CN105774800A - Collision relieving method and device between vehicles in hybrid vehicle queue - Google Patents

Abstract

The invention discloses a collision relieving method and device between vehicles in a hybrid vehicle queue. The collision relieving method between the vehicles in the hybrid vehicle queue includes the following steps that firstly, vehicle condition information of the network-connection vehicles and vehicle condition information of the non-network-connection vehicles in front of and behind the network-connection vehicles are acquired; secondly, the head vehicle and the rear vehicles in the hybrid vehicle queue are sequentially judged, if the corresponding network-connection vehicle is the head vehicle or the rear vehicles, the third step is executed, and if the corresponding network-connection vehicle is neither the head vehicle nor the rear vehicles, the first step is executed again, wherein the head vehicle is the network-connection vehicle with the braking working condition exceeding a set threshold value, and the rear vehicles are vehicles meeting the queue collision relieving requirement behind the head vehicle; thirdly, the vehicle condition information of the network-connection vehicles is received; fourthly, the expected acceleration of each network-connection vehicle in the rear vehicles is planned; fifthly, each network-connection vehicle in the rear vehicles travels according to the planned expected acceleration, and the each network-connection vehicle in the rear vehicles travels according to a vehicle following mode; and sixthly, whether each of the rear vehicles is parked or not is judged, if certain rear vehicles are parked, control is stopped, and if no rear vehicles are parked, the third step is executed again. By the adoption of the method and device, the braking space between the vehicles in the hybrid vehicle queue can be used for control.

Description

Impact-moderation method between vehicle and device in a kind of hybrid vehicle queue

Technical field

The present invention relates to automobile technical field, particularly relate in a kind of hybrid vehicle queue the impact-moderation method between vehicle and device.

Background technology

In recent years, owing to automobile pollution is skyrocketed through, the reasons such as law on road traffic safety rule relatively lag behind, driving technology training examinationization, road traffic bears unprecedented pressure.According to statistics, in China's expressway traffic accident form, collision accident proportion is the highest, almost can reach 66.76%.In in people, three, Che He road, human factors causes that vehicle accident accounts for more than the 90% of total number of accident.Therefore, how effectively to reduce collision accident, be obtained for showing great attention to of government, enterprise and research institution all the time.

Advanced drive assist system (ADAS, AdvancedDriverAssistanceSystems), owing to can overcome the disadvantages that the limitation of human driver own, obtains widely studied and application.Wherein, Active collision avoidance system, as maximally effective collision free accident technology, obtains further investigation and extensive use.This system accurately obtains the information from car and nearby vehicle identification potential safety hazard by advanced sensing technology.Under emergency work condition, system actively takes control for brake, thus collision free accident, it is ensured that driving safety.At present, Active collision avoidance system only by control from car braking avoid and front vehicles or barrier collision, braking decision-making does not consider the impact on front vehicle, in fact it could happen that braking deceleration is excessive, front braked space utilizes the phenomenons such as insufficient, thus causing the generation of rear-end collision accident.Therefore, hybrid vehicle queue impact-moderation technology is arisen at the historic moment.

The fast development of car networking technology is vehicle active safety technologies, and especially hybrid vehicle queue impact-moderation technology brings huge opportunity.In the unobstructed environment of highway and city expressway, by existing car car short-range communication technology and 4G/5G networking technology, it is mutual that many workshops can realize large-scale initiative information.The perception of environment is expanded to nearby vehicle and whole traffic environment by vehicle from the information gathering limited from car sensor, in advance front vehicles emergency can be carried out perception and prediction, and make rational control decision, compensate for because the little braking deceleration caused of traditional sensors information gathering scope is excessive, braked space utilizes the phenomenons such as insufficient, and then avoids the generation of many cars chain of rings collision accident.

Although many cars impact-moderation controls to bring obvious safety improvement under net connection environment, but the formation of networked environment also needs for a long time entirely, before this, there is net connection car and jointly participates in the part car networking stage of traffic circulation with non-connection car of netting.Current car car impact-moderation control method is mostly based on desirably full networked environment scene, and is necessarily subject to the restriction of non-net connection car in the vehicle networked scene of substantial portion.Accordingly, it would be desirable to the impact-moderation method of a kind of hybrid vehicle queue under interconnecting segment environment of design (net connection car participates in jointly with non-net connection car) and device.

Thus, it is desirable to have a kind of technical scheme overcomes or at least alleviates at least one drawbacks described above of prior art.

Summary of the invention

It is an object of the invention to provide in a kind of hybrid vehicle queue the impact-moderation method between vehicle and device overcomes or at least alleviates at least one drawbacks described above of the prior art.

For achieving the above object, the present invention provides a kind of impact-moderation method in hybrid vehicle queue between vehicle, and in described hybrid vehicle queue, the impact-moderation method between vehicle comprises the following steps: 1) gather the vehicle condition information of the adjacent non-net connection car in the net connection vehicle condition information of car and this net connection Herba Plantaginis, rear；2) the head car in hybrid vehicle queue, rear car are sequentially judged, if head car or rear car, then enter step 3), otherwise return step 1), head car is that damped condition exceedes the net connection car setting threshold value, and rear car is the vehicle that a car rear meets queue impact-moderation requirement；3) the vehicle condition information of net connection car is received；4) the expectation acceleration of each net connection car in planning rear car；5) in rear car, each net connection car travels according to the expectation acceleration planned, non-net connection car travels according to car-following model；6) judge whether each rear car stops, if so, then stop controlling；Otherwise, step 3 is returned).

Further, step 4) in, minimum for optimization aim to net connection car critical retardation power summation, the expectation acceleration of each net connection car in planning rear car.

Further, the expectation acceleration a of net connection car_i,desK () is obtained by following formula:

$\underset{a_{di}}{min}\underset{0}{\overset{T}{\∫}}(\mathrm{\Σ}\frac{m_{c\_i}}{S_{c\_i}\left(t\right)}{\[v_{c\_i}\left(t\right)-v_{c\_i-1}\left(t\right)\]}^2+\mathrm{\Σ}\frac{m_0}{S_{c\_j}\left(t\right)}{\[v_{c\_j}\left(t\right)-v_{c\_j-1}\left(t\right)\]}^2)$

Its constraints is:

$\left\{\begin{array}{c}{\stackrel{\·}{x}}_i\left(t\right)=v_i\left(t\right)\\ {\stackrel{\·}{v}}_i\left(t\right)=a_i\left(t\right)\\ {\stackrel{\·}{a}}_i\left(t\right)=\frac{1}{\mathrm{\τ}}(a_{di}\left(t\right)-a_i\left(t\right)),i\∈G\\ {\stackrel{\·}{a}}_i\left(t\right)=0,i\∈B\\ a_{i,\min }\≤a_{di}\left(t\right)\≤a_{i,\max }\end{array}\right.$

Wherein: T is prediction time domain length, G is the set of net connection car in hybrid vehicle queue, and B is the set of non-net connection car in hybrid vehicle queue, and c_i is the sequence number of net connection car, and c_j is the sequence number of non-net connection car, m_{c_i}It is the quality of the c_i net connection car, m₀For the quality of non-net connection car, S_iIt it is the following distance of i-th car and its front truck；x_iT () is the headstock position of i-th car of t, v_iT () is the speed of i-th car of t, a_iT () is the acceleration of i-th car of t, a_diT () is the expectation acceleration of i-th car, τ is the time constant characterizing vehicle operating lag, a_i,minIt is the minimum acceleration of i-th car, a_i,maxIt it is the peak acceleration of i-th car.

Further, step 2) in, the composition of hybrid vehicle queue has the feature that the time headway of two cars adjacent in 1. hybrid vehicle queue is less than preset value；2. trailer is net connection car；3. the net connection car between head car and trailer is random with non-net connection truck position distribution.

Further, the car-following model of non-net connection car is:

$a_{des}=\left\{\begin{array}{cc}0,& TTC>14.68\\ 1.411\&CenterDot;1.04\&CenterDot;({\mathrm{TTC}}^{0.72}-3^{0.72})-6.91,& 3\&le;TTC\&le;14.68\\ -6.918,& TTC<3\end{array}\right.$

In formula, TTC is the collision avoidance time.

The present invention also provides for the impact-moderation device in a kind of hybrid vehicle queue between vehicle, in described hybrid vehicle queue, the impact-moderation device between vehicle includes perception unit, communication unit, the first brak control unit, the second brak control unit and cloud computing platform, wherein: described perception unit, communication unit and the first brak control unit are mounted on net connection car, described second brak control unit is installed on non-net connection car；Described perception unit is for collecting joining the vehicle condition information of the adjacent non-net connection car in Herba Plantaginis, rear from car status information and this net and exporting of respective wire connection car；Described communication unit is used for netting connection car and carries out real-time information interaction with described cloud computing platform；Described cloud computing platform is for receiving the expectation acceleration of each net connection car in the information of described perception unit output, the head car judged in hybrid vehicle queue and rear car and planning rear car and exporting, described head car is that damped condition exceedes the net connection car setting threshold value, and described rear car is the vehicle that a car rear meets queue impact-moderation requirement；Described first brak control unit travels for Controling network connection car desirably acceleration after the expectation acceleration receiving the output of described cloud computing platform；Described second brak control unit is used for controlling non-net connection and travels according to car-following model.

Further, described perception unit includes environmental perception device, acceleration harvester, on-vehicle information harvester and positioner, wherein: described environmental perception device joins the vehicle condition information of car for the non-net that the forward and backward side of perception is adjacent and exports；Described acceleration harvester is used for the perception longitudinal acceleration from car；Described on-vehicle information harvester is used for gathering speed, brake pressure and throttle opening information；Described positioner positions information for gathering from car.

Further, the information conveyance that described perception unit is gathered by 4G/5G radio communication by described communication unit to described cloud computing platform, and for receive the planning of described cloud computing platform from car expectation acceleration.

Further, described cloud computing platform includes communicator, head car judgment means, rear car judgment means and expectation acceleration device for planning, wherein: described communicator is mutual with described communication unit information, the described communication unit and for the expectation planned acceleration is flowed in rear car on each net connection car；Described head car judgment means is for judging described head car according to the damped condition of input；Described rear car judgment means is for receiving the information of described head car, and requires to judge described rear car according to queue impact-moderation, and the vehicle in hybrid vehicle queue is numbered；Described expectation acceleration device for planning, for the driving states information according to net connection car, plans the expectation acceleration of net connection car in each rear car.

Further, described expectation acceleration device for planning includes information fusion module, model prediction module and expectation acceleration calculation module, wherein, the vehicle condition information of the described information fusion module forward and backward side's vehicle for being perceived by described environmental perception device merges, quickly determine forward and backward side's car status information, and flow to described model prediction module；Described model prediction module is predicted for the motion of each described rear car, flows to described expectation acceleration calculation module；Described expectation acceleration calculation module, for the described each rear car motion according to prediction, calculates the expectation acceleration of each rear car and flows to described communicator.

Due to the fact that and take above technical scheme, it has the advantage that 1, the present invention obtains, by 4G/5G radio communication, the non-net connection car status information that in hybrid vehicle queue, all net connection car information and net connection car perceive, minimum for controlling optimization aim to net connection car critical retardation power summation, carry out hybrid vehicle queue impact-moderation and control.Owing to having considered net connection car and non-net connection the influencing each other of car, it is thus possible to effectively utilize the braked space in each workshop in hybrid vehicle queue to be controlled, collision avoid with remission effect on to be much better than the actively collision avoidance of tradition bicycle and control；2, the present invention is based on the 4G/5G communication technology having independent intellectual property right, the short range communication technology that not only the big quantum jump of communication distance is traditional, and improve the singal reporting codes such as access delay, traffic rate, bandwidth, it may be achieved urban road is omnidistance impact-moderation control in travelling；3, the present invention is based on the China Mobile's cloud computing platform (" Herba Cistanches (BigCloud) platform ") having independent intellectual property right, excellent performance on computational efficiency, data mining capability, fully meet " the many aspects of vehicle safety applications standard of National Highway Traffic safety management office and collision avoidance standard cooperative programme organizational choice ", it may be achieved low time delay, highly reliable cloud platform are applied.

During the car car impact-moderation that the present invention can be widely applied under interconnecting segment environment in hybrid vehicle queue controls.

Accompanying drawing explanation

Fig. 1 is the hybrid vehicle queue impact-moderation control method flow chart based on Model Predictive Control provided by the present invention.

Fig. 2 is the structural representation that the hybrid vehicle queue impact-moderation based on Model Predictive Control provided by the present invention is implemented.

Fig. 3 is single unit vehicle hardware architecture diagram.

Detailed description of the invention

For making purpose of the invention process, technical scheme and advantage clearly, below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is further described in more detail.In the accompanying drawings, same or similar label represents same or similar element or has the element of same or like function from start to finish.Described embodiment is a part of embodiment of the present invention, rather than whole embodiments.The embodiment described below with reference to accompanying drawing is illustrative of, it is intended to is used for explaining the present invention, and is not considered as limiting the invention.Based on the embodiment in the present invention, all other embodiments that those of ordinary skill in the art obtain under not making creative work premise, broadly fall into the scope of protection of the invention.Below in conjunction with accompanying drawing, embodiments of the invention are described in detail.

In describing the invention; it will be appreciated that; term " orientation or the position relationship of the instruction such as " center ", " longitudinal direction ", " transverse direction ", "front", "rear", "left", "right", " vertically ", " level ", " top ", " end " " interior ", " outward " be based on orientation shown in the drawings or position relationship; be for only for ease of the description present invention and simplifying and describe; rather than instruction or imply indication device or element must have specific orientation, with specific azimuth configuration and operation, therefore it is not intended that limiting the scope of the invention.

" the hybrid vehicle queue " of indication of the present invention includes net connection car and non-net connection car, and " net connection car " refers to New Generation of Intelligent automobile, and as shown in Figure 2 with the vehicle of wifi symbol, it has a characteristic that simultaneously

The first, there is environment sensing function, by from the forward and backward side's information of vehicles of car sensor (radar and video camera) perception (including distance, speed and acceleration etc.), wherein, forward and backward side's vehicle can be net connection car, it is also possible to be non-net connection vehicle.

The second, there is GPS positioning function, it is possible to obtain car's location information in real time.

3rd, there is communication function, by vehicle-mounted short range communication device or 4G/5G means of communication, it may be achieved the information between net connection vehicle and net connection vehicle, between net connection vehicle and cloud platform is mutual.

" non-net connection car " refers to conventional truck, does not namely possess the vehicle of These characteristics, or does not at least possess the vehicle of thirdly function, as shown in Figure 2 without the vehicle of wifi symbol.

As it is shown in figure 1, the impact-moderation method between vehicle comprises the following steps in the hybrid vehicle queue that provides of present embodiment:

1) gather the vehicle condition information of the adjacent non-net connection car in the net connection vehicle condition information of car and this net connection Herba Plantaginis, rear: all net connection car collection vehicle condition information from car damped condition and its forward and backward side's vehicle in driving vehicle queue, and above-mentioned information is sent to cloud computing platform by 4G/5G communication equipment.Wherein, damped condition includes from car brake signal and corresponding braking acceleration, and brake signal can be extracted by former car CAN, once collect this signal, then can determine whether to be operated for brake pedal；Braking acceleration generally refers to the longitudinal acceleration of vehicle.

2) sequentially judge the head car in hybrid vehicle queue, rear car, be if so, then designated as a car or rear car, and enter step 3)；Otherwise, return step 1), this step particularly as follows:

Whether first, it is determined that there is the damped condition of net connection car to exceed setting threshold value, if so, then this net connection car is designated as a car；Otherwise, step 1 is returned).That is, according to step 1) in the vehicle condition information of the adjacent non-net connection car in the net connection vehicle condition information of car that collects and this net connection Herba Plantaginis, rear, cloud computing platform is braked performance analysis, if the brake signal of a certain net connection car and longitudinal acceleration signal occur simultaneously, and longitudinal acceleration exceedes setting threshold value (such as 2m/s²) time, then this net connection car is designated as a car, and is labeled.

Then, it is judged that whether the vehicle at head car rear meets queue impact-moderation requirement, rear car if so, then it is designated as；Otherwise, step 1 is returned).It is to say, according to queue impact-moderation requirement, all after cloud computing platform enemy's car meet the vehicle that queue impact-moderation requires and carry out the foundation of hybrid vehicle queue, and each car is numbered.Queue impact-moderation requires to include: 1. head car for net connection car and creates braking deceleration；2. between vehicle the time headway of rear car and front truck less than preset value (such as 3 seconds)；3. trailer is net connection car；4. the net connection car between head car and trailer is random with non-net connection truck position distribution.

3) cloud computing platform receives the vehicle condition information of net connection car in hybrid vehicle queue.

4) the expectation acceleration of each net connection car in planning rear car: cloud computing platform combine that net connection car (except head car except) in mixing hybrid vehicle queue carries from car status information and be adjacent non-net and join car Partial State Information, utilize based on MPC (ModelPredictiveControl, Model Predictive Control) hybrid vehicle queue impact-moderation method, the expectation acceleration of net connection car each in hybrid vehicle queue is carried out centralized planning, and each rear car optimum calculated is expected the net connection car that acceleration flows in hybrid vehicle queue respectively.

5) in rear car, each net connection car travels according to the expectation acceleration planned, non-net connection car travels according to car-following model.

6) judge whether each rear car stops, if so, then stop controlling；Otherwise, step 3 is returned).

Above-mentioned steps 4) in, minimum for optimization aim to net connection car critical retardation power summation, the expectation acceleration of each net connection car in planning rear car.

Above-mentioned steps 5) in, the expectation acceleration a of net connection car_i,desK () is obtained by following method:

A) hybrid vehicle queue is set up

Assuming there be m net connection car in the hybrid vehicle queue of N car, the ID set of net connection car present position in hybrid vehicle queue is:

$G=\{g_1,g_2,...,g_m\},G\&Subset;\{\mathrm{1,2},...N\}$

So need to participate in optimizing the vehicle ID set calculated and be:

C=G ∪ { g_i-1}∪{g_j+ 1}, g_i> 1, g_j< N, g_i,g_j∈G

Order

C={c_1, c_2 ... c_p}, c_1 < c_2 < ... < c_p, m≤p≤N,

Set C representated by element is optimize the string virtual vehicle queue controlled in calculating process, in actual environment these vehicles not necessarily before and after adjacent.Wherein, net connection car can accurately calculate the relative kinetic energy density from car and front truck, but not net connection car can only calculate and the relative kinetic energy density of front net connection car.If there are continuous two non-net connection cars in car team, then it is all uncontrollable, and relative kinetic energy density calculation between the two is also meaningless.

B) object function is set up

The relative kinetic energy density of two cars in virtual car team, namely in two cars, the marginal value power of rear car is:

$F_{i,i+1}\left(t\right)=\frac{1}{2}\frac{m_{i+1}}{S_i}\&CenterDot;{\&lsqb;v_i\left(t\right)-v_{i+1}\left(t\right)\&rsqb;}^2$

Wherein F_i,i+1T () is the relative kinetic energy density between t i+1 car and i-th car, m_iIt is the quality of i-th car, S_iT () represents the following distance of i-th car of t and (i-1) car, v_iThe speed of (t) respectively i-th car of t, v_i-1(t)≤v_i(t), i=1,2 ..., n-1, works as v_i-1(t) > v_iV is taken time (t)_i-1(t)=v_i(t)。

Following distance is calculated by following formula.

S_i=x_i(t)-x_i+1(t)-L_i

Wherein L_iIt is the length of i-th car, x_iT () is the headstock position of i-th car of t.

Using the critical retardation power summation of many cars as object function, using the minimum optimization aim controlled as hybrid vehicle queue impact-moderation of its value, and only consider the critical retardation power between adjacent two cars, it may be assumed that

$\begin{array}{c}J\left(t\right)=\frac{1}{2}{\&Sigma;}_{c\_i=1}^m\frac{m_{c\_i}}{S_{c\_i}\left(t\right)}\&CenterDot;{\&lsqb;v_{c\_(i-1)}\left(t\right)-v_{c\_i}\left(t\right)\&rsqb;}^2\\ +\frac{1}{2}{\&Sigma;}_{c\_j=1}^{N-1-m}\frac{m_0}{S_{c\_j}\left(t\right)}\&CenterDot;{\&lsqb;v_{c\_(j-1)}\left(t\right)-v_{c\_j}\left(t\right)\&rsqb;}^2\end{array}$

Wherein, c_i ∈ G and c_i > 1, c_j ∈ C_CG, c_j > 1 and c_ (j-1) ∈ G, namely c_i as rear car net join car ID, c_i be non-net connection car but front truck be net connection car vehicle ID.m₀Quality for non-net connection car, although its quality is immesurable, but considers type and the mass range of vehicle, it may be assumed that the quality of all non-net connection cars is same certain value when calculating.Its speed and following distance information can be passed through Adjacent vehicles radar and accurately obtain.

Owing to non-net connection car is controlled by this control method, therefore the expectation acceleration of non-net connection car is determined by driver.In car-following model, by netting the vehicle speed data of the non-net connection car measured by the perception unit joining car, the actual acceleration that difference obtains, it is approximately expectation acceleration and is substituted into calculating.

Therefore, can show that the hybrid vehicle queue impact-moderation algorithm object function based on net connection car critical retardation power is:

$\underset{a_{des}}{\min }\frac{1}{2}\underset{i=1}{\overset{N-1}{\&Sigma;}}\frac{m_{i+1}}{x_i\left(t\right)-x_{i+1}\left(t\right)-L_i}\&CenterDot;{\&lsqb;v_i\left(t\right)-v_{i+1}\left(t\right)\&rsqb;}^2$

Wherein a_desExpectation acceleration input for each net connection car.

C) constraints

Choose kinematics model as auto model, and assume that vehicle actual acceleration is first order inertial loop with expectation acceleration input.Net connection car expects that acceleration is determined by controlling system, and non-net connection car adopts current acceleration as expecting to accelerate to substitute into auto model, and assumes that the non-net connection car ID set formed is B, obtains the auto model in prediction time domain as follows:

$\left\{\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_i\left(t\right)=v_i\left(t\right)\\ {\stackrel{\&CenterDot;}{v}}_i\left(t\right)=a_i\left(t\right)\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=\frac{1}{\mathrm{\&tau;}}\left(a_{i,des}\right(t)-a_i(t\left)\right),i\&Element;G\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=0,i\&Element;B\end{array}\right.$

Wherein x_iT () is i-th car headstock position in t, v_iT () is i-th car at the headstock speed of t and a_iT () is i-th car acceleration in t, a_i,desT () is i-th car expectation acceleration in t, τ is time constant.In order to make this algorithm have solution, and also continuing to optimize the acceleration of other vehicle when some car collides, constraints only considers the stopping power of each car, and is not provided with safe distance constraint.The acceleration of each net connection car is constrained to:

a_i,min≤a_i,des(t)≤a_i,max

Wherein a_i,minIt is the minimum acceleration of i-th car, a_i,maxIt it is the peak acceleration of i-th car.

D) Optimized model is set up

In conjunction with the object function set up in b) and c) in constraints, if the vectorial a that centralized Control variable is each car expectation acceleration composition_des(t)=[a_1,des(t),a_2,des(t),...,_N,des(t)]^T, solve time range for prediction time domain Δ t, i.e. [t0, t0+ Δ t].Obtain representing based on the hybrid vehicle queue impact-moderation Controlling model such as following formula of total critical retardation power:

The expectation acceleration a of net connection car_i,desK () is obtained by following formula:

$\underset{a_{di}}{\min }\underset{0}{\overset{T}{\&Integral;}}(\&Sigma;\frac{m_{c\_i}}{S_{c\_i}\left(t\right)}{\&lsqb;v_{c\_i}\left(t\right)-v_{c\_i-1}\left(t\right)\&rsqb;}^2+\&Sigma;\frac{m_0}{S_{c\_j}\left(t\right)}{\&lsqb;v_{c\_j}\left(t\right)-v_{c\_j-1}\left(t\right)\&rsqb;}^2)$

Its constraints is:

$\left\{\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_i\left(t\right)=v_i\left(t\right)\\ {\stackrel{\&CenterDot;}{v}}_i\left(t\right)=a_i\left(t\right)\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=\frac{1}{\mathrm{\&tau;}}(a_{di}\left(t\right)-a_i\left(t\right)),i\&Element;G\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=0,i\&Element;B\\ a_{i,\min }\&le;a_{di}\left(t\right)\&le;a_{i,\max }\end{array}\right.$

Wherein: T is prediction time domain length, G is the set of net connection car in hybrid vehicle queue, and B is the set of non-net connection car in hybrid vehicle queue, and c_i is the sequence number of net connection car, and c_j is the sequence number of non-net connection car, m_{c_i}It is the quality of the c_i net connection car, m₀For the quality of non-net connection car, S_iIt it is the following distance of i-th car and its front truck；x_iT () is the headstock position of i-th car of t, v_iT () is the speed of i-th car of t, a_iT () is the acceleration of i-th car of t, a_diT () is the expectation acceleration of i-th car, τ is the time constant characterizing vehicle operating lag, a_i,minIt is the minimum acceleration of i-th car, a_i,maxIt it is the peak acceleration of i-th car.

Step 5) in the preparation method of car-following model of non-net connection car specific as follows:

A) non-net connection car car-following model is set up

For analyzing in Some vehicles networked environment, the impact on non-net connection car of the net connection car, adopt and improve driver AP (ActionPoint) model (belonging to driving mental model), simulate the behavior that the driver of non-net connection car brakes with car.Driver's desired braking deceleration that this model characterizes is determined by TTC (TimeToCollision), and this TTC is the collision avoidance time, characterizes and knocks into the back the required time with front truck under current state from car, and its form is as follows:

a_des=c e TTC^f+d(1)

In formula (1), c, d, e and f are undetermined constant.

The use of formula (1) need to meet certain precondition, and when namely front truck visual angle change rate in driver's seat reaches certain threshold value, driver just can perceive and take operation, can keep this operation eliminating driver before car and front truck relative motion.Front truck visual angle change rate in driver's seat is:

$\frac{d\mathrm{\&theta;}}{dt}\&ap;\frac{\mathrm{\&Delta;}v}{{\mathrm{\&Delta;x}}^2}---\left(2\right)$

Wherein θ is relative velocity and the vehicle headway of front truck angular field of view in driver's eye, Δ v and Δ x respectively front and back two cars.

Following distance D is expected with driver_pReplace Δ x, obtain judging the threshold of perception current of driver:

$k\&ap;\frac{\mathrm{\&Delta;}v}{D_p^2}---\left(3\right)$

As k > 6 × 10^-4Time, driver will take operation according to formula (1).

E, f and d value in formula (1): e=1.04, f=0.72, d=-6.918.Wherein e, f are that driver estimates from the parameter that car is relevant to front truck TTC, and d is the desired maximum braking deceleration of driver.From the form of (1) it can be seen that, only when TTC is reduced to 0, driver just can take expectation maximum braking deceleration, this means that driver is before the collision all without taking maximum braking, and the present embodiment it may be reasonably assumed that when two car TTC are less driver will take maximum braking, making this TTC is TTC_l, then formula (1) can be rewritten as

$a_{des}=c\&CenterDot;e\&CenterDot;({\mathrm{TTC}}^f-{\mathrm{TTC}}_l^f)+d---\left(4\right)$

On the other hand, driver will not take brake operating when TTC is very big, it is assumed that driver determines that taking the TTC of brake operating is TTC_u, work as TTC=TTC_uTime, the desired braking deceleration of driver is 0, and therefore the parameter c in (4) can determine by equation (5).

$c\&CenterDot;e\&CenterDot;({\mathrm{TTC}}_u^f-{\mathrm{TTC}}_l^f)+d=0---\left(5\right)$

TTC_uDetermining according to ongoing driver characteristics result of study, recording the TTC average meeting China's driver characteristic is 14.68s.Use this value as TTC_u, TTC_lTake 3s.Each parameter value is substituted into (1), and obtaining non-net connection car car-following model is:

$a_{des}=\left\{\begin{array}{cc}0,& TTC>14.68\\ 1.411\&CenterDot;1.04\&CenterDot;({\mathrm{TTC}}^{0.72}-3^{0.72})-6.91,& 3\&le;TTC\&le;14.68\\ -6.918,& TTC<3\end{array}\right.---\left(6\right)$

The time of driver's reaction of this car-following model simulation is longer, and braking deceleration uphill process is relatively mild, meets true operation, and final parking space also more gears to actual circumstances.

As shown in Figures 2 and 3, the present invention also provides for the impact-moderation device in a kind of hybrid vehicle queue between vehicle, and it includes perception unit 1, communication unit the 2, first brak control unit 3a, the second brak control unit 3b and cloud computing platform 4.

Perception unit 1 is installed on net connection car, and for collecting the vehicle condition information joining the adjacent non-net connection car in Herba Plantaginis, rear from car status information and this net of respective wire connection car, and it is delivered to communication unit 2.Perceptually a kind of preferred implementation of unit 1, it includes environmental perception device 11, acceleration harvester 12, on-vehicle information harvester 13 and positioner 14, wherein: environmental perception device 11 joins the vehicle condition information of car for the non-net that the forward and backward side of perception is adjacent and exports.Environmental perception device 11 is made up of vehicle-borne CCD video camera and millimetre-wave radar, the forward and backward side relative position with the non-net connection car in track is detected by vehicle-borne CCD video camera, whether detect non-net connection car in identical track, by the velocity and acceleration of vehicle-mounted millimeter wave detections of radar forward and backward Fang Fei net connection car, vehicle-borne CCD video camera and millimetre-wave radar collect all information and all flow to communication unit 2.

Acceleration harvester 12 is used for the longitudinal acceleration of its place vehicle of perception and flows to communication unit 2.Acceleration harvester 12 in present embodiment can adopt inertial sensor, and requires that the renewal frequency of inertial sensor is not less than 5Hz.

On-vehicle information harvester 13 is used for gathering speed, brake pressure and throttle opening information.Preferably, on-vehicle information harvester 13 is integrated with CAN communication chip, for gathering necessary car status information (such as vehicle brake signal) from former car CAN, and these car status information is flowed to communication unit 2.The frequency acquisition requiring CAN communication chip is not less than 5Hz, with the real-time that guarantee information updates.

Positioner 14 positions information for gathering from car.In present embodiment, that vehicle locating device 14 adopts is Beidou satellite navigation system (BeiDouNavigationSatelliteSystem, BDS), is used for obtaining vehicle position information and sending communication unit 2 to.Positioner 14 requires that renewal frequency is not less than 5Hz, and positioning precision is not less than 1m, to ensure the accurate of vehicle location and to update, thus ensureing to calculate accuracy and the real-time of vehicle headway.

Communication unit 2 is installed on net connection car, and is used for netting connection car and carries out real-time information interaction with cloud computing platform 4.In present embodiment, communication unit 2 adopts 4G/5G wireless communication module.In mixing hybrid vehicle queue impact-moderation control process, communication unit 2 passes through to send cloud computing platform 4 to from car status information and environmental information by what the perception unit 1 on net connection car gathered, and obtain from the calculated expectation acceleration of cloud computing platform 4, it is delivered to the first brak control unit 3a, first brak control unit 3a travels for Controling network connection car desirably acceleration after the expectation acceleration receiving cloud computing platform 4 output, it is achieved impact-moderation control for brake.Described 4G/5G wireless telecommunications, adopt the LTE-V communication module based on TD-LTE technical research, including cellular based communication technology and two kinds of communication modes of distributed straight-through technology, the follow-up LTE-V mechanics of communication towards 5G that may be used without in being currently in development of extension, but it is not limited to this.

First brak control unit 3a is installed on net connection car, and travel for Controling network connection car desirably acceleration after the expectation acceleration receiving cloud computing platform 4 output, it is in holding state when not receiving the expectation acceleration of described cloud computing platform output, each vehicle freely travels according to driver intention, to realize the control of net connection car in hybrid vehicle queue.Preferably, the first brak control unit 3a controls vehicle acceleration by the brake actuator in E-Gas and VSC.

Second brak control unit 3b is used for controlling non-net connection car and travels according to car-following model.Car-following model refers to when front is in an emergency, the emergency reaction that driver takes, and the car-following model of non-net connection car is:

$a_{des}=\left\{\begin{array}{cc}0,& TTC>14.68\\ 1.411\&CenterDot;1.04\&CenterDot;({\mathrm{TTC}}^{0.72}-3^{0.72})-6.91,& 3\&le;TTC\&le;14.68\\ -6.918,& TTC<3\end{array}\right.$

In formula, TTC is the collision avoidance time.

Cloud computing platform 4 is for receiving the expectation acceleration of each net connection car in the information of perception unit 1 output, the head car judged in hybrid vehicle queue and rear car and planning rear car and exporting.Described head car is that damped condition exceedes the net connection car setting threshold value.Described rear car is the vehicle that a car rear meets queue impact-moderation requirement.

As a kind of preferred implementation of cloud computing platform 4, cloud computing platform 4 includes communicator 41, head car judgment means 42, rear car judgment means 43 and expectation acceleration device for planning 44, wherein:

Communicator 41 is mutual with each communication unit 2 information, and the communication unit 2 for being flowed in rear car on each net connection car by the expectation planned acceleration.Head car judgment means 42 is for judging described head car according to the damped condition of input.Specifically, according to described brake signal and described longitudinal acceleration, head car judgment means 42 judges whether certain car is described head car.In this embodiment, when brake pedal is operated and braking acceleration is more than 2m/s²Time, remember that this car is head car.Rear car judgment means 43 is for receiving the information of a car, and requires to judge described rear car according to queue impact-moderation, and the vehicle in hybrid vehicle queue is numbered.Hybrid vehicle queue judgment means 43 is formed into columns according to criterion enemy's car front vehicle.Head car rear vehicle in, the vehicle met the following conditions can be incorporated into mixing hybrid vehicle queue in: 1. between vehicle time headway less than preset value (such as 3 seconds)；2. trailer is net connection car.It is unsatisfactory for one of above-mentioned condition, is not all incorporated into mixing hybrid vehicle queue.Expect that acceleration device for planning 44 is for the driving states information according to net connection car, plans the expectation acceleration of net connection car in each rear car.

In above-mentioned embodiment, it is desirable to acceleration device for planning 44 includes information fusion module 441, model prediction module 442 and expectation acceleration calculation module 443, wherein:

The vehicle condition information of the information fusion module 441 forward and backward side's vehicle for being perceived by described environmental perception device 11 merges, and quickly determines forward and backward side's car status information, and flows to model prediction module 442.

Model prediction module 442 is predicted for the motion of each rear car, flows to expectation acceleration calculation module 443.Specifically: model prediction module 442 is for according to speed, acceleration and position, being predicted the motion of vehicle.For certain non-net connection car, by time discretization, taking discrete time long for Δ T, collecting k, to walk the position of this vehicle be x (k), and speed is v (k), and acceleration is a (k), it is desirable to acceleration is a_des(k), then k+j step is set to x (k+j | k)=x (k+j-1 | k)+v (k+j-1 | k) * Δ T, speed is v (k+j | k)=v (k+j-1 | k)+a (k+j-1 | k) * Δ T, and acceleration is a (k+j | k)=(τ-Δ t)/τ * a (k+j-1 | k)+Δ T/ τ * a_des(k+j-1|k).Here, τ is the time constant characterizing vehicle operating lag, it can thus be appreciated that: model prediction module 442 is predicted for the position of vehicle, speed and acceleration carry out Np step.

Expect that acceleration calculation module 443 is for each rear car motion according to prediction, calculates the expectation acceleration of each rear car and flows to communicator 41.In particular, it is desirable to acceleration calculation module 443 walks the vehicle location, speed and the acceleration information that collect according to k, predicting the outcome of the Np step that combination model prediction module 442 obtains, it is desirable to minimize object function, thus optimizing the aimed acceleration of vehicle.

In sum, the device that the respective embodiments described above provide can be minimum for controlling optimization aim with vehicle critical retardation power, can effectively utilize being controlled with car space of each workshop in hybrid vehicle queue, it is achieved the reasonable layout of vehicle relative position, thus the generation of collision free as far as possible；When collision cannot be avoided, can effectively slow down the degree of injury that collision brings.

Last it is noted that above example is only in order to illustrate technical scheme, it is not intended to limit.Although the present invention being described in detail with reference to previous embodiment, it will be understood by those within the art that: the technical scheme described in foregoing embodiments still can be modified by it, or wherein portion of techniques feature is carried out equivalent replacement；And these amendments or replacement, do not make the essence of appropriate technical solution depart from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (10)

1. impact-moderation method between vehicle in a hybrid vehicle queue, it is characterised in that comprise the following steps:

1) the vehicle condition information of the adjacent non-net connection car in the net connection vehicle condition information of car and this net connection Herba Plantaginis, rear is gathered；

2) the head car in hybrid vehicle queue, rear car are sequentially judged, if head car or rear car, then enter step 3), otherwise return step 1), head car is that damped condition exceedes the net connection car setting threshold value, and rear car is the vehicle that a car rear meets queue impact-moderation requirement；

3) the vehicle condition information of net connection car is received；

4) the expectation acceleration of each net connection car in planning rear car；

5) in rear car, each net connection car travels according to the expectation acceleration planned, non-net connection car travels according to car-following model；

6) judge whether each rear car stops, if so, then stop controlling；Otherwise, step 3 is returned).

2. impact-moderation method between vehicle in hybrid vehicle queue as claimed in claim 1, it is characterised in that step 4) in, minimum for optimization aim net connection car critical retardation power summation, in planning rear car, each net joins the expectation acceleration of car.

3. impact-moderation method between vehicle in hybrid vehicle queue as claimed in claim 2, it is characterised in that the expectation acceleration a of net connection car_i,desK () is obtained by following formula:

$\underset{a_{di}}{\min }\underset{0}{\overset{T}{\&Integral;}}\left(\&Sigma;\frac{m_{c\_i}}{S_{c\_i}\left(t\right)}{\&lsqb;v_{c\_i}\left(t\right)-v_{c\_i-1}\left(t\right)\&rsqb;}^2+\&Sigma;\frac{m_0}{S_{c\_j}\left(t\right)}{\&lsqb;v_{c\_j}\left(t\right)-v_{c\_j-1}\left(t\right)\&rsqb;}^2\right)$

Its constraints is:

$\left\{\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_i\left(t\right)=v_i\left(t\right)\\ {\stackrel{\&CenterDot;}{v}}_i\left(t\right)=a_i\left(t\right)\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=\frac{1}{\mathrm{\&tau;}}\left(a_{di}\right(t)-a_i(t\left)\right),i\&Element;G\\ {\stackrel{\&CenterDot;}{a}}_i\left(t\right)=0,i\&Element;B\\ a_{i,\min }\&le;a_{di}\left(t\right)\&le;a_{i,\max }\end{array}\right.$

Wherein: T is prediction time domain length, G is the set of net connection car in hybrid vehicle queue, and B is the set of non-net connection car in hybrid vehicle queue, and c_i is the sequence number of net connection car, and c_j is the sequence number of non-net connection car, m_{c_i}It is the quality of the c_i net connection car, m₀For the quality of non-net connection car, S_iIt is the following distance of i-th car and its front truck, x_iT () is the headstock position of i-th car of t, v_iT () is the speed of i-th car of t, a_iT () is the acceleration of i-th car of t, a_diT () is the expectation acceleration of i-th car, τ is the time constant characterizing vehicle operating lag, a_i,minIt is the minimum acceleration of i-th car, a_i,maxIt it is the peak acceleration of i-th car.

4. as claimed any one in claims 1 to 3 impact-moderation method between vehicle in hybrid vehicle queue, it is characterized in that, step 2) in, the composition of hybrid vehicle queue has the feature that the time headway of two cars adjacent in 1. hybrid vehicle queue is less than preset value；2. trailer is net connection car；3. the net connection car between head car and trailer is random with non-net connection truck position distribution.

5. impact-moderation method between vehicle in hybrid vehicle queue as claimed in claim 4, it is characterised in that the car-following model of non-net connection car is:

$a_{des}=\left\{\begin{array}{cc}0,& TTC>14.68\\ 1.411\&CenterDot;1.04\&CenterDot;({\mathrm{TTC}}^{0.72}-3^{0.72})-6.91,& 3\&le;TTC\&le;14.68\\ -6.918,& TTC<3\end{array}\right.$

In formula, TTC is the collision avoidance time.

6. impact-moderation device between vehicle in a hybrid vehicle queue, it is characterized in that, including perception unit (1), communication unit (2), the first brak control unit (3a), the second brak control unit (3b) and cloud computing platform (4), wherein: described perception unit (1), communication unit (2) and the first brak control unit (3a) are mounted on net connection car, described second brak control unit (3b) is installed on non-net connection car；Described perception unit (1) is for collecting joining the vehicle condition information of the adjacent non-net connection car in Herba Plantaginis, rear from car status information and this net and exporting of respective wire connection car；Described communication unit (2) is used for netting connection car and carries out real-time information interaction with described cloud computing platform (4)；Described cloud computing platform (4) is used for receiving the expectation acceleration of each net connection car in information, the head car judged in hybrid vehicle queue and the rear car and planning rear car that described perception unit (1) exports and exports, described head car is that damped condition exceedes the net connection car setting threshold value, and described rear car is the vehicle that a car rear meets queue impact-moderation requirement；Described first brak control unit (3a) travels for the connection car desirably acceleration of Controling network after receiving the expectation acceleration that described cloud computing platform (4) exports；Described second brak control unit (3b) is used for controlling non-net connection car and travels according to car-following model.

7. impact-moderation device between vehicle in hybrid vehicle queue as claimed in claim 6, it is characterized in that, described perception unit (1) includes environmental perception device (11), acceleration harvester (12), on-vehicle information harvester (13) and positioner (14), wherein: described environmental perception device (11) joins the vehicle condition information of car for the non-net that the forward and backward side of perception is adjacent and exports；Described acceleration harvester (12) for perception from the longitudinal acceleration of car；Described on-vehicle information harvester (13) is used for gathering speed, brake pressure and throttle opening information；Described positioner (14) positions information for gathering from car.

8. impact-moderation device between vehicle in hybrid vehicle queue as claimed in claim 6, it is characterized in that, the information conveyance that described perception unit (1) is gathered by 4G/5G radio communication by described communication unit (2) is to described cloud computing platform (4), and is used for receiving that described cloud computing platform (4) plans expects acceleration from car.

9. impact-moderation device between vehicle in hybrid vehicle queue as claimed in claim 6, it is characterized in that, described cloud computing platform (4) includes communicator (41), head car judgment means (42), rear car judgment means (43) and expectation acceleration device for planning (44), wherein: described communicator (41) is mutual with described communication unit (2) information, the described communication unit (2) and for the expectation planned acceleration is flowed in rear car on each net connection car；Described head car judgment means (42) is for judging described head car according to the damped condition of input；Described rear car judgment means (43) is for receiving the information of described head car, and requires to judge described rear car according to queue impact-moderation, and the vehicle in hybrid vehicle queue is numbered；Described expectation acceleration device for planning (44), for the driving states information according to net connection car, plans the expectation acceleration of net connection car in each rear car.

10. impact-moderation device between vehicle in hybrid vehicle queue as claimed in claim 9, it is characterized in that, described expectation acceleration device for planning (44) includes information fusion module (441), model prediction module (442) and expectation acceleration calculation module (443), wherein, the vehicle condition information of the described information fusion module (441) forward and backward side's vehicle for being perceived by described environmental perception device (11) merges, quickly determine forward and backward side's car status information, and flow to described model prediction module (442)；Described model prediction module (442) is predicted for the motion of each described rear car, flows to described expectation acceleration calculation module (443)；Described expectation acceleration calculation module (443), for the described each rear car motion according to prediction, calculates the expectation acceleration of each rear car and flows to described communicator (41).