JP2016143240A - Lane change optimization device and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform lane change appropriately.SOLUTION: An information obtaining part 26 obtains a travel condition of an own vehicle and a travel condition of each of peripheral vehicles present around the own vehicle which are travelling on each lane in the direction in which the own vehicle is travelling. A peripheral vehicle behavior prediction part 28 predicts a time sequence of behaviors of each of the peripheral vehicles according to the travel condition of each of the peripheral vehicles. A lower layer optimization part 320 creates a time sequence of the acceleration of the own vehicle according to the obtained travel condition of the own vehicle and the predicted time sequence of the behaviors of each of the peripheral vehicles at every timing when the own vehicle changes a lane in such a way that a first evaluation function is optimized. An upper layer optimization part 322 calculates a combination of the time sequence of the acceleration of the own vehicle and the lane change timing of the own vehicle according to the time sequence of the acceleration of the own vehicle created by the lower layer optimization part 320 at every lane change timing in such a way that a second evaluation function is optimized.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a lane change optimization device and program.

  Conventionally, a system that automatically changes lanes is known (Patent Document 1). In the technique described in Patent Document 1, an accelerator / brake and a steering angle are automatically obtained for driving on a given route, and a lane change is automatically started after observing an obstacle, a lane end point, a slow vehicle, and the like. . This algorithm finds a possible space and changes lanes on a safe route.

  In addition, a system is known that supports a driver by finding a safe lane change route according to a traffic merging end, a lane end point, or a driver's request (Patent Document 2). In the system described in Patent Document 2, the behavior of surrounding vehicles is predicted, and time series variables until completion of merging are generated and transmitted to the driver. It also analyzes the generated time series variables and tells the driver whether there is a safe distance for lane changes.

US Patent Application Publication No. 2014/0207325 US Patent Application Publication No. 2005/0015203

  In the technique described in Patent Document 1, since the lane change is started after observing an obstacle, a lane end point, a slow vehicle, and the like, it is often necessary to decelerate the vehicle due to an observation delay, and the efficiency deteriorates. Changing lanes at a reduced speed is a cause of bottlenecks and affects traffic flow.

  Further, in the technique described in Patent Document 2, a possible space is found and the lane is changed according to some rules, but the efficiency in terms of travel time is not taken into consideration. Further, since the control is not based on long-term prediction, it is impossible to change the lane in advance without reducing the speed.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide a lane change optimization device and a program that can appropriately change lanes.

  In order to achieve the above object, the lane change optimizing device of the present invention travels in each lane having the same traveling direction as that of the host vehicle, and each of the surrounding vehicles existing around the host vehicle. Information acquisition means for acquiring a running state; and surrounding vehicle behavior prediction means for predicting a time series of the behavior of each of the surrounding vehicles based on the running state of each of the surrounding vehicles acquired by the information acquisition means; , Based on the travel state of the host vehicle acquired by the information acquiring unit and the time series of the behavior of each of the surrounding vehicles predicted by the surrounding vehicle behavior predicting unit. In addition, first optimization means for generating a time series of acceleration of the host vehicle so as to optimize a first evaluation function including an evaluation function relating to a safe lane change in traveling of the host vehicle, and the first optimization Based on a time series of the acceleration of the host vehicle generated at each lane change timing by the means, the second evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle is optimized. Second optimization means for obtaining a combination of a time series of vehicle acceleration and the timing of lane change of the host vehicle.

  The program according to the present invention uses a computer to acquire information on the traveling state of the host vehicle and the traveling states of the surrounding vehicles existing in the vicinity of the host vehicle traveling in the lanes having the same traveling direction as the host lane. A surrounding vehicle behavior predicting means for predicting a time series of the behavior of each of the surrounding vehicles based on the running state of each of the surrounding vehicles acquired by the information acquiring means, and the self-acquisition acquired by the information acquiring means. Based on the running state of the vehicle and the time series of the behavior of each of the surrounding vehicles predicted by the surrounding vehicle behavior predicting means, a safe lane change in the traveling of the own vehicle at each timing of the lane change of the own vehicle First optimization means for generating a time series of acceleration of the host vehicle so as to optimize a first evaluation function including an evaluation function relating to the vehicle, and a lane by the first optimization means Based on the time series of the acceleration of the host vehicle generated at each further timing, the acceleration of the host vehicle is optimized so as to optimize a second evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle. Is a program for functioning as a second optimizing means for obtaining a combination of the time series and the timing of changing the lane of the host vehicle.

  According to the present invention, the information acquisition means acquires the traveling state of the host vehicle and the traveling states of the surrounding vehicles that are traveling in the same lane as the own lane and that exist around the host vehicle. Then, the time series of the behaviors of the surrounding vehicles is predicted by the surrounding vehicle behavior prediction means based on the running states of the surrounding vehicles acquired by the information acquisition means.

  Then, based on the travel state of the host vehicle acquired by the information acquisition unit by the first optimization unit and the time series of the behaviors of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit, For each lane change timing, a time series of the acceleration of the host vehicle is generated so as to optimize a first evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle.

  Then, an evaluation function related to a safe lane change in traveling of the host vehicle is obtained based on the time series of the acceleration of the host vehicle generated by the second optimizer at each lane change timing by the first optimizer. A combination of the time series of the acceleration of the host vehicle and the timing of the lane change of the host vehicle is obtained so as to optimize the second evaluation function including the second evaluation function.

  In this way, the time series of the behavior of each of the surrounding vehicles is predicted based on the running state of each of the surrounding vehicles, and based on the running state of the own vehicle and the predicted time series of each behavior of the surrounding vehicles. Then, a time series of the acceleration of the host vehicle is generated so as to optimize the first evaluation function at each timing of lane change of the host vehicle, and a time series of the acceleration of the host vehicle generated at each timing of lane change Based on the above, the lane change can be appropriately performed by obtaining a combination of the acceleration time series of the host vehicle and the timing of the lane change of the host vehicle so as to optimize the second evaluation function.

  The first evaluation function related to traveling of the host vehicle according to the present invention may include an evaluation function related to safe traveling of the host vehicle.

  The first evaluation function related to traveling of the host vehicle according to the present invention may include an evaluation function related to energy efficiency of the host vehicle.

  The first optimization means according to the present invention performs, in parallel, processing for generating a time series of acceleration of the host vehicle so as to optimize the first evaluation function at each timing of lane change of the host vehicle. Can be.

  The second evaluation function related to traveling of the host vehicle according to the present invention may include the first evaluation function.

  The second evaluation function related to traveling of the host vehicle according to the present invention may be the sum of the first evaluation function and an evaluation function related to the behavior of the host vehicle in a predetermined maximum prediction step.

  The first optimizing unit according to the present invention includes a travel state of the host vehicle acquired by the information acquiring unit, and a time series of behaviors of the surrounding vehicles predicted by the surrounding vehicle behavior predicting unit. And generating a time series of acceleration of the host vehicle so as to optimize the first evaluation function for each lane change timing of the host vehicle and each lane to be changed, and the second optimization means The own vehicle is optimized so as to optimize the second evaluation function based on the time series of the lane change and the time series of the acceleration of the own vehicle generated for each lane to be changed by the first optimization means. It is possible to obtain a combination of the acceleration time series, the lane change timing of the host vehicle, and the change destination lane.

  The storage medium for storing the program of the present invention is not particularly limited, and may be a hard disk or a ROM. Further, it may be a CD-ROM, a DVD disk, a magneto-optical disk or an IC card. Furthermore, the program may be downloaded from a server or the like connected to the network.

  As described above, according to the lane change optimization device and program of the present invention, based on the running state of each surrounding vehicle, the time series of the behavior of each surrounding vehicle is predicted, Generating a time series of the acceleration of the own vehicle so as to optimize the first evaluation function at each timing of the lane change of the own vehicle based on the predicted time series of each behavior of the surrounding vehicle, Based on the time series of the acceleration of the host vehicle generated at each lane change timing, the combination of the time series of the host vehicle acceleration and the timing of the lane change of the host vehicle is optimized so as to optimize the second evaluation function. By obtaining, the effect that a lane change can be performed appropriately is acquired.

It is a block diagram which shows the lane change optimization apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows an example of the lane which each of the own vehicle and a surrounding vehicle drive | works. 3 is a block diagram illustrating a detailed configuration example of an optimization unit 32. FIG. It is explanatory drawing for demonstrating (delta) showing the timing of the lane change of the own vehicle, and the lane of a change destination. It is explanatory drawing for demonstrating the case where a parameter is set in consideration of the speed of a surrounding vehicle. It is a figure which shows an example of the lane change which considered the destination. It is a flowchart which shows the content of the lane change optimization process routine in the computer of the lane change optimization apparatus which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the content of the surrounding vehicle behavior prediction process in the computer of the lane change optimization apparatus which concerns on the 1st Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the present embodiment, the case where the present invention is applied to a lane change optimization device that estimates the combination of the time series of the acceleration of the host vehicle and the timing of the lane change of the host vehicle and outputs it to the vehicle control device. Explained as an example. In the present embodiment, a case where the host vehicle is an autonomous driving vehicle will be described as an example.

[Outline of Embodiment of the Present Invention]
In the embodiment of the present invention, the goal is to perform safe and efficient automatic driving while changing lanes according to traffic conditions. Specifically, the optimum accelerator or brake operation and the optimum timing for changing the lane are obtained in consideration of the future behavior of the surrounding vehicles traveling in the lanes around the host vehicle.

  The evaluation function for the optimization problem includes items related to driving efficiency and safe lane change, and the optimal driving behavior for a long time is calculated while satisfying constraints such as danger avoidance. As a result, lane changes and speeds are adjusted in advance to avoid traffic jams and realize highly efficient driving. In the present embodiment, a control solution is obtained at high speed by setting a problem that can be calculated in parallel and two-stage optimization.

<System configuration>
As shown in FIG. 1, a lane change optimization device 10 according to a first embodiment of the present invention includes an input operation unit 12 for a driver to input a destination, and an inter-vehicle communication unit that performs inter-vehicle communication. 14, a vehicle speed sensor 16 that sequentially detects the speed of the host vehicle, an acceleration sensor 18 that sequentially detects the acceleration of the host vehicle, a position measuring unit 20 that sequentially measures the position of the host vehicle, road network data, and map information. Based on the stored road network database 22 and the inter-vehicle communication data obtained by the inter-vehicle communication unit 14, a combination of the time series of the acceleration of the own vehicle and the timing of the lane change of the own vehicle is obtained, and the vehicle control A combination of the computer 24 output to the device 36, the time series of the acceleration of the host vehicle output by the computer 24, and the timing of the lane change of the host vehicle is realized. In so that, and a vehicle control device 36 for controlling the driving operation of the vehicle, a.

  The inter-vehicle communication unit 14 receives the traveling state of each of the surrounding vehicles existing around the host vehicle through inter-vehicle communication. FIG. 2 shows an example of each of the surrounding vehicles existing around the host vehicle. As shown in FIG. 2, in the present embodiment, each of the surrounding vehicles will be described as an example in which each vehicle travels in each lane that has the same traveling direction as the own lane.

  FIG. 2 shows an example where the lanes having the same traveling direction are three lanes. H represents the host vehicle, p1, p2,..., PN represent each of the surrounding vehicles traveling in the same lane as the host vehicle, and q1, q2,. Each of the surrounding vehicles traveling in the lane represents each of the surrounding vehicles, and r1, r2,. N is a predetermined number.

  For example, the vehicle-to-vehicle communication unit 14 is configured for each of the surrounding vehicles (p1, p2,..., PN, q1, q2,..., QN, r1, r2,..., RN) shown in FIG. As the running state, the positions and speeds of the surrounding vehicles are received by inter-vehicle communication.

  The position measurement unit 20 is configured using, for example, a GPS sensor, and measures the current position of the host vehicle.

  The computer 24 includes a CPU, a RAM, and a ROM that stores a program for executing a lane change optimization process routine to be described later, and is functionally configured as follows. The computer 24 receives the destination information input by the input operation unit 12, the position and speed of each of the surrounding vehicles received by the inter-vehicle communication unit 14, the speed and acceleration of the host vehicle detected by the vehicle speed sensor 16. An information acquisition unit 26 for acquiring the acceleration of the host vehicle detected by the sensor 18, the position of the host vehicle measured by the position measurement unit 20, and road network data and map information stored in the road network database 22; Based on the position and speed of each of the surrounding vehicles received by the information acquisition unit 26, the surrounding vehicle behavior prediction unit 28 that predicts the time series of the behavior of each of the surrounding vehicles, and the purpose acquired by the information acquisition unit 26 Generates a route from the current position to the destination based on the location information, own vehicle location, road network data, and map information Generated by the route generation unit 30, the speed and acceleration of the host vehicle acquired by the information acquisition unit 26, the time series of each behavior of the surrounding vehicle predicted by the surrounding vehicle behavior prediction unit 28, and the route generation unit 30 An optimization unit 32 that obtains a combination of the time series of the acceleration of the host vehicle and the timing of the lane change of the host vehicle based on the determined route, and the time series of the acceleration of the host vehicle calculated by the optimization unit 32 And an output unit 34 for outputting the timing of changing the lane of the host vehicle.

  The information acquisition unit 26 uses the speed of the host vehicle detected by the vehicle speed sensor 16, the acceleration of the host vehicle detected by the acceleration sensor 18, and the position of the host vehicle measured by the position measurement unit 20. Acquired as the running state. Moreover, the information acquisition part 26 acquires each position and speed of a surrounding vehicle as each driving state of a surrounding vehicle.

  The information acquisition unit 26 also includes destination information input by the input operation unit 12, the position of the host vehicle measured by the position measurement unit 20, road network data and map information stored in the road network database 22. Are output to the route generation unit 30.

  The surrounding vehicle behavior prediction unit 28 predicts the time series of the positions of the surrounding vehicles as the time series of the behaviors of the surrounding vehicles based on the positions and speeds of the surrounding vehicles acquired by the information acquisition unit 26. To do.

  For example, the surrounding vehicle behavior prediction unit 28 determines the relative distance and relative distance between the surrounding vehicle and the preceding vehicle of the surrounding vehicle based on the position and speed of the surrounding vehicle acquired by the information acquisition unit 26 for each of the surrounding vehicles. Estimate speed. Then, the surrounding vehicle behavior prediction unit 28 calculates the following expression (1) for each of the surrounding vehicles based on the speed of the surrounding vehicle acquired by the information acquisition unit 26 and the estimated relative distance and relative speed. As shown, the time series of each acceleration of surrounding vehicles is predicted.

In the above equation (1), p represents the pth lane, and k (= 1 to N) represents the kth surrounding vehicle. Therefore, “p _k ” represents the k th surrounding vehicle traveling in the p th lane. Also,

Represents the k-th peripheral vehicle p _k traveling the p-th lane, the distance to the preceding vehicle of the surrounding vehicle p _k. Also,

Represents the speed of the _kth surrounding vehicle pk traveling in the pth lane. Also,

Represents the k-th peripheral vehicle p _k traveling the p-th lane, the relative speed between the preceding vehicle of the surrounding vehicle p _k.

F _{a in the} above formula (1) is a driving model of the driver of the vehicle. For example, an intelligent driver model (IDM) shown in the following formula (2) can be used as f _a .

Here, α, β, R ₀ , v ₀ and t _h are predetermined parameters.

  In addition, the surrounding vehicle behavior prediction unit 28 predicts the time series of the positions of the surrounding vehicles according to the following equation (3) based on the predicted time series of the accelerations of the surrounding vehicles.

  The route generation unit 30 generates a route from the current position to the destination based on the destination information output by the information acquisition unit 26, the position of the host vehicle, road network data, and map information.

  The optimization unit 32 determines the acceleration of the host vehicle based on the position and speed of the host vehicle acquired by the information acquisition unit 26 and the time series of the positions of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit 28. The combination of the time series and the timing of changing the lane of the host vehicle is obtained. As shown in FIG. 3, the optimization unit 32 includes a lower layer optimization unit 320 and an upper layer optimization unit 322. The lower layer optimization unit 320 is an example of a first optimization unit, and the upper layer optimization unit 322 is an example of a second optimization unit.

  In the present embodiment, the state variable of the host vehicle is defined as shown in the following equation (4).

Note that x _h εR represents a position, v _h εR represents a speed, and l _h ε {-1, 0, +1} represents a lane in which the host vehicle travels. l _h = 0 represents the lane in which the host vehicle is currently traveling, l _h = 1 represents the lane on the right side of the lane in which the host vehicle is currently traveling, and l _h = −1 is the left side of the lane in which the host vehicle is currently traveling. Represents a lane. Y _h ∈R represents the lateral position of the vehicle.

Similarly to each of the surrounding vehicles, with respect to the behavior of the own vehicle, the speed of the own vehicle acquired by the information acquisition unit 26, the acceleration of the own vehicle, the position of the own vehicle, and a lower layer optimization unit 320 described later. Based on the time series u of the acceleration of the host vehicle calculated in step (1), the time series of the position x _{h and} the time series of the speed v _h of the host vehicle are calculated according to the above equation (3).

  Further, the lane change of the host vehicle during the discontinuous time can be expressed by the following equation (5).

Here, in the above equation (5), δ _l and δ _r are binary variables. δ _l is a variable indicating whether or not to change to the left lane, and δ _r is a variable indicating whether or not to change to the right lane. FIG. 4 is an explanatory diagram of δ _l and δ _r . As shown in FIG. 4, at each time, δ _l and δ _r take a value of 0 or 1. The time series of δ _l and δ _r represent the lane change timing and the lane to be changed. In addition, δ _l and δ _{r at} each time have a constraint _expressed by the following equation (6).

It is assumed that the lateral position y _h of the host vehicle is obtained in a time series of δ _l and δ _r . Further, the acceleration u _h , the speed v _h , and the position x _{h at} each time satisfy the constraint condition shown in the following formula (7) for speed limitation, comfortable driving, and safety.

Here, the constant t _h represents the vehicle head time, and R ₀ represents the minimum inter-vehicle distance.

  The optimization unit 32 solves the T-step horizon zone optimization problem in the framework of model predictive control. Specifically, the optimization unit 32 is based on the position and speed of the host vehicle acquired by the information acquisition unit 26 and the time series of the positions of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit 28. Then, the combination of the time series u of the acceleration of the host vehicle shown in the following formula (8) and δ representing the lane change timing of the host vehicle and the destination lane shown in the following formula (9) is obtained.

  In the present embodiment, since the solution is obtained at high speed by the two-stage optimization process of the optimization process by the lower layer optimization unit 320 and the optimization process by the upper layer optimization unit 322, the optimization problem is described below. Set as shown.

  Specifically, in the horizon zone T, as shown in the following formula (10), a constraint condition is defined such that the lane change is not performed more than once.

Then, the lower layer optimization unit 320 calculates the time series u ^* of the acceleration of the host vehicle (optimum acceleration), and the upper layer optimization unit 322 calculates the lane change timing (optimum lane change timing) and the change destination. Δ ^* representing the lane is calculated.

  Based on the position and speed of the host vehicle acquired by the information acquisition unit 26 and the time series of the positions of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit 28, the lower layer optimization unit 320 A time series of the acceleration of the host vehicle is generated so as to optimize the first evaluation function for each lane change timing and each lane to be changed. The first evaluation function includes an evaluation function related to a safe lane change during traveling of the host vehicle.

  Further, the lower layer optimization unit 320 generates a time series of the acceleration of the host vehicle so as to optimize the first evaluation function for each δ representing the timing of the lane change of the host vehicle and the lane of the change destination. Do in parallel.

Specifically, the lower layer optimization unit 320 fixes the δ represents the lane timing and change destination of lane change of the vehicle, while satisfying the constraint condition g ¹ predetermined by the following formula (11) as shown in, for calculating a time series u acceleration of the vehicle so as to minimize the first evaluation function J ^1.

In this embodiment, the constraint g ¹ in the formula (11), the speed shown in the above formula (7), the acceleration, and corresponds to the constraint on the inter-vehicle distance.

For example, the lower layer optimization unit 320 calculates the time series u of the acceleration of the host vehicle that minimizes the first evaluation function J ¹ as shown in the following formula (12).

The first evaluation function J ¹ shown in the above formula (12), the evaluation function L _a on energy efficiency of the vehicle, the evaluation function L _b for the safe running of the vehicle, for the safe lane change in the running of the vehicle And an evaluation function L _c .

The evaluation function L _a regarding the energy efficiency of the host vehicle, the evaluation function L _b regarding the safe travel of the host vehicle, and the evaluation function L _c regarding the safe lane change during the travel of the host vehicle are, for example, the following formula (13): It can represent with-(15).

In the above equation (13), V _d is a desired speed, and w _v and w _u are constants. In the above formula (14), Δx _h represents the inter-vehicle distance, t _d represents the desired vehicle head time, and α _f and w _f are constants. In the above equation (15), α _c and w _c are constants, and x to _j represent the positions of the surrounding vehicles in the change destination lane (the change destination lane is determined by δ).

Further, the lower layer optimization unit 320 sets each parameter in the above formulas (13) to (15) based on the time series of the positions of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit 28. For example, as shown in FIG. 5, when the host vehicle H changes the lane so as to interrupt between the surrounding vehicle A and the surrounding vehicle B, the lower layer optimization unit 320 uses the constant α in the above equation (15). Set _c . By setting the constant α _c , the evaluation function corresponding to the relative distance (x _h (τ) −x _j ~ (τ)) between the host vehicle and each of the surrounding vehicles is optimized.

  The upper layer optimization unit 322 includes the time series of the lane change generated by the lower layer optimization unit 320 and the acceleration of the own vehicle generated for each δ representing the lane to be changed, and the current generation generated by the route generation unit 30. Based on the route from the position to the destination, the time series u of the acceleration of the host vehicle, and δ representing the timing of the lane change of the host vehicle and the lane of the change destination so as to optimize the second evaluation function Find a combination. The second evaluation function includes an evaluation function related to a safe lane change during traveling of the host vehicle. Further, the second evaluation function includes the first evaluation function shown in the above equation (11).

Specifically, the upper layer optimization unit 322, while satisfying the constraint condition g ² regarding a lane change, to minimize the second evaluation function as shown in the following equation (16), the timing of lane change and Δ representing the lane to be changed is obtained. That is, the upper layer optimization unit 322 is based on the time series of the acceleration of the host vehicle generated by the lower layer optimization unit 320 for each δ representing the timing of the lane change of the host vehicle and the lane of the change destination. The δ representing the lane change timing of the host vehicle and the lane to be changed and the time series u of the corresponding acceleration are calculated so that the second evaluation function shown in the following equation (16) is minimized.

For example, the second evaluation function can be set to be the sum of the first evaluation function and the evaluation function E (z _h (t + T)) related to the behavior of the host vehicle in a predetermined maximum prediction step. Upper layer optimization unit 322 is, for example, to minimize the second evaluation function J ² as shown in the following equation (17), obtaining the δ represents the lane timing and change destination lane change.

Here, the evaluation function E (z _h (t + T)) relating to the behavior of the host vehicle is an expected value of the terminal cost, and is an optional part that is expanded in consideration of road conditions and routes. By setting the second evaluation function J ² as _{E (z h (t + T} )) is included, [delta] is calculated representing the lane timing and change destination of lane change in consideration of the travel time.

For example, as an example of E (z _h (t + T)), a function represented by the following formula (18) can be used.

V￣ (l _h , t + T) in the above equation (18) is a speed that can be expected after T steps, the time series of the acceleration of the host vehicle calculated by the lower layer optimization unit 320, the timing and change of the lane change It is calculated based on the combination with δ representing the previous lane and the route from the current position to the destination. FIG. 6 is a diagram for explaining V ￣ (l _h , t + T). For example, as shown in FIG. 6, the upper layer optimization unit 322 is based on a route from the current position generated by the route generation unit 30 to the destination, and a lane that leads to a different direction from the route to the destination. expected speed V¯ of the _(l h, t + T) is set to 0. By setting the expected speed V￣ (l _h , t + T) in a lane that leads to a direction different from the route to 0, the value of E (z _h (t + T)) increases, and the lane leads to the destination direction. Lane change is prompted.

  The output unit 34 outputs a combination of the time series of the acceleration of the host vehicle calculated by the optimization unit 32, the timing of changing the lane of the host vehicle, and the lane to be changed.

The vehicle control device 36 controls the driving operation of the vehicle so as to realize a combination of the time series of the acceleration of the host vehicle output by the output unit 34 and the timing of the lane change of the host vehicle and the lane to be changed. Do. Specifically, the vehicle control device 36 determines the steering angle of the vehicle from δ representing the lane change timing output by the output unit 34 and the lane to be changed, and controls the driving operation of the vehicle. Further, the vehicle control device 36 obtains engine torque based on the time series of the acceleration of the host vehicle output by the output unit 34 and controls the driving operation of the vehicle. Note that the delay with respect to the communication time and the calculation time is, for example, the constraint condition “u _h (t | _t ) = u _h (t + 1 | _t ) where the acceleration of the host vehicle obtained one step before is used at the current time. _-1 ) "can be used to eliminate delays in communication time and calculation time.

<Operation of lane change optimization device>
Next, the operation of the present embodiment will be described.

  When a vehicle equipped with the lane change optimization device 10 is running, data is received from the surrounding vehicle by the inter-vehicle communication unit 14, the speed of the host vehicle is sequentially detected by the vehicle speed sensor 16, and the acceleration sensor 18 is detected. When the destination information is input by the input operation unit 12 when the acceleration of the host vehicle is sequentially detected by the position measurement unit 20 and the position of the host vehicle is sequentially measured by the position measurement unit 20, the lane change optimization device 10 The lane change optimization process routine shown in FIG. 7 is executed.

  First, in step S100, each variable included in the above equations (1) to (18) is initialized.

  Next, in step S <b> 102, the information acquisition unit 26 detects the speed of the host vehicle detected by the vehicle speed sensor 16, the acceleration of the host vehicle detected by the acceleration sensor 18, and the host vehicle measured by the position measurement unit 20. The position is acquired as the traveling state of the host vehicle. The information acquisition unit 26 also includes destination information input by the input operation unit 12, the position of the host vehicle measured by the position measurement unit 20, road network data and map information stored in the road network database 22. Are output to the route generation unit 30.

  In step S104, the information acquisition unit 26 acquires the positions and speeds of the surrounding vehicles as the running states of the surrounding vehicles.

  In step S106, the route generation unit 30 generates a route from the current position to the destination based on the destination information output in step S102, the position of the host vehicle, road network data, and map information.

  In step S108, the surrounding vehicle behavior prediction unit 28 predicts the time series of the behaviors of the surrounding vehicles based on the positions and speeds of the surrounding vehicles acquired in step S104. Step S108 is realized by the surrounding vehicle behavior prediction processing routine shown in FIG.

<Neighboring vehicle behavior prediction processing routine>
First, in step S200, the position and speed of each of the surrounding vehicles at the current time t acquired in step S102 are set as an initial state.

  Next, in step S202, based on the initial state set in step S200, the relative distance and relative speed between the surrounding vehicle and the preceding vehicle of the surrounding vehicle are estimated for each of the surrounding vehicles. Then, based on the speed of the surrounding vehicle set as the initial state in step S200 and the estimated relative distance and relative speed, the time series of the acceleration of each surrounding vehicle is predicted according to the above equation (1).

  Next, in step S204, the position of each surrounding vehicle at the next time (t + 1) is estimated according to the above equation (3) based on the time series of the acceleration of each surrounding vehicle predicted in step S202. .

  In step S206, it is determined whether the prediction target time is after t + T, where T is the horizon zone. When the prediction target time is after t + T, the process proceeds to step S208. On the other hand, if the prediction target time is before t + T, the process returns to step S202 with the prediction target time as the next time in step S210.

  In step S208, the time series of the positions of the surrounding vehicles estimated in step S204 is output as a prediction result.

  Next, returning to the lane change optimization processing routine, in step S110, the lower layer optimization unit 320, based on the time series of the positions of the surrounding vehicles output in step S208, the above formulas (13) to (13). Each parameter in (15) is set.

  In step S112, the lower layer optimization unit 320 determines the own vehicle based on the position and speed of the own vehicle acquired in step S102 and the time series of the positions of the surrounding vehicles predicted in step S108. A time series of the acceleration of the host vehicle is generated so as to optimize the first evaluation function shown in the above equation (12) for each δ representing the timing of the lane change and the lane to be changed.

  In step S114, the upper layer optimization unit 322 determines the own vehicle after time T calculated from the time series of the acceleration of the own vehicle generated for each δ representing the lane change timing and the lane to be changed in step S112. And the time series of the acceleration of the host vehicle so as to optimize the second evaluation function shown in the equation (17) based on the behavior of the vehicle and the route from the current position to the destination generated in step S106. A combination of the lane change timing of the host vehicle and δ representing the change destination lane is obtained.

  In step S116, the output unit 34 outputs a combination of the time series of the acceleration of the host vehicle calculated in step S114 and δ representing the lane change timing of the host vehicle and the lane to be changed to the vehicle control device 36. Then, the process returns to step S100.

  By repeatedly executing the above lane change optimization processing routine, a combination of the time series of the acceleration of the own vehicle generated sequentially and δ representing the lane change timing of the own vehicle and the lane of the change destination is The vehicle is output to the control device 36, and the vehicle control device 36 performs a driving operation of the vehicle based on a combination of the time series of the acceleration of the own vehicle and δ representing the lane change timing of the own vehicle and the lane to be changed.

  As described above, according to the lane change optimization device according to the embodiment of the present invention, the time series of the behavior of each of the surrounding vehicles is predicted based on the traveling state of each of the surrounding vehicles. Based on the traveling state and the predicted time series of each of the surrounding vehicles, the time series of the acceleration of the host vehicle is optimized so as to optimize the first evaluation function at each timing of the lane change of the host vehicle. And generating a time series of acceleration of the host vehicle so as to optimize the second evaluation function based on the time series of the acceleration of the host vehicle generated for each δ representing the timing of the lane change and the lane to be changed to The lane change can be appropriately performed by obtaining a combination of the timing of changing the lane of the host vehicle and δ representing the lane to be changed.

  In addition, efficient driving on a multi-lane road can be realized, and fuel efficiency can be improved and travel time can be significantly shortened compared to conventional methods.

  In addition, the vehicle can be automatically driven in real time using an optimization algorithm capable of parallel calculation.

  Note that the present invention is not limited to the embodiment described with reference to the drawings, and various modifications can be applied without departing from the scope of the invention.

  For example, in the above-described embodiment, the case where the host vehicle is an autonomous driving vehicle has been described as an example. However, the present invention is not limited to this, and the present invention is applied even if the host vehicle is a driving support target vehicle. be able to. For example, the present invention can also be applied to a driving support system that presents a combination of the time series of the acceleration of the host vehicle, the timing of the lane change of the host vehicle, and the lane to be changed to the driver of the vehicle through a human interface.

Further, in the above embodiment, the first evaluation function, the evaluation function L _a on energy efficiency of the vehicle, a case containing the evaluation function L _b for the safe running of the vehicle is described as an example, this is not limited to, the evaluation function L _a and the evaluation function L _b is not necessarily included in the first evaluation function.

  Moreover, although the case where the route generation unit 30 generates a route from the current position to the destination has been described as an example in the above embodiment, the present invention is not limited to this. For example, a route from the current position input in advance to the destination may be used.

In the above embodiment, the case where the evaluation function E (z _h (t + T)) regarding the behavior of the host vehicle is set based on the route from the current position to the destination is described as an example. However, the present invention is not limited to this. It is not something. For example, an evaluation function E (z _h (t + T)) regarding the behavior of the host vehicle may be set based on traffic jam information and road conditions.

In the above embodiment, the case where the above equation (18) is used as the evaluation function E (z _h (t + T)) related to the behavior of the host vehicle has been described as an example. However, the present invention is not limited to this. The evaluation function may be used.

  Moreover, although the said embodiment demonstrated the case where the surrounding vehicle behavior prediction part 28 estimated the behavior of a surrounding vehicle using the driving | running model shown to said Formula (2) as an example, it is limited to this. Other driving models may be used instead.

  The program of the present invention may be provided by being stored in a storage medium.

DESCRIPTION OF SYMBOLS 10 Lane change optimization apparatus 12 Input operation part 14 Inter-vehicle communication part 16 Vehicle speed sensor 18 Acceleration sensor 20 Position measurement part 22 Road network database 24 Computer 26 Information acquisition part 28 Surrounding vehicle behavior prediction part 32 Optimization part 30 Route generation part 34 Output unit 36 Vehicle control device 320 Lower layer optimization unit 322 Upper layer optimization unit

Claims (8)

Information acquisition means for acquiring the traveling state of the own vehicle and each traveling state of the surrounding vehicles that are traveling in the same lane as the own lane and that are in the vicinity of the own vehicle;
Based on the running state of each of the surrounding vehicles acquired by the information acquisition unit, the surrounding vehicle behavior predicting unit that predicts the time series of the behavior of each of the surrounding vehicles;
Based on the travel state of the host vehicle acquired by the information acquisition unit and the time series of the behavior of each of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit, for each lane change timing of the host vehicle. First optimization means for generating a time series of acceleration of the host vehicle so as to optimize a first evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle;
Based on the time series of the acceleration of the host vehicle generated at each lane change timing by the first optimization unit, a second evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle is optimized. Second optimization means for obtaining a combination of the time series of acceleration of the host vehicle and the timing of lane change of the host vehicle,
Lane change optimization device including The lane change optimization device according to claim 1, wherein the first evaluation function related to traveling of the host vehicle includes an evaluation function related to safe traveling of the host vehicle. The lane change optimization device according to claim 1, wherein the first evaluation function related to traveling of the host vehicle includes an evaluation function related to energy efficiency of the host vehicle.   The said 1st optimization means performs the process which produces | generates the time series of the acceleration of the own vehicle in parallel so that a said 1st evaluation function may be optimized for every timing of the lane change of the own vehicle. 4. The lane change optimization device according to any one of items 3 to 4.   The lane change optimization device according to any one of claims 1 to 4, wherein the second evaluation function related to the traveling of the host vehicle includes the first evaluation function.   The second evaluation function related to the traveling of the host vehicle is a sum of the first evaluation function and an evaluation function related to the behavior of the host vehicle in a predetermined maximum prediction step. The lane change optimization device as described in the item. The first optimization means is based on the traveling state of the host vehicle acquired by the information acquisition means and the time series of the behaviors of the surrounding vehicles predicted by the surrounding vehicle behavior prediction means, Generating a time series of acceleration of the host vehicle so as to optimize the first evaluation function for each lane change timing of the host vehicle and each lane to be changed;
The second optimization means optimizes the second evaluation function based on the time series of the lane change generated by the first optimization means and the time series of the acceleration of the host vehicle generated for each lane to be changed. The lane change according to any one of claims 1 to 6, wherein a combination of a time series of the acceleration of the own vehicle, a timing of changing the lane of the own vehicle, and a lane to be changed is obtained. Optimization device. An information acquisition means for acquiring a running state of the host vehicle and each running state of surrounding vehicles existing around the host vehicle running in each lane having the same running direction as the own lane;
Peripheral vehicle behavior prediction means for predicting the time series of the behavior of each of the surrounding vehicles based on the running state of each of the surrounding vehicles acquired by the information acquisition means,
Based on the travel state of the host vehicle acquired by the information acquisition unit and the time series of the behavior of each of the surrounding vehicles predicted by the surrounding vehicle behavior prediction unit, for each lane change timing of the host vehicle. First optimization means for generating a time series of acceleration of the host vehicle so as to optimize a first evaluation function including an evaluation function relating to a safe lane change in traveling of the host vehicle, and the first optimization Based on a time series of the acceleration of the host vehicle generated at each lane change timing by the means, the second evaluation function including an evaluation function related to a safe lane change in traveling of the host vehicle is optimized. A program for functioning as a second optimization means for obtaining a combination of a time series of vehicle acceleration and the timing of lane change of the host vehicle.

Images