CN106568456B - Non-stop charging method based on GPS/Beidou positioning and cloud computing platform - Google Patents

Abstract

The non-stop charging method based on GPS/Beidou positioning and cloud computing platform belongs to the field of electronic information. The present invention uses the GPS track point as the center circle to select the road section set belonging to the circle as the selected path. The present invention adds common constraints of direction constraint conditions, and proposes how to judge the direction. Judgment of driving direction An up and down judgment algorithm based on the inflection point of the line is proposed to judge the direction. The selection of the final matching path is based on the coefficient trajectory similarity algorithm of Euclidean distance. The invention can realize the automation and simplification of road vehicle supervision under the optimization of driving track, alternative route and final matching route.

Description

Non-stop charging method based on GPS/Beidou positioning and cloud computing platform

technical field

The invention belongs to the field of electronic information, and is an automatic electronic charging method based on GPS/Beidou positioning and a cloud computing platform and applied to highways, tunnels and large bridges.

Background technique

ETC (Electronic Toll Collection, electronic toll collection system) technology, as shown in Figure 1, relies on the Internet to upload vehicle information to the application database through the workstation in the form of images or mobile keys in real time. Traditional ETC technology has problems such as slow recognition speed, high installation cost, and difficult operation and maintenance, and some toll stations are located in remote locations, making it difficult to install and maintain ETC hardware equipment. In recent years, a new generation of electronic toll collection system based on GPS and Beidou positioning system is gradually developing. It records the route and distance of vehicles through satellite positioning, automatically calculates and deducts road tolls, and effectively solves the above-mentioned drawbacks of traditional ETC technology. . At the same time, the development of technologies such as mobile Internet communication, Internet of Vehicles and GIS has also provided a corresponding technical foundation for the non-stop toll collection system based on GPS/Beidou and cloud computing platforms.

Contents of the invention

Aiming at the shortcomings of the traditional ETC, the present invention designs a non-stop charging system based on GPS/Beidou and a cloud computing platform. The system consists of a vehicle-mounted positioning terminal, GIS-based and cloud-based non-stop charging software. Among them, the vehicle positioning terminal returns the vehicle GPS data collected in real time to the cloud server system through the 3G/4G wireless network through GPS satellite/Beidou positioning. The non-stop payment software includes the path judgment strategy involved in the system, the vehicle positioning service and the fee settlement service, and the main work involved in the present invention is as follows:

(1) As shown in Figure 2, an ETC system architecture integrating mobile Internet, GPS/Beidou positioning, GIS and cloud platform is designed.

(2) In order to make the vehicle GPS data with certain errors corrected to the correct driving road in GIS, a path matching algorithm based on GPS/Beidou trajectory is designed and applied.

(3) The GIS-based vehicle positioning business process is designed.

(4) The fee settlement process of the cloud-based GIS non-parking toll system is designed.

The core algorithm of the non-stop charging system based on GPS/Beidou and cloud computing platform is as follows:

Typical GPS/Beidou sports car data is a series of ordered GPS track points containing vehicle latitude and longitude information, which is mainly used to determine the current position of the vehicle in real time. Relying solely on GPS/Beidou satellite positioning, its accuracy is easily affected by the external environment and produces errors. In the intersection road section shown in Figure 3, A, B, C, and D are road section nodes respectively, and the solid line segments connected by them represent the road trajectory; GPS ₁ ~ GPS ₄ are vehicle GPS collection points in turn, and they are connected and The resulting dotted line segment is the vehicle trajectory. Assume that GPS track points GPS ₁ to GPS ₃ can be uniquely matched to road segment AB, but when passing through point B, the number of alternative matching road segments increases due to the appearance of a fork in the road, and the distance from GPS ₄ to road segment BC and road segment BD is almost equal , at this time, it is difficult for the general matching algorithm to correctly judge the subsequent driving section of the vehicle. This problem is called the Y-junction problem.

If the above road judgment errors occur in this system, it will seriously affect the normal operation of the toll collection system. Therefore, on the basis of predecessors' research results, the present invention improves and combines the alternative route set algorithm based on traffic restrictions and the polyline similarity algorithm based on editing costs, and realizes a route matching algorithm based on GPS trajectory. The GPS track points of the vehicle are screened and matched with the surrounding road tracks, and finally the actual driving road of the vehicle can be accurately judged to ensure the normal operation of the system.

The algorithm of the present invention uses the vehicle GPS sampling point set and the path track point set as basic data, and realizes the accurate judgment of the real vehicle driving road through three steps, and the specific steps are as follows:

(1) Search the set of alternative paths

According to the continuous vehicle GPS collection points, the vehicle driving trajectory is drawn, and with GPS _i as the center, all possible driving roads within its preset radius are searched to form an alternative route set RC _i ;

(2) Filter the set of alternative paths

Based on the actual traffic restrictions and the geometric connectivity of the road network, the intersection R _insect (i) of RC _i and the subsequent feasible road section set PN _i of the overall path P _i at each stage is calculated. And R _insect (i) updates and reorganizes the set of candidate overall paths P _i+1 .

(3) Select the final matching path

Traverse all the track points in turn, and use the idea of relative editing cost to calculate the similarity r between the path in P and the GPS track path. The one with the largest r value is regarded as the closest to the actual road track, and it is taken as the final matching path.

Although the geometric shapes of the driving trajectory and the road trajectory are very close, they have obvious characteristic differences due to the different point collection methods.

First of all, the density of sampling points is different between the two. The route trajectory is drawn by manually collecting points, and the collected road nodes are evenly distributed in the trajectory; the driving trajectory is collected in real time by the vehicle positioning terminal. If the vehicle is moving slowly or stationary, then the GPS trajectory points during this period Agglomeration or even overlap will occur.

Second, the sampling points of the two have different meanings. The sampling points of the line trajectory are all key nodes of the road, most of which are key nodes that change the direction of the trajectory. Although they are rare, the entire line trajectory can be drawn simply and clearly; while the vehicle trajectory is dense, and a section of the trajectory tends to be straight. It often contains multiple GPS collection points, and most of the sampling points do not change the direction of the driving trajectory. Such points have no effect on the selection of the following path set and the comparison of road trajectory similarity.

Therefore, this algorithm needs to simplify the GPS driving trajectory of the vehicle, remove invalid sampling points, and only retain sampling points that can significantly change the direction of the trajectory and are evenly distributed, as shown in Figure 4.

Assuming that the minimum steering angle λ is defined as 30°, and the weight of the minimum sampling distance interval η is 3, when a node has a turning angle greater than λ and a distance interval greater than η from its front and rear nodes, it is called a key point. For the trajectory TR=p ₁ p ₂ p ₃ p ₄ p ₅ p ₆ p ₇ p ₈ in the figure, only the rotation angles of p ₃ , p ₄ and p ₅ are greater than λ, and the weight of the distance from the sampling points before and after it is greater than η, Then add p ₃ , p ₄ and p ₅ into the key node set. In addition, since p ₁ and p ₇ are the start point and end point of the trajectory respectively, they also need to be added to the key node set. Finally, the node set is sorted according to the node serial number, and the simplified vehicle trajectory TR _key = p ₁ p ₃ p ₄ p ₅ p ₇ formed by the connection of key nodes in the figure is obtained.

The value principles of λ and η should be consistent with the actual situation of road inflection points, so that the driving trajectory can be approximated to the line trajectory to the greatest extent. The calculation method of the movement direction change value θ of the trajectory point p is shown in the formula (1-1):

in is the vector between p _i-1 and p _i points, and Among them, p _{i, x} , p _{i, y} and p _{i-1, x} , p _{i-1, y} are the coordinates of trajectory points p _i and p _i-1 respectively.

Determination of the set of alternative paths

In the case of errors in vehicle GPS positioning, the driving trajectory depicted by GIS may be offset, as shown in Figure 5. The solid line segment in the figure is the track of the road segment, and the dashed line segment is the driving track formed by connecting the GPS track points of each vehicle in sequence.

Each track point will cover different road sections within a certain radius. For example, the road sections covered by GPS point ₁ within the radius of search_scop are a, d, and f, while the road sections covered by GPS point ₂ are a, b, and c. Then it can be said that when the system scans GPS ₁ point, road sections a, d, f are the alternative route sets of GPS ₁ point. Similarly, the road sections a, b, and c are the alternative route sets of GPS ₂ , and the road sections b, e, and g are the alternative route sets of GPS ₃ points. For each GPS track point pi, the set of all geometrically connected road segments in the positioning error area is recorded as R(p _i ).

Among them, the value of the radius Search_scop changes according to the road conditions of each road section, but the following principles should be followed: first, the search radius Search_scop should be greater than the maximum error value of the positioning accuracy of the GPS device, so as to prevent the actual driving road from being divided outside the search radius; The search radius should cover the lanes adjacent to the current GPS point, so as not to cause too little basic data due to too small a search radius.

Filtering of the set of alternative paths

In the actual highway road network, in addition to the geometric connectivity of the road, there are also driving restrictions such as driving direction restrictions, overspeed restrictions, and vehicle type restrictions.

When the set of alternative routes R( _pi ) of GPS track point p _i is determined, it is necessary to screen out possible alternative routes at the current stage from all geometrically connected paths around the vehicle according to the above driving constraints. And traverse each GPS positioning track point p _i in turn, respectively filter and update the last screening results, and finally get all the candidate path sets.

For example, in the road conditions shown in Figure 4, for a vehicle at GPS point ₁ , its candidate route set is R( _pi ), and the geometric connected paths recorded in the database for this road network include a, d and f, and according to the driving direction restrictions of each road section, the actual driving direction of the vehicle is only consistent with the direction restriction of road section a, that is, the subsequent path set of this path is only a, and R(p _i ) is obtained after filtering the driving restrictions The set of subsequent paths is denoted as NR(p _i ).

Use T _i (topo) to represent the topological constraints of the R(p _i ) section (the topological constraints are defined nodes and road paths), and use T _i (dirc) to represent the directional constraints of the R(p _i ) section, then The calculation method of the immediate subsequent feasible road segment set NR(p _i ) of R(p _i ) is shown in formula (1-2):

NR(p _i )＝{R _j ∈R(p _i )|T _i (topo)∩T _i (dirc)} (1-2)

If i=0, then the current point is the initial trajectory point, set the number of elements of R(p ₀ ) as k, and initialize k geometrically connected path sets and the overall candidate path set R(path _i ); if i≠0, Then the current point is the middle track point, calculate the intersection R _insect (i) of R(p _i ) and the direct follow-up feasible road segment set R(p _i-1 ) of the previous stage R(p _i-1 ); if R _insect ( i) is empty, the current path R(i) is an invalid path, delete it from the set of geometrically connected paths in the current stage; if R _insect (i) has one and only one road segment, then update the follow-up path set R(p _i ) in the current stage ) and R(path _i ); if R _insect (i) contains m road sections, copy and split R(path _i ) into m total alternative paths, R _insect (j)(j∈[1,m]) As the last road segment of R _insect (j), update its immediate follow-up feasible path set at the same time.

Finally, after all the GPS track points are judged, the final overall candidate path set R(path) can be obtained.

Driving direction judgment

It is worth mentioning that among all driving restrictions, geometric connectivity, speed restrictions, vehicle restrictions and other restrictions can be judged according to the node attributes stored in the database, while the judgment of driving direction restrictions is more complicated.

The invention proposes an uplink and downlink judging algorithm based on the line inflection point. The road section nodes that describe the shape of the road are stored in the database, and the points that can determine the direction of the road are regarded as road inflection points. The driving direction of the vehicle between two turning points can be determined more accurately by using the road turning point, and the driving direction of the road turning point is determined, so that the driving direction of the vehicle relative to the entire road section can be judged.

Because the points taken by the algorithm are all inflection points of the line, that is, the line segment between two inflection points is a straight line segment, so it can be considered that the scenarios where the algorithm is used are all straight lines. List all four possible line models as shown in Fig. 6, which respectively represent the relative positions of the four vehicles and the inflection point on the two inflection point lines at time t ₁ and t ₂ . Among them, S ₁ and S ₂ are the inflection points of the route, P ₁ and P ₂ are the front and rear GPS track points of the vehicle within the time t ₁ and t ₂ , and the direction of the arrow is the upward direction.

It can be seen from the figure that in case one, the front and rear positions of the vehicle are both before point S ₁ ; in case two, the front and rear positions of the vehicle are located between two inflection points; in case three, the front and rear positions of the vehicle are both after point S ₂ ; Although the position is between two points, the direction is opposite to that of Case 2.

After it is determined that the vehicle belongs to the above four situations, the driving direction can be judged according to the distance changes between the front and rear positions of the vehicle and the two inflection points respectively. Taking case 1 as an example, as shown in Figure 7, when car P travels from time t ₁ to time t ₂ , the change in distance to station S ₁ is: △L=p ₁ s ₁ -p ₂ s ₁ , car P For S ₂ station distance variation is: △L'=p ₁ s ₂ -p ₂ s ₂ .

If △L<0 and △L'>0, it can be judged that car P is away from station S ₁ and approaching station S ₂ , and the traveling direction is from S ₁ to S ₂ ; otherwise, it is approaching station S ₁ and away from station S ₂ station, the direction of travel is S ₂ to S ₁ . From this, we can know the driving direction of the vehicle relative to this section of road. If the driving direction is consistent with the direction of the road section in the database, it means that the vehicle meets the driving direction restriction of the road section, and this road section will be added to the subsequent driving route set NR(p _i ); otherwise, it means that the driving direction violates the road section direction restriction, and this road section should be deleted from the alternative route set R(p _i ).

Selection of the final matching path

After traversing all GPS track points, get the final candidate route set NR(p _n ), NR(p _i ) may contain one or more routes, and they all conform to the traffic restriction conditions of the driving track, but there is only one route is the actual driving path of the vehicle. At this time, it is necessary to compare the trajectory of the driving route with each alternative path trajectory, so as to obtain the path with the highest similarity as the final matching path.

For the method of trajectory similarity comparison, there is a path similarity judgment method based on curve similarity, but this method requires a coordinate system, and the distance between curves will affect the size of the final similarity value ScoreSim. However, the actual roads are mostly sparse trajectories connected by short straight line segments, rather than curves with uniform curvature changes, so the method used in the literature is not suitable for path trajectory comparison in the actual road conditions studied in this project.

Another existing algorithm takes into account the sparse features of driving trajectories, and proposes a sparse trajectory similarity calculation based on Euclidean distance. The edit distance (the minimum number of edits required to change from one trajectory to another ) idea is applied to the trajectory similarity calculation, that is, to calculate the editing cost spent from editing trajectory P to trajectory Q. The smaller the cost value, the higher the similarity between two sparse trajectories. However, this method only considers the absolute Euclidean distance between the two trajectories, does not unify the coordinate systems of the corresponding points of the two trajectories, and ignores the relative editing cost between the corresponding points of the two trajectories. For example, if there is another interfering trajectory between two trajectories, its editing cost will always be less than the cost of the desired trajectory, resulting in a judgment error.

Therefore, this algorithm improves the existing algorithm and proposes a relative Euclidean distance calculation method in a unified coordinate system. By calculating the coordinate offset of the track points before and after the track Q, the track P is correspondingly displaced to realize the distance of the track line segment. The coordinates are unified, so the calculated Euclidean distance is relative, and the error in the existing calculation method is avoided.

The calculation method of the Euclidean distance is to represent the trajectory of the moving object with a coordinate vector of the same dimension, and then calculate the Euclidean distance between the corresponding two trajectory points at each moment, and then synthesize these distances to obtain the trajectory Euclidean distance between. For example, in two-dimensional space, the Euclidean distance calculation method between the two trajectories R and S is shown in formula (1-3):

In the formula, _{k is} the number of sampling points of _trajectory R and S, E(R,S ₎ is the Euclidean distance between them; _{i, x} , s _{i, y} represent their respective x and y coordinates, and distance represents the Euclidean distance between track points r _i and s _i .

For the vehicle trajectory of the trajectory streamlining process, firstly, the coordinates of each key point are serialized to form a sequence P; the trajectory coordinates of the driving route are serialized to form a sequence Q. Let m and n be the lengths of P and Q respectively, p _i is the i-th element of P, q _i is the i-th element of the sequence Q, and q _i ' is the i-th element in the sequence Q converted to the P coordinate system element. In the process of calculating the relative Euclidean distance from sequence P to sequence Q, the sequence Q is used as the reference to transform P to Q. The schematic diagram of P and Q trajectory is shown in Fig. 8 .

The detailed calculation steps are as follows:

(1) Calculate the conversion offset

To calculate the relative editing cost from B to E, point E needs to be substituted into the coordinate system of trajectory P. First, it is necessary to calculate the X and Y coordinate offsets from the starting point D to E respectively, where O is the origin coordinate system, as shown in formulas (1-4) and (1-5):

(2) Coordinate replacement

Move A to E' according to the offset calculated in the first step, as the position of trajectory Q in the coordinate system of trajectory P. The calculation method of E' coordinates is shown in formula (1-6), (1-7):

substitute(p _i,x ,q _i,x )=p _i-1,x +der(q _i,x ) (1-6)

substitute(p _i,y ,q _i,y )=p _i-1,y +der(q _i,y ) (1-7)

(3) Calculate the Euclidean distance

At this point, point E in trajectory Q has been transformed into the coordinate system of trajectory P, and the geometric distance from B to E' can be regarded as the relative editing cost of editing B to E', see formula (1-8):

Among them, |p _i -q _i '| is the Euclidean distance between two coordinate points (the Euclidean distance in two-dimensional and three-dimensional space is the actual distance between two points). The editing cost between two trajectories is the sum of the editing costs of each point, see formula (1-9):

When the editing cost between two sparse trajectories is greater, it means that more editing operations are needed to realize the conversion between the two trajectories, which means that the similarity between the two trajectories is smaller; The greater the similarity. Substituting the vehicle trajectory A and the trajectory point coordinates corresponding to the path B in the final candidate path set NR(p _n ) into the formula (1-9) in turn to obtain the similarity value of each path trajectory Similarity(A,B) _i , and finally select the one with the largest value as the final matching path.

(4) GIS-based vehicle positioning business process is designed:

As shown in Figure 9, the GPS data acquisition and storage process for the vehicle positioning terminal is mainly completed by the vehicle positioning terminal servo module, which acts as a transfer station for GPS data transmission and forwards the GPS data to the GPS data module and the GIS servo module respectively. First, the cloud server system will send control messages to the vehicle positioning terminal through the servo module according to user needs. When the vehicle positioning terminal receives the control message, it performs corresponding control operations. Secondly, the vehicle GPS data collected in real time is returned to the cloud server system through the 3G/4G wireless network. During this transmission process, the server has been in a listening state, waiting for the data message from the vehicle positioning terminal. After receiving the data message, the cloud server will simultaneously send the GPS message to the GPS data module and the GIS servo module. The GIS-oriented matching process serializes real-time GPS data and provides it for GIS loading and simulation. The GIS servo module waits for the parsed GPS data transmitted by the vehicle positioning terminal servo module. The server searches the road network data around the vehicle in the road network database according to the vehicle coordinates, serializes it together with the GPS data of the vehicle trajectory, and then loads it into the GIS, and passes through the screening and editing of the path matching algorithm based on the GPS trajectory introduced above. Calculate the cost and get the path matching result. Finally, the GIS server module also needs to submit the matching result to the settlement module for its final billing and settlement work.

(5) Liquidation module process:

As shown in Figure 10, firstly, the clearing module reads the matching result, and queries the "whether toll" attribute in the road network data table according to the road section ID. If it is a toll road, it means that the vehicle has not left the toll road section, and continues to receive subsequent matching results . If it is a non-toll road section, the clearing module needs to query the vehicle withholding record table according to the vehicle ID. At this time, the vehicle may be in one of two situations: if the last record does not record the vehicle entry record, it means that the vehicle has not yet entered the toll collection Road, at this time, the vehicle ID, road section ID and driving time should be recorded; if the vehicle driving record has been recorded in the last record, it means that the vehicle has passed the toll road and left, and the clearing module should drive the vehicle away The time information is recorded, and the driver's personal account is charged and deducted according to the vehicle type and the corresponding road section toll standard. In this way, the settlement business of the system for driving vehicles is realized.

Invention effect

Through the construction of GPS/Beidou and cloud computing platform supported by mobile Internet communication technology, this system can realize the automation and simplification of road vehicle supervision under the optimization of driving trajectory, alternative path and final matching path. The non-stop charging system based on GPS/Beidou and cloud computing platform researched by the present invention provides a complete set of solutions for the information management of public transportation, which improves the intelligence level and work efficiency of road supervision work. The design method and the technology used in the implementation process also provide some new ideas for the development of intelligent transportation systems.

1. Determination of the selected path set: prior art: directly use GPS positioning and then GIS to describe the driving track

Improved algorithm: take the GPS track point as the center circle and select the road segment set belonging to this circle as the selected path.

2. Screening of the selected path set: in the prior art, the selected path set is narrowed only according to the topological constraints.

Improved Algorithm: Add direction constraints and common constraints, and propose how to judge the direction.

3. Judgment of driving direction: An uplink and downlink judgment algorithm based on the inflection point of the line is proposed to judge the direction.

4. Selection of the final matching path: prior art: (1) method for judging path similarity based on curve similarity.

(2) Coefficient trajectory similarity algorithm based on Euclidean distance.

Improved algorithm: Improvement based on the combination of two existing technologies: the relative Euclidean distance algorithm of the unified coordinate system.

Description of drawings

Figure 1 Traditional ETC technical framework diagram.

Fig. 2 is a system architecture diagram of the present invention.

Fig. 3 is a schematic diagram of the Y-junction intersection of the present invention.

Fig. 4 is a simplified schematic diagram of the trajectory of the present invention.

Fig. 5 is a schematic diagram of path matching in the present invention.

Fig. 6 shows several relative positions of the present invention.

Fig. 7 is a schematic diagram of judging the driving direction of a straight road according to the present invention.

Fig. 8 is a schematic diagram of P and Q trajectories of the present invention.

Fig. 9 is a flow chart of the vehicle positioning service based on GIS in the present invention.

Fig. 10 is a flow chart of the liquidation business of the present invention.

Detailed ways

1. The non-stop charging method based on GPS/Beidou positioning and cloud computing platform is characterized in that:

The system includes ETC integrating mobile Internet, GPS/Beidou positioning, GIS and cloud platform;

The specific steps are as follows:

(4) Search the set of alternative paths

According to the continuous vehicle GPS collection points, the vehicle driving trajectory is drawn, and with GPS _i as the center, all possible driving roads within its preset radius are searched to form an alternative route set RC _i ;

(5) Filter the set of alternative paths

Based on the actual traffic restrictions and the geometric connectivity of the road network, calculate the intersection R _insect (i) of RC _i and the subsequent feasible road segment set PN _i of the overall path P _i at each stage; and update and reorganize the alternatives by R _insect (i) Overall path set P _i+1 ;

(6) Select the final matching path

Traverse all the track points in turn, and use the idea of relative editing cost to calculate the similarity r between the path in P and the GPS track path, and the one with the largest r value is regarded as the closest to the actual road track, and it is taken as the final matching path; the details are as follows :

By calculating the coordinate offset of the trajectory points before and after the trajectory Q, the trajectory P is correspondingly displaced to realize the unification of the coordinates of the trajectory line segment; and the calculation method of the Euclidean distance is to represent the trajectory of the moving object with a coordinate vector of the same dimension , and then calculate the Euclidean distance between the corresponding two trajectory points at each moment, and then synthesize the Euclidean distance between the corresponding two trajectory points at each moment to obtain the Euclidean distance between the trajectories;

For the vehicle trajectory of the trajectory streamlining process, firstly, the coordinates of each key point are serialized to form a sequence P; the trajectory coordinates of the driving route are serialized to form a sequence Q; p _i is the i-th element of P, and q _i is the sequence Q q _i ' is the i-th element converted from the sequence Q to the P coordinate system; in the process of calculating the relative Euclidean distance from the sequence P to the sequence Q, the sequence Q is used as the benchmark, so that P Transform to Q operation;

The detailed calculation steps are as follows:

(4) Calculate the conversion offset

To calculate the relative editing cost from B to E, point E needs to be substituted into the coordinate system of trajectory P; firstly, the X and Y coordinate offsets of the starting point D to E need to be calculated respectively, where O is the origin coordinate system, As shown in formulas (1-4) and (1-5):

(5) Coordinate replacement

Move A to E' according to the offset calculated in the first step, as the position of trajectory Q in the coordinate system of trajectory P. The calculation method of E' coordinates is shown in formula (1-6), (1-7):

substitute(p _i,x ,q _i,x )=p _i-1,x +der(q _i,x ) (1-6)

substitute(p _i,y ,q _i,y )=p _i-1,y +der(q _i,y ) (1-7)

(6) Calculate the Euclidean distance

At this point, point E in track Q has been transformed into the coordinate system of track P, and the geometric distance from B to E' is regarded as the relative editing cost of editing B to E', see formula (1-8):

Among them, |p _i -q _i '| is the Euclidean distance between two coordinate points; and the editing cost between two trajectories is the sum of the editing costs of each point, see formula (1-9):

Substitute the vehicle trajectory A and the trajectory point coordinates corresponding to route B in the final candidate route set NR(p _n ) into formula (1-9) and obtain the similarity value Similariartiy(A,B) _i of each route trajectory, Finally, the one with the largest value is selected as the final matching path.

Assuming that the minimum steering angle λ is defined as 30°, when the turning angle of a certain node is greater than λ, p ₃ , p ₄ and p ₅ are added to the key node set; the starting point and end point of the trajectory also need to be added to the key node set; Finally, the node set is sorted according to the node serial number, and the simplified vehicle trajectory connected by the key nodes is obtained.

For each GPS track point p _i , record all geometrically connected road segments in the positioning error area as R(p _i );

The value of the search radius Search_scop of the alternative path should follow the following principles: first, the search radius Search_scop should be greater than the maximum error value of the positioning accuracy of the GPS device; second, the search radius should cover the lane adjacent to the current GPS point;

When the candidate path set R( _pi ) of the GPS track point p _i is determined, it is necessary to filter out the candidate paths of the current stage from all the geometrically connected paths around the vehicle according to the driving constraints; and traverse each GPS in turn Locate the track point p _i , respectively filter and update the last screening results, and finally get all the candidate path sets.

The road section nodes that describe the shape of the road are stored in the database, and the point that can determine the direction of the road is called the road inflection point; the road inflection point is used to judge the driving direction of the vehicle between the two inflection points, so as to determine the driving direction of the vehicle relative to the entire road section.

The present invention adopts following implementation scheme:

As shown in the system architecture diagram in Figure 2, the implementation process of the system mainly includes four stages: acquisition, transmission, matching, and settlement, which correspond to GPS/Beidou data acquisition, data transmission and analysis, path matching, and driver account of the vehicle positioning terminal respectively. settlement of charges.

(1) Acquisition: The vehicle positioning terminal deployed on the actual driving vehicle is the data acquisition platform of the system, and is mainly responsible for collecting real-time GPS/Beidou data of the vehicle.

(2) Transmission: The transmission of data is mainly completed by the 3G/4G module of the vehicle positioning terminal. This module encapsulates the collected data and vehicle identification information into a message with a certain format, and transmits it to the cloud platform through the 3G/4G network server system.

(3) Matching: When the server system receives the vehicle GPS data collected by the vehicle positioning terminal, it needs to match it with the road network data pre-loaded in the GIS, and correct the vehicle trajectory through the path matching algorithm, so as to find the actual driving path of the vehicle .

(4) Settlement: According to the matching result of GIS, the server system starts to call the settlement module, and combines the vehicle information and road section toll standards to perform corresponding billing and settlement work on the personal account of the vehicle driver.

Claims (4)

1. the non-stop charging method based on GPS/ Beidou positioning and cloud computing platform, it is characterised in that:

System includes the ETC of integrated mobile Internet, the positioning of GPS/ Beidou, GIS and cloud platform；Shown in specific step is as follows:

(1) alternative path collection is searched for

Vehicle driving trace is depicted according to continuous vehicle GPS collection point, with GPS_iFor the center of circle, search in its pre-set radius All possible travel forms alternative path collection RC_i；

(2) alternative path collection is screened

Based on practical drive a vehicle limitation and road network geometry connectivity, RC is calculated_iWith each stage overall path P_iLater road Section collection PN_iIntersection R_insect(i)；And by R_insect(i) it updates and recombinates alternative overall path collection P_i+1；

(3) final coupling path is chosen

All tracing points are successively traversed, the phase in path and GPS track path in P is calculated separately using the thought of opposite editor's cost Like degree r, r value the maximum be then considered as it is closest with real road track, and as final coupling path；It is specific as follows:

By calculating the coordinate shift amount of the front and back track Q tracing point, track P is subjected to corresponding trajectory displacement, to realize track The coordinate unification of line segment；And the calculation method of Euclidean distance is by the coordinate vector table of the track identical dimensional of mobile object The Euclidean distance between corresponding two tracing points is engraved when showing, then calculating each, it is then two corresponding to being engraved when each Euclidean distance between tracing point is integrated, and Euclidean distance between track is obtained；

The track of vehicle of processing is simplified for track, first by its each key point coordinate sequence, constitutes sequence P；By driving line The trajectory coordinates on road serialize, and constitute sequence Q；p_iFor i-th of element of P, q_iFor i-th of element of sequence Q, q_i' it is sequence Q In be transformed into i-th of element under P coordinate system；During sequence of calculation P is converted to the opposite Euclidean distance of sequence Q, with On the basis of sequence Q, make P to Q operational transformation；

Steps are as follows for calculating in detail:

(1) transition deviation amount is calculated

Want opposite editor's cost of calculating B to E, it is necessary to which E point is updated in the coordinate system of track P；Firstly the need of respectively The X, Y coordinates offset of starting point D to E is calculated, wherein O is origin system, as shown in formula (1-4), (1-5):

(2) coordinate is replaced

A is moved into E ' according to the offset that the first step calculates, as position of the track Q under the P coordinate system of track, E ' coordinate is calculated Method is shown in formula (1-6), (1-7):

substitute(p_i,x,q_i,x)=p_i-1,x+der(q_i,x) (1-6)

substitute(p_i,y,q_i,y)=p_i-1,y+der(q_i,y)

(1-7)

(3) Euclidean distance is calculated

The E point in the Q of track has been converted into the coordinate system of track P at this time, and the geometric distance of B to E ', which is regarded as, edits B to the phase of E ' To editor's cost, formula (1-8) is seen:

Wherein, | p_i-q_i' | the Euclidean distance between two coordinate points；And editor's cost between two tracks is then each point editor The sum of cost is shown in formula (1-9):

By vehicle driving trace A, and final alternative path collection NR (p_n) in tracing point coordinate corresponding to the B of path substitute into formula (1- 9) and the similarity value Similartiy (A, B) of each path locus is obtained_i, finally select its intermediate value the maximum as final matching Path；

(4) the vehicle positioning operation flow based on GIS is devised:

GPS data acquisition and storage process on-vehicle positioning terminal servo module towards on-vehicle positioning terminal is completed, it is as GPS GPS data, is transmitted to GPS data module and GIS servo module by the terminal of data transmission respectively；Cloud server system first Control message can be sent to on-vehicle positioning terminal by servo module according to user demand；Disappear when on-vehicle positioning terminal receives control Corresponding control operation is carried out after breath；Secondly, the vehicle GPS data acquired in real time are returned to cloud clothes by 3G/4G wireless network Business device system；In this transmit process, server is constantly in listening state, waits the data-message of on-vehicle positioning terminal；When After receiving data-message, GPS message can be sent to GPS data module and GIS servo module simultaneously by Cloud Server；Towards The matching process of GIS serializes real-time GPS data, and is supplied to GIS load and emulation；It is vehicle-mounted fixed that GIS servo module waits The GPS data by parsing of position terminal servo module transmission；Server searches vehicle according to vehicle coordinate in Traffic network database It is serialized together with vehicle driving trace GPS data, is then subsequently loaded into GIS, and by being based on by surrounding road net data The screening of the Path Matching Algorithm of GPS track and editor's cost calculate, and obtain route matching result；Last GIS servo module is also It needs for matching result to be forwarded to clearance module, last charge accounting work is carried out for it；

(5) block process is cleared:

Matching result is read by clearance module first, and " whether charging " attribute in road net data table is inquired according to road section ID, If toll road, illustrate that vehicle does not sail out of charging section also, continues to subsequent match result；If it is non-charging section, Clearance module needs to pay button record sheet according to vehicle ID enquiring vehicle, and at this moment vehicle is likely to be among two kinds of situations: if last Record is driven into without recording vehicle in primary record, then represents vehicle and does not enter the toll road also, it at this time should be by vehicle ID, road Section ID is recorded with the time is driven into；If having recorded the vehicle in last time record drives into record, represents vehicle and have already passed through charge It road and sails out of, vehicle time of departure information should be placed on record by this moment clearing module, and according to type of vehicle and corresponding Section charging standard corresponding charging carried out to driver personal account pay button work；Realize the system for Travel vehicle with this Account settlement business.

2. method according to claim 1, it is characterised in that:

Assuming that defining minimum steering angle λ is 30 °, when the front and back section corner of some node is greater than λ, by p₃、p₄And p₅It is added crucial Node collection；Track initial point and end point are also required to be added into key node collection；Finally, node collection is arranged by node ID Sequence, obtain being formed by connecting by key node simplifies vehicle driving trace.

3. method according to claim 1, it is characterised in that:

For each GPS track point p_i, the section collection of its all geometry connection in position error area is denoted as R (p_i)；

The value of the search radius Search_scop of alternative path should follow following principle: first is that search radius Search_scop is answered Greater than the worst error value of GPS device positioning accuracy；Second is that search radius, which should cover current GPS point, closes on lane；

As GPS track point p_iAlternative path collection R (p_i) after determination, it is necessary to according to driving restrictive condition from vehicle periphery institute In the path for thering is geometry to be connected to, the alternative path of current generation is filtered out；And successively traverse each GPS positioning tracing point p_i, point The other the selection result to last time is screened and is updated, and all alternative path collection are finally obtained.

4. method according to claim 1, it is characterised in that:

The point that can wherein determine road driving direction is become road by the section node that description road shape is stored in database Inflection point；Driving direction of the vehicle between two inflection points is judged using road inflection point, to judge vehicle relative to whole section Driving direction.