Apparatus, method and system for controlling automatic horn output of autonomous vehicle

Abstract

An automatic horn output control device, method, and system for an autonomous vehicle are disclosed. An automatic horn output control method for an autonomous vehicle according to one embodiment of the present invention may include: (a) acquiring surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle including an autonomous vehicle; (b) calculating a collision risk with said object based on said surrounding data; (c) determining whether a preset horn output condition is satisfied based on said collision risk; and (d) controlling the horn output of said target vehicle based on said determination result.

Description

Technology Field

[0001] The present invention relates to an automatic horn output control device, method, and system for an autonomous vehicle. Background Technology

[0002] Autonomous vehicles utilize various sensor and communication technologies to avoid collisions while driving and automatically perform driving functions such as lane keeping and speed control. While existing autonomous vehicle technologies possess automated driving functions—including lane keeping, lane changing, acceleration / deceleration, and collision avoidance—they currently lack sufficient features for warning against collisions caused by dangerous behavior of surrounding vehicles or warnings regarding surrounding objects to prevent collisions in situations where avoidance is impossible.

[0003] In the case of conventional technology related to this, a method was primarily used to output various horn sounds by measuring distance using LiDAR (Light Detection And Ranging) sensors. Furthermore, existing technology performed warning judgments solely using the vehicle's own sensors and adopted a method of selectively outputting horn sounds of varying volumes depending on the situation.

[0004] However, even when autonomous vehicles are stationary, collision risks frequently arise due to the unexpected behavior of surrounding vehicles or pedestrians, and active response capabilities to address this are currently inadequate. In particular, there is a problem in that it is difficult to notify the driver of the situation without intervention; if there is no function to provide active warnings to surrounding vehicles and objects without driver intervention in situations where collision avoidance is impossible while the autonomous driving function is operating, the safety of the autonomous vehicle and its surroundings cannot be guaranteed. Furthermore, with the recent emergence of companies transitioning from LiDAR to cameras for autonomous vehicle sensors, there is a growing need for lighter and more efficient sensor utilization methods.

[0005] The technology forming the background of the present invention is disclosed in Korean Registered Patent Publication No. 10-2790107. The problem to be solved

[0006] The present invention aims to solve the problems of the aforementioned prior art by providing an automatic horn output control device, method, and system for an autonomous vehicle that can prevent collisions with surrounding vehicles and objects by automatically outputting a horn in situations where avoidance through collision avoidance control is impossible during driving.

[0007] The present invention aims to solve the problems of the aforementioned prior art by providing an automatic horn output control device, method, and system for an autonomous vehicle that can provide warnings to pedestrians, adjacent vehicles, etc., around the target vehicle by calculating the collision risk in real time using an artificial intelligence-based analysis model.

[0008] The present invention aims to solve the problems of the aforementioned prior art and to provide an automatic horn output control device, method, and system for an autonomous vehicle that can enhance the safety and effectiveness of the autonomous driving environment through an active warning function that responds to various dangerous situations.

[0009] However, the technical problems that the embodiments of the present invention aim to solve are not limited to the technical problems described above, and other technical problems may exist. means of solving the problem

[0010] As a technical means for achieving the above-mentioned technical problem, a method for controlling the automatic horn output of an autonomous vehicle according to one embodiment of the present invention may include: (a) acquiring surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle including an autonomous vehicle; (b) calculating a collision risk with said object based on said surrounding data; (c) determining whether a preset horn output condition is satisfied based on said collision risk; and (d) controlling the horn output of said target vehicle based on said determination result.

[0011] Additionally, the above step (d) may selectively perform the horn output based on the result of determining whether the preset horn limit condition is satisfied.

[0012] In addition, the above horn restriction condition may include a condition in which the object is a vehicle and the vehicle is in a state where the hazard lights are on.

[0013] In addition, the above step (b) can be performed based on a pre-trained artificial intelligence-based analysis model.

[0014] In addition, the above analysis model may be a Temporal Convolutional Network (TCN)-based model.

[0015] Additionally, the horn output condition may include at least one of the following: a condition in which the distance to the object decreases to a preset collision risk distance or less; a condition in which the object maintains a stationary state for a preset time or longer; and a condition in which the object enters within the stopping distance of the target vehicle.

[0016] Additionally, step (a) above may acquire surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle.

[0017] In addition, the above step (b) can calculate the collision risk based on at least one of the Time to Collision (TTC) with the object, the stopping distance of the target vehicle, and the margin distance obtained by subtracting the stopping distance from the actual distance with the object.

[0018] In addition, the automatic horn output control method of an autonomous vehicle according to one embodiment of the present invention may include the step of (e) recording log data including at least one of the output time, situation information, and collision risk level after the horn is output through the target vehicle.

[0019] Meanwhile, an automatic horn output control device for an autonomous vehicle according to one embodiment of the present invention may include a data collection unit that acquires surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle including an autonomous vehicle; a calculation execution unit that calculates a collision risk level with said object based on said surrounding data; a condition determination unit that determines whether a preset horn output condition is satisfied based on said collision risk level; and an output control unit that controls the horn output of said target vehicle based on said determination result.

[0020] In addition, the output control unit may selectively perform the horn output according to the result of determining whether a preset horn limit condition is satisfied.

[0021] In addition, the above-mentioned operation unit can calculate the collision risk based on a pre-trained artificial intelligence-based analysis model.

[0022] In addition, the data collection unit can acquire the surrounding data by using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle.

[0023] Meanwhile, a control system for an autonomous vehicle having an automatic horn output function according to one embodiment of the present invention may include a sensor module for collecting surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle corresponding to the autonomous vehicle; a control module for calculating a collision risk with said object based on said surrounding data, determining whether a preset horn output condition is satisfied based on said collision risk, and generating a control signal for controlling the horn output of said target vehicle based on said determination result; and a horn module configured to output a horn signal based on said control signal.

[0024] The means for solving the problem described above are merely exemplary and should not be interpreted as intended to limit the present invention. In addition to the exemplary embodiments described above, additional embodiments may exist in the drawings and the detailed description of the invention. Effects of the invention

[0025] According to the means for solving the problem of the present invention described above, an automatic horn output control device, method, and system for an autonomous vehicle can be provided, which can prevent collisions with surrounding vehicles and objects by automatically outputting a horn under situations where avoidance through collision avoidance control is impossible during the driving of the autonomous vehicle.

[0026] According to the means for solving the problem of the present invention described above, it is possible to provide an automatic horn output control device, method, and system for an autonomous vehicle that can calculate the collision risk in real time using an artificial intelligence-based analysis model and provide warnings to pedestrians, adjacent vehicles, etc., around the target vehicle.

[0027] According to the means for solving the problem of the present invention described above, it is possible to provide an automatic horn output control device, method, and system for an autonomous vehicle that can enhance the safety and effectiveness of the autonomous driving environment through an active warning function that responds to various dangerous situations.

[0028] According to the solution to the problem of the present invention described above, collision risk can be accurately evaluated by analyzing the change patterns of time series data through an artificial intelligence model based on a Temporal Convolutional Network (TCN).

[0029] According to the solution to the problem of the present invention described above, traffic safety can be improved by preventing unnecessary horn output through horn limiting conditions while providing active warnings in actual dangerous situations.

[0030] However, the effects obtainable from this invention are not limited to those described above, and other effects may exist. Brief explanation of the drawing

[0031] FIG. 1 is a schematic diagram of a control system of an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention. FIG. 2 is a conceptual diagram illustrating the operation flow of a control system for an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention. FIG. 3 is a conceptual diagram showing an autonomous vehicle automatic horn output framework disclosed herein. FIG. 4 is a conceptual diagram illustrating an automatic horn output control process using vehicle-to-object communication (V2X) according to one embodiment of the present invention. FIG. 5 is a schematic diagram of an automatic horn output control device of an autonomous vehicle according to one embodiment of the present invention. FIG. 6 is an operation flowchart of an automatic horn output control method for an autonomous vehicle according to one embodiment of the present invention. Specific details for implementing the invention

[0032] Embodiments of the present invention are described below with reference to the attached drawings to enable those skilled in the art to easily implement the invention. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals.

[0033] Throughout this specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected," but also cases where they are "electrically connected" or "indirectly connected" with other elements interposed between them.

[0034] Throughout the entire specification, when a component is described as being located "on," "on top," "on top," "under," "on bottom," or "on bottom" of another component, this includes not only cases where the component is in contact with the other component but also cases where another component exists between the two components.

[0035] Throughout this specification, when a part is described as "comprising" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0036] The present invention relates to an automatic horn output control device, method, and system for an autonomous vehicle.

[0037] FIG. 1 is a schematic diagram of a control system of an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention.

[0038] Referring to FIG. 1, a control system (10) of an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention may include a control module (100), a sensor module (200), and a horn module (300) provided in a target vehicle (1), a roadside base station (Road Side Unit, RSU; 400), a user terminal (500), and a database (600). Meanwhile, in the description of the embodiment of the present invention, the control module (100) mounted in the target vehicle (1) may be referred to as an automatic horn output control device (100) of an autonomous vehicle, and for convenience of explanation, it will be referred to uniformly as 'control module (100)' below.

[0039] The control module (100), sensor module (200), horn module (300), roadside base station (400), user terminal (500), and database (600) can communicate with each other through a network (20). The network (20) refers to a connection structure that enables information exchange between each node, such as terminals and servers. Examples of such a network (20) include, but are not limited to, a 3GPP (3rd Generation Partnership Project) network, an LTE (Long Term Evolution) network, a 5G network, a WIMAX (World Interoperability for Microwave Access) network, the Internet, a LAN (Local Area Network), a Wireless LAN (Wireless Local Area Network), a WAN (Wide Area Network), a PAN (Personal Area Network), a Wi-Fi network, a Bluetooth network, a satellite broadcasting network, an analog broadcasting network, and a DMB (Digital Multimedia Broadcasting) network.

[0040] The user terminal (500) can be any type of wireless communication device, such as a smartphone, smartpad, tablet PC, PCS (Personal Communication System), GSM (Global System for Mobile communication), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet) terminal.

[0041] Meanwhile, in the description of the embodiment of the present invention, the roadside base station (400) is a communication device installed on the roadside that supports vehicle-to-infrastructure communication (V2I), and can provide intersection signal information, road condition information, etc. to the target vehicle (1) in real time.

[0042] These roadside base stations (400) can be deployed at intersections, major road sections, accident-prone areas, etc., and can transmit and receive data with the target vehicle (1) through the network (20) according to communication methods such as 5G network, WAVE (Wireless Access in Vehicular Environments) communication, DSRC (Dedicated Short Range Communications). For example, the roadside base station (400) can transmit information on the current status (red, yellow, green) of the traffic light and the scheduled time for signal switching in a broadcasting manner, and the control module (100) of the target vehicle (1) can receive this and use it to determine the horn output condition.

[0043] Additionally, in the description of the embodiments of the present invention, the database (600) may be a server system that collects, stores, and manages log data generated from the control module (100) of the target vehicle (1). The database (600) may store information such as horn output history, collision risk, surrounding data, and whether horn output conditions are met in a time-series format, and may be implemented as a cloud-based distributed storage system.

[0044] The data accumulated in the database (600) in this way can be used for various purposes, such as analyzing the cause of accidents, improving autonomous driving algorithms, analyzing traffic patterns, and identifying dangerous sections. In addition, the database (600) can integrate and analyze data collected from multiple target vehicles (1) to identify the traffic risk level of a specific region or time period, and provide this as training data for a TCN-based model.

[0045] Additionally, referring to FIG. 1, the target vehicle (1), including the autonomous vehicle, may be equipped with a control module (100), a sensor module (200), and a horn module (300) as described in detail below. For reference, in the description of the embodiments of the present invention, the term "autonomous vehicle" may refer to a vehicle capable of recognizing the driving environment without driver operation or with minimal intervention, and independently determining the driving path to move to a destination.

[0046] For example, autonomous vehicles can be classified from Level 0 (non-automated) to Level 5 (fully automated) according to the classification criteria of the Society of Automotive Engineers (SAE), and the subject vehicle (1) disclosed herein may include a vehicle equipped with autonomous driving functions of Level 2 (partially automated) or higher. Autonomous vehicles can perform driving control functions such as lane keeping, lane changing, and acceleration / deceleration, but in situations where collision avoidance control alone cannot respond (e.g., sudden approach of a vehicle behind while stopped, reversing of a vehicle in front while waiting at a traffic light), it is necessary to transmit warnings to surrounding vehicles and objects. In such situations, the automatic horn output control technology of the present invention can improve traffic safety by actively notifying of dangerous situations without driver intervention.

[0047] Meanwhile, among the sub-modules mounted on the target vehicle (1), the control module (100) can be installed inside the vehicle as an onboard type and can function as a central control unit that processes surrounding data received from the sensor module (200) in real time and controls the horn module (300) by determining whether to output a horn. The control module (100) can be implemented in the form of an ECU (Electronic Control Unit) and may include a high-performance processor, memory, communication interface, etc.

[0048] For example, the control module (100) can be linked with the sensor module (200) and the horn module (300) through a network (20) based on CAN (Controller Area Network) communication, LIN (Local Interconnect Network) communication, Ethernet communication, etc. Meanwhile, the control module (100) may include functional blocks such as a data collection unit (110), a calculation execution unit (120), a condition determination unit (130), an output control unit (140), and a log recording unit (150), as described below, and each functional block may be implemented in the form of a software module or a hardware module, as described in detail below.

[0049] In addition, the sensor module (200) may include a plurality of sensors mounted on the front, side, and rear of the target vehicle (1), and may measure information such as the location, distance, and speed of surrounding objects in real time. Specifically, the sensor module (200) may include at least one of an ultrasonic sensor, a radar sensor, and a camera sensor.

[0050] For example, the ultrasonic sensor can be placed on the front bumper, rear bumper, side, etc. of the target vehicle (1) to measure the distance to a nearby object (e.g., 0.2 m to 5 m), and can be particularly useful for detecting the risk of collision while driving at low speed or in a stationary state.

[0051] As another example, radar sensors can measure the distance and relative speed of objects at medium to long ranges (e.g., 1 m to 200 m) and can provide stable measurement performance even in adverse weather conditions.

[0052] As another example, camera sensors can recognize the shape, color, and movement of objects, classify object types such as vehicles, pedestrians, and bicycles through image processing algorithms, and determine the status of hazard lights and turn signals.

[0053] Additionally, the horn module (300) can function as an actuator that outputs a horn sound based on a control signal received from the control module (100). The horn module (300) can utilize a horn device that is basically installed in the target vehicle (1) and can receive a horn output signal through an electrical connection with the control module (100).

[0054] For example, the horn module (300) can receive a Pulse Width Modulation (PWM) signal or a digital control signal from the control module (100) to determine whether to output a horn sound. According to one embodiment of the present invention, the horn module (300) can output a horn sound of the same intensity that is preset, which can induce clear awareness of danger in surrounding objects by providing a consistent warning signal without performing complex volume adjustments depending on the situation, and the horn module (300) can operate independently of the vehicle's existing horn system or be integrated with the existing system to support both manual horn operation by the driver and automatic horn output.

[0055] FIG. 2 is a conceptual diagram illustrating the operation flow of a control system for an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention.

[0056] Referring to FIG. 2, a control system (10) of an autonomous vehicle equipped with an automatic horn output function according to one embodiment of the present invention can be divided into three layers according to a time-series operation flow. First, in the perception layer corresponding to the first process, surrounding environment data is collected through a sensor module (200), and the location, distance, speed, etc. of an object can be recognized.

[0057] Next, in the decision layer corresponding to the second process, the control module (100) can calculate the collision risk based on the data received from the recognition layer and determine whether to output the horn by considering the horn output conditions and the horn limit conditions.

[0058] Subsequently, in the output layer corresponding to the third process, an actual horn sound is output through the horn module (300) according to the judgment result of the decision layer, and log data can be recorded. This multi-layered process can be repeatedly performed in real time to provide continuous monitoring of the surrounding environment and active warning functions.

[0059] Below, the operation flow of the control system (10) of an autonomous vehicle equipped with an automatic horn output function disclosed herein will be explained in detail, focusing on the function of the control module (100).

[0060] FIG. 3 is a conceptual diagram showing an autonomous vehicle automatic horn output framework disclosed herein.

[0061] Referring to FIG. 3, the autonomous vehicle automatic horn output framework disclosed herein may represent a data processing flow consisting of input data, a TCN-based module, and an output stage. In the input data stage, sensor data obtained from ultrasonic sensors, radar sensors, and camera sensors, as well as vehicle speed and vehicle acceleration information of the target vehicle (1), may be collected. This raw data may be transmitted to a derived variable calculation stage and processed into derived variables such as TTC (Time to Collision), stopping distance, and margin distance.

[0062] Additionally, the processed derived variable is input into a TCN-based module (collision risk calculation module) to analyze the change pattern of the time series data, and a collision risk expressed as a value between 0 and 1 can be calculated. The calculated collision risk passes through a horn restriction condition gate, and the final decision on whether to output the horn is made only when the horn restriction condition (e.g., whether the hazard lights of the vehicle ahead are on) is not satisfied. This framework can operate in real time inside the target vehicle (1) in an onboard format.

[0063] Specifically, the control module (100) can acquire surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle (1) including an autonomous vehicle.

[0064] Specifically, among the surrounding data that can be collected from the target vehicle (1), the location information may include relative coordinate information of an object based on the target vehicle (1). For example, the location information may be expressed as x-axis (lateral) and y-axis (longitudinal) coordinate values ​​in a two-dimensional planar coordinate system and may be obtained through a radar sensor or a camera sensor. When using a camera sensor, the pixel coordinates of the object can be converted into actual spatial coordinates through an image processing algorithm, and when using a stereo camera, depth information can be additionally obtained to generate three-dimensional location information.

[0065] In addition, among the surrounding data, distance information may refer to the straight-line distance or Euclidean distance between the target vehicle (1) and the object. Distance information can be measured using the difference in transmission and reception times of sound waves via an ultrasonic sensor, and can be particularly effective for measuring the distance of a short-range object (e.g., within 5m). As another example, a radar sensor can measure the distance to a medium-range or long-range object using the reflection time of electromagnetic waves. Distance information can be used as input data for calculating TTC and determining collision risk.

[0066] Additionally, among the surrounding data, speed information may include the moving speed of the object and the relative speed with respect to the target vehicle (1). The speed information may be directly measured using the Doppler effect of the radar sensor or indirectly calculated by differentiating the amount of position change over consecutive time frames. For example, the change in the object's position in consecutive video frames acquired from a camera sensor may be tracked, and the speed may be calculated by dividing the time interval between frames. The speed information may be used to determine whether the object is approaching, calculate TTC, calculate stopping distance, etc.

[0067] In addition, according to one embodiment of the present invention, the control module (100) may acquire various surrounding data in addition to location information, distance information, and speed information. For example, acceleration information of an object may be calculated as the rate of change of speed over time and may be used to detect sudden braking or sudden acceleration behavior of the object. As another example, type information of the object (vehicle, pedestrian, bicycle, motorcycle, etc.) may be classified through a camera sensor and an image recognition algorithm. As yet another example, if the object is a vehicle, vehicle status information such as the hazard light status, turn signal operation status, and brake light status may be recognized through a camera sensor. These additional information may be used to determine horn output conditions and horn restriction conditions.

[0068] For example, the control module (100) can acquire surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle (1).

[0069] More specifically, the control module (100) can receive ambient data from the sensor module (200) periodically or on an event basis. For example, the control module (100) can receive sensor data from the sensor module (200) at a frequency of 10 Hz to 100 Hz (e.g., at intervals of 10 ms to 100 ms), which can provide sufficient time resolution for real-time collision risk detection. Data transmission between the sensor module (200) and the control module (100) can be performed via in-vehicle communication protocols such as CAN communication, Ethernet communication, or UART (Universal Asynchronous Receiver / Transmitter) communication. Additionally, the sensor module (200) can transmit timestamp information along with the sensor data, thereby enabling the control module (100) to accurately determine the synchronization and order of the time-series data.

[0070] Additionally, the control module (100) can calculate the risk of collision with objects existing around the target vehicle (1) based on acquired surrounding data. For reference, in the description of the embodiments of the present invention, 'collision risk' is an indicator that quantitatively indicates the possibility of collision between the target vehicle (1) and surrounding objects, and is expressed as a real value between 0 and 1, so that the closer the collision risk is to 0, the lower the possibility of collision, and the closer it is to 1, the higher the possibility of collision, but is not limited thereto.

[0071] According to one embodiment of the present invention, a control module (100) can determine whether the horn output condition is satisfied by comparing the calculated collision risk with a preset threshold value. For example, if the collision risk is 0.7 or higher, it may be determined that the horn output condition is satisfied, but the threshold value may be adjusted according to the driving environment, vehicle characteristics, user settings, etc. The collision risk is calculated through a TCN-based model and may be derived as a result of comprehensively considering derived variables such as TTC, stopping distance, and margin distance.

[0072] For example, the control module (100) can operate to calculate the collision risk based on a pre-trained artificial intelligence-based analysis model.

[0073] In this regard, according to one embodiment of the present invention, the control module (100) can calculate the collision risk using an analysis model corresponding to a Temporal Convolutional Network (TCN) based model.

[0074] Specifically, TCN (Temporal Convolutional Network)-based models are deep learning models based on Convolutional Neural Networks designed to process time-series data, and they can have the advantages of higher learning stability and easier parallel processing compared to Recurrent Neural Networks (RNN) or LSTM (Long Short-Term Memory). TCN-based models can effectively learn long-term dependencies by utilizing Causal Convolution and Dilated Convolution techniques, and can mitigate problems such as Gradient Vanishing or Gradient Exploding.

[0075] According to one embodiment of the present invention, a TCN-based model can be trained using a supervised learning method with prior collected training data, and the training data may include sensor data measured in various driving scenarios and information on whether an actual collision occurred. The trained TCN-based model can receive time-series change patterns of TTC, stopping distance, and safety margin as input and infer the collision risk at the current time point in real time. The structure of such a TCN-based model may specifically include a plurality of convolution layers, activation functions, normalization layers, dropout layers, etc., and the hyperparameters of the model may be optimized according to the characteristics of the training data and the required accuracy.

[0076] Additionally, according to one embodiment of the present invention, the control module (100) can calculate the collision risk based on at least one of the Time to Collision (TTC) with the object, the stopping distance of the target vehicle (1), and the margin distance obtained by subtracting the stopping distance from the actual distance with the object.

[0077] In addition, according to one embodiment of the present invention, the TTC with the object can be calculated based on the following Equation 1.

[0078] [Equation 1]

[0079]

[0080] Here, TTC is the estimated time to collision if the two objects (the target vehicle and surrounding objects) maintain their current positions, speeds, and paths at a given point in time, and d rel is the relative distance (m) between the target vehicle (1) and the object, which can be calculated as Euclidean distance. rel represents the relative speed (m / s) between the target vehicle (1) and the object, and can be calculated through vector operations. cosθ is the cosine value of the angle between the directions of travel of the two objects, and can be calculated as cosθ=1 when the two objects are moving in the same direction and cosθ=-1 when they are moving in opposite directions. In this regard, the smaller the calculated TTC value, the shorter the time to collision, which may increase the risk of collision. For example, if the TTC is 2 seconds or less, it may be judged as a high-risk situation, but it is not limited to this.

[0081] In addition, the control module (100) can calculate the margin based on the following Equation 2.

[0082] [Equation 2]

[0083]

[0084] Here, d margin is the clearance distance (m), and d actualis the distance (m) to an actual object (vehicle, pedestrian, etc.) recognized around the target vehicle (1), and d stop is the stopping distance of the target vehicle (1).

[0085] In summary, the control module (100) disclosed herein may consider various derived variables, such as TTC, stopping distance, and clearance distance, to calculate the collision risk. These derived variables can be calculated in real time based on raw data (position, distance, speed, etc.) obtained from the sensor module (200), and each derived variable can evaluate the collision risk from different perspectives.

[0086] For example, TTC represents the time to collision in terms of time, and the margin of error represents the safety margin for collision avoidance in terms of space. By comprehensively considering these multiple derived variables, the control module (100) can perform a more accurate and robust collision risk assessment compared to the case where only a single variable is used. In addition, the time series change patterns of the derived variables are input into a TCN-based model, which can enable the calculation of risk that considers not only the value at the current moment but also past change trends.

[0087] Additionally, the control module (100) can determine whether a preset horn output condition is satisfied based on the calculated collision risk.

[0088] Specifically, the control module (100) can determine whether the horn output condition is satisfied by comprehensively considering surrounding environment information along with the calculated collision risk. For example, the control module (100) can first determine that the collision risk is a candidate situation for horn output if it is greater than or equal to a preset first threshold value (e.g., 0.7). Next, the control module (100) can further determine whether the current driving situation corresponds to the horn output condition (first condition, second condition, third condition, etc.).

[0089] For example, the control module (100) can check whether the current distance information is less than or equal to the collision risk distance, whether the object's stopping time is greater than or equal to the threshold time, and whether the margin distance is negative. The control module (100) can finally decide to output a horn only when both the collision risk condition and the horn output condition are satisfied, and this multi-stage judgment structure can minimize unnecessary horn outputs caused by false positives. In addition, the control module (100) can improve the accuracy of the judgment by additionally considering V2I data (traffic light status, road conditions, etc.) received from the roadside base station (400).

[0090] For example, the control module (100) can determine whether a horn output condition is satisfied, which includes at least one of a condition in which the distance to an object is reduced to a preset collision risk distance (first condition), a condition in which the object remains in a stationary state for a preset time or longer (second condition), and a condition in which the object enters within the stopping distance of the target vehicle (1) (third condition).

[0091] In this regard, the first condition among the aforementioned horn output conditions may correspond to a situation where the distance to the object decreases to a level below a preset collision risk distance. For example, the first condition may be satisfied if, while the target vehicle (1) is stopped waiting at a traffic light, the vehicle in front slides in neutral gear or unintentionally reverses, causing the distance between the target vehicle (1) and the vehicle to continuously narrow. The reason a horn must be output in such a situation may be to prevent a contact accident by making the driver of the vehicle in front aware that the vehicle is reversing and inducing immediate braking action.

[0092] In addition, the control module (100) can determine whether the first condition is satisfied by checking whether the calculated collision risk is greater than or equal to the first threshold value (e.g., 0.8), whether the actual measured distance is less than or equal to the collision risk distance (e.g., 0.5m), and whether the distance shows a decreasing trend over time. In addition, the control module (100) can more accurately determine whether the first condition is satisfied by detecting a pattern in which the TTC value decreases rapidly through a TCN-based model.

[0093] In contrast to this, the second condition may correspond to a situation where the object remains in a stationary state for longer than a preset time. For example, the second condition may be satisfied if, at an intersection, the traffic light turns green but the vehicle ahead does not start moving for a certain period of time (e.g., 3 seconds) or longer due to driver inattention. In such a situation, the reason a horn should be sounded is to induce the vehicle ahead to start moving in order to minimize delays in traffic flow and to prevent sudden braking or additional waiting time for the vehicle behind.

[0094] In addition, the control module (100) can determine whether the second condition is satisfied by confirming that the traffic light status received from the roadside base station (400) is green, confirming through the camera sensor that there are no pedestrians, confirming that the speed of the vehicle ahead is 0, and confirming that the stopping time is greater than or equal to the threshold time (T), and comprehensively determining whether the collision risk is greater than or equal to the second threshold value (e.g., 0.6). In addition, the control module (100) can distinguish that it is a situation of simple inattention rather than an emergency situation by additionally confirming that the hazard lights of the vehicle ahead are not illuminated.

[0095] Additionally, the third condition may respond to a situation where an object enters within the stopping distance of the target vehicle (1). For example, when the target vehicle (1) is traveling at a certain speed, if a pedestrian or bicycle suddenly enters the roadway and is located closer than the stopping distance considering the target vehicle (1)'s current speed and deceleration ability, the third condition may be satisfied. The reason a horn must be sounded in such a situation may be to induce the object to quickly recognize danger in a dangerous situation where a collision cannot be avoided by braking the target vehicle (1) alone, thereby causing the object to take avoidance action on its own.

[0096] In addition, the control module (100) can determine whether the third condition is satisfied by checking whether the margin distance (actual distance - stopping distance) is negative, the collision risk is greater than or equal to the third threshold (e.g., 0.9), and the TTC value is very small (e.g., less than 1 second). In addition, the control module (100) can determine in real time whether the third condition is satisfied by detecting a time series pattern in which the margin distance decreases rapidly through a TCN-based model.

[0097] However, this is not limited to this, and the horn output conditions may be additionally defined according to various driving scenarios. For example, this may include a case where a vehicle on the side abruptly changes lanes without a turn signal and enters the driving path of the target vehicle (1) while the target vehicle (1) is driving (a side cutting-in situation).

[0098] As another example, it may include a case where a pedestrian approaches from the rear while the target vehicle (1) is reversing in a parking lot, and as yet another example, it may include a case where another vehicle approaches from the rear of the target vehicle (1) at excessive speed, posing a risk of collision. These extended embodiments may be added through software updates or user settings of the control module (100), and independent thresholds and judgment logic may be applied for each condition.

[0099] Additionally, the control module (100) can control the horn output of the target vehicle (1) based on the result of determining the horn output condition.

[0100] Specifically, the control module (100) can transmit a control signal to the horn module (300). For example, the control module (100) can transmit a digital output signal (High / Low) to the horn module (300) to control the horn output ON / OFF. For another example, the control module (100) can transmit a horn output command message via a CAN communication protocol, and the message may include information such as whether to output the horn, the duration of the output, and the output pattern. According to one embodiment of the present invention, the duration of the horn output may be set to 0.5 seconds to 2 seconds, which can provide a sufficient warning effect to the surroundings while preventing excessive noise generation.

[0101] In addition, according to one embodiment of the present invention, the control module (100) may apply a cooldown logic that suppresses re-output for a certain period of time (e.g., 5 seconds) after the horn output to prevent repetitive horn output in the same situation.

[0102] According to one embodiment of the present invention, the control module (100) can control the horn output to be selectively performed on the target vehicle (1) based on the result of determining whether a preset horn limit condition is satisfied.

[0103] For example, the control module (100) can control the target vehicle (1) not to output a horn by considering a horn restriction condition that includes a condition in which an object existing (located) around the target vehicle (1) is of the vehicle type and the vehicle is in a state where the emergency light is on.

[0104] In this regard, the aforementioned horn restriction condition may apply when the vehicle ahead has its hazard lights on. For example, if the vehicle ahead activates its hazard lights due to a breakdown, accident, or emergency situation, the driver of the vehicle is likely already aware of the abnormal situation or is in the process of responding. In such a situation, the target vehicle (1) emitting a horn may cause unnecessary stress or interfere with the emergency response, so it may be desirable to block the output of the horn.

[0105] Specifically, the control module (100) can determine whether the horn restriction condition is satisfied by photographing the rear of the vehicle ahead through a camera sensor and detecting the flashing pattern of the hazard light through an image processing algorithm. If the horn restriction condition is satisfied, the control module (100) blocks the horn output command through the output control unit (140) and can record the situation in the log data.

[0106] However, this is not limited to this, and the horn limiting conditions may be further extended according to the embodiments of the present invention. For example, when the target vehicle (1) is driving in a specific area (e.g., near a hospital, residential area, school zone, etc.), a condition to limit the horn output in consideration of noise regulations may be added. As another example, during nighttime hours (e.g., from 22:00 to 06:00), a condition to set a high horn output threshold or suppress the horn output may be applied.

[0107] As another example, a condition can be set to limit horn output if the vehicle ahead is an emergency vehicle, such as a police car, fire truck, or ambulance. These extended horn restriction conditions can be determined using GPS location information, time information, and vehicle type recognition, and can be adjusted according to user settings or traffic regulations.

[0108] Additionally, the control module (100) can record log data including at least one of the output time, situation information, and collision risk used in controlling the horn output after the horn is output through the target vehicle (1).

[0109] Specifically, the control module (100) can generate log data containing various information related to the automatically controlled horn output. At this time, the log data may include the time of horn output (timestamp), location information (GPS coordinates) of the target vehicle (1), speed of the target vehicle (1), type and location of surrounding objects, collision risk value, whether the horn output condition is satisfied, whether the horn restriction condition is satisfied, the duration of the horn output, etc.

[0110] The generated log data can be primarily stored in a storage device (e.g., SSD, flash memory, etc.) inside the target vehicle (1) and can be transmitted to a database (600) via a network (20) for permanent storage. The transmission of log data can be performed in real time or in a batch form when the target vehicle (1) arrives at a specific location (e.g., parking lot, charging station, etc.) and is connected to a Wi-Fi network. Additionally, data security can be ensured by storing and transmitting the log data in an encrypted state.

[0111] In addition, the log data collected in this way can be utilized for various analysis and improvement tasks. For example, by analyzing log data in the event of an accident to identify the collision risk immediately prior to the accident, whether the horn was sounded, and the behavior of surrounding vehicles, the cause of the accident can be determined and liability can be clarified.

[0112] As another example, log data can be statistically analyzed to identify regions, time zones, and situations where horns are frequently honked, and this information can be utilized in establishing traffic safety policies. As yet another example, log data can be used as retraining data for TCN-based models to continuously improve the accuracy of collision risk calculations by reflecting various cases that occur in actual driving environments.

[0113] As such, log data can be utilized in various ways for the overall performance evaluation and verification of autonomous driving systems, enabling manufacturers to derive software updates and improvements based on it.

[0114] Below, with reference to FIG. 4, we will explain the automatic horn output control process of a target vehicle (1) using vehicle-to-object communication (V2X).

[0115] FIG. 4 is a conceptual diagram illustrating an automatic horn output control process using vehicle-to-object communication (V2X) according to one embodiment of the present invention.

[0116] Referring to FIG. 4, according to an automatic horn output control process using Vehicle to Everything (V2X) communication in one embodiment of the present invention, first, a target vehicle (1) can detect its position and measure the distance to a vehicle ahead through a sensor module (200). Next, the control module (100) of the target vehicle (1) can receive intersection signal information (traffic light status, scheduled signal change time, etc.) from a roadside base station (400) via V2I communication.

[0117] Additionally, the control module (100) can receive information regarding the location and direction of movement of a pedestrian via V2P (Vehicle to Pedestrian) communication from a user terminal (500) carried by the pedestrian. The control module (100) can calculate the collision risk by combining sensor data and V2X communication data and perform a horn output determination step. For example, if the traffic light is green, there is no pedestrian, and the vehicle ahead remains stationary, the control module (100) can determine that the second condition is satisfied and output a horn. This V2X communication-based process can significantly improve the accuracy of the horn output determination by utilizing information (signal status, pedestrian intent, etc.) that is difficult to obtain with sensors alone.

[0118] Specifically, according to one embodiment of the present invention, the control module (100) can determine and decide whether to output a horn based on the following Equation 3.

[0119] [Equation 3]

[0120]

[0121] Here, TTC front (t) and TTC back (t) is time point t In this, it refers to the TTC (Time to Collision) with the vehicle in front and the vehicle behind, respectively, and T c is the threshold value of the TTC for collision judgment regarding the vehicle. E(t) is the point in time t It is whether the hazard lights of the vehicle in front are on (1: hazard lights ON, 0: hazard lights OFF), and s(t) is the point in time t It represents the signal status (1: green, 0: red). τ is the accumulated time that the vehicle ahead remained stationary, and T is a preset stop time threshold for triggering a warning. v front (t) is time point t Indicates the speed of the vehicle ahead, and d stop(t) is time point t At, the stopping distance of the own vehicle, d front (t) is time point t It is the actual distance to the vehicle ahead.

[0122] More specifically, the control module (100) can determine whether to output a horn at each point in time using Equation 3. The first condition is the point in time t TTC indicating the TTC with the vehicle ahead at front T where the value of (t) is the threshold c Decreases to below and indicates whether the hazard lights of the vehicle ahead are on. E(t) It can be satisfied when is 0 (hazard lights OFF). The second condition indicates the traffic light status. s(t) The value is 1 (green), and τ, the stopping duration of the vehicle ahead, is the threshold value T Ideal and speed v front (t) remains 0, signifying TTC with the vehicle behind. back (t) is the threshold T c Decrease to below, and the hazard light status of the vehicle ahead ( E(t) It can be satisfied when ) is 0 (hazard lights OFF). The third condition is the stopping distance d of the own vehicle. stop d where (t) is the actual distance to the vehicle ahead front (t) is greater than and the TTC of the vehicle ahead front (t) is the threshold T c Decrease to below, and the hazard light status of the vehicle ahead ( E(t) This condition is met when ) is 0 (hazard lights OFF). The case that limits automatic horn is the hazard lights of the vehicle ahead ( E(t) It may include cases where ) is lit (1) and all other cases.

[0123] In this regard, the control module (100) can evaluate the conditions of Equation 3 in real time and generate a horn output signal when a predetermined condition is satisfied. For example, when the traffic light is green ( s(t) = 1) and the stopping time (τ) of the vehicle ahead is a threshold value ( T, for example, exceeding 3 seconds and the vehicle ahead does not activate its hazard lights ( E(t) = 0) The velocity remains 0 (v front (t) = 0) Estimated Time to Collision with Rear Vehicle (TTC) back (t)) is the threshold (T c If it is smaller than ), the control module (100) can determine that it is a traffic obstruction situation and output a horn. In addition, the control module (100) can perform intelligent horn output control beyond simple threshold comparison by comprehensively considering the collision risk calculated from the TCN-based model along with the condition of Equation 3.

[0124] FIG. 5 is a schematic diagram of an automatic horn output control device of an autonomous vehicle according to one embodiment of the present invention.

[0125] Referring to FIG. 5, the control module (100) may include a data collection unit (110), a calculation execution unit (120), a condition determination unit (130), an output control unit (140), and a log recording unit (150).

[0126] The data collection unit (110) can acquire surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle (1) including an autonomous vehicle.

[0127] For example, the data collection unit (110) can acquire surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle (1).

[0128] The operation execution unit (120) can calculate the risk of collision with objects existing around the target vehicle (1) based on acquired surrounding data.

[0129] For example, the operation execution unit (120) can operate to calculate the collision risk based on a pre-trained artificial intelligence-based analysis model.

[0130] In this regard, according to one embodiment of the present invention, the computation execution unit (120) can calculate the collision risk using an analysis model corresponding to a TCN (Temporal Convolutional Network) based model.

[0131] Specifically, according to one embodiment of the present invention, the calculation unit (120) can calculate the collision risk based on at least one of the Time to Collision (TTC) with the object, the stopping distance of the target vehicle (1), and the margin distance obtained by subtracting the stopping distance from the actual distance with the object.

[0132] The condition determination unit (130) can determine whether the preset horn output condition is satisfied based on the calculated collision risk.

[0133] For example, the condition determination unit (130) can determine whether a horn output condition is satisfied, which includes at least one of a condition in which the distance to the object is reduced to a preset collision risk distance or less, a condition in which the object maintains a stationary state for a preset time or longer, and a condition in which the object enters within the stopping distance of the target vehicle (1).

[0134] The output control unit (140) can control the horn output of the target vehicle (1) based on the judgment result derived by the condition judgment unit (130).

[0135] According to one embodiment of the present invention, the output control unit (140) can control the output of a horn to be selectively performed on a target vehicle (1) based on the result of determining whether a preset horn limit condition is satisfied.

[0136] For example, the output control unit (140) can control the target vehicle (1) not to perform a horn output by considering a horn restriction condition that includes a condition in which an object existing (located) around the target vehicle (1) is of the vehicle type and the vehicle is in an emergency light state.

[0137] The log recorder (150) can record log data including at least one of the output time, situation information, and collision risk used for horn output control after the horn is output through the target vehicle (1).

[0138] Below, based on the details described above, we will briefly examine the operation flow of the present invention.

[0139] FIG. 6 is an operation flowchart of an automatic horn output control method for an autonomous vehicle according to one embodiment of the present invention.

[0140] The automatic horn output control method of the autonomous vehicle illustrated in FIG. 6 can be performed by the control module (100) described above. Therefore, even if the content described below is omitted, the description of the control module (100) can be equally applied to the description of the automatic horn output control method of the autonomous vehicle.

[0141] Referring to FIG. 6, in step S11, the data collection unit (110) can acquire surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle (1) including an autonomous vehicle.

[0142] For example, in step S11, the data collection unit (110) can acquire surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle (1).

[0143] Next, in step S12, the operation performing unit (120) can calculate the risk of collision with objects existing around the target vehicle (1) based on the acquired surrounding data.

[0144] For example, in step S12, the operation execution unit (120) may operate to calculate the collision risk based on a pre-trained artificial intelligence-based analysis model.

[0145] In this regard, according to one embodiment of the present invention, in step S12, the operation performing unit (120) can calculate the collision risk using an analysis model corresponding to a TCN (Temporal Convolutional Network) based model.

[0146] Specifically, according to one embodiment of the present invention, in step S12, the operation performing unit (120) can calculate the collision risk based on at least one of the Time to Collision (TTC) with the object, the stopping distance of the target vehicle (1), and the margin distance obtained by subtracting the stopping distance from the actual distance with the object.

[0147] Next, in step S13, the condition determination unit (130) can determine whether a preset horn output condition is satisfied based on the calculated collision risk.

[0148] For example, in step S13, the condition determination unit (130) can determine whether a horn output condition is satisfied, which includes at least one of the following: a condition in which the distance to the object is reduced to a preset collision risk distance or less, a condition in which the object maintains a stationary state for a preset time or longer, and a condition in which the object enters within the stopping distance of the target vehicle (1).

[0149] Next, in step S14, the output control unit (140) can control the horn output of the target vehicle (1) based on the result of the judgment in step S13.

[0150] Specifically, in step S14, the output control unit (140) can control the horn output to be selectively performed on the target vehicle (1) according to the result of determining whether a preset horn limit condition is satisfied.

[0151] For example, in step S14, the output control unit (140) can control the target vehicle (1) not to perform a horn output by considering a horn restriction condition that includes a condition in which an object existing (located) around the target vehicle (1) is of the vehicle type and the vehicle is in an emergency light state.

[0152] Next, in step S15, the log recorder (150) can record log data including at least one of the output time, situation information, and collision risk used for horn output control after the horn is output through the target vehicle (1).

[0153] In the description above, steps S11 through S15 may be further divided into additional steps or combined into fewer steps according to an embodiment of the present invention. Additionally, some steps may be omitted as necessary, and the order between steps may be changed.

[0154] A method for controlling the automatic horn output of an autonomous vehicle according to one embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either individually or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa.

[0155] In addition, the aforementioned method for controlling the automatic horn output of an autonomous vehicle can also be implemented in the form of a computer program or application executed by a computer stored on a recording medium.

[0156] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will understand that other specific forms can be easily modified without altering the technical concept or essential features of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.

[0157] The scope of the present invention is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the meaning and scope of the claims and the concept of equivalents thereof should be interpreted as being included within the scope of the present invention. Explanation of the symbols

[0158] 10: Control system of an autonomous vehicle equipped with an automatic horn output function 100: Automatic horn output control device for autonomous vehicles, control module 110: Data Collection Unit 120: Operation execution unit 130: Conditional judgment unit 140: Output control unit 150: Logbook 200: Sensor module 300: Horn Module 400: Roadside Unit (RSU) 500: User terminal 600: Database 20: Network 1: Target vehicle

Claims

Claim 1 A method for controlling the automatic horn output of an autonomous vehicle comprises: (a) acquiring surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle including the autonomous vehicle; (b) calculating a collision risk with said object based on said surrounding data through a pre-trained artificial intelligence-based analysis model; (c) determining whether a preset horn output condition is satisfied based on said collision risk; and (d) controlling the horn output of said target vehicle based on said determination result, wherein step (b) calculates the collision risk by utilizing a Temporal Convolutional Network (TCN)-based model as the analysis model based on at least one of the Time to Collision (TTC) with said object, the stopping distance of said target vehicle, and a margin distance obtained by subtracting the stopping distance from the actual distance with said object; and the horn output condition comprises: a first condition in which the distance with said object decreases to less than or equal to a preset collision risk distance; and a second condition in which said object maintains a stopped state for more than a preset time. A method comprising at least one of a third condition in which the object enters within the stopping distance of the target vehicle, wherein the satisfaction of the first condition is determined by detecting a time series pattern in which the TTC value decreases using the TCN-based model, and the satisfaction of the third condition is determined by detecting a time series pattern in which the margin distance decreases using the TCN-based model. Claim 2 A method according to claim 1, wherein step (d) selectively performs the horn output according to the result of a judgment on whether a preset horn limit condition is satisfied. Claim 3 In paragraph 2, the above horn restriction condition includes a condition in which the object is a vehicle and the vehicle is in a state where the hazard lights are on. Claim 4 delete Claim 5 delete Claim 6 A method according to claim 1, wherein step (a) is to acquire surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle. Claim 7 delete Claim 8 The method of claim 1, further comprising: (e) recording log data including at least one of the output time, situation information and collision risk level after the horn is output through the target vehicle. Claim 9 An automatic horn output control device for an autonomous vehicle comprises: a data collection unit for acquiring surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle including the autonomous vehicle; a computation execution unit for calculating a collision risk with said object based on said surrounding data through a pre-trained artificial intelligence-based analysis model; a condition determination unit for determining whether a preset horn output condition is satisfied based on said collision risk; and an output control unit for controlling the horn output of said target vehicle based on said determination result, wherein the computation execution unit calculates the collision risk by utilizing a Temporal Convolutional Network (TCN)-based model as the analysis model based on at least one of the Time to Collision (TTC) with said object, the stopping distance of said target vehicle, and a margin distance obtained by subtracting the stopping distance from the actual distance with said object, and said horn output condition comprises: a first condition in which the distance with said object decreases to less than or equal to a preset collision risk distance; and a second condition in which said object maintains a stopped state for more than a preset time. A device comprising at least one of a third condition in which the object enters within the stopping distance of the target vehicle, wherein the satisfaction of the first condition is determined by detecting a time series pattern in which the TTC value decreases using the TCN-based model, and the satisfaction of the third condition is determined by detecting a time series pattern in which the margin distance decreases using the TCN-based model. Claim 10 In claim 9, the device wherein the output control unit selectively performs the horn output according to the result of a judgment on whether a preset horn limit condition is satisfied. Claim 11 In Clause 10, the above horn restriction condition includes a condition in which the object is a vehicle and the vehicle is in a state where the hazard lights are on. Claim 12 delete Claim 13 delete Claim 14 In claim 9, the device wherein the data collection unit acquires the surrounding data using at least one of an ultrasonic sensor, a radar sensor, and a camera sensor mounted on the target vehicle. Claim 15 A control system for an autonomous vehicle equipped with an automatic horn output function comprises: a sensor module for collecting surrounding data including at least one of location information, distance information, and speed information of an object located around a target vehicle corresponding to the autonomous vehicle; a control module for calculating a collision risk with said object based on said surrounding data, determining whether a preset horn output condition is satisfied based on said collision risk, and generating a control signal for controlling the horn output of said target vehicle based on said determination result; and a horn module configured to output a horn signal based on said control signal, wherein the control module calculates the collision risk by utilizing a Temporal Convolutional Network (TCN)-based model as the analysis model based on at least one of the Time to Collision (TTC) with said object, the stopping distance of said target vehicle, and a margin distance obtained by subtracting the stopping distance from the actual distance with said object, and the horn output condition comprises: a first condition in which the distance with said object decreases to less than or equal to a preset collision risk distance; and a second condition in which said object maintains a stopped state for more than a preset time. A control system comprising at least one of a third condition in which the object enters within the stopping distance of the target vehicle, wherein the satisfaction of the first condition is determined by detecting a time series pattern in which the TTC value decreases using the TCN-based model, and the satisfaction of the third condition is determined by detecting a time series pattern in which the margin distance decreases using the TCN-based model.

Images