JP7679480B2 - Lane estimation device and lane estimation method - Google Patents

Description

The present invention relates to a lane estimation device and a lane estimation method for estimating the lane on which a vehicle is traveling.

A known example of this type of device is one that compares the road surface profile of each of a number of pre-registered lanes with a road surface profile measured while the vehicle is traveling, calculates the similarity of the road surface profiles for each of the multiple lanes, and estimates the lane with the highest similarity as the lane to be used when the vehicle is traveling (see, for example, Patent Document 1).

JP 2016-45063 A

However, the position of the tires in the width direction within the lane in which the vehicle is traveling may differ from vehicle to vehicle. For this reason, it is difficult to accurately estimate the lane when the vehicle is traveling by simply estimating the lane based on the similarity of the road surface profile, as in the device described in Patent Document 1.

A lane estimation device according to one aspect of the present invention includes a position information acquisition unit that acquires position information including positioning accuracy information obtained by a positioning sensor that receives a signal transmitted from a positioning satellite to determine the position of a vehicle, a driving information acquisition unit that acquires vehicle driving information including vehicle speed information and information on detection values of a detector that change according to a road surface profile on a road surface on which the vehicle is traveling, a reference information acquisition unit that acquires vehicle speed reference information that serves as a reference for a change in vehicle speed in each of a plurality of driving lanes that extend approximately parallel to each other, and a road map information acquisition unit that acquires road map information including road driving lane information and road surface profile information. and a driving lane identification unit that, when identifying a driving lane among a plurality of driving lanes that corresponds to a vehicle position determined by position information weighted by positioning accuracy information based on the position information acquired by the position information acquisition unit, the driving information acquired by the driving information acquisition unit, the vehicle speed reference information acquired by the reference information acquisition unit, and the road map information acquired by the road map information acquisition unit, identifies the driving lane based on the driving information and the vehicle speed reference information when there is no reference road surface profile in which the degree of match between the driving information and the road surface profile is equal to or greater than a predetermined value .

Another aspect of the present invention is a lane estimation method, which includes the steps of: acquiring position information including positioning accuracy information obtained by a positioning sensor that receives a signal transmitted from a positioning satellite to position the vehicle; acquiring vehicle driving information including vehicle speed information and information on detection values of a detector that change according to the road surface profile of the road surface on which the vehicle is driving; acquiring vehicle speed reference information that serves as a reference for vehicle speed changes in each of a plurality of driving lanes extending approximately parallel to each other; acquiring road map information including driving lane information and road surface profile information of the road ; and, when identifying a driving lane among a plurality of driving lanes that corresponds to the position of the vehicle determined by position information weighted by the positioning accuracy information based on the position information, the driving information, the vehicle speed reference information, and the road map information, identifying the driving lane based on the driving information and the vehicle speed reference information when there is no reference road surface profile in which the degree of match between the driving information and the road surface profile is equal to or greater than a predetermined value .

According to the present invention, it is possible to accurately estimate the lane in which a vehicle is traveling.

FIG. 1 is a diagram illustrating an example of a road having multiple lanes. 1 is a diagram showing an overall configuration of a lane estimation system including a lane estimation device according to an embodiment of the present invention. FIG. 3 is a diagram showing an example of a road surface profile obtained by the server device of FIG. 2 . FIG. 4 is a diagram showing an example of a vehicle speed change model when traveling in a non-congested traffic jam; FIG. 4 is a diagram showing an example of a model of a vehicle speed change when traveling in a traffic jam. 1 is a block diagram showing a functional configuration of a lane estimation device according to an embodiment of the present invention; 6 is a flowchart showing an example of a process executed by the controller of FIG. 5 .

Below, an embodiment of the present invention will be described with reference to Figures 1 to 6. A lane estimation device according to an embodiment of the present invention is configured to estimate the lane in which a vehicle is traveling when the vehicle is traveling on a road with multiple travel lanes (sometimes simply called lanes). Once the travel lane is estimated, it is possible to easily predict congestion for each lane, create a road surface profile that shows the unevenness of the road surface, estimate the location where a broken-down vehicle is stopped, and estimate vehicles traveling the wrong way.

Figure 1 is a schematic diagram showing an example of a road to which a lane estimation device according to an embodiment of the present invention is applied. In Figure 1, a road group RD is shown, which includes an elevated expressway RD1 installed on a bridge pier P and a general road RD2 installed on the ground along the bridge pier P. As shown in Figure 1, the road group RD is installed in a city of buildings with tall buildings BL. The road group RD has a plurality of driving lanes that extend parallel to each other, for example, lanes LN1 and LN2 on the expressway RD1 and lane LN3 on the general road RD2. Note that extending parallel to each other does not mean parallel in the strict sense, but means that they extend in the same direction or almost in the same direction (approximately parallel), and also includes cases where the roads are at different heights and overlap in a planar view.

In this embodiment, it is estimated which of the multiple driving lanes LN1 to LN3 shown in Figure 1, i.e., the multiple lanes LN1 to LN3 that are close to each other in a planar view, the vehicle 1 is traveling in. In the example of Figure 1, the vehicle (target vehicle) 1a that is the subject of the estimation of the driving lane (target vehicle) is traveling in lane LN3, and a vehicle 1b other than the target vehicle 1a is traveling in lanes LN1 and LN2, respectively, in the directions of the arrows.

The driving lane can be estimated by measuring the vehicle's position by receiving signals from positioning satellites (positioning satellites) such as the Global Positioning System (GPS) using a positioning sensor such as a GPS receiver (GPS sensor) mounted on the vehicle, and then comparing the measured vehicle position with the lane positions included in the map information. In other words, if the positioning accuracy when measuring the vehicle's position using a positioning sensor is high enough to identify the lane positions, the lane positions can be estimated using the positioning sensor.

However, in urban areas with many high-rise buildings or inside tunnels, the accuracy of positioning decreases, making it difficult to accurately estimate the lane position using a positioning sensor. On the other hand, the position of the lane can be identified by detecting a road surface profile indicating the characteristics of the road surface, such as unevenness, while the vehicle is traveling, and comparing the detected road surface profile with a pre-stored road surface profile for each lane. However, the position of the tires in the width direction within the lane in which the vehicle is traveling may differ from vehicle to vehicle. For this reason, it is difficult to accurately estimate the lane position just by using the road surface profile. Therefore, in this embodiment, a lane estimation device is configured as follows.

Figure 2 is a diagram showing the overall configuration of a lane estimation system including a lane estimation device according to an embodiment of the present invention. As shown in Figure 2, the lane estimation system has an on-board device 100 mounted on a vehicle 1, and a server device 3 capable of communicating with the on-board device 100 via a network 200. The vehicle 1 is, for example, a manually driven vehicle that is manually driven by a driver.

The in-vehicle device 100 has a positioning sensor 10 that receives positioning signals transmitted from a positioning satellite 2, and a communication unit 11 that communicates with a server device 3 via a network 200. The positioning satellite 2 is an artificial satellite such as a GPS satellite or a quasi-zenith satellite, and the current position (latitude, longitude, altitude) of the vehicle 1 can be calculated using the positioning information received by the positioning sensor 10 from the positioning satellite 2.

The network 200 includes not only public wireless communication networks such as the Internet and mobile phone networks, but also closed communication networks established for each specified management area, such as wireless LAN, Wi-Fi (registered trademark), Bluetooth (registered trademark), etc. The server device 3 is configured, for example, as a single server, or as a distributed server consisting of separate servers for each function. The server device 3 can also be configured as a distributed virtual server created in a cloud environment called a cloud server.

The server device 3 is configured to include a processing device having a CPU, ROM, RAM, and other peripheral circuits. The server device 3 has, as its functional configuration, a communication unit 31, a memory unit 32, a road surface profile generation unit 33, and a vehicle speed change model generation unit 34.

The communication unit 31 is configured to be capable of wireless communication with the in-vehicle device 100 via the network 200, and acquires the position information of the vehicle 1 and the driving information of the vehicle 1 via the communication unit 11 of the vehicle 1. The position information is information indicating the current position of the vehicle 1 calculated from a signal received by the positioning sensor 10 of the vehicle 1. The driving information is information indicating the driving state of the vehicle 1 acquired by various sensors mounted on the vehicle 1. The driving information includes vehicle speed information of the vehicle 1 and information on the detection value by an acceleration sensor (lateral acceleration sensor) that detects the acceleration (lateral acceleration) in the left-right direction of the vehicle 1. The communication unit 31 constantly acquires position information and driving information not only of the target vehicle 1a (Figure 1) that is the target of the driving lane estimation, but also of multiple vehicles 1b (Figure 1) other than the target vehicle 1a.

The memory unit 32 stores road map information. The road map information includes road position information, road shape (curvature, etc.) information, road gradient information, intersection and branch point position information, number of lanes information, lane width, and position information for each lane. Position information for each lane includes information on the center position of the lane and the boundaries of lane positions.

The road surface profile generating unit 33 generates a road surface profile indicating the road surface characteristics based on the position information and driving information of multiple vehicles 1b other than the target vehicle 1a acquired via the communication unit 31. FIG. 3 is a diagram showing an example of a road surface profile. The horizontal axis in the figure is the position of the vehicle 1 in the traveling direction along the traveling lane, that is, the distance, and the vertical axis is the amount of unevenness (depth or height) of the road surface, that is, the road surface roughness. In general, the greater the amount of unevenness of the road surface, the greater the lateral acceleration of the vehicle 1. Therefore, there is a predetermined correlation between the road surface characteristics and the lateral acceleration. This predetermined correlation is stored in the storage unit 32 in advance. The road surface profile generating unit 33 uses this predetermined correlation to calculate the amount of unevenness of the road surface corresponding to the vehicle position on the road from the lateral acceleration, and generates a road surface profile in the traveling direction of the vehicle 1 as shown in FIG. 3. The information of the generated road surface profile at each position on the road is stored in the storage unit 32 as part of the road map information.

When different vehicles 1 are traveling on the same lane, the positions of the tires on the road surface may differ, which may result in different road surface profiles detected by the lateral acceleration sensors of each vehicle 1. In this case, the road surface profile generator 33 may, for example, average the respective road surface profiles detected by the lateral acceleration sensors of each vehicle 1 to generate a representative road surface profile for each road surface.

The road surface profile generator 33 can also generate a road surface profile from data obtained by running a dedicated vehicle for measuring road surface properties. For example, a road surface profile can be generated without using a lateral acceleration sensor by running a dedicated vehicle equipped with a laser profiler and acquiring the measurement data at that time together with the position data of the dedicated vehicle.

Of the road map information stored in the memory unit 32, the road surface profile information is updated each time a road surface profile is generated by the road surface profile generating unit 33. Other road map information is updated at a predetermined interval or at any timing. In this embodiment, when estimating the driving lane of the vehicle 1, the road surface profile (reference road surface profile) of each of the lanes LN1 to LN3 at the driving position of the vehicle 1 is treated as already being stored in the memory unit 32.

The vehicle speed change model generation unit 34 generates a vehicle speed change model that is a reference for the vehicle speed change of the vehicle 1 traveling on each lane LN1 to LN3 based on the road map information stored in the memory unit 32 and the vehicle data acquired via the communication unit 31. The vehicle data is data unique to each vehicle 1, including the position information and vehicle speed information of the vehicle 1. The position information includes the number of satellites captured by the positioning sensor 10, the signal strength received by the positioning sensor 10, and the accuracy information of the positioning by the positioning sensor 10. The accuracy information of the positioning is, for example, information on the dilution of precision (DOP). The vehicle speed change model generation unit 34 can construct a vehicle speed change model by machine learning using the vehicle data for each lane obtained via the communication unit 31. The vehicle speed change model may be constructed by adding information on the road surface profile for each lane to the vehicle data for each lane.

4A and 4B are diagrams showing examples of vehicle speed change models generated by the vehicle speed change model generation unit 34. FIG. 4A is a vehicle speed change model when there is no traffic jam (when driving smoothly), and FIG. 4B is a vehicle speed change model when there is traffic jam. In FIG. 4A and FIG. 4B, the vehicle speed change model that is the reference for the vehicle speed of the vehicle 1 traveling on each lane is shown as a characteristic with the horizontal axis representing the position (distance) in the traveling direction of the vehicle 1 and the vertical axis representing the vehicle speed. In the figure, characteristics f1 and f3 are vehicle speed change models of the expressway RD1 (lane LN1 or LN2), and characteristics f2 and f4 are vehicle speed changes of the general road RD2 (lane LN3). These characteristics are obtained by machine learning using a large number of vehicle data for each lane, and correspond to characteristics obtained by averaging the vehicle data. In addition, the vehicle speed change model may be obtained by statistically processing the vehicle data.

As shown in FIG. 4A, when there is no traffic jam, the vehicle speed on the expressway is faster than the vehicle speed on the general road. Also, the change in vehicle speed on the expressway is smaller than the change in vehicle speed on the general road. Note that position S represents the position of an intersection on a general road. At intersections, vehicles stop frequently, so the vehicle speed is lower than at other intersections. On the other hand, as shown in FIG. 4B, when there is traffic jam, the vehicle speed on the expressway is faster than the vehicle speed on the general road, as in FIG. 4A, but the change in vehicle speed is larger than when there is no traffic jam. In this way, the vehicle speed change model on the expressway (characteristics f1, f3) is significantly different from the vehicle speed change model on the general road (characteristics f2, f4). The vehicle speed change model generated by the vehicle speed change model generation unit 34 is stored in the memory unit 32. This vehicle speed change model is updated every time the vehicle speed change model generation unit 34 generates a vehicle speed change model. In this embodiment, when estimating the driving lane of the vehicle 1, it is assumed that the vehicle speed change models for each of the lanes LN1 to LN3 at the driving position of the vehicle 1 are already stored in the storage unit 32.

Figure 5 is a block diagram showing the functional configuration of a lane estimation device 101 according to this embodiment. The lane estimation device 101 constitutes part of the in-vehicle device 100 of Figure 2. As shown in Figure 5, the lane estimation device 101 includes a positioning sensor 10, a communication unit 11, a sensor group 13, a switch group 14, and a controller 20. The positioning sensor 10, the communication unit 11, the sensor group 13, and the switch group 14 are each communicatively connected to the controller 20.

The sensor group 13 is a collective term for a plurality of sensors that detect the driving state of the vehicle 1. The sensor group 13 includes a lateral acceleration sensor 131 that detects the left-right acceleration of the vehicle 1, a vehicle speed sensor 132 that detects the vehicle speed, and a steering angle sensor 133 that detects the steering angle of the steering wheel. The switch group 14 is a collective term for a plurality of switches that detect the driving state of the vehicle 1. The switch group 14 includes a turn signal switch 141 that detects the operation of a turn signal by the driver. Note that a turn signal is a device that indicates the direction to the surroundings when the vehicle 1 turns right or left or changes course, and is composed of a turn signal lever or the like.

The controller 20 is an electronic control unit including a computer having a calculation unit such as a CPU, a storage unit such as a ROM and a RAM, and other peripheral circuits. The calculation unit of the controller 20 has, as functional components, an information acquisition unit 21 and a driving lane identification unit 25. The information acquisition unit 21 has a position information acquisition unit 211, a driving information acquisition unit 212, a reference information acquisition unit 213, and a road map information acquisition unit 214. The storage unit of the controller 20, like the storage unit 32 of the server device 3, stores a predetermined correlation between the road surface characteristics and lateral acceleration used when generating a road surface profile, threshold values for various judgments, and the like.

The position information acquisition unit 211 acquires the current position information of the vehicle 1 detected by the positioning sensor 10. The driving information acquisition unit 212 acquires driving information of the vehicle 1 including various detection values detected by the sensor group 13 and the switch group 14. The reference information acquisition unit 213 acquires a vehicle speed change model indicating reference information of the vehicle speed from the server device 3 via the communication unit 11. More specifically, the reference information acquisition unit 213 acquires a vehicle speed change model (FIGS. 4A and 4B) corresponding to non-congested and congested conditions for each of the multiple driving lanes LN1 to LN3 of the road at the current position of the vehicle 1 detected by the positioning sensor 10. The road map information acquisition unit 214 acquires road map information from the server device 3 via the communication unit 11. More specifically, the road map information acquisition unit 214 acquires road map information including lane information (driving lane information) of the road at the current position of the vehicle 1 detected by the positioning sensor 10 and information on the road surface profile of each lane LN1 to LN3.

The driving lane identification unit 25 identifies the driving lane among the multiple driving lanes LN1 to LN3 that corresponds to the current position of the vehicle 1 acquired by the position information acquisition unit 211 based on the driving information of the vehicle 1 acquired by the driving information acquisition unit 212, the vehicle speed change model acquired by the reference information acquisition unit 213, and the road map information of the road on which the vehicle 1 is traveling acquired by the road map information acquisition unit 214.

Specifically, first, the driving lane is identified based on the detection value of the lateral acceleration sensor 131 and the information on the road surface profile included in the road map information. More specifically, the amount of road surface unevenness is calculated from the lateral acceleration detected by the lateral acceleration sensor 131 using the correlation between the road surface properties and the lateral acceleration stored in advance. If the vehicle 1 is undergoing lateral acceleration while turning, the amount of road surface unevenness is corrected and the amount of road surface unevenness is calculated from the detection value of the lateral acceleration sensor 131. Then, the road surface profile that represents the change in the amount of road surface unevenness along the traveling direction of the vehicle 1, i.e., the measured road surface profile that is the actual measured value of the road surface profile, is compared with the road surface profile for each lane included in the road map information, i.e., the reference road surface profile, and the degree of agreement between the measured road surface profile and the reference road surface profile for each lane is calculated. Then, it is determined whether the degree of agreement is equal to or greater than a predetermined value, and when there is a reference road surface profile that is determined to be equal to or greater than the predetermined value, the lane having the reference road surface profile is identified as the driving lane currently being traveled.

In this case, the driving lane identification unit 25 may calculate the degree of match of the road surface profile over a predetermined distance, average the degree of match within the predetermined distance, and determine whether the degree of match is a predetermined value or more. When there are multiple reference road surface profiles with a degree of match equal to or greater than a predetermined value, the lane having the reference road surface profile with the highest degree of match may be identified as the driving lane currently being traveled. The degree of match can be calculated using a correlation coefficient, etc. The degree of match is also sometimes called similarity.

When there is no reference road surface profile with a degree of match of the road surface profile equal to or greater than a predetermined value, the driving lane identification unit 25 identifies the driving lane based on the detection value (vehicle speed detection value) of the vehicle speed sensor 132 and the vehicle speed change model. Specifically, first, based on the degree of change in the vehicle speed detection value, it is determined whether or not the road on which the vehicle 1 is traveling is congested. If it is determined that the road is not congested, it calculates the degree of match between the vehicle speed detection value and the vehicle speed change model for each lane (Figure 4A). Then, it is determined whether or not the degree of match is equal to or greater than a predetermined value, and if there is a vehicle speed change model determined to be equal to or greater than the predetermined value, it identifies the lane corresponding to the vehicle speed change model as the driving lane currently traveling. On the other hand, if it is determined that there is a traffic jam, it calculates the degree of match between the vehicle speed detection value and the vehicle speed change model for each lane (Figure 4B), and if there is a vehicle speed change model with a degree of match equal to or greater than a predetermined value, it identifies the lane corresponding to the vehicle speed change model as the driving lane currently traveling.

In this case, the driving lane identification unit 25 may calculate the degree of agreement over a predetermined distance, average the degrees of agreement within the predetermined distance, and determine whether the degree of agreement is a predetermined value or more. When there are multiple vehicle speed change models with a degree of agreement equal to or greater than a predetermined value, the lane corresponding to the vehicle speed change model with the highest degree of agreement may be identified as the driving lane currently being traveled. The degree of agreement may be calculated using a correlation coefficient, etc. The degree of agreement is also sometimes called similarity.

When identifying the driving lane, the driving lane identification unit 25 may estimate the driving lane based on the position information of the vehicle 1 obtained by the positioning sensor 10, and then compare the measured road surface profile with the reference road surface profile, or compare the vehicle speed detection value with the vehicle speed change model to determine whether the estimation of the driving lane based on the position information is correct, thereby identifying the driving lane. When estimating the driving lane using the detection value of the positioning sensor 10, the weighting of the estimation result based on the detection value of the positioning sensor 10 may be changed depending on the accuracy of the positioning (e.g., the value of DOP). For example, when the accuracy of the positioning is equal to or higher than a predetermined value, the weighting of the lane estimation based on the detection value of the positioning sensor 10 may be greater than the lane estimation based on the degree of agreement of the road surface profile and the lane estimation based on the degree of agreement of the vehicle speed.

When the degree of match of the road surface profile is not equal to or greater than a predetermined value and the degree of match of the vehicle speed is not equal to or greater than a predetermined value, the driving lane identification unit 25 determines whether or not a lane change has occurred based on the signal from the turn signal switch 141. That is, since a turn signal is generally operated when changing lanes, the driving lane identification unit 25 determines whether the vehicle 1 has changed lanes to the left or right based on the signal from the turn signal switch 141. For example, when the vehicle 1 is traveling in the center (second lane) of a three-lane road (first lane, second lane, third lane) on one side, the turn signal switch 141 detects the operation of the turn signal when the vehicle 1 changes lanes to the right or left lane, for example, by using separate switches. This makes it easy to determine whether the vehicle 1 has changed lanes to the left or right lane based on the signal from the turn signal switch 141.

The driving lane identification unit can also determine whether or not a lane change has occurred based on a signal from the steering angle sensor 133, instead of or in addition to the turn signal switch 141. That is, since the steering wheel is operated when changing lanes, the presence or absence of a lane change can be determined by determining whether or not the detection value of the steering angle sensor 133 is equal to or greater than a predetermined value. The turn signal is not always operated when changing lanes, or there is a risk that the turn signal may be operated erroneously. For this reason, by using the detection value of the steering angle sensor 133, the presence or absence of a lane change can be determined more accurately than when the turn signal switch 141 is used.

Figure 6 is a flowchart showing an example of processing executed by the controller 20 (CPU) according to a predetermined program. The processing shown in this flowchart is executed when the vehicle 1 is traveling in one of a number of travel lanes that extend approximately parallel to one another. In other words, it is executed when it is necessary to estimate the travel lane, and is repeated at a predetermined interval.

First, in step S1, current position information of the vehicle 1 detected by the positioning sensor 10, driving information of the vehicle 1 based on signals from the sensor group 13 and the switch group 14, road map information of the road on which the vehicle is traveling obtained via the communication unit 11, and a vehicle speed change model of the road on which the vehicle is traveling obtained via the communication unit 11 are obtained.

Next, in step S2, the measured road surface profile is obtained based on the detection value of the lateral acceleration sensor 131, and the degree of match (road surface profile match) between the measured road surface profile and the reference road surface profile for each lane included in the road map information is calculated. Next, in step S3, it is determined whether or not there is a reference road surface profile whose degree of match with the measured road surface profile is equal to or greater than a predetermined value, that is, whether or not there is a lane whose road surface profile match is equal to or greater than a predetermined value. If the result in step S3 is positive, the process proceeds to step S4, and if the result is negative, the process proceeds to step S5. In step S4, the lane having the reference road surface profile whose degree of match is equal to or greater than a predetermined value is estimated as the lane in which the vehicle 1 is traveling, and the process ends. The estimated lane is temporarily stored in the memory unit of the controller 20.

In step S5, the degree of match (vehicle speed match) between the detection value of the vehicle speed sensor 132 and the vehicle speed change model for each lane is calculated. Next, in step S6, it is determined whether or not there is a vehicle speed change model whose degree of match with the vehicle speed detection value is equal to or greater than a predetermined value, that is, whether or not there is a lane whose vehicle speed match is equal to or greater than a predetermined value. If the result in step S6 is positive, the process proceeds to step S7, and if the result is negative, the process proceeds to step S8. In step S7, the lane corresponding to the vehicle speed change model whose degree of match is equal to or greater than a predetermined value is estimated as the lane in which the vehicle 1 is traveling, and the process ends. The estimated lane is temporarily stored in the memory unit of the controller 20.

In step S8, it is determined whether the turn signal has been operated based on the signal from the turn signal switch 141. If the result of step S8 is positive, the process proceeds to step S9, and if the result is negative, the process proceeds to step S10. In this case, in step S9, the driving lane is estimated based on the signal from the turn signal switch 141 as a lane change from the lane estimated in step S4 or step S7, and the process ends. In step S10, it is determined whether the steering wheel has been operated by a predetermined amount or more based on the signal from the steering angle sensor 133. If the result of step S10 is positive, the process proceeds to step S9, and if the result is negative, the process proceeds to step S11. In this case, in step S9, the driving lane is estimated based on the signal from the steering angle sensor 133 as a lane change from the lane estimated in step S4 or step S7, and the process ends. In step S11, the driving lane is estimated based on the lane estimated in step S4 or step S7 as a lane change, and the process ends.

When estimating the driving lane based on the vehicle speed detection value (steps S5 to S7 in FIG. 6), an algorithm such as a support vector machine may be used to classify the vehicle speed change model into two groups, such as expressways and general roads. A model such as a recurrent neural network (RNN) or a long short-term memory (LSTM) may be used to estimate the lane from the vehicle speed detection value. In FIG. 6, the process of determining the road surface profile match is performed before the process of determining the vehicle speed match, but this is not limited to this, and the process of determining the vehicle speed match may be performed first, or the process of determining the vehicle speed match may be performed with priority over the process of determining the road surface profile match. The vehicle speed match may be weighted more heavily than the road surface profile match to estimate the driving lane.

The operation of the lane estimation device 101 according to this embodiment can be summarized as follows. As shown in FIG. 1, focus is placed on a target vehicle 1a in an area where multiple lanes LN1 to LN3 are present, extending parallel to one another and surrounded by buildings. This target vehicle 1a is traveling on a general road RD2, and therefore has a slower vehicle speed than vehicle 1b traveling on the expressway RD1. For this reason, when the vehicle speed (vehicle speed detection value) of the target vehicle 1a detected by the vehicle speed sensor 132 is compared with the vehicle speed change model of each of the lanes LN1 to LN3 obtained from the server device 3, the degree of agreement between the vehicle speed detection value and the vehicle speed change model of lane LN3 is the highest. This makes it possible to estimate that the target vehicle 1a is traveling on the travel lane LN3 (step S7).

By estimating the driving lane based on the vehicle speed detection value in this way, the driving lane can be accurately estimated even in a situation where the positioning accuracy is reduced due to being surrounded by buildings, etc. Furthermore, in this embodiment, if the degree of match between the measured road surface profile based on the signal from the lateral acceleration sensor 131 and the reference road surface profile obtained from the server device 3 is equal to or greater than a predetermined value, the lane having the reference road surface profile (e.g., lane LN3) is estimated as the driving lane (step S4). Therefore, even if the degree of match between the vehicle speed detection value and the vehicle speed change model is less than a predetermined value, the driving lane can be accurately estimated based on the degree of match of the road surface profile.

When the vehicle 1b in FIG. 1 is a target vehicle and the target vehicle 1b is traveling on the highway RD1 having multiple lanes LN1 and LN2, the turn signal is operated and the steering wheel is operated when changing lanes. In this case, it is determined that the target vehicle 1b has changed lanes based on the signal from the turn signal switch 141 or the steering angle sensor 133 (step S9). This makes it possible to determine whether or not the target vehicle 1b has changed lanes, and to accurately estimate the traveling lane using the lane change determination result. For example, if it is estimated that the target vehicle 1b is traveling on the traveling lane LN1 based on the vehicle speed detection value, and then it becomes difficult to estimate the lane based on the vehicle speed detection value (if the vehicle speed agreement falls below a predetermined value), it is determined that the target vehicle 1b has changed lanes and it is estimated that the target vehicle 1b is traveling on the lane LN2.

According to this embodiment, the following advantageous effects can be obtained.
(1) The lane estimation device 101 includes a position information acquisition unit 211 that acquires position information of the current position of the vehicle 1 obtained by a positioning sensor 10 that receives a signal transmitted from a positioning satellite 2 to measure the position of the vehicle 1, a driving information acquisition unit 212 that acquires driving information of the vehicle 1 including vehicle speed information of the vehicle 1 and information on the detection value of a lateral acceleration sensor 131 that changes depending on the road surface profile of the road surface on which the vehicle 1 is traveling, and a reference information acquisition unit 213 that acquires a vehicle speed change model that serves as a reference for vehicle speed changes in each of a plurality of driving lanes LN1 to LN3 that extend approximately parallel to each other. The vehicle speed change model acquired by the reference information acquisition unit 213 and the road map information acquired by the road surface profile acquisition unit 214 are provided, and a driving lane identification unit 25 is provided for identifying a driving lane among a plurality of driving lanes LN1 to LN3 that corresponds to the position of the vehicle 1 determined by the position information acquired by the position information acquisition unit 211, based on the driving information acquired by the driving information acquisition unit 212, the vehicle speed change model acquired by the reference information acquisition unit 213, and the road map information acquired by the road map information acquisition unit 214 (FIG. 5).

In this embodiment, the driving lane is estimated using road surface profile information and vehicle speed information, so that the driving lane can be accurately estimated even in situations where the accuracy of positioning by the positioning sensor 10 decreases, such as in areas with tall buildings or inside tunnels. That is, for example, in areas where expressways and general roads run approximately parallel to each other, multiple driving lanes LN1 to LN3 with different vehicle speed change characteristics (vehicle speed change models) extend, so the driving lane can be accurately estimated by using the vehicle speed information. In contrast, the position of the tires in the width direction within the lane in which the vehicle 1 is traveling may differ from vehicle to vehicle, so it is difficult to accurately estimate the driving lane based only on the road surface profile.

(2) A vehicle speed change model corresponding to each of the multiple driving lanes LN1 to LN3 is constructed by machine learning. This makes it possible to use a vehicle speed change model that accurately reflects the characteristics of vehicle speed change for each lane, enabling good lane estimation.

(3) The vehicle speed change model includes a vehicle speed change model when a traffic jam occurs and a vehicle speed change model when no traffic jam occurs (Figures 4A and 4B). The characteristics of vehicle speed change differ greatly between when there is a traffic jam and when there is no traffic jam, but the vehicle speed change model is prepared taking this into consideration, so that lane estimation can be performed more accurately.

(4) The driving information acquired by the driving information acquisition unit 212 further includes turn signal operation information. The driving lane identification unit 25 further identifies the driving lane based on the turn signal operation information (signal from the turn signal switch 141) (Figure 5). This allows for good determination of whether or not a lane change has occurred, and allows for more accurate lane estimation.

(5) The driving information acquired by the driving information acquisition unit 212 further includes steering angle information indicating the steering angle of the steering wheel. The driving lane identification unit 25 further identifies the driving lane based on the steering angle information of the steering wheel (the signal from the steering angle sensor 133) (Figure 5). This makes it possible to accurately determine whether or not a lane change is occurring, and further improves the accuracy of lane estimation.

(6) The lane estimation device 101 of this embodiment can also be used as a lane estimation method. The lane estimation method includes a step (step S1) of acquiring the position information of the current position of the vehicle 1 obtained by the positioning sensor 10 that receives a signal transmitted from the positioning satellite 2 to position the vehicle 1, the travel information of the vehicle 1 including the vehicle speed information of the vehicle 1 and the information of the detection value of the detector that changes according to the road surface profile of the road surface on which the vehicle 1 travels, the vehicle speed reference information that is the reference for the change in vehicle speed in each of the multiple travel lanes LN1 to LN3 that extend approximately parallel to each other, and the road map information including the lane information and the road surface profile information of the road, and a step (steps S2 to S7) of identifying the travel lane corresponding to the position of the vehicle 1 determined by the position information among the multiple travel lanes LN1 to LN3 based on the acquired travel information, the vehicle speed reference information, and the road map information (FIG. 6). This allows the lane on which the vehicle 1 is traveling to be accurately estimated.

In the above embodiment, the vehicle position is determined by the positioning sensor 10 by receiving a signal transmitted from a positioning satellite, but the vehicle position may be determined based on an inertial navigation method as well as this satellite positioning method. In the above embodiment, the driving information acquisition unit 212 acquires driving information of the vehicle 1 including information on the detection value (sensor value) of the lateral acceleration sensor 131, but driving information including information on the detection value of another detector that changes according to the road surface profile may be acquired. For example, the driving information acquisition unit may acquire driving information including information on the detection value of a sensor that detects a roll angle or roll rate, or information on the detection value of a sensor that detects vertical vehicle vibrations, together with vehicle speed information.

In the above embodiment, the reference information acquisition unit 213 acquires a vehicle speed change model (FIGS. 4A and 4B) as vehicle speed reference information that serves as a reference for vehicle speed changes in each of the multiple driving lanes LN1 to LN3 that extend approximately parallel to each other in the road group RD including the elevated expressway RD1 and the general road RD2, but the vehicle speed reference information is not limited to the above. Different vehicle speed reference information may be acquired according to the time of day, such as daytime or nighttime. Therefore, the vehicle speed change model is not limited to the vehicle speed change model when congestion occurs and the vehicle speed change model when congestion does not occur. The vehicle speed change model may be constructed by a method other than machine learning (e.g., statistical processing). In the above embodiment, the road group RD having multiple driving lanes that extend approximately parallel to each other is composed of multiple roads RD1 and RD2 with different heights, but this is not limited to this, and may be composed of multiple roads with the same height. For example, when a general road extends to the side of the expressway, the road group may be composed of this expressway and the general road.

In the above embodiment, the presence or absence of a lane change is determined based on turn signal operation information and steering wheel operation information, but this determination may also be made based on signals from other sensors or switches. In the above embodiment, the lane estimation device 101 is mounted on the vehicle 1, but some or all of the functions of the lane estimation device 101 may be provided in the server device 3. In the above embodiment, the lane estimation device 101 is applied to a manually driven vehicle, but the lane estimation device 101 of the present invention can also be applied to an autonomous vehicle.

The above description is merely an example, and the present invention is not limited to the above-mentioned embodiment and modified examples as long as the characteristics of the present invention are not impaired. It is also possible to arbitrarily combine one or more of the above-mentioned embodiment and modified examples, and it is also possible to combine modified examples together.

1 Vehicle, 10 Positioning sensor, 11 Communication unit, 20 Controller, 21 Information acquisition unit, 25 Travel lane identification unit, 101 Lane estimation device, 131 Lateral acceleration sensor, 132 Vehicle speed sensor, 133 Steering angle sensor, 141 Turn signal switch, 211 Position information acquisition unit, 212 Travel information acquisition unit, 213 Reference information acquisition unit, 214 Road map information acquisition unit

Claims (6)

a position information acquisition unit that acquires position information including positioning accuracy information obtained by a positioning sensor that receives a signal transmitted from a positioning satellite and measures the position of the vehicle;
a travel information acquisition unit that acquires travel information of the vehicle, the travel information including vehicle speed information of the vehicle and information of a detection value of a detector that changes according to a road surface profile on which the vehicle is traveling;
a reference information acquisition unit that acquires vehicle speed reference information that is a reference for a change in vehicle speed in each of a plurality of travel lanes that extend substantially parallel to one another;
a road map information acquisition unit that acquires road map information including road lane information and road surface profile information;
a lane identification unit that, when identifying a driving lane among the multiple driving lanes that corresponds to the vehicle's position determined by the position information weighted by the positioning accuracy information based on the position information acquired by the position information acquisition unit, the driving information acquired by the driving information acquisition unit, the vehicle speed reference information acquired by the reference information acquisition unit, and the road map information acquired by the road map information acquisition unit, identifies the driving lane based on the driving information and the vehicle speed reference information when there is no reference road surface profile in which the degree of match between the driving information and the road surface profile is equal to or greater than a predetermined value. The lane estimation device according to claim 1,
The lane estimation device, wherein the vehicle speed reference information includes a vehicle speed change model corresponding to each of the plurality of driving lanes constructed by machine learning. The lane estimation device according to claim 2,
The lane estimation device, wherein the vehicle speed change model includes a vehicle speed change model when a traffic jam occurs and a vehicle speed change model when no traffic jam occurs. The lane estimation device according to any one of claims 1 to 3,
The driving information acquired by the driving information acquisition unit further includes operation information of a turn signal light,
The lane estimation device is characterized in that the driving lane identification unit further identifies the driving lane based on operation information of the turn signal. The lane estimation device according to any one of claims 1 to 4,
The driving information acquired by the driving information acquisition unit further includes steering angle information indicating a steering angle of a steering wheel,
The lane estimation device is characterized in that the driving lane identification unit further identifies the driving lane based on steering angle information of the steering wheel. acquiring position information including positioning accuracy information obtained by a positioning sensor that receives a signal transmitted from a positioning satellite and measures the position of the vehicle;
acquiring travel information of the vehicle, the travel information including vehicle speed information of the vehicle and information of a detection value of a detector that changes according to a road surface profile on which the vehicle is traveling;
acquiring vehicle speed reference information serving as a reference for a change in vehicle speed in each of a plurality of travel lanes extending substantially parallel to one another;
obtaining road map information including road lane information and road surface profile information;
a step of identifying a driving lane among the multiple driving lanes that corresponds to the vehicle's position determined by the position information weighted by the positioning accuracy information based on the position information , the driving information , the vehicle speed reference information, and the road map information, when there is no reference road surface profile in which the degree of match between the driving information and the road surface profile is equal to or greater than a predetermined value, by executing the steps of:

Images