Road surface condition estimation apparatus

1 . A road surface condition estimation apparatus, the apparatus comprising:

a receiver configured to acquire travel information using a network provided in a vehicle;

a storage unit configured to store a road surface condition estimation model;

a road surface severity estimator configured to estimate, based on the travel information, severity of a condition of a road surface on which the vehicle is travelling using the road surface condition estimation model; and

a post-processor configured to perform post-processing on road surface severity information estimated by the road surface severity estimator,

wherein the post-processor is further configured to:

apply a first exponential parameter value, when the road surface severity estimator estimates the road surface as a deep road surface, and

apply a second exponential parameter value, when the road surface severity estimator estimates the road surface as a shallow road surface,

wherein the road surface condition estimation apparatus is configured to output the post-processed road surface severity information as a reference for switching a travel mode of the vehicle, wherein switching the travel mode is based on a comparison of the post-processed road surface severity information to a preset value.

2 . The apparatus of claim 1 , wherein the travel information includes at least one of engine torque, engine speed, longitudinal acceleration, lateral acceleration, yaw rate, wheel speed, gear state, steering angle, and vehicle speed.

3 . The apparatus of claim 1 , wherein the road surface condition estimation model is trained using a deep learning network.

4 . The apparatus of claim 3 , wherein

the road surface condition estimation model includes a first deep learning network and a second deep learning network, and

the first deep learning network and the second deep learning network are trained based on data on different road surfaces.

5 . The apparatus of claim 4 , wherein the first deep learning network is trained to estimate severity of a sandy road surface, and the second deep learning network is trained to estimate severity of a muddy road surface.

6 . The apparatus of claim 1 , wherein the road surface severity estimator is configured to estimate the road surface severity, only when the travel information satisfies a preset estimation start condition.

7 . The apparatus of claim 6 , wherein the preset estimation start condition is determined based on at least one of a steering angle, a gear state, and a travel speed of the vehicle.

8 . The apparatus of claim 6 , wherein the preset estimation start condition includes at least a condition for a steering angle of the vehicle, and

the road surface severity of a condition of a road surface on which the vehicle is travelling is estimated when the steering angle is 360 degrees or less.

9 . The apparatus of claim 6 , wherein the preset estimation start condition includes at least a condition for a gear state of the vehicle, and

the road surface severity of a condition of a road surface on which the vehicle is travelling is estimated when the gear state is not a reverse gear.

10 . The apparatus of claim 6 , wherein the preset estimation start condition includes at least a condition for a travel speed of the vehicle, and

the road surface severity of a condition of a road surf ace on which the vehicle is travelling is estimated when the travel speed of the vehicle is greater than or equal to a preset speed.

11 . The apparatus of claim 1 , wherein the road surface severity estimator is configured to estimate a result of estimating the road surface severity as a score within a preset range.

12 . The apparatus of claim 11 , wherein the score within the preset range has a range of 0.0 to 1.0.

13 . The apparatus of claim 12 , wherein the road surface severity estimator is configured to:

estimate the road surface as the deep road surface, when the estimated score is greater than a preset first reference score; and

estimate the road surface as the shallow road surface, when the estimated score is less than a preset second reference score.

14 . The apparatus of claim 13 , wherein an average of the first reference score and the second reference score is less than 0.5.

15 . The apparatus of claim 1 , wherein the post-processor is further configured to perform post-processing using an exponential moving average (EMA).

16 . The apparatus of claim 1 , wherein the first exponential parameter value is set to be greater than the second exponential parameter value.