Intelligent suspension control system based on complex road condition recognition

Abstract

The application provides an intelligent suspension control system based on complex road condition recognition, and relates to the technical field of suspension control.The system comprises a data acquisition module, a theoretical adjustment calculation module, an adjustment atlas generation module, a dynamic adjustment generation module and a real-time monitoring module.The data acquisition module generates structured road condition suspension matching data.The theoretical adjustment calculation module calculates a theoretical suspension adjustment benchmark value.The adjustment atlas generation module generates an initial suspension adjustment atlas.The dynamic adjustment generation module obtains a dynamic suspension adjustment atlas for a driving task.The real-time monitoring module performs real-time monitoring on the intelligent suspension control.The application classifies various complex road conditions through semantic analysis recognition, provides accurate road conditions, calculates a theoretical adjustment benchmark value, introduces a self-adaptive correction factor matrix, dynamically adjusts the adjustment benchmark and the fault tolerance bandwidth according to the real-time attitude of the vehicle, the body vibration and the suspension air pressure, guarantees stable control, and realizes dynamic balance of comfort and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of suspension control, and particularly relates to an intelligent suspension control system based on complex road condition recognition. BACKGROUND

[0002] With the rapid development of the automobile industry towards intelligence and comfort, users have higher requirements for the controllability, smoothness and safety during vehicle driving, and the adjustment performance of the suspension system as a core component connecting the vehicle body and the wheels directly determines the adaptability of the vehicle to different road conditions. The current traditional suspension system is designed with fixed parameters, which can only provide basic support for regular flat road conditions, and has obvious limitations in the face of complex road conditions.

[0003] In actual driving scenarios, vehicles often face various and complex road conditions, and may also encounter unpredictable road conditions such as sudden obstacles and temporary construction areas. The traditional suspension system lacks the ability to accurately recognize complex road conditions, and cannot dynamically adjust its parameters according to the changes in road conditions, resulting in excessive body vibration and decreased comfort on bumpy road sections, and difficulty in ensuring sufficient roll resistance in continuous curves or emergency avoidance scenarios, which affects the control stability and cannot meet the requirements of intelligent control of the suspension system in complex driving scenarios. SUMMARY

[0004] The present application provides an intelligent suspension control system based on complex road condition recognition to solve the defect that the prior art cannot meet the requirements of intelligent control of the suspension system in complex driving scenarios.

[0005] The present application provides an intelligent suspension control system based on complex road condition recognition, comprising:

[0006] A data acquisition module is configured to acquire multi-source road condition data during vehicle driving, identify complex road condition types in the multi-source road condition data through road condition semantic analysis, extract suspension adaptation characteristics under each road condition in combination with vehicle suspension parameters, and generate structured road condition suspension matching data.

[0007] A theoretical adjustment calculation module is configured to collect historical driving task data and extract road condition occurrence frequency, suspension response efficiency and driving smoothness indicators, and calculate a theoretical suspension adjustment reference value according to the road condition occurrence frequency and the suspension response efficiency.

[0008] An adjustment atlas generation module is configured to set a dynamic adjustment fault tolerance band, and generate an initial suspension adjustment atlas sorted by road condition priority in combination with the theoretical suspension adjustment reference value.

[0009] The dynamic adjustment generation module is used for collecting vehicle real-time attitude data, vehicle body vibration frequency spectrum and suspension air pressure data to construct an adaptive correction factor matrix, and injecting the adaptive correction factor matrix into an initial suspension adjustment graph to obtain a dynamic suspension adjustment graph of the driving task.

[0010] The real-time monitoring module is used for collecting real-time suspension working data curves in the driving process, comparing actual suspension adjustment average values in each road condition section, and monitoring the intelligent suspension control in real time by calculating the shape similarity of the real-time suspension working data curves and the dynamic suspension adjustment graph in each road condition time window.

[0011] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the multi-source road condition data includes road image data, road three-dimensional profile data, road obstacle data and road section road condition warning data. The vehicle suspension parameters include suspension spring stiffness coefficient, shock absorber damping adjustment range, stabilizer bar torsional stiffness, suspension stroke limit value and spring pressure adjustment threshold value. The suspension adaptation characteristics include suspension stiffness target value, shock absorber damping coefficient set value, vehicle body height adjustment value, suspension response delay threshold value and vibration suppression rate target value.

[0012] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the process of identifying the complex road condition type in the multi-source road condition data through road condition semantic analysis includes:

[0013] The road image data is subjected to image segmentation and feature extraction, and key road condition features are identified, including road damage, water accumulation, snow accumulation and gravel.

[0014] Based on the road three-dimensional profile data, the road slope, undulation and jolt frequency are calculated, and the road condition grade is divided.

[0015] The road obstacle data and the road section road condition warning data are fused to mark the sudden road condition information, including the sudden road condition type and the danger grade.

[0016] According to the key road condition features, the road condition grade and the sudden road condition information, the conventional flat road condition is filtered out, and the complex road condition type set is classified.

[0017] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the process of generating structured road condition suspension matching data includes:

[0018] According to the complex road condition type, the suspension spring stiffness coefficient and the shock absorber damping adjustment range are associated, and the basic suspension adjustment parameter is determined.

[0019] The road slope and the undulation are combined to calculate the vehicle body attitude compensation amount, and the suspension height adjustment demand value is derived.

[0020] By correlating the torsional stiffness of the stabilizer bar with the suspension travel limit, the anti-roll and anti-bump coefficients of the suspension are calibrated.

[0021] The unique road condition identifier, suspension adjustment parameter set, and attitude compensation amount are encapsulated into a structured data object to obtain structured road condition suspension matching data.

[0022] According to the present invention, an intelligent suspension control system based on complex road condition recognition extracts road condition occurrence frequency, suspension response efficiency, and ride comfort indicators through the following process:

[0023] Collect historical driving task data and filter historical driving records of the same vehicle model and driving area.

[0024] Analyze the road condition types in historical structured data objects and statistically analyze the frequency and duration of different complex road conditions.

[0025] By analyzing historical suspension sensor records, a mapping model of suspension adjustment time under road conditions is constructed to calculate suspension response efficiency.

[0026] Collect data on the vertical and lateral acceleration of the vehicle body during historical driving processes, and calculate the ride comfort index using the vibration comfort formula.

[0027] The intelligent suspension control system based on complex road condition recognition provided by the present invention includes the following process for calculating the theoretical suspension adjustment reference value according to the frequency of road condition occurrence and suspension response efficiency:

[0028] Based on the frequency of road conditions, the basic suspension adjustment parameters under different road conditions are weighted and calculated to generate initial adjustment reference values.

[0029] By combining the suspension response efficiency, a response delay correction coefficient is introduced to adjust the initial adjustment reference value, thus obtaining the adjusted adjustment reference value.

[0030] Based on the ride comfort index, a ride comfort guarantee coefficient is introduced to optimize the adjustment reference value, thereby obtaining the theoretical suspension adjustment benchmark value.

[0031] The intelligent suspension control system based on complex road condition recognition provided by the present invention includes the following process for generating an initial suspension adjustment map sorted by road condition priority by combining theoretical suspension adjustment reference values:

[0032] Create a three-dimensional coordinate system for timing, road conditions, and suspension parameters.

[0033] Based on the timing and priority of road conditions, the corresponding theoretical suspension adjustment benchmark values ​​are connected to form a suspension adjustment reference curve.

[0034] Differentiated adjustment of fault tolerance bandwidth is set based on road condition hazard level and driving speed.

[0035] Visualize the adjustment range boundary and priority identification of different road conditions, and form an initial suspension adjustment map.

[0036] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the process of constructing the adaptive correction factor matrix comprises:

[0037] According to the real-time attitude data of the vehicle, the roll angle and the pitch angle change rate of the vehicle body are calculated, and an attitude correction coefficient is generated.

[0038] According to the vibration frequency spectrum of the vehicle body, the vibration energy entropy value is calculated, and a damper damping correction coefficient is calibrated.

[0039] The fluctuation amplitude of the suspension air pressure data is analyzed, and a pressure compensation correction coefficient is generated.

[0040] The attitude correction coefficient, the damping correction coefficient and the pressure compensation correction coefficient are integrated to obtain the adaptive correction factor matrix.

[0041] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the process of obtaining the dynamic suspension adjustment map of the driving task comprises:

[0042] According to the adaptive correction factor matrix, the suspension adjustment reference value in the initial map is adjusted for each road condition.

[0043] According to the damping correction coefficient, the adjustment fault bandwidth under different road conditions is dynamically scaled.

[0044] The initial suspension adjustment map is superimposed with a correction factor value mark layer and a real-time road condition warning mark.

[0045] A real-time data interface with the vehicle CAN bus is established, and the adaptive correction factor matrix and the dynamic suspension adjustment map are updated according to the preset period.

[0046] According to the intelligent suspension control system based on complex road condition recognition provided by the application, the process of calculating the shape similarity of the real-time suspension working data curve in each road condition time window and the dynamic suspension adjustment map comprises:

[0047] The road condition duration is extracted from the structured road condition suspension matching data, the time window boundary corresponding to each complex road condition is determined, and the map reference curve is obtained.

[0048] The real-time suspension working data curve and the map reference curve are preprocessed.

[0049] The dynamic time warping algorithm is used to calculate the shape similarity of the preprocessed real-time suspension working data curve and the map reference curve.

[0050] Pearson correlation coefficient is introduced as an auxiliary verification index to calculate the linear correlation degree of the real-time suspension working data curve and the reference curve of the atlas.

[0051] The calculated morphological similarity result and the actual suspension adjustment average of each road condition section are stored in the system database.

[0052] The intelligent suspension control system based on complex road condition recognition provided by the application can comprehensively capture road surface images, three-dimensional contours, obstacles and early warning information, accurately identify various complex road conditions and classify them, avoid identification deviation caused by single sensor data, provide accurate road condition basis for subsequent suspension adjustment, and ensure that the suspension adjustment direction is highly matched with the actual road condition demand.

[0053] The road condition frequency, response efficiency and smoothness index extracted from the historical driving data are combined with the theoretical adjustment reference value calculation theory, and a self-adaptive correction factor matrix is introduced, so that the adjustment reference and the fault tolerance bandwidth are dynamically adjusted according to the real-time attitude of the vehicle, the body vibration and the suspension air pressure, and a dynamic suspension adjustment atlas is formed. The suspension parameters can adapt to the road condition change and the vehicle state in real time, the vibration can be effectively suppressed by optimizing the damping and height parameters in the bumpy road section, and the comfort is improved; in the curve or sudden road condition, the anti-roll and support performance can be enhanced, the control stability is ensured, and the dynamic balance of comfort and control is realized. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0055] Figure 1 is a process schematic diagram of an intelligent suspension control system based on complex road condition recognition provided by an embodiment of the application;

[0056] Figure 2 is a process schematic diagram of identifying complex road condition types in an embodiment of the application;

[0057] Figure 3 is a process schematic diagram of generating structured road condition suspension matching data in an embodiment of the application;

[0058] Figure 4 is a process schematic diagram of calculating a theoretical suspension adjustment reference value in an embodiment of the application;

[0059] Figure 5 is a process schematic diagram of constructing a self-adaptive correction factor matrix in an embodiment of the application. DETAILED DESCRIPTION

[0060] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only part of, rather than all of, the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort belong to the scope of protection of the present application.

[0061] The technical solutions in the embodiments of the present application will be described clearly below with reference to the drawings in the present application. Figures 1-5 An intelligent suspension control system based on complex road condition recognition is described.

[0062] Figure 1 is a structural schematic diagram of an intelligent suspension control system based on complex road condition recognition provided by the embodiments of the present application.

[0063] As shown in Figure 1 , an intelligent suspension control system based on complex road condition recognition provided by the embodiments of the present application includes a data acquisition module, a theoretical adjustment calculation module, an adjustment atlas generation module, a dynamic adjustment generation module and a real-time monitoring module.

[0064] The data acquisition module is configured to acquire multi-source road condition data in a vehicle driving process, identify complex road condition types in the multi-source road condition data through road condition semantic analysis, and extract suspension adaptation characteristics under each road condition in combination with vehicle suspension parameters to generate structured road condition suspension matching data.

[0065] The multi-source road condition data includes road surface image data collected by a vehicle-mounted camera, road surface three-dimensional contour data scanned by a laser radar, road surface obstacle data detected by a millimeter wave radar and road section road condition early warning data issued under a vehicle networking. The vehicle suspension parameters include a suspension spring stiffness coefficient, a shock absorber damping adjustment range, a stabilizer bar torsional stiffness, a suspension stroke limit value and a spring pressure adjustment threshold value. The suspension adaptation characteristics include a suspension stiffness target value, a shock absorber damping coefficient set value, a vehicle body height adjustment value, a suspension response delay threshold value and a vibration suppression rate target value.

[0066] Figure 2 is a flowchart of identifying complex road condition types in the embodiments of the present application.

[0067] As shown in Figure 2 , the process of identifying complex road condition types in the multi-source road condition data through road condition semantic analysis includes:

[0068] Image segmentation and feature extraction are performed on the road surface image data to identify key road condition features, including road surface damage, water accumulation, snow accumulation and gravel.

[0069] Based on the three-dimensional road surface profile data, the road surface slope, the undulation and the jolt frequency are calculated, and the road condition level is divided.

[0070] The road surface obstacle data and the road section road condition warning data are fused, and the sudden road condition information is marked, and the sudden road condition information includes a sudden road condition type and a danger level.

[0071] According to the key road condition characteristics, the road condition level and the sudden road condition information, the conventional flat road condition is filtered out, and a complex road condition type set is classified.

[0072] In the embodiment, the road surface image data collected by the vehicle-mounted camera is preprocessed, the MaskR-CNN model based on deep learning is used for image segmentation, the road surface damage (crack, pit), water accumulation (reflective area), snow accumulation (high brightness area), gravel (irregular small target) features are extracted through a multi-scale feature fusion network, the non-maximum suppression algorithm is used to filter the redundant detection boxes, and the feature position coordinates and confidence are output.

[0073] The three-dimensional point cloud data of the laser radar is synchronously processed, the road surface point set is separated through a ground point cloud segmentation algorithm, the height difference average of adjacent frame point clouds is calculated to obtain the undulation, the Fourier transform is used to convert the height sequence into the frequency domain to obtain the jolt frequency, and in combination with the slope calculation result, the road condition is divided into five levels according to a preset threshold, and the higher the level is, the worse the road condition is.

[0074] For the obstacle data of the millimeter wave radar, target distance, relative speed and size information are extracted, the prewarning data (construction area, accident point and the like) issued under the vehicle networking are spatiotemporally matched, the D-S evidence theory is used to fuse the multi-source information, the sudden road condition type (such as a throwaway object and a temporary obstacle) is marked, and the danger level is quantified according to the collision risk, which can include three levels of no risk, low risk and high risk.

[0075] A semantic analysis decision tree is constructed, the key feature confidence, the road condition level and the danger level are used as input nodes, the conventional road condition with a feature confidence lower than a preset threshold and a level less than or equal to 2 is filtered out, and then the remaining data is classified into a plurality of complex road condition types by using a fuzzy clustering algorithm, for example, a heavy damage + high undulation + high risk combination, so that a structured complex road condition type set is formed.

[0076] Figure 3 It is a flowchart for generating structured road condition suspension matching data in the embodiment of the application.

[0077] As shown in Figure 3 , the process of generating the structured road condition suspension matching data includes:

[0078] According to the complex road condition type, the suspension spring stiffness coefficient and the shock absorber damping adjustment range are associated, and the basic suspension adjustment parameter is determined.

[0079] The body posture compensation amount is calculated in combination with the road slope and the undulation, and the suspension height adjustment demand value is derived.

[0080] The anti-roll and anti-hop coefficients of the suspension are calibrated in association with the anti-torsion stiffness of the stabilizer bar and the suspension travel limit value.

[0081] The road condition unique identifier, the suspension adjustment parameter set and the posture compensation amount are packaged as a structured data object, and structured road condition suspension matching data is obtained.

[0082] In the embodiment, a mapping relationship library of complex road condition types and suspension hardware parameters is established, and spring stiffness coefficient intervals and shock absorber damping adjustment ranges are preset for various complex road conditions. Through a fuzzy PID control algorithm, the basic parameters are dynamically selected according to the real-time road condition level, such as high stiffness coefficient and medium-high damping for heavy damage + high undulation road conditions, to form a basic suspension adjustment parameter matrix.

[0083] Based on the road slope and the undulation obtained by the laser radar, the body posture compensation amount is calculated through a dynamics model, and a continuous suspension height adjustment curve is generated by introducing a vehicle speed correction factor.

[0084] The anti-torsion stiffness database of the stabilizer bar is called, the limit value fed back by the suspension travel sensor is combined, and the anti-roll coefficient and the anti-hop coefficient are calibrated through finite element analysis.

[0085] The data is packaged in JSON format: taking the road condition type code as the unique identifier, containing the basic parameter array (stiffness, damping), the posture compensation amount (height, inclination), the performance coefficient set (roll, hop), and marking the data collection timestamp to form structured data.

[0086] The theoretical adjustment calculation module is used to collect historical driving task data and extract road condition occurrence frequency, suspension response efficiency and ride smoothness indicators, and calculate the theoretical suspension adjustment benchmark value according to the road condition occurrence frequency and the suspension response efficiency.

[0087] The process of extracting road condition occurrence frequency, suspension response efficiency and ride smoothness indicators includes:

[0088] Collect historical driving task data and filter historical driving records of the same vehicle model and the same driving area.

[0089] Analyze the road condition types in the historical structured data object, and calculate the occurrence frequency and time length proportion of different complex road conditions.

[0090] Analyze the historical suspension sensor records, build a mapping model of road condition suspension adjustment time, and calculate the suspension response efficiency.

[0091] Collect the vehicle body vertical acceleration and lateral acceleration data in the historical driving process, and calculate the ride comfort index through the vibration comfort formula.

[0092] In this embodiment, the historical driving task data in the past year is collected, the same vehicle model data is screened through the vehicle identification code, the driving records in the same administrative region are extracted combined with the GPS positioning information, and a basic data set is formed. Abnormal values are removed by using a data cleaning algorithm, and complete road conditions and suspension interaction records are retained.

[0093] The road condition type field in the historical structured data object is analyzed, the statistical period is divided according to quarters, the number of occurrences of each complex road condition is calculated through frequency statistics, and the time length proportion of each type of road condition is obtained combined with the total driving time. A road condition frequency distribution table is established, and typical complex road condition types with high frequency are marked.

[0094] The adjustment command and execution feedback data of the historical suspension sensor record are called, the start and completion time of each suspension adjustment is determined through time series analysis, and the adjustment process time is calculated. A mapping model of road conditions and adjustment time is constructed, the ratio of adjustment time and road condition complexity is taken as the quantitative index of suspension response efficiency, and an efficiency evaluation matrix is formed.

[0095] Collect the continuous data of the vehicle body vertical and lateral acceleration sensors during the historical driving process, and extract the acceleration time domain signal according to the road condition type. The signal is calculated by the vibration comfort formula, and the acceleration amplitude and frequency characteristics are converted into the smoothness index. The lower the index, the better the ride smoothness, and accordingly a smoothness benchmark value library under different road conditions is established.

[0096] Integrate the three types of indexes into a multi-dimensional feature vector, associate the corresponding road condition type label, and form a historical performance evaluation data set.

[0097] Figure 4 It is a flowchart for calculating the theoretical suspension adjustment benchmark value in the embodiment of the application.

[0098] As Figure 4 shown, the process of calculating the theoretical suspension adjustment benchmark value according to the road condition frequency and the suspension response efficiency includes:

[0099] According to the road condition frequency weight, the basic suspension adjustment parameters under different road conditions are weighted and calculated to generate an initial adjustment reference value.

[0100] Combined with the suspension response efficiency, a response delay correction coefficient is introduced to adjust the initial adjustment reference value to obtain an adjusted adjustment reference value.

[0101] According to the ride smoothness index, a ride smoothness guarantee coefficient is introduced to optimize the adjusted adjustment reference value to obtain the theoretical suspension adjustment benchmark value.

[0102] Determine the weight coefficient based on the frequency and time length ratio of each road condition, and give higher weight to the road condition with high frequency. Weight the basic suspension adjustment parameters corresponding to different road conditions according to the weight, and generate the initial adjustment reference value covering the whole road condition scene.

[0103] Analyze the suspension response efficiency evaluation matrix, and construct a response delay correction coefficient system according to the adjustment time consumption characteristics under different road conditions. When the suspension response efficiency of a certain type of road condition is low, the corresponding correction coefficient is increased to trigger the adjustment action in advance. Through the multiplication operation of the coefficient and the initial adjustment reference value, the adjustment adjustment reference value optimized by dynamic response is obtained.

[0104] Introduce the ride comfort index as a constraint condition, and convert the comfort index into a protection coefficient. For road condition scenes with high comfort requirements, increase the weight of the protection coefficient to strengthen the vibration suppression effect. Through a multi-objective optimization algorithm, the adjustment adjustment reference value is modified twice, and a balance mechanism is established between response efficiency and comfort, and finally the theoretical suspension adjustment reference value is generated.

[0105] The adjustment map generation module is used to set the dynamic adjustment tolerance band, and generate the initial suspension adjustment map sorted by road condition priority in combination with the theoretical suspension adjustment reference value. The process includes:

[0106] Create a three-dimensional coordinate system of time sequence, road condition and suspension parameter.

[0107] According to the road condition occurrence time sequence and priority, connect the corresponding theoretical suspension adjustment reference value to form the suspension adjustment reference curve.

[0108] According to the road condition danger level and driving speed, set the differential adjustment tolerance band width.

[0109] Visualize the adjustment range boundary and priority identification under different road conditions to form the initial suspension adjustment map.

[0110] Build a three-dimensional coordinate system, and the three dimensions correspond to driving time sequence, road condition type and suspension adjustment parameter respectively. The time axis is divided into intervals according to the actual driving process, and each interval corresponds to the road condition change of a specific road section. The road condition axis is arranged according to type and corresponds to the actual detected road condition type one by one. The parameter axis includes all adjustable technical indicators of the suspension system.

[0111] According to the sequence of various road conditions in vehicle driving, combined with the emergency degree of road condition, mark the theoretical suspension adjustment reference value corresponding to each time in the coordinate system, and then connect these points in turn with a smooth curve to form a continuous suspension adjustment reference curve.

[0112] According to the dangerous degree of the road condition and the current driving speed of the vehicle, different road conditions are set with different width adjustment fault tolerance bands. When the dangerous degree is high or the driving speed is fast, the fault tolerance band is set narrower, requiring more accurate suspension adjustment. When the road condition is normal or the driving speed is low, the fault tolerance band is appropriately widened, retaining a certain adjustment flexibility.

[0113] In the coordinate system, the adjustment range boundaries corresponding to different road conditions are marked with different colors and line styles, and the road conditions with different priorities are distinguished by eye-catching signs. Integrating these elements together, a complete initial suspension adjustment map is formed, which shows the adjustment standards and the allowed fluctuation range that the suspension system should follow under different driving stages and different road conditions.

[0114] The dynamic adjustment generation module is used to collect real-time attitude data of the vehicle, body vibration frequency spectrum and suspension air pressure data to construct an adaptive correction factor matrix, and inject the adaptive correction factor matrix into the initial suspension adjustment map to obtain a dynamic suspension adjustment map for the driving task.

[0115] Figure 5 is a flowchart of constructing an adaptive correction factor matrix in an embodiment of the present application.

[0116] As shown in Figure 5 , the process of constructing the adaptive correction factor matrix includes:

[0117] According to the real-time attitude data of the vehicle, the body roll angle and the pitch angle change rate are calculated to generate an attitude correction coefficient.

[0118] According to the body vibration frequency spectrum, the vibration energy entropy value is calculated to calibrate the damper damping correction coefficient.

[0119] The fluctuation amplitude of the suspension air pressure data is analyzed to generate a pressure compensation correction coefficient.

[0120] The attitude correction coefficient, the damping correction coefficient and the pressure compensation correction coefficient are integrated to obtain the adaptive correction factor matrix.

[0121] The three-dimensional attitude information of the vehicle body is collected by the on-board gyroscope and the inclination sensor, and the real-time change data of the body roll angle and the pitch angle are extracted. The sliding window algorithm is used to smooth the continuously collected angle data to eliminate transient jitter interference, and then the change rate of the roll angle and the pitch angle is calculated by difference operation. According to the deviation degree of the change rate and the preset attitude stability threshold, the attitude correction coefficient is generated, and when the change rate exceeds the stability threshold, the coefficient is adjusted in the direction of enhancing the stability of the suspension. When the change rate is in the stable interval, the coefficient remains the base value, ensuring that the attitude correction is accurately matched with the dynamic state of the vehicle body.

[0122] The vibration sensor installed on the key parts of the vehicle body (such as the chassis and seat rails) collects vibration signals in different frequency bands, and converts the time-domain vibration data into frequency-domain spectrum through fast Fourier transform. The energy proportion of each frequency component in the spectrum is calculated, and then the vibration energy entropy value is obtained. The higher the energy entropy value, the more disordered the vibration distribution and the worse the vehicle body smoothness. According to the difference between the energy entropy value and the ideal smoothness state, the damper damping correction coefficient is calibrated: when the entropy value increases, the coefficient adjusts in the direction of increasing damping and suppressing vibration, and vice versa to reduce the damping to improve comfort.

[0123] The pressure sensor in the suspension gas circuit collects the gas pressure value in real time, records the fluctuation of the gas pressure during continuous driving, and calculates the maximum fluctuation amplitude of the gas pressure per unit time. Combined with the ideal working range of the suspension gas pressure, the influence of the fluctuation amplitude on the suspension support performance is judged. When the fluctuation amplitude exceeds the normal range, it indicates that the gas pressure stability is insufficient, and a pressure compensation correction coefficient needs to be generated. The working state of the gas pressure regulating valve is adjusted through the coefficient to reduce the fluctuation and ensure that the suspension gas pressure is maintained in the ideal range.

[0124] The posture correction coefficient, damper damping correction coefficient and pressure compensation correction coefficient are arranged according to the preset dimension, and are structured and packaged in matrix form. The row dimension of the matrix corresponds to different vehicle body dynamic states (such as steering, braking and jolting), and the column dimension corresponds to the correction coefficient type, forming an adaptive correction factor matrix. After the matrix is generated, it is synchronized to the suspension control unit in real time, providing correction basis for updating the subsequent dynamic adjustment map, and realizing adaptive optimization of the suspension adjustment strategy.

[0125] The process of obtaining the dynamic suspension adjustment map for the driving task includes:

[0126] According to the adaptive correction factor matrix, adjust the suspension adjustment reference value in the initial map for each road condition.

[0127] According to the damping correction coefficient, dynamically adjust the adjustment fault tolerance bandwidth under different road conditions.

[0128] Superimpose the correction factor value marker layer and the real-time road warning identifier on the initial suspension adjustment map.

[0129] Establish a real-time data interface with the vehicle CAN bus, and update the adaptive correction factor matrix and the dynamic suspension adjustment map according to the preset period.

[0130] In the process of obtaining the dynamic suspension adjustment map, first, based on the constructed adaptive correction factor matrix, the reference value in the initial suspension adjustment map is adjusted for each road condition. For each complex road condition type, the corresponding posture correction coefficient and pressure compensation correction coefficient are extracted from the adaptive correction factor matrix, and the suspension adjustment reference value in the initial map is operated and fused. For example, in the continuous curve road condition, the suspension stiffness reference value is adjusted in combination with the posture correction coefficient to enhance the body anti-roll capability. In the bumpy road condition, the suspension height reference value is optimized according to the pressure compensation correction coefficient to improve the road adaptability and ensure that the reference value corresponding to each road condition can match the current vehicle dynamic state.

[0131] For road conditions with high vibration energy entropy, such as gravel roads and damaged roads, the damping correction coefficient tends to enhance the damping effect, and at the same time, the synchronous narrowing adjustment tolerance bandwidth is adjusted, and the fluctuation range of the shock absorber damping is strictly controlled to avoid affecting the vibration suppression effect due to damping adjustment deviation. In the road condition with gentle vibration, the damping correction coefficient tends to be moderate, and the tolerance bandwidth is appropriately relaxed to retain certain adjustment flexibility, balancing comfort and handling.

[0132] Add a correction factor value label layer to mark each type of coefficient in the adaptive correction factor matrix next to the adjustment reference value of the corresponding road condition in a visual form, which facilitates intuitive viewing of the correction basis. Integrate real-time road condition warning marks to mark warning symbols in the corresponding period in the map according to the temporary road condition information (such as sudden construction, icy road surface) issued by the Internet of Vehicles, and remind the system to focus on the suspension adjustment accuracy of that section.

[0133] Through a special communication protocol, the system is connected to the CAN bus for data, and the latest vehicle posture, vibration, air pressure, and other real-time data are obtained from the bus at a preset period to recalculate the adaptive correction factor matrix. At the same time, the updated matrix is injected into the suspension adjustment map to complete the dynamic iteration of the map. The entire process ensures that the dynamic suspension adjustment map can reflect the vehicle operating state and road condition changes in real time, providing accurate and timely adjustment basis for the suspension system.

[0134] The real-time monitoring module is used to collect real-time suspension working data curves during driving, compare the actual suspension adjustment mean value in each road condition section, and calculate the shape similarity between the real-time suspension working data curve and the dynamic suspension adjustment map in each road condition time window to monitor the intelligent suspension control in real time. The process includes:

[0135] Extract the road condition duration from the structured road condition suspension matching data to determine the time window boundary corresponding to each complex road condition and obtain the map reference curve.

[0136] Data preprocessing is performed on the real-time suspension working data curve and the map reference curve.

[0137] The dynamic time warping algorithm is used to calculate the shape similarity of the real-time suspension working data curve and the reference curve of the atlas after pretreatment.

[0138] The Pearson correlation coefficient is introduced as an auxiliary verification index to calculate the linear correlation degree of the real-time suspension working data curve and the reference curve of the atlas.

[0139] The calculated shape similarity result and the actual suspension adjustment average of each road condition section are stored in the system database.

[0140] When the real-time monitoring module carries out monitoring work, the duration information of various complex road conditions is extracted from the structured road condition suspension matching data. Combined with the switching nodes of the real-time road condition during vehicle driving, the time window boundaries corresponding to each complex road condition are determined based on the time points of the beginning and end of the road condition. According to these boundaries, the reference curve segment of the corresponding period is intercepted from the dynamic suspension adjustment atlas, ensuring that each atlas reference curve can accurately match the current monitoring road condition period, establishing a unified time dimension benchmark for subsequent curve comparison.

[0141] For the real-time suspension working data curve, the filter algorithm is used to eliminate the high-frequency noise mixed in the sensor collection process, avoiding the influence of instantaneous interference data on the curve shape accuracy. At the same time, the curve is trend-smoothed to retain the true change trend of the suspension working state. For the reference curve of the atlas, interpolation adjustment is performed according to the sampling frequency of the real-time data to ensure the consistency of the sampling density of the two curves, eliminating the comparison deviation caused by the difference in data collection frequency, and laying a data foundation for subsequent similarity calculation.

[0142] First, the two curves are decomposed into multiple data feature points, a distance matrix between the feature points is constructed, and the optimal matching path between the two curves is found through dynamic programming, which can best fit the shape change trend of the two curves. According to the distance sum of each feature point on the optimal path, an intuitive shape similarity index is obtained, and the higher the index value, the closer the shape of the real-time suspension working data curve and the reference curve of the atlas, and the higher the consistency of the actual working state of the suspension and the ideal adjustment state.

[0143] The Pearson correlation coefficient is obtained by calculating the covariance of the real-time suspension working data curve and the reference curve of the atlas, combined with the standard deviation of the two curves. This coefficient can reflect the linear correlation degree of the two curves, and if the absolute value of the coefficient is high, it means that the overall change trend of the two curves is highly consistent, which verifies the reliability of the shape similarity result from the linear correlation perspective. If the coefficient deviates from the shape similarity index, the system will trigger preliminary troubleshooting of the data anomaly reason to ensure the rigor of the monitoring result.

[0144] Finally, the calculated morphological similarity results are associated with the actual suspension adjustment average of each road condition segment and stored. The corresponding relationship including road condition type, time window, similarity, and actual adjustment average is normalized to a structured record and written into the system database in real time.

[0145] In summary, the embodiment provides an intelligent suspension control system based on complex road condition recognition. Through multi-source road condition data fusion and semantic analysis, the road surface image, three-dimensional profile, obstacle, and early warning information can be comprehensively captured, various complex road conditions can be accurately identified and classified, the recognition deviation caused by a single sensor data can be avoided, the accurate road condition basis for subsequent suspension adjustment is provided, and the suspension adjustment direction is highly matched with the actual road condition demand.

[0146] The road condition frequency, response efficiency, and smoothness index extracted in combination with historical driving data are used to calculate the theoretical adjustment reference value, and a self-adaptive correction factor matrix is introduced. The adjustment reference and fault tolerance bandwidth are dynamically adjusted according to the real-time attitude of the vehicle, the body vibration, and the suspension air pressure, forming a dynamic suspension adjustment atlas. The suspension parameters can be adapted to the road condition changes and the vehicle state in real time. In the bumpy road section, the vibration can be effectively suppressed by optimizing the damping and height parameters, and the comfort is improved. In the curve or sudden road condition, the anti-roll and support performance can be enhanced, the control stability is ensured, and the dynamic balance of comfort and control is realized.

[0147] Through the combination of dynamic time warping algorithm and Pearson correlation coefficient, the morphological similarity of real-time suspension working curve and dynamic atlas is calculated in detail. Not only the obvious adjustment abnormalities can be quickly identified, but also the subtle morphological deviations can be captured, and the accurate monitoring of the suspension working state is realized. At the same time, the monitoring data and the actual adjustment average are associated and stored, which provides comprehensive data support for subsequent suspension adjustment strategy iteration, fault diagnosis, and parameter optimization, forms a closed-loop system, continuously improves the adaptation ability and reliability of the suspension system, further enhances the comprehensive performance of the vehicle in complex driving scenarios, and provides users with a better driving experience.

[0148] Through the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be implemented by means of software and the necessary general hardware platform, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the method described in each embodiment or some parts of the embodiment.

[0149] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit the same; and although the present application has been described in detail with reference to the foregoing embodiments, it should be appreciated by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features thereof can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent suspension control system based on complex road condition recognition, characterized in that, include: The data acquisition module is used to acquire multi-source road condition data during vehicle operation, identify complex road condition types in the multi-source road condition data through road condition semantic analysis, and extract suspension adaptation features for each road condition by combining vehicle suspension parameters to generate structured road condition suspension matching data. The multi-source road condition data includes road surface image data, road surface three-dimensional contour data, road obstacle data, and road segment condition warning data. The vehicle suspension parameters include suspension spring stiffness coefficient, shock absorber damping adjustment range, stabilizer bar torsional stiffness, suspension travel limit value, and spring pressure adjustment threshold. The suspension adaptation features include suspension stiffness target value, shock absorber damping coefficient set value, vehicle height adjustment value, suspension response delay threshold, and vibration suppression rate target value. The process of generating structured road condition suspension matching data includes: Based on the complex road condition type, the suspension spring stiffness coefficient and the shock absorber damping adjustment range are correlated to determine the basic suspension adjustment parameters; By combining road slope and undulation, the vehicle body posture compensation amount is calculated, and the required value for suspension height adjustment is derived. By correlating the torsional stiffness of the stabilizer bar with the suspension travel limit value, the anti-roll and anti-bump coefficients of the suspension are calibrated. The unique road condition identifier, suspension adjustment parameter set, and attitude compensation amount are encapsulated into a structured data object to obtain structured road condition suspension matching data; The theoretical adjustment calculation module is used to collect historical driving task data and extract road condition occurrence frequency, suspension response efficiency and driving comfort index, and calculate the theoretical suspension adjustment benchmark value based on the road condition occurrence frequency and the suspension response efficiency; The adjustment map generation module is used to set the dynamic adjustment tolerance zone and generate an initial suspension adjustment map sorted by road condition priority in combination with the theoretical suspension adjustment benchmark value. The dynamic adjustment generation module is used to collect real-time vehicle attitude data, body vibration spectrum and suspension air pressure data to construct an adaptive correction factor matrix, and inject the adaptive correction factor matrix into the initial suspension adjustment map to obtain the dynamic suspension adjustment map for the driving task. The real-time monitoring module is used to collect real-time suspension working data curves during driving, compare the actual average suspension adjustment value in each road condition segment, and perform real-time monitoring of intelligent suspension control by calculating the morphological similarity between the real-time suspension working data curve in each road condition time window and the dynamic suspension adjustment map.

2. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of identifying complex road condition types in the multi-source road condition data through road condition semantic analysis includes: The road surface image data is segmented and feature extracted to identify key road condition features, including road surface damage, water accumulation, snow accumulation, and gravel. Based on the three-dimensional contour data of the road surface, the road surface slope, undulation and bump frequency are calculated, and the road condition level is classified. By integrating the road obstacle data and road condition early warning data, emergency road condition information is marked, including the type and hazard level of the emergency road condition. Based on the key road condition characteristics, road condition levels, and sudden road condition information, regular smooth road conditions are filtered out, and a set of complex road condition types is obtained.

3. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of extracting indicators such as road condition frequency, suspension response efficiency, and ride comfort includes: Collect historical driving task data and filter historical driving records of the same vehicle type and driving area; Analyze the road condition types in historical structured data objects and count the frequency and duration of different complex road conditions. By analyzing historical suspension sensor records, a mapping model of suspension adjustment time under road conditions is constructed to calculate suspension response efficiency. Collect data on the vertical and lateral acceleration of the vehicle body during historical driving processes, and calculate the ride comfort index using the vibration comfort formula.

4. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of calculating the theoretical suspension adjustment benchmark value based on the frequency of road conditions and the suspension response efficiency includes: Based on the frequency of road conditions, the basic suspension adjustment parameters under different road conditions are weighted and calculated to generate initial adjustment reference values. Based on the suspension response efficiency, a response delay correction coefficient is introduced to adjust the initial adjustment reference value to obtain the adjustment reference value; Based on the aforementioned ride comfort index, a ride comfort guarantee coefficient is introduced to optimize the adjustment reference value, thereby obtaining the theoretical suspension adjustment benchmark value.

5. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of generating an initial suspension adjustment map sorted by road condition priority based on the theoretical suspension adjustment benchmark values ​​includes: Create a three-dimensional coordinate system for time series, road conditions, and suspension parameters; Based on the timing and priority of road conditions, the corresponding theoretical suspension adjustment benchmark values ​​are connected to form a suspension adjustment reference curve; Differentiated adjustment fault tolerance bandwidth is set based on road condition hazard level and driving speed; Visualize and label the adjustment range boundaries and priority indicators under different road conditions to form an initial suspension adjustment map.

6. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of constructing the adaptive correction factor matrix includes: Based on the real-time vehicle attitude data, calculate the rate of change of vehicle roll angle and pitch angle, and generate attitude correction coefficients. Based on the vehicle body vibration spectrum, calculate the vibration energy entropy value and calibrate the damper damping correction coefficient; Analyze the fluctuation range of the suspension air pressure data to generate a pressure compensation correction coefficient; By integrating attitude correction coefficients, damping correction coefficients, and pressure compensation correction coefficients, an adaptive correction factor matrix is ​​obtained.

7. The intelligent suspension control system based on complex road condition recognition according to claim 6, characterized in that, The process of obtaining the dynamic suspension adjustment map for the driving task includes: Based on the adaptive correction factor matrix, the suspension adjustment reference value in the initial map is adjusted for each road condition. Based on the aforementioned damping correction coefficient, the adjustable fault-tolerant bandwidth is dynamically expanded or contracted under different road conditions; A correction factor numerical marker layer and a real-time road condition warning label are superimposed on the initial suspension adjustment map; Establish a real-time data interface with the vehicle's CAN bus and update the adaptive correction factor matrix and dynamic suspension adjustment map according to a preset cycle.

8. The intelligent suspension control system based on complex road condition recognition according to claim 1, characterized in that, The process of calculating the morphological similarity between the real-time suspension operating data curve and the dynamic suspension adjustment map within each road condition time window includes: The duration of road conditions is extracted from the structured road condition suspension matching data, the time window boundary corresponding to each complex road condition is determined, and a map reference curve is obtained. Data preprocessing is performed on the real-time suspension working data curves and the reference curves. The dynamic time warping algorithm is used to calculate the morphological similarity between the preprocessed real-time suspension working data curve and the reference curve. Pearson correlation coefficient was introduced as an auxiliary verification index to calculate the linear correlation between the real-time suspension working data curve and the reference curve. The calculated morphological similarity results are stored together with the average actual suspension adjustment value for each road condition segment in the system database.

Images