An electromagnetic denial system for an autonomous driving platform

Abstract

The present invention relates to an electromagnetic rejection system for an autonomous driving platform, which includes a target recognition subsystem, a rejection decision-making subsystem, and a target rejection subsystem; the target recognition subsystem is used to identify the azimuth data of the target platform and transmit the azimuth data of the target platform to the rejection decision-making subsystem; the rejection decision-making subsystem automatically or manually makes an electromagnetic rejection decision based on the target angle and the straight-line distance. If it is determined that there is a risk for the target platform, a rejection signal is output to the target rejection subsystem; the target rejection subsystem realizes real-time tracking of the target platform within a set error range, and after receiving the rejection command, applies a GPS electromagnetic rejection signal to the target platform. The present invention can achieve the purpose of GPS electromagnetic rejection of the target autonomous driving platform within a set distance range, and supports both manual and automatic rejection methods, which is safer and more efficient when dealing with sudden situations of out-of-control autonomous driving platforms.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving, and in particular to an electromagnetic rejection system for an autonomous driving platform. Background Art

[0002] While the widespread use of unmanned platforms brings convenience to people, it also brings some potential risks. Taking unmanned driving platforms as an example, after unmanned technology is enabled in traditional cars, it will inevitably increase the physical complexity of the original car system. At the same time, when the integration of electronic systems reaches a certain level, the electromagnetic compatibility issues between various electronic devices will become more important. When the electromagnetic compatibility issues of unmanned driving systems are not considered well, the entire system may have functional abnormalities during operation. The conceivable situations include but are not limited to abnormal drone performance effects and unmanned car loss of control. Among them, the risk of unmanned car loss of control in densely populated areas is the most serious.

[0003] At present, there are three main means of countering unmanned platforms: acoustic interference, physical interference, and electromagnetic interference. The acoustic countermeasure method emits sound waves of special frequencies to cause the key components of the target gyroscope to resonate and output erroneous information, thereby causing the target function to fail, thereby achieving the purpose of countermeasure; the physical countermeasure method causes the structure of the target unmanned platform to fail through collision, scratching, and burning through the air, thereby achieving the purpose of countermeasure; the electromagnetic interference method radiates interference signals to the target unmanned platform in a direction to cause the navigation or communication system of the target unmanned platform to malfunction, thereby achieving the purpose of countermeasure. Among them, the countermeasure based on physical methods is the most widely used.

[0004] However, whether it is an unmanned platform countermeasure device based on physical methods, it often requires manual operation when applied and is mainly aimed at unmanned aerial vehicle systems. This poses a serious risk of casualties when dealing with large-target unmanned platforms such as driverless cars. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an electromagnetic rejection system for an autonomous driving platform to solve the deficiencies of the prior art.

[0006] The object of the present invention is achieved by the following technical solutions: an electromagnetic denial system for an autonomous driving platform, the system comprising a target recognition subsystem, a denial decision subsystem and a target denial subsystem;

[0007] The target identification subsystem is used to identify the position data of the target platform and transmit the position data of the target platform to the denial decision subsystem;

[0008] The denial decision subsystem automatically or manually makes an electromagnetic denial decision based on the target angle and radial distance. If it is determined that there is a risk to the target platform, a denial command is output to the target denial subsystem.

[0009] The target denial subsystem realizes real-time tracking of the target platform within a set error range, and after receiving the denial command sent by the denial decision subsystem, it applies a GPS electromagnetic denial signal to the target platform, and controls the output denial signal power in real time based on the distance information of the target platform extracted by the target recognition system.

[0010] The target recognition subsystem includes a camera and a millimeter-wave radar; the camera and the millimeter-wave radar are used to obtain in real time the distance information of the target platform within a set range directly in front of the own platform, and to obtain in real time the angle information of the target platform within a set angle range directly in front of the own platform; the target platform angle data is preliminarily extracted through a network model, and the angle and distance information output by the millimeter-wave radar are integrated to achieve accurate recognition of the target angle and distance.

[0011] The method of preliminarily extracting the angle data of the target platform through the network model and integrating the angle and distance information output by the millimeter wave radar to realize accurate recognition of the angle and distance of the target includes:

[0012] The millimeter-wave radar detects the angle, distance and speed information of all objects. After radar filtering, the angle information is compared with the azimuth of the target relative to the own platform identified by the YOLOv5 network optimization model and the angles of all objects detected by the millimeter-wave radar, and objects with angle differences less than the set range are screened out;

[0013] Then, according to the rate of change of the target angle identified by the YOLOv5 network optimization model, the approximate normal velocity of the target is calculated, the velocity information of the objects with angle differences less than the set range is compared with the approximate normal velocity of the target, and the objects with radial velocity differences less than the set range are screened out to obtain the distance and angle of the target wheel.

[0014] The denial decision process of the denial decision subsystem includes:

[0015] The denial decision subsystem receives the target position information output by the target recognition subsystem and the position information of the own platform, calculates the relative position parameter ΔP and compares it with the danger threshold P d Make comparisons;

[0016] If ΔP>P d Then it is determined that there is no danger and whether there is a human rejection command input. If so, the rejection command is output and output; if not, the decision is terminated;

[0017] If, ΔP≤P dIf the target is judged to be approaching dangerously, it is judged that there is a dangerous situation and a rejection command is generated and output.

[0018] The software interface of the denial decision subsystem includes a parameter setting module, a target identification and status detection module and a denial effect module;

[0019] The parameter setting module includes relevant parameter settings for target identification and denial systems, as well as controls for realizing the functions of inputting and visualizing human denial commands and starting the test process;

[0020] The target recognition and status detection module includes visualization of target platform angle, distance and speed information and related information of own positioning device;

[0021] The denial effect module includes the target platform's current orientation, longitude and latitude, number of satellites received and other information generated based on GPS navigation information, as well as visualization of the target platform's regional spectrum information.

[0022] The target denial subsystem includes a radiation device and a directional device; the radiation device is composed of an antenna, a radio frequency signal source and a power amplifier; the radiation device obtains the sensitivity threshold of the target GPS system under different distance conditions through testing, and obtains the GPS sensitivity threshold of the target platform at any distance through fitting, thereby determining the radiation power.

[0023] The orientation device consists of a rotating device and an antenna fixing device, which is used to locate and track the target platform. The process of the target orientation device being oriented toward the target platform is divided into tracking and pursuit modes according to the angle difference between the current turntable angle and the target angle through an adaptive incremental turntable control algorithm, and the two modes are distinguished by the size relationship between Δθ and the decision threshold t.

[0024] The two modes are distinguished by the relationship between Δθ and the decision threshold t, including:

[0025] When Δθ≥t, the turntable enters the pursuit mode. If it is the same direction pursuit, the new angular velocity w is output. z2 Take the maximum value w max , then the output angle θ z2 The increment function f depends on Δθ and Δw p3 , if it is a counter-attack, then it depends on the increment function f of the position Δθ and Δw p2 ;

[0026] When Δθ<t, the turntable enters the following mode. If it is opposite-direction following, then w z2 Take the maximum value, θ z2 Remain unchanged. If it is following in the same direction, then w z2 and θ z2 Depends on f related to Δw w1and f p1 Increment function.

[0027] The present invention has the following advantages: an electromagnetic denial system for an autonomous driving platform, which can remotely achieve the purpose of suppressing the loss of control of a target unmanned driving platform, and supports both manual and automatic intervention methods (manual selection of the target platform and automatic identification of the target platform), which is safer and more efficient when dealing with the above-mentioned emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a functional structure schematic diagram of the present invention;

[0029] Figure 2 Schematic diagram of the structure of the detection head in the optimization network model;

[0030] Figure 3 Flowchart for identifying target distance and angle for network optimization model;

[0031] Figure 4 A logical flow chart for the denial decision;

[0032] Figure 5 This is the flow chart of the turntable control algorithm based on multiple increments. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application provided below in conjunction with the drawings is not intended to limit the scope of protection of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application. The present invention is further described below in conjunction with the drawings.

[0034] The present invention is installed on an unmanned platform and can automatically or manually identify the target autonomous driving platform (hereinafter referred to as the target platform) and perform effective electromagnetic denial (electromagnetic denial denial of the target GPS system) when the target autonomous driving platform (hereinafter referred to as the target platform) suddenly loses control, causing the target autonomous driving platform to lose its navigation signal and then slow down, change the path, etc., so as to directly solve the problem or reduce the further deterioration of the situation.

[0035] like Figure 1As shown, the present invention includes three subsystems: target identification, denial decision and target denial, which respectively realize the functions of discovering the target platform (autopilot platform identification), electromagnetic denial command decision and effective electromagnetic denial for the designated target platform; wherein, after the target identification subsystem identifies the orientation of the target platform, it transmits the orientation data to the denial decision subsystem. The denial decision subsystem automatically or manually makes an electromagnetic denial decision based on the target angle and radial distance information. If it is determined that the target platform is at risk, a denial signal is output to the target denial subsystem. After receiving the denial command, the target denial subsystem performs directional GPS electromagnetic denial on the target platform.

[0036] The present invention is mainly designed for autonomous driving platforms, so the target unmanned platform selects the D-KIT development platform equipped with the apollo5.5 autonomous driving system as the target platform. The platform is equipped with a GPS signal receiver, which can realize the autonomous driving function by processing the GPS signals before and after the vehicle in the GPS navigation mode. In addition, the apollo system supports the positioning module to provide two GPS-related services, namely: positioning through GPS+IMU and positioning through multi-sensor fusion.

[0037] Our platform is the FR-09 general mobile platform from YUHESEN. FR-09 supports standardized word protocols and provides interfaces. It can use the standard CAN bus as a communication interface to connect to modules such as pipeline systems, laser radars, and differential GPS, and supports secondary development of autonomous driving systems such as Baidu Apollo and ROS.

[0038] Furthermore, the target recognition subsystem is based on cameras and millimeter-wave radars to obtain real-time distance information of target platforms within 20 meters directly in front of its own platform, and clearly present the angle information of target platforms within a range of at least ±30° directly in front of its own platform, and the data refresh rate must not be lower than 10fps.

[0039] Among them, the 77GHz ARS408-21 long-range millimeter-wave radar is used to identify the distance parameters of the target platform, and the distance information of the target platform is extracted through the ARS408-21 close-range beam and ±45° angle range scanning mode to meet the requirements of angle and distance accuracy and meet the design requirements of the target identification subsystem. The ARS408-21 millimeter-wave radar follows the CAN communication protocol, and its data reading depends on the CAN analyzer. The radar is connected through the official CAN analyzer for connection testing. After the test is completed, the radar data filtering and visualization are completed based on the official encapsulation function.

[0040] Among them, the multi-sensor fusion method is used to increase the system recognition accuracy. The AR023ZWDR model camera is combined with the YOLOV5s model to preliminarily extract the target angle data, and the millimeter-wave radar output angle information is integrated to achieve more accurate and robust recognition of the target angle.

[0041] The present invention reads camera information based on the YOLOv5s model. YOLOv5 is implemented based on the PyTorch framework. The YOLOv5s structural model consists of four parts: an input end, a backbone network, a neck, and a detection head.

[0042] like Figure 2 As shown, in order to improve the accuracy of the model in identifying the target vehicle, the present invention optimizes the above-mentioned YOLOv5 network model, that is, a CBAM module is added in front of the detection head corresponding to the medium anchor frame in the YOLOv5s network model to improve the accuracy.

[0043] like Figure 3 As shown in the figure, in order to improve the target recognition accuracy, the millimeter-wave radar detects the angle, distance and speed information of all objects. After radar filtering, the azimuth angle of the target relative to the own platform identified by the YOLOv5 network optimization model is compared with the angle of all objects detected by the millimeter-wave radar one by one, and the objects with angle difference less than the set range are screened out; then, according to the change rate of the target angle identified by the YOLOv5 network optimization model, the approximate normal velocity of the target is calculated, and the speed information of the objects with angle difference less than the set range is compared with the approximate normal velocity of the target, and the objects with radial velocity difference less than the set range are screened out to obtain the distance and angle of the target wheel.

[0044] Furthermore, the function of the denial decision subsystem is to process the data of the target identification and target denial subsystems, and control the two to achieve directional electromagnetic denial, and mainly plays the role of electromagnetic denial decision in the electromagnetic denial system.

[0045] like Figure 4 As shown in the figure, the generation of the denial command in the denial decision subsystem depends not only on the target platform position information, but also on the own platform movement position information and the manual denial command of the own decision maker. The specific process of a single decision of the system is as follows:

[0046] The denial decision subsystem receives the target position information output by the target recognition subsystem and the position information of the own platform, calculates the relative position parameter ΔP and compares it with the danger threshold P d Compare. If, ΔP>P d Then it is determined that there is no danger and whether there is a human rejection command input. If there is, a rejection command is generated and output; if not, the decision is terminated. If ΔP≤P dIf the target is judged to be approaching dangerously, it is judged that there is a dangerous situation and a rejection command is generated and output.

[0047] Denial logic visualization: The Python-based Pyqt5 package designs the software interface based on the concepts of signals and slots, and the decision logic visualization design is performed based on the Pyqt5 package; when designing the logic visualization interface, the control-logic code is separated. In the control code, the human intervention parameters required for the denial decision and the interconnection between the target recognition and denial systems are integrated into the corresponding controls and the modules are divided according to the control functions; in the logic design code, the logic between each module is compiled through the signal-slot mechanism.

[0048] The overall software interface consists of a parameter setting module, a target identification and status detection module, and a denial effect module. The parameter setting module mainly includes the relevant parameter settings for target identification and denial systems, as well as controls for realizing functions such as input and visualization of human denial commands and starting the test process; the target identification and status detection module mainly includes the visualization of information such as the angle, distance, speed, etc. of the target platform and related information of the own directional device; the target denial effect module mainly includes the visualization of the disturbance status of the target platform, including the current orientation and longitude and latitude information generated by the target platform based on GPS navigation information, and the spectrum detected at the GPS antenna. In addition, the functional visualization logic implementation process of the software and the corresponding subsystems will be reflected in the construction of the target denial subsystem and the testing of the electromagnetic denial system.

[0049] Furthermore, the target denial subsystem now tracks the target platform in real time, and after receiving the electromagnetic denial command from its own platform, it applies the GPS electromagnetic denial signal to the target platform, and can control the output of the denial signal power in real time according to the distance of the target platform. At the same time, the system tracking accuracy requires the error to be less than or equal to 1°, the radiation power change frequency is required to be less than 10ms, and it can ensure effective electromagnetic denial within 15m of its own platform. According to its system function, the target denial subsystem can be further divided into radiation devices and directional devices and designed separately.

[0050] The radiation device is mainly composed of antenna, RF signal source and power amplifier. In addition, the sensitivity threshold of the target GPS system under different distance conditions is obtained through testing, and the GPS sensitivity threshold of the target platform at any distance is obtained through fitting, so as to determine the radiation power.

[0051] Furthermore, the antenna selection is a double-ridged horn antenna that covers the GPS frequency band, has strong directivity, high gain, good stability, and richer prior knowledge.

[0052] The RF signal source is the DSG3136B produced by RIGOL. The DSG3136B supports the output of RF signals with a frequency of 0.9 to 13.6 GHz, including the GPS frequency band of 1.57542 GHz. In addition, the minimum output level of the DSG3136B is -110 dBm, the maximum output level is +20 dBm, and its level setting time is 5 ms, which meets the system's radiated power frequency requirements.

[0053] The power amplifier uses the Mini-Circuits ZHL-20W013 power amplifier with a gain of about 40dB, a maximum output power of 20W, and a power consumption of about 65W to meet the working distance requirements of the denial system.

[0054] Furthermore, the orientation device is composed of a rotating device and an antenna fixing device, and is used to locate and track the target platform.

[0055] Among them, the rotating device adopts the K103DX200J electric turntable system as the rotating device. Among them, the PC end is used as a control input device to continuously output control signals, the MT-22E control board is used as a motion controller, and the K103DX200J turntable is used to integrate the functions of the driver, motor and mechanical device. The K103DX200J electric turntable system has multiple working modes, including: positioning mode and constant speed mode. In addition, the system supports two control modes: open-loop and closed-loop control. After testing, when the system uses the most common open-loop positioning mode, under different acceleration conditions (deceleration and acceleration are the same), the MT-22E open-loop control response time is close to the theoretical response time, and the error between most theoretical and actual results is within 10 -2 The order of magnitude is close to that of the azimuth, which meets the design index of the directional device of the present invention. Therefore, the present invention adopts an open-loop positioning mode and designs a corresponding algorithm to achieve real-time and accurate positioning.

[0056] To solve the problem of positioning lag in the positioning mode of the turntable system, the present invention designs a turntable control algorithm based on adaptive increment, which divides the turntable positioning scenarios into tracking mode and locking mode. The tracking is divided into same-direction tracking and reverse tracking modes.

[0057] like Figure 5 As shown in the figure, the algorithm rotates the target orientation device (turntable) toward the target platform according to the angle difference Δθ between the current turntable angle and the target angle (the current turntable orientation angle θ z and the azimuth angle θ of the target platform relative to the own platform tar The absolute value of the previous difference) is divided into tracking and pursuit mode, and the two modes are distinguished by the size relationship between Δθ and the decision threshold t.

[0058] When Δθ≥t, the turntable enters the pursuit mode. If it is the same direction pursuit (the current speed of the turntable w z and target speed w tar The new angular velocity w is the same as z2 Take the maximum value w max , output angle θ z2 The increment function f depends on Δθ and Δw p3 ; If it is a counter-attack, it depends on the incremental function f of the position Δθ and Δw p2 .

[0059] When Δθ<t, the turntable enters the following mode. If it is opposite-direction following, then w z2 Take the maximum value, θ z2 remains unchanged; if it is following in the same direction, then w z2 and θ z2 Depends on f related to Δw w1 and f p1 Increment function.

[0060] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used for various other combinations, modifications and improvements, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art do not deviate from the spirit and scope of the present invention, and should be within the scope of protection of the claims attached to the present invention.

Claims

1. An electromagnetic denial system for an autonomous driving platform, characterized in that: The system includes a target identification subsystem, a denial decision subsystem and a target denial subsystem; The target identification subsystem is used to identify the position data of the target platform and transmit the position data of the target platform to the denial decision subsystem; The denial decision subsystem automatically or manually makes an electromagnetic denial decision based on the target angle and radial distance, and visualizes the denial logic through a software interface. If it is determined that the target platform is at risk, a denial command is output to the target denial subsystem; The target denial subsystem realizes real-time tracking of the target platform within a set error range, and after receiving the denial command, applies a GPS electromagnetic denial signal to the target platform, and controls the output denial signal power in real time according to the distance information of the target platform extracted by the target recognition system; The target denial subsystem includes a radiation device and a directional device; the radiation device is composed of an antenna, a radio frequency signal source and a power amplifier; the radiation device obtains the target GPS system sensitivity threshold under different distance conditions through testing, and obtains the GPS sensitivity threshold of the target platform at any distance through fitting, thereby determining the radiation power; The orientation device is composed of a rotating device and an antenna fixing device, and is used to locate and track the target platform; The process of the target orientation device turning toward the target platform is divided into tracking and pursuit modes according to the angle difference between the current turntable angle and the target angle through the adaptive incremental turntable control algorithm. Decision Threshold The size relationship between them is used to distinguish the two modes. is the angle difference between the current turntable angle and the target angle; Said through Decision Threshold The size relationship between them is used to distinguish the two modes, including: when When the turntable enters the pursuit mode, if it is the same direction pursuit, the new angular velocity output Take the maximum value , then the output angle depending on and The increment function , if it is an opposite pursuit, it depends on the position and The increment function ; when , the turntable enters the follow mode. If it is opposite direction follow, then Take the maximum value, Remain unchanged. If it is following in the same direction, and Depends on Related and Increment function.

2. The electromagnetic denial system for an autonomous driving platform according to claim 1, characterized in that: The target recognition subsystem includes a camera and a millimeter-wave radar; the camera and the millimeter-wave radar are used to obtain in real time the distance information of the target platform within a set range directly in front of the own platform, and to obtain in real time the angle information of the target platform within a set angle range directly in front of the own platform; the target platform angle data is preliminarily extracted through a network model, and the angle and distance information output by the millimeter-wave radar are integrated to achieve accurate recognition of the target angle and distance.

3. The electromagnetic denial system for an autonomous driving platform according to claim 2, characterized in that: The method of preliminarily extracting the angle data of the target platform through the network model and integrating the angle and distance information output by the millimeter wave radar to realize accurate recognition of the angle and distance of the target includes: The millimeter-wave radar detects the angle, distance and speed information of all objects. After radar filtering, the angle information is compared with the azimuth of the target relative to the own platform identified by the YOLOv5 network optimization model, and objects with angle differences less than the set range are screened out; Then, according to the rate of change of the target angle identified by the YOLOv5 network optimization model, the approximate normal velocity of the target is calculated, the velocity information of the objects with angle differences less than the set range is compared with the approximate normal velocity of the target, and the objects with radial velocity differences less than the set range are screened out to obtain the distance and angle of the target platform.

4. The electromagnetic denial system for an autonomous driving platform according to claim 1, characterized in that: The denial decision process of the denial decision subsystem includes: The denial decision subsystem receives the target position information output by the target recognition subsystem and the position information of the own platform, and calculates the relative position parameters and with the danger threshold Make comparisons; if, Then it is determined that there is no danger and whether there is a human rejection command input. If so, a rejection command is output; if not, the decision is terminated; if, If the target is judged to be approaching dangerously, it is judged that there is a dangerous situation and a rejection command is generated and output.

5. The electromagnetic denial system for an autonomous driving platform according to claim 1, characterized in that: The software interface of the denial decision subsystem includes a parameter setting module, a target identification and status detection module and a denial effect module; The parameter setting module includes relevant parameter settings for target identification and denial systems, as well as controls for realizing the functions of inputting and visualizing human denial commands and starting the test process; The target recognition and status detection module includes visualization of target platform angle, distance and speed information and related information of own positioning device; The denial effect module includes the visualization of the target platform's current orientation, longitude and latitude, number of satellites received, and spectrum information of the target platform area generated based on GPS navigation information.

Images