A vehicle posture perception and road surface preview system and method based on three-source fusion

Abstract

The application discloses a vehicle posture sensing and road pre-aiming system and method based on three-source fusion, and the system comprises a laser line array unit used for acquiring road elevation data on each wheel track; an ASIC processing unit connected with the laser line array unit and used for pre-processing the laser reflection signal; an IMU unit used for collecting angular velocity data and acceleration data of the vehicle to calculate first attitude data; an infrared horizon instrument unit used for detecting infrared radiation temperature difference of the sky and the ground in the corresponding direction to obtain second attitude data; and a fusion control unit connected with the ASIC processing unit, the IMU unit and the infrared horizon instrument unit and receiving the pre-processed road elevation data, the first attitude data and the second attitude data, and outputting control signals of each wheel. The technical scheme provided by the application realizes posture sensing and road pre-aiming improvement through three-source fusion, high-speed preprocessing and redundancy design, and improves vehicle driving safety and system reliability.

Description

Technical Field

[0001] This invention relates to the field of vehicle active suspension control technology, and in particular to a vehicle attitude perception and road surface prediction system and method based on three-source fusion. Background Technology

[0002] With the rapid development of automotive intelligence and connectivity, vehicle comfort, stability, and safety have become core research and development directions in the industry. Vehicle attitude perception is a key prerequisite for realizing active suspension control and autonomous driving assistance, while road surface preview can obtain information about road undulations in advance, providing sufficient response time for active control, thereby improving vehicle ride smoothness and handling stability.

[0003] In existing technologies, vehicle attitude perception often employs single-sensor or dual-sensor fusion schemes. For example, attitude data is collected solely through an IMU (Inertial Measurement Unit), or fused with an IMU and a unidirectional infrared horizon sensor. However, a single IMU suffers from integration drift, and attitude errors accumulate over time, leading to a decrease in attitude perception accuracy. Unidirectional infrared horizon sensors, under complex conditions (such as complex slopes or severe weather), cannot fully capture changes in vehicle attitude, making them prone to attitude misjudgment. Furthermore, existing road surface prediction systems often use a single road surface detection sensor, which has a limited detection range, weak anti-interference capabilities, and the prediction distance is not optimized based on vehicle dynamics parameters, making it difficult to adapt to prediction requirements under different driving conditions.

[0004] Furthermore, existing systems have poor fault tolerance for sensor failures. When any sensor fails, the entire attitude perception and aiming system may malfunction, affecting vehicle safety. Additionally, existing systems have high data processing latency, failing to meet the real-time requirements of active control, resulting in delayed control signal output and impacting control effectiveness. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention provides a vehicle attitude perception and road surface prediction system and method based on three-source fusion.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] On one hand, the present invention provides a vehicle attitude perception and road surface prediction system based on three-source fusion, comprising:

[0008] A laser linear array unit is set for each wheel and is used to emit laser beams to the road surface in front of the corresponding wheel and receive laser reflection signals in order to obtain road surface elevation data on the driving trajectory of each wheel.

[0009] An ASIC processing unit, connected to the laser linear array unit, is used to preprocess the laser reflection signal and transmit the preprocessed road elevation data to the fusion control unit.

[0010] The IMU unit is used to collect the vehicle's angular velocity data and acceleration data, calculate the first attitude data based on the angular velocity data and acceleration data, and transmit the first attitude data to the fusion control unit;

[0011] The infrared horizon unit is used to detect the infrared radiation temperature difference between the sky and the ground in the corresponding direction, calculate the second attitude data based on the infrared radiation temperature difference, and transmit the second attitude data to the fusion control unit.

[0012] The fusion control unit is connected to the ASIC processing unit, the IMU unit, and the infrared horizon unit, respectively, and receives the preprocessed road surface elevation data, first attitude data, and second attitude data. The fusion control unit is used to generate corresponding feedforward control commands based on the preprocessed road surface elevation data; calculate the real-time confidence scores of the first attitude data and the second attitude data; perform weighted fusion of the first attitude data and the second attitude data based on the real-time confidence scores of the first attitude data and the second attitude data to generate fused attitude data; and integrate the feedforward control commands with the fused attitude data to output control signals for each wheel.

[0013] Furthermore, the laser linear array units are independently set for each of the four wheels of the vehicle and are integrated with air curtain anti-fouling devices and automatic covers; the scanning point of the laser linear array unit is located 0.45-0.60m in front of the midpoint of the corresponding wheel's contact point, and this range constitutes the pre-aiming distance, and the scanning width is not less than 1.1 times the width of the corresponding tire tread.

[0014] Furthermore, the ASIC processing unit uses hardware phase-locked loop technology to process the laser reflection signal, and the processing delay from receiving the laser reflection signal to outputting the road surface elevation data does not exceed 1ms.

[0015] Furthermore, the IMU unit employs a six-axis MEMS inertial measurement sensor, which incorporates a three-axis gyroscope and a three-axis accelerometer to collect angular velocity data and acceleration data of the vehicle, respectively. The angular velocity data is converted into initial first attitude data using an integration algorithm built into the IMU unit, and the initial first attitude data is corrected using acceleration data to obtain first attitude data for attitude fusion.

[0016] Furthermore, the infrared horizon unit uses a long-wave infrared thermopile array sensor to detect the infrared radiation temperature difference between the sky and the ground in each direction, and calculates the pitch angle and roll angle through the infrared radiation temperature difference in each direction to obtain the second attitude data; the long-wave infrared thermopile array sensor consists of four sensors, which are installed on the front, rear, left and right sides of the vehicle top, respectively, and their field of view covers ±60° in the horizontal direction and ±15° in the vertical direction of the corresponding direction.

[0017] Furthermore, the process of calculating the real-time confidence level by the fusion control unit is as follows: the real-time confidence level of the first attitude data is calculated by integrating the angular velocity data of the IMU unit and correcting the integral drift by the acceleration data; the real-time confidence level of the second attitude data is obtained by calculating the real-time confidence level of each single direction by the infrared radiation temperature difference between the sky and the ground in each direction by the infrared horizon unit and taking the average value; the fusion control unit uses an FPGA chip, and its fusion calculation delay does not exceed 1ms, and the total end-to-end delay of the system from road surface detection to output control signal does not exceed 7ms.

[0018] Furthermore, the strategy for the fusion control unit to dynamically allocate fusion weights based on the real-time confidence of the first attitude data is as follows: a first threshold Th1 and a second threshold Th2 are preset, satisfying 0 < Th2 < Th1 ≤ 1. When the real-time confidence of the first attitude data > Th1, the second attitude data and the first attitude data are weighted and fused, and the fusion weight of the second attitude data is greater than the fusion weight of the first attitude data. When the real-time confidence of the first attitude data < Th2, the first attitude data is used as the only attitude source. When Th2 ≤ the real-time confidence of the first attitude data ≤ Th1, the fusion is weighted according to the confidence ratio, and the first attitude data and the second attitude data are weighted and fused to obtain the fused attitude data.

[0019] Furthermore, it also includes an ultrasonic sensor unit, which is installed on the lower edge of the bumper in front of the corresponding wheel, close to the front of the laser linear array sensor of the laser linear array unit, for emitting ultrasonic waves and receiving echoes to obtain road surface elevation envelope data; the difference between the road surface elevation envelope data of the ultrasonic sensor unit and the road surface elevation data of the laser linear array unit is compared and verified with a preset threshold. When the difference between the two exceeds the preset threshold, a self-test program is triggered to identify the failed sensor and switch the data source.

[0020] Furthermore, the failure modes of the laser linear array unit, the IMU unit, and the infrared horizon unit satisfy an orthogonal relationship. When any one or two of the laser linear array unit, the IMU unit, and the infrared horizon unit fail, the system has redundancy and fault tolerance capabilities, and maintains normal operation through the remaining sensors.

[0021] On the other hand, the present invention provides a vehicle attitude perception and road surface prediction method based on three-source fusion, applied to the system, including the following steps:

[0022] Step S1: Obtain road surface elevation data on the driving trajectory of each wheel by using the laser linear array unit set for each wheel;

[0023] Step S2: The laser reflection signal is preprocessed by the ASIC processing unit, and the preprocessed road surface elevation data of each wheel is output.

[0024] Step S3: Collect the vehicle's angular velocity and acceleration data through an IMU unit located at the vehicle's center of gravity, and output the first attitude data;

[0025] Step S4: Detect the infrared radiation temperature difference between the sky and the ground using an infrared horizon sensor unit installed on the top of the vehicle, and output the second attitude data;

[0026] Step S5: Perform the following operations through the fusion control unit:

[0027] Generate corresponding feedforward control commands based on the pre-processed road surface elevation data of each wheel;

[0028] Calculate the real-time confidence scores of the first attitude data and the second attitude data, respectively.

[0029] The first attitude data and the second attitude data are weighted and fused based on the real-time confidence scores of the first attitude data and the second attitude data to generate fused attitude data.

[0030] The feedforward control command and the fused attitude data are integrated in a coordinated manner to output control signals for each wheel.

[0031] The present invention provides a vehicle attitude perception and road surface prediction system and method based on three-source fusion, which has the following advantages:

[0032] Through a unique feedforward control and error correction architecture, employing a three-source data fusion mode of laser linear array unit, IMU unit, and infrared horizon unit, and utilizing high-speed preprocessing, dynamic weighted fusion algorithm, and multiple protection and redundancy designs of the ASIC processing unit, this approach effectively solves the passive defects of traditional solutions such as "guessing lag" and "error first, correction later." This achieves a comprehensive improvement in vehicle attitude perception and road surface prediction performance, significantly enhancing vehicle driving comfort, stability, and safety. It is suitable for various engineering application scenarios. Specifically, through independent four-wheel settings and precise parameter design of the laser linear array unit, and high-speed signal processing using the ASIC processing unit's built-in hardware phase-locked loop, road surface elevation information is acquired in advance and environmental noise interference is eliminated, improving road surface prediction accuracy and real-time performance. This solution addresses the issues of turn-by-turn aiming failure and control overshoot under high-frequency interference in traditional solutions. Through dynamic weighted fusion of three data sources, it compensates for the shortcomings of single or dual-sensor fusion, corrects IMU drift errors, and achieves attitude perception accuracy better than 0.1°. It also possesses redundancy and fault tolerance capabilities, preventing system paralysis due to single sensor failure and improving system reliability. Through hardware integration optimization and time synchronization design of each unit, the total end-to-end latency of the system does not exceed 7ms, meeting the real-time requirements of active suspension control and ensuring complete engineering implementation. Multiple protections for the laser linear array unit, four-way cross-verification of the infrared horizon sensor, and optional ultrasonic redundancy design enhance the system's anti-interference capability and stability, extend sensor lifespan, reduce maintenance costs, and adapt to various complex working conditions and vehicle models. Attached Figure Description

[0033] Figure 1 This is a structural block diagram of the vehicle attitude perception and road surface prediction system based on three-source fusion in an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram showing the installation positions of the laser linear array unit 1 and the infrared horizon unit 4 in an embodiment of the present invention.

[0035] Figure 3 For the present invention Figure 2 Enlarged side view diagram at point A in the middle;

[0036] Figure 4 For the present invention Figure 2 Enlarged top view of point A in the middle;

[0037] Figure 5 This is a scanning display diagram of the infrared horizon unit 4 in an embodiment of the present invention;

[0038] Figure 6 This is a flowchart of the method for a vehicle attitude perception and road surface prediction system based on three-source fusion in an embodiment of the present invention.

[0039] Figure descriptions: 1. Laser linear array unit; 2. ASIC processing unit; 3. IMU unit; 4. Infrared horizon unit; 5. Fusion control unit; 6. Ultrasonic sensor unit; 7. Air curtain anti-fouling device; 8. Automatic cover. Detailed Implementation

[0040] To make the technical problems, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0041] In the description of this invention, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc., are based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0042] In the description of this invention, unless otherwise explicitly defined, terms such as "setting," "installing," and "connecting" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this invention based on the specific content of the technical solution.

[0043] like Figures 1-5 As shown, this embodiment provides a vehicle attitude perception and road surface prediction system based on three-source fusion, including a laser linear array unit 1, an ASIC processing unit 2, an IMU unit 3, an infrared horizon sensor unit 4, a fusion control unit 5, and an ultrasonic sensor unit 6. The fusion control unit 5 is electrically connected to the ASIC processing unit 2, the IMU unit 3, the infrared horizon sensor unit 4, and the ultrasonic sensor unit 6, respectively. Through data interaction and fusion and high-speed preprocessing of the ASIC processing unit 2, the IMU unit 3, and the infrared horizon sensor unit 4, real-time and reliable vehicle attitude perception and road surface prediction are achieved.

[0044] like Figure 2 As shown, the laser linear array unit 1 uses a high-resolution laser linear array sensor. There are four laser linear array sensors, which are respectively set on the inner side of the bumper in front of the left front wheel, right front wheel, left rear wheel and right rear wheel of the corresponding vehicle, so as to realize independent road surface pre-aiming of the driving trajectory of each wheel of the vehicle and ensure the accuracy of road surface detection.

[0045] The laser linear array sensor emits a continuous laser beam, and the receiver receives the laser reflection signal from the road surface. Through signal analysis, it obtains the road elevation data along the corresponding wheel's trajectory. To ensure the aiming effect is adapted to the vehicle's driving conditions, the scanning point of each laser linear array sensor is located at a aiming distance in front of the midpoint of the corresponding wheel's contact point. The aiming distance is set to [value missing]. For the maximum pre-aiming distance and minimum aiming distance Perform calculations and set the maximum aiming distance. ,in, The maximum steering angle is determined by the vehicle's factory parameters and ranges from 0.3 to 0.5 rad. This refers to the tire width, ranging from 0.195 to 0.285 meters. A safety factor (range 1.1-1.3) is used to mitigate lateral errors during vehicle steering; a minimum aiming distance is set. ,in, This represents the total end-to-end latency of the system. ≤0.007s, This refers to the maximum speed of passenger vehicles, ranging from 30-50 m / s. To ensure sufficient control response time, a safety factor (range 1.5-2.0) is set.

[0046] It is understandable that the feasible solution interval for the pre-aiming distance D is... ≤ ≤ In this embodiment, 0.45m ≤ ≤0.60m.

[0047] Meanwhile, the scanning width of the laser linear array sensor in the laser linear array unit 1 Not less than 1.1 times the corresponding tire tread width, through the tire width Maximum steering angle of the vehicle and aiming distance For the scan width Perform calculations and set This setting ensures that the laser beam fully covers the tire's trajectory, avoiding the loss of elevation data due to road surface undulations, thereby improving the comprehensiveness of road surface prediction.

[0048] like Figure 3 , Figure 4As shown, in a preferred embodiment, the laser linear array unit 1 integrates an air curtain anti-fouling device 7 and an automatic cover 8. Specifically, the air curtain anti-fouling device 7 uses a blower to actively draw in air, forming an air film barrier in front of the laser window of the laser linear array unit 1 to block the adhesion of pollutants such as dust and mud. The automatic cover 8 is electromagnetically controlled; when the vehicle comes to a complete stop (i.e., vehicle speed = 0 km / h), the automatic cover 8 automatically covers the laser window of the laser linear array unit 1 for a duration of ≥3 seconds, preventing static contamination and physical damage.

[0049] like Figure 3 , Figure 4 As shown, the ASIC processing unit 2 is designed using a dedicated integrated circuit and integrated inside the laser linear array unit 1. The ASIC processing unit 2 has four built-in hardware phase-locked loops, which are connected one-to-one with the laser linear array unit 1. They are used to preprocess the raw laser reflection signal received by the laser linear array unit 1 to reduce signal transmission delay and interference. The specific processing flow is as follows:

[0050] 1) Hardware phase-locked loop (PLL) technology is used to filter environmental noise in the laser reflection signal, such as sunlight interference and ground reflection clutter. The center frequency of the PLL is consistent with the emission frequency of the laser linear array unit 1. In this embodiment, the laser emission frequency is 100kHz and the bandwidth is set to 1kHz to ensure effective noise filtering while avoiding the loss of useful signals.

[0051] 2) By detecting the time difference between the laser emission signal and the laser reflection signal The laser propagation speed c is related to the distance between the laser linear array unit 1 and the road surface detection point. ,set up ,in, .

[0052] 3) The distance Convert to road surface elevation data ,set up ,Right now ,in, The installation angle of the laser linear array unit 1 is the angle between the laser beam and the horizontal plane. In this embodiment, it is preferably... =30°, which can be adjusted according to the installation location.

[0053] 4) Transfer the preprocessed road surface elevation data The road surface elevation data is output to the fusion control unit 5. The resolution should be no less than 0.1 mm to ensure the accuracy of the road surface elevation data. Its real-time performance and accuracy.

[0054] In this embodiment, the IMU unit 3 employs a six-axis MEMS inertial measurement sensor. For example... Figure 2 , Figure 5 As shown, it is installed at the vehicle's center of gravity. This location reduces interference from localized vehicle vibrations on the sensor's data acquisition, ensuring the accuracy of the collected angular velocity data. and acceleration data Used to reflect changes in the overall attitude of the vehicle.

[0055] The IMU unit 3 integrates a three-axis gyroscope and a three-axis accelerometer. The three-axis gyroscope has a range of ±2000° / s and a zero-bias stability of ±1° / h, and is used to collect the vehicle's angular velocity data. ,set up ,in, For roll rate, For pitch angular velocity, The angular velocity is the heading angle; the triaxial accelerometer has a range of ±16g and a zero-bias stability of ±50μg, and is used to collect triaxial acceleration data of the vehicle. ,set up ,in, For longitudinal acceleration, For lateral acceleration, This is the vertical acceleration.

[0056] The data update frequency of the IMU unit 3 is 1000Hz. The angular velocity data is processed using the integration algorithm built into the IMU unit 3. Perform time integration to obtain the initial attitude data. , , ), and then based on the acceleration data After correcting the integral drift error, the first attitude data is obtained. Specifically, through the angular velocity data Data collection time interval Calculate the initial attitude data and set... , , ,in, Let n be the time for integrating the angular velocity, and n be the number of integration iterations. , =1 / 1000=0.001s, , , The roll angular velocity is the value collected in the i-th acquisition. Pitch angular velocity angular velocity of heading .

[0057] Because the triaxial gyroscope exhibits drift characteristics, the acceleration data... The difference between the theoretical gravitational component and the integral drift error is used to correct the integral drift error. The correction formula is set as follows: , , ,in, The acceleration correction factor ranges from 0.01 to 0.05, where g is the gravitational acceleration (g = 9.8 m / s²), yielding the first attitude data. Simultaneously, the first attitude data ( The data is transmitted to the fusion control unit 5 for attitude fusion.

[0058] The infrared horizon unit 4 employs a long-wave infrared thermopile array sensor, and the number of long-wave infrared thermopile array sensors is four. For example... Figure 5 As shown, the devices are installed in four directions on the top of the vehicle: front, rear, left, and right. Specifically, they are installed at the front of the roof above the windshield, at the rear of the roof above the trunk lid, at the left side of the roof above the center line, and at the right side of the roof above the center line. This four-directional installation layout can comprehensively capture the vehicle's attitude changes in roll and pitch directions, avoiding the limitations of single-direction detection.

[0059] The long-wave infrared thermopile array sensor has a wavelength of 8-14μm. Infrared radiation in this wavelength range can effectively penetrate light fog and dust, exhibiting strong resistance to environmental interference. Its field of view covers ±60° horizontally and ±15° vertically in the corresponding direction. The long-wave infrared thermopile array sensor is used to detect the infrared radiation temperature of the sky and ground in the corresponding direction in real time and outputs a temperature distribution matrix with a resolution of no less than 16×4 pixels. Let the 16×4 pixel infrared temperature distribution matrix be... :

[0060]

[0061] The first two columns show the temperature measurements of the sky area, and the last two columns show the temperature measurements of the ground area, along with the infrared radiation temperature difference in each direction. (i = 1, 2, 3, 4, representing the vehicle's forward, backward, left, and right directions, respectively), thereby achieving attitude perception and environmental compensation.

[0062] Its working principle is as follows: the long-wave infrared thermopile array sensor detects the infrared radiation temperature of the sky and the ground in the corresponding direction in real time, and outputs a temperature distribution matrix with a resolution of not less than 16×4 pixels. ; Calculate the infrared radiation temperature difference between the sky and the ground in each direction based on the temperature distribution matrix. When the vehicle's attitude changes, the ratio of sky to ground in the infrared horizon sensor's field of view changes, causing a change in the difference in radiation intensity between the upper and lower halves. This difference is approximately linearly related to the attitude angle. This difference can be analyzed... Output second attitude data, which includes pitch angle. and roll angle The specific calculation process is as follows:

[0063] The infrared radiation temperature difference between the sky and the ground is measured forward. and the infrared radiation temperature difference between the sky and the ground behind. pitch angle Perform calculations and set the pitch angle. ,in, pitch angle The conversion factor, ranging from 0.01 to 0.03 rad / ℃, is determined by the pitch angle. The vehicle status is determined, and the result is output, including:

[0064] If pitch angle If the value is positive, the vehicle is determined to be going uphill;

[0065] If pitch angle If the value is negative, the vehicle is determined to be on a downhill slope.

[0066] Through the temperature difference between the sky and the ground to the left Temperature difference between the sky and the ground to the right For roll angle Perform calculations and set ,in, This is the roll angle conversion factor, with a value ranging from 0.01 to 0.03 rad / ℃.

[0067] If roll angle If the value is positive, then the vehicle is determined to be tilted to the left.

[0068] If roll angle If the result is negative, the vehicle is determined to be tilting to the right.

[0069] When the vehicle is traveling in complex road conditions, the elevation angle is calculated by detecting the infrared radiation temperature difference between the sky and the ground using long-wave infrared thermopile array sensors in four directions. and roll angle Data fusion for second pose data ( , ) Perform calculations and set , ,in, , This is a cross-correction factor, ranging from 0.1 to 0.3, used to correct for roll angle. With pitch angle The mutual influence between them ensures the second attitude data under complex road conditions. , The accuracy of the infrared horizon sensor unit 4 is as follows: The infrared horizon sensor unit 4 outputs second attitude data (…). , And transmit it to the fusion control unit 5.

[0070] The fusion control unit 5 employs an FPGA chip, integrating a data receiving module, a confidence calculation module, a weight allocation module, an attitude fusion module, and a control command generation module to achieve data fusion, control command generation, and output. In this embodiment, the time delay from receiving data from each unit to outputting fused attitude data in the fusion control unit 5 does not exceed 1ms, and the total end-to-end delay from road surface detection to outputting the control signal does not exceed 7ms, meeting the real-time requirements of active suspension control. Specifically:

[0071] The fusion control unit 5 receives the road surface elevation data for each wheel output by the ASIC processing unit 2. By analyzing the road surface elevation variation patterns, such as pothole depth and undulation frequency, and based on the vehicle's active suspension dynamic parameters, such as suspension travel, stiffness coefficient, and damping coefficient, feedforward control commands are generated for each wheel. This allows for advance control of the suspension's extension and contraction, offsetting the impact of road surface undulations on vehicle ride comfort and achieving active anti-sight control.

[0072] The fusion control unit 5 calculates the real-time confidence scores of the first attitude data and the second attitude data, respectively. The real-time confidence score ranges from (0,1], as follows:

[0073] 1) Integration time of the angular velocity data from the IMU unit 3 For the first attitude data ( Real-time confidence level Perform calculations and set ,in, This is the drift coefficient, with a value ranging from 0.001 to 0.01 s. -1 , The integral time of the angular velocity ranges from 0 to 60 seconds, and is based on the real-time confidence level of the first attitude data. The accuracy of the first attitude data is judged, and the result is output, wherein:

[0074] When 0.2 < When the value is ≤1, the attitude data accuracy of the IMU unit 3 is determined to be normal, and the first attitude data output at this time is ( )reliable.

[0075] when If the accuracy is ≤0.2, the attitude data of IMU unit 3 is determined to be abnormal, and the first attitude data output at this time ( The data collected by the triaxial accelerometer is unreliable, has a large drift error, and the acceleration data is... Correct the integration results.

[0076] Among them, the angular velocity data of the IMU unit 3 Integral time The longer the duration, the higher the real-time confidence level of the first attitude data. The lower the value, the better the acceleration data. The more significant the correction to the integration result, the higher the real-time confidence level of the first attitude data. The higher.

[0077] 2) Temperature difference due to infrared radiation from the sky and the ground in all directions. Real-time confidence level in each unidirectional direction Perform calculations and set ,in, The infrared radiation temperature difference between the sky and the ground detected by the i-th long-wave infrared thermopile array sensor. This is an infrared confidence calibration constant (with a value range of 0.1-0.2). The fusion control unit 5 uses the real-time confidence levels in the aforementioned front, rear, left, and right directions. Real-time confidence level of the second attitude data Perform calculations and set the real-time confidence level of the second attitude data. , The value range is (0,1], where the greater the temperature difference between the sky and the ground, the higher the signal-to-noise ratio and the higher the confidence of the second attitude data.

[0078] 3) In this embodiment, the real-time confidence level based on the first attitude data Real-time confidence of second attitude data The real-time confidence level of the first attitude data is determined by setting a first threshold Th1 and a second threshold Th2 (in this embodiment, 0 < Th2 < Th1 ≤ 1, the first threshold Th1 is 0.7, and the second threshold Th2 is 0.3). The fused pose data is compared with the first threshold Th1 and the second threshold Th2, and the pose source is determined based on the comparison result. The fusion weights are then allocated based on the determination result to generate fused pose data. , , ),in:

[0079] When 0.7 < When the value is less than 1, the second attitude data is used as the attitude source to correct the drift error of the first attitude data. The fusion weight is then calculated as follows: , =1- The fused attitude angle is , , heading angle Provided only by IMU unit 3.

[0080] When 0.3≤ When the value is ≤0.7, the first attitude data and the second attitude data are weighted and fused to form the attitude source. A dynamic weighted fusion algorithm is then used to process the first attitude data output by the IMU unit 3. The infrared horizon unit 4 outputs the second attitude data. , Weighted fusion is performed, and the fusion weight is... , = The fused attitude angle is , , And generate fused pose data ( , , Based on real-time confidence and real-time confidence Real-time confidence level With real-time confidence The comparison is performed, and the fusion result is judged based on the comparison result. The result is then output, including:

[0081] like If so, the fusion result is determined to be the solution value of the infrared horizon unit 4;

[0082] like ≤ If the result is obtained, the fusion result is determined to be the integral value of the IMU unit 3.

[0083] When 0 < When the value is less than 0.3, the first pose data is determined to be the only pose source, and the fusion weight is then calculated. =0, =1, the fused attitude angle is , , To avoid fluctuations in the fused attitude data caused by sudden changes in weights, the switching of fused weights adopts a gradual change, that is, the rate of change of fused weights does not exceed 0.1 / s.

[0084] It is understandable that in this embodiment, when the vehicle enters a special environment, such as a tunnel, a tree-lined road, heavy rain, or snow, the infrared radiation temperature difference between the sky and the ground decreases or even disappears, thereby affecting the real-time confidence level of the second attitude data. The value is reduced to 0-0.3. At this point, the fusion control unit 5 automatically switches to using the first attitude data from the IMU unit 3 as the sole attitude source, ensuring the system operates normally even in scenarios where the infrared horizon unit 4 fails. Simultaneously, the system uses acceleration data... IMU integral drift is corrected to maintain attitude perception accuracy. The real-time confidence level of the second attitude data is also assessed when the vehicle leaves a special environment. When the system is restored to 0.3-1, it automatically returns to the three-source fusion mode, achieving seamless switching.

[0085] Optionally, in another embodiment, to further improve the reliability of road surface pre-aiming, the system is supplemented with four ultrasonic sensor units 6. Figure 2 , Figure 5 As shown, the sensors are respectively installed on the lower edge of the bumper in front of the corresponding wheels, close to the front of the laser linear array sensor of the laser linear array unit 1, and partially overlap with the detection range of the laser linear array unit 1. They are used to emit ultrasonic pulses. In this embodiment, the emission frequency is preferably 40kHz. This embodiment does not limit the emission frequency. Those skilled in the art can adjust it according to the actual situation, as long as it meets the requirements for road surface elevation envelope data. This transmission frequency is sufficient for the required detection needs, balancing detection distance and anti-attenuation performance, and is suitable for road surface elevation envelope data within the range of 1-3m. Detection. And receive the echo to obtain the road surface elevation envelope data. and the road surface elevation envelope data Perform calculations and set ,in, The speed of sound is 340 m / s at room temperature. The time it takes for the ultrasound to travel is 1000 rpm. In this embodiment, the installation angle of the ultrasonic sensor is... =30°, which can be adjusted according to the actual situation.

[0086] The fusion control unit 5 employs a Kalman filter. Preprocessing is performed on the preprocessed road surface elevation envelope data. and the road elevation data preprocessed by the laser linear array unit 1 Difference value The comparison and verification are performed against a preset threshold, and the verification formula is set as follows: The preset threshold is 0.01. This embodiment does not limit the specific value of the preset threshold. Those skilled in the art can adjust it according to the actual situation, as long as it meets the difference value. The comparison and verification are sufficient, and the sensor states of the laser linear array unit 1 and the ultrasonic sensor unit 6 are determined based on the comparison and verification results. The results are then output, wherein:

[0087] when When the elevation is ≤0.01m, it is determined that the sensor states of both the laser linear array unit 1 and the ultrasonic sensor unit 6 are normal, and the output is the road surface elevation data preprocessed by the laser linear array unit 1. .

[0088] when If the elevation is greater than 0.01m, then one of the sensors in the laser linear array unit 1 and the ultrasonic sensor unit 6 is determined to be abnormal. At this time, a self-test program is triggered, and the preprocessed road surface elevation data is output. and preprocessed road surface elevation envelope data average Output, that is .

[0089] Preferably, if one of the sensors detects abnormal data five times in a row, the sensor is determined to be faulty, and the data source is switched to another sensor, thereby improving the redundancy and reliability of the system.

[0090] Specifically, the failure modes of the laser linear array unit 1, IMU unit 3, and infrared horizon unit 4 are orthogonal, meaning their failure causes are independent of each other. When any one or two of the sensors in the laser linear array unit 1, IMU unit 3, and infrared horizon unit 4 fail, the system continues to operate normally through the remaining sensors. The specific fault-tolerance strategy is as follows:

[0091] For the laser linear array unit 1, if a portion of the laser linear array sensors in laser linear array unit 1 fails, the road surface elevation data from the remaining laser linear array sensors in laser linear array unit 1 shall be used. Based on the road surface elevation envelope data of the ultrasonic sensor unit 6 Generate feedforward control commands; if all laser linear array units 1 fail, use the road surface elevation envelope data from the ultrasonic sensor unit 6. Replacement, ensuring normal feedforward control.

[0092] For the IMU unit 3, the second attitude data output by the infrared horizon unit 4 is used as the sole attitude source, based on the road surface elevation data. Generate control signals to ensure proper attitude perception and control.

[0093] For the infrared horizon sensor unit 4, the first attitude data output by the IMU unit 3 is used as the sole attitude source, based on the road surface elevation data. Generate control signals, and simultaneously use the acceleration data from the IMU unit 3. Enhance drift error correction to ensure attitude perception accuracy.

[0094] If both sensors fail, specifically the laser linear array sensor of the laser linear array unit 1 and the MEMS inertial measurement sensor of the IMU unit 3, the attitude data from the infrared horizon unit 4 and the road surface elevation envelope data from the ultrasonic sensor unit 6 will be used. If the laser linear array sensor of the laser linear array unit 1 and the long-wave infrared thermopile array sensor of the infrared horizon unit 4 fail, the attitude data of the IMU unit 3 and the road surface elevation envelope data of the ultrasonic sensor unit 6 will be used. If the MEMS inertial measurement sensor of IMU unit 3 and the long-wave infrared thermopile array sensor of infrared horizon unit 4 fail, the road surface elevation data of laser linear array unit 1 shall be used. Both can maintain basic attitude awareness and aiming capabilities.

[0095] In this embodiment, the fusion control unit 5 integrates the generated feedforward control commands for each wheel with the final fused attitude data of the vehicle, and after optimization by a PID control algorithm, outputs the final control signals for each wheel. (r=1,2,3,4, corresponding to the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively), the data is transmitted to the active suspension actuator of the fusion control unit 5 to control the real-time adjustment of the suspension's extension and retraction, achieving dual control of road surface prediction and attitude compensation. Simultaneously, the fused attitude data and the feedforward control commands for each wheel are synchronously output to the autonomous driving domain controller of the fusion control unit 5, providing attitude and road surface prediction data support for autonomous driving assisted decision-making. The collaborative integration of the feedforward control commands and the fused attitude data adopts a quantitative collaborative calculation method, achieving precise integration based on vehicle dynamics parameters. The specific formula and process are as follows:

[0096] In this embodiment, the feedforward control command is set. ,in, This is the suspension stiffness coefficient, ranging from 15000-25000 N / m, used to counteract the impact of road undulations on the vehicle's attitude. The fused attitude data ( , , The feedforward control command is used to correct vehicle body attitude deviations. and fused pose data ( , , By using a vehicle suspension dynamics model for quantitative coordination, control deviations from single commands are eliminated, achieving a dual control effect of anti-aiming compensation and real-time attitude correction. Attitude correction coefficients for each wheel are calculated based on fused attitude data. , The feedforward control command Co-calculation with attitude correction coefficients yields the intermediate quantity for collaborative integration. The final control signal is then output after PID optimization. The specific calculation formula is as follows:

[0097] Attitude correction coefficient calculation: , ,in, To correct the pitch angle, the directional coefficient is set to 1 for the left front wheel and right front wheel, and -1 for the left rear wheel and right rear wheel. To correct the roll angle, the left front and left rear wheels are set to 1, and the right front and right rear wheels are set to -1. , To integrate pitch, roll, and yaw angles from attitude data Used for autonomous driving domain controller decision-making, but not for collaborative integration calculations. The collaborative integration intermediate quantity... ,in, The attitude compensation coefficient has a value range of 50-100 N / rad, and 80 N / rad is preferred in this embodiment to enhance the compensation effect of attitude deviation and avoid vehicle attitude fluctuation.

[0098] PID-optimized control signal ,in, , , The optimized PID parameters are set to the following values: =15、 =0.3、 =3, To coordinate and integrate the integral terms of intermediate quantities, which are used to eliminate steady-state control errors, To coordinate and integrate the differential terms of intermediate quantities, thereby accelerating the control response speed and suppressing control overshoot.

[0099] Example 2

[0100] This embodiment provides a method for vehicle attitude perception and road surface prediction based on three-source fusion, applied to the system described in Embodiment 1, and specifically includes the following steps, as follows: Figure 6 As shown, the detailed implementation process of each step is as follows:

[0101] Step S1: The laser linear array unit 1, corresponding to each wheel, simultaneously emits laser beams towards the road surface in front of the corresponding wheel. After reflection from the road surface, the laser beams are received by the receiving end of the laser linear array unit 1, acquiring the original laser reflection signal. The scanning point of the laser linear array unit 1 is located 0.45-0.60m in front of the midpoint of the corresponding wheel's contact point. The scanning width of the laser linear array sensor of the laser linear array unit 1... The elevation data should be no less than 1.1 times the corresponding tire tread width to ensure that the collected road surface elevation data fully covers the wheel's trajectory.

[0102] In step S2, the ASIC processing unit 2 reads the original laser reflection signal output by the corresponding laser linear array unit 1, uses hardware phase-locked loop technology to filter environmental noise, calculates the time difference Δt between laser emission and laser reflection, converts it to distance d based on the laser propagation speed c, and then calculates the distance d according to the installation angle of the laser linear array unit 1. Converted into road surface elevation data After preprocessing, the ASIC processing unit 2 will process the road surface elevation data. The output is sent to the fusion control unit 5, and the processing delay of this step does not exceed 1ms. If the system adds an ultrasonic sensor unit 6, the road surface elevation envelope data of the ultrasonic sensor unit 6 is collected synchronously. The data is then transmitted to the fusion control unit 5, which uses a Kalman filter to... Preprocessing is performed, and the output of the laser linear array unit 1 is compared with that of the laser linear array unit 1. The comparison and verification were carried out to determine the final road surface elevation data.

[0103] Step S3: The IMU unit 3, installed at the vehicle's center of gravity, collects the vehicle's angular velocity data in real time. and triaxial acceleration data The data update frequency is 1000Hz; the IMU unit 3 uses a built-in integration algorithm to process the angular velocity data. Perform time integration to obtain the initial attitude data. , , ), and then based on the acceleration data After correcting the integral drift error, the first attitude data is obtained. ), and the first attitude data ( The output is sent to the fusion control unit 5.

[0104] Step S4: Long-wave infrared thermopile array sensors installed on the four sides of the vehicle's roof detect the infrared radiation temperature of the sky and ground in real time in the corresponding directions, and output a temperature distribution matrix with a resolution of no less than 16×4 pixels; based on the temperature distribution matrix, the infrared radiation temperature difference between the sky and ground in each direction is calculated. , , and Based on the infrared radiation temperature difference between the sky and the ground in various directions, and according to the vehicle's road conditions, such as uphill, downhill, side tilt, and combined road conditions, the pitch angle is adjusted. and roll angle The second attitude data is obtained by calculation. , ), and the second attitude data ( , The output is sent to the fusion control unit 5.

[0105] Step S5: The fusion control unit 5 receives the road surface elevation data output by the ASIC processing unit 2. The first attitude data output by the IMU unit 3 ( ), the second attitude data output by infrared horizon unit 4 ( , ), perform the following operations:

[0106] S51, the fusion control unit 5 parses the road surface elevation data. It identifies road conditions and generates feedforward control commands for each wheel based on the dynamic parameters of the active suspension.

[0107] S52, the fusion control unit 5 uses the formula Calculate the real-time confidence level of the first attitude data ,in =0.005s -1 , Integrating the current angular velocity over time; using the infrared radiation temperature difference between the sky and the ground. Real-time confidence level in each unidirectional direction Calculations are performed using (i = 1, 2, 3, 4, representing the front, rear, left, and right sides of the vehicle, respectively), and the following settings are defined. Then, by taking the real-time confidence level of each unidirectional direction... The real-time confidence level of the second attitude data is calculated by averaging the values. The real-time confidence level of the second attitude data .

[0108] S53, the fusion control unit 5, based on preset thresholds Th1=0.7 and Th2=0.3, and based on the real-time confidence level of the first attitude data... Real-time confidence of second attitude data The values ​​are dynamically allocated to determine the fusion weights, and the first and second pose data are weighted and fused according to the corresponding weights to generate the optimal fused pose data. , , ).

[0109] S54, the feedforward control command and the fused attitude data are integrated collaboratively, and then processed by a PID algorithm ( =15、 =0.3、 =3) After optimization, the final control signals of each wheel are transmitted to the active suspension actuator of the fusion control unit 5, and the suspension actuator is controlled to adjust the extension and retraction in real time. The fusion calculation delay of this step does not exceed 1ms. The optimization process of the system PID algorithm is deeply adapted to the system operating conditions. When a sensor fails, the fusion control unit 5 will synchronously adjust the input data source of the PID algorithm to ensure that the optimized PID algorithm is adapted to the current system operating state and maintains the control accuracy and stability.

[0110] The total end-to-end delay of the system from step S1 to step S5 does not exceed 7ms, ensuring the real-time performance of the control signals.

[0111] The above is a description of a vehicle attitude perception and road surface prediction system and method based on three-source fusion, which is used to help understand the present invention. However, the implementation of the present invention is not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the principle of the present invention should be considered as equivalent substitutions and are included within the protection scope of the present invention.

Claims

1. A vehicle attitude perception and road surface prediction system based on three-source fusion, characterized in that, include: A laser linear array unit is set for each wheel and is used to emit laser beams to the road surface in front of the corresponding wheel and receive laser reflection signals in order to obtain road surface elevation data on the driving trajectory of each wheel. An ASIC processing unit, connected to the laser linear array unit, is used to preprocess the laser reflection signal and transmit the preprocessed road elevation data to the fusion control unit. The IMU unit is used to collect the vehicle's angular velocity data and acceleration data, calculate the first attitude data based on the angular velocity data and acceleration data, and transmit the first attitude data to the fusion control unit; The infrared horizon unit is used to detect the infrared radiation temperature difference between the sky and the ground in the corresponding direction, calculate the second attitude data based on the infrared radiation temperature difference, and transmit the second attitude data to the fusion control unit. The fusion control unit is connected to the ASIC processing unit, the IMU unit, and the infrared horizon unit, respectively, and receives the preprocessed road surface elevation data, the first attitude data, and the second attitude data. The fusion control unit is used to generate corresponding feedforward control commands based on the preprocessed road surface elevation data; and to calculate the real-time confidence of the first attitude data and the real-time confidence of the second attitude data, respectively. The first attitude data and the second attitude data are weighted and fused based on the real-time confidence scores of the first attitude data and the second attitude data to generate fused attitude data; the feedforward control command is then integrated with the fused attitude data to output control signals for each wheel.

2. The system according to claim 1, characterized in that, The laser linear array units are independently set for each of the four wheels of the vehicle and are integrated with air curtain anti-fouling devices and automatic covers. The scanning point of the laser linear array unit is located 0.45-0.60m in front of the midpoint of the corresponding wheel's contact point. This range constitutes the pre-aiming distance, and the scanning width is not less than 1.1 times the width of the corresponding tire tread.

3. The system according to claim 1, characterized in that, The ASIC processing unit uses hardware phase-locked loop technology to process the laser reflection signal, and the processing delay from receiving the laser reflection signal to outputting the road surface elevation data does not exceed 1ms.

4. The system according to claim 1, characterized in that, The IMU unit employs a six-axis MEMS inertial measurement sensor, which incorporates a three-axis gyroscope and a three-axis accelerometer to collect angular velocity and acceleration data of the vehicle, respectively. The angular velocity data is converted into initial first attitude data using an integration algorithm built into the IMU unit. The initial first attitude data is then corrected using acceleration data to obtain first attitude data, which is used for attitude fusion.

5. The system according to claim 1, characterized in that, The infrared horizon unit uses a long-wave infrared thermopile array sensor to detect the infrared radiation temperature difference between the sky and the ground in each direction. The pitch angle and roll angle are calculated based on the infrared radiation temperature difference in each direction to obtain the second attitude data. There are four long-wave infrared thermopile array sensors, which are installed on the front, rear, left and right sides of the vehicle roof. Their field of view covers ±60° in the horizontal direction and ±15° in the vertical direction of the corresponding direction.

6. The system according to claim 1, characterized in that, The process of calculating the real-time confidence level of the fusion control unit is as follows: the real-time confidence level of the first attitude data is calculated by integrating the angular velocity data of the IMU unit and correcting the integral drift by the acceleration data; the real-time confidence level of the second attitude data is obtained by calculating the real-time confidence level of each single direction by the infrared radiation temperature difference between the sky and the ground in each direction of the infrared horizon unit and taking the average value; the fusion control unit uses an FPGA chip, and its fusion calculation delay does not exceed 1ms, and the end-to-end total delay of the system from road surface detection to output control signal does not exceed 7ms.

7. The system according to claim 6, characterized in that, The strategy of the fusion control unit to dynamically allocate fusion weights based on the real-time confidence of the first attitude data is as follows: a first threshold Th1 and a second threshold Th2 are preset, and 0 < Th2 < Th1 ≤ 1 is satisfied. When the real-time confidence of the first attitude data is greater than Th1, the second attitude data and the first attitude data are weighted and fused, and the fusion weight of the second attitude data is greater than the fusion weight of the first attitude data. When the real-time confidence of the first attitude data is less than Th2, the first attitude data is used as the only attitude source. When Th2 ≤ the real-time confidence of the first attitude data ≤ Th1, the fusion is weighted according to the confidence ratio. The first attitude data and the second attitude data are weighted and fused to obtain the fused attitude data.

8. The system according to claim 1, characterized in that, It also includes an ultrasonic sensor unit, which is installed on the lower edge of the bumper in front of the corresponding wheel, close to the front of the laser linear array sensor of the laser linear array unit, for emitting ultrasonic waves and receiving echoes to obtain road surface elevation envelope data; the difference between the road surface elevation envelope data of the ultrasonic sensor unit and the road surface elevation data of the laser linear array unit is compared and verified with a preset threshold. When the difference between the two exceeds the preset threshold, a self-test program is triggered to identify the failed sensor and switch the data source.

9. The system according to claim 1, characterized in that, The failure modes of the laser linear array unit, the IMU unit, and the infrared horizon unit satisfy an orthogonal relationship. When any one or two of the laser linear array unit, the IMU unit, and the infrared horizon unit fail, the system has redundancy and fault tolerance capabilities, and maintains normal operation through the remaining sensors.

10. A vehicle attitude perception and road surface prediction method based on three-source fusion, applied to the system described in any one of claims 1-9, characterized in that, Includes the following steps: Step S1: Obtain road surface elevation data on the driving trajectory of each wheel by using the laser linear array unit set for each wheel; Step S2: The laser reflection signal is preprocessed by the ASIC processing unit, and the preprocessed road surface elevation data of each wheel is output. Step S3: Collect the vehicle's angular velocity and acceleration data through an IMU unit located at the vehicle's center of gravity, and output the first attitude data; Step S4: Detect the infrared radiation temperature difference between the sky and the ground using an infrared horizon sensor unit installed on the top of the vehicle, and output the second attitude data; Step S5: Perform the following operations through the fusion control unit: Generate corresponding feedforward control commands based on the pre-processed road surface elevation data of each wheel; Calculate the real-time confidence scores of the first attitude data and the second attitude data, respectively. The first attitude data and the second attitude data are weighted and fused based on the real-time confidence scores of the first attitude data and the second attitude data to generate fused attitude data. The feedforward control command and the fused attitude data are integrated in a coordinated manner to output control signals for each wheel.

Images