DRIVING ASSISTANCE DEVICE AND DRIVING ASSISTANCE SYSTEM

Abstract

Driving assistance device (1), comprising: a recording unit (12) that records a plurality of previous driving positions (12a, 12a') of a vehicle and position estimation information (12b) in relation to the previous driving positions (12a, 12a'); a relative position estimation unit (11) that estimates a relative position of the vehicle on a lane based on measurement data from a sensor (10) that detects an environment of the vehicle and adds the estimated relative position on that lane to a history of the relative position to the previous driving position and the measurement data of the sensor (10) used in the recording unit to the position estimation information (12b); a position estimation unit (13) which estimates the current position (P) of the vehicle in relation to the previous driving position on the basis of the measurement data of the sensor (10) recorded in the recording unit (12) and the position estimation information (12b) of the recording unit (12) and adds the current position (P) as the last previous driving position to the history of the previous driving position of the recording unit (12); a deviation determination unit (14) that determines a deviation of the current position (P) of the vehicle estimated by the position estimation unit (13) from the plurality of previous driving positions (12a, 12a') by using a distribution of the plurality of previous driving positions (12a, 12a') recorded in the recording unit (12) and a predetermined reference value; and a control unit (15) that issues a warning to a driver or controls the vehicle to eliminate the deviation when the deviation is determined.

Description

Technical field

[0001] The present invention relates to a driver assistance device and a driver assistance system. Current state of the art

[0002] As one of the driver assistance functions of an automobile, there is a lane departure warning / control system that issues a warning to a driver or initiates corrective action to eliminate a deviation when lane departure is detected. This type of lane departure warning / control system often uses a method to initiate a warning / control based on a lane marking detected by an onboard sensor; however, a problem arises in that it is difficult to execute a warning or feedback control in an environment where a lane marking is not present, such as at an intersection or the like, or in an environment where, due to the influence of blurring or the like, it is difficult for the onboard sensor to detect a lane marking even though one is present.

[0003] As an example of a method for solving this problem, descriptions such as the following are disclosed in paragraphs 0007 and 0008 of PTL 1. “An object of the present invention is to provide a vehicle control device capable of performing lane keeping assist / lane drift avoidance control from information on previous lanes in relation to data from a vehicle navigation system and of performing lane keeping assist or lane drift avoidance when it is not possible to obtain lane information from an image taken by a camera, or even when a vehicle is driving on a road where there is no map information without updating the map information.”“Thus, according to the present invention, to eliminate the previously described disadvantage, a vehicle control device performing lane keeping assistance or lane departure avoidance comprises: a roadway storage means for storing a vehicle's roadway; a position detection means for detecting the current position of the vehicle; and a lane departure determination means for determining the degree of the vehicle's deviation from a lane based on the roadway stored in the roadway storage means and the current position of the vehicle detected by the position detection means.”

[0004] This means that in PTL 1, by determining the deviation from a lane based on the previous driving position and the current position of a vehicle, a warning and feedback control can be executed at the time of the lane deviation, even in an environment where a lane marking is not present. Furthermore, PTL 2 discloses a method for driver assistance, whereby a warning is generated depending on the history and quality of the driver's lane behavior. From these parameters, the driver's state of attention is inferred in order to distinguish between intentional and unintentional lane departures and to reduce the number of false warnings. PTL 3 discloses a computer-based driver assistance system. This system receives images of lane markings from a camera and identifies lane transitions from them. The system creates a history of data regarding the locations of these lane transitions.Based on this recorded historical data, the system actively controls the steering, braking, and / or acceleration of the vehicle to perform a lane-related driving task. PTL 4 discloses a method for lane-accurate vehicle positioning. It fuses data from vehicle sensors, a satellite positioning unit, and a lane-accurate digital map. An on-board computer calculates the vehicle's position by means of a map comparison. For this purpose, an observational model is updated, and a probability distribution is used to estimate and filter the position based on data particles. PTL 5 discloses a device for estimating driving trajectories. This device acquires and stores vehicle positions under specific conditions. To estimate a trajectory, the stored positions that lie within a predefined spatial area are statistically processed by averaging.PTL 6 discloses a lane departure warning system that adapts its control characteristics to the driving situation. It estimates the expected lane change frequency for the current route. This estimate is based on data from a navigation system, infrastructure information, and / or the driving history. Depending on the estimated frequency, the activation conditions of the warning or assistance function are modified to reduce the response sensitivity during frequent lane changes. Finally, PTL 7 discloses a method for storing driving histories for individual route segments. It analyzes whether the driving histories present in a segment can be assigned to one or more types based on a predefined condition, such as driving straight ahead or turning. If multiple types are identified, the system stores them in a classified manner.This allows subsequent processes, such as self-position estimation, to access the appropriate history type, thus improving their accuracy. List of quotations Patent literature PTL 1: JP 2015-162228 A PTL 2: DE 10 2004 047 889 A1 PTL 3: DE 10 2017 103 113 A1 PTL 4: DE 10 2011 120 497 A1 PTL 5: JP 2008 - 014 870 A PTL 6: JP 2005 - 018 211 A PTL 7: US 2019 / 391 594 A1 Summary of the invention; Technical task

[0005] However, in PTL 1, as described in paragraph 0014 and the like, the previous driving position (lane) stored in the vehicle navigation system's lane memory is used. Thus, the problem arises that in an environment where the vehicle navigation's position estimation accuracy is low, such as in an environment where the accuracy of a global navigation satellite system (GNSS) is low, the quality of the previous driving position (lane) as a reference is poor, and it is not possible to determine the deviation from a lane accordingly. Furthermore, since the deviation is determined based on the average of a large number of previous driving positions, an unsuitable reference driving position, where the vehicle does not normally travel, is calculated by averaging on a multi-lane road or at an intersection.Thus, the disadvantage is that it is not possible to determine the deviation itself with reference to the unsuitable reference driving position, and it is determined, for example, that the vehicle deviates while driving on a suitable route.

[0006] Therefore, one object of the present invention is to provide a driver assistance device capable of performing lane deviation warning / control even at an intersection, on a multi-lane road and in an environment where there are no lane separation lines and GNSS accuracy is low. Technical solution

[0007] To solve the problem, a driver assistance device with the features of claim 1 and driver assistance systems with the features of claims 8 and 12 are provided. Advantageous further developments are described in the dependent claims. Advantageous effects of the invention

[0008] According to the present invention, a warning for lane deviation or execution of a feedback control can also be carried out at an intersection, on a multi-lane road and in an environment where there are no lane separation lines and the GNSS accuracy is low. Brief description of the drawings [ Fig. 1] Fig. Figure 1 shows a functional block diagram of a driver assistance device according to embodiment 1. [ Fig. 2] Fig. Figure 2 shows a diagram illustrating an example of position estimation information recorded in a recording unit. [ Fig. 3] Fig. Figure 3 shows a diagram illustrating an example of vehicle / environment information recorded in the recording unit. [ Fig. 4] Fig. Figure 4 shows a diagram illustrating an example of deviation determination on a three-lane road. [ Fig. 5] Fig. Figure 5 shows a diagram illustrating an example of deviation determination at an intersection. [ Fig. 6] Fig. Figure 6 shows a diagram illustrating an example of a deviation determination when an exceptional prior driving position is included. [ Fig. 7] Fig. Figure 7 shows a diagram illustrating an example of a deviation determination based on a speed at the intersection. [ Fig. 8] Fig. Figure 8 shows a diagram illustrating an example of a deviation determination based on a speed in front of a temporary stop line. [ Fig. 9] Fig. Figure 9 shows a diagram illustrating an example of a deviation determination based on a speed in a parking lot. [ Fig. 10] Fig. Figure 10 shows a diagram illustrating an example of a velocity distribution. [ Fig. 11] Fig. Figure 11 shows a diagram illustrating an example of a deviation determination based on the activation status of a direction indicator. [ Fig. 12] Fig. Figure 12 shows a diagram illustrating an example of a distribution using a grid map. [ Fig. 13] Fig. Figure 13 shows a diagram illustrating a block configuration of a driver assistance system in embodiment 2. [ Fig. 14] Fig. Figure 14 shows a diagram illustrating an example of a processing sequence of an actuation control unit. [ Fig. 15] Fig. Figure 15 shows a diagram illustrating a block configuration of a driver assistance system in embodiment 3. [ Fig. 16] Fig. Figure 16 shows a diagram illustrating an example of a screen presented by a user input unit. [ Fig. 17] Fig. Figure 17 shows a diagram illustrating an example of a deviation determination based on valid / invalid information. [ Fig. 18] Fig. Figure 18 shows a diagram illustrating a system configuration of a driver assistance system in embodiment 4 and a device for sharing recorded data. [ Fig. 19] Fig. Figure 19 shows a diagram illustrating a block configuration of a driver assistance system in embodiment 5. [ Fig. 20] Fig. Figure 20 shows a diagram illustrating an example of deviation determination using a target path. [ Fig. 21] Fig. Figure 21 shows a diagram illustrating a block configuration of a driver assistance system in embodiment 6. [ Fig. 22] Fig. Figure 22 shows a diagram illustrating an example of processing in one unit to estimate a similar position. Description of embodiments

[0009] A driver assistance device according to embodiments of the present invention is described below with reference to the drawings. Design 1

[0010] The following is a driver assistance device according to embodiment 1 of the present invention with regard to Fig. 1 to 12 described. (Block configuration)

[0011] Fig. Figure 1 shows a functional block diagram of a driver assistance device according to the present embodiment. As shown in Fig. As shown in Figure 1, the driving assistance device 1 comprises a relative position estimation unit 11, a recording unit 12, a position estimation unit 13, a deviation determination unit 14, and a control unit 15. Measurement data from a sensor 10 are input to the relative position unit 11 and the position estimation unit 13. In particular, the driving assistance device 1 is a computer comprising hardware of an arithmetic operating device, such as a CPU, a main memory device, such as a ROM, an auxiliary memory device, such as a RAM, a communication device, and the like. The functions described below are implemented such that the arithmetic operating device executes a program loaded into the main memory device by the auxiliary memory device. A description is given below, while a prior art method in such a computer field is omitted accordingly.

[0012] Sensor 10 is a sensor mounted on a vehicle that measures the environment around the vehicle and can be, for example, a monocular camera, a stereo camera, a LiDAR, a millimeter-wave radar, a sonar, or the like. Sensor 10 measures the three-dimensional position of an object located around the vehicle. When a monocular camera is used, the acquired data is an image, and the three-dimensional position cannot be directly determined. The three-dimensional position can be measured by using a multiple images via a prior art motion stereo method or similar techniques. In the present embodiment, a stereo camera is used as Sensor 10. In addition to three-dimensional information, the stereo camera can capture information such as lane markings and stop lines, which are necessary for driver assistance, from a single image.Sensor 10 is not limited to the stereo camera and can also be other sensors or a combination of multiple sensors, such as a monocular camera and a LiDAR. Furthermore, Sensor 10 can include a sensor that measures the vehicle's condition, in addition to the sensor that measures the environment around the vehicle. For example, a GNSS, a compass, or a gyroscope capable of measuring the vehicle's position and orientation can be used. Additionally, a sensor that acquires information such as the vehicle's position and orientation by communicating with a radio beacon installed on a road can be used.

[0013] Each component of the driver assistance device 1 is explained below, followed by a detailed description of each component.

[0014] The relative position estimation unit 11 estimates the relative position of the vehicle on a lane where the vehicle is traveling for the first time, based on the measurement data from sensor 10. Additionally, the relative position estimation unit 11 records the estimated relative position in the recording unit 12 as a first previous driving position 12a0 on this lane and the measurement data from the sensor 10 used when the first previous driving position 12a0 is estimated as first position estimation information 12b0. Here, the relative position of the vehicle refers to information indicating the relative position / orientation based on the vehicle's position / orientation at a specific point in time.

[0015] The recording unit 12 records a plurality of previous driving positions 12a, position estimation information 12b relating to the previous driving positions 12a, and vehicle / environment information 12c. The plurality of previous driving positions 12a also includes the first previous driving position 12a0, and the position estimation information 12b also includes the first position estimation information 12b0.

[0016] The position estimation unit 13 estimates the current position and orientation (hereinafter referred to as a "current position P") of the vehicle relative to the previous driving position 12a based on the measurement data from sensor 10 recorded in the recording unit 12 and position estimation information 12b. Additionally, the position estimation unit 13 adds a history of the estimated current position P as the last previous driving position to the preceding driving position 12a of the recording unit 12 and outputs the result of this addition to the deviation determination unit 14. Thus, the position estimation unit 13 essentially does not operate on a lane where the vehicle is traveling for the first time and operates exclusively on a lane with a driving history.

[0017] The deviation determination unit 14 determines the deviation from the plurality of previous driving positions 12a by using the plurality of previous driving positions 12a recorded in the recording unit 12, the current position P of the vehicle in relation to the plurality of previous driving positions 12a and a predetermined reference value Th.

[0018] If the deviation detection unit 14 detects a deviation, the control unit 15 issues a warning to a driver or controls a steering system or an acceleration / deceleration system of the vehicle to eliminate the deviation. (Operation of the relative position estimation unit 11)

[0019] The following describes in detail the estimation processing of the vehicle's relative position by the relative position estimation unit 11. As previously described, the relative position estimation unit 11 estimates the relative position of the vehicle on a lane where the vehicle is traveling for the first time, based on the measurement data from sensor 10. Additionally, the relative position estimation unit 11 records the estimated relative position as the first previous driving position 12a0 and the measurement data from sensor 10 used in this estimation as the first position estimation information 12b0 in the recording unit 12.

[0020] As previously described, the relative position estimation unit 11 activates when the previous driving position 12a is not contained in the recording unit 12 according to the current environment, that is, when the vehicle is driving in the current environment for the first time. Thus, the relative position / orientation estimated by the relative position estimation unit 11 serves as the first previous driving position 12a0 with respect to the environment (track).

[0021] For example, if a monocular camera or a stereo camera is used as the sensor 10, the relative position estimation unit 11 can use a structure-from-motion (SfM) method or a visual-simultaneous localization-and-mapping (VSLAM) method, which is a method for estimating the relative position / orientation of the camera or the three-dimensional position of a feature point by extracting the feature point and an image feature amount of the feature point from an image and linking feature points between the multitude of images by using image feature amounts.

[0022] Fig. Figure 2 shows a diagram illustrating an example of the position estimation information 12b recorded in the recording unit 12 when the sensor 10 is a camera. As in Fig. Figure 2 shows the position estimation information 12b, which is defined by a combination of the "three-dimensional position" and the "image feature value" of the feature point estimated by the SfM or VSLAM method. The three-dimensional position of the feature point has the same coordinate system as that of the previous driving position estimated by the SfM or VSLAM method.

[0023] If, however, sensor 10 is a stereo camera or a LiDAR, an iterative closest-point simultaneous localization and mapping (ICP-SLAM) method can be used to estimate the position / orientation of the sensor by combining the three-dimensional position output by sensor 10 across a multitude of time points. In this case, since the Fig. 2. The “image feature amount” shown is not essential; the position estimation information 12b is defined exclusively by the “three-dimensional position”. (Operation of position estimation unit 13)

[0024] The following describes the details of the vehicle position estimation processing performed by the position estimation unit 13. As previously described, the position estimation unit 13 estimates the current position P of the vehicle relative to the previous driving position 12a based on the measurement data from sensor 10 recorded in the recording unit 12 and position estimation information 12b. Additionally, the position estimation unit 13 adds the history of the current position P as the last previous driving position to the previous driving position 12a in the recording unit 12 and outputs the result of this addition to the deviation determination unit 14.

[0025] For example, if a monocular or stereo camera is used as the sensor 10, the position estimation unit 13 can estimate the position and orientation from the correspondence between the two-dimensional and three-dimensional positions of the feature points. In this case, the position estimation unit 13 first extracts the feature point and the image feature value of the feature point from the point currently captured by the camera and links the extracted data with the image feature value contained in the position estimation information 12b. In this way, the position estimation unit 13 determines a multitude of correspondences between the two-dimensional position of the feature point in the current image and the three-dimensional position contained in the position estimation information 12b.Subsequently, the position estimation unit 13 estimates the three-dimensional position / orientation of the camera by using a solution to a known perspective-n-point (PnP) problem to estimate the position and orientation of the camera from the correspondence between the two-dimensional and three-dimensional positions.

[0026] If, however, the sensor 10 is a stereo camera or a LiDAR, an iterative closest-point (ICP) method can be used to estimate the position and orientation between three-dimensional point groups by linking the three-dimensional point groups together.

[0027] Here, the three-dimensional position contained in the position estimation information 12b has the same coordinate system as that of the previous driving position 12a. Thus, the estimated three-dimensional position / orientation of the camera has the same coordinate system as that of the previous driving position 12a, and the current position P can be determined with respect to the previous driving position 12a. (Operation of recording unit 12)

[0028] Details of the recording processing by the recording unit 12 are described below. As previously described, the multitude of previous driving positions 12a, the position estimation information 12b, and the vehicle / environment information 12c are recorded in the recording unit 12.

[0029] The vehicle / environment information 12c is a position determined by grouping the previous driving positions estimated by the relative position estimation unit 11 and the position estimation unit 13 according to a combination of the vehicle status and the environment status during the journey. Fig. Figure 3 shows an example of the vehicle / environment information 12c. This example illustrates that the preceding driving positions 12a with driving position IDs of 1, 5, and the like are classified into a group defined by the vehicle status and the environment status in the first row, and the preceding driving positions 12a with driving position IDs of 2, 3, and the like are classified into a group defined by the vehicle status and the environment status in the second row. The vehicle status refers, for example, to a driver or a type of tire and the number of passengers and is estimated from input by a user, output from a sensor located in the chassis, a seat, a seatbelt, or the like.The environmental status refers, for example, to the presence or absence of a vehicle ahead, the status of a signal, and the weather, and is estimated from the output of sensor 10 or the like.

[0030] The vehicle status, such as vehicle speed and the activation status of a turn signal at each position of the previous driving position 12a, can be recorded together in the recording unit 12. Such a vehicle status can be acquired by a Controller Area Network (CAN). (Operation of the deviation determination unit 14)

[0031] Details of the real-time deviation determination processing performed by the deviation determination unit 14 with regard to Fig. 4 to 11 described. The deviation determination unit 14 determines the deviation from the multitude of previous driving positions in real time by using the multitude of previous driving positions 12a recorded in the recording unit 12, the current position P of the vehicle estimated by the position estimation unit 13 and the predetermined reference value Th.

[0032] First, the deviation determination unit 14 checks the driving position ID, which belongs to the same group as the current vehicle status and environment status, with respect to the vehicle / environment information 12c recorded in the recording unit 12. Then, the deviation determination unit 14 determines the deviation from a multitude of selected previous driving positions 12a by using a multitude of previous driving positions 12a according to the driving position ID, the current position P of the vehicle estimated by the position estimation unit 13, and the predetermined reference value Th. The processing of deviation determination content by the deviation determination unit 14 for each specific road environment is described below.

[0033] Fig. Figure 4 shows a diagram illustrating an example of deviation determination on a three-lane road divided by lane dividers. Here, a lane divider L1 detectable by sensor 10 is represented by a solid line, and a lane divider L1' that cannot be detected by sensor 10 due to the influence of blur or the like is represented by a dashed line. Fig. Figure 5 shows a diagram illustrating an example of deviation determination at an intersection where a temporary stop line L2 is present in addition to the lane separation line L1.

[0034] The deviation determination unit 14 determines the deviation from the previous driving position 12a, which belongs to the same group as the current vehicle status and the environment status, and the current position P estimated by the position estimation unit 13.

[0035] First, the deviation determination unit 14 establishes a straight line X in a direction perpendicular to a vehicle's direction of travel by using the current position P as a reference.

[0036] The deviation determination unit 14 then applies a distribution to the coordinates of a point where the straight line X and each preceding driving position 12a intersect. In the examples of Fig. 4 and Fig. 5 is the preceding driving position 12a concentrated near arbitrary ideal driving positions. For such preceding driving positions 12a, for example, a multimodal distribution 12d, in which three cusps are present, can be obtained by using a mixed normal distribution as the distribution, in the example of Fig. 4 can be achieved and it can be done in the example of Fig. 5. A multimodal distribution 12d is obtained, in which two peaks are present. Thus, the value of the distribution 12s at each point on the straight line represents the probability of the vehicle being present at that point. The distribution 12d is not limited to the mixed normal distribution, and other probability distribution models can also be used.

[0037] Finally, the deviation determination unit 14 compares the probability of the vehicle being at the current position P, determined from the distribution 12d, with the predetermined reference value Th. If the probability of being at the current position P is less than the reference value Th, the deviation determination unit 14 determines that the vehicle is deviating. If the turn signal or hazard warning lights are activated, even if the probability of being at the current position P is less than the reference value Th, the deviation determination unit 14 determines that the driver has intentionally performed a deviation operation, and does not determine that the driver has deviated.

[0038] By repeating such processing during the journey, the deviation detection unit 14 can constantly monitor the deviation from the lane.

[0039] Fig. Figure 6 shows a diagram illustrating an example of deviation determination when an exceptional preceding driving position 12a' is included in the preceding driving position 12a belonging to the same group as the current vehicle status and the environment status at the time the parked vehicle is avoided. The number of exceptional preceding driving positions 12a' is significantly smaller than the number of other normal preceding driving positions 12a. Thus, in the distribution 12d determined by the deviation determination unit 14, an extremely small peak is formed at a position corresponding to the preceding driving position 12a', but the probability of this is low.Therefore, in the deviation determination based on the distribution 12d, the deviation determination unit 14 can properly determine the deviation from the track based on the majority of the normal preceding driving positions 12a, without being influenced by the exceptional preceding driving positions 12a'. In . Fig. 6. The distribution 12d is determined taking into account the exceptional preceding driving position 12a', but the distribution 12d can be determined in such a way that an exceptional preceding driving position, where the probability at the point where the straight line X and each preceding driving position intersect is less than a predefined value, can be deleted from the recording unit 12, and only the preceding driving position 12a, considered normal, is used.

[0040] The deviation determination unit 14 can perform deviation determination with respect to the vehicle status in addition to deviation determination with respect to the previously described current position P. The vehicle status that can be used in the deviation determination is, for example, the speed or the activation status of the turn signal. The processing of deviation determination content based on vehicle information by the deviation determination unit 14 for each specific environment is described below.

[0041] Fig. Figure 7 shows a diagram illustrating an example of a deviation determination based on speed at an intersection. Fig. Figure 8 shows a diagram illustrating an example of a deviation determination based on the speed before the temporary stop line L2. Fig. Figure 9 shows a diagram illustrating an example of a deviation determination based on speed in a parking lot.

[0042] In the deviation determination based on the speed, which in Fig. As shown in Figures 7 to 9, the deviation determination unit 14 first selects a set G of points from the previous driving positions 12a near the current position P. The deviation determination unit 14 then records the velocity V associated with each point in the set G, applies the mixed normal distribution, and determines a velocity distribution 12d. v . Fig. Figure 10 shows a diagram illustrating an example of the velocity distribution 12d v The deviation determination unit 14 compares the probability of the presence of a current velocity v, which is derived from the velocity distribution 12d. vThe deviation is determined using the predetermined reference value Th. If the probability of the current speed v being present is lower than the reference value Th, for example if the speed is extremely higher than the standard speed, the deviation determination unit 14 determines that the vehicle deviates from the normal speed.

[0043] Fig. Figure 11 shows a diagram illustrating an example of deviation determination based on the activation status of the turn signal at an intersection. In deviation determination based on the activation status of the turn signal, the deviation determination unit 14 selects a position P. L When the direction indicator is switched on or off, it selects from the previous driving positions 12a. Here, the deviation determination unit 14 selects a switch-off position other than position P. Loff if the current direction indicator is in an on state, and selects an on position other than position P L The system switches off when the current direction indicator is in a switched-off state. Subsequently, the deviation determination unit 14 defines a straight line Y in one direction of travel from the current position P. The deviation determination unit 14 projects a multitude of positions P. L in a direction perpendicular to the straight line Y, applies the mixed normal distribution to the positions on the straight line Y and determines a distribution 12d L with regard to the on or off position of the direction indicator. The deviation determination unit 14 compares the probability of the current position P being present, which is derived from the distribution 12d. LThe direction indicator is determined using the predetermined reference value Th. If the probability of the current position P being less than the reference value Th, the deviation determination unit 14 determines that the current position P deviates from the normal on and off positions. (Operation of control unit 15)

[0044] The following describes the content of the vehicle control processing by the control unit 15. If the deviation detection unit 14 detects a deviation in a position, speed, when the turn signal is switched on or off, or the like, the control unit 15 issues a warning to the driver or controls the vehicle to correct the deviation.

[0045] When issuing a warning to the driver, the control unit 15 notifies the driver of the deviation by means of an acoustic signal, a navigation system screen, steering wheel vibration, seat vibration, or other methods. When controlling the vehicle, the control unit 15 operates the steering, brakes, accelerator pedal, turn signals, and the like to correct the deviation. The control measures for correcting the deviation are described in detail below.

[0046] For example, in Fig. 4 and Fig. 5. By determining the deviation, the control unit 15 determines that the position has deviated. In this case, the control unit 15 first determines a position on the straight line X that is closest to the current position P and at which the probability of presence exceeds the predetermined reference value Th. Alternatively, the control unit 15 determines a position that is closest to the current position P and at which the probability of presence has its maximum value. Subsequently, the control unit 15 controls the steering, brake, and accelerator pedal so that the vehicle moves from the current position P to the position exceeding the predetermined reference value Th or in the direction of the determined maximum value.

[0047] For example, in Fig. 10. The deviation determination shows that the speed has deviated. In this case, the control unit 15 first determines a speed that is closest to the current speed v and at which the probability of its presence exceeds the predetermined reference value Th. Alternatively, the control unit 15 determines a speed that is closest to the current speed v and at which the probability of its presence is at its maximum value. The control unit 15 then controls the brake and accelerator pedal so that the vehicle accelerates or decelerates from the current speed v to the determined speed exceeding the reference value Th or to the determined speed at its maximum value.

[0048] Furthermore, for example in Fig. 11. The deviation determination determines the on-state of the direction indicator. In this case, the control unit 15 controls the direction indicator to have the same on-state as the on-state of the direction indicator at the selected on or off position P. L of the direction indicator. (Effects)

[0049] According to the embodiment 1 described above, the following effects can be achieved. (1) By using the position estimation information recorded in the recording unit, even in an environment where the accuracy of the GNSS is low, the position estimation unit can estimate the current position with high accuracy in relation to the previous position, and the deviation determination unit can determine the deviation from the previous journey and issue a warning / control. (2) Even if the preceding driving position includes a driving position different from a normal driving position, such as when avoiding a parked vehicle, a deviation may be determined while the influence of the driving position different from the normal driving position is suppressed. (3) The deviation can be properly determined even in an environment where a large number of traces are present. (4) A warning and control action may be taken if the vehicle status differs from the vehicle status during the previous journey. (5) It can provide a warning and control in case of speeding at an intersection ( Fig. 7), a failure to stop at a temporary stop line ( Fig. 8), a sudden acceleration due to an accelerator / brake malfunction in a parking lot ( Fig. 9) and the like. (6) A warning and control system may be implemented if the direction indicator is forgotten to be switched on or off at an intersection or the like. (7) A change in driving position due to a change in vehicle status or environment status shall not be determined as a deviation. (8) The control unit can effectively eliminate the deviation according to the detected deviation mode. (Modification example of embodiment 1)

[0050] The previously described deviation determination unit 14 determines the deviation based on distribution 12d and the like, calculated in real time with respect to the current position P. However, the method of calculating the distribution is not limited to this. For example, a pre-calculated distribution can be recorded in the recording unit 12, and the deviation determination can be carried out using this distribution. In particular, the distribution can be determined by dividing a space corresponding to a street environment into a grid-shaped map and holding a value corresponding to the probability of presence in each grid cell.

[0051] Fig. Figure 12 shows a diagram illustrating an example of the distribution by a grid map 12e. In this example, a value is assigned to each grid corresponding to a distance from the previous driving position 12a. A distribution for a multitude of previous driving positions 12a is determined by performing a similar processing for each previous driving position 12a and adding the determined multitude of grid maps 12e.

[0052] In this case, the deviation determination unit 14 determines the deviation when the grid value corresponding to the current position P is smaller than the predetermined reference value Th. The control unit 15 controls the steering, brake, and accelerator pedal so that the vehicle moves to a position where the grid value increases relative to the current position P.

[0053] If the deviation is determined by using the velocity, the velocity distribution becomes 12d. v The deviation is calculated in advance and maintained in each grid cell of grid map 12e by processing similar to that used in the deviation determination unit 14. Furthermore, if the deviation is determined by using the on / off status of the direction indicator, a value is set for each grid cell according to the distance to the on / off position of the direction indicator.

[0054] According to the previously described modification example of embodiment 1, the following effects can be achieved. That is, by pre-calculating the distribution, the computational load of real-time processing during the vehicle's journey can be reduced, and the deviation determination processing can be performed at a higher speed. Design 2

[0055] The following is a driver assistance device according to embodiment 2 of the present invention with regard to Fig. 13 and Fig. 14. In the following description, the same components as those of embodiment 1 are designated with the same reference numerals, and the main differences are described. Items not specifically described are the same as those in embodiment 1.

[0056] A driver assistance system 2 in the present embodiment comprises, in addition to the driver assistance device 1 in embodiment 1, a lane departure warning (LDW) device 22, which performs a warning / control based on a lane separation line, a road departure warning (RDW) device 23, which performs a warning / control based on a road edge, and an actuation determination unit 21, which determines which of the LDW device and RDW device is used. (Block configuration)

[0057] Fig. Figure 13 shows a diagram illustrating a block configuration of the driver assistance system 2. As in Fig. Figure 13 shows that the driver assistance system 2 comprises the actuation control unit 21, the LDW device 22, the RDW device 23, and the driver assistance device 1. The actuation control unit 21 determines which element of the group comprising the LDW device 22, the RDW device 23, and the driver assistance device 1 is to be actuated, based on the output of sensor 10. The LDW device 22 performs deviation detection and warning / control based on the lane marking. The RDW device 23 performs deviation detection and warning / control based on the road edge. A prior art method is used for the LDW device 22 and the RDW device 23. (Operation of the actuation control unit)

[0058] The following is the content of the processing in the operating unit 21 with regard to Fig. 14 described. The actuation determination unit 21 determines which element of the group comprising the LDW device 22, the RDW device 23 and the driver assistance device 1 is to be actuated, based on the output of the sensor 10.

[0059] First, in step S1, the actuation control unit 21 determines whether a lane separation line has been detected in the output of sensor 10. If the lane separation line has been detected, the process continues with step S2. If the lane separation line has not been detected, the process continues with step S3.

[0060] In step S2, the actuation determination unit 21 defines the LDW device 22 as an actuation device and causes the LDW device 22 to execute the warning / control based on the lane separation line detected by the sensor 10.

[0061] In step S3, the actuation control unit 21 determines whether a road edge has been detected in the output of sensor 10. If the road edge has been detected, the process continues with step S4. If the road edge has not been detected, the process continues with step S5.

[0062] In step S4, the actuation determination unit 21 defines the RDW device 23 as an actuation device and causes the RDW device 23 to execute the warning / control based on the road edge detected by the sensor 10.

[0063] In step S5, the actuation determination unit 21 designates the driver assistance device 1 as the actuation device and causes the driver assistance device 1 to execute the warning / control in a situation in which neither the lane separation line nor the road edge can be detected by sensor 10. (Effects)

[0064] According to the embodiment 2 described above, the following effects can be achieved. (1) A deviation detection and warning / control can be properly carried out according to the environment. (2) In an environment where the lane separation line and the road edge are detected, the previous driving position 12a and the position estimation information 12b are not recorded in the recording unit 12. This reduces the utilization of a memory area. (Modification example of embodiment 2)

[0065] The actuation determination unit 21 described above determines the actuation device based on the detection results of the lane separation line and the road edge in the output of sensor 10. The procedure for determining the actuation device is not limited to this.

[0066] For example, a sensor such as a GNSS, which estimates a map and the carrier vehicle's position on the map, may also be present, and the actuation determination unit 21 may determine the actuation device based on the position on the map. In particular, the driver assistance device 1, which performs a warning / control based on the previous driving position, is determined as the actuation device at a location where it is known in advance that it will be difficult to detect a lane divider and road edge, such as at an intersection, and at a location where it is known in advance that it will be difficult to perform a warning / control based on a lane divider and road edge, such as on a road with variable traffic and no lane divider.The LDW device 22 is designated as the actuating device at a location where there is a high probability of the activation of a warning / control based on the lane separation line, such as on a highway or a major state motorway.

[0067] According to the previously described modification example of embodiment 2, the following effects can be achieved. That is, an actuating device suitable for any environment can be specified without being affected by the time required to process the sensor output and the detection of the lane separation line and the road edge, or by the risk of erroneous detection. embodiment 3

[0068] The following is a driver assistance device according to embodiment 3 of the present invention with regard to Fig. Sections 15 to 17 are described. In the following description, the same components as those of embodiment 1 are designated with the same reference numerals, and the main differences are described. Items not specifically described are the same as those in embodiment 1.

[0069] A driver assistance system 3 in the present embodiment is achieved by adding a user input unit to the driver assistance device 1 in embodiment 1. The user input unit receives valid / invalid information 12f for the deviation determination result, which is entered by a user. (Block configuration)

[0070] Fig. Figure 15 shows a diagram illustrating a block configuration of the driver assistance device 3. As in Fig. As shown in Figure 15, the driver assistance device 3 comprises a relative position estimation unit 11, a recording unit 12, a position estimation unit 13, a deviation determination unit 14, a control unit 15 and a user input unit 31.

[0071] The user input unit 31 receives valid / invalid information 12f for the deviation determination result, which is entered by the user. In the present embodiment, the recording unit 12 records valid / invalid information 12f received from the user input unit 31 in addition to the information equivalent to that in embodiment 1. In the present embodiment, the deviation determination unit 14 determines the deviation taking into account the valid / invalid information 12f. (Operation of the user input unit)

[0072] The following is the content of the processing in user input unit 31 with regard to Fig. 16 described. The user input unit 31 receives a valid / invalid information 12f for the deviation determination result from the user's input.

[0073] Fig. Figure 16 shows a diagram illustrating an example of a screen 31a presented by the user input unit 31. The user input unit 31 presents the user with the date, time, and location determined as a deviation by the deviation determination unit 14 using a navigation system screen or the like. The location can be displayed using a navigation system map. User inputs are valid if the result determined as the deviation is correct, and inputs are invalid if the result is incorrect. Input is provided by touching or pressing a button on the screen. This specifies a location where the deviation determination unit 14 has made an incorrect deviation determination. (Operation of the deviation determination unit)

[0074] The following is the content of the deviation determination processing in the deviation determination unit 14 in the present embodiment with regard to Fig. 17 described. The deviation determination unit 14 in the present embodiment determines whether a deviation determination is carried out on the basis of the valid / invalid information 12f or not, prior to the deviation determination processing, which is equivalent to that in embodiment 1.

[0075] Fig. Figure 17 shows a diagram illustrating an example of a status where it is determined not to perform the deviation determination based on the valid / invalid information 12f. In this example, the deviation was previously determined at position 31b; however, the user indicates that the deviation determination was invalid. Thus, if the current position P is within an area 31c of a predetermined distance from position 31b, the deviation determination unit does not perform the deviation determination. Conversely, if the current position P is not within area 31c, the processing by the deviation determination unit 14 is not carried out. (Effects)

[0076] According to the embodiment 3 described above, the following effects can be achieved. That is, since the deviation determination is not carried out within a predetermined distance from the previous deviation position identified as invalid, the erroneous deviation determination is not repeated. Design 4

[0077] The following is a driver assistance device according to embodiment 4 of the present invention with regard to Fig. 18. The same components as those of embodiment 1 are subsequently designated with the same reference numerals, and the main differences are described. The points not specifically described are the same as those in embodiment 1.

[0078] A driver assistance system 4 in the present embodiment is achieved by adding a data transmission and data reception unit 4a, which transmits and receives data, to the configuration according to the driver assistance device 1 in embodiment 1. By forming a system comprising a plurality of driver assistance devices 4 and a device 40 for sharing data, information acquired by the carrier vehicle can be forwarded to another vehicle, or the carrier vehicle can perform a deviation determination based on information acquired by another vehicle. (Block configuration)

[0079] Fig. Figure 18 shows a diagram illustrating a block configuration of the driver assistance device 4 in the present embodiment and the device 40 for sharing data transmitted by each driver assistance device.

[0080] As in Fig. Figure 18 shows that the driver assistance device 4 comprises the driver assistance device 1 in embodiment 1 and the data transmitting and receiving unit 4a.

[0081] The data-sharing device 40 comprises a data transmission and reception unit 40a and a shared recording unit 40b. Two or more driver assistance devices 4 are connected to the data-sharing device 40 via a network 41, such as a mobile phone network. The driver assistance devices 4 are mounted on different vehicles. The data-sharing device 40 is, for example, installed in a server.

[0082] The data transmission and reception unit 4a of the driver assistance device 4 transmits data recorded by the recording unit 12 of the driver assistance device 1 via the network 41 to the device 40 for data sharing. Data received by the device 40 for data sharing is recorded as data of the recording unit 12 of the driver assistance device 1.

[0083] The data transmission and reception unit 40a of the data-sharing device 40 outputs the data received from the driver assistance device 4 to the shared recording unit 40b. The data recorded by the shared recording unit 40b is sent to the driver assistance device 4. (Operation of the shared recording unit)

[0084] The following describes the content of the processing in the shared recording unit 40b. The shared recording unit 40b integrates and records the stored data received from a multitude of driver assistance devices 4.

[0085] Similar to recording unit 12, shared recording unit 40b initially groups the previous driving positions based on the vehicle / environment information 12c. Here, in shared recording unit 40b, a vehicle type can be used as a vehicle status in addition to the vehicle status used in recording unit 12. Instead of identifying the driver individually, a driver attribute, such as the driver's age, can be used. The following process is performed on each group.

[0086] Subsequently, the shared recording unit 40b unifies the coordinate systems of the stored data received by the multitude of driver assistance devices 4. Since the stored data from each driver assistance device 4 is recorded in a unique coordinate system, the coordinate systems must be unified to integrate the multitude of recorded data. To unify the coordinate systems, a harmonization of three-dimensional positions contained in the position estimation information 12b of all recorded data can be used. For example, an iterative closest-point (ICP) method can be used to estimate the position between three-dimensional point groups by joining the three-dimensional point groups together.

[0087] The shared recording unit 40b then selects a previous driving position for transmission to the driving assistance device 4 from a large number of previous driving positions with a unified coordinate system. For example, similar to the processing by the deviation determination unit 14, the current position P is virtually determined, and the distribution is applied to all previous driving positions. If a predetermined number of previous driving positions are selected and the distribution applied, the previous driving positions are chosen such that the difference between the shape of the distribution and the distribution applied to all previous driving positions is minimized. Thus, a distribution similar to that obtained when using all previous driving positions can be achieved by using a small number of previous driving positions. (Effects)

[0088] According to the embodiment 4 described above, the following effects can be achieved. (1) By using a large number of previous driving positions recorded by a large number of vehicles, deviations can be determined with greater accuracy. In addition, even at a location where the carrier vehicle has not previously traveled, deviations can be determined and a warning / control action can be taken. (2) By expressing a distribution similar to a distribution encompassing a large number of data using a small number of previous driving positions, the data capacity communicated through the network and the storage capacity of the driver assistance device can be reduced. Design 5

[0089] The following is a driver assistance device according to embodiment 5 of the present invention with regard to Fig. 19 and Fig. 20. The same components as those of embodiment 1 are subsequently designated with the same reference numerals, and the main differences are described. The points not specifically described are the same as those in embodiment 1.

[0090] A driver assistance system 5 in the present embodiment is achieved by adding a path planning unit 51, which plans the path of the vehicle based on the output of the sensor 10, to the driver assistance device 1 in embodiment 1. (Block configuration)

[0091] Fig. Figure 19 shows a diagram illustrating a block configuration of the driver assistance device 5. As in Fig. Figure 19 shows that the driver assistance device 5 comprises a relative position estimation unit 11, a recording unit 12, a position estimation unit 13, a deviation determination unit 14, a control unit 15, and the path planning unit 51. The path planning unit 51 plans the target path along which the vehicle is to travel based on the output of the sensor 10, but can also plan the target path using a map, a user-entered destination, or the like, in addition to the output of the sensor 10. In the present embodiment, the deviation determination unit 14 determines the deviation by using the path planned by the path planning unit 51, in addition to the determination method in embodiment 1. (Operation of the deviation determination unit)

[0092] The following is the content of the processing in the deviation determination unit 14 in the present embodiment with regard to Fig. 20 described.

[0093] Fig. Figure 20 shows a diagram illustrating an example of deviation determination using the target path. In the present embodiment, a target path 51a planned by the path planning unit 51 is entered into the deviation determination unit 14. Thus, the deviation determination unit 14 takes the target path 51a into account when generating the distribution 12d. In particular, the deviation determination unit 14 first selects the preceding travel position 12a from the target path 51a within a predetermined range. Thus, in the example of Fig. 20 exclusively the preceding driving position 12a of a journey on the track on the upper side of Fig. 20 selected and the previous driving position 12a of a journey on the track on the lower side of Fig. 20 is excluded. Subsequently, using the selected previous driving position 12a and the selected target path 51a, the distribution 12d is applied by processing similar to that used in the deviation determination unit 14. Here, a weight can be added to the previous driving position 12a and target path 51a in the application of distribution 12d. For example, the influence of the target path 51a on the distribution 12d can be further increased by increasing the weight for the target path 51a.

[0094] Subsequently, the deviation determination unit 14 determines the deviation on the basis of the fitted distribution 12d by processing similar to embodiment 1.

[0095] Similarly, the target route 51a planned by the track planning unit 51 can also be used in the deviation determination based on the speed and the activation status of the direction indicator. In particular, track planning unit 51 plans the target speed for the target route 51a and the activation status of the target direction indicator, and the deviation determination unit 14 generates the distribution 12d by using the target speed and the activation status of the target direction indicator. (Effects)

[0096] According to the embodiment 5 described above, the following effects can be achieved. (1) The number of previous driving positions used to apply the distribution is reduced, and the deviation determination processing is accelerated. A movement deviating from the target path can be determined as a deviation. Furthermore, even if the previous driving position includes a driving position different from a normal driving position, such as a driving position when the parked vehicle is avoided, a deviation can be determined by excluding the driving position that differs from the normal driving position. (2) By using the previous driving position in addition to the target path, a deviation from the target path can be determined taking into account the displacement in a normal range according to the position and the vehicle / environment information. Design 6

[0097] The following is a driver assistance device according to embodiment 6 of the present invention with regard to Fig. 21 and Fig. 22. The same components as those of embodiment 1 are subsequently designated with the same reference numerals, and the main differences are described. The points not specifically described are the same as those in embodiment 1.

[0098] A driving assistance device 6 in the present embodiment comprises a unit 61 for estimating a similar position, which replaces the relative position in the environment in which the vehicle was first driven with a position in relation to the similar previous driving position 12a from the output of the sensor 10, the output of the relative position estimation unit 11 and the plurality of previous driving positions 12a and the position estimation unit 12b recorded in the recording unit 12. (Block configuration)

[0099] Fig. Figure 21 shows a diagram illustrating a block configuration of the driver assistance device 6. As in Fig. Figure 21 shows that the driver assistance device 6 comprises a relative position estimation unit 11, a recording unit 12, a position estimation unit 13, a deviation determination unit 14, a control unit 15 and the unit 61 for estimating a similar position. (Operation of Unit 61 to estimate a similar position)

[0100] The following describes the processing performed by unit 61 for estimating a similar position. Unit 61 for estimating a similar position replaces the relative position in the environment where the vehicle first traveled with a position relative to the similar previous driving position 12a from the output of sensor 10, the output of the relative position estimation unit 11, the multitude of previous driving positions 12a, and the position estimation unit 12b recorded in the recording unit 12.

[0101] Fig. Figure 22 shows a diagram illustrating an example of processing by Unit 61 for estimating a similar position, and represents an environment E1 (left diagram) around the current position P and an environment E2 (right diagram) around the previous driving position 12a, recorded in Recording Unit 12. When the two environments are compared, there is a difference in that the vehicle cannot drive straight from the current position in environment E1 and cannot turn left from the current position in environment E2; but both environments are similar in that the vehicle must temporarily stop at the temporary stop line L2 before turning into the priority road from the current position.

[0102] As previously described, if the preceding driving position 12a is recorded in the recording unit 12 in the preceding environment E2 similar to the current environment E1 at the time of driving, the similar position estimation unit 61 replaces the current position P of the vehicle in the current environment E1 with a position P' in the preceding environment E2. Specifically, the similar position estimation unit 61 detects a similar object present in both the current environment E1 and the preceding environment E2. Examples of the similar object include road markings such as a lane divider line L1 and a temporary stop line L2, detected by the output of sensor 10; three-dimensional objects such as a building O1 and a tree O2; and position estimation information 12b determined by the relative position estimation unit 11. In the example of Fig.In unit 22, the temporary stop line L2 is identified as the similar object. Unit 61 then calculates the position P' in the preceding environment E2 by adding the vehicle's current position P relative to the similar object in the current environment E1 to the position of the similar object in the preceding environment E2. Finally, unit 61 outputs the position P' relative to the preceding driving position to the deviation determination unit 14. Thus, the deviation determination unit 14 can perform the deviation determination assuming the vehicle is at position P' in the preceding environment E2. (Effects)

[0103] According to the embodiment 6 described above, the following effects can be achieved. That is, even at a location where the carrier vehicle has not previously driven, the deviation can be determined by using the previous driving position in a similar environment and to execute the warning / control.

[0104] The present invention is not limited to the preceding embodiments, and various modifications may exist. For example, the preceding embodiments have been described in detail to provide an easily understandable description of the present invention, and the preceding embodiments are not necessarily limited to a single case encompassing all described configurations. Other forms considered within the scope of the technical idea of ​​the present invention are also included within the scope of the present invention. Furthermore, some components in one embodiment may be replaced by components of another embodiment, and the configuration of another embodiment may be added to the configuration of one embodiment. With respect to some components in the embodiments, other components may be added, removed, and replaced.Some or all of the configurations, functions, functional units, processing elements, and the like may be implemented in hardware, for example, using an integrated circuit. Alternatively, the aforementioned components, functions, and the like may be implemented in software by the processor, which interprets and executes a program to perform the respective functions. Information such as a program, a table, and a file that executes each function may be stored in memory, a recording device such as a hard disk drive or solid-state drive (SSD), or a recording medium such as an IC card, an SD card, or a DVD. Reference symbol list 1, 3 to 6 Driver assistance device 2 Driver assistance systems 10 Sensor 11 Relative position estimation unit 12 recording units 12a, 12a' previous driving position 12b Position Estimation Information 12c Vehicle / Environment Information 12d, 12d v , 12d L distribution 12e Grid map 12f valid / invalid information 13 Position estimation unit 14 Deviation Determination Unit 15 Control unit 21 Actuation determination unit 22 LDW device 23 RDW device 31 User input unit 4a Data transmission and reception unit 40 Data sharing device 40a Data transmission and reception unit 40b shared recording unit 51 Railway Planning Unit 61 units for estimating a similar position

Claims

[1] Driver assistance device (1), comprising: a recording unit (12) that records a plurality of previous driving positions (12a, 12a') of a vehicle and position estimation information (12b) in relation to the previous driving positions (12a, 12a'); a relative position estimation unit (11) that estimates a relative position of the vehicle on a lane based on measurement data from a sensor (10) that detects an environment of the vehicle and adds the estimated relative position on that lane to a history of the relative position to the previous driving position and the measurement data of the sensor (10) used in the recording unit to the position estimation information (12b); a position estimation unit (13) which estimates the current position (P) of the vehicle in relation to the previous driving position on the basis of the measurement data of the sensor (10) recorded in the recording unit (12) and the position estimation information (12b) of the recording unit (12) and adds the current position (P) as the last previous driving position to the history of the previous driving position of the recording unit (12); a deviation determination unit (14) that determines a deviation of the current position (P) of the vehicle estimated by the position estimation unit (13) from the plurality of previous driving positions (12a, 12a') by using a distribution of the plurality of previous driving positions (12a, 12a') recorded in the recording unit (12) and a predetermined reference value; and a control unit (15) that issues a warning to a driver or controls the vehicle to eliminate the deviation when the deviation is determined. [2] Driving assistance device (1) according to claim 1, wherein the deviation determination unit (14) uses a multimodal distribution as the distribution. [3] Driver assistance device (1) according to claim 1, wherein the recording unit (12) furthermore records a vehicle status of the vehicle, and the deviation determination unit (14) determines the deviation by using the vehicle status recorded in the recording unit (12) and a vehicle status determined during the current driving. [4] Driving assistance device (1) according to claim 3, wherein the deviation determination unit (14) determines the deviation of a vehicle speed by using the speed which is a type of vehicle status and a speed determined during the current journey. [5] Driving assistance device (1) according to claim 3, wherein the deviation determination unit (14) determines the deviation of an on-state of a direction indicator by using the on-state of the direction indicator, which is a type of vehicle status, and the on-state of the direction indicator, determined during the current journey. [6] Driving assistance device (1) according to claim 1, wherein the recording unit (12) further records vehicle / environment information (12c) comprising at least one element of the group comprising a driver of the vehicle, a tire, a number of passengers, the presence or absence of a vehicle ahead, a traffic light activation status and the weather, and the deviation determination unit (14) compares the vehicle / environment information (12c) in the recording unit (12) with the vehicle / environment information (12c) determined during the current journey, selects the plurality of previous driving positions (12a, 12a') under the same condition and determines the deviation on the basis of the plurality of selected previous driving positions (12a, 12a'). [7] Driving assistance device (1) according to claim 1, wherein a pre-calculated distribution of the plurality of previous driving positions (12a, 12a') is recorded in the recording unit (12), and the deviation determination unit (14) determines the deviation from the plurality of previous driving positions (12a, 12a') by using the distribution of the recording unit (12), the position of the vehicle in relation to the plurality of previous driving positions (12a, 12a') and the predetermined reference value. [8] Driver assistance system (2) comprising: the driver assistance device (1) according to claim 1; an LDW device (22) which determines the deviation on the basis of a lane separation line and issues a warning or controls the vehicle; an RDW device (23) which determines the deviation based on a road edge and issues a warning or controls the vehicle; and an actuation determination unit (21) which determines which element of the group comprising the LDW device (22), the RDW device (23) the driver assistance device (1) is to be actuated. [9] Driving assistance system (2) according to claim 8, wherein, if a lane separation line and a road edge are not detected from the output of the sensor (10), the actuation determination unit (21) actuates the driving assistance device (1). [10] Driving assistance system (2) according to claim 8, further comprising a sensor (10) that estimates a position of the vehicle on a map, wherein the actuation determination unit (21) actuates the driving assistance device (1) on the basis of the position of the vehicle on the map. [11] Driving assistance device (1) according to claim 1, further comprising a user input unit (31) which receives information from a user to indicate the validity / invalidity of a position determined as the deviation; wherein, if the position of the vehicle is contained at a predetermined distance from the position predetermined as the invalid deviation, the deviation determination unit (14) does not perform a deviation determination. [12] Driver assistance system (2) in which the driver assistance device (1) according to claim 1 and a device (40) for sharing data are connected to each other via a network, wherein the device (40) for sharing data comprises a shared recording unit (40b) which integrates data received from a plurality of the driver assistance devices, records the integrated data as shared data and makes the shared recording data available to each of the driver assistance devices. [13] Driver assistance system (2) according to claim 12, wherein the shared recording unit (40b) selects a previous driving position and provides the selected previous driving position for each of the driver assistance devices, so that a difference between a form of distribution applied to all previous driving positions (12a, 12a') and a form of distribution applied to the selected previous driving position is minimized. [14] Driving assistance device (1) according to claim 1, further comprising a path planning unit (51) which plans a target path of the vehicle from the output of the sensor (10), wherein the deviation determination unit (14) determines the deviation from the target path using a distribution of the preceding driving positions (12a, 12a') in a predetermined area. [15] Driving assistance device (1) according to claim 14, wherein the deviation determination unit (14) applies the distribution to the previous driving position and the target path with a weight. [16] Driving assistance device (1) according to claim 1, further comprising a unit (61) for estimating a similar position, which replaces a relative position in a current environment with a similar position with respect to the previous driving position from the output of the sensor (10), an output of the relative position estimation unit (11) and the plurality of previous driving positions (12a, 12a') and the position estimation unit (13) recorded in the recording unit (12).

Images