KR102216272B1 - System for collecting and providing road traffic danger information using smart phone and cloud server, and method for the same - Google Patents

Abstract

Provided are a road traffic danger information collection and provision system using a smartphone and a cloud server and a method thereof. The road traffic danger information collection and provision system can collect and quickly provide road traffic danger information corresponding to a shock event with respect to a road situation or a driving situation by analyzing data sensed by a GPS module, a gravitational acceleration sensor, and a gyro sensor, which are basically embedded in a smartphone, through a machine learning algorithm; can also improve a sensing rate with respect to a danger driving pattern of a driving vehicle by dualizing the analysis of a cloud server and the analysis of a smartphone application, and improve the performance of the machine learning algorithm by enabling the cloud server to continuously update a learning dataset based on pre-stored sensor data; can also minimize the probability of danger information non-detection caused by unit time splitting by overlapping analysis cycles with each other when analyzing the sensor data sensed through the smartphone sensor; and can also quickly generate road traffic danger information in real-time and reduce battery consumption of the smartphone of the driving vehicle by enabling the smartphone application to operate a machine learning algorithm of an analysis process only when the sensor data sensed through the gravitational acceleration sensor and the gyro sensor exceeds a preset value.

Description

A system for collecting and providing road traffic risk information using a smartphone and a cloud server and its method {SYSTEM FOR COLLECTING AND PROVIDING ROAD TRAFFIC DANGER INFORMATION USING SMART PHONE AND CLOUD SERVER, AND METHOD FOR THE SAME}

The present invention relates to a road traffic hazard information collection and provision system, and more specifically, a GPS module, a gravitational acceleration sensor, and a gyro sensor basically embedded in a smartphone to detect and analyze, and to detect and analyze an impact event on a road condition or a driving condition. It relates to a system and method for collecting and providing road traffic risk information using a smartphone and a cloud server for collecting and providing corresponding road traffic risk information.

According to the World Health Organization report, 1.25 million people worldwide die annually from road traffic accidents, and 56% of these road traffic accidents are reported to be caused by dangerous driving patterns by drivers.

1 is a diagram showing the occurrence of an impact event in a dangerous road traffic situation.

As shown in a) of FIG. 1, when the driving vehicle 120 and a plurality of rear vehicles 130a to 130n travel along the road driving direction on the road 110, as shown in b) of FIG. 1 Likewise, a shock event may occur due to the driver's dangerous driving pattern. For example, a shock event may occur according to a driving situation event such as a sudden braking, a sudden start, or a sudden turn, or a road situation event such as a pothole or a bump.

In this case, in the case of the first rear vehicle 130a immediately behind the driving vehicle 120, it is possible to respond by visually recognizing the danger situation caused by the driving vehicle 120, but the danger situation caused by the driving vehicle 120 In the case of rear vehicles that cannot be recognized with the naked eye, it is difficult to cope with this, so there is a risk of traffic accidents such as a collision.

On the other hand, as a prior art, Korean Patent Application Publication No. 2012-31335 discloses an invention entitled "Method of transmitting information on unexpected situations of a high-speed driving vehicle using a smartphone", which can be easily carried by a driving vehicle driver. A communication relay device that allows smartphones and roadside devices to establish a wireless LAN-based temporary communication network within the communication area, and collects relevant information and quickly delivers it to other vehicles when an unexpected situation occurs in a vehicle running at high speed. It relates to a message transmission method between vehicles using the.

According to the conventional method of transmitting information on unexpected situations of a high-speed vehicle using a smartphone according to the prior art, when an accident that may occur unexpectedly in a vehicle traveling at high speed on the road or a sudden deceleration of the vehicle due to traffic congestion occurs, this information It is possible to immediately respond to an emergency situation and prevent consecutive collisions in advance by transmitting the data to vehicles running from the rear in real time.

However, in the case of a method of transmitting information on an unexpected situation of a high-speed vehicle using a smartphone according to the prior art, there is a problem that it is practically impossible to monitor the entire road section because a number of separate roadside communication relay devices must be installed.

On the other hand, CCTVs are installed on the road to monitor traffic conditions by manpower, or recently, radar sensors are installed to collect road traffic risk information. However, the entire road is due to the cost of sensors installed on the road and maintenance of communication infrastructure. There is a problem that it is practically impossible to monitor the section.

Republic of Korea Patent No. 10-1793630 (Registration date: October 30, 2017), Invention title: "Collaboration method and cloud server between smart devices" Republic of Korea Patent No. 10-877515 (Registration date: December 30, 2008), title of invention: "Real-time traffic information collection/providing method and apparatus using vehicle navigation" Republic of Korea Patent Publication No. 2019-56695 (published date: May 27, 2019), title of invention: "Cloud-based traffic signal information providing system" Republic of Korea Patent Publication No. 2012-36674 (published date: April 18, 2012), title of the invention: "Traffic information providing system and method using a smartphone" Republic of Korea Patent Publication No. 2012-31335 (published on April 3, 2012), title of the invention: "Method of transmitting information on unexpected situations of a high-speed driving vehicle using a smartphone" Republic of Korea Patent Publication No. 2009-70679 (published date: July 1, 2009), title of invention: "Traffic information providing apparatus and method"

The technical problem to be achieved by the present invention for solving the above-described problem is to analyze data detected by a GPS module, a gravity acceleration sensor, and a gyro sensor basically embedded in a smartphone according to a machine learning algorithm, and the road situation or driving situation It is to provide a system and method for collecting and providing road traffic risk information using a smartphone and a cloud server that can quickly provide by collecting road traffic risk information corresponding to an impact event for

Another technical problem to be achieved by the present invention is that it is possible to improve the detection rate for dangerous driving patterns of a driving vehicle by dualizing the analysis of the smartphone app and the analysis of the cloud server, and the cloud server continuously utilizes the stored sensor data. It is to provide a system and method for collecting and providing road traffic risk information using a smartphone and a cloud server that can improve the performance of machine learning algorithms over time by updating the learning dataset.

As a means for achieving the above-described technical problem, the road traffic risk information collection and provision system using a smartphone and a cloud server according to the present invention collects road traffic risk information corresponding to a dangerous driving pattern of a driving vehicle running on the road. What is claimed is: 1. A system for collecting and providing road traffic hazard information, comprising: a terminal mounted on a driving vehicle running on a road, comprising: a driving vehicle smartphone including a GPS module and a communication module to upload sensor data; A smart phone sensor including a gravity acceleration sensor and a gyro sensor, and is embedded in the smart phone of the driving vehicle to detect an impact event that is a dangerous driving pattern of the driving vehicle to generate sensor data; A smart phone app installed on the driving vehicle smart phone, collecting sensor data generated by the smart phone sensor, performing an analysis process, and generating road traffic risk information; A cloud server that collects and analyzes all sensor data uploaded from at least one driving vehicle smartphone, and creates and provides a road traffic risk information map; And a rear vehicle smartphone receiving a road traffic hazard information map from the cloud server, as terminals respectively mounted on at least one rear vehicle driving behind the road, wherein a smartphone app installed on the driving vehicle smartphone And each of the cloud servers is characterized in that the analysis process is performed in duplicate according to a previously learned machine learning algorithm.

Here, the impact event of the driving vehicle includes a driving situation event and a road situation event, the driving situation event occurs due to a sudden braking, sudden start or sudden turn of the driving vehicle, and the road situation event is due to a porthole or bump on the road. It is characterized by occurring.

Here, the sensor data may include data detected by a gravity acceleration sensor and a gyro sensor of the smartphone sensor; And data generated by the GPS module of the driving vehicle smartphone that can determine the location of the driving vehicle on the road.

Here, the driving vehicle smart phone includes: a GPS module for generating location information according to a position of the driving vehicle on a road; And a communication module that communicates with the cloud server to upload sensor data sensed by the smartphone sensor.

Here, the smart phone sensor, a gravity acceleration sensor that is embedded in the smart phone of the driving vehicle and detects an acceleration applied to the vehicle body of the driving vehicle in response to an impact event, which is a dangerous driving pattern of the driving vehicle; And a gyro sensor that is embedded in the driving vehicle smartphone and detects an angular velocity applied to the vehicle body of the driving vehicle in response to an impact event that is a dangerous driving pattern of the driving vehicle.

Here, the smartphone app may include: a data storage unit for storing sensor data detected by the smartphone sensor; An analysis process execution unit for generating road traffic risk information by performing an analysis process on the sensor data according to a machine learning algorithm; And an alarm generator for generating an alarm so that road traffic hazard information generated by the analysis process execution unit is immediately provided to a driver of the driving vehicle.

Here, the analysis process performing unit is characterized in that it performs the analysis process only when the sensor data detected by the smartphone sensor exceeds a preset value.

Here, the analysis process performing unit includes: a data collection unit collecting the sensor data as raw data, a smoothing unit performing smoothing on the raw data collected by the data collecting unit at a predetermined analysis period; A feature extracting unit for extracting a feature vector of the sensor data smoothed by the smoothing performing unit; A data classification unit for classifying data to correspond to a feature vector extracted by the feature extraction unit according to a machine learning algorithm; A learning dataset downloaded from the cloud server and continuously provided to improve the machine learning algorithm when the data classification unit classifies data; And a road traffic risk information generating unit that generates road traffic risk information corresponding to an impact event of a driving vehicle from the data classified by the data classification unit.

Here, the smoothing performing unit is characterized in that the predetermined analysis period is overlapped with each other so as to minimize the non-detection of the impact event.

Here, the cloud server includes: a communication module that communicates with a communication module of a smartphone of a driving vehicle to upload the sensor data; A data storage unit for storing all sensor data uploaded from at least one driving vehicle smartphone; An analysis process execution unit for generating road traffic risk information by performing an analysis process according to a machine learning algorithm on all sensor data uploaded from at least one or more driving vehicle smartphones; And a road traffic risk information map creation unit that creates a road traffic risk information map so that the road traffic risk information generated by the analysis process execution unit is provided to a rear vehicle and a road manager in real time.

Here, the analysis process performing unit includes: a data collection unit collecting the sensor data as raw data; A smoothing performing unit for smoothing the raw data collected by the data collecting unit at a predetermined analysis period; A feature extracting unit for extracting a feature vector of the sensor data smoothed by the smoothing performing unit; A data classification unit for classifying data to correspond to a feature vector extracted by the feature extraction unit according to a machine learning algorithm; A learning dataset continuously provided to improve the machine learning algorithm when the data classification unit classifies data; A road traffic risk information generation unit that generates road traffic risk information corresponding to an impact event of the driving vehicle from the data classified by the data classification unit; And a clustering performing unit clustering to display road traffic risk information for an event occurring in a preset distance range as one event information.

On the other hand, as another means for achieving the above-described technical problem, the method for collecting and providing road traffic risk information using a smartphone and a cloud server according to the present invention is a road traffic risk corresponding to a dangerous driving pattern of a driving vehicle traveling on a road. A method for collecting and providing road traffic hazard information by collecting and providing information, the method comprising: a) generating sensor data by sensing an impact event by a smartphone sensor of a smartphone of a driving vehicle mounted in a driving vehicle; b) collecting sensor data detected by the smartphone sensor through a smartphone app installed on the smartphone of the driving vehicle; c) uploading the sensor data to a cloud server by the driving vehicle smartphone; d) the smart phone app checking whether the sensor data is greater than a preset value; e) when the sensor data is greater than a preset value, generating road traffic risk information by performing an analysis process on the sensor data by a smartphone app; f) performing, by the cloud server, an analysis process for all sensor data uploaded from at least one smartphone of a driving vehicle, and creating a road traffic hazard information map; And g) providing the road traffic risk information map to rear vehicles and road managers in real time, wherein each of the smartphone app installed on the driving vehicle smartphone and the cloud server is analyzed according to a previously learned machine learning algorithm. It is characterized in that the process is carried out in duplicate.

Here, the step e) includes: e-1) collecting the sensor data as raw data; e-2) smoothing the raw data at a predetermined analysis period; e-3) extracting a feature vector of the smoothed sensor data; e-4) classifying data corresponding to the feature vector according to a machine learning algorithm; e-5) generating road traffic hazard information corresponding to an impact event of the driving vehicle from the classified data; And e-6) immediately alerting the driving vehicle driver of the road traffic hazard information.

Here, the step f) includes: f-1) collecting sensor data uploaded from each of at least one or more smartphones of the driving vehicle as raw data; f-2) smoothing the raw data at a predetermined analysis period; f-3) extracting a feature vector of the smoothed sensor data; f-4) classifying data corresponding to the feature vector according to a machine learning algorithm; f-5) generating road traffic hazard information corresponding to an impact event of the driving vehicle from the classified data; f-6) clustering the road traffic risk information according to a preset distance range; And f-7) creating a road traffic risk information map so that the road traffic risk information is provided to a rear vehicle and a road manager in real time.

According to the present invention, by analyzing data detected by a GPS module, a gravity acceleration sensor, and a gyro sensor, which are basically built into a smartphone, according to a machine learning algorithm, road traffic risk information corresponding to an impact event on a road situation or a driving situation is analyzed. It can be collected and provided quickly.

According to the present invention, it is possible to improve the detection rate for dangerous driving patterns of a driving vehicle by dualizing the analysis of the smartphone app and the analysis of the cloud server, and the cloud server continuously updates the learning data set using the previously stored sensor data. By doing so, the machine learning algorithm performance can be improved over time.

According to the present invention, when analyzing sensor data sensed through a smartphone sensor, analysis cycles are overlapped with each other, thereby minimizing the possibility of not detecting risk information according to unit time division.

According to the present invention, the smartphone app operates the machine learning algorithm of the analysis process only when the sensor data detected through the gravity acceleration sensor and the gyro sensor exceeds a preset value, thereby reducing the battery consumption of the smartphone of the driving vehicle. It can reduce and generate road traffic risk information quickly and in real time.

1 is a diagram showing the occurrence of an impact event in a dangerous road traffic situation.
2 is a schematic configuration diagram of a road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.
3 is a detailed configuration diagram of a driving vehicle smartphone and a smartphone app in the road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.
4 is a detailed configuration diagram of a cloud server in a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention.
5 is a diagram for explaining smoothing of sensor data in a road traffic hazard information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.
6 is a view for explaining the overlap of analysis cycles in the road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.
7 is a diagram illustrating grouping of shock events in a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention.
8 is a diagram illustrating an impact event occurrence point in a road traffic hazard information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.
9 is an operation flow diagram of a method for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention.
10 is an operation flow diagram for specifically explaining the analysis process of the smart phone shown in FIG. 9.
11 is an operation flow diagram for specifically explaining the analysis process of the cloud server shown in FIG. 9.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and similar reference numerals are assigned to similar parts throughout the specification.

Throughout the specification, when a part "includes" a certain component, it means that other components may be further included rather than excluding other components unless otherwise stated. In addition, terms such as "... unit" described in the specification mean a unit that processes at least one function or operation, which may be implemented by hardware or software or a combination of hardware and software.

Hereinafter, a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention will be described with reference to FIGS. 2 to 8, and an embodiment of the present invention will be described with reference to FIGS. A method of collecting and providing road traffic risk information using a smartphone and a cloud server is described.

[Road traffic risk information collection and provision system using smartphone and cloud server]

2 is a schematic configuration diagram of a road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.

Referring to FIG. 2, a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention includes road traffic corresponding to a dangerous driving pattern of a driving vehicle 120 traveling on a road 110. As a road traffic hazard information collection and provision system that collects and provides dangerous information, a driving vehicle smartphone 200, a smartphone sensor 300, a smartphone app 400, a cloud server 500, and a trailing vehicle smartphone 600 ).

The driving vehicle smartphone 200 is a terminal mounted on the driving vehicle 120 traveling on the road 110, and includes a GPS module 210 and a communication module 220 to upload sensor data.

The smartphone sensor 300 includes a gravity acceleration sensor 310 and a gyro sensor 320, and is embedded in the driving vehicle smartphone 200 to detect an impact event, which is a dangerous driving pattern of the driving vehicle 120. Generate sensor data. Here, the shock event of the driving vehicle 120 includes a driving situation event and a road situation event, the driving situation event occurs due to a sudden braking, sudden start or sudden turn of the driving vehicle 120, and the road situation event It is caused by portholes or bumps on the top. In addition, the sensor data may include data sensed by the gravitational acceleration sensor 310 and the gyro sensor 320 of the smartphone sensor 300; And data generated by the GPS module 210 of the driving vehicle smartphone 200 that can determine the location of the driving vehicle on the road, but is not limited thereto.

The smartphone app 400 is installed on the smartphone 200 of the driving vehicle, collects sensor data generated by the smartphone sensor 300, performs an analysis process, and generates road traffic hazard information.

The cloud server 500 collects all sensor data uploaded from at least one driving vehicle smartphone 200 to perform an analysis process, and creates and provides a road traffic risk information map. Here, the cloud server 500 may be installed with a database management system (DBMS) for big data analysis.

The following vehicle smartphone 600 is a terminal mounted on at least one rear vehicle 130 traveling behind the road 110, and provides a road traffic risk information map from the cloud server 500 Receive. Here, the following vehicle smart phone 600 includes a smart phone sensor 300 and a smart phone app 400, like the above-described driving vehicle smart phone 200, so that each impact event along the driving direction on the road By detecting the sensor data can be uploaded, and road traffic hazard information can be generated. That is, the rear vehicle 130 may be the driving vehicle 120 based on the point where the impact event occurs. Similarly, when another vehicle in front of the driving vehicle 120 detects an impact event, the rear vehicle 130 ).

In the road traffic hazard information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention, the driving vehicle smartphone 200 is installed with a smartphone app 400 to analyze real-time sensor data related to road traffic. To provide road traffic risk information to the driver, and the cloud server 500 not only performs the same analysis process as the smartphone app 400 installed on the driving vehicle smartphone 200, but also stores historical data to continue It is responsible for updating the training data set.

Accordingly, since most of the drivers possess smartphones to collect road traffic risk information, it is possible to construct a road traffic risk information collection and provision system more cost-effectively than in the prior art. At this time, in the case of a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, each of the smartphone app 400 and the cloud server 500 installed on the driving vehicle smartphone 200 Can perform the analysis process in duplicate according to the previously learned machine learning algorithm.

Meanwhile, FIG. 3 is a detailed configuration diagram of a driving vehicle smartphone and a smartphone app in the road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention.

Referring to FIG. 3, in the system for collecting and providing road traffic hazard information using a smartphone and a cloud server according to an embodiment of the present invention, a driving vehicle smartphone 200 includes a GPS module 210 and a communication module 220, and , The smartphone sensor 300 includes a gravity acceleration sensor 310 and a gyro sensor 320, and the smartphone app 400 includes a data storage unit 410, an analysis process execution unit 420, and an alarm generator ( 430).

The GPS module 210 of the driving vehicle smartphone 200 generates location information according to the position of the driving vehicle 120 on the road 110. In addition, the communication module 220 of the smart phone 200 of the driving vehicle communicates with the cloud server 500 to upload sensor data sensed by the smart phone sensor 300.

The gravitational acceleration sensor 310 of the smartphone sensor 300 is embedded in the smartphone 200 of the driving vehicle, and the vehicle body of the driving vehicle 120 responds to an impact event that is a dangerous driving pattern of the driving vehicle 120 The acceleration applied to is detected. In addition, the gyro sensor 320 of the smart phone sensor 300 is embedded in the smart phone 200 of the driving vehicle, and in response to an impact event that is a dangerous driving pattern of the driving vehicle 120, the driving vehicle 120 The angular velocity applied to the vehicle body is detected.

The smartphone app 400 analyzes sensor data detected by the smartphone sensor 300 collected in real time, and immediately warns the driver when a dangerous driving pattern of the driving vehicle 120 occurs. At this time, the smart phone app 400 is a variety of wireless communication, for example, 3G, 5G, LTE, Wi-Fi sensor data detected through the smart phone sensor 300 embedded in the driving vehicle smart phone 200. It also serves to transmit to the cloud server 500 through the like.

Specifically, the data storage unit 410 of the smartphone app 400 stores sensor data detected by the smartphone sensor 300, and the analysis process execution unit 420 of the smartphone app 400 Road traffic risk information is generated by performing an analysis process on the sensor data according to a machine learning algorithm, and the alarm generator 430 of the smartphone app 400 is a road generated by the analysis process execution unit 420 An alarm is generated so that traffic hazard information is immediately provided to the driver of the driving vehicle 120. In this case, it is preferable that the analysis process execution unit 420 performs the analysis process only when the sensor data detected by the smartphone sensor 300 exceeds a preset value.

Specifically, the analysis process execution unit 420 includes a data collection unit 421, a smoothing unit 422, a feature extraction unit 423, a data classification unit 424, a learning data set 425, and a road traffic. It includes a risk information generation unit 426.

The data collection unit 421 of the analysis process execution unit 420 collects the sensor data as raw data, and the smoothing execution unit 422 of the analysis process execution unit 420 is in the data collection unit 421 The collected raw data is smoothed at a predetermined analysis cycle. Here, as shown in FIG. 5 to be described later, the smoothing performing unit 422 may overlap the predetermined analysis periods with each other so as to minimize the non-sensing of the impact event.

A feature extraction unit 423 of the analysis process performing unit 420 extracts a feature vector of the sensor data smoothed by the smoothing unit 422, and a data classification unit 424 of the analysis process performing unit 420 Classifies the data to correspond to the feature vector extracted by the feature extraction unit 423 according to a previously learned machine learning algorithm.

The learning data set 425 of the analysis process execution unit 420 is downloaded from the cloud server 500 and is continuously provided to improve the machine learning algorithm when the data classification unit 424 classifies the data.

Accordingly, the road traffic risk information generation unit 426 of the analysis process execution unit 420 includes the road traffic risk information generation unit 426 from the data classified by the data classification unit 424. Road traffic risk information corresponding to an impact event can be generated.

Meanwhile, FIG. 4 is a detailed configuration diagram of a cloud server in a system for collecting and providing road traffic risk information using a smart phone and a cloud server according to an embodiment of the present invention.

Referring to FIG. 4, in the system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, the cloud server 500 performs a communication module 510, a data storage unit 520, and an analysis process. It includes a unit 530 and a road traffic hazard information map preparation unit 540.

The communication module 510 of the cloud server 500 communicates with the communication module 220 of the driving vehicle smartphone 200 to upload the sensor data.

The data storage unit 520 of the cloud server 500 stores all sensor data uploaded from at least one or more driving vehicle smartphones 200.

The analysis process execution unit 530 of the cloud server 500 generates road traffic risk information by performing an analysis process according to a machine learning algorithm for all sensor data uploaded from the at least one or more driving vehicle smartphones 200.

Road traffic risk information map creation unit 540 of the cloud server 500 is road traffic risk information so that the road traffic risk information generated by the analysis process execution unit 530 is provided to the rear vehicle 130 and the road manager in real time. Create a map.

Specifically, the analysis process execution unit 530 includes a data collection unit 531, a smoothing execution unit 532, a feature extraction unit 533, a data classification unit 534, a learning data set 535, and road traffic. The risk information generating unit 536 and the clustering performing unit 537 are performed.

The data collection unit 531 of the analysis process execution unit 530 collects the sensor data as raw data, and the smoothing execution unit 532 of the analysis process execution unit 530 is the data collection unit 531 The raw data collected in) are smoothed at a predetermined analysis cycle. In this case, the smoothing performing unit 532 overlaps the predetermined analysis periods with each other to minimize the non-detection of the impact event.

The feature extraction unit 533 of the analysis process execution unit 530 extracts a feature vector of the sensor data smoothed by the smoothing execution unit 532, and further, a data classification unit of the analysis process execution unit 530 534 classifies data to correspond to the feature vector extracted by the feature extraction unit 533 according to a machine learning algorithm, and the learning dataset 535 of the analysis process execution unit 530 is the data When data is classified by the classification unit 534, it is continuously provided to improve the machine learning algorithm, and the road traffic risk information generation unit 536 of the analysis process execution unit 530 is used in the data classification unit 534 Road traffic risk information corresponding to the impact event of the driving vehicle 120 is generated from the classified data. In other words, the extracted feature vector corresponds to a dangerous driving pattern for the driving vehicle 120, for example, shock events such as sudden braking, sudden start, sudden turn, and porthole through a learning dataset that is a previously learned machine learning algorithm. It is classified as road traffic hazard information.

As shown in FIG. 7 to be described later, the clustering execution unit 537 of the analysis process execution unit 530 clusters to display road traffic risk information for an event occurring in a preset distance range as one event information ( Grouping).

In other words, the cloud server 500 performs the same analysis process performed by the smart phone app 400 installed on the driving vehicle smartphone 200 described above, and is transmitted from the plurality of driving vehicle smartphones 200. It also collects and stores sensor data. At this time, the stored sensor data is used as a continuous learning data set for improving machine learning algorithms. In addition, road traffic risk information classified and generated by the cloud server 500 is prepared as a road traffic risk information map and provided to the rear vehicle 130 and road manager in real time.

On the other hand, Figure 5 is a view for explaining the smoothing of sensor data in the road traffic hazard information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention, Figure 6 is a smart phone according to an embodiment of the present invention It is a diagram for explaining the overlapping of the analysis cycle in the road traffic risk information collection and provision system using a cloud server.

As shown in Figure 5, in the road traffic risk information collection and provision system using a smartphone and a cloud server according to an embodiment of the present invention, the gravity acceleration sensor 310 and the gyro built in the driving vehicle smartphone 200 Since the sensor 320 is not a high-end sensor, the vibration of the data may be quite large, and in order to mitigate this, the raw data is smoothed.

Thereafter, the smoothed sensor data is aggregated in an analysis period of a predetermined time interval, and feature vectors for sensor data sensed through the gravity acceleration sensor 310 and the gyro sensor 320 are extracted. In this case, as shown in FIG. 6, by overlapping the analysis periods with each other, it is possible to minimize undetected event information due to division.

Meanwhile, FIG. 7 is a diagram illustrating grouping of shock events in a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, and FIG. 8 is a smart phone and a smart phone according to an embodiment of the present invention. It is a diagram illustrating the occurrence point of an impact event in a road traffic risk information collection and provision system using a cloud server.

In the case of a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, the cloud server 500 is an analysis performed by the smartphone app 400 installed on the driving vehicle smartphone 200 After performing the same process, road traffic hazard information is generated. At this time, in the case of the smartphone app 400 described above, it operates only when the sensor data detected through the gravity acceleration sensor 310 and the gyro sensor 320 exceeds a preset value, but the cloud server ( 500) may provide a road traffic hazard information map by performing an analysis process on all sensor data transmitted from the driving vehicle smartphone 200. At this time, the cloud server 500, as shown in Fig. 7, in order to increase the visibility of the information user when providing a road traffic hazard information map, a road for an event occurring within a preset distance range, for example, 50 m. Traffic risk information may be grouped, and an impact event grouping area for a road traffic risk situation may be displayed as one event information.

In addition, in the case of a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, the cloud server 500 is, as shown in FIG. 8, a shock event when providing a road traffic risk information map. Potholes or bumps, which are the points of occurrence, can be displayed.

In the case of a system for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, it is possible to improve the detection rate for the dangerous driving pattern of the driving vehicle by dualizing the analysis of the smartphone app and the analysis of the cloud server. have. Specifically, since the cloud server continuously updates the training data set by using pre-stored sensor data, machine learning algorithm performance may improve as time passes.

In addition, in the case of the system for collecting and providing road traffic hazard information using a smartphone and a cloud server according to an embodiment of the present invention, the smartphone app includes a threshold value for sensor data detected through a gravity acceleration sensor and a gyro sensor. Since the machine learning algorithm of the analysis process is operated only when it exceeds, it is possible to reduce the battery consumption of the smartphone of the driving vehicle and quickly generate road traffic risk information in real time.

[Method of collecting and providing road traffic risk information using smartphones and cloud servers]

9 is an operation flow diagram of a method for collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention, and FIG. 10 is an operation flow diagram for specifically explaining the analysis process of the smartphone shown in FIG. 9 And FIG. 11 is an operation flow diagram for specifically explaining the analysis process of the cloud server shown in FIG. 9.

9 to 11, the method of collecting and providing road traffic risk information using a smartphone and a cloud server according to an embodiment of the present invention corresponds to a dangerous driving pattern of a driving vehicle 120 traveling on the road 110. As a road traffic risk information collection and provision method that collects and provides road traffic risk information, first, the smart phone sensor 300 of the driving vehicle smartphone 200 mounted in the driving vehicle 120 detects an impact event and Data is generated (S110). Here, the smart phone sensor 300 is embedded in the smart phone 200 of the driving vehicle, and is applied to the vehicle body of the driving vehicle 120 in response to an impact event, which is a dangerous driving pattern of the driving vehicle 120. Gravity acceleration sensor 310 for detecting acceleration; And a gyro sensor 320 that is embedded in the driving vehicle smartphone 200 and detects an angular velocity applied to the vehicle body of the driving vehicle 120 in response to an impact event that is a dangerous driving pattern of the driving vehicle 120. Can include.

Specifically, the impact event of the driving vehicle 120 includes a driving situation event and a road situation event, and the driving situation event occurs due to sudden braking, sudden start, or sudden turn of the driving vehicle 120, and the road situation event is It is caused by portholes or bumps on the road. In addition, the sensor data may include data sensed by the gravitational acceleration sensor 310 and the gyro sensor 320 of the smartphone sensor 300; And data generated by the GPS module 210 of the driving vehicle smartphone 200 that can determine the location of the driving vehicle on the road, but is not limited thereto.

Next, the sensor data detected by the smartphone sensor 300 is collected through the smartphone app 400 installed on the smartphone 200 of the driving vehicle (S120). Here, the smartphone app 400 includes: a data storage unit 410 for storing sensor data sensed by the smartphone sensor 300; An analysis process execution unit 420 for generating road traffic risk information by performing an analysis process on the sensor data according to a machine learning algorithm; And an alarm generator 430 that generates an alarm so that road traffic hazard information generated by the analysis process execution unit 420 is immediately provided to the driver of the driving vehicle 120, but is not limited thereto.

Next, the driving vehicle smartphone 200 uploads the sensor data to the cloud server 500 (S130).

Next, the smartphone app 400 checks whether the sensor data is greater than a preset value (S140).

Next, when the sensor data is greater than a preset value, the smartphone app 400 performs an analysis process on the sensor data to generate road traffic risk information (S150).

Specifically, referring to FIG. 10, the sensor data is collected as raw data (S151), and then, the raw data is smoothed at a predetermined analysis period (S152), and at this time, the predetermined It is desirable to overlap the analysis cycles with each other. Next, a feature vector of the smoothed sensor data is extracted (S153), and then, data corresponding to the feature vector is classified according to a machine learning algorithm (S154). Next, road traffic risk information corresponding to the impact event of the driving vehicle 120 is generated from the classified data (S155), and thereafter, the road traffic risk information is immediately alerted to the driver of the driving vehicle 120 ( S156).

Next, the cloud server 500 performs an analysis process on all sensor data uploaded from at least one driving vehicle smartphone 200, and creates a road traffic risk information map (S160).

Specifically, referring to FIG. 11, sensor data uploaded from each of the at least one driving vehicle smartphone 200 is collected as raw data (S161), and thereafter, the raw data is smoothed at a predetermined analysis period (S162). . Next, a feature vector of the smoothed sensor data is extracted (S163), and then, data corresponding to the feature vector is classified according to a machine learning algorithm (S164). Next, road traffic risk information corresponding to the impact event of the driving vehicle 120 is generated from the classified data (S165), and thereafter, the road traffic risk information is clustered according to a preset distance range (S166). . Next, a road traffic risk information map is created so that the road traffic risk information is provided to the rear vehicle 130 and the road manager in real time.

Next, referring again to FIG. 9, the road traffic risk information map is provided to the rear vehicle 130 and the road manager in real time (S170).

Therefore, each of the smart phone app 400 and the cloud server 500 installed on the driving vehicle smart phone 200 performs an analysis process in duplicate according to a previously learned machine learning algorithm.

Meanwhile, according to a study by the US Department of Transportation, it is reported that about 80% of traffic accidents caused by driver error can be prevented by providing risk information about the danger situation ahead of the accident. As a result, according to an embodiment of the present invention, it is possible to inform a driver who collects information with a smart phone about road traffic hazard information in real time, for example, sudden braking, sudden acceleration, sudden turn due to drowsy driving, and the like. According to the embodiment of, the cloud server may provide road traffic hazard information in front to a plurality of rear vehicle drivers. In addition, according to an embodiment of the present invention, the cloud server transmits road traffic risk information such as a porthole to the road manager, so that a dangerous situation on the road can be quickly removed.

The above description of the present invention is for illustrative purposes only, and those of ordinary skill in the art to which the present invention pertains will be able to understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

110: road 120: driving vehicle
130: rear vehicle
200: driving vehicle smartphone 300: smartphone sensor
400: smartphone app 500: cloud server
600: rear vehicle smartphone
210: GPS module 220: communication module
310: gravity acceleration sensor (G sensor) 320: gyro sensor
410: data storage unit 420: analysis process execution unit
430: alarm generator
421: data collection unit 422: smoothing execution unit
423: feature extraction unit 424: data classification unit
425: training dataset 426: road traffic hazard information generation unit
510: communication module 520: data storage unit
530: Analysis process execution unit 540: Road traffic hazard information map preparation unit
531: data collection unit 532: smoothing execution unit
533: feature extraction unit 534: data classification unit
535: training dataset 536: road traffic hazard information generation unit
537: Clustering Execution Department

Claims (20)

In the road traffic risk information collection and provision system that collects and provides road traffic risk information corresponding to a dangerous driving pattern of a driving vehicle traveling on a road,
A terminal mounted on the driving vehicle 120 traveling on the road 110, comprising: a driving vehicle smartphone 200 including a GPS module 210 and a communication module 220 to upload sensor data;
A smartphone that includes a gravity acceleration sensor 310 and a gyro sensor 320, and is embedded in the driving vehicle smartphone 200 to detect an impact event that is a dangerous driving pattern of the driving vehicle 120 to generate sensor data Sensor 300;
A smart phone app 400 installed on the driving vehicle smart phone 200, collecting sensor data generated by the smart phone sensor 300, performing an analysis process, and generating road traffic risk information;
A cloud server 500 that collects all sensor data uploaded from at least one driving vehicle smartphone 200 to perform an analysis process, and creates and provides a road traffic risk information map; And
A terminal mounted on at least one rear vehicle 130 traveling behind the road 110, and a rear vehicle smartphone 600 receiving a road traffic risk information map from the cloud server 500 Including,
Each of the smartphone app 400 and the cloud server 500 installed on the driving vehicle smartphone 200 provides a smartphone and a cloud server, characterized in that the analysis process is performed in duplicate according to a previously learned machine learning algorithm. Used road traffic hazard information collection and provision system. The method of claim 1,
The shock event of the driving vehicle 120 includes a driving situation event and a road situation event, the driving situation event occurs due to a sudden braking, sudden start or sudden turn of the driving vehicle 120, and the road situation event is a porthole on the road. Road traffic risk information collection and provision system using a smartphone and a cloud server, characterized in that caused by or bumps. The method of claim 1,
The sensor data may include data sensed by the gravitational acceleration sensor 310 and the gyro sensor 320 of the smartphone sensor 300; And road traffic hazard information collection and provision system using a smartphone and a cloud server including data generated by the GPS module 210 of the driving vehicle smartphone 200 that can determine the location of the driving vehicle on the road. The method of claim 1, wherein the driving vehicle smartphone 200,
A GPS module 210 that generates location information according to the location of the driving vehicle 120 on the road 110; And
Road traffic hazard information collection and provision system using a smartphone and a cloud server including a communication module 220 that communicates with the cloud server 500 to upload sensor data detected by the smartphone sensor 300. The method of claim 1, wherein the smartphone sensor 300,
A gravity acceleration sensor 310 that is embedded in the driving vehicle smartphone 200 and detects an acceleration applied to the vehicle body of the driving vehicle 120 in response to an impact event, which is a dangerous driving pattern of the driving vehicle 120; And
It is embedded in the driving vehicle smartphone 200 and includes a gyro sensor 320 that detects an angular velocity applied to the vehicle body of the driving vehicle 120 in response to an impact event that is a dangerous driving pattern of the driving vehicle 120 Road traffic risk information collection and provision system using smart phones and cloud servers. The method of claim 1, wherein the smartphone app 400,
A data storage unit 410 for storing sensor data sensed by the smartphone sensor 300;
An analysis process execution unit 420 for generating road traffic risk information by performing an analysis process on the sensor data according to a machine learning algorithm; And
Road traffic using a smartphone and a cloud server including an alarm generator 430 that generates an alarm so that the road traffic risk information generated by the analysis process execution unit 420 is immediately provided to the driver of the driving vehicle 120 Risk information collection and provision system. The method of claim 6,
The analysis process execution unit 420 performs an analysis process only when the sensor data detected by the smartphone sensor 300 exceeds a preset value, characterized in that the road using a smartphone and a cloud server Traffic risk information collection and provision system. The method of claim 6, wherein the analysis process execution unit 420,
A data collection unit 421 collecting the sensor data as raw data;
A smoothing performing unit 422 for smoothing the raw data collected by the data collecting unit 421 at a predetermined analysis period;
A feature extraction unit 423 for extracting a feature vector of the sensor data smoothed by the smoothing unit 422;
A data classification unit 424 for classifying data to correspond to a feature vector extracted by the feature extraction unit 423 according to a machine learning algorithm;
A learning dataset 425 downloaded from the cloud server 500 and continuously provided to improve the machine learning algorithm when the data classification unit 424 classifies data; And
A road using a smartphone and a cloud server including a road traffic risk information generator 426 that generates road traffic risk information corresponding to an impact event of the driving vehicle 120 from the data classified by the data classification unit 424 Traffic risk information collection and provision system. The method of claim 8,
The smoothing execution unit 422 is a road traffic risk information collection and provision system using a smartphone and a cloud server, characterized in that the predetermined analysis period is overlapped with each other so as to minimize the detection of the impact event. The method of claim 1, wherein the cloud server 500,
A communication module 510 communicating with the communication module 220 of the driving vehicle smartphone 200 to upload the sensor data;
A data storage unit 520 for storing all sensor data uploaded from at least one or more driving vehicle smartphones 200;
An analysis process execution unit 530 for generating road traffic risk information by performing an analysis process according to a machine learning algorithm on all sensor data uploaded from at least one driving vehicle smartphone 200; And
Road traffic risk information map creation unit 540 for creating a road traffic risk information map so that the road traffic risk information generated by the analysis process execution unit 530 is provided to the rear vehicle 130 and the road manager in real time. Road traffic risk information collection and provision system using a smartphone and a cloud server comprising a. The method of claim 10, wherein the analysis process execution unit (530),
A data collection unit 531 collecting the sensor data as raw data;
A smoothing performing unit 532 for smoothing the raw data collected by the data collecting unit 531 at a predetermined analysis period;
A feature extraction unit 533 for extracting a feature vector of the sensor data smoothed by the smoothing execution unit 532;
A data classification unit 534 for classifying data to correspond to a feature vector extracted by the feature extraction unit 533 according to a machine learning algorithm;
A learning dataset 535 continuously provided to improve the machine learning algorithm when the data classification unit 534 classifies data;
A road traffic risk information generation unit 536 that generates road traffic risk information corresponding to an impact event of the driving vehicle 120 from the data classified by the data classification unit 534; And
Road traffic risk information collection and provision system using a smartphone and cloud server including a clustering execution unit 537 that groups road traffic risk information for events occurring in a preset distance range to be displayed as one event information . The method of claim 11,
The smoothing execution unit 532 is a road traffic risk information collection and provision system using a smartphone and a cloud server, characterized in that the predetermined analysis period is overlapped with each other so as to minimize the undetected impact event. In the road traffic risk information collection and provision method for collecting and providing road traffic risk information corresponding to the dangerous driving pattern of the driving vehicle 120 traveling on the road 110,
a) generating sensor data by sensing an impact event by the smart phone sensor 300 of the driving vehicle smart phone 200 mounted on the driving vehicle 120;
b) collecting sensor data sensed by the smart phone sensor 300 through the smart phone app 400 installed on the driving vehicle smart phone 200;
c) uploading the sensor data to the cloud server 500 by the driving vehicle smartphone 200;
d) the smart phone app 400 checking whether the sensor data is greater than a preset value;
e) when the sensor data is greater than a preset value, generating road traffic hazard information by performing, by the smartphone app 400, an analysis process for the sensor data;
f) performing, by the cloud server 500, an analysis process for all sensor data uploaded from at least one or more driving vehicle smartphones 200, and creating a road traffic risk information map; And
g) including the step of providing the road traffic risk information map to the rear vehicle 130 and the road manager in real time,
Each of the smartphone app 400 and the cloud server 500 installed on the driving vehicle smartphone 200 provides a smartphone and a cloud server, characterized in that the analysis process is performed in duplicate according to a previously learned machine learning algorithm. How to collect and provide road traffic risk information used. The method of claim 13,
The shock event of the driving vehicle 120 includes a driving situation event and a road situation event, the driving situation event occurs due to a sudden braking, sudden start or sudden turn of the driving vehicle 120, and the road situation event is a porthole on the road. Road traffic risk information collection and provision method using a smartphone and a cloud server, characterized in that caused by or bumps. The method of claim 13,
The sensor data may include data sensed by the gravitational acceleration sensor 310 and the gyro sensor 320 of the smartphone sensor 300; And road traffic risk information collection and provision method using a smartphone and a cloud server including data generated by the GPS module 210 of the driving vehicle smartphone 200 that can determine the location of the driving vehicle on the road. The method of claim 13, wherein the smart phone sensor 300,
A gravity acceleration sensor 310 that is embedded in the driving vehicle smartphone 200 and detects an acceleration applied to the vehicle body of the driving vehicle 120 in response to an impact event, which is a dangerous driving pattern of the driving vehicle 120; And
It is embedded in the driving vehicle smartphone 200 and includes a gyro sensor 320 that detects an angular velocity applied to the vehicle body of the driving vehicle 120 in response to an impact event that is a dangerous driving pattern of the driving vehicle 120 How to collect and provide road traffic risk information using smart phones and cloud servers. The method of claim 13, wherein the smartphone app (400),
A data storage unit 410 for storing sensor data sensed by the smartphone sensor 300;
An analysis process execution unit 420 for generating road traffic risk information by performing an analysis process on the sensor data according to a machine learning algorithm; And
Road traffic using a smartphone and a cloud server including an alarm generator 430 that generates an alarm so that the road traffic risk information generated by the analysis process execution unit 420 is immediately provided to the driver of the driving vehicle 120 How to collect and provide risk information. The method of claim 13, wherein step e),
e-1) collecting the sensor data as raw data;
e-2) smoothing the raw data at a predetermined analysis period;
e-3) extracting a feature vector of the smoothed sensor data;
e-4) classifying data corresponding to the feature vector according to a machine learning algorithm;
e-5) generating road traffic hazard information corresponding to an impact event of the driving vehicle 120 from the classified data; And
e-6) A method of collecting and providing road traffic risk information using a smartphone and a cloud server comprising the step of immediately alerting the driver of the driving vehicle 120 of the road traffic risk information. The method of claim 18,
The method of collecting and providing road traffic risk information using a smartphone and a cloud server, characterized in that the predetermined analysis periods are overlapped with each other so as to minimize undetected impact events in step e-2). The method of claim 13, wherein step f),
f-1) collecting sensor data uploaded from each of the at least one driving vehicle smartphone 200 as raw data;
f-2) smoothing the raw data at a predetermined analysis period;
f-3) extracting a feature vector of the smoothed sensor data;
f-4) classifying data corresponding to the feature vector according to a machine learning algorithm;
f-5) generating road traffic hazard information corresponding to an impact event of the driving vehicle 120 from the classified data;
f-6) clustering the road traffic risk information according to a preset distance range; And
f-7) Road traffic risk information collection using a smartphone and a cloud server including the step of creating a road traffic risk information map so that the road traffic risk information is provided to the rear vehicle 130 and the road manager in real time Delivery method.

Images