KR102061320B1 - Multichannel Vehicle Communication System and Method based on Machine Learning - Google Patents

Abstract

The present invention relates to a system and a method for multi-channel vehicle communication based on machine learning. According to the present invention, the system comprises: a service server which comprises a database storing vehicle collecting data collected from vehicles, a model generating unit making a neural network model learn the vehicle collecting data stored in the database as learning data to generate a vehicle communication channel selection model, and a model providing unit providing the generated vehicle communication channel selection model to a vehicle; and the vehicle which contains a vehicle communication control unit selecting one or more in multi communication channels provided in the vehicle according to a vehicle driving environment and a communication environment to control a vehicle communication by using the vehicle communication channel selection model provided from the service server. According to the present invention, the present invention can support to select an optimum communication channel proper for receiving information necessary for a vehicle driving support while driving a vehicle and to perform the vehicle communication.

Description

Multichannel Vehicle Communication System and Method based on Machine Learning

The present invention relates to a multichannel vehicle communication system and method, and more particularly, to a machine learning based multichannel vehicle communication system and method.

An autonomous vehicle is a vehicle that can be driven by a vehicle without a driver or a passenger operating. Self-driving vehicles are driven in various road driving environments and in various situations. For example, autonomous vehicles must be able to drive safely while appropriately coping with various situations and variables such as obstacles, traffic conditions, weather, road topography, and driver's conditions encountered while driving on roads.

To this end, autonomous vehicles may be able to sense the surroundings using techniques such as radar, rider, satellite navigation system (GNSS), computer vision, and / or vehicle (V2X) communications. Autonomous vehicle control systems use V2X communications to interpret the sensing information and data acquired using V2X communications to identify obstacles and / or associated signaling systems, as well as appropriate navigation paths. One.

V2X is a technology that allows vehicles in operation to exchange and share road infrastructure, traffic information, and pedestrian information by wireless communication, and communicates with traffic infrastructure such as vehicle-to-vehicle (V2V) communications and traffic lights that inform traffic conditions and vehicle-to-vehicle access. It consists of V2I (vehicle to infrastructure) communication, V2P (vehicle to pedestrian) communication supporting pedestrian information.

Cellular Network (2G / 3G / 4G / 5G), LPWAN (LoRa, Sigfox, Microchip Bluetooth Internet of Things), WiFi-802.11x (802.11a / b / g / n), V2X (802.11p) for V2X communications Various communication channels, such as known WAVE (Wireless Access in Vehicular Environment), DSRC (Dedicated Short Range Communications), LiFi, and satellite communication, may be used.

However, most of the communication technologies used to implement autonomous vehicles still use only a single communication channel. Therefore, autonomous vehicles select the optimal communication channel by considering the characteristics of various communication channels of autonomous vehicles according to various situations and variables such as obstacles, traffic conditions, weather, road topography, driver's condition etc. As a result, the demand for technology that can optimize the performance of vehicle communication is increasing.

Accordingly, an aspect of the present invention is to provide a machine learning based vehicle communication system and method for supporting vehicle communication by selecting an optimal communication channel suitable for receiving information required for vehicle driving support in a vehicle driving situation.

Machine-learning-based vehicle communication system according to the present invention for solving the above technical problem is a vehicle communication channel by learning a neural network model with a database for storing vehicle collection data collected from a vehicle, the vehicle collection data stored in the database as training data A vehicle driving environment using a service generation server including a model generation unit generating a selection model, and a model providing unit providing the generated vehicle communication channel selection model to a vehicle, and a vehicle communication channel selection model provided from the service server. And a vehicle including a vehicle communication controller configured to control one or more of the multi communication channels provided in the vehicle to perform vehicle communication according to a communication environment.

The vehicle collection data may include driving environment data and communication environment data.

The driving environment data may include at least one of road information, weather information, and vehicle driving speed when the vehicle collection data is collected.

The communication environment data may include communication channel information and communication quality information when the vehicle collection data is collected.

The vehicle collection data may further include vehicle driving situation data.

The vehicle driving state data may include at least one of driver state information, traffic state information, and vehicle state information.

The vehicle communication controller may control a vehicle communication by selecting a communication channel suitable for receiving information required for vehicle driving support in a current vehicle driving situation.

Machine learning-based vehicle communication method according to the present invention for solving the above technical problem, storing vehicle collection data collected from a vehicle in a database, by learning the neural network model using the vehicle collection data stored in the database as training data Generating a vehicle communication channel selection model, providing the generated vehicle communication channel selection model to the vehicle, and the vehicle communication control unit using the vehicle communication channel selection model provided from the service server, and the vehicle driving environment and communication And controlling one or more vehicles to communicate with each other by selecting one or more of the multi communication channels provided in the vehicle according to the environment.

And a computer readable recording medium having recorded thereon a program for executing the method on the computer.

According to the present invention, it is possible to support vehicle communication by selecting an optimal communication channel suitable for receiving information required for vehicle driving support in a vehicle driving situation.

1 is a diagram showing the configuration of a machine learning based multi-channel vehicle communication system according to an embodiment of the present invention.
2 is a diagram illustrating electronic components of a vehicle according to an exemplary embodiment of the present invention.
3 is a diagram showing the configuration of a service server according to an embodiment of the present invention.
4 is a flowchart provided to explain the operation of a machine learning based multi-channel vehicle communication system according to an embodiment of the present invention.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention.

1 is a diagram showing the configuration of a machine learning based multi-channel vehicle communication system according to an embodiment of the present invention.

Referring to FIG. 1, a machine learning based multi-channel vehicle communication system according to the present invention may include a plurality of vehicles 100, a weather server 200, a traffic server 300, and a service server 400.

The vehicle 100, the weather server 200, the traffic server 300, and the service server 400 may exchange various information and data through the communication network 10.

The communication network 10 may include a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), the Internet, 2G, 3G, 4G, LTE mobile communication network, Bluetooth, It may include Wi-Fi, Wibro, satellite communication networks, and Low Power Wide Area (LPWA) networks such as LoRa and Sigfox, and the communication method may be wired or wireless and may be any communication method. .

The vehicle 100 may be an autonomous vehicle capable of fully autonomous driving or a vehicle equipped with an advanced driver assistance system (ADAS) capable of semiautonomous driving.

2 is a diagram illustrating electronic components of a vehicle according to an exemplary embodiment of the present invention.

2, the vehicle 100 according to the present invention includes a vehicle control unit 110, a communication unit 120, a vehicle communication control unit 130, a data collection unit 140, a learning unit 150, and a storage unit 160. And electronic components such as the vehicle sensor unit 170.

The vehicle control unit 110 performs a function of controlling a mechanical device such as an engine, a transmission, a steering device, a braking device, a suspension device, and the vehicle communication control unit through an on-board diagnostics (OBD) interface. 130 or the data collection unit 140 may be provided. The vehicle controller 110 may be implemented as an electronic control unit (ECU) of the vehicle.

The communication unit 120 provides a communication function to allow the vehicle 100 to exchange various information and data with a device inside or outside the vehicle, and to perform a communication, a transmission antenna, a reception antenna, and various communication protocols may be implemented. Frequency) and at least one of the RF element. And the communication unit 120 is Cellular Network (2G / 3G / 4G / 5G), LPWAN (LoRa, Sigfox, Microchip Bluetooth Internet of Things), WiFi-802.11x (802.11a / b / g /) for vehicle (V2X) communication n), it is possible to support communication through various communication channels such as WAVE (Wireless Access in Vehicular Environment), DSRC (Dedicated Short Range Communications), LiFi, satellite communication, and optical communication, which are known as V2X (802.11p).

V2X is a communication technology in which a vehicle in operation exchanges and shares road infrastructure, traffic information, and pedestrian information through wireless communication, and it provides traffic infrastructure such as vehicle to vehicle (V2V) communication and traffic lights to inform traffic conditions and vehicle-to-vehicle access. It may include vehicle to infrastructure (V2I) communication to communicate, vehicle to pedestrian (V2P) communication supporting pedestrian information.

The vehicle communication control unit 130 performs a function of controlling one or more of the multi communication channels provided in the vehicle 100 to perform vehicle communication according to the vehicle driving environment and the communication environment. According to an embodiment, the vehicle communication control unit 130 may be integrally implemented with the vehicle control unit 110.

The vehicle communication control unit 130 determines the vehicle driving situation by using various information and data collected from the inside or the outside of the vehicle, and uses the vehicle communication channel selection model provided from the service server 400 to determine the vehicle driving environment and communication. According to an environment, the communication unit 120 may be controlled to select one or more optimal communication channels from among the multi communication channels provided in the vehicle to communicate with the vehicle. For example, the vehicle communication control unit 130 may apply information about an accident on a road, a traffic jam, a road geometry, a driver situation, a road weather situation, a road surface condition, a pedestrian situation, and the like to optimize a vehicle communication channel selection model. One or more channels may be selected and vehicle communication may be performed on the selected communication channel.

More specifically, road geometry (e.g., road types such as crossroads, crossroads, curvature and slope road geometry, or building heights around roads), road weather conditions, road traffic (e.g. communication traffic may be proportional to traffic) The communication quality for each communication channel according to the like can be predicted using the vehicle communication channel selection model. The available communication channel may be limited according to a source for obtaining information required for autonomous vehicle driving (eg, another vehicle nearby, a road infrastructure, or a server located remotely). In addition, service priorities may be applied according to types of autonomous vehicle services (eg, safe driving support or fuel efficiency improvement). For example, an emergency message transmission related to the safety of the vehicle occupant may be selected to be communicated preferentially. In addition, a vehicle communication may be performed by selecting a communication channel suitable for the selected service.

In addition to the above description, the vehicle communication control unit 130 receives vehicle driving status data including driver status information, traffic status information, and vehicle status information, and is a communication channel suitable for receiving information required to support vehicle driving in a current vehicle driving situation. Can be selected to control vehicle communication.

The data collection unit 140 collects data generated or transmitted from various data sources such as internal and external sensors of the vehicle 100, other vehicles nearby, the weather server 200, and the traffic server 300. do.

The vehicle collection data collected by the data collector 140 may include driving environment data and communication environment data. The driving environment data may include at least one of road information, weather information, and vehicle driving speed when vehicle collection data is collected. The communication environment data may include communication channel information and communication quality information when vehicle collection data is collected. The vehicle collection data may further include vehicle driving situation data. The vehicle driving state data may include at least one of driver state information, traffic state information, and vehicle state information.

Vehicle collection data collected by the data collection unit 140 may be delivered to the storage 160 and the service server 170 and stored.

The learning unit 150 may additionally collect information and data collected separately from the vehicle 100 based on the vehicle communication channel selection model provided from the service server 400, for example, driver individual characteristic information such as driver driving habits of the vehicle. In addition, the vehicle communication channel selection model can be additionally trained according to individual driving characteristics such as the main driving route of the vehicle and the main driving region (country, city, etc.).

The storage unit 160 may store various information and data related to communication channel control of an individual vehicle. For example, the storage unit 160 stores the data collected by the data collection unit 140 and the vehicle communication channel selection model provided from the service server 400, and if necessary, the learning unit 150 and the vehicle communication control unit 130. ) And the like. The storage unit 160 may be implemented in various storage media such as a ROM, a RAM, an EPROM, a flash drive, and a hard drive in hardware.

The vehicle sensor unit 170 is mounted inside or outside the vehicle to sense a situation inside and outside the vehicle and driver information, and for this purpose, a camera, an infrared sensor, an ultrasonic sensor, a microphone, a driver biometric information sensor, a laser, Various sensing means such as a rider, a temperature and humidity sensor, a rainfall sensor, an illuminance sensor, and a GPS sensor may be provided. In addition, the vehicle sensor unit 170 may detect an inertial sensor, a collision sensor, a speed sensor, an inclination sensor, a battery sensor, a fuel sensor, a tire sensor, a steering sensor, an accelerator pedal position sensor, and a brake pedal position in order to sense the state and attitude of the vehicle. It may further include a sensor.

The vehicle controller 110 and the vehicle communication controller 130 may include application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), and field programmable gate arrays (FPGAs). It may be implemented using at least one of processors, controllers, micro-controllers, microprocessors, and electrical units for performing other functions.

Referring again to FIG. 1, the meteorological server 200 is a device operated by a meteorological office or a private meteorological service provider, and may provide information on temperature, rainfall, snowfall, wind speed, humidity, solar radiation, sunshine, and the like. The weather server 200 may directly provide weather data to the vehicle 100 or may provide the weather data through the service server 400.

The traffic server 300 is a device operated by a road management entity or an operator that provides information related to traffic of a road, and may provide road traffic information in real time. The road traffic information may include the current traffic volume and the average vehicle speed of each road. The traffic server 300 may directly provide road traffic information to the vehicle 100 or may provide the traffic server 300 through the service server 400.

The service server 400 may generate a vehicle communication channel selection model by learning a neural network model using the vehicle collection data collected from the vehicle as learning data, and provide the generated vehicle communication channel selection model to the vehicle.

3 is a diagram showing the configuration of a service server according to an embodiment of the present invention.

Referring to FIG. 3, the service server 400 may include a database 410, a model generator 430, and a model provider 450.

The database 410 stores vehicle collection data collected by the vehicle. Vehicle collection data collected while the vehicle is driving corresponds to a large amount of space-time big data. The database 410 may support a big data processing system that supports real-time collection and storage of spatiotemporal big data and distributed parallel data processing.

The model generator 430 may generate a vehicle communication channel selection model for generating a vehicle communication channel selection model by training the neural network model with the training data generated using the vehicle collection data stored in the database 410.

The neural network model learned by the model generator 430 may be in the form of a machine learning algorithm such as a convolution neural network (CNN). Convolutional Neural Networks (CNN), Deep Neural Networks (DNN), Recurrent Neural Networks (RNNs), restricted Boltzmann machines, deep trust based on inception module Neural network algorithms to which various deep learning techniques are applied, such as Deep Belief Network (DBN) and Deep Q-Network, may be used.

The service server 400 may additionally generate training data or allow the model generator 430 to update the vehicle communication channel selection model by using the training data accumulated for a predetermined period.

The model provider 450 may provide the vehicle 100 with a vehicle communication channel selection model generated by learning from the model generator 430. The model providing unit 450 may provide the updated model to the vehicle 100 whenever the vehicle communication channel selection model is updated in the model generating unit 430.

4 is a flowchart provided to explain the operation of a machine learning based multi-channel vehicle communication system according to an embodiment of the present invention.

Referring to FIG. 4, first, the service server 400 may store vehicle collection data collected from a vehicle in a database 410 to build learning data for learning a vehicle communication channel selection model (S410). The database 410 may support a big data processing system that supports real-time collection and storage of spatiotemporal big data and distributed parallel data processing.

Next, the service server 400 may generate a vehicle communication channel selection model by learning a neural network model using the vehicle collection data stored in the database 410 as training data (S420). The vehicle communication channel selection model generated in step S420 may be learned to select one or more optimal communication channels among the multi communication channels provided in the vehicle according to the vehicle driving environment and the communication environment.

Thereafter, the service server 400 may provide the generated vehicle communication channel selection model to the vehicle (S430).

Finally, the vehicle communication control unit 130 provided in the vehicle 100 uses one of the vehicle communication channel selection models provided in step S430 to establish one or more of the multi communication channels provided in the vehicle according to the vehicle driving environment and the communication environment. The vehicle communication may be selected and controlled (S440).

Embodiments of the invention include a computer readable medium containing program instructions for performing various computer-implemented operations. This medium records a program for executing the method described above. The media may include, alone or in combination with the program instructions, data files, data structures, and the like. Examples of such media include, but are not limited to, magnetic media such as hard disks, floppy disks, and magnetic tape, optical recording media such as CDs and DVDs, floppy disks and program commands such as magnetic, optical media, ROM, RAM, flash memory, and the like. Hardware devices configured to store and perform such operations. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like.

Although the preferred embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

100: vehicle
110: vehicle control unit 120: communication unit
130: vehicle communication control unit 140: data collection unit
150: learning unit 160: storage unit
170: vehicle sensor unit
200: weather server
300: transportation server
410: database 430: model generation unit
450: model provider

Claims (10)

A database for storing vehicle collection data collected from a vehicle, a model generator for generating a vehicle communication channel selection model by learning a neural network model from the vehicle collection data stored in the database as training data, and the generated vehicle communication channel selection model. A service server including a model providing unit provided to a vehicle, and
A vehicle comprising a vehicle communication control unit for controlling the vehicle communication by selecting one or more of the multi-communication channels provided in the vehicle according to the vehicle driving environment and communication environment using the vehicle communication channel selection model provided from the service server
Including,
And the vehicle collection data includes driving environment data and communication environment data. In claim 1,
The driving environment data includes at least one of road information, weather information, and vehicle driving speed when the vehicle collection data is collected.
And the communication environment data includes communication channel information and communication quality information when the vehicle collection data is collected. In claim 2,
The vehicle collection data further includes vehicle driving situation data. In claim 3,
The vehicle driving situation data,
A machine learning based vehicle communication system comprising at least one of driver status information, traffic condition information, and vehicle status information. In claim 4,
The vehicle communication control unit,
Machine learning-based vehicle communication system for controlling the vehicle communication by selecting a communication channel suitable for receiving information required for vehicle driving support in the current vehicle driving situation. Storing vehicle collection data collected by the vehicle in a database,
Generating a vehicle communication channel selection model by training the neural network model using the vehicle collection data stored in the database as training data;
Providing the generated vehicle communication channel selection model to a vehicle, and
Controlling the vehicle communication control unit to perform at least one vehicle communication by selecting one or more of the multi communication channels provided in the vehicle according to the vehicle driving environment and the communication environment using the vehicle communication channel selection model provided from the service server;
Including,
And the vehicle collection data includes driving environment data and communication environment data. In claim 6,
The driving environment data includes at least one of road information, weather information, and vehicle driving speed when the vehicle collection data is collected.
And the communication environment data includes communication channel information and communication quality information when the vehicle collection data is collected. In claim 7,
The vehicle collection data further includes vehicle driving situation data. In claim 8,
The vehicle driving situation data,
A machine learning based vehicle communication method comprising at least one of driver status information, traffic condition information, and vehicle status information. In claim 9,
The vehicle communication control unit,
Machine learning-based vehicle communication method for controlling the vehicle communication by selecting a communication channel suitable for receiving information required for vehicle driving support in the current vehicle driving situation.

Images