Intelligent vehicle control system based on ADAS system

Abstract

The invention, which relates to the technical field of intelligent vehicle control, discloses an ADAS-based intelligent vehicle control system comprising a data acquisition end, a data analysis end and a data output end. The data acquisition end is used for recording environment information of a field track in real time, carrying out primary processing on the environment information, packaging processed data into a data set and sending the data set to the data analysis end; the data analysis end is used for analyzing and processing the data set processed by the data acquisition end to obtain a simulation route model of the car model advancing along a racing track; and the data output end is used for performing comprehensive processing on state information and racing track information when the car model travels through the data analysis end by simulating management, analysis and calculation of the route model so as to ensure rationalization and feasibility of car model traveling. The method has the beneficial effects that the application of the ADAS system on the small car model is realized, and the small car model is assisted in better completing a racing track competition to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicle control, and in particular to an intelligent vehicle control system based on an ADAS system. Background Art

[0002] The ADAS system, also known as the Advanced Driver Assistance System, uses a variety of sensors installed on the vehicle (millimeter wave radar, lidar, monocular / binocular cameras, and satellite navigation) to sense the surrounding environment at any time during driving, collect data, identify, detect, and track static and dynamic objects, and combine navigation map data to perform systematic calculations and analysis, so that the driver can be aware of possible dangers in advance, effectively increasing the comfort and safety of car driving.

[0003] Currently, ADAS systems have been widely used in commercial vehicles and large-scale automobile competitions. However, due to the large size of electronic devices such as sensors used in these vehicles and the need for strong computing power of the system's core computing chips, it is impossible to directly copy such ADAS systems to small car models participating in competitions held in closed venues.

[0004] The above ADAS system was studied and found that the main reasons why it cannot be directly applied to small car models are as follows:

[0005] 1. The sensors and other electronic components used in ordinary commercial vehicles are too large and heavy, so they are not suitable for small car models;

[0006] 2. The closed-field competition for small car models has a simple environment, and high-computing-power computing chips are not suitable for calculations in simple environments.

[0007] Therefore, an intelligent vehicle control system based on ADAS system is proposed to solve the above problems. Summary of the invention

[0008] The main purpose of the present invention is to provide an intelligent vehicle control system based on an ADAS system to solve the problems raised in the above background.

[0009] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: an intelligent vehicle control system based on an ADAS system, the system comprising a data acquisition end, a data analysis end and a data output end;

[0010] The data acquisition end is used to: record the track environment information in real time, perform preliminary processing on the environment information, and package the processed data into a data set and send it to the data analysis end;

[0011] The data analysis end is used to obtain a simulated route model of the car model traveling along the track by analyzing and processing the data set processed by the data acquisition end;

[0012] The data output end is used to: manage, analyze and calculate the simulation route model, connect the data with the key parts of the car model, integrate the track data with the car model attribute information, and receive the status information fed back by the car model when it is moving in real time, send the status information of the car model to the data analysis end, and comprehensively process the status information of the car model when it is moving and the track information through the data analysis end to ensure the rationality and feasibility of the car model's movement.

[0013] Preferably, the data acquisition terminal includes a track environment information acquisition module and a vehicle model running state acquisition module;

[0014] The track environment information acquisition module includes: a video acquisition unit;

[0015] The video acquisition unit is used to collect environmental information of the track through a micro camera installed on the car model, and the environmental information includes: uphill, downhill, straight, curve, continuous straight and continuous curve.

[0016] Preferably, the vehicle model running state acquisition module includes: a vehicle model transmission information acquisition unit and a vehicle model battery information acquisition unit;

[0017] The vehicle model transmission information collection unit is used to collect the rotation speed information of the vehicle model transmission shaft through a rotor sensor arranged at the vehicle model transmission shaft;

[0018] The vehicle model battery information acquisition unit is used to: acquire voltage output information of the vehicle model battery through a current sensor arranged at the output end of the vehicle model battery.

[0019] Preferably, the track environment information collection module performs the following process:

[0020] S1: Collect video information of the track through the video acquisition unit;

[0021] S2: pre-processing the collected video information of the track;

[0022] S3: Integrate the features of the pre-processed video information and form a corresponding set.

[0023] Preferably, the step S2 is specifically:

[0024] The collected video information is filtered, denoised, binarized and edge detected.

[0025] Preferably, the step S3 is specifically:

[0026] The track features are extracted from the preprocessed video information into: uphill set S, downhill set X, straight set Z, curve set W, continuous straight set LZ and continuous curve set LW. The division of track features follows the following rules:

[0027] Set the slope threshold to ±5°. When the slope is greater than 5°, the track is marked as uphill; when the slope is less than -5°, the track is marked as downhill.

[0028] Set the track's curvature angle threshold to ±5°. When the curvature angle is greater than 5° or less than -5°, the track is marked as entering a curve.

[0029] When the track curvature angle is less than or equal to 5° and greater than or equal to -5°, and the slope is less than or equal to 5° and greater than or equal to -5°, the track is marked as a straight road. When the length of the straight road is greater than or equal to 3 meters, this section of the track is marked as a continuous straight road.

[0030] When there are two or more consecutive bends on the track, this section of the track is marked as continuous bends.

[0031] Preferably, the data analysis end includes a data processor, which is used to integrate and process the data information collected by the track environment information collection module and the vehicle model moving state collection module, and send the analysis and processing results to the data output end. The data processor performs the following process:

[0032] The complete track information matrix is ​​obtained through the track search algorithm;

[0033] The center line of the track is calculated through the track information matrix;

[0034] Obtain the optimal route for the car model to travel along the track based on the center line of the track, and perform balance control on the car model;

[0035] According to the optimal route of the track, the battery output power and drive shaft speed of different characteristic sections of the track are allocated to the car model.

[0036] Preferably, the complete track information matrix is ​​obtained by using a track search algorithm, specifically:

[0037] Extract the connected domain of the track image and find the connected domain position corresponding to the normal track;

[0038] Traverse the image, extract the boundary information and width information of the connected domain corresponding to the track, filter the boundary, and store it in an array;

[0039] After the line search is completed, a complete matrix is ​​obtained that can be processed for track analysis.

[0040] Preferably, the method of obtaining the optimal route for the car model to travel along the track according to the center line of the track is specifically as follows:

[0041] After obtaining the complete track information matrix, scan from the bottom to the top of the image, map the left and right boundaries, and obtain a mapping relationship, which is recorded as: Y z and Y y , when the track connection domain ends, the scan ends; through the obtained mapping relationship, the midpoint of the corresponding left and right boundaries is taken and recorded as Y z1 and Y y1 , calculate Y z1 To Y y1 The straight-line distance is recorded as: Y L , then the position D of the midpoint is:

[0042]

[0043] And save the position of the midpoint in an array.

[0044] Preferably, the balance control of the car model is specifically as follows:

[0045] Adopt cascade PID algorithm, which is divided into angular velocity loop, angle loop and speed loop;

[0046] In the outermost loop, the angular velocity of the heading angle is used as the velocity loop;

[0047] For the adjustment of the upright loop PID parameters, first adjust the Kp of the angular velocity loop so that there is obvious resistance when the car body falls to one side, then adjust the Kp of the angle loop so that the car body can stand upright for 3 to 4 seconds, and finally adjust the Kp of the speed loop so that the car can stand upright stably.

[0048] The present invention has the following beneficial effects:

[0049] 1. In the present invention, by setting a data collection terminal, all-round data collection is carried out on the track in which the small car model participates and the body of the small car model, thereby realizing the integration of relevant data of the small car model for the track, and by extracting the features of different sections in the track, a data set for each section is established, which is beneficial for the small car model to make corresponding voltage output and torque output adjustments for different sections.

[0050] 2. In the present invention, by setting up a data analysis terminal, the miniaturization of the commercial ADAS system is achieved, which can be more suitable for small car models and avoid the cost waste of high-computing power chips in the commercial ADAS system. At the same time, through the data analysis terminal, the data information collected by the data acquisition terminal is integrated and analyzed, and the small car model is assisted in selecting the optimal path of the track, which helps to improve the maximum full speed of the small car model on the track.

[0051] 3. In the present invention, by setting a data output terminal, real-time coordination between the car model body and the data processor is achieved, and different feedback can be applied to the car model body according to different sections of the car model on the track, thereby further improving the performance of the car model. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic diagram of the overall system architecture of an ADAS-based smart car control system according to the present invention;

[0053] Figure 2 This is a flow chart of the execution of a track environment information acquisition module of an ADAS system-based smart car control system of the present invention;

[0054] Figure 3 This is a data processing end execution flow chart of an ADAS system-based intelligent vehicle control system of the present invention;

[0055] Figure 4 The present invention is a flow chart of a track line search algorithm of an intelligent vehicle control system based on an ADAS system. DETAILED DESCRIPTION

[0056] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall within the scope limited by the application equally.

[0057] In the present invention, terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are relational words determined only for the convenience of describing the structural relationships of the various parts or elements of the present invention, and do not specifically refer to any part or element in the present invention and should not be understood as limitations on the present invention.

[0058] In the present invention, terms such as "fixed connection", "connected", "connection", etc. should be understood in a broad sense, indicating that it can be fixedly connected, integrally connected or detachably connected; it can be directly connected or indirectly connected through an intermediate medium. Relevant scientific research or technical personnel in this field can determine the specific meanings of the above terms in the present invention according to specific circumstances, and they should not be understood as limiting the present invention.

[0059] like Figures 1 to 4 As shown, this embodiment provides an intelligent vehicle control system based on an ADAS system, including a data acquisition terminal, a data analysis terminal and a data output terminal;

[0060] The data acquisition end is used to: record the track environment information in real time, perform preliminary processing on the environment information, and package the processed data into a data set and send it to the data analysis end;

[0061] The data analysis end is used to obtain a simulated route model of the car model traveling along the track by analyzing and processing the data set processed by the data acquisition end;

[0062] The data output end is used to: manage, analyze and calculate the simulation route model, connect the data with the key parts of the car model, integrate the track data with the car model attribute information, and receive the status information fed back by the car model when it is moving in real time, send the status information of the car model to the data analysis end, and comprehensively process the status information of the car model when it is moving and the track information through the data analysis end to ensure the rationality and feasibility of the car model's movement.

[0063] The data collection terminal includes a track environment information collection module and a vehicle model running state collection module;

[0064] The track environment information acquisition module includes: a video acquisition unit;

[0065] The video acquisition unit is used to collect environmental information of the track through a micro camera installed on the car model, and the environmental information includes: uphill, downhill, straight, curve, continuous straight and continuous curve.

[0066] The vehicle model running state acquisition module includes: a vehicle model transmission information acquisition unit and a vehicle model battery information acquisition unit;

[0067] The vehicle model transmission information collection unit is used to collect the rotation speed information of the vehicle model transmission shaft through a rotor sensor arranged at the vehicle model transmission shaft;

[0068] The vehicle model battery information acquisition unit is used to: acquire voltage output information of the vehicle model battery through a current sensor arranged at the output end of the vehicle model battery.

[0069] The track environment information collection module performs the following process:

[0070] S1: Collect video information of the track through the video acquisition unit;

[0071] S2: pre-processing the collected video information of the track;

[0072] S3: integrating the features of the pre-processed video information and forming a corresponding set;

[0073] Wherein, the step S2 is specifically:

[0074] Perform filtering, denoising, binarization and edge detection on the collected video information;

[0075] (1) Filtering and denoising:

[0076] Decomposing the captured video into video frames and converting the decomposed video frames into grayscale frames to reduce random noise in the video;

[0077] Through the convolutional neural network, the grayscale frame is feature extracted, the grayscale frame is mapped into a high-dimensional feature space, and the noise therein is sampled;

[0078] Perform filtering and denoising based on the sampling results;

[0079] (2) Binarization

[0080] Set a global threshold and compare the image pixel values ​​with the threshold. Values ​​greater than the threshold are set to white (255), and values ​​less than the threshold are set to black (0).

[0081] (3) Edge detection

[0082] Calculate the spatial gradient of image brightness and obtain the edge strength by taking the square root of the sum of the squares of the gradients in the horizontal and vertical directions;

[0083] Calculate the second-order derivative of the image, which is more sensitive to edge responses;

[0084] Use the Canny edge detector, which combines multi-level edge detection, non-maximum suppression, and dual threshold detection;

[0085] Morphological operations can be used to remove isolated noise points or fill small gaps to improve edge detection results;

[0086] By detecting contours on the binarized image, the shape of the object in the image can be further extracted.

[0087] In this embodiment, feature extraction is mainly achieved through a convolutional neural network, which specifically includes:

[0088] 1. Convolutional layer

[0089] Local connection and weight sharing: In the convolution layer, each neuron is only connected to a local area (receptive field) of the input image. This design reduces the number of model parameters. At the same time, the weight of each convolution kernel (or filter) is the same across the entire input image. This weight sharing mechanism further reduces the number of model parameters and enables the model to capture the spatial hierarchy in the image.

[0090] Multiple convolution kernels and feature maps: A convolution layer usually contains multiple convolution kernels, each of which is responsible for capturing different features, such as edges, corners, textures, etc. When a convolution kernel slides over the input image, it produces a feature map that represents the intensity and location of a specific feature in the input image. The outputs of multiple convolution kernels are stacked to form a new feature map that contains information about multiple features in the input image.

[0091] 2. Activation Function

[0092] Non-linear transformation: In order to increase the non-linear expression ability of the model, a non-linear activation function such as ReLU (Rectified Linear Unit) is usually applied after the convolution layer. The definition of the ReLU function is f(x) = max(0,x), which sets all negative values ​​to 0 and keeps positive values ​​unchanged. This operation not only increases the non-linearity of the model, but also helps alleviate the gradient vanishing problem and speed up training;

[0093] 3. Pooling layer

[0094] Dimensionality reduction and enhanced invariance: The main function of the pooling layer is to reduce the spatial dimension of the feature map, reduce the amount of calculation, and improve the generalization ability of the model. Pooling operations are usually divided into two types: maximum pooling and average pooling. Maximum pooling selects the maximum value in each pooling window as the output, which helps to retain the most significant features; while average pooling selects the average value, which helps to smooth the feature map and reduce the impact of noise;

[0095] 4. Fully connected layer

[0096] Feature integration and classification decision: After the convolutional layer and the pooling layer, the feature map is flattened into a one-dimensional vector and input to the fully connected layer. Each neuron in the fully connected layer is connected to all neurons in the previous layer, which enables the network to learn complex relationships between features. In the last few layers of CNN, the fully connected layer usually acts as a classifier and outputs the probability of each category. Through the activation function (such as the Softmax function), the probability distribution of each category is output to achieve classification decisions.

[0097] The step S3 is specifically:

[0098] The track features are extracted from the preprocessed video information into: uphill set S, downhill set X, straight set Z, curve set W, continuous straight set LZ and continuous curve set LW. The division of track features follows the following rules:

[0099] Set the slope threshold to ±5°. When the slope is greater than 5°, the track is marked as uphill; when the slope is less than -5°, the track is marked as downhill.

[0100] Set the track's curvature angle threshold to ±5°. When the curvature angle is greater than 5° or less than -5°, the track is marked as entering a curve.

[0101] When the track curvature angle is less than or equal to 5° and greater than or equal to -5°, and the slope is less than or equal to 5° and greater than or equal to -5°, the track is marked as a straight road. When the length of the straight road is greater than or equal to 3 meters, this section of the track is marked as a continuous straight road.

[0102] When there are two or more consecutive bends on the track, this section of the track is marked as continuous bends.

[0103] After the microcontroller collects the image signal, it needs to be processed to extract the main track information. At the same time, due to the existence of crossroads, starting lines, uneven light, reflective points outside the track, and unclear images in the distance of the track, the image effect will be greatly reduced. Therefore, the software must eliminate interference factors, effectively identify the track, and provide as much track information as possible for decision-making.

[0104] The track information we extract in image signal processing mainly includes: the position of the track in the image, track boundary point information, track width information, track centerline position, and track type identification.

[0105] Due to the characteristics of the camera itself and the visual principle that things are smaller when far away and larger when near, the image will be deformed in a trapezoidal form, which makes the track information seen by the camera different from the actual track information. Therefore, the track is used for measurement and calibration to obtain a series of parameters, and the original image captured by the camera is restored to the real track information. The original image is a two-dimensional data matrix obtained by converting the analog image through an analog circuit. Each element of the matrix corresponds to a pixel point. The first row of the image corresponds to the farthest point, about 100cm from the camera fixing pole, and the bottom row of the image corresponds to the closest point, 5cm from the fixing pole. The image at a distance is small, and the image at a close distance is large, and the black line is in a trapezoidal shape.

[0106] In the matrix returned by the camera, each pixel has a grayscale value from 0 to 255. The larger the value, the brighter the pixel. Conversely, the darker the pixel, the smaller the value. The grayscale difference between the same color is small, while the grayscale difference between different colors is large. We use a fixed threshold to binarize the pixel matrix to obtain a black and white image to distinguish the track from the background.

[0107] The camera height and depression angle were calibrated to obtain a series of parameters, and the original perspective image was subjected to inverse perspective transformation to obtain the basic image for single-chip microcomputer processing.

[0108] In this embodiment, a 140-degree wide-angle lens is used, which may cause deviation in the judgment of the track information. In order to overcome this error, the camera is calibrated, which plays a very important role in judging the track information and accurately finding the center line;

[0109] The original image reflects the image directly seen by the camera installed on the smart car, which contains image distortion. Therefore, the information on the image cannot reflect the real information of the track. The actual distance between two adjacent pixel rows on the image will change with the distance from the camera.

[0110] In order to facilitate the processing of the image information captured by the camera, the original image is subjected to an inverse perspective transformation, and an inverse perspective transformation matrix is ​​used to realize the mapping of the original image to the top view.

[0111] The data analysis end includes a data processor, which is used to integrate and process the data information collected by the track environment information collection module and the vehicle model moving state collection module, and send the analysis and processing results to the data output end. The data processor performs the following process:

[0112] The complete track information matrix is ​​obtained through the track search algorithm;

[0113] The center line of the track is calculated through the track information matrix;

[0114] Obtain the optimal route for the car model to travel along the track based on the center line of the track, and perform balance control on the car model;

[0115] According to the optimal route of the track, the battery output power and transmission shaft speed of different characteristic sections of the track are allocated to the car models;

[0116] The complete track information matrix is ​​obtained by the track search algorithm, specifically:

[0117] Extract the connected domain of the track image and find the connected domain position corresponding to the normal track;

[0118] Traverse the image, extract the boundary information and width information of the connected domain corresponding to the track, filter the boundary, and store it in an array;

[0119] After the line search is completed, a complete matrix is ​​obtained that can be processed for track analysis.

[0120] After obtaining the complete track information matrix, the center line of the track can be calculated through a certain algorithm. Based on the idea of ​​PID control algorithm, the offset between the image center of the smart car and the actual center of the track is used to control the rotation of the servo and the driving force of the motor. Track center line algorithm:

[0121] (1) Scan from the bottom to the top of the image, map the left and right boundaries, and obtain a mapping relationship. When the track connection domain ends, the scan ends;

[0122] (2) Using the mapping relationship obtained in step (1), take the midpoints of the corresponding left and right boundaries, find the position of the midpoints, and save them in an array.

[0123] For balance control, the most common are two categories. One is PID, which adds the output of the speed loop and the angle loop and transmits them to PWM to control the rotation of the motor. Many PID angle loops are 5ms and the speed loop is 10ms. The other is cascade PID, with an inner loop (angular velocity loop) of 1ms. The middle loop (angle loop) is 5ms and the outer loop (speed loop) is 10ms. Because the inner loop is 1ms, a faster balance response can be achieved.

[0124] Cascade control system is one of the effective methods to improve control quality and has been widely used in process control. The so-called cascade control is to use two controllers to work in series. The output of the outer loop controller is used as the set value of the inner loop controller. The output of the inner loop controller is used to manipulate the control valve, thereby having a better control effect on the outer loop controlled quantity. Such a control system is called a cascade system. PID cascade control means that both controllers in the cascade control are PID controllers.

[0125] In this embodiment, the control algorithm used for balancing the car model is a cascade PID algorithm, which is divided into an angular velocity loop (innermost loop), an angle loop (middle loop), and a speed loop (outermost loop).

[0126] As mentioned above, due to the poor accuracy and stability of the vehicle encoder, we use the z-axis (heading angle) angular velocity as the speed loop (outermost loop) in the outermost loop.

[0127] For the adjustment of the upright loop PID parameters, we first adjust the Kp of the angular velocity loop so that there is obvious resistance when the car body falls to one side, then adjust the Kp of the angle loop so that the car body can stand upright for 3 to 4 seconds, and finally adjust the Kp of the speed loop so that the car can stand upright stably.

[0128] Once the driving path is determined, the program will calculate the angle that the front wheels should turn through the angle control algorithm. This angle will be converted into a specific steering motor control signal to achieve precise steering of the car. Precise steering control is essential for the car to smoothly pass through curves and avoid obstacles.

[0129] In terms of speed control, the program will set the desired driving speed based on the current track conditions and the established race strategy. At this time, the speed control algorithm will dynamically adjust the output power of the motor to ensure that the car drives stably on the track and as close to the desired speed as possible. This real-time speed adjustment mechanism can not only improve the car's racing performance, but also effectively reduce the impact of track changes on driving stability.

[0130] In this embodiment, the data processor is specifically a single chip microcomputer, model: Infineon's TC264D as the core processor, which has the following characteristics:

[0131] High integration: The single-chip microcomputer integrates CPU, memory (RAM, ROM), I / O interface, timer / counter, communication interface and other functional modules on one chip, making the circuit design more concise and reducing the number of components;

[0132] Small size and low power consumption: Due to the high integration of the MCU, the size of the entire system can be reduced, which is very suitable for applications with strict requirements on size and power consumption, such as portable devices and IoT devices;

[0133] Output current: Able to provide 5A continuous output current;

[0134] Protection function: The battery voltage can be detected in real time through the A11 interface;

[0135] Good real-time performance: MCUs usually have strong real-time processing capabilities and can respond quickly to external events, making them suitable for real-time control systems such as automatic control and monitoring systems.

[0136] In this embodiment, the rotor speed at the transmission shaft of the car model is reflected in the amplitude of controlling the steering of the car model. The higher the speed, the greater the rotation amplitude. Conversely, the lower the speed, the smaller the rotation amplitude.

[0137] The vehicle model running status acquisition module and the track environment information acquisition module set in the single chip cooperate with each other to perform the following process:

[0138] To set the battery power threshold:

[0139] U1=100%, U2=80%, U3=60%, U2=50%;

[0140] Set the rotor speed at the car model's drive shaft:

[0141] s1=100%, s2=80%, s3=60%, s2=50%;

[0142] When the car model enters the straight road (because the straight road length is greater than or equal to 3 meters in the division of the above track characteristics, it is marked as a continuous straight road, which means that the length of the straight road is less than 3 meters, so there are other sections following the straight road, and the speed at this time should not be set to the lowest, so that the car model can adjust in time when entering other sections), the battery power supply is calibrated to U2=80%, and the rotor speed is calibrated to s2=60% at this time;

[0143] When the car model enters the continuous straight road, the battery power supply is calibrated to U1 = 100%, and the rotor speed is calibrated to s2 = 50%;

[0144] When the car model enters the uphill, the battery power is calibrated to U1 = 100%, and the rotor speed is calibrated to s2 = 50%;

[0145] When the car model enters the downhill, the battery power is calibrated to U1=80%, and the rotor speed is calibrated to s2=50%;

[0146] When the car model enters a curve, the battery power is calibrated to U1=60%, and the rotor speed is calibrated to s2=80%;

[0147] When the car model enters a continuous curve, the battery power supply is calibrated to U1 = 50%, and the rotor speed is calibrated to s2 = 100%;

[0148] Through the above settings, it can be ensured that the car model maintains high speed while ensuring the balance of the car body in different sections of the track.

[0149] In this embodiment, in terms of motor drive, an electric drive with DRV8701e as the core is selected, and PWM signals and a separate 2s battery are transmitted through the mainboard.

[0150] This circuit driver has two independent channels and is a dual-channel H-bridge DC motor driver board. It can realize one electric drive to control multiple motors, meeting the requirements of the two DC motors of the motorcycle to cooperate with each other. Among them, the output current can reach 3.5A, which is suitable for us to adjust the PWM within a certain range during debugging without damaging the DC motor. The input voltage range is wide and can be powered by the motherboard. It also has protection measures such as overcurrent protection, undervoltage protection, and overtemperature protection.

[0151] The MOSFET selected is TPH1R403NL, which is an N-channel power MOSFET with a withstand voltage of 40V and an extremely low on-resistance of 1.3 milliohms. Considering that the MOSFET will withstand large currents, a large area of ​​copper is laid and windows are opened to prevent the electric drive copper wire from bursting due to excessive voltage.

[0152] The power supply directly inputs 7.6v voltage through the motherboard mos, one way is input to the DRV8701e driver chip through the 1N5819HW Schottky diode, and the other way is directly supplied to the mos. At the same time, the PWM and DIR signal interfaces of the motherboard transmit control signals to the SN74HC125 buffer, which is sent to the left and right sides of the DRV driver chip to control the two DC motors in coordination with the battery DC input. In order to more conveniently observe the working status of the electric drive, an LDO circuit is added to the electric drive to control the enable LED and power LED. Because the LR8341A-T33 has good characteristics of low voltage, difference and low quiescent current, it is selected to provide a stable voltage to display the response status of the electric drive. In order to reduce the difficulty of motorcycle balance, our electric drives are all symmetrical and integrated designs.

[0153] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent vehicle control system based on an ADAS system, characterized in that: The system includes a data acquisition terminal, a data analysis terminal and a data output terminal; The data acquisition end is used to: record the track environment information in real time, perform preliminary processing on the environment information, and package the processed data into a data set and send it to the data analysis end; The data analysis end is used to obtain a simulated route model of the car model traveling along the track by analyzing and processing the data set processed by the data acquisition end; The data output end is used to: manage, analyze and calculate the simulation route model, connect the data with the key parts of the car model, integrate the track data with the car model attribute information, and receive the status information fed back by the car model when it is moving in real time, send the status information of the car model to the data analysis end, and comprehensively process the status information of the car model when it is moving and the track information through the data analysis end to ensure the rationality and feasibility of the car model's movement.

2. The ADAS-based intelligent vehicle control system according to claim 1, characterized in that: The data collection terminal includes a track environment information collection module and a vehicle model running state collection module; The track environment information acquisition module includes: a video acquisition unit; The video acquisition unit is used to collect environmental information of the track through a micro camera installed on the car model, and the environmental information includes: uphill, downhill, straight, curve, continuous straight and continuous curve.

3. The ADAS-based intelligent vehicle control system according to claim 2, characterized in that: The vehicle model running state acquisition module includes: a vehicle model transmission information acquisition unit and a vehicle model battery information acquisition unit; The vehicle model transmission information collection unit is used to collect the rotation speed information of the vehicle model transmission shaft through a rotor sensor arranged at the vehicle model transmission shaft; The vehicle model battery information acquisition unit is used to: acquire voltage output information of the vehicle model battery through a current sensor arranged at the output end of the vehicle model battery.

4. The ADAS-based intelligent vehicle control system according to claim 3, characterized in that: The track environment information collection module performs the following process: S1: Collect video information of the track through the video acquisition unit; S2: pre-processing the collected video information of the track; S3: Integrate the features of the pre-processed video information and form a corresponding set.

5. The ADAS-based intelligent vehicle control system according to claim 1, characterized in that: The step S2 is specifically: The collected video information is filtered, denoised, binarized and edge detected.

6. The ADAS-based intelligent vehicle control system according to claim 1, characterized in that: The step S3 is specifically: The track features are extracted from the preprocessed video information into: uphill set S, downhill set X, straight set Z, curve set W, continuous straight set LZ and continuous curve set LW. The division of track features follows the following rules: Set the slope threshold to ±5°. When the slope is greater than 5°, the track is marked as uphill; when the slope is less than -5°, the track is marked as downhill. Set the track's curvature angle threshold to ±5°. When the curvature angle is greater than 5° or less than -5°, the track is marked as entering a curve. When the track curvature angle is less than or equal to 5° and greater than or equal to -5°, and the slope is less than or equal to 5° and greater than or equal to -5°, the track is marked as a straight road. When the length of the straight road is greater than or equal to 3 meters, this section of the track is marked as a continuous straight road. When there are two or more consecutive bends on the track, this section of the track is marked as continuous bends.

7. The ADAS-based intelligent vehicle control system according to claim 1, characterized in that: The data analysis end includes a data processor, which is used to integrate and process the data information collected by the track environment information collection module and the vehicle model moving state collection module, and send the analysis and processing results to the data output end. The data processor performs the following process: The complete track information matrix is ​​obtained through the track search algorithm; The center line of the track is calculated through the track information matrix; Obtain the optimal route for the car model to travel along the track based on the center line of the track, and perform balance control on the car model; According to the optimal route of the track, the battery output power and drive shaft speed of different characteristic sections of the track are allocated to the car model.

8. The ADAS-based intelligent vehicle control system according to claim 7, characterized in that: The complete track information matrix is ​​obtained by the track search algorithm, specifically: Extract the connected domain of the track image and find the connected domain position corresponding to the normal track; Traverse the image, extract the boundary information and width information of the connected domain corresponding to the track, filter the boundary, and store it in an array; After the line search is completed, a complete matrix is ​​obtained that can be processed for track analysis.

9. The ADAS-based intelligent vehicle control system according to claim 8, characterized in that: The optimal route for the car model to travel along the track is obtained according to the center line of the track, specifically: After obtaining the complete track information matrix, scan from the bottom to the top of the image, map the left and right boundaries, and obtain a mapping relationship, which is recorded as: Y z and Y y , when the track connection domain ends, the scan ends; through the obtained mapping relationship, the midpoint of the corresponding left and right boundaries is taken and recorded as Y z1 and Y y1 , calculate Y z1 To Y y1 The straight-line distance is recorded as: Y L , then the position D of the midpoint is: And save the position of the midpoint in an array.

10. The ADAS-based intelligent vehicle control system according to claim 7, characterized in that: The balance control of the car model is specifically as follows: Adopt cascade PID algorithm, which is divided into angular velocity loop, angle loop and speed loop; In the outermost loop, the angular velocity of the heading angle is used as the velocity loop; For the adjustment of the upright loop PID parameters, first adjust the Kp of the angular velocity loop so that there is obvious resistance when the car body falls to one side, then adjust the Kp of the angle loop so that the car body can stand upright for 3 to 4 seconds, and finally adjust the Kp of the speed loop so that the car can stand upright stably.

Images