A vehicle-mounted multi-sensor active-passive coordinated stability enhancement device and method

Abstract

This invention discloses an onboard multi-sensor active-passive collaborative stabilization device and method, relating to the field of stability control technology for autonomous vehicle perception systems. The device includes a vehicle mounting base, a passive vibration isolation module, an active compensation module, a multi-sensor common reference support platform, an attitude and vibration sensing module, and a collaborative control module. The passive vibration isolation module attenuates mid-to-high frequency vibrations and impact loads transmitted from the vehicle body to the support platform. The active compensation module compensates for the pitch, roll, and / or yaw attitudes of the support platform in real time. The collaborative control module preprocesses disturbance information, identifies disturbances, divides frequency bands, and allocates control, ensuring that low-frequency attitude disturbances are compensated by the active compensation module, mid-to-high frequency vibrations are attenuated by the passive vibration isolation module, and outputs stabilized multi-sensor data and / or residual attitude information. This invention improves the installation stability and relative pose stability of multiple sensors such as cameras, lidar, and IMUs during vehicle operation, reduces image blurring, point cloud distortion, and attitude drift, and enhances the stability and accuracy of multi-sensor fusion perception, localization mapping, and autonomous driving systems.

Description

Technical Field

[0001] This invention relates to the field of stability control technology for autonomous vehicle perception systems, specifically to an active and passive collaborative stabilization device and method for multiple sensors such as onboard cameras, lidar, and inertial measurement units (IMUs), belonging to the field of vehicle vibration isolation, attitude compensation, and multi-sensor fusion perception support technology. Background Technology

[0002] With the rapid development of low-speed autonomous vehicles, unmanned delivery vehicles, park inspection vehicles, and special mobile platforms, environmental perception, localization mapping, and fusion decision-making technologies based on multi-source sensors such as cameras, LiDAR, and IMUs have become a key foundation for achieving autonomous vehicle operation. These sensors are typically mounted on external vehicle supports, top beams, or cabin structures. During vehicle operation, they are susceptible to factors such as engine vibration, uneven road surfaces, changes in vehicle attitude, and local resonance of the mounting structure. This can cause continuous vibration and attitude shifts in the sensor mounting platform, leading to image blurring, point cloud distortion, inertial navigation drift, and changes in the relative extrinsic parameters of multiple sensors. Ultimately, this affects the accuracy and stability of algorithms such as visual SLAM, LiDAR SLAM, and BEV fusion perception.

[0003] In existing technologies, the stabilization of vehicle-mounted sensors is typically addressed using passive vibration reduction, elastic vibration isolation pads, damping brackets, or single-sensor gimbals. While passive vibration isolation structures offer advantages such as low cost and simple structure, their ability to suppress transient impacts under low-frequency, large-displacement disturbances and complex operating conditions is limited. Furthermore, they are prone to decreased vibration isolation effectiveness or even resonance amplification in specific frequency bands. Although active stabilization methods can compensate for some attitude disturbances, most existing solutions are designed for single cameras or single measurement units, making it difficult to accommodate the shared reference installation requirements of multiple sensors such as LiDAR, cameras, and IMUs, and also difficult to guarantee the stability of the relative pose relationships between multiple sensors during stabilization. In addition, existing vehicle-mounted stabilization solutions mostly focus on vibration reduction at the mechanical structure level or control compensation of single actuators, lacking a mechanism for active and passive division of labor based on the frequency band characteristics of multi-source disturbances, and also lacking a design approach for collaborative adaptation with backend perception, localization, and fusion algorithms. This results in insufficient adaptability in application scenarios such as low-speed autonomous driving, where real-time performance, cost, weight, and integration requirements are high. Especially for multi-sensor fusion systems that require long-term stable output of high-quality images, point clouds, and inertial navigation information, existing technologies still struggle to achieve a balance between lightweight, low cost, and high stability.

[0004] Therefore, there is an urgent need to provide an active and passive collaborative stabilization device and method for vehicle-mounted multi-sensor systems, so as to effectively suppress vehicle disturbances of different frequency bands and types, improve the overall stability and data quality of the multi-sensor mounting platform, and enhance its support capability for autonomous driving perception, localization and fusion algorithms. Summary of the Invention

[0005] To address the problem that existing technologies for vehicle-mounted cameras, lidar, IMUs, and other sensors are susceptible to engine vibration, road surface excitation, mounting bracket resonance, and vehicle attitude disturbances during vehicle operation, leading to image blurring, point cloud distortion, inertial navigation drift, and instability of relative extrinsic parameters of multiple sensors, which in turn affects the accuracy of visual SLAM, lidar SLAM, and multi-sensor fusion perception, this invention provides a vehicle-mounted multi-sensor active and passive collaborative stabilization device and method to achieve hierarchical suppression and collaborative compensation for disturbances of different frequency bands and types, thereby improving the stability of the multi-sensor mounting platform, data consistency, and its support capability for autonomous driving perception and localization algorithms.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The present invention provides a vehicle-mounted multi-sensor active-passive collaborative stabilization device, including a vehicle body mounting base, a passive vibration isolation module, an active compensation module, a multi-sensor common reference bearing platform, an attitude and vibration sensing module, and a collaborative control module; the vehicle body mounting base is fixedly installed on the vehicle body or sensor support structure; the passive vibration isolation module is disposed between the vehicle body mounting base and the multi-sensor common reference bearing platform, used to attenuate the mid-to-high frequency vibrations and some impact loads transmitted from the vehicle body to the bearing platform; the active compensation module is connected to the multi-sensor common reference bearing platform, used for... The system performs real-time compensation for the pitch, roll, and / or yaw attitudes of the carrier platform. The multi-sensor common reference carrier platform is used to install one or more sensors, including cameras, lidar, and inertial measurement units (IMUs), so that the relative attitude relationships of multiple sensors remain stable during the stabilization process. The attitude and vibration sensing module is used to collect acceleration, angular velocity, attitude angle, or vibration response information from the vehicle body and the carrier platform. The collaborative control module is connected to the attitude and vibration sensing module and the active compensation module, and is used to perform collaborative allocation control of passive vibration isolation and active compensation based on the disturbance frequency, amplitude, direction, and trend of change.

[0007] Preferably, the passive vibration isolation module includes an elastic support unit, a damping energy dissipation unit, and a limit protection unit, wherein the elastic support unit is used to provide vibration isolation degrees of freedom, the damping energy dissipation unit is used to reduce resonance peak and attenuate vibration energy, and the limit protection unit is used to limit the overtravel of the bearing platform under impact conditions or large displacement conditions.

[0008] Preferably, the passive vibration isolation module is one or more combinations of a quasi-zero stiffness vibration isolation structure, a spring-damped composite structure, a flexible hinge vibration isolation structure, or a rubber-metal composite vibration isolation structure.

[0009] Preferably, the active compensation module includes a gimbal mechanism, a drive motor, a transmission mechanism, and an angle detection unit. The active compensation module has two-degree-of-freedom or three-degree-of-freedom attitude adjustment capability and is used to correct low-frequency, large-amplitude attitude disturbances in real time.

[0010] Preferably, the attitude and vibration sensing module includes a reference inertial sensor disposed at the vehicle body mounting base and a platform inertial sensor disposed at the multi-sensor common reference bearing platform. The collaborative control module determines the response relationship between disturbance input and platform output based on the measurement results of the two sensors.

[0011] Preferably, the collaborative control module includes a signal preprocessing unit, a disturbance identification unit, a frequency band division unit, a control allocation unit, and a compensation output unit; the signal preprocessing unit is used to filter, denoise, synchronize, and normalize the acquired signals; the disturbance identification unit is used to identify disturbance characteristics formed by engine vibration, road surface excitation, structural resonance, and transient impact during vehicle operation; the frequency band division unit is used to divide the disturbance into a low-frequency attitude disturbance frequency band, a mid-frequency resonance sensitive frequency band, and a high-frequency vibration frequency band; the control allocation unit is used to allocate the disturbance to the active compensation channel and the passive vibration isolation channel according to the frequency band and amplitude; the compensation output unit is used to output the residual attitude information and / or compensation parameters of the carrying platform to the perception and positioning algorithm module.

[0012] This invention also provides a vehicle-mounted multi-sensor active-passive collaborative stabilization method based on the above-mentioned device, comprising the following steps: S1, collecting acceleration, angular velocity, attitude angle, and vibration response information from the vehicle mounting end and the multi-sensor common reference bearing platform end; S2, preprocessing the collected signals and extracting disturbance frequency, amplitude, direction, and time-varying characteristics; S3, classifying and identifying the disturbance type according to the disturbance characteristics, classifying the disturbance into low-frequency attitude disturbance, mid-to-high frequency continuous vibration, and transient impact disturbance; S4, isolating and attenuating mid-to-high frequency vibration through a passive vibration isolation module, performing real-time attitude correction for low-frequency attitude disturbance through an active compensation module, and suppressing excessive displacement caused by impact through damping and limiting structures; S5, acquiring the residual attitude error and vibration response of the multi-sensor common reference bearing platform after stabilization, and outputting the compensated attitude information to a visual SLAM, laser SLAM, BEV fusion perception, or inertial navigation solution module; S6, updating the control parameters according to image clarity, point cloud distortion, feature matching rate, positioning error, or stability of fusion results to achieve adaptive closed-loop optimization.

[0013] Compared with existing technologies, this invention has at least the following beneficial effects: First, this invention adopts a combined active and passive stabilization architecture, which suppresses mid-to-high frequency vibrations through passive vibration isolation and corrects low-frequency attitude disturbances through active compensation, thus simultaneously addressing the suppression requirements of disturbances in different frequency bands and overcoming the limited applicability of single passive or single active solutions. Second, this invention sets up a multi-sensor common reference bearing platform, ensuring that multiple sensors such as cameras, lidar, and IMUs maintain a stable relative installation relationship during the stabilization process. This helps reduce extrinsic parameter drift and time alignment errors during multi-sensor fusion, improving the accuracy of fused perception and localization mapping. Third, this invention not only improves platform stability at the mechanical structure level but also outputs residual attitude information or compensation parameters to the backend perception and localization algorithms, achieving synergy between mechanical stabilization and algorithm compensation, thereby improving the overall robustness and engineering applicability of the system. Fourth, this invention is lightweight, cost-controllable, and easy to integrate, making it suitable for low-speed unmanned vehicles, unmanned delivery vehicles, park inspection vehicles, and other mobile robot platforms, with good promotional value and application prospects. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the overall process of a vehicle-mounted multi-sensor active-passive collaborative stabilization device and method according to the present invention.

[0015] Figure 2 This is the core flowchart of the active and passive frequency division coordinated stabilization control method of the present invention.

[0016] Figure 3 This is a schematic diagram of the system response effect before and after the active and passive collaborative stabilization of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the protection scope of the present invention.

[0018] It should be noted that all directional indications in the embodiments of the present invention, such as up, down, left, right, front, and back, are only used to explain the relative positional relationship and movement of the components in a specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0019] Furthermore, the use of terms such as "first" and "second" in this invention is merely for distinguishing different technical features and should not be construed as indicating or implying their relative importance, or implicitly limiting the number of indicated technical features. Therefore, features marked with "first" or "second" may explicitly or implicitly include at least one of those features. Technical solutions from various embodiments can be combined with each other, but only if they are feasible for those skilled in the art. When a combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0020] In low-speed autonomous vehicles, unmanned delivery vehicles, park inspection vehicles, and other mobile robot platforms, the stability, real-time performance, and integration of the perception system are crucial technical indicators affecting the performance of environmental perception, localization and mapping, and fusion decision-making. Faced with multi-source disturbances such as engine vibration, uneven road surface excitation, vehicle attitude changes, and mounting bracket resonance, onboard sensors such as cameras, LiDAR, and IMUs are prone to image blurring, point cloud distortion, attitude drift, and relative extrinsic parameter changes, thereby reducing the stability and accuracy of visual SLAM, LiDAR SLAM, and BEV fusion perception algorithms. Traditional static vibration damping structures or single active gimbal solutions struggle to simultaneously meet the comprehensive requirements of low-frequency attitude compensation, mid-to-high-frequency vibration isolation, and stable installation of multiple sensors on a common reference.

[0021] To address this, this invention proposes an onboard multi-sensor active-passive collaborative stabilization device and method. By optimizing the onboard multi-sensor installation chain, constructing an active-passive frequency-division collaborative stabilization architecture, designing a dual-side perception mechanism at both the vehicle and platform ends, and combining mechanical stabilization with algorithmic compensation collaborative output strategies, it achieves end-to-end stable control from disturbance perception, signal preprocessing, frequency band identification, active-passive control allocation to residual attitude compensation output. The technical system constructed by this invention not only considers lightweight structure, controllable cost, and engineering deployability, but also provides technical support for the application of high-quality multi-sensor perception systems in low-speed autonomous driving scenarios.

[0022] The various embodiments of the present invention will be described in detail below. These embodiments, in conjunction with the accompanying drawings, will illustrate the technical details from vehicle disturbance input, multi-sensor common reference mounting, active and passive coordinated stability augmentation control to response effect verification, comprehensively revealing the innovative points and application effects of the present invention. Example 1

[0023] The following is in conjunction with the appendix Figure 1 The overall system flowchart shown illustrates an application example of this invention in a low-speed unmanned delivery vehicle scenario.

[0024] This embodiment provides an onboard multi-sensor active-passive collaborative stability enhancement device and method, applicable to low-speed autonomous vehicles equipped with cameras, lidar, and inertial measurement units (IMUs). The device is installed on the vehicle's front cabin support, top beam, or other sensor support structure to sense, identify, distribute, and collaboratively suppress multi-source disturbances generated during vehicle operation, thereby improving the stability of the multi-sensor mounting platform and the data quality of the backend perception and localization algorithms.

[0025] As attached Figure 1 As shown, the overall process of this embodiment includes: vehicle disturbance input, attitude and vibration sensing module, signal preprocessing, disturbance identification and frequency band division, active and passive collaborative stabilization control, stable output of multi-sensor common reference bearing platform, multi-sensor data and residual attitude information output, sensing and positioning algorithm module, and stabilization effect evaluation and parameter update. Specifically, when a vehicle starts, brakes, turns, accelerates or decelerates, or passes over uneven road surfaces, it generates vehicle disturbance input consisting of engine vibration, road surface excitation, vehicle attitude changes, and mounting bracket resonance. The above disturbances first act on the vehicle-mounted sensor mounting positions and are transmitted to the stabilization system via the vehicle mounting base. The attitude and vibration sensing module is used to collect motion state information from the vehicle body and sensor platform. In this embodiment, the attitude and vibration sensing module includes a vehicle reference sensor set at the vehicle mounting base and a platform sensor set at the multi-sensor common reference bearing platform, used to collect acceleration, angular velocity, attitude angle, displacement, and vibration response information, respectively. Through dual-side sensing, disturbance input characteristics and platform output response characteristics can be obtained simultaneously, providing basic data for subsequent collaborative control. The acquired signals enter the signal preprocessing stage. In this stage, the system filters, denoises, synchronizes time, removes outliers, and normalizes the raw signals to reduce the impact of noise, sampling errors, and time delays on control accuracy. If necessary, state estimation methods can also be used to predict platform attitude and disturbance trends.

[0026] The preprocessed signal is sent to the disturbance identification and frequency band division module. This module is used to identify the source, frequency range, amplitude, direction of change, and duration of disturbances during vehicle operation, and classifies the disturbances into low-frequency attitude disturbances, mid-frequency resonance-sensitive disturbances, and high-frequency continuous vibrations based on the identification results. Low-frequency attitude disturbances are usually caused by vehicle undulations, braking pitch, and cornering; mid-frequency resonance-sensitive disturbances are usually formed by modal amplification of the mounting bracket or load-bearing platform structure; and high-frequency continuous vibrations are usually formed by the continuous transmission of minor excitations from the engine or road surface.

[0027] After completing disturbance identification and frequency band allocation, the system enters the active-passive coordinated stabilization control phase. For low-frequency attitude disturbances, the coordinated control module drives the active compensation module to perform real-time attitude correction in the pitch and roll directions for the multi-sensor common reference bearing platform. For mid-to-high frequency continuous vibrations, attenuation is mainly achieved through the elastic support unit and damping energy dissipation unit in the passive vibration isolation module. For large displacement or impact conditions, the limit protection structure restricts the platform's overtravel. Through this active-passive division of labor and coordination mechanism, the system can simultaneously meet the requirements for low-frequency disturbance compensation and mid-to-high frequency vibration suppression while maintaining its lightweight and low-cost characteristics.

[0028] After active and passive coordinated stabilization control, the multi-sensor shared reference platform enters a stable output state. In this state, the camera, lidar, and IMU mounted on the platform maintain a relatively stable pose relationship, thereby reducing the problem of multi-sensor extrinsic parameter changes caused by platform jitter, structural deformation, or asynchronous compensation. Subsequently, the system outputs multi-sensor data and residual attitude information. The multi-sensor data includes image data, point cloud data, and inertial navigation data, while the residual attitude information includes small attitude errors, micro-vibration responses, or time synchronization correction parameters that still exist on the stabilized platform. The above information is jointly output to the perception and localization algorithm module. The perception and localization algorithm module may include one or more of the following: visual SLAM module, lidar SLAM module, BEV fusion perception module, and inertial navigation solution module. By receiving the stabilized multi-sensor data and residual attitude information, the backend algorithm can further perform software-level secondary compensation to improve image feature matching accuracy, point cloud registration stability, multi-sensor spatiotemporal consistency, and localization mapping accuracy.

[0029] Finally, based on the output of the perception and localization algorithm module, the system evaluates the stabilization effect and performs parameter updates. The stabilization effect evaluation and parameter updates can be based on indicators such as image clarity, point cloud distortion, feature matching rate, localization error, mapping stability, or fusion result stability. The updated results are then fed back to the disturbance identification and frequency band division stage or the active-passive collaborative stabilization control stage to achieve closed-loop optimization control.

[0030] The beneficial effects of this embodiment are as follows: by constructing an overall closed-loop process of "disturbance input - perception - analysis - active and passive collaborative stabilization - stable output - algorithm compensation - evaluation feedback", the mechanical stability, external parameter stability and data quality of multi-sensor systems in low-speed autonomous driving scenarios can be effectively improved, thereby enhancing the robustness and engineering application effect of perception and localization algorithms. Example 2

[0031] The following is in conjunction with the appendix Figure 2 The flowchart of the active and passive frequency division cooperative stabilization control method shown below illustrates the core control method of this invention.

[0032] This embodiment provides a vehicle-mounted multi-sensor active-passive collaborative stabilization method, applied to the aforementioned vehicle-mounted multi-sensor active-passive collaborative stabilization device. This method is applicable to vehicle-mounted multi-sensor common reference support platforms equipped with sensors such as cameras, lidar, and inertial measurement units (IMUs), and is used to identify, classify, control, allocate, and compensate for multi-source disturbances during vehicle operation.

[0033] As attached Figure 2 As shown, the method in this embodiment includes the following steps: Step S201: Collect the disturbance response signal.

[0034] Acceleration, angular velocity, attitude angle, displacement, and vibration response signals are collected during vehicle operation via a vehicle reference sensor mounted on the vehicle body mounting base and a platform sensor mounted on a multi-sensor common reference bearing platform. The collected data includes both input disturbance information from the vehicle body and output response information from the platform, thus characterizing the disturbance transmission characteristics and the platform's dynamic response state.

[0035] Step S202: Perform signal preprocessing.

[0036] The acquired disturbance response signal is input into the collaborative control module, where the original signal is filtered, denoised, time-synchronized, outlier-removed, and normalized to improve the accuracy of subsequent disturbance identification and control allocation. In an optional implementation, state estimation can also be performed on the preprocessed signal to obtain more stable attitude and vibration state parameters.

[0037] Step S203: Extract perturbation features.

[0038] Based on the preprocessed signal, the collaborative control module extracts characteristic parameters of the disturbance, such as the dominant frequency, amplitude, direction of change, duration, and time-varying trend. Through time-domain analysis, frequency-domain analysis, or joint time-frequency analysis, it can distinguish the characteristics of different disturbances caused by vehicle starting, braking, turning, road unevenness excitation, engine vibration, and bracket resonance, providing a basis for subsequent disturbance classification.

[0039] Step S204: Disturbance type identification and frequency band division are performed.

[0040] Based on the feature parameters extracted in step S203, the disturbances are identified by type and divided into frequency bands, classifying them into low-frequency attitude disturbances, mid-to-high-frequency continuous vibrations, and transient impacts or large displacement disturbances. Low-frequency attitude disturbances are mainly characterized by slow changes in the platform's pitch, roll, and / or yaw angles; mid-to-high-frequency continuous vibrations are mainly characterized by continuous vibration responses caused by engine or road surface excitations or structural resonances; and transient impacts or large displacement disturbances are mainly characterized by high-impact responses or platform overtravel trends occurring within a short period.

[0041] Step S205: Low-frequency attitude disturbances are entered into the active compensation module to perform attitude correction.

[0042] When the collaborative control module determines that the disturbance is a low-frequency attitude disturbance, it sends a control command to the active compensation module. The active compensation module then drives the gimbal mechanism or attitude adjustment mechanism to perform pitch, roll and / or yaw compensation on the multi-sensor common reference bearing platform to reduce the platform tilt and attitude drift caused by the low-frequency attitude disturbance.

[0043] In step S206, the medium-to-high frequency continuous vibration enters the passive vibration isolation module to perform vibration attenuation.

[0044] When the collaborative control module determines that the disturbance is a medium-to-high frequency continuous vibration, the elastic support unit and damping energy dissipation unit in the passive vibration isolation module isolate and attenuate the vibration, thereby reducing the response amplitude of the platform in the resonance frequency band and the high-frequency vibration frequency band, and reducing image blurring and point cloud distortion.

[0045] Step S207: Transient impact or large displacement triggers limit and protection processing.

[0046] When the collaborative control module determines that the disturbance is a transient impact or a large displacement disturbance, the platform motion is constrained by the limit protection structure and / or damping protection structure to prevent overtravel displacement, severe impact, or structural damage to the multi-sensor common reference bearing platform. This step can be executed independently or in parallel with steps S205 and S206.

[0047] Step S208: Output the platform stabilization results.

[0048] After active compensation, passive vibration isolation, and limiting and protection, the multi-sensor common reference bearing platform outputs the stabilized attitude state and vibration response results, so that the platform as a whole maintains a low vibration amplitude and a small attitude deviation, and maintains a stable relative pose relationship between each sensor.

[0049] Step S209, residual attitude error detection.

[0050] The platform's sensors are used to continue detecting the residual attitude error, angular velocity deviation, and micro-vibration response after stabilization, in order to obtain information on the minute disturbances that still exist after the active-passive combined stabilization. This step reflects the actual performance of the stabilization system and provides input parameters for subsequent software-level compensation.

[0051] Step S210: The compensation information is output to the perception algorithm.

[0052] The residual attitude error, time synchronization correction parameters, and / or attitude compensation parameters obtained in step S209 are output to the perception and localization algorithm module to perform software-level secondary compensation on the image data, point cloud data, and inertial navigation data. The perception and localization algorithm module may include one or more of the following: a visual SLAM module, a laser SLAM module, a BEV fusion perception module, and an inertial navigation solution module.

[0053] Step S211: Closed-loop parameters updated.

[0054] Based on the processing results of the perception and localization algorithm module, the stabilization effect is evaluated. Based on indicators such as image clarity, point cloud distortion, feature matching rate, localization error, mapping stability or fusion result stability, the control parameters of the active compensation module, disturbance identification threshold, frequency band division boundary and / or the equivalent parameters of the passive vibration isolation module are updated to form a closed-loop adaptive optimization.

[0055] The beneficial effects of this embodiment are as follows: by classifying and processing multi-source disturbances on the vehicle according to frequency band and type, low-frequency attitude disturbances are preferentially corrected by the active compensation module, mid-to-high frequency continuous vibrations are preferentially attenuated by the passive vibration isolation module, and transient impacts or large displacements are constrained by limiting and protecting structures, thereby forming a stabilization control mechanism that combines active and passive frequency division. Compared with a single vibration isolation method or a single active compensation method, this embodiment can more effectively improve the overall stability of the multi-sensor common reference bearing platform and enhance the robustness and applicability of the back-end sensing and positioning algorithms. Example 3

[0056] The following is in conjunction with the appendix Figure 3 The system response diagram shown illustrates the simulation verification method of the active-passive collaborative stabilization scheme of the present invention.

[0057] To intuitively verify the superiority of the present invention over rigid installation and passive vibration isolation only, this embodiment uses MATLAB to establish a frequency response simulation model and compares and analyzes the system vibration transmissibility or attitude response amplitude under different installation methods. During the simulation, vehicle disturbances are simplified to a sweep frequency excitation within the range of 1Hz to 200Hz, and three comparative conditions are set: the first is the rigid installation condition, corresponding to the sensor platform being directly mounted on the vehicle body bracket; the second is the passive vibration isolation condition, corresponding to an elastic damping vibration isolation structure being set between the vehicle body bracket and the sensor platform; the third is the active-passive coordinated stabilization condition proposed in this invention, corresponding to the superposition of low-frequency attitude active compensation on the basis of passive vibration isolation.

[0058] As attached Figure 3 As shown, under a set of schematic simulation parameters, the response curve A corresponding to the rigid installation condition shows a significant peak near the resonance region, indicating that the vehicle body disturbance is significantly amplified in this frequency band; the response curve B corresponding to the passive vibration isolation condition is lower than that of the rigid installation condition, indicating that the passive vibration isolation structure has a certain suppression effect on mid-to-high frequency vibrations, but there is still a significant response in the low-frequency attitude disturbance region and the resonance sensitive region; the response curve C corresponding to the active-passive synergistic stabilization condition of the present invention is lower than the previous two conditions in general, especially the suppression effect in the low-frequency disturbance region and the resonance region is more obvious, indicating that the active-passive synergistic structure can simultaneously achieve low-frequency attitude correction and mid-to-high frequency vibration isolation attenuation.

[0059] In this embodiment, the appendix Figure 3 The graph can be drawn using black and white lines, with the rigid installation response curve represented by a solid line, the passive isolation response curve by a dashed line, and the active-passive synergistic stabilization response curve represented by a dotted line. The horizontal axis represents frequency, and the vertical axis represents vibration transmissibility or attitude response amplitude. This graph can serve as a simulation effect diagram for verifying the principle of this invention, and also as a template for comparing the results of subsequent vibration table experiments or real vehicle tests.

[0060] Furthermore, in an optional implementation, verification can also be performed by constructing a small vibration table experimental platform. The camera, lidar, and IMU are mounted on a multi-sensor common reference bearing platform 4. The image blurring, point cloud distortion, and attitude drift amplitude under different mounting schemes are compared, and the measured results are converted to values ​​comparable to those of the attached... Figure 3 The corresponding frequency response curves are used to further verify the effectiveness of the present invention in engineering applications.

[0061] In summary, this invention effectively solves the problems of insufficient low-frequency disturbance suppression, unstable relative pose of multiple sensors, and disconnection from back-end algorithms in existing vehicle sensor stabilization solutions by adopting a technical approach of multi-sensor common reference installation, active and passive frequency division collaborative control, and mechanical stabilization and algorithm compensation linkage output. It has good practicality and promotion value. Technical terms that require explanation and are helpful in understanding the present invention

[0062] (1) Vehicle-mounted multi-sensor: refers to a combination of two or more sensors installed on a vehicle platform for environmental perception, localization and mapping, attitude measurement or fusion decision-making, including but not limited to cameras, lidar, inertial measurement units (IMUs), millimeter-wave radar, depth sensors and ultrasonic sensors. In this invention, a combination of camera, lidar and IMU is preferred.

[0063] (2) Active-passive synergistic stabilization: refers to a stabilization control method that combines passive vibration isolation technology with active attitude compensation technology. Passive vibration isolation is mainly used to attenuate mid-to-high frequency continuous vibrations and some impact disturbances, while active compensation is mainly used to correct low-frequency, large displacement or low-frequency attitude change disturbances. The two work together through control strategies to improve the overall stability of the sensor mounting platform.

[0064] (3) Passive vibration isolation: refers to a technical method that does not rely on external drive actuators, but only on mechanical components such as springs, rubber, flexible supports, dampers, flexible hinges or quasi-zero stiffness structures to achieve vibration isolation and energy dissipation. Its characteristics are relatively simple structure and low cost, but its ability to suppress low-frequency large-amplitude disturbances is limited.

[0065] (4) Active compensation: refers to a technical method that corrects the position or attitude of a sensor platform in real time based on attitude changes or vibration signals detected by sensors, using motors, actuators, gimbal mechanisms, or other controllable drive components. In this invention, active compensation is preferably used for dynamic adjustment of pitch and roll directions, and can also be used for yaw direction compensation when necessary.

[0066] (5) Multi-sensor common reference bearing platform: refers to a unified rigid or quasi-rigid structure platform used to install multiple vehicle sensors, so that the relative pose relationship of multiple sensors remains basically unchanged during the stabilization process, thereby reducing the problem of external parameter changes caused by installation structure deformation, platform shaking or asynchronous compensation actions during the multi-sensor fusion process.

[0067] (6) Vehicle body mounting base: refers to the mounting component that is fixedly installed on the vehicle frame, roof, cabin, crossbeam or other load-bearing structure, used to bear the vibration, impact and attitude disturbance during vehicle operation, and as the input connection end of the stabilization device.

[0068] (7) Attitude and vibration sensing module: refers to the detection unit used to collect motion state information of the vehicle body and sensor platform, including but not limited to accelerometer, gyroscope, IMU, angle encoder, displacement sensor and attitude detection module, etc., to obtain acceleration, angular velocity, displacement, attitude angle and vibration response parameters.

[0069] (8) Cooperative control module: refers to the hardware circuit, embedded controller, industrial control unit or software algorithm module used to preprocess, extract features, identify frequency bands, control and allocate, and compensate output the acquired disturbance signals. This module is the core control unit for realizing active and passive cooperative stabilization.

[0070] (9) Frequency band division: According to the frequency range, amplitude and variation characteristics of vehicle disturbances, the input disturbances are divided into different types such as low frequency attitude disturbances, mid frequency resonance sensitive disturbances and high frequency continuous vibrations, so that different suppression methods such as active compensation, passive vibration isolation or limit protection can be adopted respectively.

[0071] (10) Low-frequency attitude disturbance: refers to low-frequency angular motion or gradual attitude deviation of the platform caused by vehicle starting, braking, turning, pitching, rolling, load changes, etc., which usually manifest as changes in pitch angle, roll angle or yaw angle. In this invention, this type of disturbance is mainly corrected by the active compensation module.

[0072] (11) Mid-to-high frequency vibration: refers to higher frequency mechanical vibration caused by engine vibration, road surface roughness excitation, tire-transmitted vibration, local resonance of the support, etc., which usually manifests as continuous shaking or local amplification response of the installation platform. In this invention, this type of disturbance is mainly attenuated by the passive vibration isolation module.

[0073] (12) Quasi-zero stiffness vibration isolation: refers to a type of vibration isolation technology that uses special structural design to enable the vibration isolation system to have sufficient support capacity in the static bearing direction, while exhibiting an equivalent stiffness close to zero in the dynamic working range, thereby achieving better vibration isolation effect in a lower frequency range.

[0074] (13) Residual attitude error: refers to the small attitude deviation, angular velocity deviation or micro-vibration response that still exists on the sensor platform after active and passive collaborative stabilization processing. This parameter can be further output to the back-end sensing, positioning or fusion algorithm module for software-level compensation.

[0075] (14) Extrinsic parameter stability: refers to the ability of multiple sensors to maintain a constant relative position and attitude relationship. The higher the extrinsic parameter stability, the higher the accuracy of time alignment, spatial registration and feature association in multi-sensor fusion algorithms.

[0076] (15) Visual SLAM: refers to a technical method that uses image information acquired by a camera to achieve simultaneous localization and map construction of a vehicle or mobile platform. This method has high requirements for image clarity, feature stability and spatiotemporal consistency, and is therefore easily affected by vehicle vibration and attitude changes.

[0077] (16) Laser SLAM: refers to the technical method of using lidar point cloud data for environmental mapping and pose estimation. Laser SLAM is sensitive to point cloud distortion, scanning posture changes and platform stability, so reliable stabilization support is required.

[0078] (17) BEV Fusion Perception: This refers to a technical method that maps data from multiple sources such as cameras, lidar, and IMUs to a bird's-eye view space for target detection, environmental understanding, or scene fusion analysis. This type of method has high requirements for the stability of external parameters, time synchronization, and data quality among multiple sensors.

[0079] (18) Vibration transmissibility: refers to the ratio of the amplitude of the vibration response at the output end to the amplitude of the vibration excitation at the input end. It is used to measure the ability of a vibration isolation structure or stabilization system to suppress vibration transmission. The lower the vibration transmissibility, the better the vibration isolation or stabilization effect.

[0080] (19) Attitude response amplitude: refers to the magnitude of pitch, roll or yaw angular displacement response of the sensor-supported platform under external disturbance. It can be used to characterize the stability of the platform and the effect of active compensation.

[0081] (20) Limit protection structure: refers to the mechanical limit component installed in the vibration isolation or compensation mechanism, which is used to limit the overtravel of the sensor platform under large displacement, strong impact or abnormal working conditions, and prevent structural damage or sensor detachment.

Claims

1. A vehicle-mounted multi-sensor active-passive collaborative stability enhancement device, characterized in that, It includes a vehicle body mounting base, a passive vibration isolation module, an active compensation module, a multi-sensor common reference bearing platform, an attitude and vibration sensing module, and a collaborative control module; The vehicle body mounting base is fixedly installed on the vehicle body or sensor support structure; The passive vibration isolation module is disposed between the vehicle body mounting base and the multi-sensor common reference bearing platform, and is used to attenuate the vibration and impact loads transmitted from the vehicle body to the multi-sensor common reference bearing platform; The active compensation module is connected to the multi-sensor common reference bearing platform and is used to perform real-time compensation on the pitch attitude, roll attitude and / or yaw attitude of the multi-sensor common reference bearing platform. The multi-sensor common reference bearing platform is used to install two or more vehicle-mounted sensors, so that the relative pose relationship of each vehicle-mounted sensor remains stable during the stabilization process. The attitude and vibration sensing module is used to collect acceleration, angular velocity, attitude angle, displacement and / or vibration response information of the vehicle body and the multi-sensor common reference bearing platform. The collaborative control module is connected to the attitude and vibration sensing module and the active compensation module, respectively. It is used to preprocess the collected disturbance information, identify the disturbance, divide the frequency band and control the allocation. According to the division result, the low-frequency attitude disturbance is compensated by the active compensation module and the mid-to-high frequency vibration is attenuated by the passive vibration isolation module, so as to output the stabilized multi-sensor data and / or the residual attitude information of the multi-sensor common reference bearing platform.

2. The vehicle-mounted multi-sensor active-passive collaborative stabilization device according to claim 1, characterized in that, The passive vibration isolation module includes an elastic support unit, a damping energy dissipation unit, and a limit protection unit. The elastic support unit provides vibration isolation degrees of freedom, the damping energy dissipation unit reduces resonance peak value and attenuates vibration energy, and the limit protection unit restricts the overtravel motion of the multi-sensor common reference bearing platform under large displacement or impact conditions.

3. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The passive vibration isolation module is one or more combinations of quasi-zero stiffness vibration isolation structure, spring-damped composite structure, flexible hinge vibration isolation structure, and rubber-metal composite vibration isolation structure.

4. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The active compensation module includes a gimbal mechanism, a drive motor, a transmission mechanism, and an angle detection unit; the active compensation module has two-degree-of-freedom or three-degree-of-freedom attitude adjustment capability, and is used to perform pitch direction compensation, roll direction compensation, and / or yaw direction compensation on the multi-sensor common reference bearing platform.

5. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The vehicle-mounted sensor includes two or more of the following: camera, lidar, inertial measurement unit (IMU), millimeter-wave radar, depth sensor, and ultrasonic sensor; preferably, the vehicle-mounted sensor includes at least a camera, lidar, and inertial measurement unit (IMU).

6. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The attitude and vibration sensing module includes a vehicle reference sensor disposed at the vehicle body mounting base and a platform sensor disposed at the multi-sensor common reference bearing platform; the cooperative control module determines the relationship between vehicle body disturbance input and platform response output based on the measurement results of the vehicle reference sensor and the platform sensor.

7. The vehicle-mounted multi-sensor active-passive collaborative stabilization device according to claim 1, characterized in that, The collaborative control module includes a signal preprocessing unit, a disturbance identification unit, a frequency band division unit, a control allocation unit, and a compensation output unit. The signal preprocessing unit is used to filter, denoise, synchronize, and normalize the acquired signals. The disturbance identification unit is used to identify disturbance characteristics formed by engine vibration, road surface excitation, structural resonance, and transient impact. The frequency band division unit is used to divide the disturbance into a low-frequency attitude disturbance band, a mid-frequency resonance sensitive band, and a high-frequency vibration band. The control allocation unit is used to allocate the disturbance to the active compensation channel and the passive vibration isolation channel according to the disturbance frequency band, amplitude, and direction.

8. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The compensation output unit is used to output the residual attitude information, time synchronization correction parameters and / or compensation parameters of the multi-sensor common reference bearing platform to the perception and positioning algorithm module, so as to perform software-level secondary compensation on the multi-sensor data.

9. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The perception and localization algorithm module includes one or more of the following: visual SLAM module, laser SLAM module, BEV fusion perception module, and inertial navigation solution module.

10. The vehicle-mounted multi-sensor active-passive collaborative stability enhancement device according to claim 1, characterized in that, The vehicle body mounting base and the multi-sensor common reference bearing platform are made of aluminum alloy, high-strength steel, carbon fiber composite material or a combination thereof. The multi-sensor common reference bearing platform is a rigid or quasi-rigid structure, used to reduce the relative pose changes between multiple sensors during the stabilization process.

Images