JP2016177487A - Vehicle search system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress delay in vehicle search.SOLUTION: In an on-vehicle device 12, a communication section 130 transmits vehicle information related to a vehicle to a server 14. A vehicle action prediction section 132 repetitively predicts the prediction route of the vehicle on the basis of the vehicle information until predetermined prediction time. When a search condition transmitted from the server 14 is received, an inference section 134 calculates a confidence factor on the basis of the predicted route of the vehicle. Then, the communication section 130 transmits the confidence factor to the server 14. In the server 14, a simple prediction section 146 selects the candidates of the vehicle satisfying the search condition on the basis of the search condition and the vehicle information transmitted from each one of multiple on-vehicle devices 12. A communication section 142 transmits the search condition to each one of the on-vehicle devices 12 of the vehicles selected as the candidates of the vehicle. Then, a vehicle search section 148 searches for the vehicle satisfying the search condition on the basis of the confidence factors transmitted from each one of the on-vehicle devices 12.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a vehicle search system, and more particularly to a vehicle search system that searches for a vehicle that satisfies a search condition.

Conventionally, a search technique for searching for a vehicle that satisfies a certain condition from a database is known.
For example, a technique for searching for past vehicle travel history data is known (Patent Document 1).

  Further, a technique for receiving destination information and route information of a car navigation system on the server side and displaying an appropriate advertisement on the car navigation system based on the information is known (Patent Document 2).

  Moreover, there exists a prior art regarding a vehicle behavior, particularly a route prediction model (Patent Document 3, Non-Patent Document 1, and Non-Patent Document 2).

JP 2000-163690 A JP 2010-102685 A JP 2013-120526 A

Brian D. Ziebart, Andrew L. Maas, Anind K. Dey, and J. Andrew Bagnell, “Navigate Like a Cabbie: Probabilistic Reasoning from Observed Context-Aware Behavior”, UbiComp'08 Proceedings of the 10th international conference on Ubiquitous computing John Krumm, “A Markov Model for Driver Turn Prediction”, SAE 2008 World Congress

  Here, “Future vehicle behavior search” is a complicated problem involving uncertainty, and is difficult to implement in a normal database system that handles past moving body information, such as Patent Document 1 described above. For example, when an ordinary database system is used in response to an inquiry “List cars that are likely to pass through road A in the future up to time T,” a “destination and route input to car navigation system” can be displayed as a number of vehicles. It is stored in a database that manages the input to the car navigation system in real time via communication and is searched. Such an idea can be found in, for example, Patent Document 2 described above. Such a technique is effective for a vehicle in which the driver inputs the destination to the car navigation system, but there is a problem that the route of the other vehicle is uncertain and cannot be searched. .

  On the other hand, future vehicle behavior can be predicted by using a “vehicle behavior prediction model”. As a prediction model for predicting the traveling direction and route of a vehicle, for example, there are those described in Non-Patent Document 1 or Non-Patent Document 2. The prediction of the vehicle behavior is expressed as y = f (x). However, y is a predicted value, x is a situation until now, and f is a predicted model. The future vehicle behavior search is simply a task of “listing vehicles satisfying a certain condition”, but f (x) is calculated for all vehicles, and the result is used as a condition. Enumerating matching vehicles increases the computational load, causing problems such as delay or increased computational resources. This problem is because a normal prediction model focuses on “forward” prediction, and the use of “reverse” prediction that the prediction result satisfies a certain condition is not assumed. While ordinary database technology speeds up the search by building an index, “Future Vehicle Behavior Search” is an extremely dynamic and uncertain subject, and it keeps changing, so the index It is considered difficult to configure at present.

  In order to avoid the above problem of increase in computational resources, a distributed process on the vehicle side as shown in Patent Document 3 can be considered. However, in the method disclosed in Patent Document 3, calculation (prediction) is started after the vehicle receives a request for processing, so that a delay occurs. Further, the method of Patent Document 3 is based on the concept of preferentially requesting distributed processing for a vehicle having an idle resource. However, when considering the prediction of vehicle behavior, the vehicle is requested to be predicted by another unrelated vehicle. This means that the history of the destination and route of the vehicle is transmitted to other unrelated vehicles, and privacy problems may occur. In order to solve these problems, an approach different from that of Patent Document 3 is required.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a vehicle search system that can suppress delay in vehicle search.

  In order to achieve the above object, a vehicle search system according to a first aspect of the present invention is a vehicle search system including a plurality of in-vehicle devices provided in a plurality of vehicles, and an information processing device, Vehicle information transmission means for transmitting vehicle information related to the vehicle to the information processing apparatus, and vehicle behavior prediction means for repeatedly predicting the time series of the behavior of the vehicle until after a predetermined prediction time based on the vehicle information. And whether or not the search condition is satisfied based on the time series of the behavior of the vehicle predicted by the vehicle behavior prediction means when the search condition regarding the position and time transmitted from the information processing apparatus is received. Search information generating means for generating information representing the above and information representing whether or not the search condition generated by the search information generating means is satisfied is sent to the information processing apparatus. Search information transmission means, and the information processing device is configured to detect each of the plurality of vehicles based on the input search condition and the vehicle information transmitted from each of the plurality of in-vehicle devices. A search condition for transmitting the search condition to each of the vehicle selection means for selecting the vehicle candidate satisfying the search condition and the in-vehicle device of the vehicle selected as the vehicle candidate by the vehicle selection means. Transmission means and vehicle search means for searching for a vehicle that satisfies the search condition based on information indicating whether or not the search condition is transmitted from each of the in-vehicle devices.

  According to the in-vehicle device of the first invention, the vehicle information transmission means transmits vehicle information related to the vehicle to the information processing device. Then, the vehicle behavior prediction means repeatedly predicts the time series of the vehicle behavior until after a predetermined prediction time based on the vehicle information.

  Then, when the search information generation unit receives the search condition regarding the position and time transmitted from the information processing device, the search condition is satisfied based on the time series of the vehicle behavior predicted by the vehicle behavior prediction unit. Whether or not is generated. Then, the search information transmission means transmits information indicating whether or not the search condition generated by the search information generation means is satisfied to the information processing apparatus.

  Further, according to the information processing apparatus of the first invention, from each of the plurality of vehicles based on the search condition input by the vehicle selecting means and the vehicle information transmitted from each of the plurality of in-vehicle devices. Select candidate vehicles that meet the search criteria.

  Then, the search condition transmission means transmits the search condition to each of the in-vehicle devices of the vehicle selected as the vehicle candidate by the vehicle selection means. Then, the vehicle search means searches for a vehicle that satisfies the search condition based on information indicating whether or not the search condition transmitted from each of the in-vehicle devices is satisfied.

  As described above, the in-vehicle device repeatedly predicts the time series of the behavior of the vehicle based on the vehicle information until a predetermined prediction time and receives the search condition transmitted from the information processing device. Information indicating whether or not the search condition is satisfied is generated based on the time series of the behavior of the vehicle, and the information indicating whether or not the generated search condition is satisfied is transmitted to the information processing apparatus. Based on the input search condition and vehicle information transmitted from each of the plurality of in-vehicle devices, a vehicle candidate satisfying the search condition is selected from each of the plurality of vehicles, and the vehicle selected as the vehicle candidate The search condition is transmitted to each of the vehicle-mounted devices, and the vehicle search is suppressed by searching the vehicle based on the information indicating whether the search condition transmitted from each of the vehicle-mounted devices is satisfied. be able to.

  A vehicle search system according to a second invention is a vehicle search system including a plurality of in-vehicle devices provided in a plurality of vehicles and an information processing device, wherein the in-vehicle device stores vehicle information related to the vehicles. Vehicle information transmission means for transmitting to the information processing apparatus, vehicle behavior prediction means for repeatedly predicting a time series of the behavior of the vehicle until after a predetermined prediction time based on the vehicle information, and the information processing Vehicle behavior information transmission means for transmitting, to the information processing apparatus, a time series of the behavior of the vehicle predicted by the vehicle behavior prediction means when a prediction result transmission request transmitted from the apparatus is received. The information processing device is configured to input each of the plurality of vehicles based on the input search condition related to the position and time and the vehicle information transmitted from each of the plurality of in-vehicle devices. To the vehicle selection means for selecting the vehicle candidate satisfying the search condition, and the prediction result transmission request to each of the in-vehicle devices of the vehicle selected as the vehicle candidate by the vehicle selection means. Generating information indicating whether or not the search condition is satisfied for each of the in-vehicle devices, based on the prediction result transmission request transmitting means and the time series of the behavior of the vehicle transmitted from each of the in-vehicle devices. Search information generation means, and vehicle search means for searching for a vehicle that satisfies the search condition based on information indicating whether or not the search condition is generated by the search information generation means.

  According to the in-vehicle device of the second invention, the vehicle information transmission means transmits the vehicle information related to the vehicle to the information processing device. Then, the vehicle behavior prediction means repeatedly predicts the time series of the vehicle behavior until after a predetermined prediction time based on the vehicle information.

  And when the prediction result transmission request | requirement transmitted from the information processing apparatus is received by the vehicle behavior information transmission means, the time series of the vehicle behavior predicted by the vehicle behavior prediction means is transmitted to the information processing apparatus.

  Further, according to the information processing apparatus of the second invention, the plurality of vehicles are based on the search condition input by the vehicle selection means and the vehicle information transmitted from each of the plurality of in-vehicle devices. A vehicle candidate satisfying the search condition is selected from each of the above. And a prediction result transmission request | requirement transmission means transmits a prediction result transmission request | requirement with respect to each vehicle-mounted apparatus of the vehicle selected as a vehicle candidate by the vehicle selection means.

  Then, the search information generating means generates information indicating whether or not the search condition is satisfied for each of the in-vehicle devices based on the time series of the behavior of the vehicle transmitted from each of the in-vehicle devices. Then, the vehicle search means searches for a vehicle that satisfies the search condition based on information indicating whether the search condition generated by the search information generation means is satisfied.

  As described above, when the in-vehicle device repeatedly predicts the time series of the behavior of the vehicle based on the vehicle information until a predetermined prediction time and receives the prediction result transmission request transmitted from the information processing device. The time series of the predicted vehicle behavior is transmitted to the information processing device, and the information processing device is configured to output the plurality of vehicles based on the input search condition and the vehicle information transmitted from each of the plurality of in-vehicle devices. A candidate vehicle satisfying the search condition is selected from each, a prediction result transmission request is transmitted to each in-vehicle device of the vehicle selected as the vehicle candidate, and the behavior of the vehicle transmitted from each in-vehicle device is transmitted. Delay in vehicle search by calculating information indicating whether or not the search condition is satisfied for each of the in-vehicle devices based on the time series, and searching for the vehicle based on the information indicating whether or not the search condition is satisfied Suppress It is possible.

The search condition according to the present invention is a search condition including a travel position of a vehicle and a time section,
The vehicle behavior prediction means repeatedly predicts a time series of the traveling position of the vehicle based on the vehicle information until after a predetermined prediction time as an upper limit of a time interval included in the search condition, and the search information The generation unit is configured to determine the travel position included in the search condition within the time interval included in the search condition based on the time series of the travel position of the vehicle predicted by the vehicle behavior prediction unit. Information indicating whether or not to travel can be calculated.

  The information indicating whether or not the search condition according to the present invention is satisfied may be a certainty factor indicating the degree of satisfying the search condition.

  According to the present invention, for each of the searched vehicles, the information processing apparatus is a task corresponding to information indicating whether the search condition is satisfied, and indicates instruction information for instructing a task to be executed by the vehicle And a task transmission unit that transmits the instruction information generated by the task generation unit to the in-vehicle device of the vehicle, the in-vehicle device from the information processing device Task processing means for executing the task based on the transmitted instruction information may be further provided.

  According to the present invention, for each of the searched vehicles, the task generation means instructs the presence / absence of information or the type of presentation information according to information indicating whether or not the search condition is satisfied. The information may be generated, and the task processing unit may execute a task of presenting information to an occupant of the vehicle based on the instruction information transmitted from the information processing apparatus.

  As described above, according to the first invention, the in-vehicle device repeatedly predicts the time series of the behavior of the vehicle based on the vehicle information until after a predetermined prediction time, and is transmitted from the information processing device. Information indicating whether or not the search condition is satisfied is generated based on the predicted time series of vehicle behavior and information indicating whether or not the generated search condition is satisfied is received. The information processing device selects a candidate vehicle satisfying the search condition from each of the plurality of vehicles based on the input search condition and the vehicle information transmitted from each of the plurality of in-vehicle devices. Then, the search condition is transmitted to each on-vehicle device of the vehicle selected as the vehicle candidate, and the vehicle is searched based on the information indicating whether the search condition transmitted from each of the on-vehicle devices is satisfied. Of vehicle search It is possible to suppress the extension, the effect is obtained that.

  According to the second invention, the in-vehicle device repeatedly predicts the time series of the behavior of the vehicle based on the vehicle information until after a predetermined prediction time, and transmits the prediction result transmitted from the information processing device. When the request is received, the time series of the predicted vehicle behavior is transmitted to the information processing device, and the information processing device is based on the input search condition and the vehicle information transmitted from each of the plurality of in-vehicle devices. Then, a vehicle candidate satisfying the search condition is selected from each of the plurality of vehicles, a prediction result transmission request is transmitted to each of the in-vehicle devices of the vehicle selected as the vehicle candidate, and transmitted from each of the in-vehicle devices. Calculating information indicating whether or not the search condition is satisfied for each of the in-vehicle devices based on the time series of the behavior of the vehicle, and searching for the vehicle based on the information indicating whether or not the search condition is satisfied By vehicle search It is possible to suppress the delay, the effect is obtained that.

It is a block diagram which shows the structure of the vehicle search system which concerns on the 1st Embodiment of this invention. It is explanatory drawing for demonstrating route prediction of a vehicle. It is explanatory drawing for demonstrating discretizing map data with a mesh. 3 is a conceptual diagram for explaining a hash table H. FIG. 4 is a conceptual diagram for explaining the update of a hash table H. FIG. 4 is a conceptual diagram for explaining an auxiliary hash table G. FIG. 6 is a conceptual diagram for explaining setting of a region Q. FIG. It is explanatory drawing for demonstrating the operation example of the server of a vehicle search system. It is a flowchart which shows the content of the prediction process routine in the vehicle-mounted apparatus which concerns on embodiment of this invention. It is a flowchart which shows the content of the vehicle behavior prediction process routine in the vehicle-mounted apparatus which concerns on embodiment of this invention. It is a flowchart which shows the content of the data storage process routine in the server which concerns on embodiment of this invention. It is a flowchart which shows the content of the search process routine in the server which concerns on embodiment of this invention. It is a flowchart which shows the content of the simple prediction process routine in the server which concerns on embodiment of this invention. It is a flowchart which shows the content of the inference process routine in the vehicle-mounted apparatus which concerns on embodiment of this invention. It is a flowchart which shows the content of the certainty calculation process routine in the vehicle-mounted apparatus which concerns on embodiment of this invention. It is a block diagram which shows the structure of the vehicle search system which concerns on the 2nd Embodiment of this invention.

  Embodiments of the present invention will be described below with reference to the drawings.

[First Embodiment]
<Configuration of vehicle search system 210 according to the first embodiment>
As illustrated in FIG. 1, the vehicle search system 10 according to the first exemplary embodiment of the present invention includes a plurality of in-vehicle devices 12 provided in a plurality of vehicles and vehicle information transmitted from the plurality of in-vehicle devices 12. And a server 14 that searches for a vehicle that receives and satisfies the search condition. The plurality of in-vehicle devices 12 and the server 14 are connected via a network 16 such as the Internet. The server 14 is an example of an information processing device.

In the present embodiment, the user asks the vehicle search system 10 a “question (query) about future vehicle behavior” such as “a vehicle likely to pass the road segment A by T _q minutes and its certainty”. " And the vehicle equipment 12 and the server 14 of the vehicle search system 10 output the search result with respect to a query. In the following, an example will be described in which “a vehicle that is likely to pass through road segment A by T _q minutes”, which conditions the behavior of the vehicle, is used as a search condition for the position and time interval.

[In-vehicle device 12]
The in-vehicle device 12 includes a position measuring unit 120 that sequentially measures position information of the own vehicle, a vehicle sensor 124 that sequentially detects the speed of the own vehicle, position information of the own vehicle measured by the position measuring unit 120, and a vehicle sensor. And a computer 126 that predicts the behavior of the host vehicle based on the speed of the host vehicle detected by 124. The in-vehicle device 12 is mounted on a vehicle.

  The position measurement unit 120 is configured using, for example, a GPS sensor, and measures position information (X, Y) of the current time t of the host vehicle. The position information includes latitude X and longitude Y at which the vehicle is located at the current time t.

  The computer 126 includes a CPU, a RAM, and a ROM that stores a program for executing each processing routine described later, and is functionally configured as follows. The computer 126 includes an information acquisition unit 128 that acquires position information (X, Y) of the current time t of the host vehicle measured by the position measurement unit 120 and the speed of the host vehicle detected by the vehicle sensor 124, and information The vehicle information acquired by the acquisition unit 128 is sequentially transmitted to the server 14 together with the vehicle ID, and the communication unit 130 that transmits / receives data to / from the server 14, and the position information of the host vehicle acquired by the information acquisition unit 128 Based on the speed of the host vehicle and the vehicle behavior predicting unit 132 that predicts the predicted route of the host vehicle, and when the search condition transmitted from the server 14 is received, information indicating whether the search condition is satisfied is generated. And a task processing unit 136 that executes the task when the task instruction information transmitted from the server 14 is received. The communication unit 130 is an example of vehicle information transmission means and search information transmission means. The inference unit 134 is an example of a search information generation unit.

  The information acquisition unit 128 acquires the position information (X, Y) of the current time t of the host vehicle measured by the position measurement unit 120 as vehicle information related to the vehicle. The information acquisition unit 128 acquires the speed of the host vehicle at the current time t detected by the vehicle sensor 124.

  The communication unit 130 sequentially transmits the position information (X, Y) of the current vehicle t acquired by the information acquisition unit 128 to the server 14. In addition, the communication unit 130 receives the search condition regarding the position and time interval transmitted from the server 14. Further, when the communication unit 130 receives the search condition transmitted from the server 14, the communication unit 130 transmits information indicating whether the search condition generated by the inference unit 134 described later is satisfied to the server 14. Further, the communication unit 130 receives the instruction information transmitted from the server 14 and instructing a task to be executed by the vehicle.

The vehicle behavior predicting unit 132 determines that the host vehicle has reached a predetermined prediction limit based on the position information (X, Y) of the host vehicle's current time t acquired by the information acquiring unit 128 and the speed of the host vehicle. The prediction route of is repeatedly predicted. Note that the predicted route of the host vehicle is an example of a time series of the behavior of the host vehicle.
Even if the vehicle behavior prediction unit 132 does not receive the search condition from the server 14, the vehicle behavior prediction unit 132 sequentially updates the prediction every few milliseconds to several seconds to several minutes, for example. By sequentially updating the prediction of the behavior of the vehicle, it is possible to respond faster than in the case where the prediction is performed after receiving the search condition, and the problem of delay can be alleviated. In addition, it is possible to smooth the load by sequentially predicting the route of the vehicle and scheduling so as to be performed when there is a margin in calculation resources.

  In the present embodiment, an example will be described in which a route on which the host vehicle travels is predicted after T minutes (for example, about 1 to 10 minutes) (hereinafter, T is referred to as a prediction limit). Note that the prediction limit T is predetermined and corresponds to the upper limit of the time interval included in the search condition.

FIG. 2 is a diagram for explaining a route prediction model for predicting the route of the host vehicle.
As shown in FIG. 2, at time t, the history of the route traveled by the host vehicle (list of road segment IDs) was r _t−1 , r _t−2 ,..., R _t−N. To do. At this time, the road segment at the next time can be predicted based on the conditional probability p (r _t | r _t−1 , r _t−2 ,..., R _t−N ) estimated in advance. The conditional probability can be obtained by statistically learning from an individual traveling history or a traveling history of all vehicles. The conditional probability is a road segment that is connected to the road segment A when the host vehicle is currently traveling on the road segment A, and any of the next road segments on which the host vehicle travels has a higher probability. This is a model of “a straight line has a higher probability than a right turn”, “a second consecutive right turn is less likely to occur”, and the like.
In addition, when a destination is input to the car navigation system, it may be considered that the possibility of using the route presented by the car navigation system is high.

In this embodiment, a case where a path is predicted by sampling will be described as an example. That is, the vehicle behavior predicting unit 132 sets one road segment to travel next with a probability proportional to the conditional probability p (r _t | r _t−1 , r _t−2 ,..., R _t−N ). Sampling. _R = the history of the path _{{r t-1, r t} -2, ···, r t-N} and represents an operation for obtaining a _{r t} is sampled at the time of history _{R r t} ← S (R) It expresses. Further, if the average required travel time of the r _t to be expressed as T (r _t), it can be predicted with the procedure as shown in the following pseudocode. Here, the average required travel time T (r _t) of r _t is calculated by including the speed limit of the history or road segment speed of the vehicle.

[Pseudo code]
(Algorithm / path sampling)
Input: route history R = {r _t−1 , r _t−2 ,..., R _t−N }, prediction limit T
Output: predicted route _{R '= {r t + K} , r t + K-1, ···, r t}

R ′ ← {} // empty array T ₀ ← 0 // complete when prediction limit T is exceeded T ₀ <T
r ← S (R'∪R) // Sampling from the travel history so far R '← R'∪ {r} // Add to predicted route T ₀ ← T ₀ + T (r) // Required travel time T End while adding (r)
return R '
[End of pseudo code]

  The above algorithm samples one prediction path up to the prediction limit T minutes. By executing this algorithm a sufficient number of times, a plurality of predicted paths are obtained. The list of predicted paths can be regarded as a probabilistic path prediction up to approximately T minutes later. In the present embodiment, a case where the prediction algorithm is executed J times and J prediction paths are obtained will be described as an example.

  When the inference unit 134 receives the search condition transmitted from the server 14, the inference unit 134 searches as information indicating whether the search condition is satisfied based on the predicted route of the host vehicle predicted by the vehicle behavior prediction unit 132. A certainty factor representing the degree of satisfying the condition is calculated. The search condition includes the travel position of the vehicle and the time interval.

Specifically, the inference unit 134 converts the predicted route of the host vehicle into a format required by the server 14. The format after conversion depends on the search conditions of the server 14, but in the following, on the server 14 side, “list vehicles that have passed the road segment A with a certainty P or more within T _q minutes” (however, T _q < A case where the query is T; T is a vehicle prediction limit) will be described.

By processing the list of predicted routes predicted by the vehicle behavior predicting unit 132 using the following algorithm, “the certainty of passing the road segment A within T _q minutes” is calculated.

[Pseudo code]
[Calculation of road segment passage certainty in algorithm / inference unit 134]
Input: List of predicted routes L = {R ₁ , R ₂ ,..., R _K }, road segment A, time T _q
Output: P of certainty that the vehicle will pass road segment A within T _q minutes
P <−0 // Confidentiality K <−length (L) // Number of predicted paths Foreach (R in L): // Included in the predicted path list L
Loop T _c <−0 // T _c > T _q for all routes R
Consider A not included Foreach (r in R): // Included in route R
Loop for road segment r
if (r == A): // When route A includes A
P <-P + 1 / K // Increase confidence
break this this inner loop
// Because processing of route R is over
Consider next route R '
end if
T _c <−T _c + T (r) // Time is the required time T (r) for r
Only increase
if (T _c > T _q ): // End when T _q is exceeded
break this this inner loop
end if
end inner for loop
end outer for loop
return P
[End of pseudo code]

The calculation of the certainty factor depends on the road segment A and the time _Tq and is unknown until the search condition is received, and therefore cannot be calculated in advance.

In addition, about the calculation of the said certainty factor, although the query demonstrated as an example the case where a query is " _List the vehicle which passes the road segment A more than the certainty factor P within _Tq minutes", also about another type of query, too Basically, the certainty factor can be calculated with an algorithm based on the same idea as described above. Further, since the vehicle behavior prediction process is common to all types of queries, it is not necessary to perform another type of vehicle behavior prediction process in order to increase the types of queries. This is because the sequential vehicle behavior prediction processing and on-demand inference processing (confidence calculation processing) are separated. Further, when a new type of query is executed, a new inference processing program may be transmitted to the vehicle side and executed.

  In addition, since there is no idle resource on the in-vehicle device 12 side of the vehicle, there may be a case where the prediction of the route of the vehicle that should have been sequentially updated on the vehicle side has become old. This is notified to the server 14 side, and the server 14 side receives it and determines whether or not to predict and infer the route on the server 14 side in consideration of the calculation resources and the like.

  The task processing unit 136 executes the task based on the task instruction information transmitted from the server 14. In the present embodiment, a case will be described as an example where task processing unit 136 executes a task of presenting information to a vehicle occupant based on instruction information transmitted from server 14. For example, the task processing unit 136 displays an advertisement on the display unit of the car navigation system when stopped by a signal.

[Server 14]
The server 14 is composed of a server including a CPU, a ROM that stores a program for realizing each processing routine described later, a RAM that temporarily stores data, a memory as a storage unit, a network interface, and the like. Functionally, it can be expressed by a configuration including a reception unit 140, a communication unit 142, a data storage unit 144, a database 145, a simple prediction unit 146, a vehicle search unit 148, and a task generation unit 150. The communication unit 142 is an example of a search condition transmission unit and a task transmission unit, and the simple prediction unit 146 is an example of a vehicle selection unit.

The accepting unit 140 accepts a query input to the server 14. The query is, for example, “a question regarding future vehicle behavior” such as “a vehicle likely to pass the road segment A by T _q minutes and its certainty”. As described above, the search condition included in the query includes the travel position of the vehicle and the time interval. In the above description, the inquiry “vehicles likely to pass through road segment A by T _q minutes” has been considered. However, the search conditions in the present embodiment are as follows: Can think.

(1) The road segment A passes within ₁ minute T, then the road segment B vehicles (2) which passes within ₂ minutes T passing through the road segment A within ₁ minute T, then the road segment B T ₂ vehicle turning right road segment a non vehicle (3) the vehicle (4) to pass through the road segment a or road segment B within _one minute T T within _one minute passes within a minute, and the like.

  The communication unit 142 receives vehicle position information (X, Y) transmitted from each of the plurality of in-vehicle devices 12. In addition, the communication unit 142 transmits a search condition included in the query received by the receiving unit 140 to each of the in-vehicle devices 12 of the vehicle selected as a vehicle candidate by the simple prediction unit 146 described later. Further, the communication unit 142 receives the certainty factor transmitted from each of the in-vehicle devices 12 that has received the search condition. Moreover, the communication part 142 transmits the instruction information produced | generated by the task production | generation part 150 mentioned later with respect to the vehicle-mounted apparatus 12 of a vehicle.

  The data storage unit 144 stores the combination of the vehicle position information (X, Y) received from the communication unit 142 from each of the plurality of in-vehicle devices 12 and the vehicle ID of the vehicle in the database 145.

  Specifically, the data storage unit 144 acquires (vehicle ID, X (latitude), Y (longitude)) received by the communication unit 142. In the present embodiment, as shown in FIG. 3, the latitude and longitude on the map are discretized into meshes, and a mesh ID is assigned to each of the discretized meshes. In the following, the vehicle ID is represented by V.

The data storage unit 144 stores (vehicle ID, X (latitude), Y (longitude)) received by the communication unit 142 in a hash table H (data storage location) using the mesh ID as a key. If the mesh ID is m, all vehicle data included in the mesh m is stored in H [m] as a “hash table using the vehicle ID as a key”. That is, the structure of the hash table H is realized as a double hash table as shown in FIG. Incidentally, _{m n} represents a mesh ID, _{V n} represents the vehicle _ID, represents a _(X n, _{Y n)} latitude and longitude.

Also, because of the restriction that only the latest data received by the communication unit 142 is stored in H, when storing the vehicle with the vehicle ID (V), as shown in FIG. It is necessary to erase from the hash table H. Therefore, when a vehicle ID is given, it is necessary to specify which mesh ID stores the data. Therefore, as shown in FIG. 6, an auxiliary hash table G that enables retrieval of vehicle ID → mesh ID is prepared. In the hash table of FIG. 4 above, it can be seen that V ₁ belongs to m ₁ , but because of the nature of the hash table, the reverse lookup (in the direction opposite to the arrow) is extremely inefficient. Using table G, it is possible to efficiently handle the vehicle ID to the mesh ID. By using the auxiliary hash table G, as shown in FIG. 5 above, it can be seen that V ₂ belongs to m ₁ , so the data of V ₂ can be erased from H [m ₁ ].

  In summary, the data storage process is shown in the following pseudo code.

[Pseudo code]
[Algorithm / Data storage]
Input: (Vehicle ID, Latitude, Longitude) (hereinafter referred to as (V, X, Y))
oldMeshID <-G [v] // Vehicle ID (V) data is current
Remove the data of the specific vehicle V from H [oldMesh] to find out which mesh is included // Figure 5
newMeshID <-Discretize (X, Y) // Calculate mesh ID
(Figure 3)
H [newMeshID] [V] <-(X, Y) // Mesh newMeshID
A new vehicle IDV
G [V] <-newMeshID // Update auxiliary hash table [end of pseudo code]

  The database 145 stores (vehicle ID, X (latitude), Y (longitude)) for each vehicle. Further, the database 145 stores each data by the hash table H and the auxiliary hash table G.

  The simple prediction unit 146 performs a search from each of the plurality of vehicles based on the search conditions included in the query received by the reception unit 140 and the position information (X, Y) of the plurality of vehicles stored in the database 145. Select candidate vehicles that meet the conditions.

  It is a waste of vehicle resources to issue a command to the in-vehicle devices 12 of all the vehicles to "calculate the certainty level satisfying the search condition". Even if a strict prediction is not used, a more efficient calculation can be performed by using a simple prediction considering physical constraints. For example, when searching for “a vehicle whose vehicle position after 5 minutes is in the vicinity of the point A”, assuming that the vehicle is 120 km / h or less per hour, only for vehicles that currently exist within 10 km from the point A It is sufficient to calculate confidence.

  Therefore, the requirement required for the simple prediction unit 146 is that a vehicle that may match the search condition can be identified at high speed using the information on the position information (X, Y) of a plurality of vehicles stored in the database 145 ( It can be extracted in real time.

  Specifically, first, the simple prediction unit 146 uses the hash table H of the database 145 to determine the inside of a certain geographical area Q (in the above-described example of “a vehicle existing within 10 km from the road segment A” Region Q is a 10 km circle centered on road segment A). FIG. 7 is a diagram for explaining the relationship between the region Q and the mesh. In FIG. 7, a fine shaded portion represents a region Q, and a coarse shaded portion represents a mesh set including the region Q. As shown in FIG. 7, when a region Q is given, it is desired to obtain a mesh set (as small as possible) that includes the region Q. This means that one point in the region Q is appropriately selected, converted into a mesh ID, a mesh around the mesh is searched, and if there is a common part with the region Q, it is added as a mesh set. It can be obtained by repeating. When searching for the area Q, it is sufficient to search for the area Q and the mesh set M corresponding to the rough shaded portion. However, in the present embodiment, in order to search for vehicles in the area Q without excess or deficiency. In addition, the vehicle existing at a position not included in the region Q is removed from the rough shaded portion.

  Once the mesh set M corresponding to the region Q is obtained, H [m] is enumerated for all the element meshes m, and whether each vehicle included in the H [m] is really included in the region Q. By checking, the purpose of listing up vehicle IDs traveling inside the area Q is achieved.

  In summary, the database search is shown in the following pseudo code.

[Pseudo code]
[Algorithm / Simple prediction]
Input: Area Q
Output: List W of vehicles traveling in area Q
W <-{}
Acquisition of mesh set including M <-Q Foreach (min in M):
Foreach (vin in H [m]):
(X, Y) <-H [m] [V]
if ((X, Y) is included in Q):
W <-W∪ {V}
end if
end inner for loop
End outer for loop
[End of pseudo code]

The vehicle search unit 148 searches for a vehicle that satisfies the search condition based on the certainty factor transmitted from each of the in-vehicle devices 12. Specifically, the vehicle search unit 148 searches for a vehicle that satisfies the search condition based on information representing a combination of the vehicle ID and the certainty factor received by the communication unit 142. For example, when the search condition is “a vehicle that is likely to pass the road segment A after T _q minutes”, the vehicle search unit 148 sets a vehicle having a certainty factor of 50% or more as a vehicle that satisfies the search condition. In addition, when the search query is, for example, “ _List vehicles that pass through road segment A with a certainty P or more within T _q minutes”, the vehicle search unit 148 selects vehicles having a certainty P or more. A vehicle that satisfies the search conditions.

For the processing of the communication unit 142, the simple prediction unit 146, and the vehicle search unit 148, a query “List vehicles that are likely to pass the road segment A by T _q minutes and their certainty” as shown in FIG. When the server 14 is input to the server 14, a specific procedure will be described as to how the server 14 returns a result.

(1) First, an area in which a vehicle that can pass through the road segment A by T _q minutes later can be determined by considering the physical speed of the vehicle. Let this be region Q. By performing the area query process in the simple prediction unit 146 with the area Q as an input, the vehicles traveling in the current area Q are listed.
The communication unit 142 transmits a query “notify the certainty of passing through the road segment A by T _q minutes” to the vehicle-mounted device 12 for all the vehicles traveling on the listed Q.

(2) On the server 14 side, if the calculation results of the certainty factor of all vehicles are returned, this is the result, but even if the results are not complete within a predetermined time, only the returned ones are used. And In some cases, it is possible to obtain a list of vehicles having a high degree of certainty by removing those having a certainty degree or less from the result list.

  Based on the information representing the combination of the vehicle ID and the certainty level received by the communication unit 142, the task generation unit 150 generates a task corresponding to the certainty level and indicating the task to be executed by the vehicle. To do. In the present embodiment, a case will be described as an example where task generation unit 150 generates instruction information that indicates the presence or absence of information presentation or the type of presentation information in accordance with the certainty factor.

  In FIG. 8 (3), an advertisement is given as an example of a specific task. By notifying vehicles that are likely to pass a certain road segment, such as sales information of stores along the street, it becomes possible to make more targeted and appropriate advertisements. In this example, the presence / absence or type of an advertisement is determined according to the certainty factor P. This strategy may be performed according to a predefined database.

<Operation of vehicle search system>
Next, the operation of the vehicle search system 10 according to the embodiment of the present invention will be described. When each vehicle equipped with the in-vehicle device 12 is traveling, the in-vehicle device 12 of each vehicle executes a prediction processing routine shown in FIG. 9, and the server 14 executes a search processing routine shown in FIG.

  First, the prediction process routine which the vehicle-mounted apparatus 12 of each vehicle performs is demonstrated. The in-vehicle device 12 is configured such that when the position information (X, Y) of the current time t of the own vehicle is measured by the position measuring unit 120 and the speed of the own vehicle at the current time t is detected by the vehicle sensor 124, FIG. The prediction processing routine shown in FIG.

<Prediction processing routine>
First, in step S100, the information acquisition unit 128 acquires position information (X, Y) of the current time t of the host vehicle measured by the position measurement unit 120 as vehicle information. The information acquisition unit 128 acquires the speed of the host vehicle at the current time t detected by the vehicle sensor 124.

  In step S <b> 102, the communication unit 130 transmits the position information (X, Y) of the current vehicle t acquired in step S <b> 100 and the vehicle ID to the server 14.

  In step S104, the vehicle behavior prediction unit 132 performs a predetermined prediction based on the position information (X, Y) of the current time t of the host vehicle acquired in step S100 and the speed of the host vehicle. The predicted route of the host vehicle is predicted until after time. Step S104 is realized by the vehicle behavior prediction processing routine shown in FIG.

<Vehicle behavior prediction processing routine>
In step S200, the history R of the route traveled by the host vehicle is acquired based on the time series of the position information (X, Y) of the host vehicle acquired in step S100.

  In step S201, 1 is substituted into a variable j related to the number of predicted paths, and initialization is performed.

In step S202, the predicted route R ′ and the time variable T ₀ are initialized.

In step S204, it determines whether the time variable _{T 0} is less than the predicted limit T. If the time variable T ₀ is smaller than the prediction limit T, the process proceeds to step S206. On the other hand, if the time variable T ₀ is greater than or equal to the prediction limit T, the process proceeds to step S211.

In step S206, the route history R acquired in step S200, the route R ′ predicted in step S206 up to the previous time, and the conditional probability p (r _t | r _t−1 , r estimated in advance). _{based on (t-2} ,..., rt _-N )), the route r predicted that the host vehicle will travel at the next time is sampled.

  In step S208, the route r sampled in step S206 is added to the predicted route R ′.

In step S210, the added r average required travel time T a (r) in the above step S208, is added to the time variable _{T 0.}

  In step S211, the predicted route R 'obtained in steps S204 to S210 is stored in a memory (not shown).

  In step S212, it is determined whether or not the predetermined number J of predicted paths is equal to the variable j corresponding to the number of predicted paths R 'stored in the memory in step S211. If the number J of predicted routes is equal to the variable j, the process proceeds to step S216. On the other hand, if the number J of predicted routes is different from the variable j, the variable j is incremented in step S214, and the process returns to step S202.

  In step S216, the plurality of predicted routes stored in the memory in step S211 are output as a result, and the vehicle behavior prediction processing routine is terminated.

Next, a search processing routine executed by the server 14 will be described. When the server 14 receives the vehicle ID and the vehicle position information (X, Y) transmitted from at least one in-vehicle device 12, the server 14 executes a data storage processing routine shown in FIG.
When the server 14 receives a query input from the user, the server 14 executes a search processing routine shown in FIG.

<Data storage processing routine>
In step S300, the received vehicle ID and position information (X, Y) of the vehicle are received.

  In step S302, using the auxiliary hash table G, the mesh ID to which the vehicle ID currently belongs is identified from the vehicle ID received in step S300.

  In step S304, the data corresponding to the vehicle ID received in step S300 is removed from the hash table corresponding to the mesh ID specified in step S302.

  In step S306, a new mesh ID is calculated based on the vehicle position information (X, Y) acquired in step S300.

  In step S308, the data corresponding to the vehicle ID received in step S300 is stored in the hash table corresponding to the mesh ID calculated in step S306.

  In step S310, the auxiliary hash table G is updated based on the new mesh ID calculated in step S306, and the data storage processing routine ends.

<Search processing routine>
First, in step S <b> 400, the reception unit 140 receives a query input to the server 14.

  Next, in step S402, the simple prediction unit 146 includes the search conditions included in the query received in step S400 and the positional information (X, Y) of the plurality of vehicles stored in the database 145 in the data storage processing routine. Based on the above, a vehicle candidate that satisfies the search condition is selected from each of the plurality of vehicles. Step S402 is realized by the simple prediction processing routine shown in FIG.

<Simple prediction processing routine>
In step S500, an area Q in a range that is centered on the position (road segment) included in the search condition and that is predetermined according to the time interval included in the search condition is received.

  In step S502, the vehicle list W is initialized.

  In step S504, a mesh set including the region Q received in step S500 is acquired.

  In step S506, one mesh m is set from the mesh set acquired in step S504.

  In step S508, one vehicle ID is set out of each of the vehicle IDs included in the mesh m set in step S506 based on each of the position information (X, Y) stored in the database 145.

  In step S510, vehicle position information (X, Y) corresponding to the vehicle ID set in step S508 is acquired from each of the position information (X, Y) stored in the database 145.

  In step S512, based on the vehicle position information (X, Y) acquired in step S510, whether the latitude X and longitude Y where the vehicle is located are included in the region Q acquired in step S500. Determine. When the latitude X and longitude Y where the vehicle is located are included in the region Q, the process proceeds to step S514. On the other hand, when the latitude X and longitude Y where the vehicle is located are not included in the region Q acquired in step S500, the process proceeds to step S516.

  In step S514, the vehicle ID set in step S508 is stored in the vehicle list W.

  In step S516, it is determined whether or not the processes in steps S508 to S514 have been executed for all vehicles belonging to the mesh m set in step S506. If the processes of steps S508 to S514 have been executed for all the vehicles belonging to the mesh m set in step S506, the process proceeds to step S518. On the other hand, for vehicles belonging to the mesh m set in step S506, if there is a vehicle that has not executed the processes in steps S508 to S514, the process returns to step S508.

  In step S518, it is determined whether or not the processes in steps S506 to S516 have been executed for all the meshes included in the mesh set acquired in step S504. If the processes in steps S506 to S516 have been executed for all the meshes included in the mesh set acquired in step S504, the process proceeds to step S520. On the other hand, for the meshes included in the mesh set acquired in step S504, if there is a mesh that has not been subjected to the processing in steps S506 to S516, the process returns to step S506.

  In step S520, each vehicle ID stored in the vehicle list W in step S514 is output as a candidate vehicle satisfying the search condition, and the simple prediction processing routine is terminated.

  Next, returning to the search processing routine, in step S404, the search condition included in the query accepted in step S400 is transmitted to each on-vehicle device 12 of the vehicle selected as the vehicle candidate in step S402.

  Here, when each of the in-vehicle devices 12 of the vehicle selected as a vehicle candidate receives the search condition transmitted from the server 14 in step S404 by the communication unit 130, the inference unit 134 of the in-vehicle device 12 displays FIG. The inference processing routine shown in FIG. Hereinafter, after describing the inference processing routine executed in the in-vehicle device 12, the processing after step S406 of the search processing routine executed by the server 14 will be described.

<Inference processing routine>
In step S600, the plurality of prediction paths output in step S104 of the prediction processing routine and the search conditions received by the communication unit 130 are acquired.

  In step S602, the inference unit 134 calculates a certainty factor based on the plurality of predicted routes and search conditions of the host vehicle acquired in step S600. Step S602 is realized by the certainty factor calculation processing routine shown in FIG.

<Confidence calculation processing routine>
In step S700, time _Tq and road segment A included in the search conditions acquired in step S600 are acquired.

  In step S702, the certainty factor P is initialized.

  In step S704, the number K of predicted routes acquired in step S600 is acquired.

  In step S706, one predicted route R is set among the plurality of predicted routes acquired in step S600.

In step S708, the time variable _Tc is initialized.

  In step S710, one road segment r is set among the road segments included in the predicted route R set in step S706.

  In step S712, it is determined whether or not the road segment r set in step S710 matches the road segment A acquired in step S700. If the road segment r matches the road segment A, the process proceeds to step S713. On the other hand, if the road segment r does not match the road segment A, the process proceeds to step S714.

  In step S713, the certainty factor P initialized in step S702 or the certainty factor P updated in the previous step S713 is increased according to the increase formula P ← P + 1 / K.

In step S714, the initialization time variable _{T c} or previous time variable _{T c} updated in the step S714 in the above step S 708, the average required travel time of a road segment r set in step S710 T Add (r _t ).

In step S716, it determines the time variable _{T c} updated in step S714 is, the greater or not than the time _{T q} obtained in step S700. When the time variable _{T c} is greater than the time _{T q} returns to step S706. On the other hand, if the time variable T _c is less than or equal to the time T _q , the process proceeds to step S718.

  In step S718, it is determined whether or not the processing in steps S710 to S716 has been executed for all road segments included in the predicted route R set in step S706. When the processes of steps S710 to S716 have been executed for all road segments included in the predicted route R set in step S706, the process proceeds to step S720. On the other hand, for road segments included in the predicted route R set in step S706, if there is a road segment for which the processes in steps S710 to S716 are not performed, the process returns to step S710.

  In step S720, it is determined whether or not the processes in steps S706 to S718 have been executed for all of the K predicted paths acquired in step S704. If the processes in steps S706 to S718 have been executed for all the K predicted paths acquired in step S704, the process proceeds to step S722. On the other hand, for the K predicted routes acquired in step S704, if there is a predicted route for which the processes in steps S706 to S718 are not performed, the process returns to step S706.

  In step S722, the certainty factor P updated in step S713 is output as a result, and the certainty factor calculation processing routine is terminated.

  Next, returning to the inference processing routine, in step S604, the certainty factor P output in step S602 is transmitted to the server 14 by the communication unit 130, and the inference processing routine is terminated.

  Next, processing after step S406 of the search processing routine executed by the server 14 will be described.

  In step S <b> 406, the communication unit 142 of the server 14 receives the certainty factor P transmitted from each of the in-vehicle devices 12.

  In step S408, it is determined whether or not the certainty factor P has been received from all the vehicles that have transmitted the search conditions in step S404. When the certainty factor P is received from all of the vehicles that have transmitted the search conditions, the process proceeds to step S410. On the other hand, if the certainty factor P has not been received from all of the vehicles that have transmitted the search conditions, the process returns to step S406. Even if the certainty factor P has not been received from all of the vehicles that have transmitted the search conditions, if a predetermined time has elapsed, the process proceeds to step S410.

  In step S410, the vehicle search unit 148 searches for a vehicle that satisfies the search condition included in the query, based on each of the certainty factors P received in step S406.

  In step S412, for each of the vehicles searched in step S410, instruction information indicating a task to be executed by the vehicle is generated according to the certainty factor transmitted from the vehicle.

  In step S414, each of the instruction information for instructing the task to be executed by the vehicle generated in step S412 is transmitted to each of the vehicles, and the search processing routine is ended.

  Each of the in-vehicle devices 12 that has received the instruction information performs a task by the task processing unit 136 based on the instruction information received by the communication unit 130.

  As described above, according to the vehicle search system of the first embodiment of the present invention, the in-vehicle device repeatedly predicts the predicted route of the vehicle until after a predetermined prediction time based on the vehicle information. When the search condition transmitted from the server is received, a certainty factor is generated based on the predicted route of the predicted vehicle, and the generated certainty factor is transmitted to the server. Based on the vehicle information transmitted from each of the plurality of in-vehicle devices, a vehicle candidate satisfying the search condition is selected from each of the plurality of vehicles, and for each of the in-vehicle devices of the vehicle selected as the vehicle candidate By transmitting the search condition and searching for the vehicle based on the certainty factor transmitted from each of the in-vehicle devices, the delay of the vehicle search can be suppressed.

  In addition, because future position prediction is performed sequentially, the result can be returned if only the inference process (confidence level calculation process) is performed at the stage of the inquiry from the server side. And a complicated search process with uncertainty of “search for future vehicle position” for a large number of candidate vehicles can be realized with a reduced delay.

  Also, since all processing is not performed on the server side and resources on the vehicle side are used in a distributed manner, a small load on the server side can be realized, and even if the number of vehicles increases, server side resource enhancement is not required In addition, the vehicle can be searched.

  In addition, by reducing the number of vehicles on which inference processing is to be performed, the amount of communication between the vehicle and the server and the average calculation load per vehicle can be reduced, and the total amount of inference processing load can be reduced. can do.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The second embodiment is different from the first embodiment in that an inference unit for calculating a certainty factor is provided on the server side. In the first embodiment, the case where the inference unit for calculating the certainty factor is mounted on the in-vehicle device has been described as an example. However, the inference unit can be mounted on the server side. When the inference unit is mounted on the server side, the communication unit of the server receives the predicted route output by the vehicle behavior prediction unit of the in-vehicle device by communication, and calculates the certainty factor by the inference unit on the server side.

<Configuration of vehicle search system 210 according to the second embodiment>
As shown in FIG. 16, the vehicle search system 210 according to the second embodiment of the present invention uses a plurality of in-vehicle devices 212 provided in a plurality of vehicles and vehicle information transmitted from the plurality of in-vehicle devices 212. A server 214 that receives and searches for a vehicle that satisfies the search condition. The plurality of in-vehicle devices 212 and the server 214 are connected via a network 16 such as the Internet.

[In-vehicle device 212]
The in-vehicle device 212 is based on the position measuring unit 120, the vehicle sensor 124, the position information of the own vehicle measured by the position measuring unit 120, and the speed of the own vehicle detected by the vehicle sensor 124. And a computer 226 for predicting.

  The computer 226 includes a CPU, a RAM, and a ROM that stores a program for executing processing routines to be described later, and is functionally configured as follows. The computer 226 includes an information acquisition unit 128 that acquires position information (X, Y) of the current time t of the host vehicle measured by the position measurement unit 120 and the speed of the host vehicle detected by the vehicle sensor 124, and information The vehicle information acquired by the acquisition unit 128 is sequentially transmitted to the server 214 together with the vehicle ID, and the communication unit 230 that transmits and receives data to and from the server 214, and the position information of the host vehicle acquired by the information acquisition unit 128 And a vehicle behavior prediction unit 132 that predicts the predicted route of the host vehicle based on the speed of the host vehicle, and a task processing unit 136 that executes a task when the instruction information of the task transmitted from the server 214 is received. I have. The communication unit 230 is an example of a vehicle behavior information transmission unit and a vehicle information transmission unit.

  The communication unit 230 sequentially transmits the position information (X, Y) of the current vehicle t acquired by the information acquisition unit 128 to the server 214. Further, the communication unit 230 receives the prediction result transmission request transmitted from the server 214. Further, when the communication unit 230 receives the prediction result transmission request transmitted from the server 214, the communication unit 230 transmits the predicted route predicted by the vehicle behavior prediction unit 132 to the server 214. Further, the communication unit 230 receives instruction information transmitted from the server 214 and instructing a task to be executed by the vehicle.

[Server 214]
The server 214 is composed of a server including a CPU, a ROM that stores a program for realizing each processing routine to be described later, a RAM that temporarily stores data, a memory as a storage unit, a network interface, and the like. Functionally, the reception unit 140, the communication unit 242, the data storage unit 144, the database 145, the simple prediction unit 146, the inference unit 234, the vehicle search unit 248, and the task generation unit 150 can be represented. The communication unit 142 is an example of a prediction result transmission request transmission unit, and the simple prediction unit 146 is an example of a vehicle selection unit. The inference unit 234 is an example of search information generation means.

  The communication unit 242 receives vehicle position information (X, Y) transmitted from each of the plurality of in-vehicle devices 212. In addition, the communication unit 242 transmits a prediction result transmission request to each of the in-vehicle devices 212 of the vehicle selected as a vehicle candidate by the simple prediction unit 146 described later. In addition, the communication unit 242 transmits instruction information generated by the task generation unit 150 described later to the in-vehicle device 212 of the vehicle.

  The inference unit 234 determines the degree of satisfying the search condition as information indicating whether or not the search condition included in the query is satisfied for each of the in-vehicle devices 212 based on each predicted route of the vehicle received by the communication unit 242. Is calculated.

  The vehicle search unit 248 searches for a vehicle that satisfies the search condition based on the certainty factor generated by the inference unit 234.

<Operation of vehicle search system>
Next, the operation of the vehicle search system 210 according to the second embodiment of the present invention will be described. When each of the vehicles equipped with the in-vehicle device 212 travels and the server 214 is activated, the in-vehicle device 212 of each vehicle executes the prediction processing routine shown in FIG. 9, and the server 214 executes the search processing routine shown in FIG. Execute. Further, the inference processing routine shown in FIG. 14 and the certainty factor calculation processing routine shown in FIG. 15 are executed on the server 214 side.

  In addition, about the other structure and effect | action of the vehicle search system which concern on 2nd Embodiment, since it is the same as that of 1st Embodiment, description is abbreviate | omitted.

  As described above, according to the vehicle search system of the second embodiment of the present invention, the in-vehicle device repeatedly predicts the predicted route of the vehicle until after a predetermined prediction time based on the vehicle information. When the prediction result transmission request transmitted from the server is received, the predicted route of the predicted vehicle is transmitted to the server, and the server transmits the input search condition and vehicle information transmitted from each of the plurality of in-vehicle devices. And selecting a candidate vehicle satisfying the search condition from each of the plurality of vehicles, transmitting a prediction result transmission request to each of the in-vehicle devices of the vehicle selected as the vehicle candidate, Based on the predicted route of the vehicle transmitted from each, the certainty factor is calculated for each of the in-vehicle devices, and by searching for the vehicle based on the certainty factor, a delay in the vehicle search can be suppressed.

  Note that the present invention is not limited to the embodiment described with reference to the drawings, and various modifications can be applied without departing from the scope of the invention.

  For example, in the above embodiment, the case where the route of the vehicle is predicted has been described as an example, but the present invention is not limited to this. For example, when the vehicle sensor 124 is configured using a car navigation system, a destination is input to the car navigation system, and the driving route of the vehicle is indicated to the driver, the driver is highly likely to use the car navigation system. Therefore, an improvement such as making the route a predicted route is also conceivable.

  In the above embodiment, the case where the predicted route is predicted as the time series of the vehicle behavior is described as an example. However, the present invention is not limited to this. For example, the destination is predicted as the time series of the vehicle behavior. May be.

  Further, in the above embodiment, the simple prediction unit has been described as an example of calculating the certainty factor based on the position information of the vehicle, but is not limited to this. For example, the simple prediction unit is not limited to the traveling direction of the vehicle. The certainty factor can also be calculated based on angle information such as whether or not the road segment A (point A) is ahead of the vector.

  In the above embodiment, the task generation unit 150 generates a task according to the certainty factor, and the task processing unit 136 executes the task. However, the present invention is not limited to this. There is no need to generate a task or execute a task.

Further, in the above-described embodiment, an example in which detecting the speed of the vehicle by the vehicle sensor 124, calculates the T (r _t) of the average required travel time r _t with the speed of the vehicle Although described, the present invention is not limited to this. For example, T (r _t ) may be used as an average required travel time predetermined for r _t .

  The storage medium for storing the program of the present invention is not particularly limited, and may be a hard disk or a ROM. Further, it may be a CD-ROM, a DVD disk, a magneto-optical disk or an IC card. Furthermore, the program may be downloaded from a server or the like connected to the network.

DESCRIPTION OF SYMBOLS 10,210 Vehicle search system 12,212 Vehicle-mounted apparatus 14,214 Server 16 Network 120 Position measurement part 124 Vehicle sensor 126,226 Computer 128 Information acquisition part 130,230 Communication part 132 Vehicle behavior prediction part 134,234 Inference part 136 Task processing Unit 140 reception unit 142, 242 communication unit 144 data storage unit 145 database 146 simple prediction unit 148, 248 vehicle search unit 150 task generation unit

Claims (6)

A vehicle search system comprising a plurality of in-vehicle devices provided in a plurality of vehicles and an information processing device,
The in-vehicle device is
Vehicle information transmission means for transmitting vehicle information related to the vehicle to the information processing device;
Vehicle behavior prediction means for repeatedly predicting the time series of the behavior of the vehicle based on the vehicle information until after a predetermined prediction time;
Whether or not the search condition is satisfied based on a time series of the behavior of the vehicle predicted by the vehicle behavior prediction means when the search condition regarding the position and time transmitted from the information processing apparatus is received. Search information generating means for generating information to represent;
Search information transmitting means for transmitting to the information processing apparatus information indicating whether or not the search condition generated by the search information generating means is satisfied,
The information processing apparatus includes:
Vehicle selection means for selecting a candidate for the vehicle that satisfies the search condition from each of the plurality of vehicles based on the input search condition and the vehicle information transmitted from each of the plurality of in-vehicle devices. When,
Search condition transmission means for transmitting the search condition to each of the in-vehicle devices of the vehicle selected as the vehicle candidate by the vehicle selection means;
A vehicle search system comprising: vehicle search means for searching for a vehicle that satisfies the search condition based on information indicating whether or not the search condition is transmitted from each of the in-vehicle devices. A vehicle search system comprising a plurality of in-vehicle devices provided in a plurality of vehicles and an information processing device,
The in-vehicle device is
Vehicle information transmission means for transmitting vehicle information related to the vehicle to the information processing device;
Vehicle behavior prediction means for repeatedly predicting the time series of the behavior of the vehicle based on the vehicle information until after a predetermined prediction time;
Vehicle behavior information transmission means for transmitting, to the information processing apparatus, a time series of the behavior of the vehicle predicted by the vehicle behavior prediction means when receiving a prediction result transmission request transmitted from the information processing apparatus; With
The information processing apparatus includes:
Based on the input search conditions related to the position and time and the vehicle information transmitted from each of the plurality of in-vehicle devices, the candidate of the vehicle satisfying the search conditions is selected from each of the plurality of vehicles. Vehicle selection means;
Prediction result transmission request transmission means for transmitting the prediction result transmission request to each of the in-vehicle devices of the vehicle selected as the vehicle candidate by the vehicle selection means;
Search information generating means for generating information indicating whether or not the search condition is satisfied for each of the in-vehicle devices based on the time series of the behavior of the vehicle transmitted from each of the in-vehicle devices;
A vehicle search system comprising: vehicle search means for searching for a vehicle that satisfies the search condition based on information indicating whether or not the search condition is generated generated by the search information generation means. The search condition is a search condition including a travel position of a vehicle and a time section,
The vehicle behavior prediction means repeatedly predicts a time series of the travel position of the vehicle until after a predetermined prediction time as an upper limit of a time interval included in the search condition based on the vehicle information,
The search information generation unit is configured to determine the travel position included in the search condition within the time interval included in the search condition based on a time series of the travel position of the vehicle predicted by the vehicle behavior prediction unit. The vehicle search system according to claim 1, wherein information indicating whether the vehicle travels is calculated.   The vehicle search system according to any one of claims 1 to 3, wherein the information indicating whether or not the search condition is satisfied is a certainty factor indicating a degree of satisfying the search condition. For each of the retrieved vehicles,
The information processing apparatus includes:
A task according to information indicating whether or not the search condition is satisfied, and task generation means for generating instruction information for instructing a task to be executed by the vehicle;
Task transmission means for transmitting the instruction information generated by the task generation means to the in-vehicle device of the vehicle,
The in-vehicle device is
The vehicle search system according to claim 1, further comprising task processing means for executing the task based on the instruction information transmitted from the information processing apparatus. For each of the searched vehicles, the task generation means generates the instruction information indicating whether or not to present information or the type of presentation information according to information indicating whether or not the search condition is satisfied. ,
The vehicle search system according to claim 5, wherein the task processing unit executes a task of presenting information to an occupant of the vehicle based on the instruction information transmitted from the information processing apparatus.

Images