Intelligent trolley device with prospective marker

Abstract

The invention discloses an intelligent car device with prospective markers, and aims to solve the problems of insufficient prospective performance, lagging track judgment, poor hardware stability, low control precision and the like caused by height limitation of a camera of a traditional car in national college student intelligent car competition. And advanced recognition of subsequent types of racing tracks is realized. The device is composed of a trolley body, a control unit, a sensing module, a driving module and a power management module, the sensing module collects marker images, the control unit combines data of an encoder and an attitude sensor and generates cooperative control signals through a cascade PID algorithm, and the driving module adjusts the speed and steering of the trolley in real time. Through a three-stage voltage reduction circuit, a reinforced mechanical structure and an anti-interference image processing algorithm, the stability and the turning smoothness under high-speed driving are remarkably improved, and the method can be expanded to application scenes such as autonomous navigation robots.

Description

Technical Field

[0001] This invention relates to the fields of intelligent control, robotics, and automotive electronics, specifically to an intelligent vehicle device with forward-looking identification features. Background Technology

[0002] The National Undergraduate Intelligent Vehicle Competition, an engineering practice competition advocated by the Ministry of Education, focuses on intelligent vehicles and encompasses knowledge from multiple disciplines, including automatic control, pattern recognition, sensing technology, electrical engineering, computer science, mechanics, and automotive engineering. It aims to cultivate undergraduates' practical innovation abilities and teamwork spirit. The competition requires participating vehicles to autonomously follow a designated track at high speed. The track typically includes straightaways, curves, loops, obstacles, intersections, and other elements, placing extremely high demands on the vehicle's environmental perception, real-time decision-making, and motion control capabilities.

[0003] However, existing intelligent vehicles have many technical shortcomings in practical applications, making it difficult to balance high speed and stability: 1. Insufficient foresight and delayed track identification: Due to the limitations imposed by competition rules on camera height (such as the height restriction for high-speed photoelectric groups), the field of view of cameras used in traditional cars is limited, and there is a wide-angle distortion problem. When the vehicle speed reaches 4m / s or above, the images captured by the camera can only reflect the current and nearby track information, and cannot identify subsequent track types (such as sudden curves, loops, obstacles) in advance. This leads to untimely decision-making and response, forcing the vehicle speed to be limited, and even causing the risk of running off the track.

[0004] 2. Shortcomings in hardware design: In terms of power management, existing solutions mostly use simple step-down circuits, resulting in poor output voltage stability. This makes it difficult to meet the power supply requirements of different modules such as control units, sensors, and motors, and voltage fluctuations can easily lead to insufficient motor power and distorted sensor data. In terms of motor drive, some solutions use driver chips with low rated current and poor commutation accuracy, which cannot provide sufficient and stable power to the vehicle, especially in high-speed cornering and hill climbing scenarios. In terms of sensor installation, the mounting structure of components such as cameras and encoders lacks reinforcement design, and vibrations during vehicle operation can easily cause component displacement, affecting data acquisition accuracy.

[0005] 3. Insufficient algorithm optimization and low control precision: In terms of image processing, traditional solutions mostly use a single line scanning method and fixed threshold binarization processing, which has weak anti-interference ability and is easily affected by changes in lighting and track noise, resulting in inaccurate extraction of the track center line; In terms of control strategy, most cars use a single PID control without differentiated design for speed and direction, which easily leads to problems such as lag in servo response and untimely speed adjustment when cornering, affecting the smoothness of cornering and overall driving stability.

[0006] 4. Inadequate mechanical structure design: The brushless motor support structure of some cars is poorly designed, resulting in low efficiency of negative pressure transmission generated by the fan, which limits the speed limit of the car; problems such as unbalanced body center of gravity and insufficient component installation precision also further affect the high-speed operation stability of the car.

[0007] In response to the shortcomings of the existing technologies, there is an urgent need for an intelligent vehicle device that can improve forward-looking perception capabilities and optimize the collaborative performance of software and hardware, so as to solve the core problems of lagging track identification and poor operational stability when traditional vehicles are traveling at high speeds. Summary of the Invention

[0008] In view of the shortcomings of the prior art, the purpose of this invention is to provide an intelligent vehicle device with forward-looking identification to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, a specific embodiment of the present invention provides an intelligent vehicle device with forward-looking markers, including a vehicle body, a control unit, a sensing module, a drive module, and a power management module; the sensing module, drive module, and power management module are all electrically connected to the control unit; the sensing module is used to collect image information of forward-looking markers in the forward-looking area of ​​the track, and the forward-looking markers are used to characterize the subsequent type of the track; the control unit is configured to identify the forward-looking markers based on the image information and generate control signals according to the identification results; the drive module is drivenly connected to the walking components of the vehicle body and is used to adjust the vehicle's driving state in response to the control signals.

[0010] An intelligent vehicle device with forward-looking markers according to an embodiment of this application achieves early recognition of track types and high-speed stable tracking of the vehicle through an original forward-looking marker design, optimized hardware selection, reasonable mechanical structure and efficient control algorithm.

[0011] In addition, the intelligent vehicle device with forward-looking identification proposed in this application may also have the following additional technical features: In one embodiment of this application, the forward-looking markers include multiple types, each corresponding to a subsequent type of track, such as a straight road, a left curve, a right curve, a left loop, a right loop, a left obstacle, a right obstacle, and a straight road at an intersection.

[0012] In one embodiment of this application, the various types of forward-looking markers are distinguished by preset differentiated visual features, including preset shapes, colors, and patterns, for accurate identification by the perception module.

[0013] In one embodiment of this application, the sensing module includes a color camera, a 1024-line high-precision encoder, and a six-axis attitude sensor; the color camera is installed at the front of the vehicle body and is used to collect forward-looking landmarks and track images; there are two encoders, which are respectively connected to the drive wheels of the vehicle body to detect the wheel speed; the attitude sensor is used to collect the angular velocity and acceleration information of the vehicle, and its signal output terminal is electrically connected to the control unit.

[0014] In one embodiment of this application, the color camera is mounted on a carbon fiber rod with a vertical reinforcing rod at the bottom. The camera is fixed at a fixed angle and has a height of 20cm. Its image output end is bound to the data interface of the control unit.

[0015] In one embodiment of this application, the control unit receives the rotation speed signal from the encoder and the attitude signal from the attitude sensor, and combines the recognition result of the forward-looking marker to generate a speed control signal and a direction control signal, which together regulate the drive module.

[0016] In one embodiment of this application, the drive module includes a sensorless brushless motor drive circuit, a brushed motor drive circuit, and a vertical servo motor. The vertical servo motor is used to drive the front wheels to steer in response to a direction control signal. The brushless motor drive circuit achieves commutation through back EMF detection and is electrically connected to the brushless motor of the vehicle body. The brushed motor drive circuit adopts an H-bridge topology, has a rated operating current of not less than 100A, and is electrically connected to the drive motor of the vehicle body.

[0017] In one embodiment of this application, the power management module includes a lithium polymer battery and a three-stage buck circuit; the three-stage buck circuit outputs 5V, 3.3V and 6V voltages respectively, which power the control unit, the sensing module and the servo motor respectively, and its input terminal is electrically connected to the lithium polymer battery.

[0018] In one embodiment of this application, the vehicle body is provided with a brushless motor support structure, which is a 3D-printed I-shaped column that is vertically fixed between the vehicle chassis and the brushless motor for transmitting negative pressure.

[0019] The control method for the intelligent vehicle device includes the following steps: S1: System initialization, completing the startup configuration of the sensing module, drive module and control unit; S2: The perception module collects image information of the track's forward-looking area and transmits it to the control unit; S3: The control unit identifies the type of forward-looking markers based on image information and determines the subsequent type of track. S4: The control unit combines the wheel speed signal and the vehicle attitude signal to generate speed control signal and direction control signal; S5: The drive module responds to control signals and adjusts the vehicle's speed and steering angle; S6: Trigger a stop when the departure line feature is detected and the vehicle speed is greater than 1m / s; otherwise, repeat S2-S5.

[0020] The advantages of this invention compared to existing technologies are: (1) By designing eight different and forward-looking markers, the track type can be identified in advance, which solves the core pain point of decision lag when driving at high speed due to the limited field of view of the camera in traditional intelligent cars, and effectively breaks through the bottleneck of speed improvement.

[0021] (2) In terms of hardware, the three-stage step-down power supply module provides stable power supply for each functional module, the high rated current drive chip ensures sufficient power output, and the sensor reinforcement installation structure reduces data acquisition errors during the operation of the vehicle; in terms of software, algorithms such as double line scanning and median filtering optimize the image processing accuracy, and the cascade PID control strategy realizes precise coordination of speed and direction, significantly improving the smoothness of cornering and the stability of continuous driving.

[0022] (3) A USB high frame rate color camera is used, which can provide richer track information compared with traditional grayscale cameras. Combined with anti-interference image processing algorithm, it can adapt to the lighting changes in different competition environments. The reasonable vehicle body structure design enhances the anti-vibration capability and adapts to the complex scene of the competition track.

[0023] (4) The overall design is fully compatible with the requirements of the Loongson track in the high-speed optoelectronic group of the National Undergraduate Intelligent Vehicle Competition. After actual competition testing, the operation effect is reliable. At the same time, the technical solution can be extended to various autonomous tracking scenarios such as autonomous navigation robots and intelligent inspection vehicles, and has a wide range of application prospects.

[0024] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A three-dimensional representation of an intelligent vehicle device with forward-looking identifiers in one embodiment of the present invention. Figure 1 ; Figure 2A three-dimensional representation of an intelligent vehicle device with forward-looking identifiers in one embodiment of the present invention. Figure 2 ; Figure 3 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 1 ; Figure 4 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 2 ; Figure 5 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 3 ; Figure 6 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 4 ; Figure 7 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 5 ; Figure 8 This is the circuit board principle of a smart car device with forward-looking identification in one embodiment of the present invention. Figure 6 ; Figure 9 This is a software flowchart of an intelligent vehicle device with a forward-looking identifier in one embodiment of the present invention; Figure 10 This is a system overview and energy flow diagram of an intelligent vehicle device with forward-looking identifiers according to an embodiment of the present invention; Figure 11 This is a core control and drive link diagram of an intelligent vehicle device with forward-looking identifiers in one embodiment of the present invention; Figure 12 This invention provides the mechanical structure and installation relationship of an intelligent vehicle device with a forward-looking identifier in one embodiment. Figure 13 This is a flowchart of a control method for an intelligent vehicle device with forward-looking identifiers according to an embodiment of the present invention; Figure 14 This is a flowchart of a forward-looking identifier recognition algorithm for an intelligent vehicle device with forward-looking identifiers according to an embodiment of the present invention; Figure 15 This is a flowchart illustrating the image processing and track centerline extraction process of an intelligent vehicle device with forward-looking markers in one embodiment of the present invention. Figure 16 This is a flowchart illustrating the generation of cascaded PID control signals for an intelligent vehicle device with forward-looking identifiers in one embodiment of the present invention. Figure 17 This is a timing diagram of hardware collaboration and forward control of an intelligent vehicle device with forward-looking identifiers in one embodiment of the present invention.

[0027] Explanation of reference numerals in the attached figures: 1. Vehicle body; 2. Control unit; 3. Sensing module; 4. Drive module; 5. Power management module; 104. Brushless motor; 105. Drive motor; 106. Brushless motor support structure; 301. Color camera; 302. Encoder; 303. Attitude sensor; 304. Carbon fiber rod; 305. Vertical reinforcing rod; 401. Brushless motor drive circuit; 402. Brushed motor drive circuit; 403. Vertical servo motor; 501. Lithium polymer battery; 502. Three-stage step-down circuit. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] like Figures 1 to 17 As shown, an intelligent vehicle device with a forward-looking identifier according to an embodiment of the present invention is composed of a vehicle body 1, a control unit 2, a sensing module 3, a drive module 4, and a power management module 5. The selection, structural design, and connection relationship of each component are as follows: (a) Structural design and component installation of vehicle body 1 Vehicle body 1 is a self-made model car, with overall dimensions of 30cm in length, 26cm in width, and 35cm in height, weighing 1200g. The chassis is made of lightweight, high-strength materials, with a centrally located and low center of gravity to improve high-speed stability. The key mounting structure and component layout of vehicle body 1 are as follows: PCB mounting area: A rectangular PCB mounting area is reserved in the middle of the vehicle body 1 to fix the Loongson 2K0300 Jiujiupai development board of the control unit 2. The development board is connected to the chassis through 4 copper pillars with a height of 5mm to ensure that there is a heat dissipation gap between the development board and the chassis.

[0030] Motor bracket installation: The original C-model motor bracket is installed on both sides of the rear wheel of the vehicle body 1 to fix the brushed drive motor 105 and encoder 302. The motor bracket is fastened to the chassis with M3 screws. During installation, adjust the gear meshing clearance between the motor shaft and the encoder 302 to 0.1-0.2mm to reduce power transmission loss.

[0031] Brushless motor support structure 106: A 3D-printed I-shaped support column is installed on the upper middle part of the vehicle body 1. The support column is made of PLA and is 15mm high. The bottom surface is fixed to the chassis with 4 M2.5 screws, and the top surface is in contact with the mounting bracket of the brushless motor 104 (model RS2205) to ensure that the negative pressure generated by the fan is directly transmitted to the vehicle body 1, avoiding torque decomposition and negative pressure loss.

[0032] Battery compartment 107: A battery compartment is provided at the rear of the vehicle body 1 for storing lithium polymer battery 501 (3S 2200mAh 45C). The battery is fixed with Velcro, and the battery output wire is connected to the power management module 5 through the wire hole. The wire hole is provided with a rubber sleeve to prevent the wire from being worn.

[0033] (ii) Configuration and connection of control unit 2 Control unit 2 uses the Loongson 2K0300 microcontroller independently developed by Loongson Technology. This microcontroller is a 64-bit architecture with high performance and low power consumption, which is suitable for the competition's requirements for control chips.

[0034] Core parameters and peripheral circuits: The Loongson 2K0300 microcontroller's GPIO interface is expanded to 96 channels. The interfaces connected to the sensing module 3 include: the data interface of the color camera 301 (GPIO066, GPIO065), the pulse input interface of the encoder 302 (GPIO050, GPIO048), and the I2C interface of the attitude sensor 303 (GPIO062, GPIO063). The interfaces connected to the drive module 4 include: the PWM output interface of the brushless motor drive circuit 401 (GPIO088, GPIO089), the control interface of the brushed motor drive circuit 402 (GPIO020, GPIO021), and the PWM control interface of the vertical servo motor 403 (GPIO073).

[0035] Development and Debugging Environment: The development environment is based on WSL - Ubuntu 22.04 LTS subsystem, with the host Windows 11 providing a complete Linux compilation chain via WSL2. Code writing and debugging are done using VS Code, with the IntelliSense plugin and CMake Tools enabling one-click cross-compilation. The Loongson ELF executable file is generated using loongarch64-linux-gnu-gcc 10.3 and pushed to the target board via Wi-Fi SCP. The board runs a MobaXterm SSH terminal, with commands controlling file execution and stopping. Its serial port forwarding function is used for real-time log printing and online parameter adjustment.

[0036] (III) Structure and Installation of Sensing Module 3 The sensing module 3 includes a color camera 301, a 1024-line high-precision encoder 302, and a six-axis attitude sensor 303. The specific configuration, installation method, and function of each component are as follows: Color Camera 301: Utilizing a 110° distortion-free USB high frame rate color camera, this provides richer track color information compared to grayscale cameras, adapting to various lighting conditions during the race. The color camera 301 is mounted centrally at the front of the vehicle body 1 via a carbon fiber rod 304. The carbon fiber rod 304 is 8mm in diameter and 20cm long, with a vertical reinforcing rod 305 (6mm in diameter and 10cm in length) welded to its bottom. The reinforcing rod is fixed to the chassis with M2 screws, forming a stable "T"-shaped bracket. The color camera 301 is secured to the top of the carbon fiber rod 304 with clips, the lens facing the track at a 15° angle to the horizontal. The clips are sealed with hot melt adhesive to prevent angle shift during driving. The height of the color camera 301 strictly adheres to the competition rules, set at 20cm to ensure a field of view covering a 10cm near-field and a 100cm far-field field of view within the track area.

[0037] During debugging, a vibration test platform was used to simulate high-speed operation. The camera displacement was monitored by an acceleration sensor, and the torque of the reinforcing rod screws was fine-tuned until the displacement was ≤0.1mm.

[0038] The USB interface of the color camera 301 is connected to the USB Host interface of the control unit 2. Image data is transferred to the buffer of the control unit 2 via DMA. The buffer size is set to 640×480 pixels to ensure that the latest frame image is stored in real time.

[0039] Encoder 302: Two Zhufei 1024 linear direction mini encoders are selected and installed on the motor brackets of the two rear wheels respectively. The gears of encoder 302 mesh with the output gears of the motor (transmission ratio 1:1). The encoder 302 is powered by 3.3V and is connected to the 3.3V power interface of control unit 2. It outputs 1024 pulses per revolution. Control unit 2 captures the pulse signal through the GPIO interface, calculates the pulse frequency to obtain the wheel speed, and then derives the vehicle's speed. The measurement accuracy can reach 2mm (near look) and 13mm (far look), with a detection frequency of 50Hz.

[0040] The six-axis attitude sensor 303 uses an IMU660RA six-axis accelerometer and gyroscope. Based on the BMI270 chip, this sensor integrates a 16-bit three-axis gyroscope and a 16-bit three-axis accelerometer, featuring an ultra-low power consumption design with a 3.3V supply voltage. It communicates with control unit 2 via an I2C interface. The sensor is installed next to the PCB mounting area in the middle of vehicle body 1, no more than 5cm away from control unit 2 to reduce signal interference. Its main function is to collect the vehicle's angular velocity and acceleration information, calculate the vehicle's attitude changes on the circular track by integrating the angular velocity, and accurately identify circular elements to avoid the problem of the vehicle being unable to exit after entering the circular track at high speed.

[0041] (iv) Circuit design and connection of driver module 4 Drive module 4 includes a sensorless brushless motor drive circuit 401, a brushed motor drive circuit 402, and a vertical servo motor 403, which are used to convert the control signals of control unit 2 into mechanical motion. The specific design is as follows: The sensorless brushless motor drive circuit 401 drives the brushless motor 104. Its core chips are the FD6288Q pre-driver chip and the STC15F2K60S2 microcontroller. The FD6288Q chip preprocesses the PWM signal output from the control unit 2 to enhance the driving capability. Its three-phase output is connected to the three-phase windings of the brushless motor 104 via a TPH1R403NL MOSFET to form an H-bridge circuit. The STC microcontroller detects the back electromotive force of the brushless motor 104 and determines the commutation timing through zero-crossing detection, achieving sensorless commutation. The circuit also includes a voltage detection module that collects the battery voltage through a voltage divider resistor. When the voltage drops below 11.1V, the control unit 2 reduces the power of the brushless motor 104 to prevent insufficient negative voltage due to insufficient power supply. The power input of the brushless motor drive circuit 401 is connected to the 12V output of the power management module 5, allowing for a working current of up to 20A, meeting the power requirements of the brushless motor 104.

[0042] Brushed motor drive circuit 402: This circuit drives the brushed drive motor 105. The core chip is the DRV8701E driver chip, which supports high current output with a rated operating current of no less than 100A. It can control four NMOS transistors to form an H-bridge circuit, enabling forward and reverse rotation and speed regulation of the motor. The PWM input interface of the DRV8701E is connected to the PWM output interface of the control unit 2 to receive speed control signals; the current detection interface collects the motor current through a sampling resistor to achieve overcurrent protection. The power input terminal of the brushed motor drive circuit 402 is directly connected to the lithium polymer battery 501 to ensure sufficient power. The output terminal is connected to both ends of the brushed drive motor 105, driving the rear wheel to rotate.

[0043] Vertical Servo 403: An SU400 servo is selected and vertically mounted on the bracket of the front wheel steering mechanism of vehicle body 1. The servo's output shaft is connected to the front wheel steering arm via a connecting rod, facilitating steering by pushing or pulling the front wheels. The vertical servo 403 is powered by 6V and connected to the 6V output terminal of the power management module 5. Its control signal input terminal is connected to the GPIO073 interface of control unit 2 to receive direction control signals. The servo's steering angle range is -30° to +30°, and its response time is ≤0.1s, ensuring rapid direction adjustment when cornering.

[0044] (v) Circuit design and power distribution of power management module 5 Power management module 5 provides stable power to all modules. Its core components include a lithium polymer battery 501 and a three-stage buck circuit 502, as detailed below: Lithium polymer battery 501: Uses a 3S Grignard 2200mAh 45C battery with a nominal voltage of 11.1V, a fully charged voltage of 12.6V, and a capacity of 2200mAh, sufficient for continuous operation of the vehicle for over 30 minutes. The battery's positive and negative terminals are connected to the three-stage step-down circuit 502 via the main power switch. When the switch is closed, the battery voltage is input to each step-down chip.

[0045] 502 Three-stage step-down circuit: 12V-5V step-down circuit: Utilizing a TPS54302 DC-DC regulator chip, the input voltage range is 9V-15V, the output voltage is 5V, and the output current is 3A. A 22μF electrolytic capacitor and a 0.1μF ceramic capacitor are connected in parallel at the input for filtering, and a 47μF electrolytic capacitor and a 0.1μF ceramic capacitor are connected in parallel at the output for voltage stabilization. This circuit powers control unit 2, the buzzer, and the color camera 301.

[0046] 5V-3.3V step-down circuit: Utilizes an RT9013-3.3GB linear regulator chip, with an input voltage of 5V, an output voltage of 3.3V, and an output current of 500mA. A 100Ω current-limiting resistor is connected in series at the input, and a 1μF ceramic capacitor is connected in parallel at the output for filtering, supplying power to the six-axis attitude sensor 303, the display screen, and the encoder 302.

[0047] 12V-6V step-down circuit: Utilizing the LM1085-ADJ adjustable voltage regulator chip, the input 12V voltage is reduced to 6V by adjusting the ratio of the feedback resistors, with an output current of 2A. A 22μF electrolytic capacitor is connected in parallel at the input, and a 10μF electrolytic capacitor and a 0.1μF ceramic capacitor are connected in parallel at the output to power the vertical servo motor 403.

[0048] II. Core Algorithms and Implementation Process of the Software System The software system of control unit 2 is developed based on the C language. Its core components include system initialization, image acquisition and processing, forward-looking sign recognition, PID control algorithm, and parking judgment modules. The implementation details and interaction flow of each module are as follows: (I) System Initialization After the system is powered on, control unit 2 first executes the initialization program to complete the following configuration: 1. Hardware peripheral initialization: Initialize the motor PWM output channel and set the PWM frequency to 10kHz; initialize the 0-degree reference direction of the vertical servo 403, at which point the front wheel is in a straight-line state; initialize the registers of the color camera 301, set the screen baud rate to 115200, and the resolution to 640×480 pixels; initialize the CPU process priority, set the image processing process to the highest priority to ensure real-time response.

[0049] 2. Software parameter initialization: Initialize the image buffer, set the buffer read / write flag variable to "not ready"; initialize the PID control parameters, including the proportional coefficient K of the speed loop incremental PID. p =0.8, integral coefficient Kᵢ=0.1, derivative coefficient Kd=0.2, proportional coefficient K of the directional loop position PID (without integral term) p =1.2, differential coefficient Kd=0.3; initialize global variable Global_road_type, default value is "unknown track"; initialize parking judgment flag stop=0.

[0050] (II) Image Acquisition After initialization, control unit 2 activates color camera 301 to acquire images. The specific process is as follows: 1. The control unit 2 sends acquisition commands to the color camera 301 via the USB interface. The CMOS photoelectric sensor of the color camera 301 converts the track optical image into electronic signals, which are then converted into digital image data by the ADC analog-to-digital converter.

[0051] Control unit 2 sets task priorities: image acquisition is the highest priority, and PID calculation is the second highest priority; DMA transfer is used to reduce CPU usage and ensure that the latency is <10ms at a processing frequency of 50Hz.

[0052] 2. Digital image data is written to the preset buffer of control unit 2 in real time via DMA transfer. After the transfer is completed, the buffer read / write flag variable is set to "ready", triggering an image processing interrupt.

[0053] (iii) Proactive Identification of Markers After detecting that the buffer is "ready", control unit 2 first executes the forward-looking marker recognition procedure, the process of which is as follows: 1. The program scans the image data in the buffer and focuses on the upper half of the image (look-ahead area), which is the forward-looking location of the track, where forward-looking landmarks are set.

[0054] 2. Identification based on the positional characteristics of the white arrow on the marker: A preset grayscale threshold of 180 for white pixels is used (fixed threshold for blocks, adaptable to different lighting conditions). Pixel areas with grayscale values ​​≥180 in the image are scanned to determine if they form an arrow shape. Eight marker types are distinguished by the arrow's direction (left, right, straight) and additional features (such as circular markers, obstacle markers). Straight line marker: White arrow pointing horizontally to the right, with no additional markings; Left turn sign: White arrow tilted 45° to the upper left; Right turn sign: White arrow tilted 45° to the upper right; Left ring marker: White arrow tilts to the upper left, with a circular mark at the end of the arrow; Right-hand sign: A white arrow tilts upwards and to the right, with a circular mark at the end of the arrow; Left obstacle marker: White arrow tilted to the upper left, with a square mark in front of the arrow; Right obstacle marker: White arrow tilted to the upper right, with a square mark in front of the arrow; Straight ahead sign at intersection: White arrow pointing horizontally to the right, with two parallel horizontal lines below the arrow.

[0055] 3. After identification is complete, the identifier type is stored in the global variable Global_road_type, and the buffer read / write flag variable is updated to "not ready" to prepare for the next collection.

[0056] (iv) Image processing and centerline extraction After identifying forward-looking landmarks, control unit 2 performs subsequent processing on the same frame image to extract the track centerline: 1. Binarization: The image is divided into 16×16 blocks using a block-based fixed threshold method. Each block is assigned an independent grayscale threshold (determined through pre-race calibration). Pixels with grayscale values ​​higher than the threshold are set to 255 (white, track area), and pixels with grayscale values ​​lower than the threshold are set to 0 (black, background area), thus achieving image binarization.

[0057] Threshold calibration method: Under pre-race lighting conditions, sample images of the track are collected, and the grayscale histogram of each 16×16 block is statistically analyzed. The valley value of the histogram is used as the threshold. The calibration data is stored in the non-volatile memory of the control unit 2 for runtime access.

[0058] 2. Double-scanning method for edge line extraction: The first scan uses the "climbing line" method, starting from the known track edge line at the bottom of the image (near-sight region) and tracing the edge line position line by line upwards; the second scan uses the "longest white column scan" method, which counts the continuous length of white pixels line by line, and the edge of the longest continuous segment is the track edge line. The intersection of the two scan results is taken to improve the accuracy of edge line recognition.

[0059] 3. Median Filtering Noise Reduction: The edge coordinate data obtained from the scanning process is processed using a median filtering algorithm. The formula is as follows: Where w is a 3×3 neighborhood range, the median is selected to filter out the coordinates of noise points with random abrupt changes, ensuring the smoothness of the edge.

[0060] 4. Centerline Calculation: Take the midpoint of the coordinates of the left and right sidelines to obtain the coordinate sequence of the track centerline, which is stored in the array mid_line_list as the reference for directional control.

[0061] (v) PID control calculation and control signal generation Control unit 2 combines the value of Global_road_type, the rotational speed signal from encoder 302, and the attitude signal from attitude sensor 303 to generate speed control and direction control signals using a cascaded PID algorithm. 1. Speed ​​control signal generation: The pulse signal from encoder 302 is input through the GPIO interface of control unit 2. The program calculates the pulse count every 10ms and derives the current wheel speed based on the pulse count and encoder 302 parameters (1024 lines / revolution). .

[0062] Set the target speed based on Global_road_type : When the indicator is "Go straight" or "Go straight at an intersection", the corresponding motor PWM value is 90%. When marked as a left turn or a right turn The corresponding motor PWM value is 70%; When marked as left loop, right loop, left obstacle, or right obstacle, The corresponding motor PWM value is 80%.

[0063] The speed loop uses an incremental PID algorithm to calculate the control increment: Where e(k) = The ∆u(k) is superimposed on the current PWM value to obtain the final speed control signal, which is then output to the brushed motor drive circuit 402.

[0064] 2. Direction control signal generation: Calculate the deviation between the track centerline and the image center. (Unit: pixels) Combined with the angular velocity information collected by the attitude sensor 303, the deviation value is corrected (such as in a circular track, the deviation caused by centrifugal force is compensated by the integral of angular velocity).

[0065] The direction loop uses a positional PID algorithm without an integral term to calculate the control output: u(k) corresponds to the steering angle of the vertical servo 403. u(k) is converted into a PWM signal and output to the vertical servo 403 to adjust the steering of the front wheel.

[0066] Data fusion employs a weighted average method: Let the encoder 302's velocity measurement weight be α = 0.7, and the attitude sensor 303's angular velocity integral weight be β = 0.3, with the final deviation... ,in For centerline deviation, This is the integral value of the angular velocity.

[0067] 3. Differential adjustment: The output u(k) of the steering loop is synchronously fed back to the speed loop. When the absolute value of u(k) is greater than 10 (corresponding to a steering angle of 15°), the PWM difference between the left and right rear wheels is adjusted. The PWM value of the inner wheel is reduced by 5%-10%, realizing differential cornering and avoiding going off the track.

[0068] (vi) Drive module response and vehicle movement 1. After receiving the speed control signal, the brushed motor drive circuit 402 uses the DRV8701E chip to control the switching frequency of the NMOS transistor in the H-bridge circuit, thereby adjusting the speed of the brushed drive motor 105 and realizing the speed regulation of the trolley.

[0069] 2. After receiving the direction control signal, the vertical servo 403 drives the gear set to rotate, which in turn drives the connecting rod to push and pull the front wheel to steer. The steering angle is linearly related to the PWM duty cycle of the control signal.

[0070] 3. The sensorless brushless motor drive circuit 401 receives the voltage detection signal and PWM signal from the control unit 2. After preprocessing by the FD6288Q chip, it drives the brushless motor 104 to rotate. The fan generates negative pressure, which is transmitted to the vehicle body 1 through the I-shaped support to improve the vehicle's grip.

[0071] (vii) Parking judgment After each image acquisition, the program additionally scans the bottom of the image (near-sight region) to determine if there are departure line features (multiple parallel white horizontal lines): 1. When the departure line feature is detected, the program reads the current vehicle speed. ,like If the speed is greater than 1m / s, then set stop=1 to generate a parking control signal, set the motor PWM value to 0, and the vertical servo motor 403 returns to 0 degrees, stopping the car.

[0072] 2. If no departure line characteristics are detected or If the speed is ≤1m / s, the image acquisition and subsequent steps will be executed repeatedly until the parking conditions are met.

[0073] III. Equipment Commissioning and Optimization To ensure stable operation of the device, hardware debugging and software parameter optimization are required. The specific process is as follows: 1. Hardware debugging: Power supply debugging: Use a multimeter to measure the output voltage of the 502 three-stage step-down circuit, and ensure that the 5V output error is ≤ ±0.1V, the 3.3V output error is ≤ ±0.05V, and the 6V output error is ≤ ±0.1V; Motor drive debugging: Send test PWM signals through serial port debugging assistant to observe whether the speed of brushed drive motor 105 is stable and whether the commutation of brushless motor 104 is smooth, without any stuttering or abnormal noise. Sensor debugging: Check whether the image transmission of the color camera 301 is stable and without frame loss; rotate the encoder 302 and observe whether the number of pulses received by the control unit 2 is accurate; shake the trolley and observe whether the output data of the attitude sensor 303 changes continuously.

[0074] 2. Software parameter optimization: Using VSCode's online debugging function, run the image processing program step by step, adjust the binarization threshold and scanning parameters, and ensure accurate edge line recognition. PID parameters are tuned on an actual racetrack using an engineering tuning method, starting with adjusting K. p To make the car respond quickly, Kd is adjusted to suppress oscillations, and finally Kᵢ is adjusted to eliminate static errors, until the car can drive smoothly under all track types. Optimize the recognition algorithm for forward-looking markers, adjust the recognition threshold for arrow features, and improve the recognition accuracy under different lighting conditions.

[0075] To improve environmental adaptability, the markers can use reflective materials or add infrared features; the software algorithm adds an adaptive threshold adjustment module to dynamically update the binarization threshold in real time according to the ambient light.

[0076] IV. Key Technological Highlights and Implementation Results 1. Forward-looking marker design: By using 8 different markers to indicate the track type in advance, the problem of insufficient forward-looking capability caused by the limited height of the camera is solved, enabling the car to adjust its speed and direction in advance when driving at high speed, thus breaking through the speed bottleneck.

[0077] 2. Hardware and software co-optimization: The hardware adopts a modular design, ensuring stable and reliable circuitry and a reasonably reinforced mechanical structure; the software combines algorithms such as double-scanning and median filtering to improve image processing accuracy, and the cascaded PID algorithm achieves precise coordination of speed and direction, significantly improving cornering smoothness and driving stability.

[0078] 3. High scalability: The core technology of this device can be directly extended to scenarios such as autonomous navigation robots and intelligent inspection vehicles. Only by adjusting the type of markers and control parameters according to actual needs, autonomous tracking in different scenarios can be achieved.

[0079] Appendix: Main Technical Parameters of Four-Wheel Steering Vehicle It should be noted that the control method in the embodiments of this application can be automatically controlled by a controller. The control method of the controller can be implemented by simple programming by those skilled in the art, which is common knowledge in the field. Furthermore, this application is mainly used to protect the entire device including mechanical structure, software core algorithm and process design, so this application will not explain the control method and circuit connection in detail.

[0080] Obviously, the above-described embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A smart cart device with a forward looking marker, characterized by, The application relates to a vehicle body (1), a control unit (2), a sensing module (3), a driving module (4) and a power management module (5); the sensing module (3), the driving module (4) and the power management module (5) are electrically connected with the control unit (2); the sensing module (3) is used for collecting image information of a forward-looking marker in a forward-looking area of a race track, and the forward-looking marker is used for representing a subsequent type of the race track; the control unit (2) is configured to identify the forward-looking marker based on the image information and generate a control signal according to an identification result; and the driving module (4) is in transmission connection with a walking component of the vehicle body (1) and is used for adjusting a traveling state of the vehicle in response to the control signal.

2. The smart trolley device with a forward looking marker of claim 1, wherein, The forward-looking marker comprises multiple types, and different types correspond to one of a straight track, a left curve, a right curve, a left ring, a right ring, a left obstacle, a right obstacle and a straight track at an intersection.

3. The smart cart apparatus with forward looking markers of claim 2, wherein, The multiple types of forward-looking markers are distinguished by preset differentiated visual features, and the visual features include preset shapes, colors and patterns and are used for accurate identification by the sensing module.

4. The smart cart apparatus with forward looking markers of claim 1, wherein, The sensing module (3) comprises a color camera (301), a 1024-line band direction high-precision encoder (302) and a six-axis attitude sensor (303); the color camera (301) is installed at a front part of the vehicle body (1) and is used for collecting images of the forward-looking marker and the race track; the encoder (302) is in transmission cooperation with driving wheels of the vehicle body (1) and is used for detecting wheel speeds; and the attitude sensor (303) is used for collecting angular velocity and acceleration information of the vehicle and is electrically connected with the control unit (2) through a signal output end.

5. The smart cart device with a forward looking marker of claim 1, wherein, The color camera (301) is installed through a carbon rod (304), the bottom of the carbon rod (304) is provided with a vertical reinforcing rod (305), the camera (301) is fixed at an angle and has a height of 20 cm, and an image output end of the camera (301) is bound with a data interface of the control unit (2).

6. The smart cart apparatus with forward looking markers of claim 1, wherein, The control unit (2) receives speed signals of the encoder (302) and attitude signals of the attitude sensor (303), generates speed control signals and direction control signals in combination with an identification result of the forward-looking marker, and the speed control signals and the direction control signals cooperatively control the driving module (4).

7. The smart cart apparatus with forward looking markers of claim 1, wherein, The driving module (4) comprises an inductance-free brushless motor driving circuit (401), a brush motor driving circuit (402) and a vertical rudder (403); the vertical rudder (403) is used for driving front wheels to turn in response to the direction control signals; the brushless motor driving circuit (401) realizes commutation through back electromotive force detection and is electrically connected with a brushless motor (104) of the vehicle body (1); the brush motor driving circuit (402) adopts an H-bridge topology structure and has a rated working current of not less than 100 A and is electrically connected with a driving motor (105) of the vehicle body (1).

8. The smart cart apparatus with forward looking markers of claim 1, wherein, The power management module (5) comprises a lithium polymer battery (501) and a three-stage voltage reduction circuit (502); the three-stage voltage reduction circuit (502) outputs 5V, 3.3V and 6V voltages respectively, and supplies power for a control unit (2), a sensing module (3) and a steering engine (403), and the input ends thereof are electrically connected with the lithium polymer battery (501).

9. The smart cart apparatus with forward looking markers of claim 1, wherein, The vehicle body (1) is provided with a brushless motor support structure (106), the support structure (106) is a 3D printing H-shaped column, is vertically fixed between the chassis of the vehicle body (1) and the brushless motor (104), and is used for conducting negative pressure.

10. A control method of the intelligent trolley device according to any one of claims 1 to 9, characterized by, The method comprises the following steps: S1: system initialization, completing the start configuration of the sensing module, the driving module and the control unit; S2: the sensing module collects image information of a front-looking area of a track and transmits the image information to the control unit; S3: the control unit identifies the type of a front-looking marker based on the image information, and determines the subsequent type of the track; S4: the control unit generates a speed control signal and a direction control signal in combination with a wheel rotation speed signal and a car posture signal; S5: the driving module adjusts the driving speed and the steering angle of the car in response to the control signal; S6: when the feature of a starting line is detected and the speed of the car is greater than 1m / s, the car is triggered to stop, otherwise, S2-S5 are circularly executed.

Images