CN105501216A - Internet of vehicles based hierarchical energy management control method for hybrid vehicle - Google Patents

Abstract

The invention discloses an internet of vehicles based hierarchical energy management control method for a hybrid vehicle. An upper controller acquires a range of a target speed of the hybrid vehicle on the basis of traffic signal light timing and sets the target speed of the hybrid vehicle as an upper limit of the range of the target speed, and a fast model prediction algorithm is provided for predicating an optimal target speed sequence within a given time window. Compared with the prior art, the internet of vehicles based hierarchical energy management control method for the hybrid vehicle has the advantages that by predication of the optimal target speed of the hybrid vehicle on the basis of traffic signal light timing, stopping or collision of the hybrid vehicle at a red light can be effectively avoided; by acquisition of the optimal target speed sequence according to the fast model prediction algorithm, time cost of model prediction is reduced on the premise that predicated data precision and fuel economy of the hybrid vehicle are guaranteed, and program execution efficiency is effectively improved.

Description

Based on the layering energy management control method of the hybrid vehicle of car networking

Technical field

The present invention relates to a kind of hybrid vehicle energy management control method, in particular a kind of layering energy management control method of the hybrid vehicle based on car networking.

Background technology

The energy management control method of hybrid vehicle directly affects the dynamic property of car load, economy, traveling comfort and discharge, is the Focal point and difficult point of field of hybrid electric vehicles research.At present, what realized commercial application is rule-based control method, but rule-based control method relies on expertise, does not possess good adaptability for working condition, and therefore, scholars' primary study is based on the control method optimized.

Global optimization and instantaneous optimization two kinds is mainly comprised based on the control method optimized.In global optimization approach, dynamic programming (dynamicprogramming, DP), quadratic programming (quadraticprogramming, and the classic calculus of variations (classicalvariationalmethod QP), etc. VP) method all needs known state of cyclic operation, and in the process of automobile actual travel, state of cyclic operation is unknown.Wherein, dynamic programming based on the reverse optimizing of bellman principle is the best global optimization approach of search capability, but program organization is very complicated, and online optimizing needs to adopt model prediction (modelpredictivecontrol, MPC) algorithm obtains state of cyclic operation, add the time cost of calculating, the requirement of discontented full car application, and research shows, traditional model prediction algorithm can only be used in the dynamic system of " enough slow ", inapplicable for the hybrid power system of on-line optimization.In order to save the time that program is run, scholars have studied the instantaneous optimization algorithm based on minimal principle, as equivalent fuel oil consumption minimum principle (equivalentconsumptionminimizationstrategy, ECMS) and Pang Teyajin minimal principle (Pontryagin ' sminimumstrategy, PMP) etc., these algorithms can obtain approximate globally optimal solution and not need known state of cyclic operation.But relative to the arithmetic capability of vehicle-mounted micro controller system, these algorithms still cannot realize real-time control.The object of current hybrid vehicle energy management research is a car, and namely energy management optimized algorithm can only ensure that the economy of single vehicle is optimum, for many hybrid vehicle energy management optimization problems in traffic system, there is no relevant report at present.In addition, at present for the control algorithm of single hybrid vehicle energy management and optimization do not consider traffic signal impact and influencing each other between car and car, the optimal solution under energy management optimal solution now and actual conditions has a certain distance.

Along with the development of intelligent transportation system, the real-time optimal control adopting car networking technology to solve many hybrid vehicles in intelligent transportation system becomes possibility.Based on the hybrid vehicle layering energy management control method of car networking, new thinking can be provided for the problem of the real-time power management and optimization that solve many hybrid vehicles.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of layering energy management control method of the hybrid vehicle based on car networking is provided.

In order to solve the problems of the technologies described above, the present invention adopts following technical scheme: a kind of layering energy management control method of the hybrid vehicle based on car networking, comprises the steps:

Step (1), based on car networked environment, hybrid vehicle carries out car by Dedicated Short Range Communications (DSRC), RF identification (RFID), bluetooth, ZIGBEE or WI-FI and to communicate with car and car is implemented to communicate with traffic;

Step (2), set up top level control device math modeling math modeling, comprise the steps:

(21) car load Longitudinal Dynamic Model is set up

(22) utilize traffic signal lamp timing, obtain the scope of target vehicle speed and the upper limit of the target vehicle speed setting hybrid vehicle vehicle speed range for this reason;

(23) based on the target vehicle speed of hybrid vehicle, the optimal objective speed of a motor vehicle sequence of accelerated model prediction algorithm prediction specified time window is adopted;

(24) optimal objective speed of a motor vehicle sequence is fed back to the chaufeur of each car by transmission over radio form, chaufeur carries out accelerating or braking according to optimal objective speed of a motor vehicle sequence;

Step (3), lower floor's controller are according to acceleration or braking information, obtain optimum torque or the power division of current time driving engine and motor, and optimum torque or power division instruction are sent to engine controller, electric machine controller, gearbox control and electrokinetic cell controller by transmission over radio;

Step (4), each power part controller control corresponding power part according to the control command received and perform relevant output function, and the actual output feedack of power part is carried out Closed-cycle correction to lower floor's controller.

As the further optimization of such scheme, the car load Longitudinal Dynamic Model set up in step (21), as formula (1):

$\left\{\begin{array}{c}{\stackrel{\·}{x}}_i=f_i\left(x_i,u_i\right)\\ f_i\left(x_i,u_i\right)=\left[\begin{array}{c}{v}_i\\ -\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2-\mathrm{\μ}g-g\mathrm{\θ}+u_i\end{array}\right]\\ x_i=\[s_i,v_i\]\end{array}\right.---\left(1\right)$

In formula (1), x _iit is the state vector of i-th car; s _ibe the position of i-th car, state with coordinate; v _ibe the speed of i-th car, unit is m/s; u _ibe the control variable of i-th car, namely any time unit mass tractive force or braking force, unit is N/kg; M _ithe quality of i-th car, unit is kg; C _dfor controlling drag coefficient; ρ _afor density of air, unit is kg/m ³; A _fibe the wind area of i-th car, unit is m ²; μ is coefficient of rolling resistance; θ is the gradient;

As the further optimization of such scheme, in step (22), utilize traffic signal lamp timing, obtain the upper and lower bound of scope of target vehicle speed and the upper limit of target setting speed of a motor vehicle scope for this reason, as formula (2):

In formula (2), v _ilfor the lower limit of target vehicle speed, unit is m/s; v _ihfor the upper limit of target vehicle speed, unit is m/s; d _ia(t _d) be the position s of i-th car _iwith the distance of traffic signal lamp a, unit is m; K _wfor the cycle number of signal lamp, round numbers; t _g, t _rbe respectively the time length of red light and green light, unit is s; t _cbe the time in a traffic lights cycle, unit is s; t _dfor the time of running car, unit is s; v _imaxbe the maxim that i-th hybrid electric vehicle sails speed, unit is m/s; v _iobjfor hybrid vehicle target vehicle speed, unit is m/s;

As the further optimization of such scheme, according to the scope of target vehicle speed, obtain the control variable u of t i-th car _ithe constraint condition of (t), shown in (3), control variable u _it () meets constraint condition, target vehicle speed v _iobjnamely [v is limited at _il, v _ih] scope in, hybrid vehicle can avoid red parking;

$\left\{\begin{array}{c}\[\frac{v_{il}\left(t_d\right)-v_i(t-1)}{\mathrm{\δ}t}\]\≤a_i\left(t\right)\≤\[\frac{v_{ih}\left(t_d\right)-v_i(t-1)}{\mathrm{\δ}t}\]\\ a_i\left(t\right)=-\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2\left(t\right)-\mathrm{\μ}g-g\mathrm{\θ}+u_i\left(t\right)\\ v_i\left(t\right)\≤v_{ih}\left(t_d\right)\\ u_{i\min }\≤u_i\left(t\right)\≤u_{imax}\end{array}\right.---\left(3\right)$

In formula (3), u _imin, u _imaxbe respectively the minimum of control variable and maxim, unit is N/kg; δ t is material calculation, and unit is s; a _it () is the longitudinal acceleration of current time i-th car, unit is m/s ²; u _it () is the control variable of t i-th car, unit is N/kg;

Based on the Forecasting Methodology of the optimal objective speed of a motor vehicle sequence of the hybrid vehicle of car networking, comprise the steps:

Step4A: the majorized function setting up accelerated model prediction, and adopt the majorized function that Novel Algorithm solving model is predicted, the majorized function of described accelerated model prediction, as represented by formula (5):

$\left\{\begin{array}{c}\underset{u_i\left(t\right)}{\min }\left\{\underset{t=t_d}{\overset{t_d+T-1}{\mathrm{\Σ}}}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\α}}_1{\[v_i\left(t\right)-v_{iobj}\left(t_d\right)\]}^2+{\mathrm{\α}}_2\left(t\right)S_{ij}{\left(t\right)}^2+{\mathrm{\α}}_3{\[u_i\left(t\right)-u_{id}\left(t\right)\]}^2\right\}\\ S_{ij}=S_0+t_h{v}_i\left(t\right)-\[s_j\left(t\right)-s_i\left(t\right)\]\\ u_{id}\left(t\right)=\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i{\left(t\right)}^2+\mathrm{\μ}g+g\mathrm{\θ}\\ v_{il}\left(t_d\right)\≤v_i\left(t\right)\≤v_{ih}\left(t_d\right)\\ u_{i\min }\≤u_i\left(t\right)\≤u_{i\max }\end{array}\right.---\left(5\right)$

In formula (4), T is given time window, and unit is s; S _ijbe the distance of i-th car and a jth car, unit is m; α _i(i=1,2,3) are weights coefficient, u _idt () is desirable control variable, unit is N/kg; ; s _i(t) and s _jt () is respectively i-th car and the position of a jth car when time t, represent with coordinate; t _hfor the interval time of front and back two car preset, unit is s; S ₀for the safety distance preset, unit is m; v _iminfor the minimum value of automobile driving speed, unit is m/s; N is the quantity of hybrid vehicle in fleet;

Step4B: the hybrid vehicle Longitudinal Dynamic Model Longitudinal Dynamic Model of hybrid vehicle being converted into linear forms, shown in (6),

$\left\{\begin{array}{c}{x}_i(k+1)=A_i\left(x_i\right)x_i\left(k\right)+B_i{u}_i\left(k\right)\\ A_i\left(x_i\right)x_i\left(k\right)=\left[\begin{array}{c}{v}_i\\ -\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2-\mathrm{\μ}g-g\mathrm{\θ}\end{array}\right]\\ B_i={\left[\begin{array}{cc}0& 1\end{array}\right]}^T\end{array}\right.---\left(6\right)$

In formula (6), A _i(x _i) be the parameter matrix relevant to state; B _iit is a constant column matrix; K is time step, and unit is s;

Step4C: the majorized function majorized function of the model prediction described in step4A being changed into the model prediction of quadratic programming form, shown in (7),

$\left\{\begin{array}{c}min\[\frac{1}{2}{(y_i-y_{iobj})}^T{Q}_i(y_i-y_{iobj})\]\\ P_i{y}_i\≤q_i\\ C_i{y}_i=b_i\\ y_i={\[u_i\left(k\right),x_i(k+1),u_i(k+1),x_i(k+2)...,u_i(k+T-1),x_i(k+T)\]}^T\end{array}\right.---\left(7\right)$

In formula (7), y _ifor comprising the state variable of target vehicle speed and desirable control variable; Q _ifor diagonal matrix; y _iobjfor the expected value of state variable, P _i, q _i, C _i, b _ibe the matrix of coefficient relevant to state variable;

Step4D: the Lagrangian fit solution formula obtaining the majorized function of the model prediction of the quadratic programming form described in step4C, shown in (8),

$L_i(y_i,{\mathrm{\λ}}_i,v_i)=\frac{1}{2}{(y_i-y_{iobj})}^T{Q}_i(y_i-y_{iobj})+{\mathrm{\λ}}_i(C_i{y}_i-b_i)+v_i(P_i{y}_i-q_i)---\left(8\right)$

In formula (8), λ _iand υ _ifor Lagrange multiplier;

Step4D1: single order Ku En-Plutarch (KKT) the optimal conditions equation obtaining Lagrangian fit solution formula, shown in (9),

$\left\{\begin{array}{c}{F}_i\left(y_i,{\mathrm{\λ}}_i,{\mathrm{\ν}}_i\right)=\left[\begin{array}{c}{Q}_i\left(y_i-y_{iobj}\right)+C_i^T{\mathrm{\λ}}_i+P_i^T{v}_i\\ P_i{y}_i+s_i-q_i\\ C_i{y}_i-b_i\\ {\mathrm{\γ}}_i{S}_{i}e\end{array}\right]=0\\ s_i=q_i-P_i{y}_i\\ {\mathrm{\γ}}_i=diag\left({\mathrm{\ν}}_i\right)\\ S_i=diag\left(s_i\right)\end{array}\right.---\left(9\right)$

In formula (9), s _ifor slack variable; γ _iand S _ibe respectively Lagrange multiplier ν _iwith slack variable s _ithe column vector of the elements in a main diagonal composition; E is unit column vector;

Step4D2: adopt Newton iteration method to solve single order Ku En-Plutarch optimal conditions equation, the iterative equation of Newton iteration method is such as formula shown in (10):

$\left[\begin{array}{c}{y}_i\left(t+1\right)\\ {\mathrm{\λ}}_i\left(t+1\right)\\ {\mathrm{\ν}}_i\left(t+1\right)\\ s_i\left(t+1\right)\end{array}\right]=\left[\begin{array}{c}{y}_i\left(t\right)\\ {\mathrm{\λ}}_i\left(t\right)\\ {\mathrm{\ν}}_i\left(t\right)\\ s_i\left(t\right)\end{array}\right]+{\mathrm{\β}}_i\left[\begin{array}{c}{\mathrm{\Δy}}_i\left(t\right)\\ {\mathrm{\Δ\λ}}_i\left(t\right)\\ {\mathrm{\Δv}}_i\left(t\right)\\ {\mathrm{\Δs}}_i\left(t\right)\end{array}\right]---\left(10\right)$

In formula (10), [Δ y _iΔ λ _iΔ ν _iΔ s _i] ^tfor the direction of search of described Newton iteration method optimized variable; β _ifor the iteration step length of described Newton iteration method, described iteration step length can ensure Lagrange multiplier and slack variable on the occasion of;

The solving equation of the direction of search of Newton iteration method optimized variable such as formula shown in (11),

$\left[\begin{array}{cccc}{Q}_i& C_i^T& P_i^T& 0\\ C_i& 0& 0& 0\\ P_i& 0& 0& 0\\ 0& 0& 0& 0\end{array}\right]\left[\begin{array}{c}{\mathrm{\Δy}}_i\\ {\mathrm{\Δ\λ}}_i\\ {\mathrm{\Δ\ν}}_i\\ {\mathrm{\Δs}}_i\end{array}\right]=F_i\left(y_i,{\mathrm{\λ}}_i,{\mathrm{\ν}}_i\right)=-\left[\begin{array}{c}{R}_{y_i}\\ R_{{\mathrm{\λ}}_i}\\ R_{{\mathrm{\ν}}_i}\\ R_{s_i}\end{array}\right]---\left(11\right)$

In formula (11), R _yi, R _{λ i}, R _{υ i}, R _sithe residual error of position Kuhn-Tucker condition;

The solving equation of the direction of search of optimized variable in Newton iteration method is reduced to reduced equation through linear change by Step4D3, as shown in (12),

$\left\{\begin{array}{c}\left[\begin{array}{cc}{Q}_i+P_i^T{W}_i^{-2}{P}_i& C_i^T\\ C_i& 0\end{array}\right]\left[\begin{array}{c}{\mathrm{\Δy}}_i\\ {\mathrm{\Δ\λ}}_i\end{array}\right]=\left[\begin{array}{c}{R}_{y_i}+P_i^T{W}_i^{-2}\left(R_{{\mathrm{\ν}}_i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)\\ R_{{\mathrm{\λ}}_i}\end{array}\right]\\ {\mathrm{\γ}}_i^{-1}{R}_{s_i}=W_i^T{W}_i\end{array}\right.---\left(12\right)$

The direction of search of formula (12) is [Δ y _iΔ λ _i] ^t; W _ifor the transformation matrix relevant to slack variable;

Step4D4: the direction of search adopting the equation solution reduced equation based on Qiao Lisiji (Cholesky) factorization, based on the equation that Qiao Lisiji (Cholesky) factorization obtains, shown in (13),

$\left\{\begin{array}{c}{C}_i{\left(Q_i+P_i^T{W}_i^{-2}{P}_i\right)}^{-1}{C}_i^T{\mathrm{\Δ\λ}}_i=C_i{\left(Q_i+P_i^T{W}_i^{-2}{P}_i\right)}^{-1}\[R_{y_i}+P_i^T{W}_i^{-2}\left(R_{{\mathrm{\ν}}_i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)\]-R_{{\mathrm{\λ}}_i}\\ \left(Q_i+P_i^T{W}_i^{-2}{P}_i\right){\mathrm{\Δy}}_i=R_{yi}+P_i^T{W}_i^{-2}\left(R_{\mathrm{\ν}i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)-C_i^T{\mathrm{\Δ\λ}}_i\end{array}\right.---\left(13\right)$

The equation (13) that Step4E obtains according to the iterative equation (10) of described Newton iteration method and Qiao Lisiji (Cholesky) factorization of Step4D4 solves optimum state variable y _i;

Step4F is according to optimum state variable y _isolve the optimal objective speed of a motor vehicle sequence of specified time window.

Compared to existing technology, the beneficial effect of the layering energy management control method of a kind of hybrid vehicle based on car networking provided by the invention is embodied in:

1, the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, car networking technology is incorporated into the energy management of hybrid vehicle, from the angle of macroscopic view, communicated with signal lamp by truck traffic and car, control state and the time gap of traffic signal lamp, minimum for target with the gross energy that hybrid electric vehicle each in flow of traffic consumes, obtain the speed of a motor vehicle sequence of each car optimum.Control method of the present invention can solve field of traffic hybrid vehicle queue of macroscopic view, i.e. the problem of many hybrid vehicle fuel oil consumption optimums.

2, the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention adopts car networking technology, take into full account and the driving intention of real-time Transmission chaufeur, true pavement conditions, the gradient, automobile load and influencing each other between car and car, while reduction fuel oil consumption, the collision between vehicle can be avoided, improve its driving safety.

3, the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, the multi-layer controller structure actv. of design improves the execution efficiency of program.Top level control device can avoid hybrid vehicle red parking by actv. based on traffic signal lamp timing.Utilize the data that real-time truck traffic and car communicate with signal lamp, adopt accelerated model prediction to obtain optimal objective speed of a motor vehicle sequence, reduce the time cost of model prediction under ensureing the prerequisite of predicted data precision, be convenient to real-time control.

4, the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, avoid hybrid vehicle red parking, adopt accelerated model prediction (F-MPC) to realize the optimum speed prediction close with MPC and fuel economy, and be reduced to 7.2 by 100 of MPC the relative computing time of each time step; The WL-ECMS energy management control method that the present invention proposes realizes the good speed of a motor vehicle and follows, and fuel consumption of 100km is suitable with ECMS, and is reduced to 1.48 by 100 of ECMS the relative computing time of each time step.Research technique of the present invention provides new thinking for solving the management of mixed power vehicle real-time power and optimizing.

Accompanying drawing explanation

Fig. 1 is the schematic diagram that the layering energy management of a kind of hybrid vehicle based on car networking of the present invention controls.

Fig. 2 (a)-Fig. 2 (j) is the vehicle speed trajectory schematic diagram that a car-No. ten cars adopt the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention respectively.

Fig. 3 (a)-Fig. 3 (j) be respectively a car-No. ten cars rule-based, based on WL-ECMS, based on the electrokinetic cell SOC track schematic diagram under the control policy of ECMS.

Fig. 4 is be the path curves schematic diagram of a car-No. ten cars.

Detailed description of the invention

Elaborate to embodiments of the invention below, the present embodiment is implemented under premised on technical solution of the present invention, give detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

Fig. 1 is the schematic diagram that the layering energy management of a kind of hybrid vehicle based on car networking of the present invention controls.Based on a layering energy management control method for the hybrid vehicle of car networking, comprise the steps

Step (1), based on car networked environment, hybrid vehicle carries out car by RF identification, bluetooth, ZIGBEE or WI-FI and to communicate with car and car is implemented to communicate with traffic;

Under car networked environment, automobile is by Dedicated Short Range Communications (dedicatedshortrangecommunication, DSRC), the wireless communications method such as RF identification (radiofrequencyidentificationdevices, RFID), bluetooth (Bluetooth), ZIGBEE, Wi-Fi or Cellular Networks (cellularnetwork) realizes truck traffic and car communicates with traffic facilities.Each car is all furnished with communication module, can receive the signal from front truck, rear car, traffic signal lamp and other sender unit in certain distance.

Step (2), top level control device calculate higher limit and the lower limit of target vehicle speed based on traffic signal lamp timing.According to the information obtained that to communicate with signal lamp based on truck traffic and car, adopt accelerated model prediction algorithm, solve the optimal objective speed of a motor vehicle and the optimum speed of a motor vehicle calculated is sent to chaufeur, chaufeur carries out accelerating or braking according to the optimal objective speed of a motor vehicle.

Set up top level control device math modeling, comprise the steps:

(21) car load Longitudinal Dynamic Model is set up

(22) utilize traffic signal lamp timing, obtain the scope of target vehicle speed and the upper limit of the target vehicle speed setting hybrid vehicle vehicle speed range for this reason;

Traffic signal lamp timing refers to the phase place of traffic signal lamp and the opportunity of each phase place appearance, phase place refers to the state of traffic signal lamp, i.e. red light or green light (not considering amber light in this preferred embodiment), the opportunity of signal lamp refers to the moment that red light or green light continue to occur and lasting time.

(23) based on the target vehicle speed of hybrid vehicle, the optimal objective speed of a motor vehicle sequence of accelerated model prediction algorithm prediction specified time window is adopted;

(24) optimal objective speed of a motor vehicle sequence is fed back to the chaufeur of each car by transmission over radio form, chaufeur carries out accelerating or braking according to optimal objective speed of a motor vehicle sequence.

Step (3), lower floor's controller are according to acceleration or braking information, obtain optimum torque or the power division of current time driving engine and motor, and optimum torque or power division instruction are sent to engine controller, electric machine controller, gearbox control and electrokinetic cell controller by transmission over radio;

Step (4), each power part controller control corresponding power part according to the control command received and perform relevant output function, and the actual output feedack of power part is carried out Closed-cycle correction to lower floor's controller.

Wherein, Longitudinal Dynamic Model integrated in top level control device is obtained in step (21), as formula (1):

$\left\{\begin{array}{c}{\stackrel{\·}{x}}_i=f_i\left(x_i,u_i\right)\\ f_i\left(x_i,u_i\right)=\left[\begin{array}{c}{v}_i\\ -\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2-\mathrm{\μ}g-g\mathrm{\θ}+u_i\end{array}\right]\\ x_i=\[s_i,v_i\]\end{array}\right.---\left(1\right)$

In formula (1), x _iit is the state vector of i-th car; s _ibe the position of i-th car, state with coordinate; v _ibe the speed of i-th car, unit is m/s; u _ibe the control variable of i-th car, namely any time unit mass tractive force or braking force, unit is N/kg; M _ithe quality of i-th car, unit is kg; C _dfor controlling drag coefficient; ρ _afor density of air, unit is kg/m ³; A _fibe the wind area of i-th car, unit is m ²; μ is coefficient of rolling resistance; θ is the gradient;

The Longitudinal Dynamic Model of hybrid vehicle is mainly used in the state variable solving automobile, the i.e. automobile position of any time and speed thereof, when other condition is constant, the position of automobile and speed are determined by its tractive force or braking force, for any given vehicle traction or braking force, can can be solved the speed of automobile by Newton's second law, the distance of its form can be solved to rate integrating, i.e. its position.

Utilize traffic signal lamp timing, obtain the upper and lower bound of scope of target vehicle speed and the upper limit of target setting speed of a motor vehicle scope for this reason, as formula (2):

In formula (2), v _ilfor the lower limit of target vehicle speed, unit is m/s; v _ihfor the upper limit of target vehicle speed, unit is m/s; d _ia(t _d) be the position s of i-th car _iwith the distance of traffic signal lamp a, unit is m; K _wfor the cycle number of signal lamp, round numbers; t _g, t _rbe respectively the time length of red light and green light, unit is s; t _cbe the time in a traffic lights cycle, unit is s; t _dfor the time of running car, unit is s; v _imaxbe the maxim that i-th hybrid electric vehicle sails speed, unit is m/s; v _iobjfor hybrid vehicle target vehicle speed, unit is m/s;

Above-mentioned target vehicle speed v _iobjscope to solve be based on traffic signal lamp timing, utilize the timing of traffic signal, hybrid vehicle fully can be avoided to run into red light through traffic lights, namely avoid hybrid vehicle red light to stop, thus reduce its fuel oil consumption.

According to the scope of target vehicle speed, obtain the control variable u of t i-th car _ithe constraint condition of (t), shown in (3), control variable u _it () meets constraint condition, target vehicle speed v _iobjnamely [v is limited at _il, v _ih] scope in, hybrid vehicle can avoid red parking;

$\left\{\begin{array}{c}\[\frac{v_{il}\left(t_d\right)-v_i(t-1)}{\mathrm{\δ}t}\]\≤a_i\left(t\right)\≤\[\frac{v_{ih}\left(t_d\right)-v_i(t-1)}{\mathrm{\δ}t}\]\\ a_i\left(t\right)=-\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2\left(t\right)-\mathrm{\μ}g-g\mathrm{\θ}+u_i\left(t\right)\\ v_i\left(t\right)\≤v_{ih}\left(t_d\right)\\ u_{i\min }\≤u_i\left(t\right)\≤u_{imax}\end{array}\right.---\left(3\right)$

In formula (3), u _imin, u _imaxbe respectively the minimum of control variable and maxim, unit is N/kg; δ t is material calculation, and unit is s; a _it () is the longitudinal acceleration of current time i-th car, unit is m/s ²; u _it () is the control variable of t i-th car, unit is N/kg;

In this formula (3), the scope of the longitudinal acceleration of the current time of i-th car can by the function representation of vehicle speed range, and longitudinal acceleration is also the function of control variable, therefore, can be obtained the constraint condition of control variable by vehicle speed range.

Red light and green light are referred to as traffic signal light condition, and the state of traffic signal lamp can represent with formula (4),

In the solution formula of described signal lamp state, that represent is t _ddivided by t _cthe remainder obtained.

The invention also discloses a kind of hybrid vehicle based on car networking the Forecasting Methodology of optimal objective speed of a motor vehicle sequence.Based on the layering energy management control method of the hybrid vehicle of car networking, based on the hybrid vehicle target vehicle speed that traffic signal lamp timing obtains, the optimal objective speed of a motor vehicle sequence in specified time window is predicted in employing accelerated model prediction algorithm (F-MPC), specifically comprises the steps:

Step4A: the majorized function setting up accelerated model prediction, and adopt the majorized function that Novel Algorithm solving model is predicted, the majorized function of described accelerated model prediction, as represented by formula (5):

$\left\{\begin{array}{c}\underset{u_i\left(t\right)}{\min }\left\{\underset{t=t_d}{\overset{t_d+T-1}{\mathrm{\Σ}}}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\α}}_1{\[v_i\left(t\right)-v_{iobj}\left(t_d\right)\]}^2+{\mathrm{\α}}_2\left(t\right)S_{ij}{\left(t\right)}^2+{\mathrm{\α}}_3{\[u_i\left(t\right)-u_{id}\left(t\right)\]}^2\right\}\\ S_{ij}=S_0+t_h{v}_i\left(t\right)-\[s_j\left(t\right)-s_i\left(t\right)\]\\ u_{id}\left(t\right)=\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i{\left(t\right)}^2+\mathrm{\μ}g+g\mathrm{\θ}\\ v_{il}\left(t_d\right)\≤v_i\left(t\right)\≤v_{ih}\left(t_d\right)\\ u_{i\min }\≤u_i\left(t\right)\≤u_{i\max }\end{array}\right.---\left(5\right)$

In formula (5), T is given time window, and unit is s; S _ijbe the distance of i-th car and a jth car, unit is m; α _i(i=1,2,3) are weights coefficient, u _idt () is desirable control variable, unit is N/kg; ; s _i(t) and s _jt () is respectively i-th car and the position of a jth car when time t, represent with coordinate; t _hfor the interval time of front and back two car preset, unit is s; S ₀for the safety distance preset, unit is m; v _iminfor the minimum value of automobile driving speed, unit is m/s; N is the quantity of hybrid vehicle in fleet;

The difference of the difference of the optimal objective speed of a motor vehicle of hybrid vehicle and actual vehicle speed and target vehicle speed, the relative distance between car and car and actual control variable and desirable control variable has relation, therefore solving of the optimum speed of a motor vehicle is appreciated that the problem of following, namely actual vehicle speed remains unchanged close to target vehicle speed, relative distance between car and car, control variable and desirable control variable be close to being, the actual output speed of a motor vehicle of hybrid vehicle is exactly the optimum speed of a motor vehicle, and the majorized function of therefore accelerated model prediction can represent with above-mentioned formula (5).

Step4B: the hybrid vehicle Longitudinal Dynamic Model Longitudinal Dynamic Model of hybrid vehicle being converted into linear forms, shown in (6),

$\left\{\begin{array}{c}{x}_i(k+1)=A_i\left(x_i\right)x_i\left(k\right)+B_i{u}_i\left(k\right)\\ A_i\left(x_i\right)x_i\left(k\right)=\left[\begin{array}{c}{v}_i\\ -\frac{1}{2M_i}{C}_D{\mathrm{\ρ}}_a{A}_{fi}{v}_i^2-\mathrm{\μ}g-g\mathrm{\θ}\end{array}\right]\\ B_i={\left[\begin{array}{cc}0& 1\end{array}\right]}^T\end{array}\right.---\left(6\right)$

In formula (6), A _i(x _i) be the parameter matrix relevant to state; B _iit is a constant column matrix; K is time step, and unit is s;

Step4C: the majorized function majorized function of the model prediction described in step4A being changed into the model prediction of quadratic programming form, shown in (7),

$\left\{\begin{array}{c}min\[\frac{1}{2}{(y_i-y_{iobj})}^T{Q}_i(y_i-y_{iobj})\]\\ P_i{y}_i\≤q_i\\ C_i{y}_i=b_i\\ y_i={\[u_i\left(k\right),x_i(k+1),u_i(k+1),x_i(k+2)...,u_i(k+T-1),x_i(k+T)\]}^T\end{array}\right.---\left(7\right)$

In formula (7), y _ifor comprising the state variable of target vehicle speed and desirable control variable; Q _ifor diagonal matrix; y _iobjfor the expected value of state variable, P _i, q _i, C _i, b _ibe the matrix of coefficient relevant to state variable;

Step4D: the Lagrangian fit solution formula obtaining the majorized function of the model prediction of the quadratic programming form described in step4C, shown in (8),

$L_i(y_i,{\mathrm{\λ}}_i,v_i)=\frac{1}{2}{(y_i-y_{iobj})}^T{Q}_i(y_i-y_{iobj})+{\mathrm{\λ}}_i(C_i{y}_i-b_i)+v_i(P_i{y}_i-q_i)---\left(8\right)$

In formula (8), λ _iand υ _ifor Lagrange multiplier;

Step4D1: single order Ku En-Plutarch (KKT) the optimal conditions equation obtaining Lagrangian fit solution formula, shown in (9),

$\left\{\begin{array}{c}{F}_i\left(y_i,{\mathrm{\λ}}_i,{\mathrm{\ν}}_i\right)=\left[\begin{array}{c}{Q}_i\left(y_i-y_{iobj}\right)+C_i^T{\mathrm{\λ}}_i+P_i^T{v}_i\\ P_i{y}_i+s_i-q_i\\ C_i{y}_i-b_i\\ {\mathrm{\γ}}_i{S}_{i}e\end{array}\right]=0\\ s_i=q_i-P_i{y}_i\\ {\mathrm{\γ}}_i=diag\left({\mathrm{\ν}}_i\right)\\ S_i=diag\left(s_i\right)\end{array}\right.---\left(9\right)$

In formula (9), s _ifor slack variable; γ _iand S _ibe respectively Lagrange multiplier ν _iwith slack variable s _ithe column vector of the elements in a main diagonal composition; E is unit column vector;

Step4D2: adopt Newton iteration method to solve single order Ku En-Plutarch optimal conditions equation, the iterative equation of Newton iteration method is such as formula shown in (10):

$\left[\begin{array}{c}{y}_i\left(t+1\right)\\ {\mathrm{\λ}}_i\left(t+1\right)\\ {\mathrm{\ν}}_i\left(t+1\right)\\ s_i\left(t+1\right)\end{array}\right]=\left[\begin{array}{c}{y}_i\left(t\right)\\ {\mathrm{\λ}}_i\left(t\right)\\ {\mathrm{\ν}}_i\left(t\right)\\ s_i\left(t\right)\end{array}\right]+{\mathrm{\β}}_i\left[\begin{array}{c}{\mathrm{\Δy}}_i\left(t\right)\\ {\mathrm{\Δ\λ}}_i\left(t\right)\\ {\mathrm{\Δv}}_i\left(t\right)\\ {\mathrm{\Δs}}_i\left(t\right)\end{array}\right]---\left(10\right)$

In formula (10), [Δ y _iΔ λ _iΔ ν _iΔ s _i] ^tfor the direction of search of described Newton iteration method optimized variable; β _ifor the iteration step length of described Newton iteration method, described iteration step length can ensure Lagrange multiplier and slack variable on the occasion of;

The solving equation of the direction of search of Newton iteration method optimized variable such as formula shown in (11),

$\left[\begin{array}{cccc}{Q}_i& C_i^T& P_i^T& 0\\ C_i& 0& 0& 0\\ P_i& 0& 0& 0\\ 0& 0& 0& 0\end{array}\right]\left[\begin{array}{c}{\mathrm{\Δy}}_i\\ {\mathrm{\Δ\λ}}_i\\ {\mathrm{\Δ\ν}}_i\\ {\mathrm{\Δs}}_i\end{array}\right]=F_i\left(y_i,{\mathrm{\λ}}_i,{\mathrm{\ν}}_i\right)=-\left[\begin{array}{c}{R}_{y_i}\\ R_{{\mathrm{\λ}}_i}\\ R_{{\mathrm{\ν}}_i}\\ R_{s_i}\end{array}\right]---\left(11\right)$

In formula (11), R _yi, R _{λ i}, R _{υ i}, R _sithe residual error of position Kuhn-Tucker condition;

The solving equation of the direction of search of optimized variable in Newton iteration method is reduced to reduced equation through linear change by Step4D3, as shown in (12),

$\left\{\begin{array}{c}\left[\begin{array}{cc}{Q}_i+P_i^T{W}_i^{-2}{P}_i& C_i^T\\ C_i& 0\end{array}\right]\left[\begin{array}{c}{\mathrm{\Δy}}_i\\ {\mathrm{\Δ\λ}}_i\end{array}\right]=\left[\begin{array}{c}{R}_{y_i}+P_i^T{W}_i^{-2}\left(R_{{\mathrm{\ν}}_i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)\\ R_{{\mathrm{\λ}}_i}\end{array}\right]\\ {\mathrm{\γ}}_i^{-1}{R}_{s_i}=W_i^T{W}_i\end{array}\right.---\left(12\right)$

The direction of search of formula (12) is [Δ y _iΔ λ _i] ^t; W _ifor the transformation matrix relevant to slack variable;

Step4D4: the direction of search adopting the equation solution reduced equation based on Qiao Lisiji (Cholesky) factorization, based on the equation that Qiao Lisiji (Cholesky) factorization obtains, shown in (13),

$\left\{\begin{array}{c}{C}_i{\left(Q_i+P_i^T{W}_i^{-2}{P}_i\right)}^{-1}{C}_i^T{\mathrm{\Δ\λ}}_i=C_i{\left(Q_i+P_i^T{W}_i^{-2}{P}_i\right)}^{-1}\[R_{y_i}+P_i^T{W}_i^{-2}\left(R_{{\mathrm{\ν}}_i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)\]-R_{{\mathrm{\λ}}_i}\\ \left(Q_i+P_i^T{W}_i^{-2}{P}_i\right){\mathrm{\Δy}}_i=R_{yi}+P_i^T{W}_i^{-2}\left(R_{\mathrm{\ν}i}-{\mathrm{\γ}}_i^{-1}{R}_{s_i}\right)-C_i^T{\mathrm{\Δ\λ}}_i\end{array}\right.---\left(13\right)$

The equation (13) that Step4E obtains according to the iterative equation (10) of described Newton iteration method and Qiao Lisiji (Cholesky) factorization of Step4D4 solves optimum state variable y _i;

Step4F is according to optimum state variable y _isolve the optimal objective speed of a motor vehicle sequence of specified time window.

Lower floor's controller utilizes the chaufeur received to accelerate and braking information, obtains optimum torque or the power division of current time power part, then optimal control instruction is sent to power part controller.Each power part controller controls power part according to control command and performs associative operation, and its actual output feedack is carried out Closed-cycle correction to lower floor's controller.

The optimum speed of a motor vehicle that lower floor's controller obtains according to top level control device, carries out the energy management of hybrid vehicle.

In order to realize the approximate real-time optimal energy management of hybrid vehicle, the present embodiment adopts a kind of ECMS control method of simplification, i.e. WL-ECMS.The groundwork of WL-ECMS is the WillansLine model utilizing driving engine and motor, by the ECMS searching method approximate rule depending on interpolation He table look-up, thus be and rule-based similar algorithm greatly reduce the complexity of program to save the time cost calculated by the ECMS simplify of arithmetic based on search.

Driving engine Willans-Line model is such as formula shown in (14).

P _ef＝a _eP _em+b _e(14)

In formula (14), P _efand P _embe respectively oil inflame power and effective engine power, unit is W; a _eand b _ebe respectively and represent the inverse of indicated efficiency of engine and the coefficient of regression of loss due to friction, be the function of engine speed.

The Willans-Line model of motor is such as formula shown in (15).

$P_{me}=\left\{\begin{array}{cc}{a}_{m1}{P}_{mm}+b_{m1}& P_{mm}>0\\ 0& P_{mm}=0\\ a_{m2}{P}_{mm}+b_{m2}& p_{mm}<0\end{array}\right.---\left(15\right)$

In formula (15), P _me, P _mmbe respectively electrokinetic cell power and motor effective power, unit is W; a _m1and b _m1be respectively coefficient of regression when motor effective power is greater than zero; a _m2and b _m2for coefficient of regression when motor effective power is less than zero.A _m1, b _m1, a _m2, b _m2be the function of motor speed.

In any sampling instant, the hybrid vehicle drive pattern energy management control policy based on WillansLine can represent with formula (16), and the energy management control policy based on ECMS is reduced to and rule-based similar control policy by this formula.

In formula (16), for the optimum net output of motor, unit is W; with be respectively the optimum horsepower output of motor and driving engine, unit is W; P _reqfor chaufeur request power, unit is W.

Electric-only mode and hybrid mode are equilibrium mode, when the drive pattern of hybrid vehicle switches between electric-only mode and hybrid mode, the transformation point of chaufeur request power is the electric-only mode demand power point equal with hybrid mode demand power.When car load demand power is less than critical power, the drive pattern of hybrid vehicle is electric-only mode; Otherwise, be then hybrid mode.

The layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, (this preferred embodiment adopts computing module to be HPDL580 to the large server Palmetto of employing U.S. Clemson University, treater is 24 core IntelXeon7542, RAM is 505G) calculate the optimal objective speed of a motor vehicle of top level control device, and the speed of a motor vehicle after optimizing is saved as Matlab data format, for the off-line hardware-in-the-loop test of lower floor's energy management control method, test platform is dSPACE, and test period is 450s.

In test program, the hybrid electric vehicle that has 10 same models is set in fleet and all on same track; The initial position of automobile is [1211099885706045.566630.229315.91960.8724], and unit is m; The initial speed of a motor vehicle is [14.5161516.316.712.0813.004714.178810.373012.0473], and unit is m/s; The time window of model prediction is 6s, and the step-length of calculating is 0.5s; Signalization lamp quantity is 15, and red light time length is 40s, and green light time length is 15s, and the distance of two traffic signal lamp is 500m; Maximum, the minimum speed of a motor vehicle arranging automobile is respectively 20m/s and 0.The complete vehicle curb weight arranging each car is 1750kg, and wind area is 2.36m ², aerodynamic drag factor is 0.32, and coefficient of rolling resistance is 0.015, and road grade is 0, and engine power is 103kW, and the rating horsepower of motor is 40kW, and peak power is 80kW.[b is set ₀, b ₁, b ₂, b ₃]=[0.1569,0.0245 ,-7.415 × 10 ^-4, 5.975 × 10 ^-5], [c ₀, c ₁, c ₂]=[0.07224,0.09681,1.075 × 10 ^-3].

The vehicle speed trajectory schematic diagram of the layering energy management control method of a car-No. ten cars employings a kind of hybrid vehicle based on car networking of the present invention respectively see Fig. 2 (a)-Fig. 2 (j), Fig. 2 (a)-Fig. 2 (j).Comprise in figure top level control device adopt the contrast of the optimum speed of a motor vehicle of F-MPC and MPC and top level control to be F-MPC and lower floor's controller is WL-ECMS time the speed of a motor vehicle follow curve.From Fig. 2 (a)-Fig. 2 (j), adopt the top level control device optimum prediction speed of a motor vehicle of F-MPC roughly the same with adopting the optimum prediction speed of a motor vehicle of MPC.In addition, for a time step, the relative computing time of top level control device is reduced to 7.2 of F-MPC by 100 of MPC, and computing time, cost significantly reduced.The F-MPC demonstrating the present embodiment proposition can significantly reduce computing time, realize the basis of control in real time realizes the control effects close with MPC.Curve is followed from the speed of a motor vehicle of lower floor's controller, the optimum prediction speed of a motor vehicle of following the speed of a motor vehicle and top level control device based on lower floor's controller of WL-ECMS is consistent substantially, illustrate that the lower floor's controller based on WL-ECMS can ensure that the good speed of a motor vehicle is followed, realize the energy management that hybrid vehicle is basic.

See Fig. 3 (a)-Fig. 3 (j), Fig. 3 (a)-Fig. 3 (j) be respectively No.-ten, a car rule-based-, based on WL-ECMS, based on ECMS control policy strategy under electrokinetic cell SOC track schematic diagram.As seen from the figure, under three kinds of different lower coating control methods, the fluctuation range of electrokinetic cell SOC remains in rational scope, illustrates that three kinds of lower floor's energy management control methods all can realize the equilibrium of power battery for hybrid electric vehicle, realizes the energy management that hybrid vehicle is basic.The SOC track contrasted under lower coating control methods different in each subgraph is known, generally speaking, instantly coating control method be followed successively by rule-based, based on WL-ECMS, based on ECMS time, electrokinetic cell SOC fluctuation range reduces successively, instantly coating control method is followed successively by WL-ECMS with during based on ECMS, and the fluctuation range of SOC is suitable.The lower coating control method illustrating based on WL-ECMS can realize with based on the suitable balancing energy control effects of ECMS, and the control effects of the two is all better than rule-based control method.In addition, for a time step, relative computing time by be reduced to based on 100 of ECMS based on WL-ECMS 1.48, computing time, cost significantly reduced.

Be be the path curves schematic diagram of a car-No. ten cars see Fig. 4, Fig. 4.In figure, the real horizontal line parallel with transverse axis represents red time window, and the white space in the middle of two red time windows represents green time window.As seen from the figure, under the testing program of the present embodiment setting, what hybrid vehicle ran into through traffic signal lamp is green light.Therefore, the present embodiment patent of the present invention based on car networking top level control device can avoid hybrid vehicle red parking by actv..In addition, the geometric locus of ten cars does not have intersection point, illustrate the present embodiment based on car networking top level control device can hybrid vehicle be avoided to collide by actv..

Table 1 hybrid vehicle fuel consumption of 100km

Table 1 is respectively the fuel consumption of 100km of ten hybrid vehicles under the control method of different the upper and lower.As shown in Table 1, when top level control device adopts F-MPC or MPC, the average fuel consumption of 100km based on WL-ECMS is only 2.26% or 3.57% relative to the average fuel consumption of 100km rate of rise of ECMS.When lower floor's controller adopts WL-ECMS or ECMS, the average fuel consumption of 100km based on F-MPC is only 0.96% or 2.25% relative to average hundred kilometers of rate of rises based on MPC.

Can be obtained by this preferred embodiment; of the present invention a kind of layering energy management control method of the hybrid vehicle based on car networking: top level control device is consistent substantially based on F-MPC and the optimum vehicle speed trajectory obtained based on MPC; for a time step; the relative computing time of top level control device is reduced to 7.2 of F-MPC by 100 of MPC, and top level control device adopts F-MPC to differ less with fuel consumption of 100km when adopting MPC.During lower floor controller employing WL-ECMS, electrokinetic cell SOC fluctuation is in rational scope, and speed of a motor vehicle tracking error is less, and fuel consumption of 100km is close to the fuel consumption of 100km based on ECMS.For a time step, the relative computing time based on WL-ECMS lower floor energy management control method can be reduced to 1.48 by 100 of ECMS.By the layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, solve the optimal objective speed of a motor vehicle, hybrid vehicle can avoid red parking by actv. through traffic signal lamp.

The layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, predict the top level control device of the optimum speed of a motor vehicle based on F-MPC and all under significantly reducing program runtime, realizing the real-time prerequisite controlled, the fuel economy that hybrid vehicle is good can be ensured based on lower floor's controller of WL-ECMS.The layering energy management control method of a kind of hybrid vehicle based on car networking of the present invention, for the real-time power management and optimization solving hybrid vehicle provides new thinking.

To those skilled in the art, obviously the invention is not restricted to the details of above-mentioned one exemplary embodiment, and when not deviating from spirit of the present invention or essential characteristic, the present invention can be realized in other specific forms.Therefore, no matter from which point, all should embodiment be regarded as exemplary, and be nonrestrictive, scope of the present invention is limited by claims instead of above-mentioned explanation, and all changes be therefore intended in the implication of the equivalency by dropping on claim and scope are included in the present invention.Any Reference numeral in claim should be considered as the claim involved by limiting.

In addition, be to be understood that, although this specification sheets is described according to embodiment, but not each embodiment only comprises an independently technical scheme, this narrating mode of specification sheets is only for clarity sake, those skilled in the art should by specification sheets integrally, and the technical scheme in each embodiment also through appropriately combined, can form other embodiments that it will be appreciated by those skilled in the art that.

Claims (6)

1., based on a layering energy management control method for the hybrid vehicle of car networking, it is characterized in that, comprise the steps:

Step (1), based on car networked environment, hybrid vehicle carries out car by Dedicated Short Range Communications (DSRC), RF identification (RFID), bluetooth, ZIGBEE or WI-FI and to communicate with car and car is implemented to communicate with traffic;

Step (2), set up top level control device math modeling math modeling, comprise the steps:

(21) car load Longitudinal Dynamic Model is set up;

(22) utilize traffic signal lamp timing, obtain the scope of target vehicle speed and the upper limit of the target vehicle speed setting hybrid vehicle vehicle speed range for this reason;

(23) based on the target vehicle speed of hybrid vehicle, the optimal objective speed of a motor vehicle sequence of accelerated model prediction algorithm prediction specified time window is adopted;

(24) optimal objective speed of a motor vehicle sequence is fed back to the chaufeur of each car by transmission over radio form, chaufeur carries out accelerating or braking according to optimal objective speed of a motor vehicle sequence.

2. the layering energy management control method of the hybrid vehicle based on car networking according to claim 1: also comprise the steps:

Step (3), lower floor's controller are according to acceleration or braking information, obtain optimum torque or the power division of current time driving engine and motor, and optimum torque or power division instruction are sent to engine controller, electric machine controller, gearbox control and electrokinetic cell controller by transmission over radio;

Step (4), each power part controller control corresponding power part according to the control command received and perform relevant output function, and the actual output feedack of power part is carried out Closed-cycle correction to lower floor's controller.

3. the layering energy management control method of the hybrid vehicle based on car networking according to claim 1 and 2, is characterized in that, the car load Longitudinal Dynamic Model set up in step (21), as formula (1):

In formula (1), x _iit is the state vector of i-th car; s _ibe the position of i-th car, state with coordinate; v _ibe the speed of i-th car, unit is m/s; u _ibe the control variable of i-th car, namely any time unit mass tractive force or braking force, unit is N/kg; M _ithe quality of i-th car, unit is kg; C _dfor controlling drag coefficient; ρ _afor density of air, unit is kg/m ³; A _fibe the wind area of i-th car, unit is m ²; μ is coefficient of rolling resistance; θ is the gradient.

4. the layering energy management control method of the hybrid vehicle based on car networking according to claim 1 and 2, it is characterized in that: in step (22), utilize traffic signal lamp timing, obtain the upper and lower bound of scope of target vehicle speed and the upper limit of target setting speed of a motor vehicle scope for this reason, as formula (2):

In formula (2), v _ilfor the lower limit of target vehicle speed, unit is m/s; v _ihfor the upper limit of target vehicle speed, unit is m/s; d _ia(t _d) be the position s of i-th car _iwith the distance of traffic signal lamp a, unit is m; K _wfor the cycle number of signal lamp, round numbers; t _g, t _rbe respectively the time length of red light and green light, unit is s; t _cbe the time in a traffic lights cycle, unit is s; t _dfor the time of running car, unit is s; v _imaxbe the maxim that i-th hybrid electric vehicle sails speed, unit is m/s; v _iobjfor hybrid vehicle target vehicle speed, unit is m/s.

5. the layering energy management control method of the hybrid vehicle based on car networking according to claim 4, is characterized in that: according to the scope of target vehicle speed, obtains the control variable u of t i-th car _ithe constraint condition of (t), shown in (3), control variable u _it () meets constraint condition, target vehicle speed v _iobjnamely [v is limited at _il, v _ih] scope in, hybrid vehicle can avoid red parking;

In formula (3), u _imin, u _imaxbe respectively the minimum of control variable and maxim, unit is N/kg; δ t is material calculation, and unit is s; a _it () is the longitudinal acceleration of current time i-th car, unit is m/s ²; u _it () is the control variable of t i-th car, unit is N/kg.

6., based on the Forecasting Methodology of the optimal objective speed of a motor vehicle sequence of the hybrid vehicle based on car networking according to claim 4, it is characterized in that: comprise the steps:

Step4A: the majorized function setting up accelerated model prediction, and adopt the majorized function that Novel Algorithm solving model is predicted, the majorized function of described accelerated model prediction, as represented by formula (5):

In formula (5), T is given time window, and unit is s; S _ijbe the distance of i-th car and a jth car, unit is m; α _i(i=1,2,3) are weights coefficient, u _idt () is desirable control variable, unit is N/kg; ; s _i(t) and s _jt () is respectively i-th car and the position of a jth car when time t, represent with coordinate; t _hfor the interval time of front and back two car preset, unit is s; S ₀for the safety distance preset, unit is m; v _iminfor the minimum value of automobile driving speed, unit is m/s; N is the quantity of hybrid vehicle in fleet;

Step4B: the hybrid vehicle Longitudinal Dynamic Model Longitudinal Dynamic Model of hybrid vehicle being converted into linear forms, shown in (6),

In formula (6), A _i(x _i) be the parameter matrix relevant to state; B _iit is a constant column matrix; K is time step, and unit is s;

Step4C: the majorized function majorized function of the model prediction described in step4A being changed into the model prediction of quadratic programming form, shown in (7),

In formula (7), y _ifor comprising the state variable of target vehicle speed and desirable control variable; Q _ifor diagonal matrix; y _iobjfor the expected value of state variable, P _i, q _i, C _i, b _ibe the matrix of coefficient relevant to state variable;

Step4D: the Lagrangian fit solution formula obtaining the majorized function of the model prediction of the quadratic programming form described in step4C, shown in (8),

In formula (8), λ _iand υ _ifor Lagrange multiplier;

Step4D1: single order Ku En-Plutarch (KKT) the optimal conditions equation obtaining Lagrangian fit solution formula, shown in (9),

In formula (9), s _ifor slack variable; γ _iand S _ibe respectively Lagrange multiplier ν _iwith slack variable s _ithe column vector of the elements in a main diagonal composition; E is unit column vector;

Step4D2: adopt Newton iteration method to solve single order Ku En-Plutarch optimal conditions equation, the iterative equation of Newton iteration method is such as formula shown in (10):

In formula (10), [Δ y _iΔ λ _iΔ ν _iΔ s _i] ^tfor the direction of search of described Newton iteration method optimized variable; β _ifor the iteration step length of described Newton iteration method, described iteration step length can ensure Lagrange multiplier and slack variable on the occasion of;

The solving equation of the direction of search of Newton iteration method optimized variable such as formula shown in (11),

In formula (11), R _yi, R _{λ i}, R _{υ i}, R _sithe residual error of position Kuhn-Tucker condition;

The solving equation of the direction of search of optimized variable in Newton iteration method is reduced to reduced equation through linear change by Step4D3, as shown in (12),

The direction of search of formula (12) is [Δ y _iΔ λ _i] ^t; W _ifor the transformation matrix relevant to slack variable;

Step4D4: the direction of search adopting the equation solution reduced equation based on Qiao Lisiji (Cholesky) factorization, based on the equation that Qiao Lisiji (Cholesky) factorization obtains, shown in (13),

The equation (13) that Step4E obtains according to the iterative equation (10) of described Newton iteration method and Qiao Lisiji (Cholesky) factorization of Step4D4 solves optimum state variable y _i;

Step4F is according to optimum state variable y _isolve the optimal objective speed of a motor vehicle sequence of specified time window.