KR102058741B1 - System for selecting of lane change feasible space(lcfs) and method for the same - Google Patents

Abstract

According to an exemplary embodiment of the present invention, a space selection system capable of changing a vehicle may include a determination unit that determines a driver's propensity to change a lane to an actual driver, and each space between a driver's vehicle and a surrounding vehicle driving a different lane. A safety score calculation unit for calculating a safety score for determining the stability as an objective value based on a stability variable relating to the stability of the vehicle running, and based on an energy variable relating to the efficiency of the vehicle running for each space, An energy score calculation unit for calculating an energy score to be determined as an objective value, and the calculated safety score and energy score are added to calculate a lane change score for each space, and the space having the maximum score can be changed to the vehicle. It includes a space selection unit for selecting the space.

Description

Space selection system that can be changed by car and method thereof {SYSTEM FOR SELECTING OF LANE CHANGE FEASIBLE SPACE (LCFS) AND METHOD FOR THE SAME}

The present invention relates to a system and method for automatically selecting a space that can be changed by a car, and more particularly, to a system and method for selecting a space that can be changed by a car in consideration of both the stability and efficiency of vehicle operation.

Conventional lane change assist technologies include blind spot detection (BSD) systems and advanced lane change assist (ALCA) systems that alert drivers when blind spot vehicles are invisible when changing lanes. In addition, there is a Lane Departure Warning System (LDWS) or Lane Keeping Assist System (LKAS) that provides warning when a driver inadvertently steps on or crosses a lane and then returns to the main lane.

In the conventional Advanced Driver Assistance System (ADAS) developed so far, most of the technologies do not consider the surrounding vehicles but only the controlled vehicles. For example, ACC, ASCC, LKAS, etc., which consider only the driver's convenience, and most of them are indirect driving assistance systems such as BSD and ALCA, which only recognize the vehicle in blind spots and give warning or notification to the driver without direct control of the vehicle. .

Recently, a longitudinal and transverse integrated lane change controller has been developed that changes the auto lane considering the driving characteristics of not only the vehicle under control but also the surrounding vehicles.However, the controller selects only the space between two vehicles closest to the vehicle to be controlled. There is a limitation to change.

Accordingly, in order to further secure driver's safety, technologies that assist in changing to an auto lane in consideration of driving characteristics of a wider range of surrounding vehicles are required.

Related prior arts include Republic of Korea Patent Publication No. 10-1480652 (name of the invention: lane change control device and its change control method, registration date: January 2, 2015).

Embodiments of the present invention consider both safety and energy aspects of the space between successive vehicles within a certain range based on the driver's vehicle to help the driver having difficulty in changing lanes on a road having a variable driving situation. The present invention provides a space selection system and a method capable of changing a car to select a space that can be optimally changed to a car.

The problem to be solved by the present invention is not limited to the problem (s) mentioned above, and other object (s) not mentioned will be clearly understood by those skilled in the art from the following description.

According to an exemplary embodiment of the present invention, a space selection system capable of changing a vehicle may include a determination unit that determines a driver's propensity to change a lane to an actual driver, and each space between a driver's vehicle and a surrounding vehicle driving a different lane. A safety score calculation unit for calculating a safety score for determining the stability as an objective value based on a stability variable relating to the stability of the vehicle running, and based on an energy variable relating to the efficiency of the vehicle running for each space, An energy score calculation unit for calculating an energy score to be determined as an objective value, and the calculated safety score and energy score are added to calculate a lane change score for each space, and the space having the maximum score can be changed to the vehicle. It includes a space selection unit for selecting the space.

In addition, the determination unit according to an embodiment of the present invention may receive behavior information regarding the vehicle driving behavior and the lane change behavior of the driver from the driver, and determine the propensity of the driver based on the behavior information.

In addition, the determination unit according to an embodiment of the present invention determines the propensity of the driver by performing a questionnaire about the vehicle driving behavior of the driver, the questionnaire survey also ranks for the stability variable and the energy variable , And a preference survey for a space changeable to a driver's preferred vehicle for a plurality of situations that may occur when driving a vehicle according to a combination of the stability variable and the energy variable.

In addition, the stability variable according to an embodiment of the present invention is the vehicle head spacing, which is the distance between the surrounding vehicle, the collision warning index which is a possibility of collision between the vehicle of the driver and the surrounding vehicle, the vehicle of the driver and the preceding vehicle in each space A first speed deviation, which is a speed deviation between the second vehicle, a second speed deviation, which is a speed deviation between the driver's vehicle, and a trailing vehicle in each space, and the distance between the driver's vehicle and the exit; It may include the amount of time required and the cumulative acceleration that is the time spent entering each space.

In addition, the safety score calculation unit according to an embodiment of the present invention obtains the constant input constant function of the triangular function type and the output constant function of the constant form reflecting the weight according to the ranking check for the stability variable, The safety score may be calculated according to a fuzzy rule generated using an input belonging function and the obtained output belonging function.

In addition, the space selection system that can be changed to the lane according to an embodiment of the present invention defines a virtual vehicle that is virtually generated by projecting the peripheral vehicle in the lane of the driver's vehicle, the driver's vehicle is the virtual The method may further include an energy variable estimating unit estimating the energy variable including a time required to enter the space between the vehicles and a magnitude of the cumulative acceleration.

The energy variable estimating unit may estimate the energy variable based on Equation 24 and Equation 20 when the driver's vehicle is located in front of or behind the space between the virtual vehicles. have.

[Formula 24]

Equation 20

Here, y ^* is an estimated value of the energy variable, K (·) means a kernel function, x ^* is the driver's vehicle measured as the driver's vehicle is located in front or rear in each space between the virtual vehicle and Input data including the speed and position of the virtual vehicle, x _i is output for the pre-trained training input data and the trained training input data as the driver's vehicle is located in front of or behind each space between the virtual vehicles. B means data, and b is a distribution degree of the trained data, and means the density of the training input data and the training output data.

The energy variable estimator may estimate the energy variable using a multilayer neural network structure when the driver's vehicle is located in the center of the space between the virtual vehicles.

In addition, the energy score calculation unit according to an embodiment of the present invention obtains the input constant function in the form of a trigonometric function and the weight constant according to the importance investigation for the energy variable, and obtains the constant output constant function, the obtained input The energy score may be calculated according to the generated fuzzy rule using the membership function and the obtained output function.

In addition, the space selection unit according to an embodiment of the present invention, if the speed of the vehicle of the driver is the same as the speed of the peripheral vehicle, but the head spacing is the same, if the distance between the vehicle of the driver and the exit is greater than a predetermined threshold The space between the vehicle closest to the driver's vehicle and the nearest vehicle is selected as a space that can be changed into the vehicle, and if the distance between the driver's vehicle and the exit is less than a predetermined threshold, the space between the vehicle surrounding the vehicle located behind the driver's vehicle One may be selected as a space that can be changed into the difference.

In addition, the space selection unit according to an embodiment of the present invention when the head spacing is the same but the distance between the vehicle of the driver and the exit is more than a predetermined threshold, the speed of the vehicle of the driver is greater than the average speed of the surrounding vehicle Selecting one of the spaces between the surrounding vehicles located in front of the vehicle of the driver as a space that can be changed into the vehicle, and when the speed of the vehicle is less than the average speed of the surrounding vehicle, the vehicle of the driver As a reference, one of the spaces between the surrounding vehicles located behind the vehicle may be selected as a space that can be changed into the vehicle.

In addition, the space selection unit according to an embodiment of the present invention, if the speed of the vehicle of the driver is the same as the speed of the peripheral vehicle, but the distance between the vehicle and the exit of the driver is more than a predetermined threshold, the head gap is greater than a predetermined threshold. One of the spaces between the surrounding vehicles may be selected as a space that can be changed into the difference.

In addition, the space selection system capable of changing the lane according to an embodiment of the present invention further includes a sensor for detecting the position information and the speed information of the surrounding vehicle from a sensor provided in the driver's vehicle and a sensor installed around the lane Can be.

In addition, according to an embodiment of the present invention, the space selection method capable of changing the lane includes: performing a survey for reflecting the driver's disposition on changing the lane to the actual driver, driving a different lane from the driver's vehicle; Calculating a safety score for determining the stability as an objective value for each space between the surrounding vehicles, based on an energy variable for efficiency of vehicle operation for each space. Calculating an energy score for determining the efficiency as an objective value, and calculating the lane change score for each space by adding the calculated safety score and the energy score, and changing the space having the maximum score to the car. The step may include selecting as much space as possible.

In addition, the space selection method that can be changed to the lane according to an embodiment of the present invention defines a virtual vehicle that is virtually generated by projecting the peripheral vehicle in the lane of the driver's vehicle, the driver's vehicle is the virtual The method may further include estimating the energy variable including a time required to enter the space between the vehicles and a magnitude of the cumulative acceleration.

In addition, the space selection method that can change the lane according to an embodiment of the present invention further comprises the step of detecting the position information and the speed information of the surrounding vehicle from a sensor provided in the driver's vehicle and a sensor installed around the lane Can be.

Specific details of other embodiments are included in the detailed description and the accompanying drawings.

According to embodiments of the present invention, a space that can be optimally changed into a car may be selected in consideration of both safety and energy aspects among spaces between consecutive surrounding vehicles within a predetermined range based on the driver's vehicle.

According to embodiments of the present invention, a driver-cooperative advanced technology that assists in changing lanes in difficult and complicated road classification and specialized sections may have high expandability, and may have high problems such as high accident rate and It can be used to solve social problems such as traffic jams, reducing traffic accident rates, reducing social costs and increasing convenience.

1 is a schematic diagram of a space selection system capable of changing a lane according to an embodiment of the present invention.
2 is a block diagram of a space selection system capable of changing a lane according to an embodiment of the present invention.
3 is a block diagram illustrating a space selection system capable of changing a lane according to an embodiment of the present invention.
4 is a schematic diagram showing the stability and energy variables to be considered when performing a survey in an embodiment of the present invention.
Figure 5a is a graph showing the results of the ranking survey of the survey in an embodiment of the present invention.
5B to 5D are graphs showing results of a preference survey in a survey in accordance with an embodiment of the present invention.
6 is a graph showing the membership function for the stability variable in one embodiment of the present invention.
FIG. 7 is a schematic diagram illustrating the concept of a virtual vehicle defined when estimating an energy variable, according to an embodiment of the present invention.
FIG. 8 is a schematic diagram illustrating a concept of a neural network of a multilayer structure used when estimating an energy variable in an embodiment of the present invention.
9 is a graph showing a membership function for an energy variable according to an embodiment of the present invention.
10 is a flowchart illustrating a space selection method capable of changing a lane according to an embodiment of the present invention.

Advantages and / or features of the present invention and methods for achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only the embodiments, the disclosure of the present invention is complete, the common knowledge in the art to which the present invention belongs It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims. Like reference numerals refer to like elements throughout.

Hereinafter, with reference to the accompanying drawings will be described embodiments of the present invention;

Prior to describing the embodiments, the present invention will be described in brief.

1 and 2, in the present invention, in order to help a driver having difficulty in changing lanes on a road having a variable driving situation, a search range of a predetermined range, ie, a lane change space, is based on a vehicle of a driver that is a controlled vehicle. The optimal space that can be changed to a car is selected by considering both safety and energy aspects among the spaces between successive neighboring vehicles in the target lane.

In detail, in selecting an optimal space, a survey is conducted to consider the behavior change of the actual driver's car. Based on this, fuzzy rules are generated by using the membership functions for the stability and energy variables. In addition, in determining the safety score and the energy score, which are output variables for the input variables of the fuzzy rule, weights are assigned according to the importance of each input variable. Finally, the space with the highest score is selected as the lane change feasible space (LCFS) by comparing the lane change score values output from each space through the fuzzy rule.

For reference, FIG. 1 is a schematic diagram of a space selection system in which a lane can be changed according to an embodiment of the present invention, and FIG. 2 is a configuration diagram of a space selection system in which a lane can be changed according to an embodiment of the present invention.

3 is a block diagram illustrating a space selection system capable of changing a lane according to an embodiment of the present invention.

Referring to FIG. 3, a space selection system capable of changing a vehicle according to an exemplary embodiment of the present invention includes a determination unit 310, a safety score calculator 320, an energy score calculator 330, and a space selector 340. It includes.

The determination unit 310 may determine the driver's propensity to change the lane with respect to the actual driver.

In an embodiment, the determination unit 310 may receive behavior information regarding the vehicle driving behavior and the lane change behavior from the driver, and determine the propensity of the driver based on the received behavior information.

That is, the determination unit 310 directly receives behavioral information including information on conditions that are important when changing a habit of driving a vehicle or a lane from an individual driver or received through communication with a separate device. The driver's disposition may be determined based on the received behavior information.

In another embodiment, the determination unit 310 may perform a survey consisting of a total of four questions to reflect the vehicle driving behavior of the driver's lane change. In other words, through the questionnaire, the ranking of the variables that the driver considers to be important when changing to the car is examined, and the preferences of the spaces that are considered optimal for changing to the car for typical situations that may occur when driving the vehicle are examined. Degrees and preferences can be analyzed. For reference, in the present embodiment, a survey was conducted on 100 drivers who have had three years or more of actual driving experience.

Before performing the survey, the determination unit 310 may preset five stability variables related to the stability of the vehicle operation and two energy variables related to the efficiency of the vehicle operation, among the variables considered to be important when changing the lane.

For example, the stability variable may be defined as the head gap, which is the distance between the driver's vehicle and the surrounding vehicle driving in different lanes, the collision warning index, which is the possibility of collision between the driver's vehicle and the surrounding vehicle, and the preceding vehicle in each space between the driver's vehicle and the surrounding vehicle. The speed difference may include a first speed deviation, a second speed deviation that is a speed deviation between the driver's vehicle and a trailing vehicle in each space, and a distance between the driver's vehicle and the exit.

For example, the energy variable may include a time required for the driver's vehicle to enter each space and a magnitude of cumulative acceleration required for the driver's vehicle to enter each space.

First, the determination unit 310 may examine the ranking of the safety variable and the energy variable.

The determination unit 310 may examine the ranking of the stability variable including the head gap, the collision warning index, the first speed deviation, the second speed deviation, and the distance to the exit.

Specifically, referring to FIG. 4, the vehicle head spacing represents a distance between two consecutive vehicles among driving vehicles driving in a neighboring target lane and a collision warning index indicating that the vehicle of the driver is a neighboring vehicle. Assuming that the vehicle is located in the center of each space, it represents the degree of collision with the preceding vehicle and the following vehicle, and the first speed deviation is the preceding vehicle in the i-th space S _i of each space between the driver's vehicle and the surrounding vehicle. Represents a speed deviation between the vehicle and the second speed deviation represents a speed deviation between the trailing vehicle in the i-th space S _i of each space between the driver's vehicle and the surrounding vehicle, and 5) the distance from the driver's vehicle to the exit It can represent the distance. For reference, in the drawing, the driver's vehicle is indicated as Sub within the current vehicle.

The determination unit 310 may examine the ranking of the required time and the magnitude of the cumulative acceleration.

In detail, the required time may indicate a time spent for the driver's vehicle to enter each space, and the magnitude of the cumulative acceleration may indicate a magnitude of the acceleration accumulated for the driver's vehicle to enter each space.

Next, the determination unit 310 may examine the preference for the changeable space with the driver's preferred car with respect to a plurality of situations that may occur when driving the vehicle according to the combination of the stability variable and the energy variable.

For example, referring to Table 1, the determination unit 310 may investigate a preference for selecting which space to change to a car according to the remaining distance to the exit when the driver changes the car from his position. . For reference, in the present embodiment, the distance to the exit was set to two cases of 2.5 km or more and 0.5 km or more.

This is to determine how sensitive the driver is to the distance to the exit when changing to a car.

TABLE 1

In the case of the questionnaire, it is assumed that the speed of the driver's vehicle is the same as the speed of the surrounding vehicle, but the head gap indicating the density between the surrounding vehicles in the target lane is the same.

As another example, referring to Table 2, the determination unit 310 is a case where the distance to the exit is sufficient, and the deviation of the vehicle is somewhat different from the average speed of the surrounding vehicles in the target lane to be entered by changing the speed of the driver's vehicle. If present, that is, faster or slower, the preference for which space to select and change to a car can be investigated. In this embodiment, the speed deviation is set to 20 km / h.

This is to find out how sensitive the driver is to the speed deviation in changing to a car.

TABLE 2

In the case of the survey, it is assumed that the head spacing indicating the density between the surrounding vehicles in the target lane is the same but the distance between the driver's vehicle and the exit is more than a predetermined threshold. In this embodiment, the predetermined threshold is set to 2.5 km.

As another example, referring to Table 3, the determination unit 310 has a sufficient distance to the exit, but the density between the surrounding vehicles in the target lane is not uniform, that is, the density between the surrounding vehicles located in front of the surrounding vehicles is high or In the case of high density between neighboring vehicles located in the rear or high density between neighboring vehicles located in the center, the preference for selecting which space to change to a car can be investigated.

This is to find out whether the driver pursues safety and energy efficiency when changing a car because the safety distance between the vehicles is secured even in a high density part. In addition, if the space having the same density to move to the front relatively to change to a car, or to move to the rear to find a preference for changing to a car.

TABLE 3

For reference, in Tables 1 to 3, V _sub refers to the speed of the driver's vehicle, and V _i refers to the speed of the surrounding vehicle.

The results of the ranking and preference survey are as shown in Figs. 5a to 5d.

In the ranking survey, in the case of safety variables, 5 points were selected for the 1st variable, 4 points for the 2nd variable, 3 points for the 3rd variable, 2 points and 5th for the 4th variable. One point was given to the selected variable, and in the case of energy variables, five points were given to the variable selected as the first rank and one point was selected to the variable selected as the second rank.

Referring to FIG. 5A, the actual driver corresponding to the character of the figure shows' distance to exit-> head gap-> collision warning index-> in relation to safety when changing to a car. The second speed deviation-> first speed deviation is considered to be important, and in relation to energy efficiency, it can be seen that the importance is in order of 'time required-> magnitude of cumulative acceleration'.

For reference, in the case of the ranking survey, the results obtained after the sum of all the rankings of the individual respondents corresponding to the actual driver were converted into importance scores and then analyzed.

The result of the preference survey on the changeable space by the driver is as follows.

First, referring to FIG. 5B, when the distance between the driver's vehicle and the exit is 2.5 km or more, more than 52% of the respondents select the next space (see S4 in FIG. 4). In addition, it was confirmed that the response to select a space located relatively forward (see S2 and S3 in FIG. 4) was more than 40% because the distance to the exit was sufficient. On the other hand, when the distance between the driver's vehicle and the exit is 0.5 km, there is a response that more than 67% will select a space located relatively rearward (see S5 to S7 in FIG. 4). However, even in this case, it can be confirmed that there are a large number of 20% of the responses to select the immediately next space (see S4 of FIG. 4).

Next, referring to FIG. 5C, when the speed of the driver's vehicle is about 20 km / h faster than the average speed of the surrounding vehicle, 96% of the snowballs are located relatively forward (see S1 to S3 of FIG. 4). The response was to choose. This can be interpreted as a result of the driver's driving behavior to minimize the instability and energy loss caused by the sudden brake. On the other hand, when the speed of the driver's vehicle is about 20km / h slower than the average speed of the surrounding vehicle, it can be seen that the rear space (refer to S5 to S7 of FIG. 4) is preferred.

Next, referring to FIG. 5D, even though each space between the surrounding vehicles in the target lane satisfies the safety distance, most drivers responded that they would select a space having a relatively low density between the surrounding vehicles rather than the nearest space. . This can be interpreted as a result of placing more emphasis on safety rather than energy loss when selecting lane change spaces. However, it can be seen from the results corresponding to FIG. 5C that the space located in the front is preferred if the density is equally low.

For reference, Figure 5a is a graph showing the results of the ranking survey in the survey in one embodiment of the present invention, Figures 5b to 5d is a graph of the preference survey in the survey in an embodiment of the present invention, It is a graph showing the results.

The safety score calculation unit 320 may calculate a safety score as an output variable for determining stability as an objective value by using the stability variable as an input variable for each space between neighboring vehicles that drive different lanes.

That is, the safety score calculation unit 320 includes 'car head spacing', 'collision alarm index', 'first speed deviation', 'second speed deviation' and 'distance to exit' for each space between surrounding vehicles. The safety score, which is an output variable, can be calculated for each space.

Specifically, the safety score calculation unit 320 obtains the constant input function and the output constant function of the constant type reflecting the weight of the triangular function type and the importance investigation for the stability variable, and obtains the obtained input and output function The safety score can be calculated according to the generated fuzzy rules. For reference, the fuzzy rule developed by TSK (Takagi-Sugeno-Kang) was applied to the fuzzy rule.

As described above, the stability variables in the present exemplary embodiment may be five, and may be input variables of the safety score calculation unit 320 for calculating the safety score, and may be represented by five input variables as follows.

1.Input Variable 1 (In _One )

The first input variable can be expressed as Equation 1 as the head spacing. The unit is m.

[Equation 1]

Here, i and i + 1 denote the preceding vehicle and the trailing vehicle in each space between the surrounding vehicles, x _i denotes the position of the preceding vehicle, and x _{i + 1} denotes the position of the following vehicle.

2. Input Variable 2 (In ₂ )

The second input variable is the collision warning index. Each collision alarm calculated by the driver's vehicle between the preceding vehicle and the following vehicle assuming that the driver's vehicle is located in the center of each space S _i between the surrounding vehicles in the target lane. Choose the smaller of the indices.

First, the collision warning index between the driver's vehicle and the preceding vehicle may be expressed as Equation 2.

[Formula 2]

Here, In _{2, i} means the collision warning index between the driver's vehicle and the preceding vehicle, rd _i means the head spacing between the driver's vehicle and the preceding vehicle, d _{br, Sub} and d _{w, Sub} are respectively It means the minimum stopping distance and the minimum safety distance of the vehicle.

At this time, the minimum stopping distance of the driver's vehicle is equal to Equation 3, and the minimum safe distance of the driver's vehicle is equal to Equation 4.

[Equation 3]

[Equation 4]

Here, v _Sub is the speed of the driver's vehicle, T _{s, delay, Sub} is the operation delay for the brake system of the driver's vehicle and is set to 0.2 seconds considering the brake characteristics, and v _{rel, i} is the driver's vehicle. And a _max means the relative speed between the preceding vehicle, a _max is set to -0.4g as a brake considering the ride comfort, T _{h, delay, Sub} is 0.9 seconds as the time delay before the driver to apply the brake to recognize the preceding vehicle Set.

Next, the collision warning index between the driver's vehicle and the following vehicle is expressed by Equation 5.

[Equation 5]

Here, In _{2, i + 1} means the collision warning index between the driver's vehicle and the trailing vehicle, rd _{i + 1} means the head gap between the driver's vehicle and the trailing vehicle, d _{br, i +1} and d _{w , i + 1} means the minimum stopping distance and the minimum safety distance for the following vehicles, respectively.

At this time, the minimum stopping distance of the following vehicle is the same as Equation 6, the minimum safety distance of the following vehicle is the same as Equation 7.

[Equation 6]

[Formula 7]

은 후행차량의 속도를 의미하고, T_{s,delay,i + 1}는 후행차량의 브레이크 시스템에 대한 작동지연으로서 브레이크 특성을 고려하여 0.2초로 설정하였고, v_{rel,i + 1}는 후행차량 및 운전자의 차량 간의 상대속도를 의미하고, d_{w,i + 1}는 후행차량의 최소안전거리이고, T_{h,delay,i +1}은 운전자가 후행차량을 인지하여 브레이크를 밟을 때까지의 시간지연으로서 0.9초로 설정하였다.Where d _{br, i +1} is the minimum stopping distance of the following vehicle,

Is the speed of the trailing vehicle, T _{s, delay, i + 1} is the operation delay for the brake system of the trailing vehicle and is set to 0.2 seconds considering the brake characteristics, and v _{rel, i + 1} is the speed of the trailing vehicle and the driver. Relative speed between vehicles, d _{w, i + 1} is the minimum safety distance of the trailing vehicles, T _{h, delay, i +1} is the time delay before the driver recognizes the trailing vehicle and brakes. Set.

Finally, In ₂ becomes a smaller value among Equations 2 and 5, as shown in Equation 8.

Equation 8

3. Input Variable 3 (In ₃ )

The third input variable may be expressed as Equation 9 as the first speed deviation. The unit is km / h.

[Equation 9]

Here, v _i is the speed of the preceding vehicle in each space between the surrounding vehicles, and v _sub is the speed of the driver's vehicle.

4. Input Variable 4 (In ₄ )

The fourth input variable can be expressed as Equation 10 as the second speed deviation. The unit is km / h.

Equation 10

Here, v _sub is the speed of the driver's vehicle, and v _{i + 1} is the speed of the trailing vehicle in each space between the surrounding vehicles.

5. Input Variable 5 (In ₅ )

The fifth input variable means the distance from the original driver's vehicle to the exit, but it is calculated as the distance from the exit to the preceding vehicle for individual scoring in each space between the surrounding vehicles in the target lane. Can be. The unit is m.

[Equation 11]

Here, x _exit means the position of the exit, and x _i means the position of the preceding vehicle in each space between the surrounding vehicles.

Meanwhile, the safety score calculator 320 may obtain an input belonging function for the stability variables corresponding to the five input variables.

Specifically, referring to FIG. 6, input variables 1 and 2 are 'Small (S)', 'Medium (M)', and 'Large (L)' for input variables 1 and 2 as input language values of each stability variable. In the case of 'Negative (N)', 'Zero (Z)', and 'Positive (P)', three membership functions are obtained respectively.In case of input variable 5, 'Small (S)', 'Large (L)' You can get 2 membership functions by separating In this embodiment, a trigonometric function is used as the belonging function. For reference, Figure 6 is a graph showing the membership function for the stability variable in an embodiment of the present invention.

In more detail, the membership function for the input variable 1 is classified based on the safety distance L, which is the head distance between the neighboring vehicles according to the speed of the trailing vehicle in each space between the surrounding vehicles, and can be expressed as Equation 12.

Equation 12

Here, L means the safety distance which is the head distance between the surrounding vehicles, i means the i-th space of each space between the surrounding vehicles, d is a safety distance including the full length of the vehicle is set to 5m in this embodiment , h is the concept of the distance between the surrounding vehicles as a speed, and means the time it takes for the trailing vehicle to pass after passing the preceding vehicle among the surrounding vehicles. The farther the head spacing is, the higher the safety of the space is.

The membership function for input variable 2 is classified based on the collision alarm index 1. The larger the collision alarm index, the greater the margin for braking, which means less collision risk. It means.

The membership functions for the input variables 3 and 4 were divided by giving 1 km / h in the deviation size between the speed of the driver's lane and the average speed of the surrounding vehicle in the target lane, which can be expressed as Equation 13. In this case, the reason for allowing 1 km / h is that when all the vehicles have the same speed, v becomes 0, and the meaning of dividing the membership function into N, Z, and P disappears.

Equation 13

Where n is the total number of vehicles in the target lane, and the more stable the velocity deviation of the input variables 3 and 4 is, the more stable the space becomes. It means easy to change by car.

The affiliation function for input variable 5 takes into account that the road junction sign begins to guide for the first time 2.5 km ahead of the junction. In this case, the reason why the belonging function L is set to 2 km or more is that 2.5 km from the driver's vehicle, that the preceding vehicle in the space farthest away from the driver's vehicle in the target lane is 2 km away from the exit. Because.

On the other hand, the safety score calculation unit 320 may obtain an output belonging function for the stability variables corresponding to the five input variables. Here, the membership function takes the form of a constant of the form Sugeno.

To this end, based on the results of the ranking survey in the survey, a weighting system is constructed by giving different weights (S, M, L or N, Z, P) depending on which part of the belonging function belongs to. In other words, the safety score calculation unit 320 in the present embodiment may reflect the weight when determining the output belonging function for the five safety variables in consideration of the importance of the stability variables analyzed through the survey. This can be represented as shown in Table 4.

TABLE 4

In the case of input variable In ₅ with the highest importance ranking, there is a 10-point gap for each belonging function. Was affected a lot. The gaps between sets of input variables 1 to 4 may be calculated through Equations 14 to 17, respectively.

[Equation 14]

Equation 15

[Formula 16]

[Equation 17]

Meanwhile, a fuzzy rule may be generated based on the obtained input belonging function and the output belonging function.

In this embodiment, the number of fuzzy rules for the safety variable can be expressed as shown in Table 5 as a total of 162. The easier and safer to change to a car _{, the} higher the z _{n, 1} output for the fuzzy rule. For each input variable, the AND rule can be applied as follows.

As an example, If In ₁ is 'L' AND In ₂ is 'L' AND In ₃ is 'P' AND In ₄ is 'P' AND In ₅ is 'L', THEN z _1,1 = 100

In the following example, If In ₁ is 'S' AND In ₂ is 'S' AND In ₃ is 'N' AND In ₄ is 'N' AND In ₅ is 'S', THEN z _162,1 = 67.4

TABLE 5

The output value z _{n, 1} for the fuzzy rule set in Table 2 is multiplied by the weight ω _{n, 1} defined as Equation 18, respectively.

Equation 18

Therefore, the safety score Fs _i , which is the final output value when all the fuzzy rules for safety variables are considered, can be expressed by Equation 19.

Equation 19

Here, Fs _i is the safety score, ω _{n, 1} is the weight according to the ranking of the stability variable survey, z _{n, 1} is the output value for the fuzzy rule.

According to an embodiment of the present invention, the space selection system capable of changing the vehicle may further include an energy variable estimating unit 325.

The energy variable estimator 325 may estimate the energy variable in real time to calculate a safety score for determining the energy variable as an objective value. That is, the energy variable estimator 325 may estimate the amount of time required and the cumulative acceleration, which is a time spent for the driver's vehicle to enter each space between the surrounding vehicles.

Specifically, referring to FIG. 7, the energy variable estimator 325 defines a virtual vehicle (GV) that is virtually generated by projecting all the current lanes, which are the lanes in which the driver's vehicle is driven, and the driver's vehicle. It is possible to estimate the amount of time required and cumulative acceleration, which is the time spent entering the space between the virtual vehicles. For reference, FIG. 7 is a schematic diagram illustrating the concept of a virtual vehicle defined when estimating an energy variable in an embodiment of the present invention.

To this end, the energy variable estimator 325 may use RBF KR (Radial Basis Function Kernel Regression) and ANN (Artificial Neural Networks) structures as estimators for estimating energy variables based on previously constructed training input / output data. . For reference, a total of 1120 training input / output data are constructed through simulation based on Matlab and Carsim.

In detail, the energy variable estimator 325 calculates a simulation-based total by considering the case where the driver's vehicle is located in front of the space between the virtual vehicles (①), in the center (②) and in the rear (③), respectively. 1120 training input and output data can be represented as shown in Table 6.

In the case of ①, as the input data, the required time and cumulative acceleration, which are output data, can be stored based on the speed and position of the trailing virtual vehicle in each space. This is because the driver's vehicle moves to a safe place in the front space because it is relatively influenced by a trailing virtual vehicle close to the driver's vehicle in each space.

On the contrary, in the case of ③, since it is greatly influenced by the preceding virtual vehicle in each space, output data can be stored based on the position and speed of the preceding virtual vehicle as input data.

Lastly, in case of ②, since the driver's vehicle is located in the center, it is affected by both the preceding virtual vehicle and the following virtual vehicle, so as to store the output data based on the position and speed of the preceding virtual vehicle and the following virtual vehicle as input data. Can be.

For reference, the output data with respect to the input data in the case of ① to ③ means the magnitude of the required time and the cumulative acceleration, which are energy variables to be estimated in the present embodiment.

TABLE 6

RBF Kernel Regression in the estimator of the present embodiment can estimate the output data for the training input and output data in the case of ①, ③.

Kernel function can be expressed as Equation 20 for RBF (Radial Basis Function).

Equation 20

Here, x _i is a case in which the training input data in advance of ③ 3X1 matrix, [v v _{sub lag lag} x] ^T, ① 3X1 matrix, [v _sub For v _lead x _lead ] ^T , where x ^* represents the input or measured data of interest, and b is the distribution of the trained data, that is, the density of the training input and training output data, estimated according to this value. The ability is different.

In this example, Bowman A.W. and Azzalini's proposed method, which can be expressed as Equation 23 by using Equation 21 and Equation 22.

[Equation 21]

는 각 차원에 속하는 입력데이터의 중간 값을 의미한다. Where p _x is the number of dimensions of the training input data, and in the case of ① and ③, 3 is 3, j means each dimension, n is the number of input data of each dimension,

Means the median value of the input data belonging to each dimension.

Formula 22

는 각 차원에 속하는 출력데이터의 중간 값을 의미한다.Here, p _y is the number of dimensions of the training output data, and in the case of ① and ③, 3 is 3, j means each dimension, n is the number of output data of each dimension,

Means the median value of the output data belonging to each dimension.

Formula 23

Meanwhile, x ^* measured in real time calculates an estimated value y ^* based on Equation 24 through comparison with x _i trained in Table 3.

[Formula 24]

Here, y ^* is an estimated value for the cases 1 and 3, and y _i is [TΣ | a |] ^{T which} is a 2X1 matrix as training output data.

Among the estimators in this embodiment, the ANN can estimate the output data for the training input / output data in the case of ②.

In this case, the input data is a 5 × 1 matrix [v _sub v _lead v _lag x _lead x _lag ] ^T , and the output data is [TΣ | a |] ^{T, which} is a 2 × 1 matrix.

In addition, the 'nnTool' basically provided by Matlab was used, and as shown in FIG. 8, in order to estimate the output data by using two hidden layers having 13 nodes, a multi-layer feedforward neural network was used. ) Design.

In addition, in using ANN, we selected sigmoid function as activation function and fed-forward back propagation method using the steepest descent method as neural network learning method. .

Accordingly, according to the present embodiment, in estimating an energy variable for calculating an energy score, the driving behavior of the driver is more realistically reflected by estimating which position is located in each space between the virtual vehicles of the driver's vehicle. You can.

The energy score calculator 330 may calculate an energy score for determining the efficiency as an objective value based on an energy variable related to the efficiency of vehicle operation in each space of the surrounding vehicle.

That is, the energy score calculation unit 330 may calculate an energy score, which is an output variable, for each space by using an energy variable including 'time required' and '큭 of cumulative acceleration' for each space between surrounding vehicles as input variables. have.

Specifically, the energy score calculation unit 330 obtains the constant input power function of the triangular function and the constant output power function reflecting the weights according to the importance investigation for the energy variable, and obtains the obtained input power function and output power function The energy score can be calculated according to the generated fuzzy rules. For reference, the fuzzy rule developed by TSK (Takagi-Sugeno-Kang) was applied to the fuzzy rule.

As described above, two energy variables in the present embodiment may be input variables of the energy score calculator 330 for calculating an energy score, and may be represented by two input variables as follows.

6. Input variable 6 (In ₆ )

The sixth input variable is the total time taken to move from the position of the driver's vehicle to the safe place in each space between the surrounding vehicles, which is estimated in real time by the estimator in the energy variable estimator 325.

7. Input Variable 7 (In ₇ )

The seventh input variable is the magnitude of the cumulative acceleration required to move from the position of the driver's vehicle to the safe place in each space between the surrounding vehicles. Like the input variable 6, the seventh input variable is a value estimated in real time by the estimator in the energy variable estimator 325. .

Meanwhile, the energy score calculator 330 may obtain an input belonging function for energy variables corresponding to the two input variables.

Specifically, referring to FIG. 9, input variables 6 and 7 as language values of energy variables of each variable are divided into 'Small (S)', 'Medium (M)', and 'Large (L)', respectively. Acquisition of membership functions. In this embodiment, a trigonometric function is used as the belonging function. For reference, FIG. 9 is a graph showing a membership function for an energy variable in an embodiment of the present invention.

In more detail, the membership function for input variable 6 is classified based on the largest time (T) of travel time to each space between surrounding vehicles, and the shorter the time, the less energy consumption, which is the S membership function. The more you belong, the easier it is to change to a car.

The membership function for input variable 7 is classified based on the largest cumulative acceleration magnitude (Σ | a |) needed to move to each space between the surrounding vehicles. The smaller the acceleration magnitude, the lower the energy consumption. It means that it is easier to change the car to belong to the belonging function.

Meanwhile, the energy score calculator 330 may obtain an output belonging function for energy variables corresponding to the two input variables. Here, the membership function takes the form of a constant of the form Sugeno.

To this end, based on the results of the ranking survey in the survey, a weighting system is constructed by giving different weights (S, M, L) according to which part of the belonging function belongs to. In other words, the energy score calculator 330 according to the present exemplary embodiment may reflect the weight when determining the output belonging function for the two energy variables in consideration of the importance of the energy variables analyzed through the survey. This can be represented as shown in Table 7.

TABLE 7

In the case of input variable In ₆ with the highest importance ranking, there is a five-point gap for each belonging function, and the lower the importance rank, the smaller the score gap so that it is affected by the variable with high importance score.

Therefore, it can be calculated as shown in Equation 25 in consideration of the ratio with the score for the input variable 6 as the gap for each set of input variable 7.

[Equation 25]

Meanwhile, a fuzzy rule may be generated based on the obtained input belonging function and the output belonging function.

The fuzzy rules for the safety variable in this embodiment can be expressed as shown in Table 8 as a total of nine. The easier it is to change the car and the lower the energy consumption _{, the} higher the z _{n, 2} output for the fuzzy rule. For each input variable, the AND rule can be applied as follows.

As an example, If In ₆ is 'S' AND In ₇ is 'S', THEN z _1,2 = 100

As another example, If In ₆ is 'L' AND In ₇ is 'L', THEN z _9,2 = 85.8

TABLE 8

The output value z _{n, 2} for the fuzzy rule set in Table 8 is multiplied by the weight ω _{n, 2} defined as Equation 26, respectively.

[Equation 26]

Therefore, the energy score Fe _i , which is the final output value when all the fuzzy rules for the energy variables are considered, can be expressed by Equation 27.

[Equation 27]

Here, Fe _i is the energy score, ω _{n, 2} is the weight according to the ranking of the energy variable survey, z _{n, 2} is the output value for the fuzzy rule.

The space selecting unit 340 may calculate the change score of the lane for each space between the surrounding vehicles by summing the safety score and the energy score, and select the space having the maximum score as the space that can be changed into the car.

In this case, the space selector 340 may reflect a predetermined weight in adding up the safety score and the energy score. In other words, the space selector 340 may calculate the change score by reflecting a greater weight on the safety score than the energy score, based on the fact that the system of the present invention places more weight on stability than energy efficiency. For example, as in Equation 28, the energy score may be given a weight of 0.23 but the safety score may be given a weight of 0.77.

 [Equation 28]

Here, F _i is the change score by the car, Fs _i means the safety score, Fe _i means the energy score.

In addition, when the speed of the vehicle of the driver is faster than the speed of the surrounding vehicle, the space selecting unit 340 may add the speed weight by 1% to the lane change score as the space corresponding to the front. That is, the nearest front space of the driver's vehicle is weighted 1%, and the next closest front space is weighted 2%.

On the contrary, when the speed of the driver's vehicle is slower than the speed of the surrounding vehicle, the space corresponding to the rear side may add a speed weight of 0.5% to the lane change score.

Accordingly, the driver's vehicle selects the space having the highest lane change score among the lane change scores calculated in each space as the optimal space that can be changed into the car.

Hereinafter, some embodiments in which the space selecting unit 340 selects an optimal space that can be changed by a car will be described.

1. The speed of the driver's vehicle is the same as that of the surrounding vehicle. Spacing  Same case

In the above case, if the distance between the driver's vehicle and the exit is greater than or equal to the predetermined threshold, the space between the driver's vehicle and the nearest neighboring vehicle is selected as a space that can be changed by car, and if the distance between the driver's vehicle and the exit is less than the predetermined threshold, the driver's One of the spaces between the surrounding vehicles located behind the vehicle may be selected as a space that can be changed by a car.

For example, referring to FIG. 1, when the distance between the driver's vehicle and the exit is 2.5 km or more, one of S2, S3, and S4 is selected as a space that can be changed to a car, and the distance between the driver's vehicle and the exit is less than 2.5 km. , S5 and S6 can be selected as a space that can be changed by car.

2. Spacing  The same distance between the driver's vehicle and the exit If above

In the above case, when the speed of the driver's vehicle exceeds the average speed of the surrounding vehicle, one of the spaces between the surrounding vehicles located in front of the driver's vehicle is selected as a space that can be changed into a car, and the speed of the driver's vehicle is When the average speed of the surrounding vehicle is less than one, one of the spaces between the surrounding vehicles located behind the driver's vehicle may be selected as a space that can be changed into a car.

For example, referring to FIG. 1, when the speed of the driver's vehicle is about 20 km / h faster than the average speed of the surrounding vehicle, one of S1 and S2 is selected as a space that can be changed into a car, and the speed of the vehicle is around. If it is 20 km / h slower than the average speed of the vehicle, one of S5 and S6 can be selected as a space that can be changed by car.

3. When the speed of the driver's vehicle is the same as that of the surrounding vehicle, but the distance between the driver's vehicle and the exit is more than a predetermined threshold.

In this case, one of the spaces between the surrounding vehicles having a head gap greater than or equal to a predetermined threshold may be selected as a space that can be changed into a car.

On the other hand, according to an embodiment of the present invention, the space selection system that can be changed to a lane further includes a sensor 315 for detecting the position information and speed information of the surrounding vehicle from a sensor provided in the driver's vehicle and a sensor installed around the lane. It may include.

The sensing unit 315 may detect location information and speed information of surrounding vehicles existing in the target lane from the environmental sensor and the road infrastructure included in the driver's vehicle. At this time, the detection unit 315 detects the surrounding vehicles running in the target lane through environmental sensors and road infrastructure such as Radar, RiDAR, Camera, etc. attached to the driver's vehicle, and the position information of each surrounding vehicle through V2X communication. Speed information can be detected.

10 is a flowchart illustrating a space selection method capable of changing a lane according to an embodiment of the present invention.

Referring to FIG. 10, according to an embodiment of the present disclosure, the space selection method capable of changing a lane according to an embodiment of the present disclosure may include determining a driver's propensity to change a lane to a real driver (S1110). Calculating a safety score for determining the stability as an objective value based on the stability variable for the stability of the vehicle operation in each space between the driving vehicles (S1120), and the energy variable for the efficiency of the vehicle operation for each space. On the basis of the step of calculating the energy score for determining the efficiency as an objective value (S1130) and the calculated safety score and the energy score by adding the calculated change score for each space, change the space having the maximum score to the car The step S1140 may be selected as the possible space.

In the determining operation S1110, the driver may receive behavior information regarding his / her vehicle driving behavior and lane change behavior from the driver, and determine the driver's disposition based on the received behavior information.

In addition, in the determining step (S1110) it is possible to determine the driver's disposition by performing a questionnaire on the driving behavior of the driver, specifically, the stability variable and the energy variable for the ranking investigation, and the stability variable And with respect to the plurality of situations that can occur when driving the vehicle according to the combination of the energy variable, the preference survey for the changeable space to the preferred car can be performed.

Here, the stability variables are the head gap, which is the distance between the surrounding vehicles, the collision warning index, which is the possibility of collision between the driver's vehicle and the surrounding vehicle, the first speed deviation, which is the speed deviation between the driver's vehicle and the preceding vehicle, and the driver's vehicle and each The second speed deviation, which is the speed deviation between the trailing vehicles in the space, and the distance between the driver's vehicle and the exit, and the energy variable may include the amount of time and cumulative acceleration, which is the time spent for the driver's vehicle to enter each space. Can be.

Computing the safety score (S1120) is to obtain an input-dependent function of the triangular function and the constant output output function reflecting the weight according to the rank degree survey for the stability variable, the input input function and the obtained output belonging The safety score can be calculated according to the generated fuzzy rules using the function.

On the other hand, according to an embodiment of the present invention, the space selection method that can be changed lanes define a virtual vehicle that is virtually generated by projecting the surrounding vehicle in the lane of the driver's vehicle, and the driver's vehicle into the space between the virtual vehicles The method may further include estimating an energy variable including a time required to enter and a magnitude of the cumulative acceleration (S1125).

In the estimating of the energy variable (S1125), when the driver's vehicle is located at the front or the rear in the space between the vehicle, the energy variable may be estimated based on Equation 24 and Equation 20 below.

[Formula 24]

Equation 20

Here, y ^* is an estimated value of the energy variable, K () means a kernel function, and x ^* indicates the driver's vehicle and the virtual vehicle measured as the driver's vehicle is located in front of or behind each space between the virtual vehicles. Input data including the speed and position, x _i means the pre-trained training input data, the output data for the trained training input data as the driver's vehicle is located in the front or rear in each space between the virtual vehicle, b is a distribution degree of the trained data, and means the density of the training input data and the training output data.

In operation S1125, when the driver's vehicle is located in the center of the space between the virtual vehicles, the energy variable may be estimated using the multilayer neural network structure.

In calculating the energy score (S1130), an input output function having a triangular function and a constant output function which reflects weights according to the importance investigation are obtained for the energy variable, and the obtained input control function and the obtained output control function are obtained. The energy score may be calculated according to the generated fuzzy rule.

On the other hand, according to an embodiment of the present invention, the space selection method that can change lanes further comprises the step (S1115) of detecting the position information and the speed information of the surrounding vehicle from a sensor provided in the driver's vehicle and a sensor installed around the lane It may include.

While specific embodiments of the present invention have been described so far, various modifications are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims below, but also by the equivalents of the claims.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above-described embodiments, which can be variously modified and modified by those skilled in the art to which the present invention pertains. Modifications are possible. Accordingly, the spirit of the present invention should be understood only by the claims set forth below, and all equivalent or equivalent modifications thereof will belong to the scope of the present invention.

310: judgment unit
315: detector
320: safety score calculation unit
325: energy variable estimation unit
330: energy score calculation unit
340: space selection unit

Claims (15)

Determination unit for determining the driver's propensity to change the lane to the actual driver;
A safety score calculator configured to calculate a safety score for determining the stability as an objective value based on a stability variable relating to the stability of the vehicle operation in each space between the driver's vehicle and the surrounding vehicles traveling on different lanes;
An energy score calculator configured to calculate an energy score for determining the efficiency as an objective value based on an energy variable relating to the efficiency of vehicle operation for each space; And
A space selecting unit which calculates a change score of a lane for each space by summing the calculated safety scores and energy scores and selects a space having a maximum score as a space that can be changed by a car.
Including,
Defines a virtual vehicle that is virtually generated by projecting the peripheral vehicle into a lane in which the driver's vehicle runs, and calculates the amount of time required and cumulative acceleration, which is the time required for the driver's vehicle to enter the space between the virtual vehicles. Further comprising an energy variable estimating unit for estimating the energy variable including;
The energy variable estimating unit
And when the driver's vehicle is located in the center of the space between the virtual vehicles, estimating the energy variable using a multi-layer neural network structure.
The method of claim 1,
The determination unit
Receiving behavior information on the vehicle driving behavior and lane change behavior of the driver from the driver, and determining the driver's propensity based on the behavior information.
The method of claim 1,
The determination unit
Determine the driver's disposition by performing a questionnaire on the driving behavior of the driver;
The survey
Ranking of the stability variable and the energy variable; And
Investigation of preferences for a space that can be changed into a vehicle preferred by a driver for a plurality of situations that may occur when driving a vehicle according to a combination of the stability variable and the energy variable
Space selection system that can be changed by a car comprising a.
The method of claim 1,
The stability variable is
A head gap that is a distance between the surrounding vehicles, a collision warning index that is a possibility of collision between the driver's vehicle and the surrounding vehicle, a first speed deviation that is a speed deviation between the driver's vehicle and a preceding vehicle in each space, and the driver's vehicle A second speed deviation, which is a speed deviation between trailing vehicles in each space, and a distance between the driver's vehicle and an exit;
The energy variable is
And a magnitude of time required and cumulative acceleration, which is a time spent for the driver's vehicle to enter each of the spaces.
The method of claim 1,
The safety score calculation unit
The output variable function obtained by reflecting the input variable function in the form of a triangular function and the weight according to the driving behavior of the driver with respect to the stability variable is obtained, and using the obtained input function and the obtained output function And a vehicle changeable space selection system, wherein the safety score is calculated according to the generated fuzzy rule.
delete The method of claim 1,
The energy variable estimating unit
When the vehicle of the driver is located in the front or rear in the space between the virtual vehicle, the space selection system that can be changed to a vehicle, characterized in that for estimating the energy variable based on the following equation (24) and (20).
[Formula 24]

Equation 20

Here, y ^* is an estimated value of the energy variable, K () means a kernel function, x ^* is the driver's vehicle and virtual measured as the driver's vehicle is located in front or rear in each space between the virtual vehicle Input data including the speed and position of the vehicle, x _i is the pre-trained training input data as the driver's vehicle is located in the front or rear in each space between the virtual vehicle, y _i is the training data input B means the density of the training input data and the training output data as a distribution degree of the trained data.
delete The method of claim 1,
The energy score calculation unit
Obtaining an input variable function in the form of a trigonometric function and a weighted constant according to the importance of the energy variable is obtained for the energy variable, and using the obtained input function and the obtained output function And a space changeable system according to claim 1, wherein the energy score is calculated according to the generated fuzzy rule.
The method of claim 4, wherein
The space selecting unit is the speed of the vehicle of the driver is the same as the speed of the surrounding vehicle, but the head spacing is the same,
If the distance between the vehicle of the driver and the exit is greater than or equal to a predetermined threshold, the space between the driver's vehicle and the surrounding vehicle closest to the vehicle is selected as a space that can be changed into the vehicle.
When the distance between the driver's vehicle and the exit is less than a predetermined threshold, one of the spaces between the surrounding vehicles located behind the driver's vehicle is selected as the space that can be changed into the vehicle system.
The method of claim 4, wherein
The space selecting unit is the same as the head spacing but the distance between the driver's vehicle and the exit is more than a predetermined threshold,
If the speed of the vehicle of the driver exceeds the average speed of the surrounding vehicle, select one of the spaces between the surrounding vehicles located in front of the vehicle of the driver as the space that can be changed to the vehicle,
When the speed of the vehicle of the driver is less than the average speed of the surrounding vehicle, the change to the lane, characterized in that for selecting one of the spaces between the surrounding vehicles located in the rear with respect to the vehicle of the driver as a space that can be changed to the vehicle Possible space selection system.
The method of claim 4, wherein
The space selecting unit is the speed of the vehicle of the driver is the same as the speed of the surrounding vehicle, but the distance between the vehicle of the driver and the exit is more than a predetermined threshold,
And selecting one of the spaces between the surrounding vehicles having the headspace interval greater than or equal to a predetermined threshold as a space that can be changed into the vehicle.
The method of claim 1,
Sensing unit for detecting the position information and the speed information of the surrounding vehicle from the sensor provided in the driver's vehicle and the sensor installed around the lane
Space selection system that can be changed by the car further comprises a.
Performing a survey for reflecting the driver's propensity to change the lane with respect to the actual driver;
Calculating a safety score for determining the stability as an objective value, based on a stability variable relating to the stability of the vehicle operation in each space between the driver's vehicle and the surrounding vehicle traveling on different lanes;
Calculating an energy score for determining the efficiency as an objective value based on an energy variable relating to the efficiency of vehicle operation for each space; And
Calculating the change score for each lane by summing the calculated safety scores and energy scores, and selecting a space having the maximum score as a space that can be changed to a vehicle;
Including,
Defines a virtual vehicle that is virtually generated by projecting the peripheral vehicle into a lane in which the driver's vehicle runs, and calculates the amount of time required and cumulative acceleration, which is the time required for the driver's vehicle to enter the space between the virtual vehicles. Estimating the energy variable comprising:
Estimating the energy variable
And when the driver's vehicle is located in the center of the space between the virtual vehicles, estimating the energy variable using a multi-layer neural network structure.
delete

Images