CN104157167A - Vehicle collision preventing method based on collaborative relative positioning technologies - Google Patents

Abstract

The invention relates to the field of intelligent transportation, in particular to a vehicle collision preventing method based on collaborative relative positioning technologies. The positioning information measurement and exchange on the basis of a V2V network, an enhanced relative positioning method on the basis of pseudorange, a vehicle self-prediction positioning and navigation method based on Kalman filtering, and a vehicle collision preventing judgment mechanism based on the relative distance are included in the method. The vehicle collision preventing method based on the collaborative relative positioning technologies includes the steps that information between near vehicles is transmitted through the V2V network, the relative distance between the vehicles is calculated on the basis of the pseudorange or Kalman filtering, and a vehicle alarm mechanism is triggered when the relative distance is smaller than a safety threshold value. The vehicle collision preventing method based on the collaborative relative positioning technologies has the advantages that a collaborative relative positioning model is built according to the multipath transmission phenomenon of urban roads, satellite signals with low spatial correlation are removed according to the correlation, and the vehicle position is predicted according to a Kalman filtering method when a GPS signal interrupts. By means of the collaborative relative positioning method, the probability of occurrence of traffic accidents of urban roads can be reduced, and prediction accuracy is improved.

Description

A kind of vehicle collision avoidance method based on collaborative relative positioning technology

Technical field

The present invention relates to intelligent transportation field, relate in particular to a kind of vehicle collision avoidance method based on collaborative relative positioning technology.

Background technology

GPS (Global Positioning System) GPS is the global navigation satellite system that the U.S. sets up, and can realize high precision, round-the-clock, real-time navigation and test the speed.GPS technology has been widely used in the every field such as Mapping remote sensing technology, Aero-Space, communication, traffic in recent years, and has changed deeply people's life.

Locating information analytical applications based on gps system progressively expands to intelligent transportation system ITS (Intelligent Transportation System).In modern ITS system, can carry out vehicle self-navigation and safe driving by locating information.According to the principle of satellite navigation and location system, GPS location technology comprises position estimation and the position estimation based on pseudorange based on carrier phase, although the former location is very accurate, has the problem of integer ambiguity, although the latter's degree of accuracy is lower, there is good real-time.But it is several as follows that the factor that affects setting accuracy when vehicle GPS receiving satellite signal comprises: gps satellite self error (comprising satellite clock error, satellite ephemeris error), transmission delay (comprising ionosphere time delay and troposphere time delay), GPS receiver error (comprising receiver clock error, thermonoise and ground multidiameter delay etc.).In addition, at a ground, strengthen system (GBAS, Ground-Based Augmentation System) in, a reference base station that is positioned at known exact position can be to vehicle broadcasting satellite precise position information, GPS receiver calculates after vehicle and intersatellite pseudorange and to revise observation data with broadcasting after precise position information compares again according to himself receiving satellite positioning signal, and the method for this DGPS can effectively reduce identical propagation delay time error and satellite error.But, in urban road both sides, there is buildings, this means that the sight line path between gps satellite and vehicle may be blocked, GPS receiver can only receive the reflected signal of satellite.The different paths reflected signal that derives from different buildingss has formed the Multipath Transmission that receives signal, and its error causing may be up to tens of rice, and this has reduced the precision of absolute fix undoubtedly.For DGPS system, because GNSS reference base station is usually based upon in open environment, therefore and the reception signal between adjacent vehicle there is the error that spatial coherence can reduce this Multipath Transmission.

Summary of the invention

The present invention overcomes above-mentioned weak point, thereby object is to provide a kind of relative position by road adjacent vehicle to detect and relative position can reduce the vehicle collision avoidance method of city vehicle collision and traffic accident generation over the aposematic mechanism of threshold value.

The present invention achieves the above object by the following technical programs: a kind of vehicle collision avoidance method based on collaborative relative positioning technology, GPS sending/receiving machine is installed on positioned vehicle, and the direct or indirect positioned vehicle of several satellite in orbit, comprises the following steps:

1) analyze the first vehicle of relative positioning, the satellite information that the second vehicle receives, determine the shared satellite in orbit of the first vehicle, the second vehicle, set up collaborative relative positioning system model;

2) the first vehicle and the second vehicle are set up V2V network, and by sharing satellite in orbit transmission packet, packet information comprises: vehicle identification ID, vehicle destination absolute fix coordinate, vehicle three-dimensional velocity, shared satellite signal to noise ratio (S/N ratio) and shared satellite pseudorange;

3) when gps signal is stablized, adopt the enhancing relative positioning method based on pseudorange to calculate relative distance between the first vehicle, the second vehicle;

4) when gps signal interrupts, adopt vehicle based on Kalman filtering from predicting that positioning navigation method predicts relative distance between the first vehicle, the second vehicle;

5) according to step 3) or step 4) relative distance that obtains carries out the vehicle collision avoidance judgement based on relative distance, if in the time that relative distance being less than or equal to safe distance, triggering vehicle carried device and trigger emergency brake of vehicle mechanism.

As preferably, the enhancing relative positioning method based on pseudorange comprises the following steps:

(11) according to formula (1) compute location vehicle v, use together the pseudorange between satellite s in-orbit:

$p_v^{\left(s\right)}={\mathrm{\ρ}}_v^{\left(s\right)}+s\×(\mathrm{\Δ}{t}_v-\mathrm{\Δ}{t}^{\left(s\right)})+d_{m,v}^{\left(s\right)}+d_{n,v}^{\left(s\right)}+{\mathrm{\ϵ}}_v^{\left(s\right)}---\left(1\right)$

Pseudorange carries out second order difference: $\▿\mathrm{\Δ}{p}_{a,b}^{k,j}=(p_a^k-p_b^k)-(p_a^j-p_b^j)=\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}+({\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k)-({\mathrm{\ϵ}}_a^j-{\mathrm{\ϵ}}_b^j)---\left(2\right);$

Wherein, represent the propagation delay time that bring in ionosphere, represent the propagation delay time that bring in troposphere, s represents the light velocity, Δ t _vrepresent positioned vehicle clocking error, Δ t ^(s)represent satellite clock error, represent the impact that thermonoise and multipath are interfered, represent actual range, a represents the first vehicle, and b represents the second vehicle, and k, j represent all shared satellite in orbit;

(12) calculate value and compare with predetermined threshold value, when calculated value is less than predetermined threshold value, judge corresponding shared satellite in orbit property relevant to positioned vehicle, retain the high shared satellite in orbit of relevance.

As preferably, positioned vehicle a, b receive the direct signal transmission that shares satellite in orbit j, pseudorange second order difference result is $\▿\mathrm{\Δ}{p}_{a,b}^{k,j}-\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}\≈{\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k.$

As preferably, the vehicle based on Kalman filtering is from predicting that positioning navigation method comprises the following steps:

(21) determine the initial value of state vector, error covariance matrix;

(22) predict next state vector, error covariance matrix value constantly:

(23) state upgrades, and next arrives constantly, uses the locating information prediction pseudorange receiving and try to achieve observing matrix;

(24) use error covariance matrix, in conjunction with the gain matrix of observing matrix calculating filter;

(25) calculate the actual distance of vehicle and shared satellite observed reading and predicted value;

(26) estimated value of computing mode vector sum error covariance matrix.

As preferably, the described vehicle collision avoidance judgement based on relative distance comprises:

$D=\left\{\begin{array}{c}|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)-\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|\\ \left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)+\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)\\ \mathrm{Min}\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2,|v_b|t+\frac{1}{2}{s}_b{t}^2)\end{array}\right.$

Wherein, | v _a| the mapping of the three-dimensional velocity of expression vehicle a on direction vector, | v _b| the mapping of the three-dimensional velocity of expression vehicle b on direction vector, s _athe acceleration that represents vehicle a, S _bthe acceleration that represents vehicle b, t represents driver's burst traffic hazard reaction and the average experience time of processing;

When vehicle a and vehicle b in the same way and row two vehicle heading vector angles be less than 90 °, now there is rear-end collision, $D=|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)-\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|;$

As vehicle a and vehicle b, going in the same direction is that two cars travel direction vector angle is greater than 90 °, before now occurring, hits accident, $D=|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)+\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|;$

When vehicle a and vehicle b side direction and row is two cars travel direction vector angle equals 90 °, now occur that side hits accident, $D=\mathrm{Min}\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2,|v_b|t+\frac{1}{2}{s}_b{t}^2).$

Beneficial effect of the present invention is: for the Multipath Transmission phenomenon in urban road, the collaborative relative positioning model of setting up can be rejected the satellite-signal that spatial coherence is low according to correlativity, and is occurring that gps signal is used the method prediction vehicle location of Kalman filtering when interrupting.By the application of collaborative relative positioning method, can reduce the probability of happening of traffic hazard in urban road.Calculate the relative distance between vehicle, can subdue error, thereby make calculated value substantially equal vehicle actual distance, improve predictablity rate.

Accompanying drawing explanation

Fig. 1 is the flow chart of steps of the inventive method;

Fig. 2 is the collaborative relative positioning system model schematic diagram of the embodiment of the present invention;

Fig. 3 is the relative positioning schematic diagram between neighbour's vehicle of the embodiment of the present invention.

Embodiment

Below in conjunction with specific embodiment, the present invention is described further, but protection scope of the present invention is not limited in this:

Embodiment 1: as shown in Figure 1, a kind of vehicle collision avoidance method based on collaborative relative positioning technology, comprise: the locating information based on V2V network is measured with exchange, the enhancing relative positioning method based on pseudorange, the vehicle based on Kalman filtering and certainly predicted positioning navigation method, the vehicle collision avoidance judgment mechanism based on relative distance, specifically comprises the following steps:

1) analyze the first vehicle of relative positioning, the satellite information that the second vehicle receives, determine the shared satellite in orbit of the first vehicle, the second vehicle, set up collaborative relative positioning system model;

2) the first vehicle and the second vehicle are set up V2V network, and by sharing satellite in orbit transmission packet, packet information comprises: vehicle identification ID, vehicle destination absolute fix coordinate, vehicle three-dimensional velocity, shared satellite signal to noise ratio (S/N ratio) and shared satellite pseudorange;

3) when gps signal is stablized, adopt the enhancing relative positioning method based on pseudorange to calculate relative distance between the first vehicle, the second vehicle;

4) when gps signal interrupts, adopt vehicle based on Kalman filtering from predicting that positioning navigation method predicts relative distance between the first vehicle, the second vehicle;

5) according to step 3) or step 4) relative distance that obtains carries out the vehicle collision avoidance judgement based on relative distance, if in the time that relative distance being less than or equal to safe distance, triggering vehicle carried device and trigger emergency brake of vehicle mechanism.

Suppose as shown in Figure 2 satellite in orbit S ₁-S ₅to carry GPS sending/receiving machine vehicle a, vehicle b transmit positioning signal, and all right the sailing both sides of vehicle a, vehicle b has in the urban traffic road of various buildingss, due to stopping of buildings, between vehicle and satellite, directly transmission path is blocked, so for vehicle a, vehicle b, receive direct signal, have and receive reflected signal, do not receive in addition satellite-signal.Wherein vehicle a can receive S ₁to S ₄signal, and S ₁, S ₂and S ₄directly to receive signal, S ₃it is reflected signal; Vehicle b can receive S ₁, S ₃, S ₄and S ₅signal, and S ₄and S ₅the signal directly receiving, S ₁and S ₃reflected signal.Obvious S ₂can only be received and S by vehicle a ₅can only be received by vehicle b, so when vehicle a and vehicle b position respectively separately, absolute fix coordinate departs from very large, complicated road environment can cause great measuring error in city directly by the absolute coordinates position of two cars, to locate to calculate its relative position, locates the distance l then calculating between vehicle by absolute coordinates ₂compare the two actual distance l ₁increased error.In collaborative relative positioning system model of the present invention, first determine the shared satellite in orbit of vehicle a and vehicle b, i.e. S ₁, S ₃and S ₄.By the satellite-signal receiving is analyzed, known S ₁for vehicle a, be yet that direct signal is reflected signal for vehicle b, this is low spatial correlativity while carrying out relative positioning for vehicle.Although vehicle a and b receive satellite S ₃be all reflected signal, but when distance between two cars is enough short their positioning error be identical substantially.Therefore in collaborative relative positioning model, only use the shared satellite S with certain space correlativity ₃and S ₄the positioning signal of transmission, therefore the relative actual position in two cars absolute coordinates position there is roughly the same offset direction and depart from big or small in, thereby the relative position that calculates the two can be subdued same deviation value and makes calculated value substantially equal vehicle actual distance.

Before adjacent vehicle exchanging orientation information, the observation satellite locator data that each vehicle need to receive comprises: the orbit coordinate of the signal to noise ratio snr of satellite transmission signal, the elevation angle of observation satellite, satellite ephemeris data and the observation satellite that obtained by ephemeris.Obviously calculating observation satellite is to the pseudorange of GPS receiver accordingly, and meanwhile each car also will be measured himself three-dimensional velocity and travel direction.To sum up, after vehicle has been set up V2V network, the packet information that collaborative relative positioning system model transmits comprises: vehicle identification ID, and vehicle destination absolute fix coordinate, vehicle three-dimensional velocity, shares satellite signal to noise ratio (S/N ratio) and shared satellite pseudorange.

The packet of the packet content that each car can send by himself and the adjacent vehicle receiving can determine the shared satellite in orbit of two cars.The consideration based on spatial coherence to the measurement of relative positioning herein, only selects the locating information that shares satellite for relative positioning.Table 1 is packet content.

Table 1

Enhancing relative positioning method based on pseudorange calculates vehicle to be measured relative distance between any two according to a known N observation satellite and N vehicle coordinate to be measured, according to observation satellite Calculation for Ephemerides pseudorange, with the difference of accurate distance and pseudorange, upgrade observation data and improve positioning precision.Determine that satellite s 3, satellite s 4, for calculating the relative pseudorange of vehicle a, vehicle b, as shown in Figure 3, are designated as k by s3, s4 is designated as j, represent the actual range between satellite k and vehicle a, represent the actual distance between satellite k and vehicle b, represent the actual distance between satellite j and vehicle a, represent the actual distance between satellite j and vehicle b, θ _krepresent that satellite k is with respect to the elevation angle degree of vehicle a, θ _jrepresent that satellite j is with respect to the elevation angle degree of vehicle a.

Pseudorange between vehicle v and satellite s can calculate according to following formula:

$p_v^{\left(s\right)}={\mathrm{\ρ}}_v^{\left(s\right)}+s\×(\mathrm{\Δ}{t}_v-\mathrm{\Δ}{t}^{\left(s\right)})+d_{m,v}^{\left(s\right)}+d_{n,v}^{\left(s\right)}+{\mathrm{\ϵ}}_v^{\left(s\right)}---\left(1\right)$

When enough near between two cars time, we have reason to think and can eliminate with reference to the method for traditional difference DGPS the propagation delay time of bringing due to ionosphere and troposphere with and satellite clock error delta t ^(s), Δ t _vrepresent positioned vehicle clocking error, represent the impact that thermonoise and multipath are interfered, represent actual range., for a shared satellite, the pseudo range difference of two cars (v=a, b) can be calculated by following formula:

$\mathrm{\Δ}{p}_{a,b}^s=\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^s+c\×\mathrm{\Δ}{t}_{a,b}+\mathrm{\Δ}{\mathrm{\ϵ}}_{a,b}^s---\left(2\right)$

According to formula (2) by pseudorange being carried out to the second order difference different GPS receiver clock error delta t of two cars that can disappear _v:

$\▿\mathrm{\Δ}{p}_{a,b}^{k,j}=(p_a^k-p_b^k)-(p_a^j-p_b^j)=\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}+({\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k)-({\mathrm{\ϵ}}_a^j-{\mathrm{\ϵ}}_b^j)---\left(3\right)$

In its computation of pseudoranges, compare other factor thermonoises conventionally enough little to such an extent as to negligible, therefore the second order difference of through type three can be converted into the calculating of relative distance, a plurality of shared satellite multipaths are interfered to the calculating of the factor.Due to the transmission environment of city zones of different complexity, for a shared satellite likely car receive its direct signal transmission and another car receives is reflected signal.In other words, the error that exists Multipath Transmission to cause to same satellite positioning signal, therefore error be generally the Multipath Transmission error causing due to different buildings reflected signals, need to detect the correlativity of the pseudorange of a shared satellite.When two cars is all received the direct signal transmission that shares satellite j, now only have the impact of reflected signal, namely we can obtain as lower approximate value:

$\▿\mathrm{\Δ}{p}_{a,b}^{k,j}-\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}\≈{\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k---\left(4\right)$

Using and share satellite j as basis of reference, when value than a default empirical value hour, we have reason to think that the positioning signal that two cars receives from satellite k has correlativity.Reference data satellite choose the usually elevation angle with respect to satellite based on vehicle.In the shared satellite of vehicle, the satellite of elevation angle maximum will be chosen for reference data satellite, because the probability that can receive its direct signal for this satellite vehicle is maximum.

Vehicle based on Kalman filtering is from predicting positioning navigation method:

Vehicle in travelling in urban area often can run into the situation that gps signal interrupts, and more owing to using shared satellite to cause this situation to occur.Kalman filtering can be from one group of limited, that comprise noise, the observation sequence of object space (may have deviation) to be doped to object space coordinate and speed.Therefore can be when predicting that by Kalman filtering gps signal interrupts in conjunction with vehicle coordinate prediction and GPS locating information the position coordinates of vehicle.Kalman filtering is a kind of Linear Minimum Variance algorithm for estimating, use the thought of Recursive Filtering to introduce the concept of state space, thereby can according to the observed reading of the state value of previous moment and current time, estimate next state value constantly according to the state transition equation of system.Use pseudorange as the observed reading of Kalman filtering herein, the fundamental equation of Kalman filtering is shown below:

X _k＝Φ _k,k-1X _k-1+W _k (5)

Z _k＝H _kX _k+V _k (6)

Formula (5) is state equation, and formula (6) is observation equation, wherein X _kfor state vector, Φ _{k, k-1}for the state-transition matrix of system, W _kfor system noise sequence and suppose that it meets that average is zero, covariance matrix is Q _kmultivariate normal distribution, i.e. V _k～N (0, Q _k), Z _kfor to time of day X _ka measurement, H _kfor observing matrix, it becomes observation space time of day spatial mappings, V _kfor its average of observation noise sequence is zero, covariance matrix is R _kand Normal Distribution, i.e. V _k～N (0, R _k).According to the observation model of formula (1) formula to pseudorange, known about state vector X _knonlinear equation, use state vector is got to local derviation.

At forecast period, wave filter uses the estimation of laststate, makes the estimation to current state.In new stage more, the predicted value that wave filter utilization obtains at forecast period the observed reading optimization of current state, to obtain more accurate new estimated value.The error covariance matrix P of Kalman filtering so _kcan represent with following formula:

$P_{k,k-1}=X_k{P}_{k-1,k-1}{X}_k^T+Q_k---\left(7\right)$

To sum up, use the vehicle of Kalman filtering from predicting that location navigation comprises following process:

1) determine the initial value of state vector, error covariance matrix;

2) predict next state vector, error covariance matrix value constantly;

3) state upgrades, and next arrives constantly, uses the locating information prediction pseudorange receiving and try to achieve observing matrix;

4) use error covariance matrix, in conjunction with the gain matrix of observing matrix calculating filter;

5) calculate the actual distance of vehicle and shared satellite observed reading and predicted value;

6) estimated value of computing mode vector sum error covariance matrix.

The next state renewal process that the second step to the of said process six steps are Kalman filter, in the situation that the initial value of the first step is provided, the mode that can constantly update with iteration obtains next predicted value constantly, does not need the historical information of hourly observation or estimation.

Vehicle collision avoidance judgment mechanism based on relative distance:

Vehicle a, the position of b when moment t can be estimated to obtain by its position at moment t-1, speed and acceleration prediction, the three-dimensional coordinate position that shares satellite (s=k, j) and vehicle (v=a, b) in the time of can be according to moment t is calculated and is learnt, for other shared satellites of vehicle (v=a, b), and the second order difference of the pseudorange of surveying and estimate the second order difference of finding range thereby result of calculation and empirical value compare and choose the shared satellite with high spatial correlativity.

Thus, rejected the low shared satellite of correlativity, the packet of the V2V Internet Transmission providing according to step 1, the speed of known vehicle acceleration m _awith enforcement direction we suppose that driver's burst traffic hazard reaction and the average experience time of processing are t.For the packet receiving, first judge the travel direction of vehicle, if vehicle a and vehicle b go in the same way, be that two vehicle heading vector angles are less than 90 °, what now may occur is rear-end collision; If it is that two cars travel direction vector angle is greater than 90 ° that vehicle a and vehicle b go in the same direction, now may occur be before hit accident; If if vehicle a and vehicle b side direction and row is two cars travel direction vector angle equals 90 °, what now may occur is that side is hit accident.Suppose | v _a| and | v _b| be respectively the mapping on direction vector of the three-dimensional velocity of vehicle a and vehicle b, above three kinds of safe distances of anticollision mechanism separately of classifying can be simplified shown as:

$D=\left\{\begin{array}{c}|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)-\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|\\ \left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)+\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)\\ \mathrm{Min}\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2,|v_b|t+\frac{1}{2}{s}_b{t}^2)\end{array}\right.---\left(8\right)$

In the time that relative distance is less than or equal to safe distance, thereby triggers vehicle carried device and automatically trigger the safety that emergency brake of vehicle mechanism ensures occupant.

Described in above, be specific embodiments of the invention and the know-why used, if the change of doing according to conception of the present invention, when its function producing does not exceed spiritual that instructions and accompanying drawing contain yet, must belong to protection scope of the present invention.

Claims (5)

1. the vehicle collision avoidance method based on collaborative relative positioning technology, is provided with GPS sending/receiving machine on positioned vehicle, and the direct or indirect positioned vehicle of several satellite in orbit, is characterized in that comprising the following steps:

1) analyze the first vehicle of relative positioning, the satellite information that the second vehicle receives, determine the shared satellite in orbit of the first vehicle, the second vehicle, set up collaborative relative positioning system model;

2) the first vehicle and the second vehicle are set up V2V network, and by sharing satellite in orbit transmission packet, packet information comprises: vehicle identification ID, vehicle destination absolute fix coordinate, vehicle three-dimensional velocity, shared satellite signal to noise ratio (S/N ratio) and shared satellite pseudorange;

3) when gps signal is stablized, adopt the enhancing relative positioning method based on pseudorange to calculate relative distance between the first vehicle, the second vehicle;

4) when gps signal interrupts, adopt vehicle based on Kalman filtering from predicting that positioning navigation method predicts relative distance between the first vehicle, the second vehicle;

5) according to step 3) or step 4) relative distance that obtains carries out the vehicle collision avoidance judgement based on relative distance, if in the time that relative distance being less than or equal to safe distance, triggering vehicle carried device and trigger emergency brake of vehicle mechanism.

2. a kind of vehicle collision avoidance method based on collaborative relative positioning technology according to claim 1, is characterized in that, the enhancing relative positioning method based on pseudorange comprises the following steps:

(11) according to formula (1) compute location vehicle v, use together the pseudorange between satellite s in-orbit:

$p_v^{\left(s\right)}={\mathrm{\ρ}}_v^{\left(s\right)}+s\×(\mathrm{\Δ}{t}_v-\mathrm{\Δ}{t}^{\left(s\right)})+d_{m,v}^{\left(s\right)}+d_{n,v}^{\left(s\right)}+{\mathrm{\ϵ}}_v^{\left(s\right)}---\left(1\right)$

Pseudorange carries out second order difference: $\▿\mathrm{\Δ}{p}_{a,b}^{k,j}=(p_a^k-p_b^k)-(p_a^j-p_b^j)=\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}+({\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k)-({\mathrm{\ϵ}}_a^j-{\mathrm{\ϵ}}_b^j)---\left(2\right);$

Wherein, represent the propagation delay time that bring in ionosphere, represent the propagation delay time that bring in troposphere, s represents the light velocity, Δ t _vrepresent positioned vehicle clocking error, Δ t ^(s)represent satellite clock error, represent the impact that thermonoise and multipath are interfered, represent actual range, a represents the first vehicle, and b represents the second vehicle, and k, j represent all shared satellite in orbit;

(12) calculate value and compare with predetermined threshold value, when calculated value is less than predetermined threshold value, judge corresponding shared satellite in orbit property relevant to positioned vehicle, retain the high shared satellite in orbit of relevance.

3. a kind of vehicle collision avoidance method based on collaborative relative positioning technology according to claim 2, is characterized in that, positioned vehicle a, b receive the direct signal transmission that shares satellite in orbit j, pseudorange second order difference result is $\▿\mathrm{\Δ}{p}_{a,b}^{k,j}-\▿\mathrm{\Δ}{\mathrm{\ρ}}_{a,b}^{k,j}\≈{\mathrm{\ϵ}}_a^k-{\mathrm{\ϵ}}_b^k.$

4. a kind of vehicle collision avoidance method based on collaborative relative positioning technology according to claim 1, is characterized in that, the vehicle based on Kalman filtering is from predicting that positioning navigation method comprises the following steps:

(21) determine the initial value of state vector, error covariance matrix;

(22) predict next state vector, error covariance matrix value constantly:

(23) state upgrades, and next arrives constantly, uses the locating information prediction pseudorange receiving and try to achieve observing matrix;

(24) use error covariance matrix, in conjunction with the gain matrix of observing matrix calculating filter;

(25) calculate the actual distance of vehicle and shared satellite observed reading and predicted value;

(26) estimated value of computing mode vector sum error covariance matrix.

5. a kind of vehicle collision avoidance method based on collaborative relative positioning technology according to claim 1, is characterized in that, the described vehicle collision avoidance judgement based on relative distance comprises:

$D=\left\{\begin{array}{c}|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)-\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|\\ \left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)+\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)\\ \mathrm{Min}\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2,|v_b|t+\frac{1}{2}{s}_b{t}^2)\end{array}\right.$

Wherein, | v _a| the mapping of the three-dimensional velocity of expression vehicle a on direction vector, | v _b| the mapping of the three-dimensional velocity of expression vehicle b on direction vector, s _athe acceleration that represents vehicle a, S _bthe acceleration that represents vehicle b, t represents driver's burst traffic hazard reaction and the average experience time of processing;

When vehicle a and vehicle b in the same way and row two vehicle heading vector angles be less than 90 °, now there is rear-end collision, $D=|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)-\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|;$

As vehicle a and vehicle b, going in the same direction is that two cars travel direction vector angle is greater than 90 °, before now occurring, hits accident, $D=|\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2)+\left(\right|v_b|t+\frac{1}{2}{s}_b{t}^2)|;$

When vehicle a and vehicle b side direction and row is two cars travel direction vector angle equals 90 °, now occur that side hits accident, $D=\mathrm{Min}\left(\right|v_a|t+\frac{1}{2}{s}_a{t}^2,|v_b|t+\frac{1}{2}{s}_b{t}^2).$