164] The delay compensation link is composed of a buffer, a compensator and a DMPC module. The buffer receives data from other vehicles according to the information topology, and selects the latest data by comparing with the buffered data. The delay compensator predicts the hypothetical state with delay, and obtains the estimated value of the hypothetical state within the prediction horizon. In the DMPC module, the nonlinear longitudinal dynamics model of the vehicle is established, the control input is calculated, and the hypothetical state of the ego vehicle is sent to the adjacent vehicles. This control structure is suitable for all controlled vehicles in the platoon. Next, the buffer and compensator that handle time delay are designed.

[0165] When transmitting the states of the preceding vehicle and adjacent vehicles using a non-ideal communication network, the characteristics such as time delay and packet loss have a great impact on the performance of the platoon. When packet loss occurs, the vehicle cannot receive information. Therefore, it can be regarded as an infinite delay, and its unified feature is the timestamp of the transmitted data. The data transmitted between vehicles include timestamp, vehicle code and hypothetical state trajectory, which are generated by DMPC and represent the possible trajectory in the future according to the optimal solution. The prediction horizon of the hypothesis is N p , and there is a certain time delay t ij (t) when vehicle j transmits information to vehicle i through the communication network. Vehicle i receives the hypothetical state sequence w ij ij a (k|t-t ij (t)) of vehicle j sent by vehicle j before t

[0166]

[0167] where: is the hypothetical state of vehicle j, which is defined as the vehicle state of vehicle j at t-t ij (t) + k, obtained by vehicle j performing vehicle control according to the optimal control sequence calculated by vehicle j at t-t ij (t).

[0168] ​The buffer and compensator select and predict the state to obtain the estimated value available for DMPC calculation. In the buffer, at time t, the newly arrived data is compared with the data at initial time t j0 The new data is stored and transmitted to the compensator according to the timestamp of the data, and the old data is deleted. When no data arrives at time t, the buffered data is output

[0169] In the compensator, the error caused by time delay is compensated to obtain the estimated state of the leading vehicle and the adjacent vehicle in the prediction time domain [t, t+N p -1], that is, the estimated state. The estimated state in the prediction range [t, t-t ij (t) + N p -1] is retained, and the part exceeding the prediction range is compensated by the prediction algorithm. For the short-term prediction problem of 1 s time, the kinematic-based prediction method can make accurate prediction at a reasonable calculation speed. In addition, since the accurate dynamics model of the adjacent vehicle cannot be directly obtained, the state prediction using the kinematic model can adapt to most working conditions. Therefore, the constant speed method is adopted to predict the future state of the vehicle. The basic trajectory prediction equation based on the constant speed method is as follows:

[0170]

[0171] In the formula, p(t) is the position of the vehicle; △t is the interval time; and v(t) is the vehicle speed.

[0172] There is a certain error between the estimated state of the other vehicle and the true state, and the error size is affected by the prediction algorithm. However, considering the limitations of vehicle speed, acceleration, torque and prediction time domain, the estimation error can be controlled within a certain range.

[0173] Therefore, there is an upper limit to the estimation error. The estimated state at time t+k can be expressed as:

[0174]

[0175] In the formula, is the estimated nominal state of vehicle j at time t-t ij (t) + k; is the estimated state of vehicle j at time t-t ij (t) + k; is the estimated state prediction value of vehicle j at time t-t ij (t) + k, and its expression is as follows:

[0176]

[0177] ​Subsequently, the vehicle i obtains the assumed state sequence estimate as follows:

[0178]

[0179] where: is the estimate of the assumed state sequence of vehicle j received by vehicle i at time t-t ij (t)+k.

[0180] After the raw data with time delay passes through the buffer, the influence of repeated oscillation of the control input caused by time disorder under time delay can be avoided. The compensator further estimates the current state to reduce the influence of the state error caused by time delay on the controller. In particular, when no new data is received for continuous multiple samplings, the compensator can obtain a reasonable current state estimate according to the data generation time and the prediction algorithm, instead of using an incorrect constant vehicle state. Subsequently, the estimate can be directly used in the DMPC controller to obtain an acceptable error control input.

[0181] Step four, solve the objective function by using the simulated annealing and particle swarm optimization algorithm.

[0182] For the nonlinear and discontinuous target value economic indicators in the objective function, a lookup table is needed to obtain the regenerative braking power under the current speed and braking intensity. The traditional quadratic programming algorithm is difficult to solve. In this case, a heuristic algorithm can solve the problem according to the probability method. The present application uses the simulated annealing and particle swarm optimization algorithm to solve the objective function, and the flow chart is as shown in Figure 3 The specific steps are as follows:

[0183] (1) Initialize each particle parameter. The acceleration value of each vehicle in the vehicle platoon constitutes a particle group p1={a1, a2, …, a N ,}, and the particle group size N is equal to the number of vehicle platoon members. When initialized, set the random update factor c1, the annealing rate λ, the maximum iteration number r max and the minimum fitness value d stop , where d stop is set to a large constant, such as 2000;

[0184] (2) Set the fitness function. The objective function is part of the fitness function; the other part is the constraint set as the penalty function C(k). Calculate and record the fitness function value of the initialized particle under the fitness function, let d stop equal to the fitness value, and record the corresponding particle group as p g (r), r is the iteration number, and p g (r) is composed of the acceleration value of each vehicle at this time.

[0185] (3) Let the initial temperature be the minimum fitness value, i.e. t0=d stop ;

[0186] (4) Use a random update factor c1, an acceleration range and a particle population p g (r) to obtain a new particle population p g (r+1), then r is incremented by 1 until the maximum number of iterations r max is reached, as shown in equation (19);

[0187] p g (r+1) = p g (r) + c1[a min ,a max ] N (19)

[0188] (5) Record the fitness function value d temp at the current temperature T temp . If d stop ≤ d f , the acceleration particle population p temp is accepted and is equal to d g corresponding to the acceleration particle population p temp . If d stop > d f , p g is accepted with a probability of , p f = p g (r+1), or with a probability of , p g remains unchanged, as shown in equations (20) and (21);

[0189] if d temp ≤ d stop : p f = p g (r+1) (20)

[0191] if d temp > d stop :

[0192]

[0193] (6) Return to step (4) until r reaches r max , then let r = 1, p g (r) = p f ;

[0194] (7) Perform a cooling operation T k+1 = λT kwhere λ is the annealing rate, taken as 0.96, T k is the temperature after k annealings;

[0195] (8) return to step (4) after cooling until the lowest temperature is reached, at which point the acceleration particles and global optimum are output.

[0196] Based on the above control method, an embodiment of the present application further provides a control device, which can be a central controller of a vehicle or the like, and which is capable of executing the above control method.

[0197] The above detailed description is only a specific description of the feasible embodiments of the present application, and is not intended to limit the protection scope of the present application. Any equivalent means or changes made without departing from the technical spirit of the present application shall be included in the protection scope of the present application.

Claims

1. A heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions, characterized in that, Comprise the following: S1: Establish the dynamics model of each vehicle in the commercial vehicle queue and the queue system model; The dynamics model of each vehicle in the S1 is as follows: The acceleration is written as an equation of the driving power, the braking force, and the resistance force, and the relationship between the acceleration , the driving force , and the resistance force is expressed as Due to the driving force Equation (1) is rewritten as wherein is the mass of the is the mass of the is the speed ratio of the is the speed ratio of the is the rolling radius of the wheels of the is the rolling radius of the wheels of the is the motor output torque of the is the motor output torque of the is the mechanical efficiency of the driveline of the is the resistance of the vehicle, which is composed of rolling resistance, air resistance, and slope resistance, and is expressed as wherein and are the gravitational acceleration and the rolling coefficient, respectively, is the slope on which the vehicle is located, is the air resistance coefficient of the first vehicle, is the air density, is the frontal area of the first vehicle; S2: Design the controller of the heterogeneous commercial vehicle queue based on the distributed model predictive control method, establish the cost function and the constraint condition; The implementation of the S2 comprises defining three types of variables, as follows: (1) System prediction variables: : the first vehicle at the predicted position of the vehicle at the moment : the first vehicle at the predicted speed of the vehicle at the moment : the first vehicle at the moment of the prediction of the acceleration of the second vehicle at the moment of the prediction of the acceleration of the second vehicle : first vehicle at moment of time : the first vehicle at the moment of the predicted control output; (2) Optimal variables: : first : optimal control input for the vehicle at the moment; (3) Assumed variables: : first : second : third : the vehicle at the assumed speed at the instant : the : the : the : first : second : third : first vehicle at the moment of assumed control output; The cost function in the S2 comprises the following: (1) Following error cost function First, a first vehicle and a following error cost function of the lead vehicle, ; Let the system output be where is the predicted position of the ith vehicle at time , is the predicted speed of the ith vehicle at time , is the predicted motor torque of the ith vehicle at time , and the desired state of the lead vehicle is , where is the assumed position of the lead vehicle at time , d is the inter-vehicle distance, is the assumed speed of the lead vehicle at time , is the assumed motor torque of the lead vehicle at time , then the following-car error cost function with the lead vehicle is: In the formula: For the first The error weighting coefficient matrix between the vehicle and the lead vehicle is taken as follows: ; Secondly, establish the first vehicle ( The following error cost function with respect to the vehicle in front; the desired state with respect to the vehicle in front is defined as follows. ,in For the car in front The assumed position at a given time, where d is the vehicle spacing. For the car in front The assumed velocity at time [time] For the lead car in Assuming the motor torque at a given time, the following error cost function with respect to the vehicle in front is: In the formula: For the first The error weighting coefficient matrix between the vehicle and the vehicle in front is taken as follows: ; (2) Economy cost function In the formula: For the first The energy consumption weighting coefficient matrix of the vehicle, taking ; For the first The power of the vehicle's motor; (3) Comfort cost function The vehicle comfort cost function is ensured by controlling the torque change rate to be at a small level to ensure the comfort of each vehicle in the queue when driving, and thus the vehicle comfort cost function is: In the formula, is the comfort weight coefficient matrix of the vehicle, and is taken as ; ; is the vehicle torque when the vehicle travels at a constant speed at the vehicle speed, as shown in formula (11): ​ S3: Add a delay compensation link to the control system; S4: Solve the objective function by using the simulated annealing and particle swarm optimization algorithm.

2. The heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions according to claim 1, characterized in that, The queue system model in the S1 is as follows: Let the system state variable be the vehicle displacement, vehicle speed and motor output torque, that is, the state variable , the control variable is the vehicle motor torque, that is , Then the state equation is Rewrite formula (4) as In the formula: is the corresponding number for each vehicle; is the displacement of the th vehicle; is the speed of the th vehicle; is the actual torque of the electric machine of the th vehicle; is the desired torque of the electric machine of the th vehicle; is the time delay parameter of the power system of the th vehicle; The dynamics equations of the system are obtained in discrete form, where Time step calculated for each step: 。 3. The heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions according to claim 1, characterized in that, The controller of the heterogeneous commercial vehicle queue in the S2 is as follows: In the formula: For the i-th car ( The objective function; The minimum speed limit for commercial vehicles on highways; The maximum speed limit for commercial vehicles on highways; This is the minimum torque that the vehicle motor can achieve; This is the maximum torque that the vehicle motor can achieve. This represents the prediction time domain length for DMPC.

4. The heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions according to claim 3, characterized in that, The constraint condition of the S2 is designed as a penalty function, as follows: where: is the penalty function for the i-th vehicle (vi). is the penalty function for the i-th vehicle (vi).

5. The heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions according to claim 3, characterized in that, The delay compensation link of the S3 comprises a buffer, a compensator and a DMPC controller module; the buffer receives data from other vehicles according to the information topology, and selects the latest data by comparing with the buffered data; the compensator predicts the assumed state with delay, and obtains the estimated value of the assumed state within the prediction horizon; in the DMPC module, the nonlinear longitudinal dynamics model of the vehicle is established, the control input is calculated, and the assumed state of the ego vehicle is sent to the adjacent vehicle; the specific process is as follows: The data transmitted between vehicles includes timestamps, vehicle codes, and assumed state trajectories, which are generated by the DMPC controller and represent future possible trajectories according to the optimal solution; the assumed prediction domain is , the vehicle transmits information to the vehicle through a communication network, and there is a certain time delay , the vehicle receives the vehicle assumed state sequence sent by the vehicle in the past , and the vehicle assumed state sequence is: wherein: is the assumed state of the vehicle defined as the vehicle If the vehicle is controlled according to the optimal control sequence calculated at the time instant the vehicle state at the prediction horizon is obtained. The buffer and compensator select and predict the state to obtain an estimated value that can be used in the DMPC controller calculation; in the buffer, at time The newly arrived data is compared with the cached data arriving at the initial time , and the new data is stored and transmitted to the compensator according to the time stamp of the data, while the old data is deleted; when no data arrives at time t, the cached data is output. In the compensator, the error caused by time delay is compensated to obtain a predicted time domain The assumed state estimation of the preceding vehicle and the adjacent vehicle, i.e. the assumed state estimation; the assumption The assumed state within the prediction range is compensated by the prediction algorithm, and the part exceeding the prediction range is compensated. The future state of the vehicle is predicted by using the constant speed method, and the basic trajectory prediction equation based on the constant speed method is as follows: In the formulae: is the position in which the vehicle is located; is the speed of the vehicle; There is a certain error between the estimated state and the true state of other vehicles, and the estimated error is controlled within a certain range considering the speed, acceleration, torque and the limitation of prediction time domain, so there is an upper limit for the estimated error, The assumed state estimation at the time can be expressed as: where: is the vehicle at the assumed nominal state at time is the assumed state at time is the vehicle at the assumed state at time is the assumed state at time is the vehicle at the assumed state prediction at time is the assumed state prediction at time Subsequently, the vehicle The assumed state sequence estimate is obtained as follows: In the formula: a vehicle a received vehicle at an estimated value of the self-vehicle assumed state sequence transmitted at the time instant The estimated value for the DMPC controller to obtain an error-acceptable control input.

6. The heterogeneous commercial vehicle fleet multi-objective optimization control method under time delay conditions according to claim 1, characterized in that, The implementation of the S4 comprises: S4.1 initialize each particle parameter; the acceleration value of each vehicle in the platoon constitutes a particle group , the particle group size N is equal to the number of platoon members, and a random update factor is set during initialization , annealing rate , maximum iteration number , and minimum fitness value , such as 2000; S4.2 Set up fitness function; take objective function as part of fitness function, another part is penalty function of constraint set; calculate and record fitness value of initialization particle under fitness function, equal to the fitness value, record corresponding particle population as , iteration number, from which the acceleration value of each vehicle is composed; S4.3 Let the initial temperature t0 be the minimum fitness value, i.e. ; S4.4 Utilizing a random update factor acceleration range and particle population resulting in a new set of particle populations subsequently incrementing by one until a maximum number of iterations is reached as shown in equation (19); S4.5 record current temperature fitness function value under if acceleration particle population accept and equal to corresponding acceleration particle population if then with probability accept such that or with probability accept unchanged, as in equations (20)-(21); S4.6 returns to step S4.4 until reached and then let , ; S4.7 performing a cooling operation wherein is the annealing rate, is the annealing rate, is the temperature after the n-th annealing. S4.8 After cooling, return to step S4.4 until the lowest temperature is reached, at which time the acceleration particles and the global optimal value are output.

7. A multi-objective optimization control device for a heterogeneous commercial vehicle fleet, characterized in that, The control device can perform the control method of claim 1.

Images