JP2013108972A - Congestion prediction device, and congestion prediction system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a congestion prediction device capable of predict a congestion situation that changes every moment due to an amount of rainfall.SOLUTION: A rainfall information collection part executes rainfall information calculation processing (S100) to calculate the amount of rainfall at the center coordinates of a link. A traffic information collection part executes unit travel time calculation processing (S110) to calculate a unit travel time. A control part calculates an error amount from a ratio between a predicted link travel time being the previous predicted value and a link travel time being a true value (S120), corrects a gain set by the amount of rainfall and the unit travel time with the error amount (S130), and outputs the current predicted link travel time (S140).

Description

  The present invention relates to a traffic jam prediction technique for forecasting and guiding traffic jams caused by rainfall.

Conventionally, in order to make guidance in a navigation device more appropriate, various information is provided to the navigation device by performing data communication with the center.
Among them, the rainfall information has a great influence on the occurrence of traffic jams, and a system for performing processing based on the rainfall information has been proposed. As an example, a system that can obtain rainfall information in real time based on various information transmitted from a vehicle has been proposed (see, for example, Patent Document 1). As another example, a system that acquires rainfall information and executes processing for displaying an optimum route has also been proposed (see, for example, Patent Document 2).

JP 2002-62368 A JP 2003-58982 A

  By the way, in the invention described in Patent Document 2, the traffic volume is measured by a traffic volume monitoring sensor provided as a road management facility on the main road, so to speak, the traffic volume information at the current congestion point is acquired, A process for displaying the optimum route is being executed.

  However, the situation of traffic congestion due to rainfall varies from moment to moment depending on the amount of rainfall. Therefore, for example, even if a road with no traffic jam is selected at the time of setting the destination, there is a place where the road is jammed when the road is reached.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a traffic jam prediction device capable of predicting the traffic jam status that changes every moment according to the amount of rainfall.

In order to achieve the above object, a traffic jam prediction apparatus according to claim 1 includes rainfall information collection means, traffic information collection means, gain setting means, and predicted value output means.
The rainfall information collecting means collects rainfall information and calculates a rainfall amount of the link based on the rainfall information. The rainfall amount of the link may be, for example, the rainfall amount at the intermediate position of the link. The traffic information collecting means collects traffic information and outputs a link travel time that is a time required for the passage of the link based on the traffic information.

  Here, particularly in the present invention, the gain setting means sets the gain based on the rainfall and the link travel time. It is conceivable that the gain is derived from a rainfall amount and a link travel time using a predetermined map. The predicted value output means multiplies the link travel time by the gain set by the gain setting means, and outputs a predicted link travel time at a future predicted time.

  That is, in the present invention, a predicted link travel time at a future predicted time is obtained by setting a gain based on the rainfall amount of the link and the link travel time. As a result, it is possible to predict the traffic situation that changes every moment depending on the amount of rainfall.

  Incidentally, as described above, it is conceivable that the gain setting means sets the gain using a predetermined map or the like. However, as shown in claim 2, the gain setting means may include a gain correction means. The gain correcting unit corrects the gain set by the gain setting unit based on the predicted link travel time output by the predicted value output unit and the actual link travel time at the predicted time. In this way, since the gain is corrected based on the actual link travel time at the predicted time, a more appropriate predicted link travel time can be output.

  Specifically, as shown in claim 3, the gain correction means may correct the gain by adding an error amount based on a ratio between the predicted link travel time at the predicted time and the actual link travel time. Conceivable. In this way, the gain set by the gain setting means can be corrected relatively easily.

As described above, the gain setting means sets the gain based on the rainfall and the link travel time. More specifically, as shown in claim 4, the traffic information collection means includes the link travel time, It is conceivable that the unit travel time required to move the unit distance on the link based on the link travel time is calculated, and the gain setting means sets the gain using the rainfall amount and the unit travel time as parameters. Here, the unit travel time may be an average value of a plurality of vehicles, and is an index indicating the degree of traffic congestion. The unit travel time rT (t) at time t is obtained by the following equation 1, for example.

rT (t) = ΣT (t) / L · N Equation 1

Here, T (t) is the link travel time of each vehicle, L is the link driving distance, and N is the number of vehicles passing through the link.

By the way, although it is possible to acquire rainfall information from weather information organizations, such as a weather station, it is not limited to this.
For example, as shown in claim 5, the rainfall information collecting means includes at least one of the rainfall information from the weather information agency, the wiper information from the probe car, and the content of the contributor's remarks from the WEB system. Can be considered. Here, the wiper information is embodied as the number of times the wiper reciprocates per unit time. Further, the content of the utterance of the poster is a content of utterance related to rain on a bulletin board, for example. Rainfall information collected from a probe car or a WEB system is converted into rainfall by a predetermined table or the like. By doing so, the rainfall amount of the link calculated by the rainfall information collecting means becomes appropriate.

  However, while the rainfall information from the weather information agency has a relatively high reliability, the rainfall information from the WEB system has a relatively low reliability. Therefore, as shown in claim 6, it is preferable that the rainfall information collecting means performs correction by weighting the rainfall amount for each rainfall information to calculate the rainfall amount of the link. By doing so, the rainfall amount of the link calculated by the rainfall information collecting means becomes more appropriate.

  Moreover, although the rainfall information collection means calculates, for example, the amount of rainfall at the intermediate position of the link for each link, there is a distance from the observation site of the rain information. Therefore, as shown in claim 7, it is conceivable that the rainfall information collecting means performs correction based on the distance between the observation location of the rainfall information and the link, and calculates the rainfall amount of the link. For example, the weighting coefficient based on the distance from the observation site is multiplied by the rainfall amount at the observation site to correct it. It is conceivable that the weight coefficient is obtained for each region by statistically processing the raining radius. By doing so, the rainfall amount of the link calculated by the rainfall information collecting means becomes more appropriate.

  Furthermore, the rainfall information collecting means calculates the rainfall amount at time t for each link, but there is a time lag from the observation time of the rain information. Therefore, as shown in claim 8, it is conceivable that the rainfall information collecting means performs correction based on the elapsed time from the observation time of the rainfall information and calculates the rainfall amount of the link. For example, the weighting factor based on the elapsed time from the observation time is corrected by multiplying the rainfall amount at the observation site. It is conceivable that the weighting factor is obtained for each region by statistically processing the time until rain stops once it starts to rain. By doing so, the rainfall amount of the link calculated by the rainfall information collecting means becomes more appropriate.

  As described above, the invention has been described as a traffic jam prediction device. As shown in claim 9, a navigation device capable of performing data communication between the traffic jam prediction device and the traffic jam prediction device and providing guidance based on the predicted link travel time. It can also be realized as a traffic jam prediction system with The guidance here includes not only route guidance including traffic jam prediction performed by route search, but also guidance for displaying traffic jam information on a map. In such a traffic jam prediction system, the same effect as the traffic jam prediction device is produced.

It is a block diagram showing a schematic structure of a traffic jam prediction system of an embodiment. It is a logic diagram which shows the function of a center. It is explanatory drawing which visualizes and shows the gain to be corrected. It is a flowchart which shows the prediction process performed in a center. It is a flowchart which shows the rainfall calculation process in a prediction process. (A) is explanatory drawing which shows the correspondence of various rainfall information, (b) is explanatory drawing which shows an example of the reliability coefficient with respect to rainfall. It is explanatory drawing which shows calculation of the rainfall in a link from the rainfall acquired at each time and each place. It is a flowchart which shows the unit travel time calculation process in a prediction process. It is a timing chart which shows the timing of gain correction in prediction processing.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a schematic configuration of a traffic jam prediction system. The traffic jam prediction system includes a center 10 and a user vehicle 20. In addition, information from the weather information organization 30, the WEB system 40, and the probe car 50 is input to the center 10.

  The center 10 performs service based on the traffic jam prediction information for the user vehicle 20 by performing data communication with the user vehicle 20. For example, when the destination is set in the user vehicle 20, a so-called route search is performed based on the current location and the destination of the user vehicle 20 transmitted to the center 10. At this time, the route based on the traffic jam prediction information is used. Search is possible.

  The user vehicle 20 has a navigation device (“navigation device” in the figure) 21, and data communication with the center 10 is possible via the navigation device 21. Thereby, the navigation device 21 transmits the current location and destination of the vehicle to the center 10 when the destination is set. On the other hand, the searched route is transmitted from the center 10.

  The meteorological information organization 30 outputs the rainfall amount measured at a weather station in each region as rainfall information at intervals of several minutes (in this embodiment, at intervals of 5 minutes). For example, the average rainfall per hour is output at intervals of several minutes together with predetermined coordinates and time. In this embodiment, since it is a predetermined coordinate, it is rain information at a predetermined point. However, as another form, it is possible to output rain information of a certain area.

  The WEB system 40 is embodied as an SNS (social network service) or the like, and outputs rainfall information based on written information such as a bulletin board. In this case, it becomes the rain information in the poster's area on the bulletin board or the like. This regional information may be determined based on the registered information of the poster, or the regional information may be adopted if the regional information is included in the written information.

  The probe car 50 is a vehicle that periodically performs data communication with the center 10, and transmits rainfall information in addition to traffic information to the center 10. The rainfall information is transmitted to the center 10 as the number of times the wiper reciprocates per minute, for example. At this time, the coordinate value and time of the probe car 50 are also transmitted. On the other hand, the traffic information is transmitted to the center 10 as the link travel time at time t. Note that the user vehicle 20 may also serve as the probe car 50.

The center 10 described above includes a rainfall information collection unit 11, a traffic information collection unit 12, a control unit 13, and a communication unit 14. FIG. 2 is a logic diagram showing functions of the center 10.
The rain information collection unit 11 processes the rain information transmitted from the weather information organization 30, the WEB system 40, and the probe car 50 (see FIG. 1). Thereby, the rainfall information collection unit 11 outputs the rainfall amount M (t) of each link per hour.

The traffic information collection unit 12 processes traffic information transmitted from the probe car 50 (see FIG. 1). As a result, the traffic information collection unit 12 sends the link travel time T (t) at time t, the link travel time T (t + Δ) at time (t + Δ), and the time for traveling the unit distance at time t (hereinafter “unit travel”). RT (t) is output. The unit travel time is calculated by the following equation 1. Here, L is the driving distance of the link, N is the number of vehicles that have passed through the link, and ΣT (t) is the total link travel time for the number of vehicles that have passed.

rT (t) = ΣT (t) / L · N Equation 1

The control unit 13 predicts at the time (t + Δ) based on the rainfall amount M (t) from the rainfall information collection unit 11, the link travel time T (t) from the traffic information collection unit 12, and the unit travel time rT (t). The link travel time hT (t + Δ) is obtained. Then, the gain is corrected using T (t + Δ) actually output from the traffic information collecting unit 12 at time (t + Δ).

  When the function of the control unit 13 is represented by logic, a filter unit 61 and a predicted value output unit 62 are included as shown in FIG. It also has various functions such as a route search function (not shown).

  The filter unit 61 has a gain output block 63 and an error addition block 64. On the other hand, the predicted value output unit 62 includes an error amount calculation block 66 and a predicted value calculation block 65.

The gain output from the gain output block 63 is expressed by the following equations 2.1 and 2.2.

MD (t) = K1 · M (t) + K2 · ΔM (t) Equation 2.1

In Expression 2.1, weighting of the obtained M (t) and the change amount ΔM (t) for each sample time is performed using the weighting coefficients K1 and K2.

Gain (rT, MD, n) Equation 2.2

Note that n is a calculation cycle that is incremented each time the rainfall amount M (t) from the rainfall information collection unit 11 is updated. In the present embodiment, n is a cycle of 5 minutes.

  Then, the predicted value calculation block 65 multiplies the link travel time T (t) at time t by the gain output from the gain output block 63 to obtain the predicted link travel time hT (t + Δ) at a future time (t + Δ). ) Is calculated.

When the time (t + Δ) is actually reached, the link travel time T (t + Δ) is output from the traffic information collection unit 12, so that the error amount calculation block 66 calculates the error amount with this as a true value. It is as shown in the following expression 3.

error = hT (t + Δ) / T (t + Δ) −1 Equation 3

Then, the error amount calculated by the error amount calculation block 66 is added to the gain at time t by the error addition block 64. It is as shown in the following formula 4.

Gain (rT, M, n + 1) = Gain (rT, M, n) + error · K Equation 4
When the gain is visualized, it becomes as shown in FIG. That is, determining the gain determines the height with respect to the rainfall M (t) and the unit travel time rT (t), and the calculation cycle is 1, 2, 3,... N-1, n. , N + 1,..., The height of each part of the surface shown in FIG. In the case of the calculation cycle “1”, it is conceivable that the planes have heights “1” at various places.

  With the above function, the control unit 13 of the center 10 calculates the predicted link travel time hT (t + Δ) at the time t + Δ at the time t. Therefore, the center 10 searches for an optimum route based on the predicted link travel time hT (t + Δ) in the route search process based on the current location and the destination transmitted from the user vehicle 20.

Next, the prediction process in the center 10 of this embodiment is demonstrated. FIG. 4 is a flowchart showing the prediction process. This prediction process is repeatedly executed at the center 10.
In the first S100, a rainfall amount calculation process is executed. The rainfall amount calculation processing is realized as a function of the rainfall information collecting unit 11, and as shown in FIG. 5, the weather information is acquired in S101, the wiper information is acquired in S102, and the WEB information is acquired in S103. In S104, the rainfall is calculated.

  First, the process of S101-103 is demonstrated concretely. In S101, the rainfall per hour is acquired from the weather information organization 30 at intervals of 5 minutes. For example, the rainfall acquired at time t is the average rainfall over the last hour. This rainfall is obtained together with the coordinates of the observation point of the weather information agency.

  In S102, the number of reciprocations of the wiper per minute is acquired from the probe car 50. The number of times the wiper reciprocates is not only standardized “Lo” and “Hi” but can be arbitrarily set by the user. For example, as shown in FIG. 10 ”,“ 50 (Lo) ”,“ 72 (Hi) ”,“ 100 ”,“ 120 ”and so on. The number of reciprocations of the wiper is converted into rainfall by the correspondence table shown in FIG. In S102, the position coordinates of the probe car 50 are also acquired.

  In S <b> 103, the content of the poster's comment is acquired from the WEB system 40. Specifically, the content of remarks indicating the rainfall situation is acquired. For example, as shown in Fig. 6 (a), "Is it raining?" It is a condition of acquiring “such as” and “like a waterfall”. The content of this statement is converted into rainfall by the correspondence table shown in FIG. If the poster's position coordinates can be specified by GPS or the like, the position coordinates are also acquired. Further, if information indicating the poster's position is being posted, the position information is acquired. Furthermore, if information indicating the position of the poster is registered as member information or the like, the position information is acquired. At this time, if the position information is “Nagoya City” indicating an area, the position information is converted into the position coordinates of the central part of Nagoya.

  As described above, by the processing of S101 to S103, the amount of rainfall from the weather information organization 30, the amount of rainfall based on the number of wiper reciprocations from the probe car 50, and the amount of rainfall based on the remarks from the WEB system 40 are acquired. Each amount of rainfall is the amount of rainfall at a certain position coordinate at a certain time.

  However, the rainfall amount from the meteorological information organization 30 has a relatively high reliability, and the rainfall amount based on the content of the statement from the WEB system 40 has a relatively low reliability. Therefore, in the present embodiment, each rainfall amount is weighted by multiplying by the reliability coefficient Ks shown in FIG. Specifically, the rainfall amount from the weather information organization 30 is multiplied by “1.0”, the rainfall amount based on the number of wiper reciprocations of the probe car 50 is multiplied by “0.9”, The rainfall based on the content is multiplied by “0.1”.

  Next, the process of S104 in FIG. 5 will be specifically described. In S104, the rainfall is calculated. The rainfall calculated here is the rainfall at time t in the center coordinates (xc, yc) of each link.

However, the amount of rainfall obtained in S101 to S103 is at each observation point, and is at each observation time.
Therefore, as shown in FIG. 7A, the rainfall actually used, that is, the rainfall M (t) = M (xc, yc, t) at the time t in the link center coordinates (xc, yc). Ask. Note that the rainfall amount observed at coordinates (x1, y1) at time t1 (rainfall amount at observation time 1 at observation point 1) is U1 (x1, y1, t1), and coordinates (x2, y2) at time t2. Let U2 (x2, y2, t2) be the rainfall observed at (observation time 2 at the observation point 2).

  This is converted based on the distance L from the observation points 1 and 2 and the elapsed time ΔTK from the observation times 1 and 2. Specifically, as a map as shown in FIG. 7B, a weighting factor Kpt for the elapsed time ΔTK and a weighting factor KpL for the distance L are set. The weighting coefficient Kpt is obtained for each region by statistically processing the time until rain stops once it starts to fall. Further, the weighting coefficient KpL is obtained for each region by statistically processing the raining radius.

  If the map shown in FIG. 7B is used, M (t) is expressed by the following Expression 5.

  Details of Equation 5 will be described. The first calculation process (ΣKpt · Um (Xm, Ym, tm)) multiplies the rainfall amount at each observation time in one observation coordinate by a weighting coefficient Kpt corresponding to that time, and further accumulates it for a predetermined time. Thus, the first calculation processing result is obtained. The first calculation processing result is calculated for each observation coordinate. Then, the second calculation processing is performed by multiplying the first calculation processing result of each observation coordinate by the weighting coefficient KpL corresponding to the observation coordinate and the reliability coefficient Ks, and further accumulating a predetermined number of observation coordinates. The second calculation processing result is obtained (ΣKs · KpL · first calculation processing result). In the third calculation process, the result of the second calculation process is divided by the weighting coefficient Kpt, the weighting coefficient KpL, and the reliability coefficient Ks used in the first calculation process and the second calculation process. That is, it is a normalization process.

Thus, in S104, the rainfall amount M (t) at the time t of the center coordinates of the link is calculated by Equation 5.
Returning to the description of FIG. 4, in S110, a unit travel time calculation process is executed. This unit travel time calculation process is realized as a function of the traffic information collection unit 12. FIG. 8 is a flowchart showing the unit travel time calculation process.

In the first S111, based on information from the probe car 50, a link travel time T (t) that is a time required to travel on a certain link is acquired.
In the next S112, a unit travel time is calculated. In this process, rT (t) is obtained using Equation 1 above.

  In S120 in FIG. 4, the error amount is calculated. This process is realized as a function of the control unit 13. Specifically, the amount of error at the time (t + Δ) is obtained using Equation 3 above.

  In subsequent S130, the gain is corrected. This process is also realized as a function of the control unit 13. Specifically, the gain is corrected by using the above equation 4 and the error amount calculated in S120.

  In the next S140, the predicted link travel time as a predicted value is output. This process is also realized as a function of the control unit 13. Specifically, the predicted link travel time is calculated by multiplying the link travel time by the gain corrected in S130, and the predicted link travel time is output.

Thereby, in this embodiment, based on prediction link travel time, the control part 13 performs the service based on traffic congestion prediction information by performing route search.
That is, in this embodiment, as shown in FIG. 9, in the calculation cycle n at time t, the predicted link travel time hT (t) obtained in the calculation cycle (n−1) at time (t−Δ) and the time An error amount is obtained from the actual link travel time T (t) of t (S120 in FIG. 4). The gain is corrected by this error amount (S130), and the predicted link travel time hT (t + Δ) is output at time t (S140). Similarly, in the calculation cycle (n + 1) at time (t + Δ), from the predicted link travel time hT (t + Δ) obtained in calculation cycle n at time t and the actual link travel time T (t + Δ) at time (t + Δ). Then, an error amount is obtained (S120 in FIG. 4). The gain is corrected based on the error amount (S130).

  As described above, in this embodiment, the rainfall information collection unit 11 executes the rainfall information calculation process (S100 in FIG. 4), and calculates the rainfall amount at the link center coordinates (S104 in FIG. 5). ). The traffic information collection unit 12 executes a unit travel time calculation process (S110 in FIG. 4), and calculates a unit travel time (S112 in FIG. 8). The control unit 13 calculates an error amount from the ratio between the predicted link travel time that is the previous predicted value and the link travel time that is the true value (S120 in FIG. 4), and is set by the rainfall amount and the unit travel time. The gain is corrected by the error amount (S130), and the current predicted link travel time is output (S140). As a result, it is possible to predict the traffic situation that changes every moment depending on the amount of rainfall.

  In the present embodiment, the error amount is calculated from the ratio between the predicted link travel time that is the previous predicted value and the actual link travel time at the predicted time (S120 in FIG. 4), and the rainfall amount and unit travel time are calculated. The set gain is corrected with the error amount (S130). Thereby, since the gain is corrected based on the actual link travel time at the predicted time, a more appropriate predicted link travel time can be output.

At this time, since the error amount is calculated from the ratio between the predicted link travel time and the actual link travel time, the gain can be corrected relatively easily.
Furthermore, in this embodiment, the rainfall information collection part 11 performs the rainfall amount calculation process shown in FIG. 5, and acquires the rainfall amount as weather information from the weather information organization 30 (S101). Further, wiper information from the probe car 50 is acquired (S102). Furthermore, the content of the poster's remarks is acquired as WEB information from the WEB system 40 (S103). Thereby, the rainfall amount (S104) of the link calculated by the rainfall information collection unit 11 becomes appropriate.

  Here, the rainfall information collection unit 11 performs correction by weighting the rainfall amount for each rainfall information, and calculates the rainfall amount of the link (S104). Thereby, the rainfall amount of the link calculated by the rainfall information collection unit 11 becomes more appropriate.

  In addition, the rainfall information collection unit 11 performs correction based on the distance between the observation location of the rain information and the link and also performs correction based on the elapsed time from the observation time of the rain information to calculate the rainfall amount of the link (S104). . Thereby, the rainfall amount of the link calculated by the rainfall information collection unit 11 becomes more appropriate.

It should be noted that the center 10 in this embodiment constitutes a “congestion prediction device” in the scope of claims, the rainfall information collection unit 11 constitutes a “rainfall information collection unit”, and the traffic information collection unit 12 provides “traffic information collection unit”. The control unit 13 constitutes “gain setting means”, “predicted value output means”, and “gain correction means”.
Specifically, the gain output block 63 of the filter unit 61 constitutes a “gain setting unit”, the prediction value calculation block 65 of the prediction value output unit 62 constitutes a “prediction value output unit”, the error amount calculation block 66 and the filter The error amount addition block 64 of the unit 61 constitutes “gain correction means”. Further, the navigation device 21 constitutes a “navigation device”, and the user vehicle 20 and the center 10 on which the navigation device 21 is mounted constitutes a “congestion prediction system”.

  Further, the rainfall amount calculation process shown in FIG. 5 corresponds to the process as the function of the “rainfall information collecting means”, and the unit travel time calculation process shown in FIG. 8 is the process as the function of the “traffic information collection means”. 4, S120 and S130 in FIG. 4 correspond to processing as a function of “gain correction means”, and S140 corresponds to processing as a function of “predicted value output means”.

As described above, the present invention is not limited to the above-described embodiment, and can be implemented in various forms without departing from the technical scope thereof.
(A) In the above embodiment, the center 10 performs a route search using the predicted link travel time by transmitting the current location and the destination from the user vehicle 20 to the center 10. On the other hand, the center 10 may transmit the predicted link travel time as traffic jam information to the user vehicle 20, and the navigation device 21 may perform a route search using the predicted link travel time. Further, the present invention is not limited to the route search, and the traffic jam prediction display on the map based on the predicted link travel time may be performed. That is, the guidance may simply be to display the traffic jam prediction on the map.

  (B) In the above embodiment, the rainfall M (t) is based on the actual measurement value, but the rainfall is predicted by the same method as the link travel time, and the predicted rainfall (t + Δ ′) is used. The predicted link travel time (t + Δ) may be calculated. In this case, if Δ is 5 minutes, Δ ′ may be 5 minutes. At this time, in addition to or instead of the traffic jam prediction display based on the predicted link travel time, the rain prediction display may be performed on the map based on the predicted rainfall amount.

  (C) In the above embodiment, the average rainfall per hour is obtained every 5 minutes. On the other hand, the average rainfall per 5 minutes may be acquired every 5 minutes. Moreover, it is good also as a structure which acquires the rainfall amount of a certain area (area) instead of the rainfall amount of a certain point.

  DESCRIPTION OF SYMBOLS 10 ... Center, 11 ... Rainfall information collection part, 12 ... Traffic information collection part, 13 ... Control part, 14 ... Communication part, 20 ... User vehicle, 21 ... Navigation apparatus, 30 ... Weather information organization, 40 ... WEB system, 50 ... Probe car, 61 ... Filter section, 62 ... Predicted value output section, 63 ... Gain output block, 64 ... Error addition block, 65 ... Predicted value calculation block, 66 ... Error amount calculation block

Claims (9)

Rain information collecting means for collecting rainfall information and calculating a rainfall amount of the link based on the rain information;
Traffic information collecting means for collecting traffic information and outputting a link travel time that is a time required for a link to pass based on the traffic information;
Gain setting means for setting a gain based on the rainfall and the link travel time;
A predicted value output means for multiplying the link travel time by the gain set by the gain setting means to output a predicted link travel time at a future predicted time;
Congestion prediction device characterized by comprising. In the traffic jam prediction device according to claim 1,
Gain correction means for correcting the gain set by the gain setting means based on the predicted link travel time output by the predicted value output means and the actual link travel time at the predicted time. Congestion prediction device characterized by. In the traffic jam prediction device according to claim 2,
The traffic jam prediction device, wherein the gain correction means corrects the gain by adding an error amount based on a ratio between the predicted link travel time and the actual link travel time at the predicted time. In the traffic jam prediction device according to any one of claims 1 to 3,
The traffic information collecting means calculates a unit travel time required to move a unit distance on the link based on the link travel time together with the link travel time;
The traffic jam prediction device, wherein the gain setting means sets a gain using the rainfall amount and the unit travel time as parameters. In the traffic jam prediction device according to any one of claims 1 to 4,
The rain information collection means collects at least one of the rainfall information from a weather information agency, wiper operation information from a probe car, and the content of a poster's remarks from a WEB system as the rain information. A traffic jam prediction device. In the traffic jam prediction device according to any one of claims 1 to 5,
The traffic jam prediction device, wherein the rainfall information collection unit performs correction by weighting the rainfall amount for each of the rainfall information, and calculates the rainfall amount of the link. In the traffic jam prediction device according to any one of claims 1 to 6,
The traffic jam prediction device, wherein the rainfall information collection means performs correction based on a distance between an observation place of the rain information and a link, and calculates a rainfall amount of the link. In the traffic jam prediction device according to any one of claims 1 to 7,
The traffic jam prediction device, wherein the rainfall information collection means performs a correction based on an elapsed time from an observation time of the rain information and calculates a rainfall amount of the link. The traffic jam prediction device according to any one of claims 1 to 8,
A navigation device that performs data communication with the traffic jam prediction device and is capable of providing guidance based on the predicted link travel time;
Congestion prediction system characterized by having.