Automobile armrest dynamic damping adjusting system and method based on multi-sensor fusion

Abstract

The invention discloses an automobile armrest dynamic damping adjustment system and method based on multi-sensor fusion, particularly relates to the technical field of automobile armrest adjustment, and is used for solving the problem of poor armrest damping self-adaptive adjustment precision. Multi-source data such as vehicle postures, power behaviors, contact behaviors and hand torque are collected, resampling, denoising and phase compensation are carried out, a fusion state vector is constructed, the future state of the vehicle is predicted based on a short-time historical state sequence, state increment is extracted, and a dynamic risk index is formed; a current state and a risk trend are combined to construct an optimization model integrating supporting, smoothing and restraining, controller parameters are dynamically adjusted through a fuzzy reasoning mechanism, self-adaptive adjustment of armrest damping is achieved, and in the damping execution process, cooperative features and dynamic correction strategies of a control instruction and feedback output are analyzed, so that the damping performance is improved. Therefore, the response precision and stability of the dynamic damping of the automobile armrest under the complex working condition are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile armrest damping adjustment, and more particularly to an automobile armrest dynamic damping adjustment system and method based on multi-sensor fusion. Background Art

[0002] With the rapid development of urban public transportation, buses frequently start, slow down, and turn during driving, especially in the event of sudden braking or complex road conditions. Passengers on board face a high risk of body posture imbalance and falls. For vertical handrails at the aisle steps, due to the combined effects of ground height changes and vehicle body shaking, passengers often need to rely on handrails for stable support when getting on the bus, walking, or changing positions. However, existing handrail systems are mostly rigid structures that lack dynamic response capabilities and are unable to adjust damping according to the real-time driving status of the vehicle and the passenger's gripping behavior. As a result, when the vehicle suddenly stops or accelerates, the passenger's hand is subjected to excessive instantaneous impact force, which can easily cause unstable grip, body displacement, or even fall, posing an obvious safety hazard.

[0003] Deficiencies in existing technologies: Most automotive armrest damping systems use a fixed damping design or a linear adjustment method based on single sensor feedback. These systems are unable to adapt to complex and changing driving conditions and driver behavior in real time, leading to problems such as poor human-machine interaction comfort, delayed posture support response, and insufficient fatigue compensation. Under dynamic conditions such as cornering, acceleration, and bumps, passengers may encounter sudden inertial disturbances, and fixed or roughly adjusted damping systems struggle to provide stable support, and may even exacerbate hand slippage and operational errors. Furthermore, there is a general lack of feedforward perception and predictive adjustment capabilities for the vehicle's future motion trends. Consequently, the armrest damping response cannot be adjusted in advance based on upcoming dynamic risks (such as sharp turns and potholes), which can easily lead to adjustment lags and reduce passenger handling stability. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a vehicle armrest dynamic damping adjustment system and method based on multi-sensor fusion to solve the problem of poor accuracy of adaptive adjustment of armrest damping in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A vehicle armrest dynamic damping adjustment method based on multi-sensor fusion includes the following steps:

[0007] Collect vehicle posture, dynamic behavior, contact behavior, and hand torque data, resample and pseudo-synchronize the collected data, use structure-preserving residual filtering to suppress high-frequency interference, perform nonlinear normalization and response phase compensation, and output a fused state vector;

[0008] Based on the fused state vector, the current vehicle speed, posture, and dynamic variables are extracted for modeling. The future state is predicted based on the short-term historical sequence. Based on the state increment and change trend, a dynamic risk index is formed to determine the disturbance intensity and trend evolution of the current vehicle operation.

[0009] Based on the vehicle's current state and predicted trends, an optimization function for mechanical support, response smoothness, and risk suppression is constructed. The controller parameters are dynamically adjusted based on the fuzzy inference mechanism to regulate the armrest damping output.

[0010] During the real-time adjustment of the damper, the collaborative characteristic information generated by the command behavior and mechanical feedback during the adjustment process is analyzed, and the strategy is adjusted based on the analysis results.

[0011] In a preferred embodiment, vehicle posture, power behavior, contact behavior and hand torque data are collected, and the collected data are resampled and pseudo-synchronized. The specific process is as follows:

[0012] Deploy a steering wheel angle sensor to obtain vehicle lateral steering trend data; deploy a three-axis acceleration sensor to sense the vehicle's longitudinal, lateral, and vertical dynamic behavior data; deploy a pressure distribution matrix sensor to obtain hand posture, grip strength, and center of gravity offset data; deploy an armrest torque sensor to obtain active or passive arm adjustment process data;

[0013] Construct a reference time axis as a unified sampling reference frame to resample signal streams with inconsistent original sampling frequencies;

[0014] All processed signals after resampling are aligned to form a pseudo-synchronous data frame.

[0015] In a preferred embodiment, a structure-preserving residual filter is used to suppress high-frequency interference, perform nonlinear normalization and response phase compensation, and output a fusion state vector. The specific process is as follows:

[0016] Use structure-preserving residuals for filtering, define the original and estimated signal residuals, and construct a filtering objective function to retain the sudden hand force and vehicle shaking signals and output the noise-suppressed signal;

[0017] The noise suppressed signal is mapped using a nonlinear monotonic mapping to output a normalized signal set;

[0018] Dynamic phase compensation based on the response estimation model is applied to the normalized signal set. After all signal channels are compensated, a fusion state vector is output.

[0019] In a preferred embodiment, the current vehicle speed, posture, and dynamic variables are extracted based on the fused state vector for modeling, and the future state is predicted based on the short-term historical sequence. The specific steps are as follows:

[0020] The output fusion state vector is converted into a vehicle dynamic state representation to form a set of state variables including the vehicle's lateral and longitudinal posture and dynamic characteristics. The selected vehicle state variables include current speed, lateral acceleration, heading angular velocity, vertical acceleration, and acceleration change rate;

[0021] Carry out vehicle state variable mapping modeling and transform vehicle state variables into dynamic state variables, including armrest pressure concentration to determine driving posture tilt, steering wheel angle to determine yaw trend, and body acceleration signal to restore the vehicle's real-time inertial behavior;

[0022] Carry out state evolution trend prediction to predict the changing trend of vehicle state in the future.

[0023] In a preferred embodiment, a state evolution trend prediction is performed to predict the changing trend of the vehicle state in the future. The specific process is as follows:

[0024] Construct a time window and collect the vehicle state vectors at the most recent sampling moment into a set of historical sequences;

[0025] Construct a long short-term memory network model to extract temporal correlation and state transition patterns in historical sequences;

[0026] The state value is predicted through the state transfer equation, and the state derivative and change trend are derived.

[0027] In a preferred embodiment, a dynamic risk index is formed based on the state increment and change trend to determine the disturbance intensity and trend evolution of the current vehicle operation. The specific process is as follows:

[0028] The state difference between the current moment and the predicted moment is used as the state increment value;

[0029] Define risk indicators to measure the extent to which the current system is on the brink of dynamic change;

[0030] The obtained state increment value is used to identify whether there is an obvious trend change at present, and the controller is used to adjust the strength, delay and mode of the control response according to the dynamic risk index.

[0031] In a preferred embodiment, based on the current state of the vehicle and the predicted trend, an optimization function for mechanical support, response smoothness, and risk suppression is constructed, and the controller parameters are dynamically adjusted based on the fuzzy inference mechanism to adjust the armrest damping output. The specific process is as follows:

[0032] The mechanical support index refers to whether the armrest needs to provide additional stabilizing support when the vehicle turns, accelerates, or bumps. The damping response smoothness index refers to the driver discomfort caused by damping changes. The dynamic risk control index refers to whether the system determines that the vehicle state is in a stage of drastic fluctuations and should quickly increase the damping to cope with possible external impacts.

[0033] An optimization function is set to optimize mechanical support, response smoothness, and risk suppression, and a fuzzy reasoning mechanism is introduced to adjust the parameters of the PID controller in real time.

[0034] The fuzzy inputs of the PID controller are the damping error, error change rate, and vehicle state trend;

[0035] After the fuzzy reasoning outputs the PID parameters, they are brought into the PID controller to analyze and calculate the handrail damping.

[0036] In a preferred embodiment, during the real-time adjustment process of the damper, the collaborative feature information generated by the command behavior and the mechanical feedback during the adjustment process is analyzed. The specific process is as follows:

[0037] Obtain the response behavior sequence between the control signal, current signal and damping output, and analyze the collaborative characteristic information between the command behavior and mechanical feedback during the adjustment process;

[0038] The collaborative feature information includes the response monotonic violation count index, and the collaborative feature information includes the error cross-threshold persistence index;

[0039] The response monotonic violation count index indicates whether the direction of change of the control command and the actual output force is consistent during the damping adjustment process;

[0040] The error threshold persistence index indicates the extent to which the damping control error continues to exceed the acceptable range during the operating cycle.

[0041] In a preferred embodiment, the strategy is adjusted according to the analysis results. The specific process is as follows:

[0042] Set the response monotonic violation threshold and the response monotonic violation count index for comparative analysis, and set the error cross-threshold persistence threshold and the error cross-threshold persistence index for comparative analysis;

[0043] When the response monotonic violation count index and the error threshold-crossing persistence index do not exceed the corresponding set thresholds, the current fuzzy controller parameters and output logic are maintained without additional intervention;

[0044] If only the response monotonic violation count index exceeds the threshold, the response consistency optimization strategy is triggered to modify the control mapping function, recalibrate the electronic control drive curve, or temporarily introduce direction restriction logic to prevent high-frequency reverse;

[0045] If only the error cross-threshold persistence index exceeds the error cross-threshold persistence threshold, the steady-state error enhancement compensation strategy is activated to increase the controller's proportional adjustment capability, adjust the fuzzy rule response range, or enable the error recovery interval speed limit control to suppress error accumulation;

[0046] When both the response monotonic violation count index and the error cross-threshold persistence index exceed their respective thresholds, the global protection strategy is activated to lock the control frequency, limit the maximum damping change rate, and record the current status data for diagnostic analysis.

[0047] A vehicle armrest dynamic damping adjustment method based on multi-sensor fusion is used to implement the above-mentioned vehicle armrest dynamic damping adjustment method based on multi-sensor fusion, including:

[0048] The armrest multi-source data acquisition module is used to collect vehicle posture, dynamic behavior, contact behavior, and hand torque data. It resamples and pseudo-synchronizes the collected data, uses structure-preserving residual filtering to suppress high-frequency interference, performs nonlinear normalization and response phase compensation, and outputs a fused state vector.

[0049] The operation trend analysis module is used to extract the current vehicle speed, posture and dynamic variables based on the fused state vector for modeling, predict the future state based on the short-term historical sequence, and form a dynamic risk index based on the state increment and change trend to determine the current vehicle operation disturbance intensity and trend evolution;

[0050] The armrest damping control module is used to construct an optimization function for mechanical support, response smoothness, and risk suppression based on the vehicle's current state and predicted trends. It also dynamically adjusts controller parameters and adjusts the armrest damping output based on a fuzzy inference mechanism.

[0051] The strategy adjustment module is used to analyze the collaborative feature information generated by the command behavior and mechanical feedback during the real-time adjustment of the damper, and adjust the strategy according to the analysis results.

[0052] Technical effects and advantages of the present invention:

[0053] The present invention collects multi-source sensor signals from the vehicle's driving status and the passenger's gripping behavior to perceive in real time the vehicle's dynamic characteristics under acceleration, deceleration, cornering, or bumpy road conditions. It then constructs a fused state vector based on data such as hand posture, grip force distribution, and arm torque. Based on this, it implements intelligent adjustment of the handrail's damping. Compared to traditional fixed-rigidity mechanical handrail structures, the present invention has the following beneficial effects:

[0054] Firstly, the system can predict in advance the risk of posture imbalance that passengers may experience based on the vehicle's movement trends in situations such as sudden braking, starting, or uneven roads. It can also proactively provide flexible support by adaptively adjusting the damping force of the armrest, significantly reducing the probability of hand slippage or body toppling due to inertial impact, thereby improving driving safety.

[0055] Secondly, the present invention adopts a control strategy that combines fuzzy logic with state prediction, so that the handrail can distinguish whether the passenger is actively holding it or passively relying on support, dynamically adjust the damping response speed and strength, and realize personalized support capabilities that vary according to the person, time, and vehicle conditions, thereby enhancing the comfort and stability of the human-computer interaction of the entire vehicle. The use of signal processing technologies such as dynamic phase compensation and residual filtering can effectively suppress the influence of high-frequency vibrations, electromagnetic interference, etc. generated during the operation of the bus on the sensor data, thereby ensuring the real-time and accuracy of the damping control. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a flow chart of a method for adjusting dynamic damping of an automobile armrest based on multi-sensor fusion.

[0057] Figure 2 The figure is a schematic structural diagram of the automobile armrest dynamic damping adjustment system based on multi-sensor fusion of the present invention. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] Example 1: Figure 1 As shown, the dynamic damping adjustment method of the automobile armrest based on multi-sensor fusion includes the following steps:

[0060] Collect vehicle posture, dynamic behavior, contact behavior, and hand torque data, resample and pseudo-synchronize the collected data, use structure-preserving residual filtering to suppress high-frequency interference, perform nonlinear normalization and response phase compensation, and output a fused state vector;

[0061] Based on the fused state vector, the current vehicle speed, posture, and dynamic variables are extracted for modeling. The future state is predicted based on the short-term historical sequence. Based on the state increment and change trend, a dynamic risk index is formed to determine the disturbance intensity and trend evolution of the current vehicle operation.

[0062] Based on the vehicle's current state and predicted trends, an optimization function for mechanical support, response smoothness, and risk suppression is constructed. The controller parameters are dynamically adjusted based on the fuzzy inference mechanism to regulate the armrest damping output.

[0063] During the real-time adjustment of the damper, the collaborative characteristic information generated by the command behavior and mechanical feedback during the adjustment process is analyzed, and the strategy is adjusted based on the analysis results.

[0064] Step 1: Collect and pre-process multi-source vehicle data. The specific process is as follows:

[0065] State data is collected from multiple sources (vehicle posture, dynamic behavior, contact behavior, and micromechanical state). The following sensors are deployed, including a steering wheel angle sensor to detect the vehicle's lateral steering tendency; a three-axis accelerometer to sense the vehicle's longitudinal, lateral, and vertical dynamic behavior; a pressure distribution matrix sensor (placed on the armrest surface) to sense behavioral factors such as hand posture, grip force, and center of gravity offset; and an armrest torque sensor to identify active or passive arm adjustment processes. This includes armrest surface pressure distribution, contact duration recording, and grip switching detection logic. This sensor is used to identify whether the passenger is holding the armrest, the hand force application area, and the frequency and rhythm of armrest grasping, thereby constructing behavioral characteristic parameters reflecting the passenger's degree of dependence.

[0066] The operating frequencies of multiple sensors vary significantly. For example, steering wheel angle sensors often sample at low frequencies (10–20 Hz), while acceleration and pressure matrices can reach hundreds of Hz. Furthermore, the response delays of each sensor vary, so uniform resampling and pseudo-synchronization are required for nonlinear time series.

[0067] Construct a reference time axis T = {t0, t1, ..., t k}, as the unified sampling reference frame of the entire control system, resamples the signal stream with inconsistent original sampling frequency, and resamples any channel s i (t), and use window interpolation to construct the resampled value: Among them, κ i (·) is the interpolation kernel of channel i, which supports nonlinear sampling recovery (such as B-spline or local support trigonometric function), r is the interpolation window width; t j is the reference time point;

[0068] All processed signals after resampling are aligned to form a pseudo-synchronous data frame;

[0069] Vehicle motion and road disturbances introduce high-frequency vibration and electromagnetic interference noise, especially acceleration and pressure signals. Direct processing will seriously affect state modeling and damping adjustment misjudgment, requiring strong noise suppression. However, sudden changes in driving behavior (such as sharp turns and skidding) must be retained and cannot be filtered out as anomalies. Therefore, denoising filtering is performed based on residual constraints. The specific steps are as follows:

[0070] Using the structure-preserving residual regularized filtering strategy, the residuals of the original and estimated signals are defined as: in, is the original signal value after interpolation or alignment of the i-th channel at time t, Represents the estimated value or filtering result of the i-th channel at time t;

[0071] Construct the filtering objective function: Among them, C 2 x(t) is the second-order difference of the signal, reflecting the local curvature mutation of the signal; λ is the adjustment factor between the residual and the structural smoothness;

[0072] Suppress unstructured noise, retain physical event signals (such as sudden hand force and vehicle shaking), and output noise-suppressed signals.

[0073] The physical dimensions of different signals are inconsistent (such as acceleration and torque), and the signal amplitudes differ by several times. Directly inputting the controller will cause the dominant dimension to dominate the control mechanics, which can easily cause numerical explosion or suppress the response. It is necessary to introduce a unified amplitude expression and retain the nonlinear dynamic characteristics of the signal. Use nonlinear monotonic mapping to output the normalized signal set;

[0074] Dynamic phase compensation and time alignment are performed on the normalized signal. Dynamic phase compensation based on the response estimation model is adopted, and the sensor response delay is modeled as: Among them, β i is the maximum response delay of the channel, measured by the device, YI i is the response time.

[0075] Perform compensation operation, the expression is: in, To normalize the signal, all signal channels are compensated and the fusion state vector is output.

[0076] Step 2: Vehicle state prediction. This involves building a short-term evolution model of vehicle operation dynamics based on the multimodal fusion state vector pre-processed in step 1. This allows for prediction of vehicle state trends within a short period of time (e.g., 1–2 seconds). The specific process is as follows:

[0077] Based on the input of multimodal fusion (fused state vector), a continuous-time state evolution model of the vehicle's dynamic behavior is constructed to predict the state at the next moment. The output fused state vector is converted into a representation of the vehicle's dynamic state, forming a set of state variables that cover the vehicle's lateral and longitudinal posture and dynamic characteristics. The selected vehicle state variables include but are not limited to the following: current speed (vehicle speed), lateral acceleration (used to determine the intensity of the curve and whether the handrail will swing sideways), heading angular velocity (describes the steering trend and determines whether a sharp turn is required), vertical acceleration (reflects bumpy road conditions and is directly related to hand tremor), and acceleration rate of change (reflects vehicle dynamic fluctuations);

[0078] The above variables are combined to form the vehicle state variable of the vehicle at the current moment;

[0079] Carry out vehicle state variable mapping and build a state evolution model. This involves mapping the fused and pre-processed multi-source sensor data into physically meaningful dynamic state variables. This involves transforming vehicle state variables into dynamic state variables. This includes determining driving posture tilt based on armrest pressure concentration, estimating yaw trends based on steering wheel angles, and restoring the vehicle's real-time inertial behavior based on vehicle body acceleration signals.

[0080] For example, the analysis of steering wheel angle and vehicle speed can infer lateral acceleration, and the armrest pressure deviation trend and vertical acceleration can infer the vehicle's bumpy state.

[0081] The vehicle's state evolution trend is predicted, i.e., the vehicle has a continuous evolution trend in complex working conditions. The change trend of the vehicle state in a short period of time in the future is predicted to achieve early adjustment of the damping force. The details are as follows:

[0082] Construct a time window and collect the vehicle state vectors of the latest K sampling moments into a set of historical sequences: H(t) = {x veh (t-K+1),...,x veh (t)}, where x veh (t) is the vehicle state vector at time t;

[0083] It should be noted that the sequence length K is determined according to the sampling frequency and the control response time limit (for example, if 20 samples are taken in 1 second, then K = 20);

[0084] Build a time recursive model or state convolution model to extract the time correlation and state transformation pattern in the historical sequence, and thus predict the state at the next time point: Where P represents the state evolution prediction model, which can adopt recursive neural network, long short-term memory network, etc.

[0085] Taking LSTM as an example, the state value can be predicted through the state transition equation, and the state derivative and change trend can also be derived.

[0086] Damping adjustment is not only related to the current state of the vehicle, but also depends on the speed and intensity of state changes. Therefore, after obtaining the predicted state, the state increment and dynamic risk characteristics are further calculated to determine the state change rate.

[0087] Calculate the state difference between the current time and the predicted time, and determine the state increment value Δx(t). The expression is:

[0088] Define the risk indicator R(t) to measure the degree to which the current system is on the edge of dynamic change: R(t) = ∑ i ρ i ·|Δx i (t)| FD , where ρ i represents the influence weight of each state variable on the damping adjustment (e.g., the lateral acceleration weight is higher than the vertical acceleration); FD is the nonlinear amplification coefficient, which is used to enhance the sudden change response; R(t) is the system dynamic risk index, which is used to classify the response level in the subsequent control stage (e.g., the comparison between ordinary adjustment and emergency control); Δx i (t) represents the state increment value of the i-th vehicle state variable at the current time t.

[0089] Furthermore, while constructing the vehicle's dynamic state vector, passenger interaction behavior characteristics are extracted, such as grip force change rate, handrail contact stability index, arm sliding trend, etc., to determine whether the passenger is currently in an unstable balance state or is relying on the handrail to maintain posture stability; the passenger state mark is used as an auxiliary prediction input to enhance the system's warning and identification capabilities for risk situations.

[0090] The future state prediction value is used for feedforward damping calculation, and the obtained state increment value is used to identify whether there is an obvious trend change at present. Finally, the controller is used to adjust the strength, delay and mode of the control response based on the dynamic risk index.

[0091] Step three: Perform dynamic damping adjustment and mode strategy fusion control. This involves dynamically adjusting the electronically controlled damper of the vehicle armrest based on the fused state information and predicted state trends, achieving intelligent responses to different driving conditions, driver operating behaviors, and vehicle motion disturbances. The specific process is as follows:

[0092] Dynamically adjust the damping response mode based on whether the passenger's grip is active or passive. For example, if the passenger is actively applying force, the damping response is appropriately weakened to avoid restricting the freedom of movement; if the passenger is passively relying on support, the damping force is increased to improve the stability and support effect. The passenger state and the vehicle dynamic state are used as inputs to the control optimization function to improve the adaptability and flexibility of the overall adjustment.

[0093] Three types of indicators are evaluated, including mechanical support requirements, which refers to whether the armrests need to provide additional stabilizing support when the vehicle turns, accelerates, or bumps; damping response smoothness, which refers to the discomfort caused to passengers by damping changes; and dynamic risk control, which refers to the system's determination that the vehicle's state is experiencing drastic fluctuations and that damping should be increased rapidly to cope with possible external impacts.

[0094] An optimization objective function J(t) is defined that integrates these three types of control objectives. It minimizes the difference between the current damping behavior and the desired state through constraints, while limiting the damping change rate and strengthening the constraint response in high-risk states.

[0095] Set the optimization function and optimize the problem as follows: Among them, τ des (t) is the expected armrest support force based on the predicted vehicle state and the pressure concentration area; τ actual (t) is the actual reaction force calculated from the current damping D(t) and the displacement velocity of the handrail; is the damping change rate; Z(R(t)) is the dynamic risk adjustment function (for example, exponentially improving high-risk response); α and β are the adjustment factors for response smoothness and risk suppression in the controller.

[0096] The fuzzy reasoning mechanism is introduced to adjust the parameters of the PID controller in real time, giving the controller dynamic adaptability. By analyzing the current damping error and vehicle state change trend, multiple fuzzy inputs (such as error amplitude, error change rate, vehicle dynamic intensity) are defined and mapped to the PID parameters (K p , K i , K d ) to dynamically adjust the output and realize the fuzzy self-adaptation of controller parameters;

[0097] The fuzzy inputs are defined as damping error, error change rate, and vehicle status trend;

[0098] The controller core form is: Where e(t) = τ des (t)-τ actual (t) is the damping deviation, K p , K i , K d The proportional, integral and differential coefficients are dynamically adjusted by fuzzy rules;

[0099] Assume that the input fuzzy variables are: state change rate and current damping error. The fuzzy classification of current damping error is small error, medium error, and sudden change.

[0100] The fuzzy rule table is defined as follows: If the state changes suddenly and the damping error is large, then K p With Kd Improve; if the state is stable and the damping error is small, then K p Drop K i improve;

[0101] After the fuzzy reasoning outputs the PID parameters, they are brought into the PID controller to complete the next step of handrail damping analysis and calculation.

[0102] Step 4: During the real-time damper adjustment process, analyze the synergistic characteristics of the command behavior and mechanical feedback during execution. The specific steps are as follows:

[0103] During the real-time adjustment process of the damper, the response behavior sequence between the control signal, current signal and damping output is obtained, and the collaborative feature information of the command behavior and mechanical feedback during the adjustment process is analyzed. The collaborative feature information includes the response monotonic violation count index and the error cross-threshold persistence index.

[0104] The response monotonicity violation count index is used to measure whether the direction of change of the control command and the actual output force is consistent during the damping adjustment process. Specifically, when the control signal is on an upward trend, the corresponding damping force should increase accordingly; conversely, a decrease in the control signal should cause the damping force to decrease. If the system violates the direction of change of the control signal and the direction of the output damping response, it is considered a response monotonicity violation. This index reflects the response consistency of the actuator by counting the cumulative number of such violation events per unit time and weighting them by the time interval spanning each event.

[0105] The higher the response monotonic violation count index, the more likely it is that there is hysteresis, delay, or nonlinear disturbance in the actual execution of the controller command by the damper, and the control path is unstable. This helps to determine whether the execution system has potential problems such as electromagnetic inertial response delay, control mismatch, and structural aging, and provides a quantitative basis for controller correction strategy and health diagnosis.

[0106] The logic for obtaining the response monotonic violation count index is as follows:

[0107] Get each time sampling point t within the damping control period i The control command value U(t i+1 ) and the next moment t i+1 The control command value U(t i ), calculate the change of adjacent control instructions: ΔU(t i )=U(t i+1 )-U(t i );

[0108] Get the damper output damping force and the output damping force at the next moment Calculate the actual output damping change:

[0109] Obtain the response monotonic violation behavior, and number all time periods that meet the violation conditions as the violation set V, and count the time interval corresponding to each violation segment: δt i =t i+1 -t i , calculate the total duration of the overall violation: T v =∑ i∈V δt i ; Get all control instruction changes ΔU(t i ) and the corresponding output damping change ΔZB(t i ), calculate the response direction offset ratio of each segment: i∈V, extract the maximum response shift ratio Ratio of minimum response deviation Get the total change range of the control command during the entire control cycle The final calculated response monotonic violation count index is:

[0110] It should be noted that the control command value is the target damping adjustment command output by the controller in each control cycle based on the real-time perception state and prediction results. This value is not an external input, but is dynamically generated by the adaptive control module inside the system.

[0111] The error cross-threshold persistence index is used to assess the extent to which the damping control error continues to exceed the acceptable range during the operating cycle. During the execution of the control task, there is always a certain error between the target output damping and the actual feedback damping. However, when the error exceeds the set threshold and fails to recover for a period of time, it indicates that the control system may not be able to effectively absorb interference or correct errors. The error cross-threshold persistence index is constructed by counting the number of time points when the error exceeds the limit continuously and combining the amplitude changes within the error fallback interval to construct an enhanced expression, thereby comprehensively measuring the error convergence speed and control steady-state capability. The higher the error cross-threshold persistence index, the longer the control system is in a non-steady state, or there are problems such as insufficient output and feedback distortion. This index can be used to determine in real time whether the control system has entered a transition abnormality zone and drive the controller to enter enhanced gain, switching mode, or trigger a fault diagnosis mechanism;

[0112] The error threshold persistence index can quantify the instability risk of the connection structure under dynamic thermal environment, assist in determining early failure modes such as interface fatigue, loosening or microcrack propagation, improve the dynamic tracking and classification assessment capabilities of connection reliability degradation, and provide support for the formulation of intelligent maintenance strategies.

[0113] The error persistence index directly affects the ability to detect structural degradation and fatigue evolution of copper busbar connections during multiple load cycles. The higher the error persistence index, the greater the energy loss at the connection interface during heating and cooling, the more asymmetric the path resistance change, and the more obvious the geometric stability offset of the hysteresis loop, reflecting the presence of irreversible fatigue damage, loosening or microcrack propagation within the connection interface. When the error persistence index continues to rise and exceeds the preset safety interval, it usually means that the connection has experienced serious structural stability degradation. Risk intervention or local structural reinforcement should be carried out in a timely manner in combination with the operation strategy to avoid cascading failures of the electrical system or the expansion of safety hazards due to connection fatigue damage.

[0114] The logic for obtaining the error threshold persistence index is as follows:

[0115] Get each time sampling point T within the control cycle i The expected damping force Q at des (T i ) and the corresponding actual output damping force S act (T i ), calculate the damping error: ZN(T i )=|Q des (T i )-S act (T i )|; Extract all sampling point index sets that satisfy the damping error greater than the upper limit threshold of the error from the error sequence, and record them as the threshold-crossing set KJ;

[0116] Get the error value N(T) corresponding to each threshold point i ) and time point T i , calculate the sum of the lengths of consecutive time intervals in the threshold-crossing state: ZH = ∑ i∈KJ (T i+1 -T i );

[0117] Get all non-threshold moments T j Corresponding error values, extract the maximum error value and the minimum error value, and calculate the error interval range: Calculate the construction of cross-threshold persistent fluctuations, the calculation expression is:

[0118] Extract the maximum error jump amplitude Δe max and the minimum error jump amplitude Δe min , construct jump unstable value: Calculate the error threshold persistence index, the calculation expression is: DETPI = log (1 + BD (t) 2 (1+BW e )).

[0119] The purpose of setting the response monotonicity violation threshold and the error cross-threshold persistence threshold is to provide the system with clear stability criteria and dynamic response boundaries, which are used to identify whether the damping control system has entered an abnormal state. Among them, the response monotonicity violation threshold is used to measure the consistency of the change direction between the control command and the damper output. When the system issues a command to increase damping but the actuator actually decreases, or the controller intends to decrease damping but the actuator increases in the opposite direction, it indicates that there is a directional violation in the command execution. If such violations accumulate to the preset response monotonicity violation threshold, it can be determined that the current system is in a state of response lag or actuator hysteresis. At this time, the system can automatically adjust controller parameters, enter protective regulation mode, or trigger health diagnosis and alarm mechanisms based on the strategy switching mechanism to prevent operational instability and structural resonance risks caused by inconsistent execution.

[0120] The error cross-threshold duration threshold is used to monitor how long the control error fails to return to the expected range, reflecting the system's steady-state control capability and error absorption speed. When the damping error continues to exceed the allowable range and the overlimit time or the cumulative number of sampling points reaches the threshold, it indicates that the current controller may have problems such as model drift, adjustment failure or poor sensor response. At this time, the enhanced anti-disturbance strategy can be activated, the fuzzy controller structure can be adjusted, or the error correction strength can be increased to ensure that the damper output returns to a reasonable range as soon as possible. The threshold can also be used as an early warning signal for the error drift trend and is used to trigger the system's fault-tolerant mechanism. The synergistic effect of the thresholds, on the one hand, ensures the correctness of the direction of the execution response, and on the other hand, ensures that the error does not accumulate out of control over time, thereby building a robust and self-adjustable closed-loop control protection boundary system.

[0121] According to the response monotonic violation count index and the error threshold persistence index, the comparison between the two thresholds is analyzed and judged, and different strategy adjustments are made:

[0122] When the response monotonic violation count index and the error threshold persistence index do not exceed the corresponding set thresholds, it means that the current damping control system is operating stably, the control command is consistent with the actual output direction, the error fluctuates within an acceptable range, and the system is in a good regulation state. At this time, the current fuzzy controller parameters and output logic are maintained without additional intervention;

[0123] If only the response monotonic violation count index exceeds the threshold, while the error cross-threshold persistence index remains at a normal level, it means that although the control command has been violated multiple times by the actuator, the error has not continued to expand, and the system still has a certain error absorption capacity. This state is usually caused by internal hysteresis of the actuator, electromagnetic inertia lag, or drive model deviation. At this time, the response consistency optimization strategy should be triggered to correct the control mapping function, recalibrate the electronic control drive curve, or temporarily introduce direction restriction logic to prevent high-frequency reversal.

[0124] If only the error cross-threshold persistence index exceeds the error cross-threshold persistence threshold, while the response monotonic violation index remains within the safe range, it means that the execution direction of the control command is generally consistent, but the error has not been effectively corrected for a long time. This may be caused by control parameter drift, uncompensated disturbances, or sensor feedback deviation. In this case, the steady-state error enhancement compensation strategy should be activated, such as increasing the controller's proportional adjustment capability, adjusting the fuzzy rule response range, or enabling error recovery interval speed limit control to suppress error accumulation.

[0125] When both indexes exceed their respective thresholds, there is a problem that the control command direction is not responded to by the actuator, and there is also a problem of continuous deviation of the damping error, which indicates that the current system may be on the verge of response instability. At this time, the global protection strategy should be activated to lock the control frequency, limit the maximum damping change rate, and switch the control mode to fault-tolerant mode or high damping protection state to ensure that the system is not further disturbed in a short period of time. At the same time, the current status data should be recorded for subsequent diagnostic analysis.

[0126] In addition, during the dynamic adjustment of the damping control strategy, the system synchronously monitors the interactive behavior characteristics between the passenger and the armrest, including indicators such as changes in grip strength, hand contact stability, slip trends and contact duration. When it is detected that the passenger is in a state of passive reliance on the armrest for support, and the vehicle dynamic disturbance or control error indicators are abnormal, the system will increase the armrest damping response priority, enhance the support rigidity and feedback sensitivity, and prioritize the stability of the passenger posture and impact resistance. If it is identified as an active grip state or a short-term transitional contact, the damping adjustment rate will be appropriately slowed down to avoid excessive damping interfering with the freedom of hand movements, thereby improving the naturalness of human-computer interaction. Through the joint judgment of the dual indicators of passenger status and control status, the armrest system can realize personalized and situational adjustment of complex working conditions and individual behaviors, thereby improving the overall safety, comfort and intelligent response capabilities of operation.

[0127] The present invention collects multi-source sensor signals from the vehicle's driving status and the passenger's gripping behavior to perceive in real time the vehicle's dynamic characteristics under acceleration, deceleration, cornering, or bumpy road conditions. It then constructs a fused state vector based on data such as hand posture, grip force distribution, and arm torque. Based on this, it implements intelligent adjustment of the handrail's damping. Compared to traditional fixed-rigidity mechanical handrail structures, the present invention has the following beneficial effects:

[0128] Firstly, the system can predict in advance the risk of posture imbalance that passengers may experience based on the vehicle's movement trends in situations such as sudden braking, starting, or uneven roads. It can also proactively provide flexible support by adaptively adjusting the damping force of the armrest, significantly reducing the probability of hand slippage or body toppling due to inertial impact, thereby improving driving safety.

[0129] Secondly, the present invention adopts a control strategy that combines fuzzy logic with state prediction, so that the handrail can distinguish whether the passenger is actively holding it or passively relying on support, dynamically adjust the damping response speed and strength, and realize personalized support capabilities that vary according to the person, time, and vehicle conditions, thereby enhancing the comfort and stability of the human-computer interaction of the entire vehicle. The use of signal processing technologies such as dynamic phase compensation and residual filtering can effectively suppress the influence of high-frequency vibrations, electromagnetic interference, etc. generated during the operation of the bus on the sensor data, thereby ensuring the real-time and accuracy of the damping control.

[0130] Example 2: Automobile armrest dynamic damping adjustment system based on multi-sensor fusion, such as Figure 2 As shown, specifically including:

[0131] The armrest multi-source data acquisition module is used to collect vehicle posture, dynamic behavior, contact behavior, and hand torque data. It resamples and pseudo-synchronizes the collected data, uses structure-preserving residual filtering to suppress high-frequency interference, performs nonlinear normalization and response phase compensation, and outputs a fused state vector.

[0132] The operation trend analysis module is used to extract the current vehicle speed, posture and dynamic variables based on the fused state vector for modeling, predict the future state based on the short-term historical sequence, and form a dynamic risk index based on the state increment and change trend to determine the current vehicle operation disturbance intensity and trend evolution;

[0133] The armrest damping control module is used to construct an optimization function for mechanical support, response smoothness, and risk suppression based on the vehicle's current state and predicted trends. It also dynamically adjusts controller parameters and adjusts the armrest damping output based on a fuzzy inference mechanism.

[0134] The strategy adjustment module is used to analyze the collaborative feature information generated by the command behavior and mechanical feedback during the real-time adjustment of the damper, and adjust the strategy according to the analysis results.

[0135] The above formulas are all dimensionless and calculated numerically. Specific dimension removal can be achieved by various means such as standardization, which will not be elaborated here. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0136] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, an ATA hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0137] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0138] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0140] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, and may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment as needed.

[0141] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0142] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A vehicle armrest dynamic damping adjustment method based on multi-sensor fusion, characterized in that: The steps include: Collect vehicle posture, dynamic behavior, contact behavior, and hand torque data, resample and pseudo-synchronize the collected data, use structure-preserving residual filtering to suppress high-frequency interference, perform nonlinear normalization and response phase compensation, and output a fused state vector; Based on the fused state vector, the current vehicle speed, posture, and dynamic variables are extracted for modeling. The future state is predicted based on the short-term historical sequence. Based on the state increment and change trend, a dynamic risk index is formed to determine the disturbance intensity and trend evolution of the current vehicle operation. Based on the vehicle's current state and predicted trends, an optimization function for mechanical support, response smoothness, and risk suppression is constructed. The controller parameters are dynamically adjusted based on the fuzzy inference mechanism to regulate the armrest damping output. During the real-time adjustment of the damper, the collaborative characteristic information generated by the command behavior and mechanical feedback during the adjustment process is analyzed, and the strategy is adjusted based on the analysis results.

2. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 1 is characterized in that: Collect vehicle posture, dynamic behavior, contact behavior, and hand torque data, and perform resampling and pseudo-synchronization on the collected data. The specific process is as follows: Deploy a steering wheel angle sensor to obtain vehicle lateral steering trend data, deploy a three-axis acceleration sensor to sense the vehicle's longitudinal, lateral, and vertical dynamic behavior data, and deploy a pressure distribution matrix sensor to obtain hand posture, grip strength, and center of gravity offset data; Deploy armrest torque sensors to obtain data on the active or passive adjustment process of the arm; Construct a reference time axis as a unified sampling reference frame to resample signal streams with inconsistent original sampling frequencies; All processed signals after resampling are aligned to form a pseudo-synchronous data frame.

3. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 2 is characterized in that: The structure-preserving residual filter is used to suppress high-frequency interference, perform nonlinear normalization and response phase compensation, and output the fusion state vector. The specific process is as follows: Use structure-preserving residuals for filtering, define the original and estimated signal residuals, and construct a filtering objective function to retain the sudden hand force and vehicle shaking signals and output the noise-suppressed signal; The noise suppressed signal is mapped using a nonlinear monotonic mapping to output a normalized signal set; Dynamic phase compensation based on the response estimation model is applied to the normalized signal set. After all signal channels are compensated, a fusion state vector is output.

4. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 3 is characterized in that: Based on the fusion state vector, the current vehicle speed, posture, and dynamic variables are extracted for modeling, and the future state is predicted based on the short-term historical sequence. The specific steps are as follows: The output fusion state vector is converted into a vehicle dynamic state representation to form a set of state variables including the vehicle's lateral and longitudinal posture and dynamic characteristics. The selected vehicle state variables include current speed, lateral acceleration, heading angular velocity, vertical acceleration, and acceleration change rate; Carry out vehicle state variable mapping modeling and transform vehicle state variables into dynamic state variables, including armrest pressure concentration to determine driving posture tilt, steering wheel angle to determine yaw trend, and body acceleration signal to restore the vehicle's real-time inertial behavior; Carry out state evolution trend prediction to predict the changing trend of vehicle state in the future.

5. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 4 is characterized in that: Carry out state evolution trend prediction to predict the changing trend of vehicle state in the future. The specific process is as follows: Construct a time window and collect the vehicle state vectors at the most recent sampling moment into a set of historical sequences; Construct a long short-term memory network model to extract temporal correlation and state transition patterns in historical sequences; The state value is predicted through the state transfer equation, and the state derivative and change trend are derived.

6. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 5 is characterized in that: Based on the state increment and change trend, a dynamic risk index is formed to determine the disturbance intensity and trend evolution of the current vehicle operation. The specific process is as follows: The state difference between the current moment and the predicted moment is used as the state increment value; Define risk indicators to measure the extent to which the current system is on the brink of dynamic change; The obtained state increment value is used to identify whether there is an obvious trend change at present, and the controller is used to adjust the strength, delay and mode of the control response according to the dynamic risk index.

7. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 6, characterized in that: Based on the vehicle's current state and predicted trends, an optimization function for mechanical support, response smoothness, and risk suppression is constructed. The controller parameters are dynamically adjusted based on the fuzzy inference mechanism to adjust the armrest damping output. The specific process is as follows: The mechanical support index refers to whether the armrest needs to provide additional stabilizing support when the vehicle turns, accelerates, or bumps. The damping response smoothness index refers to the driver discomfort caused by damping changes. The dynamic risk control index refers to whether the system determines that the vehicle state is in a stage of drastic fluctuations and should quickly increase the damping to cope with possible external impacts. An optimization function is set to optimize mechanical support, response smoothness, and risk suppression, and a fuzzy reasoning mechanism is introduced to adjust the parameters of the PID controller in real time. The fuzzy inputs of the PID controller are the damping error, error change rate, and vehicle state trend; After the fuzzy reasoning outputs the PID parameters, they are brought into the PID controller to analyze and calculate the handrail damping.

8. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 7, characterized in that: During the real-time adjustment of the damper, the collaborative characteristic information generated by the command behavior and mechanical feedback during the adjustment process is analyzed. The specific process is as follows: Obtain the response behavior sequence between the control signal, current signal and damping output, and analyze the collaborative characteristic information between the command behavior and mechanical feedback during the adjustment process; The collaborative feature information includes the response monotonic violation count index, and the collaborative feature information includes the error cross-threshold persistence index; The response monotonic violation count index indicates whether the direction of change of the control command and the actual output force is consistent during the damping adjustment process; The error threshold persistence index indicates the extent to which the damping control error continues to exceed the acceptable range during the operating cycle.

9. The vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to claim 8, characterized in that: Adjust the strategy based on the analysis results. The specific process is as follows: Set the response monotonic violation threshold and the response monotonic violation count index for comparative analysis, and set the error cross-threshold persistence threshold and the error cross-threshold persistence index for comparative analysis; When the response monotonic violation count index and the error threshold-crossing persistence index do not exceed the corresponding set thresholds, the current fuzzy controller parameters and output logic are maintained without additional intervention; If only the response monotonic violation count index exceeds the threshold, the response consistency optimization strategy is triggered to modify the control mapping function, recalibrate the electronic control drive curve, or temporarily introduce direction restriction logic to prevent high-frequency reverse; If only the error cross-threshold persistence index exceeds the error cross-threshold persistence threshold, the steady-state error enhancement compensation strategy is activated to increase the controller's proportional adjustment capability, adjust the fuzzy rule response range, or enable the error recovery interval speed limit control to suppress error accumulation; When both the response monotonic violation count index and the error cross-threshold persistence index exceed their respective thresholds, the global protection strategy is activated to lock the control frequency, limit the maximum damping change rate, and record the current status data for diagnostic analysis.

10. A vehicle armrest dynamic damping adjustment system based on multi-sensor fusion, used to implement the vehicle armrest dynamic damping adjustment method based on multi-sensor fusion according to any one of claims 1 to 9, characterized in that: include: The armrest multi-source data acquisition module is used to collect vehicle posture, dynamic behavior, contact behavior, and hand torque data. It resamples and pseudo-synchronizes the collected data, uses structure-preserving residual filtering to suppress high-frequency interference, performs nonlinear normalization and response phase compensation, and outputs a fused state vector. The operation trend analysis module is used to extract the current vehicle speed, posture and dynamic variables based on the fused state vector for modeling, predict the future state based on the short-term historical sequence, and form a dynamic risk index based on the state increment and change trend to determine the current vehicle operation disturbance intensity and trend evolution; The armrest damping control module is used to construct an optimization function for mechanical support, response smoothness, and risk suppression based on the vehicle's current state and predicted trends. It also dynamically adjusts controller parameters and adjusts the armrest damping output based on a fuzzy inference mechanism. The strategy adjustment module is used to analyze the collaborative feature information generated by the command behavior and mechanical feedback during the real-time adjustment of the damper, and adjust the strategy according to the analysis results.

Images