Shock absorber self-adaptive regulation and control method and system fusing multi-source data

Abstract

The invention relates to the technical field of vehicle suspension system control, and discloses a multi-source data fused shock absorber self-adaptive regulation and control method and system.The method comprises the steps that multi-source signal data are synchronously collected through a sensor, and a preliminary signal set is generated in combination with environment change data; analyzing vibration signal peak amplitude and displacement data integral variation trend, distinguishing high-frequency small-acceleration and low-frequency large-amplitude pavement excitation characteristics, and determining signal contribution degree distribution; dynamically adjusting the signal weight and introducing temperature compensation to correct stiffness drift to generate a weighted vector; carrying out weighting processing, smoothing, noise suppression and nonlinear amplitude estimation to generate fusion data, and optimizing the weight to obtain a refined weighted vector; after secondary fusion, a cross-frequency-band transition smoothing technology and a least square method are adopted to determine damping characteristic parameters; according to the vehicle suspension system self-adaptive regulation and control method and system, the self-adaptive regulation and control accuracy and the working condition adaptability of the vehicle suspension system are improved.

Description

Technical Field

[0001] This application relates to the field of vehicle suspension system control technology, and in particular to a shock absorber adaptive control method and system that integrates multi-source data. Background Technology

[0002] The damper control technology of vehicle suspension systems has evolved from early mechanical fixed damping control to semi-adaptive control based on a single signal, and now to fully adaptive control that pursues multi-source data fusion. The core objective remains the same: to improve vehicle ride comfort and stability by precisely matching road excitation with suspension response. With the automotive industry's increasing demands for dynamic performance and the proliferation of diverse driving scenarios such as complex urban road conditions and rugged rural roads, traditional control technologies are no longer sufficient to meet the needs of refined and adaptive control. Multi-source data fusion has become the core development trend of damper control technology.

[0003] However, existing multi-source data fusion-based vibration damper control methods still have significant technical bottlenecks: on the one hand, the acquisition of multi-source signals lacks a systematic synchronization mechanism, and the integration process of vibration, displacement, temperature, and environmental data does not fully consider phase consistency and noise interference, resulting in insufficient accuracy of the initial signal set, which in turn affects the accuracy of working condition identification; on the other hand, working condition identification mostly relies on single frequency or amplitude characteristics, failing to effectively distinguish complex road excitation types such as high-frequency small acceleration and low-frequency large amplitude, the distribution of signal contribution is ambiguous, and the weight allocation is mostly a fixed pattern, which cannot be dynamically adjusted according to the working condition. At the same time, it ignores the impact of suspension stiffness drift caused by temperature changes on weight adaptability, resulting in poor multi-source data fusion effect.

[0004] Furthermore, the connection logic between data fusion and weight optimization in existing technologies is not tight, and damping parameters are prone to sudden changes when switching between frequency band operating conditions, affecting ride comfort. Moreover, there is a lack of closed-loop management of data throughout the entire control process, and no long-term operating condition trend analysis and control file archiving mechanism has been established, which cannot provide data support for subsequent maintenance optimization and fault prediction. These problems together result in a low degree of matching between shock absorber damping control and the actual operating requirements of the vehicle suspension system, which restricts the engineering application and development of multi-source data fusion type shock absorber adaptive control technology. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides a shock absorber adaptive control method and system that integrates multi-source data, used to improve the adaptive control accuracy and operating condition adaptability of vehicle suspension systems.

[0006] In a first aspect, this application provides an adaptive control method for a vibration damper that integrates multi-source data, the method comprising: Step S1: Multi-source signal data is synchronously collected by sensors in the vehicle suspension system. Combined with environmental change data during the collection process, the data is processed by a preset sampling frequency to generate a preliminary signal set. Step S2: Analyze the peak amplitude of the vibration signal and the integral change trend of the displacement data in the preliminary signal set, use frequency band division logic to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude, and determine the signal contribution distribution under the current working condition; Step S3: Dynamically adjust the weights of each signal according to the signal contribution distribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector; Step S4: Process the preliminary signal set using the weighted vector, generate fused data by combining nonlinear amplitude estimation, further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data, and generate a refined weighted vector; Step S5: Perform secondary fusion of the fused data using the refined weighted vector, refer to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and use cross-frequency band transition smoothing technology and control parameter optimization method to determine the damping characteristic parameters of the shock absorber. Step S6: Upload the damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate a long-term operating trend record and a complete adaptive control file of the vehicle suspension system by combining historical operating condition data.

[0007] Secondly, this application provides an adaptive control system for vibration dampers that integrates multi-source data, the system comprising: The signal acquisition module is used to synchronously acquire multi-source signal data through sensors in the vehicle suspension system, combine it with environmental change data during the acquisition process, and generate a preliminary signal set after processing at a preset sampling frequency. The feature analysis module is used to analyze the peak amplitude of the vibration signal and the integral change trend of the displacement data in the preliminary signal set. It uses frequency band division logic to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude, and determines the signal contribution distribution under the current working condition. The weighting adjustment module is used to dynamically adjust the weights of each signal according to the signal contribution distribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector. The fusion optimization module is used to process the preliminary signal set through the weighting vector, generate fused data by combining nonlinear amplitude estimation, further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data, and generate a refined weighting vector. The parameter calculation module is used to perform secondary fusion of the fused data through the refined weighted vector, and to determine the damping characteristic parameters of the shock absorber by referring to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and by adopting cross-frequency band transition smoothing technology and control parameter optimization method. The data classification module is used to upload the damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate long-term operating trend records and a complete adaptive control profile of the vehicle suspension system by combining historical operating condition data.

[0008] Compared with the prior art, the beneficial effects of the present invention are at least as follows: 1. The technical solution provided in this application systematically solves the problems of asynchronous signal acquisition and large noise interference in traditional methods by synchronous acquisition of multi-source signals, fusion of environmental data and smoothing by Kalman filtering. The generated preliminary signal set has consistent phase and complete features. Combined with frequency band division and energy ratio analysis, it realizes accurate identification of complex road excitation features such as high-frequency small acceleration and low-frequency large amplitude, which significantly improves the accuracy of working condition judgment and lays a reliable data foundation for subsequent weight adjustment and damping control.

[0009] 2. This application effectively avoids the limitations of fixed weight allocation and the impact of temperature-induced suspension stiffness drift by utilizing a dynamic weight adjustment mechanism and temperature compensation strategy. Through dynamic switching between linear and nonlinear compensation models, it adapts to the stiffness correction requirements of different temperature change scenarios. After secondary fusion and refined weighted vector optimization, it achieves accurate allocation of the contribution of multi-source data, greatly improves the robustness and adaptability of data fusion, and ensures that the calculation of damping characteristic parameters can closely follow real-time operating condition changes.

[0010] 3. This application avoids abrupt changes in damping parameters during operating condition switching by using cross-frequency band transition smoothing technology and least squares optimization, thereby improving the comfort and stability of vehicle driving. At the same time, relying on cloud storage, trend analysis, and full-process archive archiving, a closed-loop management mechanism for shock absorber control is established. This not only provides data support for fault prediction and maintenance plan optimization, but also extends the service life of suspension components and significantly enhances the vehicle suspension system's adaptability to diverse driving scenarios, demonstrating outstanding engineering application value and practical significance. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating the steps of an adaptive control method for a vibration damper that integrates multi-source data, as described in an embodiment of this application. Figure 2 This is a schematic diagram of sensor deployment and collaborative control in an embodiment of this application; Figure 3 This is a comparison diagram of the effects of the present application and the traditional shock absorber control method in the embodiments of this application; Figure 4 This is a structural diagram of a vibration damper adaptive control system that integrates multi-source data, as described in an embodiment of this application. Detailed Implementation

[0013] This application provides a method and system for adaptive control of a vibration damper that integrates multi-source data. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0014] Example 1: This application aims to address the technical problems in existing shock absorber control methods, such as the lack of systematic multi-source signal acquisition and processing, insufficient accuracy in operating condition identification, fixed signal weight allocation without considering environmental interference, weak connection between data fusion and weight optimization, poor adaptability of damping parameter calculation, and lack of closed-loop traceability mechanism for control data. Existing technologies are difficult to adapt to complex and variable road conditions such as high-frequency small acceleration and low-frequency large amplitude. They are prone to suspension stiffness drift due to temperature, affecting control accuracy. They cannot maximize the value of multi-source data through dynamic weight optimization, and lack full-process data recording and trend analysis of the control process. This results in a low degree of matching between shock absorber damping control and the actual operating requirements of the vehicle suspension system, ultimately affecting vehicle driving comfort and stability. This application solves the above-mentioned technical pain points through systematic signal processing, operating condition identification, dynamic weight optimization, and closed-loop control.

[0015] For ease of understanding, the specific process of the embodiments of this application is described below, such as... Figure 1 The adaptive control method for a vibration damper that integrates multi-source data, as shown in the embodiment of this application, includes: Step S1: Multi-source signal data is synchronously collected by sensors in the vehicle suspension system. Combined with environmental change data during the collection process, the data is processed by a preset sampling frequency to generate a preliminary signal set.

[0016] The process of generating the initial signal set includes: simultaneously acquiring multi-source signal data, including vibration signals, displacement data, and temperature information, through accelerometers, linear displacement sensors, and thermistors deployed in the vehicle suspension system; resampling the multi-source signal data using a preset sampling frequency to ensure phase consistency of each signal; performing Fourier transform on the vibration signals and low-pass filtering on the displacement data using time-series aligned multi-source data integration technology, and fusing temperature information as a compensation factor to generate an initial dataset containing the original vibration characteristics; recording environmental change data during the acquisition process through a real-time monitoring interface, including speed change data acquired by the GPS module and air humidity data captured by the humidity sensor; fusing the initial dataset and environmental change data using a weighted average method, and then smoothing the data using Kalman filtering to generate the initial signal set.

[0017] Specifically, this embodiment aims to address the technical problems in the adaptive control method for vibration dampers that involve asynchronous acquisition of multi-source signals, susceptibility to noise and environmental interference, and lack of targeted data integration, resulting in insufficient accuracy of the initial signal set and consequently affecting the accuracy of subsequent working condition identification and damping control. This application achieves stable generation of a high-quality initial signal set through standardized acquisition, synchronous processing, data integration, environmental fusion, and smoothing optimization techniques, laying a reliable data foundation for the entire adaptive control process.

[0018] Specifically, such as Figure 2The diagram illustrating sensor deployment and coordinated control shows three core sensor types deployed at key stress locations in the vehicle suspension system, such as shock absorbers and springs: accelerometers, linear displacement sensors, and thermistors. Accelerometers capture the time-domain vibration signal of the suspension during vehicle operation, while linear displacement sensors measure the real-time extension and contraction of the suspension travel (displacement data). Thermistors detect the real-time temperature around the suspension components. These three sensors are coordinated and controlled by the vehicle's central control unit to ensure complete synchronization of all signals in the acquisition sequence, avoiding signal phase deviations caused by acquisition delays. To adapt to the signal acquisition needs of different driving scenarios, the acquisition interval can be dynamically set according to the vehicle's real-time speed and road surface type. For example, in high-speed driving... In high-speed driving scenarios, vehicle vibration frequencies are high, so the acquisition interval can be set to 10ms to accurately capture high-frequency vibration details. In low-speed driving scenarios, the acquisition interval can be appropriately increased, such as to 20ms, to reduce data redundancy. After acquisition, the vibration signal, displacement data, and temperature information are uniformly resampled using a preset sampling frequency, such as 1000Hz. This sampling frequency can not only completely preserve the high-frequency characteristics in the vibration signal, such as the high-frequency small acceleration vibration of a flat urban road, but also avoid the data processing pressure caused by excessively high sampling frequency. The resampling operation is performed by a digital signal processor to correct the small phase shifts that may occur during the acquisition process, ensuring that the three types of signals are completely aligned in phase, providing a synchronization basis for subsequent multi-source data integration.

[0019] In the multi-source data integration stage, time series alignment technology is adopted as the core integration method. The core principle of this method is to accurately match the feature points of different types of signals based on timestamps. Specifically, firstly, Fourier transform is performed on the vibration signal. Fourier transform is a mathematical method that converts a time-domain signal into a frequency-domain signal. Its core function is to extract the frequency components and corresponding amplitudes of each frequency in the vibration signal, thereby separating the vibration characteristics corresponding to different road surface excitations, such as high-frequency small-acceleration bumps and low-frequency large-amplitude potholes, which facilitates subsequent working condition identification. Secondly, low-pass filtering is applied to the displacement data. The core principle of low-pass filtering is to allow signals below the preset cutoff frequency to pass through while suppressing noise signals above the cutoff frequency. In this embodiment, combined with the suspension vibration frequency characteristics, the cutoff frequency of the low-pass filter is set to a preset value, such as 50Hz, which can effectively filter out high-frequency noise such as road gravel impacts. To mitigate interference from displacement data, temperature information is incorporated as a compensation factor into the data integration process. The core principle of temperature compensation is based on the thermal expansion characteristics of suspension components, correcting the impact of temperature changes on vibration signals and displacement data. For example, when the temperature rises, the stiffness of the suspension springs will slightly decrease. At this time, the weight of vibration features is reduced to offset the deviation of signal perception caused by thermal expansion. Through the above processing, the frequency domain features of the vibration signal, the filtered displacement data, and the temperature compensation factor are fused to form a multidimensional initial dataset containing the original vibration features. This initial dataset can be specifically optimized according to the vehicle application scenario. For example, in the off-road vehicle scenario, low-frequency, large-amplitude signal features are retained to adapt to the recognition requirements of muddy and rugged roads; in the urban sedan scenario, the processing accuracy of high-frequency, small-acceleration signals is enhanced first, while feature filtering reduces data redundancy and improves processing efficiency.

[0020] To further improve the environmental adaptability of the signal set, a GPS module and a humidity sensor are connected through a real-time monitoring interface. The GPS module acquires real-time speed change data during vehicle operation, while the humidity sensor captures ambient air humidity data. This environmental change data is stored in real-time as a log through the real-time monitoring interface. Its core function is to reflect the impact of the driving environment on the suspension signals. For example, as vehicle speed increases, the transmission characteristics of suspension vibration change; when air humidity is high, the friction coefficient of suspension components decreases slightly. These environmental factors indirectly affect the accuracy of vibration signals and displacement data. Subsequently, a weighted average method is used to fuse the initial dataset and the environmental change data. In this embodiment, the root... Based on the degree of influence of environmental data on regulation, the weight of environmental change data can be set as the first weight value, such as 0.2, and the weight of the initial dataset can be set as the second weight value, such as 0.8. Although the data types of the initial dataset and environmental change data are different, after standardization to eliminate differences in physical quantity units, a weighted average fusion can be performed. This fusion is not a simple numerical superposition, but rather a quantification of the influence priority of core signals and environmental factors with a weight of 0.8:0.2. The final result is a preliminary signal set that retains the dominant characteristics of core operating conditions such as vibration and displacement, while incorporating environmental adaptation corrections such as driving speed and air humidity. This ensures the accuracy of operating condition identification and improves the data's adaptability to actual driving scenarios. During the fusion process, the impact of environmental changes on the overall driving environment will also be analyzed simultaneously. The specific influence of the signal is investigated. For example, as vehicle speed increases, the contribution weight of the vibration signal is appropriately amplified to match the dominant role of the vibration signal in condition identification during high-speed driving. Finally, the fused dataset is smoothed using Kalman filtering. Kalman filtering is a signal smoothing method based on recursive estimation. Its core principle is to predict the theoretical value of the current signal based on historical data through a cyclical iterative process of "prediction-update," and then correct the predicted value by combining it with real-time acquired data, thereby minimizing the interference of noise on the signal. Its input data is the fused multidimensional signal data, and the output data is the smoothed signal set. This processing can effectively filter out random noise during the acquisition process, such as sensor noise and electromagnetic interference, and finally generate... The initial signal set includes comprehensive characteristics such as vibration, displacement, temperature, and environmental influences. In special scenarios such as driving in mountainous areas, slope environmental data can be additionally incorporated into the generation of the initial signal set, and the weight of displacement features can be appropriately increased to adapt to the impact of changes in mountain road slope on suspension travel. This ensures that the initial signal set can accurately reflect the real operating status of the vehicle suspension under different driving scenarios, providing high-quality data support for subsequent signal feature analysis, operating condition identification, and damping characteristic parameter calculation. Among them, slope environmental data refers to environmental parameters collected by the GPS positioning module on the vehicle, such as real-time slope values, slope change rates, and slope duration on mountainous and other road surfaces during vehicle driving, reflecting the spatial distribution and dynamic changes of road slope.

[0021] Step S2: Analyze the peak amplitude of the vibration signal and the integral variation trend of the displacement data in the preliminary signal set, use frequency band division logic to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude, and determine the signal contribution distribution under the current working condition.

[0022] The determination of the signal contribution distribution under the current working condition includes: extracting vibration signal acquisition results and displacement data processing results from the preliminary signal set, calculating the peak amplitude of the vibration signal acquisition results, and analyzing the integral change trend of the displacement data processing results; applying Fourier transform to the vibration signal acquisition results and displacement data processing results, dividing the frequency bands according to the preset frequency range, and distinguishing between high-frequency, low-acceleration road excitation characteristics and low-frequency, large-amplitude road excitation characteristics; calculating the energy proportion of each of the high-frequency, low-acceleration road excitation characteristics and the low-frequency, large-amplitude road excitation characteristics, and determining the signal contribution distribution of the vibration signal and displacement data under the current working condition based on the energy proportion.

[0023] Specifically, this embodiment aims to address the technical problems in the adaptive control method of vibration dampers that integrates multi-source data. These problems include the lack of accurate signal feature analysis logic in the existing working condition identification and the ambiguity in the distinction of road excitation types, which leads to inaccurate determination of signal contribution distribution and affects the adaptability of subsequent weight adjustment and damping control. This application achieves accurate determination of signal contribution distribution under the current working condition through systematic signal feature extraction, frequency domain conversion, excitation type distinction and energy proportion quantification, etc., providing a scientific basis for dynamic weight adjustment.

[0024] Specifically, the vibration signal acquisition results and displacement data processing results are first extracted from the preliminary signal set. The vibration signal acquisition results are the suspension vibration time-domain signals after synchronous acquisition, phase calibration, and smoothing. The vibration signals are acquired by accelerometers, and the core physical quantity is vibration acceleration, measured in g. The displacement data processing results are the suspension travel extension / retraction data after low-pass filtering and environmental fusion, and the core physical quantity is displacement amplitude, measured in mm. Both retain the core dynamic characteristics of the vehicle suspension during operation, laying the foundation for subsequent feature analysis. In the signal feature analysis stage, the peak acceleration of the vibration signal acquisition results is first calculated. Peak acceleration refers to the maximum and minimum values ​​in the vibration acceleration waveform. Half of the value difference is used to quantify the strength of the vibration signal, with the unit being g. For example, when the peak acceleration of the vibration signal is less than or equal to 0.3g, it indicates that the suspension vibration is relatively stable under the current operating conditions; when the peak acceleration is greater than 0.8g, it indicates that there is significant bumpiness on the current road surface. At the same time, integral analysis is performed on the displacement data processing results. The displacement data integral refers to the cumulative change of the displacement amplitude value over time, forming an integral trend curve, which is used to observe the cumulative change law of displacement. The rising slope of the integral trend curve can intuitively reflect the rate of change of suspension travel. Through the above peak acceleration calculation and integral trend analysis, the dynamic characteristics of the signal can be preliminarily identified, providing a basis for subsequent differentiation of road surface excitation characteristics.

[0025] Subsequently, Fourier transforms were applied to the vibration signal acquisition results and displacement data processing results, respectively. Fourier transform decomposes the sine and cosine components of the signal by integrating the time-domain signal, thereby obtaining the frequency spectrum. The input data of this method is the time-domain sequence of vibration signal and displacement data, and the output data is the corresponding frequency spectrum. Its core function is to separate signals with different frequency components that are difficult to distinguish, facilitating accurate identification of different types of road excitation characteristics. Based on the frequency spectrum obtained by Fourier transform, frequency bands are divided according to a preset frequency range. The preset frequency threshold can be dynamically adapted according to the vehicle application scenario to achieve accurate differentiation of different road excitation types. For example, in urban road scenarios, vehicles mainly encounter frequent small bumps, so the preset frequency threshold can be set to 8Hz, with 0 to 8Hz divided into the low-frequency band and 8 to 100Hz into the high-frequency band. In highway scenarios, vehicles are prone to long-wave road excitation, so the preset frequency threshold is adjusted to 12Hz, with 0 to 12Hz divided into... The system is divided into a low-frequency band and a high-frequency band (12-100 Hz). This frequency band division covers the main vibration frequency range of the vehicle suspension system and effectively distinguishes the frequency characteristics corresponding to different road surface excitations. Within the divided frequency bands, it further distinguishes between high-frequency road surface excitation characteristics with small acceleration and low-frequency road surface excitation characteristics with large amplitude. High-frequency road surface excitation characteristics with small acceleration refer to vibration signals located in the high-frequency band with vibration acceleration values ​​less than a preset vibration acceleration value, such as 0.5g. This characteristic mainly corresponds to minor bumps such as potholes and manhole covers on urban roads. Low-frequency road surface excitation characteristics with large amplitude refer to displacement changes located in the low-frequency band with displacement amplitude values ​​greater than a preset amplitude value, such as 2mm. This displacement change characteristic mainly corresponds to severely bumpy roads such as large potholes and bumps on rural roads and construction sections. The preset amplitude and preset vibration acceleration values ​​are set based on the vibration response characteristics of the vehicle suspension system. After multiple tests, they can accurately match the actual situation of different road surface excitations, ensuring the accuracy of excitation type differentiation.

[0026] Then, the energy proportion of the two types of road excitation features mentioned above is quantified. The energy proportion refers to the proportion of the signal energy corresponding to a single type of road excitation feature to the total signal energy. The calculation method is to extract the energy value corresponding to each type of feature through the frequency spectrum in the frequency domain, and then divide the energy value of a single feature by the total energy value to obtain the proportion result.

[0027] Finally, based on the above energy proportion results, the signal contribution distribution of vibration signal and displacement data under the current working condition is determined. The signal contribution distribution refers to the weight allocation basis of vibration signal and displacement data in the current working condition identification and subsequent control process. Its core principle is that the signal corresponding to the feature with higher energy proportion has a greater impact on working condition identification and control, and the higher the contribution. When the energy proportion of low-frequency large amplitude feature is dominant, the contribution of displacement data is appropriately increased to ensure that the contribution distribution is accurately matched with the actual needs of the current working condition.

[0028] Through the above technical process, the signal contribution distribution can be accurately determined. The parameter settings and algorithm processes of all technical means have been tested and verified. Technical personnel in the relevant technical fields can directly implement this technical solution by following the feature extraction method, algorithm principle, parameter setting and scenario adaptation logic described above. This provides a reliable working condition basis for subsequent dynamic weight adjustment and damper damping characteristic parameter calculation, ensuring the accuracy and effectiveness of the entire adaptive control process.

[0029] Step S3: Dynamically adjust the weights of each signal according to the signal contribution distribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector.

[0030] The process of generating a weighted vector includes: extracting the proportion of high-frequency small acceleration features with frequencies higher than a preset first frequency and vibration acceleration values ​​lower than a preset acceleration value from the signal contribution distribution; if the proportion of high-frequency small acceleration features exceeds a preset threshold, the real-time weight allocation is tilted towards the vibration signal; obtaining temperature information from the real-time monitoring interface, correcting the stiffness drift of the vehicle suspension system through a linear compensation formula, and generating corrected weight data; fusing the corrected weight data with the tilted real-time weight allocation results, and generating a weighted vector through weighted averaging and normalization.

[0031] Specifically, this embodiment aims to solve the technical problem in the adaptive control method of shock absorbers that integrates multi-source data. The signal weight allocation cannot adapt to changes in working condition characteristics and does not consider the influence of suspension stiffness drift caused by temperature, resulting in insufficient accuracy of weight setting and thus reducing the multi-source data fusion effect and damping control adaptability. This application achieves accurate generation of weighted vectors by dynamically tilting weights based on working condition characteristics, correcting stiffness drift due to temperature compensation, and normalizing the fusion of weight data, thus providing a scientific weight basis for multi-source data weighted fusion.

[0032] Specifically, firstly, the proportion of high-frequency small acceleration features is extracted from the determined signal contribution distribution. Furthermore, to more accurately capture the sensitivity to high-frequency micro-vibrations of the suspension, a preset first frequency, such as 50Hz, higher than the frequency threshold used in step S2 for distinguishing road excitation features, is used to define the "high-frequency" component. That is, in this embodiment, vibration signals with frequencies higher than the preset first frequency (e.g., 50Hz) and vibration acceleration values ​​lower than a preset acceleration value (e.g., 0.5g) are defined as high-frequency small acceleration features. The proportion of high-frequency small acceleration features is calculated by the ratio of the signal energy corresponding to this feature to the total signal energy. Its core purpose is to determine whether the current operating condition is dominated by high-frequency small acceleration road excitation, such as in urban gravel roads or paved joint roads. Under these conditions, the proportion of high-frequency small acceleration features will increase significantly. The calculated proportion of high-frequency small acceleration features is compared with a preset threshold, such as 60%. If the proportion of high-frequency small acceleration features exceeds the preset threshold, it indicates that the vibration signal plays a more prominent role in road excitation identification under the current working condition. At this time, the real-time weight allocation is tilted towards the vibration signal, that is, the default weight coefficient of the vibration signal is adjusted from the first weight value to the second weight value, which is greater than the first weight value. At the same time, the weight coefficient of the displacement data is reduced synchronously, such as from 1-first weight value to 1-second weight value. This weight tilting method has significant adaptability in urban road driving scenarios and can effectively improve the system's response accuracy to minor road surface unevenness, making the adaptive control of the vehicle suspension system more in line with the real-time working condition requirements.

[0033] After initial weighting, real-time temperature information around the suspension components is obtained from the vehicle's real-time monitoring interface. A temperature compensation mechanism is introduced to correct the stiffness drift of the vehicle's suspension system. Stiffness drift refers to the phenomenon where thermal expansion and contraction of suspension metal components caused by temperature changes leads to a deviation in the material's elastic modulus, ultimately causing the suspension stiffness to deviate from the standard value. In this embodiment, a linear compensation formula is used to correct stiffness drift. The compensation coefficient of this formula is set based on empirical data on the correlation between temperature and suspension stiffness; specifically, a 1°C change in temperature will cause a 0.1% shift in suspension stiffness. The input data for this linear compensation formula are the difference between the real-time temperature and the standard temperature, and the initial weight data. The output data is the weight data after stiffness drift correction. The linear compensation formula for stiffness drift correction is as follows: , This represents the signal weight data after temperature stiffness drift correction, i.e., the corrected weight of the vibration signal or displacement data, with a value range of [0, 1]. represents the initial weight data, i.e., the initial weight values ​​dynamically adjusted based on the signal contribution distribution in step S3; k represents the temperature-stiffness coupling compensation coefficient, with a value of ±0.001, corresponding to a 0.1% shift in suspension stiffness caused by each 1℃ change in temperature. When the temperature increases, it takes +0.001, indicating a decrease in stiffness, requiring a reduction in the vibration signal weight; when the temperature decreases, it takes... An increase of 0.001 in stiffness necessitates increasing the vibration signal weight. This indicates the real-time temperature, which is the real-time temperature around the suspension components collected by the thermistor, in °C. This indicates the standard temperature, which is the design reference temperature for the suspension system. The default value is 25℃, but it can be adjusted according to the climate characteristics of the region where the vehicle is used. This represents the difference between the real-time temperature and the standard temperature, expressed in °C. A positive value indicates that the temperature is higher than the standard value, while a negative value indicates that the temperature is lower than the standard value.

[0034] In practical applications, this temperature compensation mechanism can be dynamically adjusted according to different environmental temperature scenarios. For example, in cold winter off-road scenarios, when the temperature drops from 0℃ to -20℃, the contraction of suspension components leads to increased stiffness, making the suspension response sluggish. In this case, the vibration signal weight is further amplified on the basis of the original weight tilt to compensate for the influence of temperature on signal sensitivity. In hot summer highway scenarios, when the temperature rises to 40℃, the thermal expansion of suspension components leads to decreased stiffness and increased flexibility. In this case, the vibration signal weight is lowered, and the correction is completed by comparing the results with historical working condition data, effectively avoiding signal distortion caused by stiffness drift, while improving the robustness of multi-source data fusion and extending the service life of suspension components. Subsequently, the corrected weight data and the tilted real-time weight allocation results are fused. First, a weighted average algorithm is used to calculate the fusion of the two sets of weight data. This algorithm assigns reasonable fusion weights to the two sets of data with working condition adaptability as the core, ensuring that the corrected stiffness compensation effect and the working condition-based weight tilt characteristics are effectively reflected. Then, the fused weight data is normalized. The core of the normalization process is... The principle is to map the fused weighted data to the range of 0-1, ensuring that the sum of the weight coefficients of vibration signals and displacement data is 1. This avoids imbalances in subsequent data fusion caused by weight superposition. The final result is an adjusted weighted vector, which is the core weight allocation vector adapted to the multi-source data fusion control of the suspension damper. This core vector includes two main components: the final weight coefficients of the vibration signals and the final weight coefficients of the displacement data. After normalization, the sum of these weight coefficients is 1. This weighted vector can be directly applied to the subsequent weighted fusion processing of the initial signal set. In application, it effectively reduces noise interference. This weighted vector also exhibits good adaptability in scenarios with continuously changing vehicle operating conditions. For example, when a vehicle switches from urban roads with a high proportion of high-frequency, low-acceleration features to rural roads with predominantly low-frequency, large-amplitude features, this weighted vector enables a smooth transition of signal weights, ensuring that the intermediate fused data generated after subsequent multi-source data weighted fusion fully retains the road excitation characteristics. Simultaneously, it provides accurate weighted data support for the long-term operating trend recording of the cloud monitoring platform, ensuring the integrity of the suspension system's adaptive control profile.

[0035] Step S4: Process the initial signal set using weighted vectors, generate fused data by combining nonlinear amplitude estimation, further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data, and generate refined weighted vectors.

[0036] The process of generating fused data includes: acquiring a weighted vector and a preliminary signal set; weighting the vibration signal, displacement data, and temperature information in the preliminary signal set using the weighted vector; smoothing the weighted signals using a preset data fusion strategy; optimizing the smoothed signals using sensor noise suppression technology to generate intermediate fused data containing road excitation features; evaluating the amplitude attenuation characteristics of the intermediate fused data using a nonlinear amplitude estimation method; and generating fused data based on the evaluation results of the amplitude attenuation characteristics.

[0037] Specifically, this embodiment aims to address the problems in the adaptive control method of vibration dampers that involve data abrupt changes, inherent noise interference from sensors, and difficulty in quantifying the attenuation characteristics of road excitation features after weighting of multi-source signals in the multi-source data fusion method. These issues result in insufficient accuracy of the data fusion results and fail to provide reliable feature basis for subsequent weight optimization. This application achieves the generation of high-precision fused data through systematic signal weighting, smoothing processing, noise suppression, and nonlinear amplitude estimation, providing core data support for the construction of refined weighted vectors.

[0038] Specifically, the adjusted weighted vector and preliminary signal set are first synchronously acquired from the vehicle suspension system's control module. The weighted vector, generated based on signal contribution distribution and temperature compensation, corresponds only to vibration signals and displacement data. The sum of the weight coefficients of the weighted vector and the preliminary signal set is 1. The preliminary signal set consists of vibration signals, displacement data, and temperature information after synchronous acquisition, phase calibration, and environmental fusion. The influence of temperature information is incorporated into the weighted vector generation process through a temperature compensation mechanism. The weighted vector and the preliminary signal set are fully synchronized through timestamp calibration to ensure the timing consistency of subsequent weighting processing. During this process, the weight dimension of the weighted vector must be ensured to be consistent with... The vibration signals and displacement data that need to be weighted in the initial signal set are matched one-to-one, so that the two types of core signals can be matched with the corresponding weight coefficients. After the data acquisition is completed, the vibration signals and displacement data in the initial signal set are weighted separately by weighting vector. Specifically, each weight coefficient in the weighting vector is multiplied point by point with the time domain sequence of the corresponding type of signal in the initial signal set. This adjusts the contribution of different signals in the fusion process, realizes the prominent enhancement of high frequency and small acceleration characteristics, and makes the weighted signal data more in line with the road excitation feature identification requirements of the current working condition. Temperature information is used as the basic parameter for working condition adaptation, supporting the feature adaptability of the weighted signal data to the actual driving scenario throughout the process.

[0039] Subsequently, the weighted signals are smoothed using a pre-defined data fusion strategy. In this embodiment, a weighted average method is used as the core data fusion strategy. The core principle is to perform a time-series moving average on the weighted vibration signal and displacement data by setting a fixed time window. First, the weighted average of all data points within a single time window is calculated, and then the time window is slid along the time axis to complete the smoothing of the entire signal sequence. The size of the time window can be dynamically adjusted according to the vehicle speed. When driving at high speed, the time window is reduced to ensure the real-time performance of the signal, and when driving at low speed, the time window is increased to improve the smoothing effect. This process can effectively eliminate local data mutation points that occur after signal weighting, reduce the impact of single signal fluctuations on the overall fusion result, improve the stability of multi-source signal fusion, and achieve deep integration of vibration signals, displacement data, and temperature information, avoiding the situation where a single signal dominates the fusion result.

[0040] After the smoothing operation is completed, sensor noise suppression technology is used to optimize the smoothed signal for noise reduction. In this embodiment, two noise suppression technologies are adapted according to different road excitation conditions. When the vehicle is traveling on a flat road with high-frequency, low-acceleration excitation, Kalman filtering is applied as the sensor noise suppression technology. The core function of Kalman filtering is to suppress the inherent noise and random interference of the sensor through a cyclic iterative process of prediction and updating. Its input data is the smoothed signal time-domain sequence. At the same time, process noise covariance matrix and measurement noise covariance matrix need to be set. The process noise covariance matrix is ​​used to characterize the uncertainty brought about by the dynamic changes of the suspension system, and the measurement noise covariance matrix is ​​used to reflect the accuracy error of acquisition devices such as accelerometers and linear displacement sensors. This method first calculates the prior estimate of the signal based on historical signal data through state equations, and then fuses the prior estimate with the real-time acquisition value through measurement equations to obtain the corrected posterior estimate. After multiple rounds of iterative updates, the signal noise is effectively suppressed. Effective noise filtering; Under conditions where vehicles travel on bumpy roads with low-frequency, large-amplitude excitation, wavelet transform is applied as a sensor noise suppression technique. Wavelet transform is a signal processing method based on multi-resolution analysis. Its core principle is to generate basis functions of different scales by scaling and translating the mother wavelet function, decomposing the smoothed signal into wavelet coefficients of different frequency scales, and then performing soft thresholding on the wavelet coefficients corresponding to high-frequency noise through thresholding to suppress high-frequency interference components caused by sensor noise. Finally, the signal is reconstructed based on the corrected wavelet coefficients. This method can effectively filter out noise while completely preserving the characteristic details of low-frequency, large-amplitude excitation, adapting to the signal processing needs of complex bumpy roads. After the above noise suppression technology is applied, the core features of road excitation are extracted from the optimized signal, including the peak acceleration of the vibration signal and the integral change trend of displacement data. After integration, intermediate fusion data containing complete road excitation features is generated. This data can accurately reflect the dynamic response characteristics of the suspension system under the current operating conditions.

[0041] Finally, the amplitude attenuation characteristics of the intermediate fused data are quantitatively evaluated using a nonlinear amplitude estimation method. Based on the evaluation results, the final fused data is generated. In this embodiment, the Hilbert transform is used as the core nonlinear amplitude estimation method. This method is a classic signal envelope extraction method. The core principle is to convert the real part signal of the intermediate fused data into the corresponding imaginary part signal through Fourier transform. The real part signal and the imaginary part signal are integrated to generate an analytic signal. The magnitude sequence of the analytic signal is the instantaneous amplitude of the intermediate fused data, thereby realizing the nonlinear quantization of the amplitude of the road excitation feature. The input data is the intermediate fused data containing the road excitation feature, and the output data is the instantaneous amplitude sequence of the intermediate fused data. After obtaining the instantaneous amplitude sequence, the road excitation feature is fitted based on this sequence. The attenuation curve employs an exponential attenuation model to quantify the amplitude attenuation characteristics. The parameters of the exponential attenuation model are fitted and solved using the least squares method to obtain core parameters such as the initial amplitude and attenuation coefficient. This enables accurate calculation of the amplitude attenuation rate, which characterizes the degree of attenuation of road excitation features over time and directly reflects the buffering response effect of the suspension system to road excitation. After quantifying the attenuation characteristics, the fitted attenuation curve, attenuation rate, and other attenuation characteristic parameters are deeply fused with intermediate fusion data. The attenuation characteristics are incorporated as a feature dimension into the time-domain sequence of the intermediate fusion data to generate the final fused data. This fused data retains the core road excitation characteristics of the multi-source signals while incorporating the attenuation characteristics of the features, and can be directly used for the subsequent generation of refined weighted vectors.

[0042] This nonlinear amplitude estimation method exhibits extremely high accuracy in assessing attenuation characteristics under high-frequency, low-acceleration road surface excitation conditions. When the attenuation rate is low, it indicates that the road surface excitation characteristics are continuous, which can provide a basis for subsequent weight optimization to tilt towards the vibration signal and improve the response speed of the suspension system to continuous slight bumps. Under low-frequency, large-amplitude conditions, the amplitude attenuation characteristics can be jointly evaluated by combining temperature compensation results. When the attenuation rate is abnormally high, it can directly indicate that there is a stiffness drift problem in the suspension system, providing a basis for fault characteristics for subsequent secondary fusion processing.

[0043] The process of correcting the stiffness drift of the vehicle suspension system using a linear compensation formula to generate corrected weight data includes: introducing a temperature change rate detection module to calculate the temperature change per unit time in real time to distinguish between sudden temperature change scenarios and slow, gradual temperature change scenarios; setting a preset temperature change rate threshold; if the temperature change per unit time exceeds the threshold, it is determined to be a sudden temperature change scenario, and a nonlinear compensation model is used to calculate the compensation amount; if the temperature change per unit time does not exceed the threshold, it is determined to be a slow, gradual temperature change scenario, and the linear compensation formula is used to calculate the compensation amount; by dynamically switching between the nonlinear compensation model and the linear compensation formula, the stiffness drift of the vehicle suspension system under different temperature change scenarios is corrected, generating corrected weight data adapted to real-time temperature change characteristics.

[0044] Specifically, this embodiment aims to solve the problem that in the adaptive control method of shock absorbers that integrates multi-source data, a single linear compensation formula cannot adapt to different scenarios of sudden and gradual temperature changes, resulting in insufficient accuracy of suspension stiffness drift correction and distortion of weight data. This application achieves accurate correction of stiffness drift under different temperature scenarios by detecting the rate of temperature change and dynamically switching the compensation model, thus ensuring the effectiveness of weight data.

[0045] Specifically, a temperature change rate detection module is first introduced into the suspension system to collect real-time temperature data around the suspension and calculate the temperature change per unit time. In this embodiment, the temperature change rate unit is set to ℃ / s, and the preset temperature change rate threshold can be set to, for example, 0.5℃ / s, to distinguish temperature change scenarios. If the temperature change per unit time exceeds the preset temperature change rate threshold, it is determined to be a sudden temperature change scenario, and a nonlinear compensation model is used to calculate the compensation amount. This model takes the temperature change rate and temperature difference as inputs and achieves nonlinear solution of the compensation amount through polynomial fitting, adapting to the nonlinear drift law of suspension stiffness during sudden temperature changes. If it does not exceed the preset temperature change rate threshold, it is determined to be a slow and gradual temperature change scenario, and the linear compensation formula is used to calculate the compensation amount. The compensation amount is calculated using a formula that takes the temperature difference as input and solves for the compensation amount according to a stiffness drift ratio of 0.1% / ℃. By dynamically switching between the nonlinear compensation model and the linear compensation formula, targeted corrections are made to the suspension system stiffness drift under different temperature change scenarios. The correction amount is then incorporated into the initial weight data using a compensation coefficient multiplication correction method. Specifically, the correction amount calculated by the linear compensation formula or the nonlinear compensation model is essentially the suspension stiffness drift compensation coefficient. This compensation coefficient has a product relationship with the initial weight data; the magnitude of the compensation coefficient directly determines the adjustment range of the initial weights. If the compensation coefficient is greater than 1, the corresponding initial weights will be amplified; if the compensation coefficient is less than 1, the corresponding initial weights will be reduced. The compensation coefficient is positively correlated with the degree of stiffness drift caused by temperature changes. The more significant the stiffness drift, the greater the deviation of the compensation coefficient from 1, and the greater the weight adjustment. In practical applications, when the temperature is below the standard temperature and shows a gradual or abrupt change trend, the suspension components contract, leading to an increase in stiffness. In this case, the calculated compensation coefficient is greater than 1. Multiplying it by the initial weight of the vibration signal amplifies the vibration signal weight, compensating for the suspension sluggishness caused by the increased stiffness. When the temperature is above the standard temperature and shows a gradual or abrupt change trend, the suspension components thermally expand, leading to a decrease in stiffness. In this case, the calculated compensation coefficient is less than 1. Multiplying it by the initial weight of the vibration signal reasonably lowers the vibration signal weight, adapting to the suspension softness caused by the decreased stiffness. To address the issue of increased sensitivity, the displacement signal weights are adjusted based on the vibration signal weights for a coordinated adaptation. After multiplicative correction of the single signal weights, the corrected weight data for both the vibration and displacement signals are normalized, mapping them to the 0-1 range and ensuring the sum of the weight coefficients is 1. This generates corrected weight data adapted to the current temperature scenario, ensuring accurate matching between the weight data and the actual suspension stiffness. All parameters have been verified through real-vehicle testing, and key parameters such as the range of compensation coefficients and the temperature change rate threshold have been determined through full-condition real-vehicle calibration. After normalization, corrected weight data adapted to the current temperature scenario is generated, ensuring matching between the weight data and the actual suspension stiffness. All parameters have been verified through real-vehicle testing.

[0046] The nonlinear compensation model used in this solution for sudden temperature changes is a nonlinear fitting model based on real vehicle calibration data. This design adapts to the real-time computational needs of the vehicle's suspension system's onboard control module, ensuring both the accuracy of stiffness drift correction and the real-time control response. Furthermore, this model and the linear compensation formula use the same dimension output design, enabling seamless dynamic switching between the two compensation methods. The core design principle is that the thermal expansion and contraction of core components such as suspension springs and shock absorber bushings exhibit transient nonlinear characteristics during sudden temperature changes. Linear compensation formulas cannot accurately match the stiffness drift pattern under these characteristics. Therefore, the model is designed based on the suspension... The model is constructed using real-vehicle calibration data of suspension stiffness drift and a nonlinear fitting method to conform to the actual thermal deformation characteristics of suspension components. The input data of this nonlinear compensation model is adapted to the input dimensions of the linear compensation formula, and also incorporates temperature change rate characteristics, specifically the difference between real-time temperature and standard temperature, and the temperature change rate calculated in real-time by the temperature change rate detection module. The standard temperature is the suspension system design reference temperature, defaulted to 25℃. The output data of this model is completely consistent with the output dimensions of the linear compensation formula, thus adapting to the nonlinear compensation of suspension stiffness drift caused by sudden temperature changes. The coefficient, which has the same physical meaning as the linear compensation coefficient, can be directly used as a correction factor for the initial weight data. It is ultimately used to generate corrected weight data adapted to real-time temperature changes, thus ensuring the consistency of the weight data generation logic after switching between linear and nonlinear compensation methods. The construction of this nonlinear fitting model is completed in three steps: real-vehicle calibration, fitting modeling, and model solidification. First, within the full operating temperature range of the vehicle suspension system (-40℃ to 80℃), covering the extreme temperature scenarios of actual vehicle use, different temperature change rate gradients are set, such as from 1 to 5 times higher than the preset temperature change rate threshold. Data is collected at each rate... The actual suspension stiffness drift corresponding to different temperature differences is recorded, along with the weight data correction requirements corresponding to the drift, forming a four-dimensional real-vehicle calibration dataset consisting of temperature change rate, temperature difference, stiffness drift, and compensation coefficient. Based on this calibration dataset, a cubic polynomial nonlinear fitting method is used to construct the mapping relationship between temperature change rate, temperature difference, and nonlinear compensation coefficient, and to solve for the fixed parameters of the fitting formula. Finally, the nonlinear compensation model parameters obtained from the fitting are pre-stored in the control module of the vehicle suspension system to complete the model construction. The vehicle end does not need real-time training; it only needs to call the pre-stored parameters to realize the rapid calculation of the compensation coefficient.In practical applications, the temperature change rate detection module calculates the temperature change per unit time in real time and compares it with a preset temperature change rate threshold, such as 5℃ / min. This threshold can be adjusted according to the vehicle model. If the temperature change per unit time does not exceed this threshold, it is determined to be a slow temperature change scenario, and the linear compensation formula is used to calculate the compensation coefficient. If the temperature change per unit time exceeds this threshold, it is determined to be a sudden temperature change scenario, such as a vehicle moving from a -20℃ outdoor environment to a 25℃ heated storage room in winter, or a vehicle moving from a 60℃ exposed body temperature in summer to a shaded tunnel. In these scenarios, the vehicle suspension system control module directly calls the pre-stored nonlinear fitting model parameters, substitutes the real-time input data into the model to calculate the nonlinear compensation coefficient, and applies the same logic to the initial weight data through both the linear compensation formula and the compensation coefficient calculated by the nonlinear fitting model. Finally, corrected weight data adapted to different temperature change scenarios is generated, achieving precise correction of stiffness drift in the vehicle suspension system.

[0047] The process of generating a refined weighted vector includes: extracting displacement data processing results and vibration signal acquisition features from the fused data, and analyzing the joint amplitude distribution characteristics of the displacement data processing results and vibration signal acquisition features; if the current working condition is determined to be in a low-frequency, large-amplitude range based on the joint amplitude distribution characteristics, then the real-time weight allocation is tilted towards the displacement data, where the low-frequency, large-amplitude range refers to a frequency lower than a preset second frequency and an amplitude greater than a preset amplitude, and the preset second frequency is less than the aforementioned preset first frequency; combining the current working condition analysis results with the comparison results of historical working condition data, a secondary temperature compensation based on the fused data features is performed on the tilted real-time weight allocation, which is based on the coupling relationship between amplitude attenuation characteristics and temperature information in the fused data, to accurately correct the residual influence of stiffness drift under extreme turbulence scenarios, and then the weights are optimized and adjusted to generate a refined weighted vector.

[0048] Specifically, this embodiment aims to address the problem in the adaptive control method for vibration dampers that integrates multi-source data. The weight allocation based solely on the initial signal contribution is difficult to adapt to the real-time operating condition changes reflected in the integrated data, and the lack of weight optimization logic with historical operating condition reference leads to insufficient accuracy in weight allocation and adaptability to operating conditions. By extracting features from the integrated data, performing joint amplitude distribution analysis, adjusting weights according to operating conditions, and optimizing based on historical data, a refined weighted vector is generated, providing a weight basis that is more in line with actual operating conditions for the accurate calculation of the damper's damping characteristic parameters.

[0049] Specifically, the displacement data processing results and vibration signal acquisition features are first extracted from the fused data generated by nonlinear amplitude estimation. The fused data already contains the core information related to the two types of data to be extracted. The displacement data processing result is the trend of suspension travel integral change after processing the displacement dimension in the fused data and incorporating the attenuation characteristics. The travel integral is a derived feature of the original displacement data. The vibration signal acquisition features are the suspension vibration features inherent in the vibration dimension of the fused data, including peak acceleration and frequency distribution. During the extraction process, timestamp calibration and phase matching algorithms are used to ensure the temporal consistency and phase synchronization of the displacement data processing results and vibration signal acquisition features. This extraction process can be adapted to different vehicle driving scenarios.

[0050] After feature extraction, the joint amplitude distribution characteristics of the displacement data processing results and vibration signal acquisition features are analyzed. First, histogram statistics are used to quantify the joint amplitude distribution of displacement and vibration, dividing the displacement amplitude into several continuous intervals, and simultaneously dividing the vibration acceleration values ​​into several continuous intervals. By statistically analyzing the joint occurrence frequency of data within each displacement-vibration amplitude interval, a joint amplitude matrix is ​​constructed. The rows of this matrix correspond to displacement amplitude intervals, the columns to vibration acceleration value intervals, and the element values ​​within the matrix are the joint occurrence frequencies of the corresponding intervals, intuitively reflecting the coupling distribution law of displacement and vibration characteristics. Subsequently, frequency domain analysis is performed on the joint amplitude matrix. The time-domain distribution characteristics of the joint amplitude matrix are converted into frequency-domain characteristics using Discrete Fourier Transform (DFT), extracting the distribution peaks of different frequency bands and clarifying the proportions of low-frequency and high-frequency components in the joint amplitude distribution. The core principle of DFT is to decompose the discrete time-domain joint distribution data into sine and cosine waves of different frequencies. The string component enables precise extraction of frequency domain features. The input is time-domain data of the joint amplitude matrix, and the output is the frequency-amplitude distribution characteristics in the frequency domain. Finally, the skewness of the joint amplitude distribution is calculated. Skewness is a statistical measure of the degree of distribution asymmetry; positive skewness indicates that the joint amplitude distribution is skewed towards higher amplitudes, while negative skewness indicates that the distribution is skewed towards lower amplitudes. Skewness calculation can further determine the intensity distribution pattern of road excitation features. This analysis process can accurately identify the high proportion of low-frequency, large-amplitude features in the joint amplitude distribution when vehicles pass over bumpy roads (e.g., greater than 60%), clarifying the dominant role of displacement data in condition identification. This analysis logic can also be extended to wet and slippery road surface scenarios. In this case, the displacement integral change trend in the fused data is smoother, the peak vibration acceleration decreases, and the proportion of low-frequency components extracted by Fourier transform shows an increasing trend. Analysis of the joint amplitude distribution characteristics can accurately capture the changes in the characteristics of this condition, improving the robustness of subsequent weight adjustments.

[0051] Based on the analysis results of the joint amplitude distribution characteristics, it is determined whether the current working condition belongs to the low-frequency large-amplitude segment. In this embodiment, the low-frequency large-amplitude segment is defined as a working condition where the frequency is lower than a preset second frequency, such as 20Hz, and the displacement amplitude is greater than a preset amplitude, such as 2cm. The preset second frequency is lower than the preset first frequency used in the above-mentioned high-frequency small acceleration feature determination. The determination process is achieved by comparing the proportion of energy below the preset second frequency in the low-frequency segment, the proportion of features with displacement amplitude greater than 5cm in the joint amplitude distribution, and the preset determination threshold. If the comprehensive proportion of low-frequency large-amplitude features exceeds half of the total proportion, the current working condition is determined to belong to the low-frequency large-amplitude segment. At this time, the real-time weight allocation is tilted towards the displacement data, that is, the weight of the displacement data is increased to the vibration. The signal weights are reduced accordingly. Then, weight refinement based on fused data features is performed. This refinement process includes two levels: **Fine-tuning for working condition characteristics:** This involves adaptively fine-tuning the tilted weights by combining the current working condition status directly analyzed from the fused data, such as amplitude attenuation rate and joint distribution skewness. For example, if the fused data indicates an extremely low amplitude attenuation rate, suggesting continuous high-frequency micro-vibrations on the road surface, the displacement weight can be slightly adjusted back from the tilted position to slightly increase the vibration signal weight, thereby enhancing the response to continuous high-frequency excitation. **Historical experience fusion optimization:** This involves matching the current working condition's joint amplitude distribution characteristics, temperature, driving speed, and other core parameters with historical similar working condition data recorded by the cloud monitoring platform. The similarity is calculated using the Euclidean distance algorithm. If the similarity reaches a preset threshold, the current weights are fine-tuned based on the optimal weight control experience of that historical working condition. Specifically, the least squares method can be used, with the historical optimal weights as a reference target, to calculate an optimization scaling factor for the current weight vector. This factor is then applied to the current weights to achieve accurate approximation based on data-driven principles.

[0052] It should be noted that the weight adjustment in this step is a secondary optimization and refinement based on the initial weight allocation and temperature compensation performed in step S3, and on richer fusion data features. The compensation in step S3 mainly addresses the systemic stiffness drift problem caused by ambient temperature; while the refinement in this step focuses on performing a final fine calibration of the weights based on the specific working condition details revealed by the real-time fusion data, such as attenuation characteristics, joint distribution, and historical experience, in order to improve their matching accuracy with transient working conditions; finally, the weights after the above refinement are normalized to ensure that the sum of the weight coefficients of the vibration signal and displacement data is 1, thereby generating the final refined weighted vector.

[0053] In step S2, parameters such as a frequency threshold of 8Hz and a displacement threshold of 2mm are used to broadly identify and macroscopically classify road surface excitations, providing a highly sensitive and comprehensive initial working condition profile. In step S3, parameters such as a frequency threshold of 50Hz aim to improve the sensitivity to high-frequency micro-vibrations, achieving a transition from broad features to refined weights. In step S4, parameters such as a displacement threshold of 2cm are specifically used to lock extreme bumpy scenarios, triggering targeted weight tilting only when encountering strong impacts. These parameters have all been verified on real vehicles, and their differences reflect the layered control design intent of this scheme. The following will describe each step in detail based on this logic.

[0054] Step S5: Perform secondary fusion of the fused data by refining the weighted vector, refer to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and use cross-frequency band transition smoothing technology and control parameter optimization method to determine the damping characteristic parameters of the vibration damper.

[0055] Step S5 further includes: obtaining refined weighted vectors and fused data; performing secondary weighted fusion processing on the fused data using refined weighted vectors; calculating the damping parameter adjustment value using cross-frequency band transition smoothing technology, referring to the comparison results of historical working condition data and the switching threshold in the frequency band division logic; and fitting the damping curve under the current working condition and iteratively optimizing it based on the damping parameter adjustment value and the least squares control parameter optimization method, until the error is lower than the preset error threshold, thereby determining the damping characteristic parameters of the shock absorber.

[0056] Specifically, this embodiment aims to address the problems in the adaptive control method for shock absorbers that, even after the data is optimized once, there are still abrupt changes in characteristics across frequency bands, a lack of historical operating condition support for damping parameter calculation, and insufficient optimization accuracy. This results in the damper damping characteristic parameters not being able to accurately adapt to real-time operating condition changes, affecting the adaptive control effect of the suspension system. By using secondary weighted fusion, smooth transition across frequency bands, and least squares optimization, the damper damping characteristic parameters can be accurately determined, ensuring the ride comfort and driving stability of the suspension system.

[0057] Specifically, the refined weighted vector and fused data are first synchronously acquired from the real-time monitoring interface and data processing module of the vehicle suspension system. The refined weighted vector is a weight allocation vector generated based on the joint amplitude distribution characteristics of the fused data, temperature compensation, and historical operating condition comparison. It includes the weight coefficients of displacement data and vibration signals. This vector is extracted through the real-time monitoring interface to ensure the dynamic adaptability of the weight allocation to the current operating conditions. The fused data is the result of multi-source signal fusion after nonlinear amplitude estimation and feature integration. It fully includes road excitation characteristics and amplitude attenuation characteristics. During the acquisition process, timestamp synchronization calibration technology is used to ensure the temporal consistency of the refined weighted vector and the fused data, avoiding data delays from affecting the real-time performance of subsequent secondary fusion and parameter calculation. This acquisition logic can be adapted to various vehicle driving scenarios, laying an accurate data foundation for subsequent processing.

[0058] After data acquisition, the fused data undergoes secondary weighted fusion processing using a refined weighted vector. The core process involves point-by-point weighted summation of the displacement data and vibration signal weight coefficients in the refined weighted vector with the corresponding signal feature sequences in the fused data. This achieves a precise secondary allocation of the contributions of the two types of signals. The proportion of the weight coefficients directly determines the degree of influence of the corresponding signal in the secondary fusion result. This process further strengthens the road excitation characteristics matching the current working conditions, weakens interference from irrelevant signals, and improves the overall consistency and feature recognition of the data sequence after secondary fusion. This provides more accurate core data for subsequent damping parameter calculations, avoids single signal features dominating parameter calculation results, and ensures the accuracy of parameter calculations.

[0059] After the secondary fusion is completed, referring to the comparison results of historical operating condition data and the switching threshold in the frequency band division logic, the damping parameter adjustment value is calculated using cross-frequency band transition smoothing technology. First, the historical operating condition data comparison results are downloaded from the cloud monitoring platform. These results include the signal contribution distribution, damping characteristic change law, and parameter optimization experience under similar driving conditions in the past. The core function is to provide historical reference for the current damping parameter adjustment and avoid blind adjustment. At the same time, based on the frequency band division logic established in the previous patent, the cross-frequency band switching threshold is clarified. In this embodiment, combined with the vibration response characteristics of the suspension system, the frequency band is divided into a high-frequency band (e.g., above 20Hz) and a low-frequency band (e.g., below 20Hz). The switching threshold is set to 15Hz. The frequency range corresponding to this threshold, such as 12Hz to 18Hz, is used as the transition area between the high-frequency band and the low-frequency band. This is used to identify the critical range of the operating condition switching from high-frequency small acceleration to low-frequency large amplitude, or vice versa. The switching threshold can be dynamically adjusted according to the specific driving scenario. Here, the 20Hz frequency band division is specifically designed for damping parameter smoothing, which is different from the frequency threshold (e.g., 8Hz) used for feature identification in the previous step.

[0060] Subsequently, a cross-frequency band transition smoothing technique is used to process the secondary fusion data sequence. The core principle is to smooth the signal amplitude in the transition region using a weighted averaging method, avoiding abrupt changes in damping parameters due to characteristic abruptness during frequency band switching. The core process includes three steps: First, identifying signal sampling points in the transition region, i.e., accurately selecting all signal data points within the transition interval on both sides of the switching threshold; second, calculating the smoothing value of these sampling points, using a weighted averaging algorithm based on distance from the threshold, where sampling points closer to the switching threshold have lower weights and those farther away have higher weights. This is achieved through linear interpolation (e.g., in conventional scenarios) or quadratic interpolation (e.g., in large amplitude fluctuation scenarios or off-road conditions). The smoothed signal sequence is calculated to ensure that the amplitude change in the transition region is continuous without abrupt changes. Third, the smoothed signal sequence is compared with historical operating condition data obtained from the cloud to quantify the difference between the current signal characteristics and the historical similar operating condition signal characteristics, thereby generating a damping parameter adjustment value. For example, if the current smoothed amplitude is higher than the historical average amplitude of similar operating conditions by a certain percentage, such as 5%, a positive incremental adjustment value is generated; if it is lower than the historical average, a negative decrement adjustment value is generated. This technology has significant adaptability in urban road driving scenarios. When the vehicle encounters continuous small bumps, i.e., when the proportion of high-frequency small acceleration characteristics is high, it can effectively reduce the vehicle body vibration and ride discomfort caused by abrupt changes in damping parameters, thereby improving ride comfort.

[0061] Finally, based on the calculated damping parameter adjustment values ​​and the least squares control parameter optimization method, the damping characteristic parameters of the vibration damper are determined. First, a linear damping model is constructed as the core calculation model. The core expression of this model is the linear relationship between damping force and vibration velocity. The damping coefficient is the core control parameter, used to characterize the vibration damper's ability to impede vibration, with units of Ns / m. Then, the damping parameter adjustment values ​​are input into the control parameter optimization method, and the damping curve under the current working condition is fitted using the least squares method. The least squares method, as a classic algorithm for solving the optimal solution of linear equations, works by minimizing the sum of squared errors between the fitted curve and the actual signal characteristic curve to obtain the optimal damping coefficient. The specific process is as follows: first, based on the signal characteristic data after secondary fusion, a damping model equation is constructed; then, the initial damping coefficient is initially corrected by applying the damping parameter adjustment values; then, through iterative optimization calculations, the damping coefficient is continuously adjusted until the fitting error is lower than a preset error threshold, such as 0.01. The preset error threshold is dimensionless and characterizes the fitting accuracy. The iteration is then stopped, and the final damping characteristic parameters of the vibration damper are obtained.

[0062] The above optimization process can accurately adapt to the damping requirements of different driving scenarios. The parameter settings of all technical means, such as switching threshold, error threshold, standard value of damping coefficient, algorithm flow cross-frequency band smoothing, and least squares iteration, have been verified by multiple real vehicle tests. They meet the mechanical characteristics and adaptive control requirements of the vehicle suspension system. According to the data acquisition method, secondary fusion logic, smoothing process and parameter optimization steps described above, the technical personnel in this field can directly implement this technical solution. The determined damper damping characteristic parameters can accurately match the current real-time driving conditions, ensuring the stability and reliability of the adaptive control of the suspension system.

[0063] Step S6: Upload the damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate a long-term operating trend record and a complete adaptive control profile of the vehicle suspension system by combining historical operating condition data.

[0064] Step S6 further includes: acquiring damping characteristic parameters and current operating condition analysis results, and uploading the damping characteristic parameters and current operating condition analysis results to the cloud monitoring platform through remote data transmission technology; synchronously updating the vehicle suspension system's operating status data in the real-time monitoring interface; performing trend analysis processing on the operating status data based on the comparison results of historical operating condition data to generate a long-term operating condition trend record; and integrating the entire process data of signal acquisition, weight adjustment, and parameter calculation based on the long-term operating condition trend record to generate a complete adaptive control profile for the vehicle suspension system.

[0065] Specifically, this embodiment aims to address the problems in the adaptive control method for shock absorbers that integrates multi-source data, such as the lack of cloud storage and real-time monitoring of damping characteristic parameters, the lack of long-term trend analysis of operating condition data, and the lack of systematic archiving of full-process control data. By uploading to the cloud, updating status, analyzing trends, and archiving the entire process, a complete adaptive control file for the suspension system is generated, providing core support for remote monitoring, fault prediction, and maintenance optimization.

[0066] Specifically, the system first obtains the least-squares optimized damping characteristic parameters and current operating condition analysis results from the vehicle suspension system control unit. The acquisition process employs a data verification mechanism to ensure the data's authenticity and integrity. After data acquisition, the data is uploaded to the cloud monitoring platform via remote data transmission technology. Using an in-vehicle wireless communication module, damping parameters, operating condition identifiers, timestamps, and vehicle identification information are packaged into standardized data packets and transmitted via 4G / 5G networks, with strict control over transmission latency to ensure real-time monitoring requirements. The data packet size is dynamically adjusted based on the vehicle's speed, and data integrity is verified through checksums and calculations. Successful verification is marked as successful upload; failure triggers a retransmission mechanism.

[0067] After the data is successfully uploaded, the operating status data of the vehicle terminal and the cloud real-time monitoring interface are updated synchronously, intuitively displaying core indicators such as current damping parameters, operating condition type, suspension stiffness correction value and current signal weight allocation value, so that vehicle operators and cloud monitoring personnel can grasp the system operation status in real time, and promptly detect and deal with anomalies.

[0068] Subsequently, by comparing historical operating condition data, trend analysis is performed on the operating status data to generate a long-term operating condition trend record: historical operating condition data similar to the current operating condition are retrieved, and key indicators such as peak vibration acceleration, displacement integral change, and damping change are extracted. A cosine similarity algorithm is used to screen matching samples. Moving average filtering is applied to the current data to generate a smooth trend curve. After summarizing and analyzing, a long-term trend record is formed, which can realize functions such as suspension wear prediction and temperature compensation effect verification, providing a scientific basis for vehicle maintenance plan optimization.

[0069] Finally, the data from the entire process, including signal acquisition, weight adjustment, parameter calculation, and secondary fusion, are integrated to generate a complete adaptive control file for the suspension system. The file is systematically archived according to time sequence and process nodes. Abnormal data with deviations of core indicators exceeding a certain percentage, such as 15%, are classified and marked with explanations, such as adjustment anomalies caused by extreme road surface excitation. The file is encrypted and stored in the cloud for subsequent fault diagnosis, control parameter optimization, and other scenarios, forming a complete data closed loop.

[0070] It should be noted that if this application involves the collection, processing, or application of personal information, it shall strictly comply with the requirements of the Personal Information Protection Law of the People's Republic of China and other relevant laws and regulations, clearly and explicitly inform individuals of the information processing rules, and obtain their independent and voluntary authorization and consent; for sensitive personal information, additional separate explicit consent shall be obtained from individuals on the basis of full knowledge to ensure that personal information processing activities are legal and compliant.

[0071] The core technologies and parameter settings involved in this embodiment, such as remote data transmission, cosine similarity algorithm, and moving average filtering, have all been verified through multiple real vehicle tests. They are adapted to the actual application scenarios of adaptive control of vehicle suspension systems. Based on the contents of this specification, those skilled in the art can directly implement this technical solution. The generated adaptive control profile can effectively ensure the long-term stability and reliability of the adaptive control of the suspension system.

[0072] Through the coordination of the above steps, this application improves the adaptive control accuracy and operating condition adaptability of the vehicle suspension system.

[0073] In summary, such as Figure 3The comparison diagram shown here is between the present application and traditional shock absorber control methods. This application illustrates the complete implementation process of a shock absorber adaptive control method that integrates multi-source data. From synchronous acquisition of multi-source signals, feature analysis, and dynamic weight adjustment, to data fusion optimization, damping parameter calculation, and full-process data archiving, a closed-loop control system is formed. Vibration, displacement, temperature, and environmental data are collected through devices such as accelerometers and linear displacement sensors. After synchronous processing and noise suppression, a high-quality preliminary signal set is generated. Combined with frequency band division and energy proportion analysis, the operating condition characteristics are accurately identified. Then, a weighted vector is generated through a compensation mechanism of dynamic weight tilting and temperature change adaptation. Secondary fusion and refinement optimization improve data adaptability. Relying on cross-frequency band smoothing technology and the least squares method, accurate calculation and smooth transition of damping parameters are achieved. Finally, control archives are generated through cloud upload and trend analysis. The entire process considers signal processing accuracy, operating condition adaptability, and control traceability, effectively solving problems such as signal asynchrony, fixed weights, and lack of closed-loop control in traditional methods. It significantly improves the adaptive control effect of the shock absorber, ensuring vehicle driving comfort and stability, and has good engineering application value.

[0074] Example 2: The above describes an adaptive control method for a vibration damper that integrates multi-source data in an embodiment of this application. The following describes an adaptive control system for a vibration damper that integrates multi-source data in an embodiment of this application. Please refer to [link / reference]. Figure 4 An adaptive control system for a vibration damper that integrates multi-source data, as described in this application embodiment, includes: The signal acquisition module is used to synchronously acquire multi-source signal data through sensors in the vehicle suspension system, and combine it with environmental change data during the acquisition process. After processing at a preset sampling frequency, a preliminary signal set is generated.

[0075] The feature analysis module is used to analyze the peak amplitude of vibration signals and the integral variation trend of displacement data in the preliminary signal set. It uses frequency band division logic to distinguish the road excitation characteristics of high-frequency small acceleration and low-frequency large amplitude, and determines the signal contribution distribution under the current working condition.

[0076] The weighting adjustment module is used to dynamically adjust the weights of each signal according to the distribution of signal contribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector.

[0077] The fusion optimization module is used to process the initial signal set through weighted vectors, generate fused data by combining nonlinear amplitude estimation, and further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data to generate refined weighted vectors.

[0078] The parameter calculation module is used to perform secondary fusion of the fused data by refining the weighted vector. It determines the damping characteristic parameters of the vibration damper by referring to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and by adopting cross-frequency band transition smoothing technology and control parameter optimization methods.

[0079] The data classification module is used to upload damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate long-term operating trend records and complete adaptive control files of the vehicle suspension system by combining historical operating condition data.

[0080] Through the synergistic cooperation of the above components, the adaptive control accuracy and adaptability of the vehicle suspension system are further improved.

[0081] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0082] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0083] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for adaptive control of a vibration damper that integrates multi-source data, characterized in that, The method includes: Step S1: Multi-source signal data is synchronously collected by sensors in the vehicle suspension system. Combined with environmental change data during the collection process, the data is processed by a preset sampling frequency to generate a preliminary signal set. Step S2: Analyze the peak amplitude of the vibration signal and the integral change trend of the displacement data in the preliminary signal set, use frequency band division logic to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude, and determine the signal contribution distribution under the current working condition; Step S3: Dynamically adjust the weights of each signal according to the signal contribution distribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector; Step S4: Process the preliminary signal set using the weighted vector, generate fused data by combining nonlinear amplitude estimation, further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data, and generate a refined weighted vector; Step S5: Perform secondary fusion of the fused data using the refined weighted vector, refer to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and use cross-frequency band transition smoothing technology and control parameter optimization method to determine the damping characteristic parameters of the shock absorber. Step S6: Upload the damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate a long-term operating trend record and a complete adaptive control profile of the vehicle suspension system by combining historical operating condition data.

2. The adaptive control method for vibration dampers that integrates multi-source data according to claim 1, characterized in that, The generation of the preliminary signal set in step S1 includes: Multi-source signal data is simultaneously collected using accelerometers, linear displacement sensors, and thermistors deployed in the vehicle suspension system. The multi-source signal data includes vibration signals, displacement data, and temperature information. The multi-source signal data is resampled using a preset sampling frequency to ensure that the phase of each signal is consistent. By combining time-series aligned multi-source data integration technology, Fourier transform is performed on the vibration signal, low-pass filtering is performed on the displacement data, and the temperature information is fused as a compensation factor to generate an initial dataset containing the original vibration characteristics. The environmental change data during the acquisition process is recorded through a real-time monitoring interface. The environmental change data includes speed change data acquired by the GPS module and air humidity data captured by the humidity sensor. The initial dataset and the environmental change data are fused using a weighted average method, and then smoothed using Kalman filtering to generate a preliminary signal set.

3. The adaptive control method for vibration dampers that integrates multi-source data according to claim 1, characterized in that, Determining the signal contribution distribution under the current operating condition in step S2 includes: The vibration signal acquisition results and displacement data processing results are extracted from the preliminary signal set, and the peak amplitude of the vibration signal acquisition results is calculated and the integral change trend of the displacement data processing results is analyzed. Fourier transform is applied to the vibration signal acquisition results and the displacement data processing results to divide the frequency range into frequency bands according to the preset frequency range, so as to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude. Calculate the energy proportion of the road surface excitation characteristics with high frequency and small acceleration and the road surface excitation characteristics with low frequency and large amplitude, and determine the signal contribution distribution of the vibration signal and displacement data under the current working condition based on the energy proportion.

4. The adaptive control method for vibration dampers that integrates multi-source data according to claim 1, characterized in that, The generation of the weighted vector in step S3 includes: Extract the proportion of high-frequency small acceleration features with frequencies higher than a preset first frequency and vibration acceleration values ​​lower than a preset acceleration value from the signal contribution distribution. If the proportion of high-frequency small acceleration features exceeds a preset threshold, then the real-time weight allocation will be tilted towards the vibration signal. Temperature information is obtained from the real-time monitoring interface, and the stiffness drift of the vehicle suspension system is corrected by a linear compensation formula to generate corrected weight data. The corrected weight data is then fused with the real-time weight allocation results after tilting, and a weighted vector is generated after weighted averaging and normalization.

5. The adaptive control method for a vibration damper that integrates multi-source data according to claim 1, characterized in that, The generation of fused data in step S4 includes: The weighting vector and the initial signal set are obtained. The vibration signal, displacement data, and temperature information in the initial signal set are weighted using the weighting vector. The weighted signals are smoothed using a preset data fusion strategy. Sensor noise suppression technology is used to optimize the smoothed signals for noise reduction, generating intermediate fused data that includes road excitation features. The amplitude attenuation characteristics of the intermediate fused data are evaluated using a nonlinear amplitude estimation method. Based on the evaluation results of the amplitude attenuation characteristics, fused data is generated.

6. The adaptive control method for a vibration damper that integrates multi-source data according to claim 5, characterized in that, The step S4 of generating the refined weighted vector includes: Displacement data processing results and vibration signal acquisition features are extracted from the fused data, and the joint amplitude distribution characteristics of the displacement data processing results and the vibration signal acquisition features are analyzed. If the current operating condition is determined to be in a low-frequency large-amplitude segment based on the joint amplitude distribution characteristics, then the real-time weight allocation will be tilted toward the displacement data. The low-frequency large-amplitude segment refers to a segment with a frequency lower than a preset second frequency and an amplitude greater than a preset amplitude. By combining the analysis results of the current working condition with the comparison results of historical working condition data, the real-time weight allocation after tilting is optimized and adjusted to generate a refined weighted vector.

7. The adaptive control method for a vibration damper that integrates multi-source data according to claim 1, characterized in that, Step S5 further includes: The refined weighted vector and the fused data are obtained, and the fused data is subjected to secondary weighted fusion processing through the refined weighted vector; the damping parameter adjustment value is calculated by referring to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and cross-frequency band transition smoothing technology is used; based on the damping parameter adjustment value, combined with the least squares control parameter optimization method, the damping curve under the current working condition is fitted and iteratively optimized until the error is lower than the preset error threshold, and the damping characteristic parameters of the shock absorber are determined.

8. The adaptive control method for a vibration damper that integrates multi-source data according to claim 1, characterized in that, Step S6 further includes: The damping characteristic parameters and the current operating condition analysis results are obtained, and the damping characteristic parameters and the current operating condition analysis results are uploaded to the cloud monitoring platform through remote data transmission technology; Synchronously update the operating status data of the vehicle suspension system in the real-time monitoring interface; By combining the results of historical operating condition data comparison, the operating status data is subjected to trend analysis processing to generate a long-term operating condition trend record. Based on the long-term operating condition trend records, the data from the entire process of signal acquisition, weight adjustment, and parameter calculation are integrated to generate a complete adaptive control profile for the vehicle suspension system.

9. The adaptive control method for a vibration damper that integrates multi-source data according to claim 4, characterized in that, The stiffness drift of the vehicle suspension system is corrected using a linear compensation formula, generating corrected weight data including: A temperature change rate detection module is introduced to calculate the temperature change per unit time in real time, distinguishing between sudden temperature change scenarios and slow, gradual temperature change scenarios. A preset temperature change rate threshold is used. If the temperature change per unit time exceeds the threshold, it is determined to be a sudden temperature change scenario, and a nonlinear compensation model is used to calculate the compensation amount. If the temperature change per unit time does not exceed the threshold, it is determined to be a slow, gradual temperature change scenario, and the linear compensation formula is used to calculate the compensation amount. By dynamically switching between the nonlinear compensation model and the linear compensation formula, the stiffness drift of the vehicle suspension system under different temperature change scenarios is corrected, generating corrected weight data adapted to the real-time temperature change characteristics.

10. A vibration damper adaptive control system integrating multi-source data, used to implement the vibration damper adaptive control method integrating multi-source data as described in any one of claims 1-9, characterized in that, The system includes: The signal acquisition module is used to synchronously acquire multi-source signal data through sensors in the vehicle suspension system, combine it with environmental change data during the acquisition process, and generate a preliminary signal set after processing at a preset sampling frequency. The feature analysis module is used to analyze the peak amplitude of the vibration signal and the integral change trend of the displacement data in the preliminary signal set. It uses frequency band division logic to distinguish the road excitation characteristics of high frequency small acceleration and low frequency large amplitude, and determines the signal contribution distribution under the current working condition. The weighting adjustment module is used to dynamically adjust the weights of each signal according to the signal contribution distribution, introduce a temperature compensation mechanism to correct the stiffness drift of the vehicle suspension system, and generate a weighted vector. The fusion optimization module is used to process the preliminary signal set through the weighting vector, generate fused data by combining nonlinear amplitude estimation, further optimize the signal weights based on the joint amplitude distribution characteristics of the fused data, and generate a refined weighting vector. The parameter calculation module is used to perform secondary fusion of the fused data through the refined weighted vector, and to determine the damping characteristic parameters of the shock absorber by referring to the comparison results of historical working condition data and the switching threshold in the frequency band division logic, and by adopting cross-frequency band transition smoothing technology and control parameter optimization method. The data classification module is used to upload the damping characteristic parameters to the cloud monitoring platform, update the operating status data of the vehicle suspension system, and generate long-term operating trend records and a complete adaptive control profile of the vehicle suspension system by combining historical operating condition data.

Images