JP4933991B2 - Traffic information collection / distribution method, traffic information collection / distribution system, center apparatus and in-vehicle terminal apparatus - Google Patents

Description

  The present invention relates to a traffic information collection / distribution method, a traffic information collection / distribution system, a center apparatus, and an in-vehicle terminal apparatus that collect and distribute road traffic information based on probe data acquired by a sensor mounted on a vehicle.

  Conventionally, probe cars are often used to obtain road traffic information. A probe car is equipped with an in-vehicle device that includes various sensors, communication devices, etc., collects data such as vehicle position, travel speed, and travel distance (hereinafter referred to as probe data) by the various sensors, and collects the collected probes. A vehicle that transmits data to a predetermined traffic information center. As a probe car, for example, a taxi is often used in cooperation with a taxi company.

  On the other hand, the traffic information center processes probe data transmitted from the probe car, and collects traffic information such as travel time between intersections, traffic jam locations, and traffic jam length. However, in reality, since there are not enough probe cars, the accuracy of collected traffic information is a problem. Thus, in order to increase the number of probe cars, for example, there is a concept of causing a vehicle equipped with a navigation device having a communication function to play a role as a probe car. Many current vehicles are already equipped with sensors necessary for collecting traffic information, and the number of navigation devices having a communication function is expected to increase further in the future.

  Thus, when a large number of vehicles become probe cars and probe data is transmitted from the large number of probe cars to the traffic information center, a problem different from the conventional case arises. First, the first problem is that probe data is transmitted from a large number of probe cars, so that the communication load of the communication line of the traffic information center and the processing load of the computer become enormous. The second problem is that the different probe data transmitted from a plurality of probe cars for the same event on the road (for example, traffic jam at a certain point) can be identified and classified by any method. It is a question of what should be done.

Patent Document 1 discloses an example of a probe car that reduces data to be transmitted to a traffic information center by detecting an event called SS / ST by an in-vehicle device of the probe car. SS (Short Stop) means a stop state where the vehicle is below a predetermined speed, and ST (Short Trip) means a traveling state where the vehicle is at or above a predetermined speed. Probe data such as position and vehicle speed is uplinked (data transfer from the in-vehicle device to the traffic information center). That is, SS / ST is an event-driven uplink method, and this method has an effect of compressing uplink data.
JP 2003-296891 A

  For the second problem, for example, adaptive resonance theory (ART) can be applied as a general method for solving the same problem. That is, a computer that processes probe data performs learning using preset teaching data, and forms a cluster based on the similar data. The probe data input in real time is matched with the cluster to detect and classify the event.

  However, event detection by SS / ST according to Patent Document 1 is different in event detection conditions (such as vehicle running conditions and surrounding conditions), vehicle types and individual differences, sensor types and individual differences, There may be a large difference in the probe data, which may make it difficult to integrate events on the traffic information center side. Even if the information is compressed by SS / ST, the probe data is uplinked when all events occur. Therefore, as long as the number of probe cars, the number of sensor types, and the time resolution improve, the information to be uplinked The amount cannot be reduced.

  In addition, in the application of adaptive resonance theory, in the case of probe data obtained by a vehicle, the characteristics and order of the target data are various and short depending on the number of vehicles, individual differences of vehicles, road driving characteristics, etc. Change in time. Therefore, unlike the application in which the sensor to be determined is limited, it is not possible to easily set teaching data or form a cluster. At least, it is difficult to detect and classify the event in real time for probe data input in real time.

  Accordingly, an object of the present invention is to reduce probe data uplinked from a probe car, and extract similar feature data from a large number of probe data for a certain road section. A traffic information collection / distribution method, a traffic information collection / distribution system, a center device, and an in-vehicle terminal device capable of performing real-time processing for adding and distributing event information related to traffic conditions to a corresponding road section It is to provide.

  The present invention includes an in-vehicle terminal device that is mounted on a vehicle and obtains probe data from a sensor included in the vehicle, and a temporary storage unit that receives and temporarily stores information transmitted from the in-vehicle terminal device. Is a traffic information collection / distribution system configured to be communicably connected to a center device that acquires event information relating to road traffic conditions based on the information obtained, and used for the traffic information collection / distribution system. Traffic information collection / distribution method, center device, and in-vehicle terminal device. And in this invention, the said vehicle-mounted terminal device and the said center apparatus operate | move as follows.

(1) When the vehicle-mounted terminal device detects a predetermined event from the probe data, the probe data related to the road section in which the event is detected and event identification information for identifying the event are sent to the center device. Send.
(2) The center device (2-1) receives the probe data and the event identification information transmitted from the in-vehicle terminal device, and temporarily stores the received probe data and the event identification information. (2-2) Among the probe data stored in the temporary storage means, a plurality of probe data that share a part of the corresponding road section with each other are characterized by principal component analysis. (2-3) a change point in the direction of the feature space vector is detected from the feature space vector obtained by the feature space projection process; and (2-4) the plurality of the change points by the detected change point. (2-5) Each of the plurality of probe data including the road section is divided into each of the divided road sections. Allocates one of the event identification information corresponding to, respectively, distributes the segment information indicating the location of the (2-6) the allocation event identification information and the road section to the vehicle terminal.
(3) The in-vehicle terminal device receives the event identification information and the section information, and is indicated by the event identification information when the current position of the vehicle is included in a road section indicated by the section information. Event information is displayed on the display device.

  According to the present invention, the amount of probe data uplinked from a probe car is reduced, and similar feature data is extracted from a large number of probe data for a certain road section, and the extracted feature data corresponds to the road. Event information related to traffic conditions can be added to a section in real time and distributed.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a diagram illustrating an example of a functional block configuration of a traffic information collection / distribution system according to the first embodiment of the present invention. As shown in FIG. 1, the traffic information collection / distribution system 1 includes a center device 10 and an in-vehicle terminal device 30 mounted on a vehicle. Here, the center device 10 and the in-vehicle terminal device 30 are communicably connected to each other via a communication network (not shown) such as a mobile phone line or the Internet. The in-vehicle terminal device 30 is connected to various sensors 50, a display device 60 and the like mounted on the vehicle.

  Here, the center apparatus 10 includes a probe data reception unit 11, a probe data update unit 12, a principal component score vector transmission unit 13, a feature space projection processing unit 14, a change point detection unit 15, an event section division unit 16, and an event allocation unit. 17, an event data distribution unit 18, a current probe data storage unit 21, a principal component score vector storage unit 22, an event data storage unit 23, and the like.

  The in-vehicle terminal device 30 includes a probe data acquisition unit 31, an event detection unit 32, a probe data transmission unit 33, a probe data division unit 34, an orthogonal component decomposition unit 35, an uplink determination unit 36, and a principal component score vector reception unit 37. , An event data receiving unit 38, an event data display unit 39, a probe data storage unit 41, and other functional blocks.

  As the sensor 50, various sensors generally mounted on a vehicle can be used. For example, vehicle speed sensors, distance sensors, acceleration sensors, brake sensors, accelerator sensors, steering angle sensors, position sensors such as GPS (Global Positioning System) receivers, slip sensors included in ABS (Antilock Braking System), radars, etc. Any of the obstacle sensors may be used.

  In the in-vehicle terminal device 30, the probe data acquisition unit 31 acquires probe data input from various sensors 50 and stores the acquired probe data in the probe data storage unit 41. In addition, the event detection unit 32 detects an event from the probe data acquired by the probe data acquisition unit 31 and adds an event label indicating the type of the detected event, for example, information such as “traffic jam” to the probe data. Then, the probe data transmission unit 33 transmits (uplinks) the probe data to which the event label is added to the center apparatus 10 based on the instruction information from the probe data division unit 34 or the uplink determination unit 36.

  In this embodiment, the event detection by the event detection unit 32 can be detected by monitoring probe data from a single sensor or a plurality of sensors 50. For example, a “congestion” event is detected when the vehicle speed is equal to or lower than a predetermined speed, and a “congestion” event label is added to the probe data obtained at that time. At this time, a plurality of event labels may be added to the probe data. The portion of the probe data to which the event label is added is called an event section.

  In the in-vehicle terminal device 30, the principal component score vector receiving unit 37 receives the principal component score vector (described later in detail) transmitted from the center device 10, and the probe data dividing unit 34 is stored in the probe data storage unit 41. The stored probe data is divided into a portion included in the received principal component score vector section and a portion included outside the section. Then, the probe data transmitting unit 33 is instructed to uplink the probe data of the portion included outside the section. Further, the orthogonal component decomposition unit 35 extracts data of a component (orthogonal component) different from the principal component score vector by performing orthogonal component decomposition on the probe data of the portion included in the section of the principal component score vector. Then, the uplink determination unit 36 determines whether or not the data of the orthogonal component exists, and when the data of the different component exists, uploads the data of the orthogonal component to the center device 10 in the probe data. The probe data transmission unit 33 is instructed to link.

  In the in-vehicle terminal device 30, the event data receiving unit 38 receives event data distributed from the center device 10. Since the event data is attached with section information indicating which road section the event data corresponds to, the event data display unit 39 includes the road section in which the vehicle is traveling in the received event data. When there is something, the event data is displayed on the display device 60. Note that the display device 60 is configured by, for example, an LCD (Liquid Crystal Display) or the like, and may share the display device of the navigation device.

  Next, in the center device 10, the probe data receiving unit 11 receives the probe data transmitted from the in-vehicle terminal device 30, and stores the received probe data in the current probe data storage unit 21. Further, the probe data update unit 12 removes the probe data stored in the current probe data storage unit 21 from the date and time information (time stamp) attached to the probe data that is out of the current time window. Here, the current time window means a period from a time before a predetermined time (for example, 5 minutes before) to the current time. That is, the probe data update unit 12 removes old data from the probe data stored in the current probe data storage unit 21.

  The feature space projection processing unit 14 performs principal component analysis processing on the probe data stored in the current probe data storage unit 21, calculates a feature space vector and a principal component score vector, and further calculates the calculated principal component score vector. Are stored in the principal component score vector storage unit 22. Then, the change point detection unit 15 detects a change point in the direction of the calculated feature space vector. The event section dividing unit 16 divides the road section based on the change point, and the event assigning unit 17 assigns the event label attached to the probe data to the divided road section, and road section information And the event label are stored in the event data storage unit 23 as event data. The event data distribution unit 18 distributes the event data stored in the event data storage unit 23 to the in-vehicle terminal device 30.

  The function of each functional block below the feature space projection processing unit 14 in the center apparatus 10 will be described in detail later.

  In FIG. 1, a center device 10 is configured by a computer including a CPU (Central Processing Unit) and a storage device (not shown), and the function of each functional block of the center device 10 is stored by the CPU. This is realized by executing a predetermined program stored in the apparatus. The storage device includes a RAM (Random Access Memory), a flash memory, a hard disk device, and the like.

  Similarly, the in-vehicle terminal device 30 is configured by a computer that includes a CPU and a storage device (not shown), and the function of each functional block of the in-vehicle terminal device 30 is stored in the storage device. This is realized by executing a predetermined program. The storage device is constituted by a RAM, a flash memory, a hard disk device, or the like.

  FIG. 2 is a diagram illustrating a configuration of probe data transmitted from the in-vehicle terminal device to the center device and event data distributed from the center device to the in-vehicle terminal device in the first embodiment.

  In FIG. 2, the date and time information of the probe data is information representing the date and time when the probe data is acquired. The section information is information related to a road section in the road link from which the probe data has been acquired (a road connecting intersections is referred to as a road link). The road link identification information, the event occurrence position information in the road link, It includes information such as distance information of a section where probe data related to the event exists. The event label is information for identifying the type of the event added by the event detection unit 32.

  If the in-vehicle terminal device 30 does not have road map information including road link identification information or position information, the in-vehicle terminal device 30 cannot attach road link identification information. In this case, latitude / longitude information obtained from a GPS receiver or the like may be used as event occurrence position information, and road link identification information may be added by the center device 10.

The probe data body is data d _ij (j = 1,..., N: n is the number of data) acquired from sensor #i (i = 1,..., S: where s is the number of sensors). Consists of. Note that d _ij acquired from the sensor #i is generally acquired as time-series data, but in the present embodiment, the time-series data is combined with, for example, data of a travel distance sensor. It is assumed that the data is converted into data having the travel distance as the main axis, for example, data obtained every time the vehicle travels 1 m.

When uplink is instructed by the uplink determination unit 36, the probe data to be _uplinked is not the data _dij itself acquired from the sensor #i (i = 1,..., S) but the orthogonal component. The orthogonal component data for the principal component score vector extracted by the decomposition unit 35 is obtained.

  The event data distributed from the center device 10 to the in-vehicle terminal device 30 includes date / time information, section information, and an event label. The section information in this case includes road link identification information and position information of at least two points on the road link. The event label is an event label assigned to the section by the event assigning unit 17, and a plurality of event labels may be assigned.

  The event data storage unit 23 stores a large number of event data configured as described above. And the event data distribution part 18 may distribute the said event data separately for every vehicle-mounted terminal device 30, Usually, it is the area with respect to the several vehicle-mounted terminal device 30 located in a predetermined area. Event data having section information of road sections included in the is simultaneously distributed by multicast or the like.

  FIG. 3 is a diagram showing an outline of a process flow in the traffic information collection / distribution system according to the first embodiment. That is, the flow of this process is as follows. The in-vehicle terminal device 30 acquires probe data from the sensor 50, detects an event from the probe data, and uplinks only the necessary minimum probe data to the center device 10. 10 shows the processing from the time when the uplink probe data is integrated or separated from the event data held so far, to generate new event data, and to distribute the held and generated event data. is there. In this process, it is assumed that there are a plurality (large number) of in-vehicle terminal devices 30.

  As shown in FIG. 3, when the in-vehicle terminal device 30 detects an event such as a traffic jam from the probe data acquired from the sensor 50 (step S10), it transmits an “uplink notification” to the center device 10 (step S10). S11). The “uplink notification” is a notification that informs the center device 10 that the in-vehicle terminal device 30 is about to uplink probe data. In the present embodiment, the “uplink notification” may be information that requests the center device 10 to transmit a principal component score vector. The “uplink notification” is attached with the current position information of the vehicle on which the in-vehicle terminal device 30 is mounted.

  When the center device 10 receives the “uplink notification”, the center device 10 refers to the principal component score vector storage unit 22 based on the attached current position information of the vehicle, and the event integrated section is included in the road link on which the vehicle is traveling. It is determined whether or not there is (step S12). Here, the event-integrated section means a road section associated with the principal component score vector, and details thereof will be described later.

  If the result of the determination is that there is an event integrated section (Yes in step S12), the center apparatus 10 transmits a principal component score vector including the integrated section information to the in-vehicle terminal device 30 (step S13). ). When there is no event integrated section (No in Step S12), the center apparatus 10 transmits an “unconditional uplink request” to the in-vehicle terminal apparatus 30 (Step S14).

  On the other hand, when the data received at that time is an “unconditional uplink request” (Yes in step S15), the in-vehicle terminal device 30 uploads all probe data related to the event to the center device 10 including the event label. Link (step S16). Further, when the received data is not an “unconditional uplink request” (No in step S15), that is, when the received data is a principal component score vector including integrated section information, the in-vehicle terminal device 30 divides the probe data into sections based on the integrated section information (step S17).

  When a new section that is not included in the integrated section is generated by the section division, probe data of the new section is extracted. For the probe data included in the integrated section, orthogonal component decomposition is performed using a principal component score vector (step S18), and the orthogonal component is extracted as a new event component. When a new section or a new component (orthogonal component) of probe data is extracted in this way (Yes in step S19), the in-vehicle terminal device 30 includes the event label, and the probe data of the extracted portion is sent to the center device 10 (Step S20). Further, when neither a new section nor a new component is extracted (No in step S19), the probe data is not uplinked.

  Next, upon receiving the uplink probe data (all probe data, new section or new component probe data) from the in-vehicle terminal device 30, the center apparatus 10 temporarily stores the probe data in the current probe data storage unit 21. After that, by performing feature space projection processing on the probe data belonging to the new event integrated section including the event integrated section and the new section, the event section is determined by detecting the change point of the resulting feature space vector. Divide (step S21). Then, the center device 10 assigns an event label to the divided event section, associates the section information of the event section with the event label, and stores them in the event data storage unit 23 (step S22).

  Further, the center device 10 distributes the event data stored in the event data storage unit 23 to the in-vehicle terminal device 30 at predetermined time intervals such as every 5 minutes (step S23). The in-vehicle terminal device 30 receives the distributed event data, collates the section information included in the received event data with the current position of the vehicle, and has the section information including the current position. If there is data, the event data is displayed on the display device (step S24).

  As described above, the center device 10 can collect event data, that is, traffic information from the in-vehicle device 30 mounted on each of a plurality of vehicles traveling on the road by the processing shown in FIG. The traffic information can be distributed to the in-vehicle device 30. Therefore, the in-vehicle device 30 can acquire and display event data detected by the other in-vehicle devices 30, that is, traffic information, even before the in-vehicle device 30 detects it.

  Next, the process that characterizes the present embodiment in the above process will be described in more detail using an example.

  FIG. 4 is a diagram schematically illustrating an example of feature space projection processing in principal component analysis according to the first embodiment. Principal component analysis is a technique for extracting data correlated to each other from a large number of data and integrating the data to reduce the amount of data and make it easier to grasp the characteristics of the data. Therefore, since it is considered that there is a high correlation between the probe data acquired from the in-vehicle terminal devices 30 of a plurality of vehicles for the same event, the probe data is integrated by applying principal component analysis. It can be done.

  For example, as illustrated in FIG. 4, it is assumed that probe data of data A, data B, and data C is acquired by a plurality of in-vehicle terminal devices 30. At this time, the correlation between the data A, B, and C is very high. Therefore, it can be said that these are data formed by the same event. However, the data C has a large data change rate due to, for example, individual differences between vehicles, and the correlation with the data A and data B is slightly reduced.

  When principal component analysis is performed on such data A, B, and C, it can be converted into so-called feature space data in which the number of types of data is reduced. Such data transformation is often referred to as feature space projection. In FIG. 4, data A, B, and C are converted into data X that is a component in which data A, B, and C change in correlation and data Y that is a component in which data C changes uncorrelated with data A and B. Has been. That is, the coordinate axis of the feature space is selected so that highly correlated data can be represented.

  In the center apparatus 10 of this embodiment, the feature space projection as described above is performed by the feature space projection processing unit 14. Data X and data Y obtained by projecting data A, B, and C in the probe data space onto the feature space are called principal component score vectors. In other words, the principal component score vector is history information of coordinate values for each coordinate axis of data in the feature space. The principal component score vector obtained by the feature space projection is stored in the principal component score vector storage unit 22.

  The coordinate value represented by the data in the feature space is called a feature space vector. In this embodiment, a change in the direction of the feature space vector is captured and determined as an event change point. For example, in the example of FIG. 4, the data X is data having a large norm (absolute value of the vector magnitude) and a large contribution of the event. The data Y is data having a small norm and a small contribution of the event. In the synthesis of a plurality of vectors, the direction of the vector to be synthesized is affected by the direction of the vector having a large norm, and the direction of the feature space vector in this case is greatly influenced by the value of the data X. Therefore, in the example of FIG. 4, since the value of the data X changes greatly in the vicinity of the position indicated by the broken line, the feature space vector also changes greatly in the vicinity of the position of the broken line. Therefore, it is determined that the event is changing near the broken line.

  Therefore, in this embodiment, the change point of the feature space vector is detected, and the event section is divided by the change point. Incidentally, in the example of FIG. 4, the section before the broken line is set as event α, and the section after the broken line is set as event β. In the center apparatus 10, these processes are performed by the change point detection unit 15 and the event section division unit 16.

Note that the principal component score vector and the feature space vector are different from each other. For example, d _ij (i = 1,..., S, j = 1,..., N) is used as probe data, and c _kj (k = 1,..., U, j = 1,..., N, u <s) is set. It is assumed that the probe data is feature space data converted by feature space projection. At this time, if c _kj is an element of the matrix C, the column vector (c _1j , c _2j ,..., C _uj ) ^t (j = 1,..., N) of the matrix C is the feature space vector, and the row vector ( c _k1 , c _k2 ,..., c _kn ) (k = 1,..., u) are the principal component score vectors.

  FIG. 5 is a diagram showing the concept of event integration section determination in the center device according to the first embodiment. When new probe data is uplinked from the in-vehicle terminal device 30 to the center device 10, the current probe data storage unit 21 of the center device 10 already has the probe data uplinked from the in-vehicle terminal device 30 of the preceding vehicle. Often exists. In FIG. 5, the probe data is represented as existing probe data # 1, # 2,.

  Further, as shown in FIG. 5, the event sections of these existing probe data # 1, # 2,..., #M cause a positional shift due to road congestion, vehicle running conditions, and the like. However, as long as the event sections overlap, the center apparatus 10 performs the feature space projection process by integrating the probe data into sections including all the event sections, assuming that the probe data originates from the same event.

  Therefore, when new probe data is uplinked, the event sections are once integrated by the existing probe data # 1, # 2,. Therefore, in this embodiment, an event interval obtained by integrating event intervals of probe data that already exists when new probe data is uplinked is referred to as an event integrated interval.

  When new probe data is uplinked, the center apparatus 10 forms a new event integrated section by logical sum of the event integrated section and the event section of the new probe data, and the existing probe data # 1, # 2,..., #M and new probe data are subjected to feature space projection processing. At this time, if there is old probe data that has passed a predetermined time or more in the event integrated section, the old probe data is excluded from the target of the feature space projection process.

  When the feature space projection process is performed on probe data having different event section ranges as shown in FIG. 5, each probe data is a missing value with respect to the event integration section subject to the feature space projection process. Will have. On the other hand, in this embodiment, although the amount of calculation increases, a principal component analysis method with a missing value that estimates and interpolates a value of a missing section is used.

  FIG. 6 is a diagram illustrating an example of event label allocation in the center device according to the first embodiment.

  Upon receiving the probe data uplink, the center apparatus 10 performs a feature space projection process on the new event integration section as described above, detects a change point of the feature space vector, and detects the event section. Split. Then, the center device 10 assigns an event label to each of the divided event sections.

  In assigning event labels, it is determined whether or not the same event label has been assigned to all probe data to be integrated. When all the same event labels are attached, the event labels are assigned to the divided event sections. When all event labels are not attached, that is, when different event labels are mixed, the most event label is selected by majority decision, and the most event label is assigned to the event section.

  By the way, in FIG. 6, the event label attached to the probe data in the event section of event # 1 is the same “congestion”, so the event label “congestion” is assigned to the event section of event # 1. It has been. In addition, event labels of “obstacle” and “slip” are assigned to the event sections of event # 3 and event # 4, respectively, by majority vote.

  FIG. 7 is a diagram illustrating an example of probe data division and orthogonal component decomposition in the in-vehicle terminal device according to the first embodiment, and FIG. 8 is a diagram illustrating a basic concept of the orthogonal component decomposition.

  As described above, the in-vehicle terminal device 30 transmits “uplink notification” to the center device 10 when uplinking probe data. On the other hand, the center device 10 determines whether or not there is an existing event integrated section on the road link on which the vehicle on which the in-vehicle terminal device 30 is mounted is running, and when there is an existing event integrated section, The principal component score vector already obtained by the feature space projection process for the existing probe data included in the event integrated section is transmitted to the in-vehicle terminal device 30.

  The in-vehicle terminal device 30 receives the principal component score vector, and compares the event interval of the received principal component score vector with the event interval of probe data to be uplinked (hereinafter referred to as a new uplink interval). Then, as shown in FIG. 7, the new uplink section is divided into a section (1) included in the event section of the principal component score vector and a section (2) not included in the event section of the principal component score vector. (Probe data dividing unit 34). On the other hand, for the probe data in the section (1), the probe data is projected onto the basis vector of the received principal component score vector, and is orthogonal to the projection component (a) and the basis vector of the principal component score vector. It decomposes into the orthogonal component (b) (orthogonal component decomposition unit 35). The probe data in the section (2) is considered to be an uplink target because it may have event information that is not included in the previous event integrated section.

8, the probe data, such as (Example 1) Y, when the principal component score vector is A, is decomposed into a component Y _B which is orthogonal thereto and projection component Y _A to its basis vectors. That is, since the orthogonal component Y _B means that it has some event information not included in the principal component score vector A, the orthogonal component Y _B is the target of uplink. However, as shown in FIG. 8 (Example 2) Y, if the orthogonal component Y _B is the norm even if small, it is determined that the event corresponding thereto is not present, is not used for the uplink. The in-vehicle terminal device 30 appropriately determines the presence / absence of an orthogonal component by providing a threshold value and comparing the norm of the extracted orthogonal component Y _B with the threshold value.

  As described above, when the norm of the orthogonal component does not reach the predetermined threshold, the in-vehicle terminal device 30 includes new information other than the event information that the center device 10 already has in the probe data of the portion. It is determined that there is no probe data, and the probe data is excluded from the uplink target. Therefore, the amount of probe data uplinked from the in-vehicle terminal device 30 to the center device 10 can be reduced.

  FIG. 9 is a diagram illustrating an example of a display method of event data in the in-vehicle terminal device according to the first embodiment.

  The in-vehicle terminal device 30 receives the event data distributed from the center device 10 and displays the received event data on the display device of the in-vehicle terminal device 30. At this time, as shown in FIG. 9, the center apparatus 10 distributes the event # 1, # 2, # 3 assigned by the existing event integrated section with an “existing flag”, and creates a new Events # 4 and # 5 assigned by the probe data are distributed with a “new flag” attached. The in-vehicle terminal device 30 displays the event data with the “existing flag” and the event data with the “new flag” on the display device so that they can be distinguished from each other by their color, shape, or the like. .

  In addition, for an event detected by the in-vehicle terminal device 30 itself, when there is an event that is not included in the distributed event data, the in-vehicle terminal device 30 refers to the event detected by the in-vehicle terminal device 30 as the color of the distributed event. It is displayed on the display device so that it can be identified by the shape and the like.

  FIG. 10 is a diagram illustrating an operation flow of the center device according to the first embodiment. As shown in FIG. 10, when the center device 10 receives an uplink notification transmitted from the in-vehicle terminal device 30 (step S31), the center device 10 is based on the current position information of the vehicle attached to the uplink notification. With reference to the score vector storage unit 22, it is determined whether or not there is an event integrated section on the road link on which the vehicle is traveling (step S32). When there is an event integrated section in the road link (Yes in step S32), the center apparatus 10 transmits a principal component score vector related to the event integrated section to the in-vehicle terminal device 30 (step S33). On the other hand, when there is no event-integrated section (No in step S32), an unconditional uplink request is transmitted to the in-vehicle terminal device 30 (not shown: corresponding to step S14 in FIG. 3).

  Next, when the center device 10 receives the probe data transmitted from the in-vehicle terminal device 30 (step S34), the center device 10 stores the received probe data in the current probe data storage unit 21 (step S35). Then, the center apparatus 10 confirms the time stamp of the data (probe data) stored in the current probe data storage unit 21 (step S36), and determines whether there is data outside the current time window width. (Step S37). As a result of the determination, if there is data outside the current time window width (Yes in step S37), the data outside the current time window width is removed from the current probe data storage unit 21 (step S38).

  Subsequently, the center apparatus 10 performs feature space projection processing of principal component analysis on the probe data included in the event integrated section formed by the event section of the received probe data and the event integrated section (step S39). . Then, the center device 10 detects a change point of the feature space vector obtained by the feature space projection process (step S40), and further divides the event section based on the change point (step S41).

  Next, the center apparatus 10 assigns an event label to each of the divided event sections by performing a loop process from step S42 to step S46 (see FIG. 6). In the loop processing, the center device 10 collates the event section with the event detection position attached to the probe data (step S43), and totals the event labels attached to the probe data in the event event section (step S43). Step S44). Then, among the event labels in the event section, the largest number of event labels are assigned as representative event labels of the event section (step S45).

  Finally, the center device 10 distributes the event data labeled by the above event label assignment to the in-vehicle terminal device 30 (step S47). In addition, although the structure of one event data at the time of delivery is based on FIG. 2, as shown in FIG. 9, when distinguishing event data based on existing probe data and data based on new probe data, Identification flags (“existing” flag and “new” flag) are attached.

  FIG. 11 is a diagram illustrating an operation flow of the in-vehicle terminal device according to the first embodiment. As illustrated in FIG. 11, when the in-vehicle terminal device 30 detects an event from the probe data acquired by the sensor 50 (step S51), the in-vehicle terminal device 30 transmits an “uplink notification” to the center device 10 (step S52). Then, since a principal component score vector or an unconditional uplink request is transmitted from the center device 10, the in-vehicle terminal device 30 determines whether or not it is a principal component score vector (step S53).

  If the result of the determination is that it is not a principal component score vector (No in step S53), that is, if it is an unconditional uplink request, the in-vehicle terminal device 30 centers the probe data and the event label. Uplink to the device 10 (step S54). On the other hand, when it is a principal component score vector (Yes in step S53), as shown in FIG. 5, the probe data is divided based on the information of the event integrated section included in the principal component score vector. (Step S55).

  Next, the in-vehicle terminal device 30 performs orthogonal component decomposition on the probe data of the portion included in the existing section (event-integrated section) of the divided probe data based on the received principal component score vector (step S56: FIG. 8), it is determined whether the norm of the orthogonal component is equal to or greater than a predetermined threshold (step S57). As a result of the determination, when the norm of the orthogonal component is equal to or greater than a predetermined threshold (Yes in step S57), the in-vehicle terminal device 30 displays the probe data of the new section, the orthogonal component of the existing section, and the event label as the center device. Uplink to 10 (step S58). On the other hand, when the norm of the orthogonal component does not reach the predetermined threshold value (No in step S57), the in-vehicle terminal device 30 uplinks the probe data and the event label of the new section to the center device 10 (step S59). .

  Subsequently, the in-vehicle terminal device 30 receives the event data distributed from the center device 10 (step S60), and displays the received event data on the display device (step S61). The display method is as shown in FIG.

  As described above, according to the first embodiment, when the in-vehicle terminal device 30 detects an event, the in-vehicle terminal device 30 does not transmit all the probe data to the center device 10, but (1) the principal component score vector is transmitted from the center device 10. (2) when there is an orthogonal component with the principal component score vector transmitted from the center device 10, and (3) when it is outside the interval of the principal component score vector transmitted from the center device 10. The orthogonal component of the probe data or the principal component score vector is transmitted to the center device 10. In other words, the in-vehicle terminal device 30 does not transmit the probe data to the center device 10 when the probe data has the same characteristics as the probe data of the center device 10. Therefore, even if the same event is detected by the in-vehicle terminal devices 30 of a plurality of vehicles, if the probe data is similar, the probe data is not transmitted to the center device 10 in duplicate. Absent. Therefore, the amount of probe data transmitted from the in-vehicle terminal device 30 to the center device 10 can be reduced, and as a result, the processing load on the center device 10 is also reduced.

  Further, according to the first embodiment, similar feature data is extracted from a plurality of probe data by the feature space projection process of the principal component analysis method, and an event is assigned to a road section where the extracted feature data exists. ing. Since this principal component analysis method only extracts feature data correlated with each other from a large number of data, there is no need for teaching data as in adaptive resonance theory, and the type of probe data, that is, A result that does not depend on the type of sensor 50 or its individual difference can be obtained. Therefore, in the first embodiment, even if probe data are successively input to the center device 10, the center device 10 extracts feature data from the probe data in real time and continuously, and the feature data. Can be assigned to the event, and the assigned event can be distributed to the in-vehicle terminal device.

  Note that the first embodiment described above can be variously modified. For example, the principal component score vector is periodically distributed from the center device 10, and when the in-vehicle terminal device 30 detects an event, transmission of the “uplink notification” to the center device 10 may be omitted. Absent. In this case, the same effect as that of the first embodiment can be obtained.

  Moreover, the vehicle-mounted terminal device 30 in 1st Embodiment demonstrated above is realizable as some car navigation apparatuses which have a communication function. In this case, the in-vehicle terminal device 30 can easily display the event data distributed from the center device 10 on the map. Therefore, the in-vehicle terminal device 30 does not have to limit the event data received to be displayed when the own vehicle hits the road link where the event has occurred. Traffic information can be displayed on a map.

(Second Embodiment)
Next, a second embodiment of the present invention will be described in detail with reference to FIG. Here, FIG. 12 is a diagram showing an example of the functional block configuration of the traffic information collection / distribution system according to the second embodiment of the present invention.

  As shown in FIG. 12, the traffic information collection / distribution system 101 according to the second embodiment includes a center device 110, a probe car on-board terminal device 130 mounted on a probe car, and a portable device mounted on a general vehicle. Type navigation terminal device 140. Here, the center device 110 and the probe car in-vehicle terminal device 130 or the portable navigation terminal device 140 are connected to be communicable with each other via a communication network (not shown) such as a mobile phone line or the Internet.

  The portable navigation terminal device 140 is generally called a PND (Personal Navigation Device) and is an inexpensive navigation device that can be easily attached to and detached from the vehicle and has a simple function. Therefore, the minimum sensors required for navigation, for example, a GPS (Global Positioning System) position sensor, a direction sensor such as a gyro, and an acceleration sensor are only connected to the PND. However, these sensors are important sensors for detecting the traveling state of the vehicle, and are represented as a motion sensor 152 in FIG.

  In contrast, the probe car on-board terminal device 130 is mounted on a probe car that acquires traffic information, and as shown in FIG. 12, not only the motion sensor 152 but also the environment sensor 151 is connected to the probe car on-board terminal device 130. The Here, the environment sensor 151 is a sensor that mainly detects the situation around the vehicle. For example, an on-vehicle radar, an infrared camera, a slip sensor, and the like can be assumed. In the present embodiment, the output information of the environment sensor 151 is used as information for identifying normal / abnormal vehicle running conditions, that is, events on the road.

  As shown in FIG. 12, the probe car in-vehicle terminal device 130 includes a probe data acquisition unit 131, a probe data storage unit 133, and a probe data transmission unit 132.

  Here, the probe data acquisition unit 131 acquires the probe data from the environment sensor 151 and the motion sensor 152 every predetermined time or every time the vehicle travels a predetermined distance, and the acquired probe data is probed. Temporarily stored in the data storage unit 133. Then, the probe data transmission unit 132 transmits the probe data temporarily stored in the probe data storage unit 133 to the center apparatus 110 for each trip.

  Here, one trip means that the vehicle travels in a predetermined road section (for example, a road section connecting an intersection and an intersection, hereinafter referred to as a road link). One-trip probe data refers to probe data obtained by a vehicle traveling on a certain road link at a certain time. Accordingly, when the same vehicle travels on the same road link at different dates and times, they become probe data of different trips.

  As described above, in this embodiment, the probe car on-vehicle terminal device 130 only needs to have a function of transmitting probe data acquired from the environment sensor 151 and the motion sensor 152 to the center device 110. Therefore, the probe car in-vehicle terminal device 130 is not limited to the one mounted on the so-called probe car, and has a configuration in which the environment sensor 151 and the motion sensor 152 are connected and can communicate with the center device 110. If so, it may be a navigation device mounted on a general vehicle.

  Next, the portable navigation terminal device 140 includes an exercise probe data acquisition unit 141, an exercise probe data transmission unit 142, an event data reception unit 143, an event data display unit 144, an exercise probe data storage unit 145, an event And a data storage unit 146.

  Here, the motion probe data acquisition unit 141 acquires motion probe data from the motion sensor 152 every predetermined time or every time the vehicle travels a predetermined distance, and the acquired motion probe data is acquired as the motion probe data. Temporarily stored in the storage unit 145. Then, the motion probe data transmission unit 142 transmits the motion probe data temporarily stored in the motion probe data storage unit 145 to the center device 110 every trip.

  Further, the event data receiving unit 143 receives event data associated with a road link distributed from the center device 110 and stores the received event data in the event data storage unit 146. Further, the event data display unit 144 displays the event data stored in the event data storage unit 146 on the display device 160 by superimposing it on map data or the like.

  As described above, in the present embodiment, the portable navigation terminal device 140 need not be connected to the environment sensor 151, but is connected to the motion sensor 152, and the motion probe data acquired from the motion sensor 152 is the center device 110. It has a function as an in-vehicle terminal device of a probe car to transmit to. The portable navigation terminal device 140 also has a function as a navigation terminal device that displays road link event data transmitted from the center device 110 on the display device 160.

  Next, the configuration and operation of the center device 110 will be described. As shown in FIG. 12, the center apparatus 110 is roughly divided into a probe data accumulation processing unit 110a and a real-time event processing unit 110b. Hereinafter, the configuration and operation of the center apparatus 110 will be described separately for the probe data accumulation processing unit 110a and the real-time event processing unit 110b.

  In the center device 110, the probe data accumulation processing unit 110a is based on the motion probe data and the environment probe data transmitted from the probe car on-board terminal device 130, and the feature space of the motion probe data in a situation where there is no event on the road. Are generated, and an event detection threshold value for detecting an event is calculated based on a residual with the feature space of the motion probe data, and the event detection threshold value is accumulated in the basis vector storage unit 125 and the event detection threshold value storage unit 126. Responsible for processing.

  In addition, the real-time event processing unit 110b receives the motion probe data transmitted from the portable navigation terminal device 140, the motion probe data, the basis vectors of the feature space accumulated in the basis vector storage unit 125, The event detection threshold stored in the event detection threshold storage unit 126 detects an event, registers the detected event in a predetermined storage device, and stores the event data registered in the storage device as a portable navigation terminal Responsible for processing to be distributed to the device 140.

  As shown in FIG. 12, the probe data accumulation processing unit 110a of the center device 110 includes a probe data receiving unit 111, a motion probe data separating unit 112, a feature space generating unit 113, an abnormal residual detecting unit 114, and a normal residual detecting unit. 115, processing function blocks such as the event detection threshold calculation unit 116, the environmental probe data storage unit 121, the motion probe data storage unit 122, the abnormal probe data storage unit 123, the normal probe data storage unit 124, the basis vector storage unit 125, the event And a storage function block such as a detection threshold value storage unit 126.

  The probe data receiving unit 111 receives the probe data transmitted from the probe car in-vehicle terminal device 130, adds a trip ID capable of uniquely identifying the probe data to the received probe data, and further, a time stamp at that time Or a link ID for identifying the road link from which the probe data is acquired. Then, the probe data receiving unit 111 appropriately classifies the probe data to which the trip ID or the like is added into environmental probe data and motion probe data, and the sorted probe data is respectively stored in the environment probe data storage unit 121 and the motion probe data. Temporarily stored in the probe data storage unit 122.

  The motion probe data separation unit 112 has the same trip ID as the motion probe data temporarily stored in the environmental probe data storage unit 121 for normality or abnormality of the motion probe data temporarily stored in the motion probe data storage unit 122. Judge based on environmental probe data. That is, the motion probe data separation unit 112 determines that the motion probe data is abnormal when an abnormal value such as an obstacle or slip is detected in the environmental probe data, and when the abnormal value is not detected. The motion probe data is determined to be normal.

  The motion probe data separation unit 112 accumulates the motion probe data in the normal probe data storage unit 124 when it is determined that the motion probe data is normal, and when it is determined as abnormal, The motion probe data is accumulated in the abnormal probe data storage unit 123. That is, the motion probe data when there is no event on the road link (that is, when normal) is accumulated in the normal probe data storage unit 124, and when any event occurs on the road link (that is, when there is an abnormality). The motion probe data is accumulated in the abnormal probe data storage unit 123.

  When abnormal probe data is accumulated in the abnormal probe data storage unit 123, the abnormal probe data includes event identification information (for example, fault information) determined according to the abnormal value of environmental probe data corresponding to the abnormal probe data. There is information on things, lane restrictions, road freezing etc.).

  Next, the feature space generation unit 113 sorts the normal probe data stored in the normal probe data storage unit 124 for each link ID, that is, for each road link, and performs principal component analysis on the normal probe data for each sorted road link. Implement and generate its feature space. Then, the feature space generation unit 113 stores the generated basis vectors of the feature space in the basis vector storage unit 125 in association with the respective road links. Note that the base vector of the feature space corresponds to a feature representing the motion characteristics of the vehicle traveling on the road link when there is no event on the road link.

  The normal residual detection unit 115 calculates a residual between the normal probe data for each road link and the feature space generated from the normal probe data. Similarly, the abnormal residual detection unit 114 calculates a residual between the abnormal probe data for each road link and the feature space generated from the normal probe data of the road link. Note that the vector of the difference between the vector represented by the probe data and the projection vector of the vector onto the feature space is called the residual vector, and the absolute value (norm) of the residual vector is called the residual. Hereinafter, the residual vector is abbreviated and may be simply referred to as a residual.

  Next, the residual will be described with reference to FIGS. 13 and 14. Here, FIG. 13 is a schematic diagram for explaining the residual of normal probe data, and FIG. 14 is a schematic diagram for explaining the residual of abnormal probe data.

  FIG. 13A shows an example of normal probe data accumulated in the normal probe data storage unit 124. The normal probe data here is, for example, the movement probe data with respect to the traveling direction of the vehicle. It represents the acceleration in the lateral direction. As described above, in this embodiment, the probe data is acquired as data at intervals of 10 m, for example, for each trip of the vehicle. Therefore, it can be said here that the probe data of the acceleration of one trip is a vector in which a change history according to the position of acceleration in a certain road section (road link) is expressed as a discrete value. For example, if the road section of the trip is 200 m, 21 pieces of acceleration data are obtained at intervals of 10 m. Therefore, the acceleration data of the trip constitutes a 21-dimensional vector space.

  When the feature space generation unit 113 performs principal component analysis, the probe data of the acceleration to be analyzed is sorted for each link ID. That is, the principal component analysis is performed on the probe data of the acceleration of each sorted road link. Therefore, if the probe data of the acceleration of the road link is a 21-dimensional vector and the probe data of the n trip acceleration is acquired for the road link, the feature space generation unit 113 has n rows × 21. Perform principal component analysis on the matrix of columns.

  In FIG. 13B, examples of basis vectors of the feature space obtained by the principal component analysis are represented as a basis 1 and a basis 2. That is, in this example, the feature space is represented by a space (in this example, a two-dimensional plane) spanned by the base 1 vector and the base 2 vector. Note that the feature space is not limited to two dimensions.

  These basis vectors can be said to be components of an acceleration change history pattern common to n-trip acceleration probe data on the road link. Therefore, when there is no event such as the appearance of an obstacle or road freezing on the road link, the probe data of the acceleration of the trip can be approximately represented by a vector obtained by appropriately combining the base vectors.

  In FIG. 13B, white circles indicate the positions of the points represented by the acceleration probe data vector included in FIG. 13A. A black circle is a projected point when the point represented by the acceleration probe data vector is projected onto a feature space (a space spanned by the base 1 vector and the base 2 vector). At this time, the residual is expressed as the distance between the position represented by the white circle (point represented by the probe data vector) and the position represented by the black circle (projection point). In FIG. 13, since the acceleration probe data is normal probe data in the absence of an event, the residual does not normally increase. Therefore, when the probe data is normal probe data, the projection point can be said to be an approximate point representing the probe data on the feature space.

  FIG. 14A shows an example of the abnormal probe data stored in the abnormal probe data storage unit 123. The abnormal probe data here is, for example, the movement probe data in the traveling direction of the vehicle. It represents the acceleration in the lateral direction. The difference between abnormal probe data and normal probe data is whether or not there was an event on the road link when the probe data was acquired. is there.

  In FIG. 14B, since the basis 1 and the basis 2 which are basis vectors of the feature space are generated by normal probe data, the residuals of abnormal probe data often increase. In other words, when the vehicle performs an obstacle avoidance operation or slips due to freezing, the change in acceleration history is naturally different from that when there is no event. Therefore, there are many cases where the abnormal probe data cannot be represented by simply combining the basis vectors. That is, in FIG. 14B, the distance between the point (white circle) represented by the vector of the abnormal probe data and the projected point (black circle) often increases.

  As described above, due to the facts described in FIGS. 13 and 14, the residual with respect to the feature space of normal probe data (hereinafter referred to as normal residual) is usually small, and the residual with respect to the feature space of abnormal probe data (hereinafter referred to as abnormal residual). The difference) is usually large.

  Next, the operation of the event detection threshold value calculation unit 116 will be described with reference to FIGS. 12 and 15. FIG. 15 is a diagram showing an example of a normal residual and abnormal residual histogram created by the event detection threshold value calculation unit 116.

  The event detection threshold value calculation unit 116 creates a frequency distribution (histogram) of normal residuals for each road link for the normal residuals detected by the normal residual detection unit 115, and further detects abnormal residuals. For the abnormal residual detected by the unit 114, a frequency distribution of the abnormal residual is created for each road link. Then, as shown in FIG. 15, normal residuals are distributed on the smaller residual side, and abnormal residuals are distributed on the larger residual side.

  Next, the event detection threshold value calculation unit 116 sets an event detection threshold value between the normal residual and the abnormal residual. This event detection threshold is a threshold for determining whether or not the motion probe data should belong to normal probe data or to abnormal probe data when certain motion probe data is acquired. . Therefore, when it is determined by the determination that the motion probe data should belong to the abnormal probe data, it is estimated that some event exists in the road link from which the motion probe data is acquired. The

  As shown in FIG. 15, the normal residual histogram and the abnormal residual histogram generally have overlapping portions. Therefore, in the event detection based on the event detection threshold, it is usually impossible to avoid false or missing information. Therefore, the false alarm rate and the false alarm rate are defined here.

  The false alarm rate is the probability of determining that an event has occurred even though no event has occurred, and its value is the frequency of normal residuals that is greater than the event detection threshold for the total frequency of normal residuals. Can be defined by the ratio of numbers. The unreported rate is the probability of determining that an event has not occurred even though an event has occurred, and its value is smaller than the event detection threshold for the total frequency of abnormal residuals. It can be defined by the ratio of frequency numbers of differences.

  As can be seen from FIG. 15, when the event detection threshold is increased, the false alarm rate decreases, but the misreport rate increases. On the other hand, when the event detection threshold is reduced, the false alarm rate decreases, but the false alarm rate increases. Therefore, in the process of the event detection threshold value calculation unit 116, for example, the event detection threshold value is set so that the ratio between the false alarm rate and the false alarm rate is equal to a predetermined ratio. At this time, if the ratio is “1”, it means that the false alarm rate and the false alarm rate are equal. Alternatively, an upper limit value may be set for the false alarm rate or the false alarm rate, and the event detection threshold value may be set so as to satisfy the upper limit value.

  The event detection threshold calculation unit 116 sets the event detection threshold for each road link as described above, and stores the set event detection threshold in the event detection threshold storage unit 126 in association with the road link. .

  Although the description of the configuration and operation of the probe data accumulation processing unit 110a of the center apparatus 110 is completed as described above, the processing performed by the probe data accumulation processing unit 110a functions to operate the real-time event processing unit 110b described below. This is a preliminary preparation process. This preparation process can be implemented, for example, by running a probe car on which the probe car on-board terminal device 130 is mounted on a road in a predetermined target area for a month, for example. As a result of the traveling, the number of pieces of probe data that can be statistically processed for each road link is obtained for each road link, and as a result, for each road link, An event detection threshold is determined.

  Next, functions and operations of the real-time event processing unit 110b of the center apparatus 110 will be described with reference to FIG. As shown in FIG. 12, the real-time event processing unit 110 b includes processing function blocks such as an exercise probe data receiving unit 117, a residual detection unit 118, an event detection unit 119, and an event data distribution unit 120, and an exercise probe data storage unit 127. And a storage function block such as the event data storage unit 128.

  The motion probe data receiving unit 117 receives the motion probe data transmitted from the portable navigation terminal device 140, and temporarily stores the received motion probe data in the motion probe data storage unit 145. At this time, the motion probe data receiving unit 117 obtains the link ID of the road link from which the motion probe data is acquired based on the GPS position data included in the motion probe data, and temporarily stores the link ID. Append to data. In addition, you may perform addition of link ID in the exercise | movement probe data transmission part 142 of the portable navigation terminal device 140. FIG.

  Next, the residual detection unit 118 refers to the basis vector storage unit 125 and reads a basis vector corresponding to a road link having a link ID added to the motion probe data. Then, the residual detection unit 118 detects the residual of the motion probe data with respect to the feature space spanned by the read basis vectors.

  Next, the event detection unit 119 refers to the event detection threshold value storage unit 126 and reads an event detection threshold value corresponding to the road link having the link ID. Then, the event detection unit 119 detects the presence or absence of an event as described in FIG. 15 based on the residual detected by the residual detection unit 118 and the read event detection threshold. As a result, when an event is detected, the event is associated with the road link and registered in the event data storage unit 128 as event data.

  Further, the event data distribution unit 120 refers to the event data storage unit 128, and for the road link in a predetermined region including the road link specified by the link ID attached to the acquired motion probe data, the road Event data registered corresponding to each link is read, and the read event data is distributed to the portable navigation terminal device 140.

  FIG. 16 is a schematic diagram for explaining a residual of motion probe data acquired from the portable navigation terminal device 140. FIG. 16A shows an example of probe data temporarily stored in the motion probe data storage unit 127. The motion probe data here is the same as in the case of FIG. 13 and FIG. It represents the acceleration in the lateral direction with respect to the direction, and is represented by, for example, a 21-dimensional vector having 21 discrete values.

  In FIG. 16B, the basis vectors represented by the basis 1 and the basis 2 are obtained by reading out the basis vector corresponding to the road link from which the motion probe data is obtained from the basis vector storage unit 125. is there. In this case, the feature space is expressed in two dimensions (plane). In FIG. 16B, the white circles indicate the positions represented by the motion probe data, and the black circles indicate the projection points of the motion probe data onto the feature space.

  At this time, the residual is represented by the distance between the position represented by the probe data and the projection point, and is obtained by the residual detection unit 118. The residual is usually small when there is no event and large when there is an event. Therefore, the event detection unit 119 reads the event detection threshold obtained in advance from the event detection threshold storage unit 126, and compares the residual obtained by the residual detection unit 118 with the event detection threshold, thereby Detect the presence or absence.

  In FIG. 12, the real-time event processing unit 110b detects an event for the motion probe data sequentially transmitted from a large number of portable navigation terminal devices 140, and when an event is detected, the event is detected as event data. Is registered in the event data storage unit 128.

  In general, since an event occurring on a road changes from moment to moment, the event data registered in the event data storage unit 128 is deleted when a predetermined time elapses. Further, when the event detection unit 119 detects different results of “with event” or “without event” for the same road link within a predetermined time, for example, the event detection unit 119 determines the presence or absence of an event by majority vote.

  This is the end of the description of the configuration and operation of the real-time event processing unit 110b of the center apparatus 110.

  The center device 110 described above is realized by a computer having a CPU and a storage device (not shown). In that case, a predetermined program is stored in the storage device, and when the CPU executes the program, the probe data reception unit 111, the motion probe data separation unit 112, the feature space generation unit 113, and the abnormal residual detection unit 114 The functions of the processing function blocks such as the normal residual detection unit 115, the event detection threshold calculation unit 116, the motion probe data reception unit 117, the residual detection unit 118, the event detection unit 119, and the event data distribution unit 120 are realized. In addition, the environmental probe data storage unit 121, the motion probe data storage unit 122, the abnormal probe data storage unit 123, the normal probe data storage unit 124, the basis vector storage unit 125, the event detection threshold storage unit 126, the motion probe data storage unit 127, Storage function blocks such as the event data storage unit 128 are configured on the storage device.

  Next, with reference to FIGS. 12, 17, and 18, the operation of the center apparatus 110 described above will be described in the form of a flow chart. Here, FIG. 17 is a diagram illustrating an operation flow of the probe data accumulation processing unit 110a, and FIG. 18 is a diagram illustrating an operation flow of the real-time event processing unit 110b.

  First, the operation flow of the probe data accumulation processing unit 110a will be described with reference to FIG.

  The center device 110 receives environmental probe data and motion probe data transmitted from the probe car on-board terminal device 130 as processing of the probe data reception unit 111 (step S101), and receives the received environmental probe data and motion probe data. These are temporarily stored in the environmental probe data storage unit 121 and the motion probe data storage unit 122, respectively.

  Next, the center device 110 performs a motion probe temporarily stored in the motion probe data storage unit 122 based on the environment probe data temporarily stored in the environment probe data storage unit 121 as a process of the motion probe data separation unit 112. The data is separated into normal probe data and abnormal probe data (step S102), and the separated normal probe data and abnormal probe data are accumulated in the normal probe data storage unit 124 and the abnormal probe data storage unit 123, respectively. As described above, in the separation processing, the motion probe data is classified as abnormal probe data when an abnormal value such as an obstacle or slip is detected in the environmental probe data corresponding to the motion probe data. If no abnormal value is detected, the probe is classified as normal probe data.

  Next, the center device 110 generates a feature space of normal probe data for each road link as a process of the feature space generation unit, and stores the basis vector in the basis vector storage unit 125 (step S103).

  Next, as a process of the normal residual detection unit 115, the center device 110 performs a process for the normal probe data stored in the normal probe data storage unit 124 for each road link, with respect to the feature space of the normal probe data. The difference is detected (step S104). In addition, the center device 110 performs, as the processing of the abnormal residual detection unit 114, for the abnormal probe data stored in the abnormal probe data storage unit 123, for each road link, the residual with respect to the feature space of the abnormal probe data. Is detected (step S105).

  Next, the center device 110, as the processing of the event detection threshold value calculation unit 116, for each road link, a histogram of residuals (normal residuals) detected by the normal residual detection unit 115 and an abnormal residual detection unit Based on the histogram of the residual (abnormal residual) detected by 114, an event detection threshold is calculated, and the calculated event detection threshold is stored in the event detection threshold storage unit 126 (step S106).

  Next, the operation flow of the real-time event processing unit 110b will be described with reference to FIG.

The center device 110 receives the motion probe data transmitted from the portable navigation terminal device 140 as the processing of the motion probe data reception unit 117 (step S111), and temporarily stores the received motion probe data in the motion probe data storage unit 145. Remember.

  Next, the center device 110 refers to the basis vector storage unit 125 as processing of the residual detection unit 118, and acquires the basis vector of the road link (the road link from which the received motion probe data is acquired) ( Step S112). Then, the center apparatus 110 detects a residual between the feature space spanned by the basis vector and the received motion probe data (step S113).

  Next, the center device 110 refers to the event detection threshold value storage unit 126 as a process of the event detection unit 119, and acquires the event detection threshold value of the road link (step S114). The center device 110 detects an event of the road link based on the residual detected in step S113 and the event detection threshold acquired in step S114 (step S115), and further detects the detected event as an event data storage unit. It registers in 146 (step S116).

  Further, as a process of the event data distribution unit 120, the center device 110 refers to the event data storage unit 146 and acquires an event registered in a road link in a predetermined area including the road link (step S117). Then, the event registered in the road link is distributed as event data to the portable navigation terminal device (step S118).

  As described above, in the description of the second embodiment so far, the residual is the absolute value (magnitude) of the residual vector for easy understanding. In that case, since only one event detection threshold is set for each road link, the event detection unit 119 can detect the presence or absence of an event, but cannot detect the type of event. . Therefore, hereinafter, it will be described that the type of event can be detected also in the second embodiment. In the following, in order to make it possible, the residual is considered as a residual vector which is the original definition.

  FIG. 19 is a schematic diagram for explaining the basic concept for detecting the type of event. In FIG. 19, a base 1 and a base 2 are basis vectors of the feature space, a white circle is a position represented by a vector of abnormal probe data, and a black circle is a projection point of the abnormal probe data onto the feature space. At this time, the residual vector is represented as a vector of a difference between the abnormal probe data vector and the projection vector represented by the projection point.

  In general, vehicle motion probe data has different characteristics depending on events that occur in road links such as obstacle avoidance, lane restrictions, and road freezing. It is. That is, as illustrated in FIG. 19, the directions of the residual vectors are different when the event types are different, and are the same when the event types are the same. In FIG. 19, the residual vector orthogonal to the feature space is displayed so as not to be orthogonal. However, since the vector in the multidimensional space is originally displayed in the three-dimensional space, it is unavoidably displayed as such. Is. Conceptually, the feature space and the residual vector are orthogonal.

  In this case, it is necessary to associate the direction of the residual with the event in advance. In order to perform the association, the abnormal probe data stored in the abnormal probe data storage unit 123 is sorted for each event (event identification information). Then, a histogram as shown in FIG. 15 is created for each sorted event. At this time, the horizontal axis of the histogram is not the magnitude of the residual vector but the direction of the residual vector.

  In general, since the residual vector is a multidimensional vector, the direction of the residual vector is represented by a multidimensional vector. Therefore, the histogram is represented as a histogram in a multidimensional space. In such a case, when the histogram is viewed in a cross section along each coordinate axis of the multidimensional space, an event detection threshold value along each coordinate axis can be set in the same manner as described in FIG. . For example, in the case of an m-dimensional residual vector, m event detection thresholds can be set.

  Therefore, the event detection threshold value detection unit 116 calculates an event detection threshold value for each dimension of the residual vector (for example, m) for each road link and for each event type, and calculates the event detection threshold value. For example, m event detection threshold values are stored in the event detection threshold value storage unit 126. Then, the event detection unit 119 compares the direction of the residual vector of the motion probe data acquired from the portable navigation terminal device 140 with m event detection thresholds for each event type, thereby determining the event. Determine the type. At this time, even if the residual vector of the acquired motion probe data does not satisfy all of the m event detection threshold values, for example, if the event detection threshold value of 2/3 or more is satisfied, the direction is the same. You may determine that it is facing the direction.

  As described above, the center device 110 can detect the type of the event by determining the direction of the residual vector of the motion probe data acquired from the portable navigation terminal device 140.

  As described above, according to the second embodiment of the present invention, only the motion probe data can be acquired by generating the feature space of the vehicle travel for each road link in advance and calculating the event detection threshold. Even with motion probe data received from a portable navigation terminal device 140 such as a PND, an event occurring on the road link can be detected from the motion probe data. The detected event can be registered in the storage device for each road link, and the registered road link event can be distributed to the portable navigation terminal device 140. That is, according to the present embodiment, even the portable navigation terminal device 140 that can acquire only motion probe data can be used as a probe car that collects road link events.

It is the figure which showed the example of the structure of the functional block of the traffic information collection and delivery system which concerns on the 1st Embodiment of this invention. In the 1st Embodiment of this invention, it is the figure which showed the structure of the probe data transmitted to a center apparatus from a vehicle-mounted terminal device, and the event data delivered from a center apparatus to a vehicle-mounted terminal device. It is the figure which showed the outline | summary of the flow of the process in the traffic information collection and delivery system which concerns on the 1st Embodiment of this invention. It is the figure which showed typically the example of the feature space projection process in the principal component analysis which concerns on the 1st Embodiment of this invention. It is the figure which showed the idea of the event integrated area determination in the center apparatus which concerns on the 1st Embodiment of this invention. It is the figure which showed the example of event label allocation in the center apparatus which concerns on the 1st Embodiment of this invention. It is the figure which showed the example of the division | segmentation of probe data and orthogonal component decomposition | disassembly in the vehicle-mounted terminal device which concerns on the 1st Embodiment of this invention. It is the figure which showed the basic idea of orthogonal component decomposition | disassembly of the probe data in the vehicle-mounted terminal device which concerns on the 1st Embodiment of this invention. It is the figure which showed the example of the display method of the event data in the vehicle-mounted terminal device which concerns on the 1st Embodiment of this invention. It is the figure which showed the operation | movement flow of the center apparatus which concerns on the 1st Embodiment of this invention. It is the figure which showed the operation | movement flow of the vehicle-mounted terminal device which concerns on the 1st Embodiment of this invention. It is the figure which showed the example of the structure of the functional block of the traffic information collection and delivery system which concerns on the 2nd Embodiment of this invention. It is a schematic diagram for demonstrating the residual of normal probe data which concerns on the 2nd Embodiment of this invention. It is a schematic diagram for demonstrating the residual of the abnormal probe data which concern on the 2nd Embodiment of this invention. It is the figure which showed the example of the histogram of the normal residual and the abnormal residual produced by the event detection threshold value calculation part which concerns on the 2nd Embodiment of this invention. It is a schematic diagram for demonstrating the residual of the exercise | movement probe data acquired from the portable navigation terminal device which concerns on the 2nd Embodiment of this invention. It is the figure which showed the operation | movement flow of the probe data accumulation | storage process part which concerns on the 2nd Embodiment of this invention. It is the figure which showed the operation | movement flow of the real-time event processing part which concerns on the 2nd Embodiment of this invention. It is a schematic diagram explaining the basic idea for detecting the kind of event which concerns on the 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Traffic information collection / delivery system 10 Center apparatus 11 Probe data receiving part 12 Probe data update part 13 Principal component score vector transmission part 14 Feature space projection process part 15 Change point detection part 16 Event area division part 17 Event allocation part 18 Event data Distribution unit 21 Current probe data storage unit 22 Principal component score vector storage unit 23 Event data storage unit 30 In-vehicle terminal device 31 Probe data acquisition unit 32 Event detection unit 33 Probe data transmission unit 34 Probe data division unit 35 Orthogonal component decomposition unit 36 Up Link determination unit 37 Principal component score vector reception unit 38 Event data reception unit 39 Event data display unit 41 Probe data storage unit 50 Sensor 60 Display device 101 Traffic information collection / distribution system 110 Center device 110a Professional Data storage processing unit 110b real-time event processing unit 111 probe data reception unit 112 motion probe data separation unit 113 feature space generation unit 114 abnormal residual detection unit 115 normal residual detection unit 116 event detection threshold calculation unit 117 motion probe data reception unit 118 Residual detection unit 119 Event detection unit 120 Event data distribution unit 121 Environmental probe data storage unit 122 Motion probe data storage unit 123 Abnormal probe data storage unit 124 Normal probe data storage unit 125 Base vector storage unit 126 Event detection threshold storage unit 127 Motion Probe data storage unit 128 Event data storage unit 130 Probe car in-vehicle terminal device 131 Probe data acquisition unit 132 Probe data transmission unit 133 Probe data storage unit 140 Type navigation terminal apparatus 141 motion probe data acquisition unit 142 motion probe data transmitting unit 143 the event data receiving unit 144 the event data display unit 145 motion probe data storage unit 146 the event data storage unit 151 environmental sensor 152 motion sensor 160 display

Claims (21)

A vehicle-mounted terminal device that is mounted on a vehicle and obtains probe data from a sensor included in the vehicle, and a temporary storage unit that receives and temporarily stores information transmitted from the vehicle-mounted terminal device, the temporarily stored information And a center device that acquires event information relating to traffic conditions on the road is connected to be communicable,
The in-vehicle terminal device is
When a predetermined event is detected from the probe data, the probe data related to the road section in which the event is detected, and event identification information for identifying the event are transmitted to the center device,
The center device is
Receiving the probe data and the event identification information transmitted from the in-vehicle terminal device, temporarily storing the received probe data and the event identification information in the temporary storage means;
Of the probe data stored in the temporary storage means, for a plurality of probe data, such as to share a part of the corresponding road section to each other, I rows feature space projection processing based on principal component analysis, the road Get the feature space vector at each road position in the section ,
For the unit vector of the acquired feature space vector at each road position, a vector representing the amount of change of the unit vector with respect to the change in the road position is obtained, and when the magnitude of the obtained vector exceeds a predetermined magnitude Detecting the road position as a change point in the direction of the feature space vector ;
Dividing a road section related to the plurality of probe data by the detected change point,
Assigning one of the event identification information corresponding to each of the plurality of probe data including the road section to each of the divided road sections;
Distributing the allocated event identification information and section information indicating the location of the road section to the in-vehicle terminal device,
The in-vehicle terminal device is
A traffic information collection / distribution method comprising: receiving the event identification information and the section information; and displaying the event information and the section information indicated by the received event identification information on a display device. The center device further includes:
A principal component score vector is obtained by the feature space projection process, and the principal component score vector is transmitted to the in-vehicle terminal device,
When the in-vehicle terminal device receives the principal component score vector,
The acquired probe data is divided into intra-section data included in a road section corresponding to the principal component score vector and out-of-section data not included in the road section,
For the intra-section data, extract orthogonal component data for the principal component score vector,
When the orthogonal component data is equal to or greater than a predetermined threshold, the orthogonal component data and the out-of-section data are transmitted to the center device as probe data,
The traffic information collection / distribution method according to claim 1, wherein when the orthogonal component data is less than a predetermined threshold, the out-of-section data is transmitted as probe data to the center device. The in-vehicle terminal device further includes
After acquiring the probe data from the sensor, before transmitting the probe data to the center device, transmit principal component score vector request information for requesting the center device to transmit the principal component score vector,
The center device is
The said principal component score vector which concerns on the road area in which the present position of the said vehicle-mounted terminal device is included is transmitted to the said vehicle-mounted terminal device, when the said principal component score vector request information is received. Traffic information collection and distribution method. The center device is
The event identification information is allocated to each of the divided road sections, and the largest number of event identification information is allocated among a plurality of event identification information corresponding to a plurality of probe data including the road section. The traffic information collection / distribution method according to 1. The in-vehicle terminal device is
When the event identification information of the event detected by the probe data acquired by itself is not included in the distributed event identification information for the road section where the vehicle is traveling, the event indicated by the distributed event identification information The traffic information collection / distribution method according to claim 1, wherein event information of an event detected by the user is displayed on a display device together with the information. A vehicle-mounted terminal device that is mounted on a vehicle and obtains probe data from a sensor included in the vehicle, and a temporary storage unit that receives and temporarily stores information transmitted from the vehicle-mounted terminal device, the temporarily stored information A traffic information collection / distribution system configured to be communicably connected to a center device that acquires event information related to road traffic conditions,
The in-vehicle terminal device is
When a predetermined event is detected from the probe data, the probe data related to the road section in which the event is detected, and event identification information for identifying the event are transmitted to the center device,
The center device is
Receiving the probe data and the event identification information transmitted from the in-vehicle terminal device, temporarily storing the received probe data and the event identification information in the temporary storage means;
Of the probe data stored in the temporary storage means, for a plurality of probe data, such as to share a part of the corresponding road section to each other, I rows feature space projection processing based on principal component analysis, the road Get the feature space vector at each road position in the section ,
For the unit vector of the acquired feature space vector at each road position, a vector representing the amount of change of the unit vector with respect to the change in the road position is obtained, and when the magnitude of the obtained vector exceeds a predetermined magnitude Detecting the road position as a change point in the direction of the feature space vector ;
Dividing a road section related to the plurality of probe data by the detected change point,
Assigning one of the event identification information corresponding to each of the plurality of probe data including the road section to each of the divided road sections;
Distributing the allocated event identification information and section information indicating the location of the road section to the in-vehicle terminal device,
The in-vehicle terminal device is
A traffic information collection / distribution system that receives the event identification information and the section information, and displays the event information and the section information indicated by the received event identification information on a display device. The center device further includes:
A principal component score vector is obtained by the feature space projection process, and the principal component score vector is transmitted to the in-vehicle terminal device,
When the in-vehicle terminal device receives the principal component score vector,
The acquired probe data is divided into intra-section data included in a road section corresponding to the principal component score vector and out-of-section data not included in the road section,
For the intra-section data, extract orthogonal component data for the principal component score vector,
When the orthogonal component data is equal to or greater than a predetermined threshold, the orthogonal component data and the out-of-section data are transmitted to the center device as probe data,
The traffic information collection / distribution system according to claim 6, wherein when the orthogonal component data is less than a predetermined threshold, the out-of-section data is transmitted as probe data to the center device. The in-vehicle terminal device further includes
After acquiring the probe data from the sensor, before transmitting the probe data to the center device, transmit principal component score vector request information for requesting the center device to transmit the principal component score vector,
The center device is
The said principal component score vector which transmits the said principal component score vector which concerns on the road area in which the present position of the said vehicle-mounted terminal device is contained is received when the said principal component score vector request | requirement information is received. Traffic information collection and distribution system. The center device is
The event identification information is allocated to each of the divided road sections, and the largest number of event identification information is allocated among a plurality of event identification information corresponding to a plurality of probe data including the road section. 6. The traffic information collection / distribution system according to 6. The in-vehicle terminal device is
When the event identification information of the event detected by the probe data acquired by itself is not included in the distributed event identification information for the road section where the vehicle is traveling, the event indicated by the distributed event identification information The traffic information collection / distribution system according to claim 6, wherein event information of an event detected by the device itself is displayed on a display device together with the information. A vehicle-mounted terminal device that is mounted on a vehicle and obtains probe data from a sensor included in the vehicle, and a temporary storage unit that receives and temporarily stores information transmitted from the vehicle-mounted terminal device, the temporarily stored information A center device used for a traffic information collection / distribution system configured to be communicably connected to a center device that acquires event information related to road traffic conditions,
Receiving the probe data and the event identification information of the event transmitted when the in-vehicle terminal device detects a predetermined event by the probe data;
Temporarily storing the received probe data and the event identification information in the temporary storage means;
Of the probe data stored in the temporary storage means, for a plurality of probe data, such as to share a part of the corresponding road section to each other, I rows feature space projection processing based on principal component analysis, the road Get the feature space vector at each road position in the section ,
For the unit vector of the acquired feature space vector at each road position, a vector representing the amount of change of the unit vector with respect to the change in the road position is obtained, and when the magnitude of the obtained vector exceeds a predetermined magnitude Detecting the road position as a change point in the direction of the feature space vector ;
Dividing a road section related to the plurality of probe data by the detected change point,
Assigning one of the event identification information corresponding to each of the plurality of probe data including the road section to each of the divided road sections;
The center apparatus, wherein the allocated event identification information and section information indicating a location of the road section are distributed to the in-vehicle terminal apparatus. The center device according to claim 11, wherein a principal component score vector is obtained by the feature space projection processing, and the principal component score vector is transmitted to the in-vehicle terminal device. When the principal component score vector request information transmitted from the in-vehicle terminal device is received, the principal component score vector relating to a road section including the current position of the in-vehicle terminal device is transmitted to the in-vehicle terminal device. The center device according to claim 12. The event identification information is allocated to each of the divided road sections, and the largest number of event identification information is allocated among a plurality of event identification information corresponding to a plurality of probe data including the road section. Item 12. The center device according to Item 11. A vehicle-mounted terminal device that is mounted on a vehicle and obtains probe data from a sensor included in the vehicle, and a temporary storage unit that receives and temporarily stores information transmitted from the vehicle-mounted terminal device, the temporarily stored information Based on the on-vehicle terminal device used in the traffic information collection / distribution system configured to be communicably connected to the center device that acquires event information related to traffic conditions on the road,
When a predetermined event is detected from the probe data, the probe data related to the road section in which the event is detected and event identification information for identifying the event are transmitted to a center device, and the center device Receiving probe data and event identification information, temporarily storing the received probe data and event identification information in the temporary storage means, and corresponding to the probe data stored in the temporary storage means a plurality of probe data, such as to share a part of the road section to each other, each I rows feature space projection processing based on principal component analysis, it obtains a feature space vector at each road location of the road section, and the obtained For the unit vector of the feature space vector at the road position, the road position of the unit vector Obtains a vector representing the variation with respect to the change, the road position when magnitude of the calculated vector exceeds a predetermined magnitude, is detected as the direction of the change point of the feature space vector, the change point that the detected Dividing a road section related to the plurality of probe data, and obtaining and distributing each of the divided road sections by assigning one of event identification information corresponding to each of the plurality of probe data including the road section Receiving the section information indicating the location of the road section and the event identification information assigned to the road section,
An in-vehicle terminal device characterized in that event information and section information indicated by the received event identification information are displayed on a display device. When the principal component score vector by the feature space projection process transmitted by the center device is received,
The acquired probe data is divided into intra-section data included in a road section corresponding to the principal component score vector and out-of-section data not included in the road section,
For the intra-section data, extract orthogonal component data for the principal component score vector,
When the orthogonal component data is equal to or greater than a predetermined threshold, the orthogonal component data and the out-of-section data are transmitted to the center device as probe data,
The in-vehicle terminal device according to claim 15, wherein when the orthogonal component data is less than a predetermined threshold, the out-of-interval data is transmitted as probe data to the center device. After the probe data is acquired from the sensor, before transmitting the probe data to the center device, principal component score vector request information for requesting the center device to transmit the principal component score vector is further transmitted. ,
The in-vehicle terminal device according to claim 16. When the event identification information of the event detected by the probe data acquired by itself is not included in the distributed event identification information for the road section where the vehicle is traveling, the event indicated by the distributed event identification information The in-vehicle terminal device according to claim 15, wherein event information of an event detected by the device itself is displayed on a display device together with the information. A first in-vehicle terminal device that is mounted on a first vehicle and obtains motion probe data and environmental probe data from a sensor included in the first vehicle; and a second vehicle that is mounted on a second vehicle. A second in-vehicle terminal device that acquires motion probe data from a sensor and storage means, and acquired from the motion probe data and environment probe data acquired from the first in-vehicle terminal device and the second in-vehicle terminal device And a center device that detects an event related to a traffic condition on a road based on the motion probe data and distributes event information of the detected event to the second in-vehicle terminal device, and is communicably connected. Configured,
The center device is
Based on the environmental probe data acquired together with the motion probe data, the motion probe data acquired from the first in-vehicle terminal device is separated into normal probe data and abnormal probe data, and the separated normal probe data And abnormal probe data are stored in the storage means for each predetermined road section,
Performing principal component analysis on the accumulated normal probe data for each road section, storing the basis vectors of the feature space obtained by the principal component analysis in the storage means,
For each road section, event detection is performed based on frequency information of residuals between the obtained feature space and the normal probe data and frequency information of residuals between the obtained feature space and the abnormal probe data. Calculating a threshold, storing the calculated event detection threshold in the storage means,
afterwards,
Obtaining the motion probe data from the second in-vehicle terminal device;
With reference to the storage means, the basis vector corresponding to the road section from which the motion probe data was acquired and the event detection threshold,
Based on the residual of the feature space spanned by the motion probe data and the basis vector, and the event detection threshold, detects an event in the road section from which the motion lobe data was acquired,
Event information of the detected event is associated with the road section and accumulated in the storage means,
Referring to the storage means, obtain event information associated with the road section and the surrounding road sections stored in the storage means,
Event information associated with the acquired road section and surrounding road sections is distributed to the second in-vehicle terminal device,
The second in-vehicle terminal device is
Receiving event information associated with the distributed road section and surrounding road sections, and displaying the received event information associated with the road section and surrounding road sections on a display device; Traffic information collection / distribution method characterized by A first in-vehicle terminal device that is mounted on a first vehicle and obtains motion probe data and environmental probe data from a sensor included in the first vehicle; and a second vehicle that is mounted on a second vehicle. A second in-vehicle terminal device that acquires motion probe data from a sensor and storage means, and acquired from the motion probe data and environment probe data acquired from the first in-vehicle terminal device and the second in-vehicle terminal device Based on the motion probe data, an event related to a traffic situation on a road is detected, and a center device that distributes event information of the detected event to the second in-vehicle terminal device is connected to be communicable. A configured traffic information collection / distribution system,
The center device is
Based on the environmental probe data acquired together with the motion probe data, the motion probe data acquired from the first in-vehicle terminal device is separated into normal probe data and abnormal probe data, and the separated normal probe data And abnormal probe data are stored in the storage means for each predetermined road section,
Performing principal component analysis on the accumulated normal probe data for each road section, storing the basis vectors of the feature space obtained by the principal component analysis in the storage means,
For each road section, event detection is performed based on frequency information of residuals between the obtained feature space and the normal probe data and frequency information of residuals between the obtained feature space and the abnormal probe data. Calculating a threshold, storing the calculated event detection threshold in the storage means,
afterwards,
Obtaining the motion probe data from the second in-vehicle terminal device;
With reference to the storage means, the basis vector corresponding to the road section from which the motion probe data was acquired and the event detection threshold,
Based on the residual of the feature space spanned by the motion probe data and the basis vector, and the event detection threshold, detects an event in the road section from which the motion lobe data was acquired,
Event information of the detected event is associated with the road section and accumulated in the storage means,
Referring to the storage means, obtain event information associated with the road section and the surrounding road sections stored in the storage means,
Event information associated with the acquired road section and surrounding road sections is distributed to the second in-vehicle terminal device,
The second in-vehicle terminal device is
Receiving event information associated with the distributed road section and surrounding road sections, and displaying the received event information associated with the road section and surrounding road sections on a display device; Traffic information collection / distribution system characterized by A first in-vehicle terminal device that is mounted on a first vehicle and obtains motion probe data and environmental probe data from a sensor included in the first vehicle; and a second vehicle that is mounted on a second vehicle. A second in-vehicle terminal device that acquires motion probe data from the sensor, and a communicably connected to the second in-vehicle terminal device; and storage means, the motion probe data and the environmental probe data acquired from the first in-vehicle terminal device; Based on the motion probe data acquired from the second in-vehicle terminal device, a center device that detects an event relating to a traffic situation on a road and distributes event information of the detected event to the second in-vehicle terminal device Because
Based on the environmental probe data acquired together with the motion probe data, the motion probe data acquired from the first in-vehicle terminal device is separated into normal probe data and abnormal probe data, and the separated normal probe data And abnormal probe data are stored in the storage means for each predetermined road section,
Performing principal component analysis on the accumulated normal probe data for each road section, storing the basis vectors of the feature space obtained by the principal component analysis in the storage means,
For each road section, event detection is performed based on frequency information of residuals between the obtained feature space and the normal probe data and frequency information of residuals between the obtained feature space and the abnormal probe data. Calculating a threshold, storing the calculated event detection threshold in the storage means,
afterwards,
Obtaining the motion probe data from the second in-vehicle terminal device;
With reference to the storage means, the basis vector corresponding to the road section from which the motion probe data was acquired and the event detection threshold,
Based on the residual of the feature space spanned by the motion probe data and the basis vector, and the event detection threshold, detects an event in the road section from which the motion lobe data was acquired,
Event information of the detected event is associated with the road section and accumulated in the storage means,
Referring to the storage means, obtain event information associated with the road section and the surrounding road sections stored in the storage means,
Event information associated with the acquired road section and surrounding road sections is distributed to the second in-vehicle terminal device.

Images