# US 2021/0171004 A1 · System of modeling anti-lock brake system controller of vehicle

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0171004 A1
- Document ID: US20210171004A1
- Status: Published application
- Kind: A1
- Application number: 16/994,820
- Filed: 2020-08-17
- Published: 2021-06-10
- Earliest claimed: 2019-12-10
- Assignees: Hyundai Motor Company; Kia Motors Corporation
- Applicants: Hyundai Motor Company; Kia Motors Corporation
- Inventors: Wook Hyun Han
- Web page: https://patentgrep.com/patent/US20210171004A1

## CPC

- B60T 8/174: characterised by using special control logic, e.g. fuzzy logic {, neural computing}
- B60T 8/171: Detecting parameters used in the regulation; Measuring values used in the regulation
- B60T 8/1761: responsive to wheel or brake dynamics, e.g. wheel slip, wheel acceleration or rate of change of brake fluid pressure
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/049: Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
- G06N 3/09: Supervised learning
- B60T: VEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL (electrodynamic brake systems for vehicle, in general B60L; brakes per se, i.e. devices where braking effect occurs, including ultimate brake actuators, F16D); ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OF VEHICLES; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A system of modeling an anti-lock brake system (ABS) controller of a vehicle includes: a vehicle speed estimation network configured to estimate a vehicle speed through machine learning using wheel speed data of each of a plurality of wheels of the vehicle; a wheel speed state estimation network configured to estimate a time series characteristic of the wheel speed through machine learning using information on the wheel speed data and information on whether a brake pedal is depressed; and a classification network configured to estimate a braking mode for controlling an increase, a decrease, or steady state of a braking pressure of each wheel through machine learning using speed estimation data of the vehicle estimated by the vehicle speed estimation network and time series characteristic estimation data of the wheel speed estimated by the wheel speed state estimation network.

## Claims

### Claim 1

A system of modeling an anti-lock brake system (ABS) controller of a vehicle, the system comprising:
a vehicle speed estimation network configured to estimate a vehicle speed through machine learning using wheel speed data of each of a plurality of wheels of the vehicle;
a wheel speed state estimation network configured to estimate a time series characteristic of the wheel speed through machine learning using information on the wheel speed data and information on vehicle whether a brake pedal is depressed; and
a classification network configured to estimate a braking mode for controlling an increase, a decrease, or steady state of a braking pressure of each wheel through machine learning using speed estimation data of the vehicle estimated by the vehicle speed estimation network and time series characteristic estimation data of the wheel speed estimated by the wheel speed state estimation network.

### Claim 2 (depends on 1)

The system of claim 1, further comprising:
a concatenate layer configured to concatenate the speed estimation data estimated by the vehicle speed estimation network with the time series characteristic estimation data of the wheel speed estimated by the wheel speed state estimation network to provide the concatenated data to the classification network.

### Claim 3 (depends on 1)

The system of claim 1, wherein the vehicle speed estimation network includes a deep learning network having a plurality of one-dimensional convolution layers configured to receive the wheel speed data of each wheel of the vehicle and extract a feature corresponding to the vehicle speed from the received wheel speed data.

### Claim 4 (depends on 3)

The system of claim 3, wherein the wheel speed data input to the vehicle speed estimation network is obtained by sequentially measuring several times each wheel speed for each wheel of the vehicle at predetermined time intervals.

### Claim 5 (depends on 1)

The system of claim 1, wherein the wheel speed state estimation network includes a plurality of long short term memory (LSTM) networks configured to perform machine learning using a plurality of pieces of the wheel speed data sequentially measured at predetermined time intervals for each wheel of the vehicle and information on whether a brake pedal of the vehicle is depressed.

### Claim 6 (depends on 5)

The system of claim 5, wherein each of the plurality of LSTM networks estimates data on an increase or decrease state of each wheel speed as the time series characteristic.

### Claim 7 (depends on 1)

The system of claim 1, wherein the classification network includes at least one fully connected (FC) layer configured to perform machine learning for estimating a braking mode of each wheel using the speed estimation data of the vehicle estimated by the vehicle speed estimation network and the time series characteristic estimation data of the wheel speed estimated by the wheel speed state estimation network.

### Claim 8 (depends on 7)

The system of claim 7, further comprising:
an output layer configured to use a softmax function, which outputs an output value of the at least one FC layer as a probability value for increasing/decreasing/maintaining the braking pressure of each wheel, as an activation function.

### Claim 9 (depends on 7)

The system of claim 7, wherein the classification network sets learning importance when the braking pressure of the wheel is decreased during the machine learning to be higher than that when the braking pressure of the wheel is maintained.

### Claim 10 (depends on 7)

The system of claim 7, wherein, when the braking pressure of the wheel is decreased or maintained at steady state during the machine learning, the classification network performs the machine learning, and when the output estimated value does not correspond to the decrease or the steady state of the braking pressure of the wheel, the classification network determines the output estimated value as the increase of the braking pressure of the wheel.

### Claim 11

A system of modeling an anti-lock brake system (ABS) controller of a vehicle, the system comprising:
a vehicle speed estimation network including a plurality of one-dimensional convolution layers configured to estimate a vehicle speed through machine learning using data obtained by sequentially measuring several times a wheel speed for each wheel of a vehicle at predetermined time intervals;
a wheel speed state estimation network including a plurality of long short term memory (LSTM) networks configured to estimate a time series characteristic of the wheel speed through machine learning using a plurality of pieces of wheel speed data sequentially measured at the predetermined time intervals for each wheel of the vehicle and information on whether a brake pedal of the vehicle is depressed;
a concatenate layer configured to concatenate the speed estimation data estimated by the vehicle speed estimation network with the time series characteristic estimation data of the wheel speed estimated by the wheel speed state estimation network to provide the concatenated data to the classification network; and
a classification network including at least one fully connected (FC) layer configured to estimate a braking mode for controlling an increase, a decrease, or steady state of a braking pressure of each wheel through machine learning using the speed estimation data of the vehicle and the time series characteristic estimation data of the wheel speed, which are provided from the concatenate layer.

### Claim 12 (depends on 11)

The system of claim 11, wherein each of the plurality of LSTM networks estimates data on an increase or decrease state of each wheel speed as the time series characteristic.

### Claim 13 (depends on 11)

The system of claim 11, further comprising:
an output layer configured to use a softmax function, which outputs an output value of the at least one FC layer as a probability value for increasing/decreasing/maintaining the braking pressure of each wheel, as an activation function.

### Claim 14 (depends on 11)

The system of claim 11, wherein the classification network sets learning importance when the braking pressure of the wheel is decreased during the machine learning to be higher than that when the braking pressure of the wheel is maintained.

### Claim 15 (depends on 11)

The system of claim 11, wherein, when the braking pressure of the wheel is decreased or maintained at steady state during the machine learning, the classification network performs the machine learning, and when the output estimated value does not correspond to the decrease or the steady state of the braking pressure of the wheel, the classification network determines the output estimated value as the increase of the braking pressure of the wheel.
