Vehicle six-degree-of-freedom motion simulation device and method
By using data clustering and dynamic adjustment of the second-order cutoff frequency in a vehicle six-degree-of-freedom motion simulation device, the problem of balancing realism and safety in simulation under different driving conditions in traditional devices is solved, thus improving both safety and realism under different conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WUHAN ZHONGJIAO TRAFFIC ENG CO LTD
- Filing Date
- 2026-01-20
- Publication Date
- 2026-05-01
AI Technical Summary
Traditional six-degree-of-freedom vehicle motion simulation devices struggle to simultaneously achieve both realism and safety under different driving conditions. A fixed second-order cutoff frequency filters out valuable low-frequency motion information under stable driving conditions, resulting in a lack of acceleration and deceleration sensations. Under severe driving conditions, it cannot effectively filter out strong low-frequency signals, causing the platform displacement to accumulate rapidly, exceeding the limits of the motion space and threatening the safety of equipment and personnel.
Through multiple vehicle driving simulation tests, acceleration and angular velocity signals were collected. Combined with a pre-set second-order cutoff frequency washing algorithm, high-frequency signals were obtained. The test dataset was clustered, and the motion evaluation value and over-limit risk of each cluster were calculated. The second-order cutoff frequency was dynamically adjusted, and the six-degree-of-freedom motion platform was controlled in real time to perform motion simulation.
It improves the realism and safety of motion simulation under different driving conditions by dynamically adjusting the second-order cutoff frequency to prevent motion from exceeding limits, ensure platform safety, and enhance the driving realism.
Smart Images
Figure CN121954508A_ABST
Abstract
Description
A vehicle six-degree-of-freedom motion simulation device and method Technical Field
[0001] This application relates to the field of vehicle driving simulation technology, specifically to a vehicle six-degree-of-freedom motion simulation device and method. Background Technology
[0002] Simulated driving is a crucial aspect of automotive research and development (R&D) and testing. Through simulated driving, not only can vehicle handling performance and driving stability be analyzed, providing strong support for automotive product development and optimization, but it can also significantly reduce testing costs and development cycles. A six-degree-of-freedom motion platform plays a key role in simulated driving, precisely controlling the platform's motion to simulate dynamic behavior in a real environment, providing the driver with an immersive driving experience. The rinsing algorithm is the core technology connecting vehicle motion and platform motion. It uses a high-pass filter to remove low-frequency components from the vehicle signal, preventing the platform from exceeding its physical motion boundaries due to the continuous accumulation of displacement and attitude angles.
[0003] To reproduce more motion patterns within a limited space, traditional washout algorithms typically use a fixed second-order cutoff frequency to filter out low-frequency signals. However, vehicles experience vastly different driving conditions in actual operation, and these conditions have varying requirements for the second-order cutoff frequency. For example, under stable driving conditions, an excessively high second-order cutoff frequency can over-filter out valuable low-frequency motion information, resulting in a lack of acceleration and deceleration sensations and weakening the simulation's realism. Under severe driving conditions, an excessively low second-order cutoff frequency cannot effectively filter out strong low-frequency signals, causing rapid accumulation of platform displacement, which can easily exceed the platform's motion space limitations, damaging immersion and even threatening equipment and personnel safety. This makes it difficult to simultaneously ensure both the simulation's realism and safety under different driving conditions. Summary of the Invention
[0004] To address the aforementioned technical problems, a vehicle six-degree-of-freedom motion simulation device and method are provided to resolve the existing issues.
[0005] The solution to the technical problem of this application is to provide a vehicle six-degree-of-freedom motion simulation device and method, including the following steps:
[0006] In a first aspect, embodiments of this application provide a method for simulating the six degrees of freedom motion of a vehicle, the method comprising the following steps:
[0007] Through multiple vehicle driving simulation tests, acceleration and angular velocity signals of different control cycles were collected. Combined with the preset second-order cutoff frequency washing algorithm, high-frequency acceleration and angular velocity signals were obtained. The tilt angle of the six-degree-of-freedom motion platform at each moment in each control cycle and the elongation of each telescopic rod were collected to form an experimental dataset.
[0008] Feature vectors reflecting motion fluctuations and average levels are extracted from acceleration and angular velocity signals, and all control cycles in the experimental dataset are clustered. Based on the difference between the elongation of the telescopic rod and the elongation limit in each cluster and the motion fluctuations, the motion evaluation value of each cluster is calculated.
[0009] The energy differences between high-frequency acceleration signals and high-frequency angular velocity signals in the frequency domain of each control cycle are evaluated. The low-frequency residual of each control cycle is calculated. Combined with the motion evaluation value, the risk degree of exceeding the limit of each cluster is obtained.
[0010] By analyzing the differences in the elongation and tilt angle of the same telescopic rod between different control cycles within each cluster, the response difference of each cluster is calculated. Combined with the risk of exceeding limits, the clusters are classified into different categories. The second-order cutoff frequency of the corresponding washing algorithm for each type of cluster is then adjusted using the response difference.
[0011] In real time, the feature vectors of the current control cycle are clustered, and the shuffling algorithm is run by calling the adjusted second-order cutoff frequency corresponding to the cluster to control the six-degree-of-freedom motion platform to perform motion simulation.
[0012] Preferably, the process of constructing the feature vector is as follows:
[0013] For each control cycle in the experimental dataset, the dispersion and mean of acceleration at all times in the acceleration signal of each control cycle are calculated and denoted as the first dispersion and the first mean, respectively; the dispersion and mean of angular velocity at all times in the angular velocity signal of each control cycle are calculated and denoted as the second dispersion and the second mean, respectively.
[0014] The sum of the first and second discrete values is used as the motion discrete value for each control cycle; the first mean, the second mean, and the motion discrete value are used to form the feature vector for each control cycle.
[0015] Preferably, the calculation of the motion evaluation value for each cluster includes:
[0016] Obtain the ultimate elongation of the telescopic rod on the six-degree-of-freedom motion platform; select the maximum value of the elongation of all telescopic rods at all times under all control cycles in each cluster, and record it as the maximum extension value; take the difference between the ultimate elongation and the maximum extension value as the limit deviation of each cluster.
[0017] The mean of the motion discrete values of all control cycles within each cluster is calculated as the motion volatility of each cluster; the ratio of motion volatility to limit deviation is used as the motion evaluation value of each cluster.
[0018] Preferably, the calculation of the low-frequency residual for each control cycle includes:
[0019] Frequency domain analysis was performed on the acceleration signal, angular velocity signal, high-frequency acceleration signal, and high-frequency angular velocity signal for each control cycle, and the spectrum diagrams were obtained respectively. The sum of the energies of all frequency components in the spectrum diagrams was recorded as the total energy.
[0020] Calculate the ratio of the total energy between the high-frequency acceleration signal and the acceleration signal in each control cycle, and record it as the first ratio; calculate the ratio of the total energy between the high-frequency angular velocity signal and the angular velocity signal in each control cycle, and record it as the second ratio.
[0021] The low-frequency residual is the average of the first ratio and the second ratio.
[0022] Preferably, the risk level exceeding the limit is the product of the mean of the low-frequency residuals of all control cycles within each cluster and the motion assessment value.
[0023] Preferably, the calculation of the response difference of each cluster includes:
[0024] The elongation of each telescopic rod at all times within each control cycle is used to form an elongation sequence. For each cluster, the difference in the elongation sequence of the same telescopic rod between any two control cycles is calculated and recorded as the difference value. The mean of the difference values of all telescopic rods between any two control cycles is calculated and used as the first difference between any two control cycles.
[0025] The tilt angles of the six-degree-of-freedom motion platform at all times within each control cycle are compiled into an angle sequence; the difference between the angle sequences between any two control cycles is calculated and denoted as the second difference; the sum of the first difference and the second difference is taken as the relative difference between any two control cycles.
[0026] The response difference is the mean of the relative differences between any two control periods within each cluster.
[0027] Preferably, the classification of clusters into different classes includes:
[0028] The product of the response difference and the risk of exceeding the limit is used as the discriminant coefficient of each cluster; the mean and standard deviation of the discriminant coefficients of all clusters are calculated, and the first threshold interval is obtained according to the Laida criterion; the clusters whose discriminant coefficients are not distributed in the first threshold interval are recorded as the exceeding clusters.
[0029] Calculate the mean and standard deviation of the response variability of all clusters except the overlimit clusters, and obtain the second threshold interval according to the Laida criterion. The clusters whose response variability is not distributed in the second threshold interval are denoted as downregulated clusters.
[0030] All clusters except for the overlimit cluster and the down-regulation cluster are denoted as other clusters.
[0031] Preferred, the first Each cluster corresponds to an adjusted second-order cutoff frequency. The calculation formula is: ,in, For the first Each cluster corresponds to the second-order cutoff frequency before adjustment. For the first The response difference corresponding to each cluster For normalization function, Let be the set of all transfinite clusters. For the set of all downregulated clusters, For the set of all other clusters, Indicates belonging to, among which, , The universal set represents the set of all clusters. This indicates finding the union of sets.
[0032] Preferably, the motion simulation of the six-degree-of-freedom motion platform includes: obtaining the feature vector of the current control cycle based on the acceleration and angular velocity signals of the current control cycle; clustering the feature vector of the current control cycle and the feature vectors of all control cycles in the test dataset; using the adjusted second-order cutoff frequency corresponding to the cluster to which the current control cycle belongs as the second-order cutoff frequency of the washout algorithm corresponding to the current control cycle; using the washout algorithm to obtain the target posture that the six-degree-of-freedom motion platform needs to reproduce; performing inverse kinematics on it to solve for the elongation of each telescopic rod on the six-degree-of-freedom motion platform; and based on the elongation, the servo controller drives the servo motor to control the six-degree-of-freedom motion platform to perform motion simulation.
[0033] Secondly, embodiments of this application also provide a vehicle six-degree-of-freedom motion simulation device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described vehicle six-degree-of-freedom motion simulation methods.
[0034] This application has at least the following beneficial effects:
[0035] This application constructs feature vectors and clusters all control cycles within the experimental dataset. Its beneficial effect is that by extracting the intensity and motion characteristics of the vehicle's driving state under each control cycle, different control cycles are clustered, thereby automatically dividing massive amounts of driving data into different representative typical driving conditions. The motion evaluation value of each cluster is calculated, which has the beneficial effect of quantifying the intensity of vehicle motion and the degree of platform over-limit risk under the driving conditions represented by each cluster from the time domain dimension, so as to assess the possibility of motion over-limit of the motion platform at a fixed second-order cutoff frequency. Calculating the low-frequency residual for each control cycle reveals the impact of the high-pass filter on low-frequency signal components at a fixed second-order cutoff frequency, allowing for an assessment of the platform's potential to exceed motion boundaries due to accumulated low-frequency displacement. Obtaining the over-limit risk for each cluster provides a multi-dimensional assessment of the platform's over-limit risk at a fixed second-order cutoff frequency. Calculating the response difference for each cluster considers the differences in actual motion feedback of the six-DOF motion platform under the same driving condition, reflecting variations in driving behavior. Under conditions of similar characteristics, the possibility of inconsistent platform motion responses due to unreasonable setting of the second-order cutoff frequency is revealed, demonstrating the matching between the setting of the second-order cutoff frequency and the driving conditions represented by different clusters. Distinguishing all clusters is beneficial because it identifies which clusters have excessive risk and which lack realism, allowing for dynamic adjustment of the second-order cutoff frequency accordingly. The second-order cutoff frequency of the corresponding algorithm for each cluster is adjusted using response difference, providing real-time feedback on the feature vectors of the current control cycle. The data is clustered, and the corresponding adjusted second-order cutoff frequency is called to run the washout algorithm. This controls the six-degree-of-freedom motion platform to perform motion simulation. The beneficial effect is that by identifying the driving condition to which the current control cycle belongs, the matching second-order cutoff frequency is dynamically called to run the washout algorithm. This allows the six-degree-of-freedom motion platform to prevent motion exceeding limits and ensure safety under severe conditions by raising the second-order cutoff frequency; and under stable conditions by lowering the second-order cutoff frequency to retain more low-frequency motion information and enhance realism. Thus, the realism and safety of motion simulation are improved under different driving conditions. Attached Figure Description
[0036] The following description, in conjunction with the accompanying drawings, provides a more detailed explanation of a six-degree-of-freedom motion simulation method for a vehicle according to this application.
[0037] Figure 1 is a flowchart of a vehicle six-degree-of-freedom motion simulation method provided in an embodiment of this application;
[0038] Figure 2 is a flowchart of the steps of the method for obtaining the response difference of each cluster provided in the embodiment of this application. Detailed Implementation
[0039] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of a vehicle six-degree-of-freedom motion simulation device and method proposed in this application, in conjunction with the accompanying drawings and embodiments, is provided. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit the scope of this application.
[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0041] Please refer to Figure 1, which shows a flowchart of a vehicle six-degree-of-freedom motion simulation method according to an embodiment of this application. The method includes the following steps:
[0042] Step 1: Through multiple vehicle driving simulation tests, acceleration and angular velocity signals of different control cycles are collected. Combined with the preset second-order cutoff frequency washing algorithm, high-frequency acceleration and angular velocity signals are obtained. The tilt angle of the six-degree-of-freedom motion platform at each moment in each control cycle and the elongation of each telescopic rod are collected to form an experimental dataset.
[0043] A six-degree-of-freedom motion platform is a device that integrates mechanical transmission and precision servo control technologies. It features high load-bearing capacity, fast response rate, high control accuracy, and good stability, and is widely used in driving simulation. The six-degree-of-freedom motion platform consists of an upper platform, a lower platform, and six electric cylinders. The upper platform is connected to the piston rod end of the electric cylinders using six sets of Hooke hinges, and the cylinder body end of the electric cylinders is also connected to the lower platform using six sets of Hooke hinges. The servo motor drives the extension and retraction of the electric cylinders, changing their extension and retraction length, enabling the upper platform to perform six postures: lifting, forward movement, lateral movement, yaw, roll, and pitch.
[0044] A six-DOF motion platform has six telescopic rods that can be extended and retracted. The extension of each of the six rods corresponds one-to-one with the six degrees of freedom of the upper platform in space. Even if the extension of five rods is fixed, the remaining rod can still affect the position and attitude data of the six degrees of freedom through extension and retraction. Therefore, the length of the telescopic rod determines the six degrees of freedom of the upper platform in space, that is, it determines the position and attitude of the upper platform. In platform control, the upper platform must achieve the ideal position and attitude by extending and retracting the six telescopic rods connecting the upper and lower platforms. A fixed position and attitude of the upper platform corresponds to a unique extension of the telescopic rod. That is, once the extension of the telescopic rod is determined, the problem of determining the position and attitude of the upper platform by finding the lengths of the six telescopic rods is the inverse kinematics solution of the platform. Conversely, the problem of determining the position and attitude of the upper platform by knowing the lengths of the telescopic rods is the forward kinematics solution of the platform.
[0045] The six-degree-of-freedom motion platform, through its washout algorithm, can reproduce the motion state of a real vehicle in space. The washout algorithm is a driving algorithm that converts the acceleration and angular velocity generated by the vehicle's motion into simulator motion within the constraints of simulator space. The washout algorithm consists of a high-frequency acceleration channel, a low-frequency acceleration channel, and a high-frequency angular velocity channel. To avoid signal interference and coupling, each channel uses a filter to frequency-divide and adjust the input signal, resulting in a corresponding signal for each channel. The high-frequency filtering channel primarily washes out the high-frequency components of the input signal to reproduce instantaneous changes in motion. The low-frequency filtering channel, also known as the tilt coordination channel, is used to reproduce continuous motion, providing a continuous sense of movement. The high-frequency angular velocity channel mainly processes the input angular velocity signal, retaining the high-frequency components. From the structure of the washout algorithm, it can be seen that the filters are the main factor affecting signal washing. The second-order cutoff frequency of the high-pass filter is the primary parameter affecting its filtering effect; different choices of the second-order cutoff frequency significantly impact the filtering effect, thus affecting the washout algorithm's overall performance.
[0046] Based on the above analysis, a simulated driving scenario is created using ADS (Advanced Design System) simulation software. In this driving scenario, the motion of the vehicle is simulated, and the acceleration and angular velocity signals of the vehicle are generated in real time.
[0047] In this embodiment, the simulated scenario can be a highway tunnel, a series of curves, a muddy road section, a water crossing section, etc. This embodiment does not impose any special restrictions on this; the ADS simulation software is a well-known technology and will not be described in detail here.
[0048] Using the washing algorithm, the real-time acquired acceleration and angular velocity signals are scaled, filtered, and processed to obtain the target posture that the six-degree-of-freedom motion platform needs to reproduce. The inverse kinematics solution is then performed to solve the elongation of each telescopic rod on the six-degree-of-freedom motion platform.
[0049] The acceleration signal and angular velocity signal filtered by the high-pass filter in the washing algorithm are respectively denoted as high-frequency acceleration signal and high-frequency angular velocity signal;
[0050] It should be noted that the washing algorithm and inverse kinematics are well-known techniques and will not be elaborated here. In the washing algorithm, the high-pass filter is set to a fixed second-order cutoff frequency of 2.2. As other implementation methods, implementers can set it according to their actual situation.
[0051] By deploying an inertial measurement unit (IMU) on a six-degree-of-freedom motion platform, the tilt angle of the six-degree-of-freedom motion platform is collected in real time;
[0052] In this embodiment, the data acquisition frequency is 500Hz. As for other implementation methods, the implementer can set the frequency according to the actual situation.
[0053] Multiple moments are treated as a control cycle, and the maximum-minimum method is used to normalize all the collected data to obtain the acceleration signal, angular velocity signal, high-frequency acceleration signal, high-frequency angular velocity signal, and tilt angle of the six-degree-of-freedom motion platform at different moments and the elongation of each telescopic rod at different moments.
[0054] In this embodiment, the duration of the control cycle is 3 seconds. As for other implementation methods, the implementer can set it according to the actual situation. The maximum and minimum value method is a well-known technology and will not be described in detail here.
[0055] Using a pre-set second-order cutoff frequency wash-out algorithm, multiple simulation experiments were conducted to collect acceleration signals, angular velocity signals, high-frequency acceleration signals, high-frequency angular velocity signals, and the tilt angle of the six-degree-of-freedom motion platform at different times and the elongation of each telescopic rod at different times, forming an experimental dataset.
[0056] In this embodiment, the preset second-order cutoff frequency is 2.2; secondly, data from 6000 control cycles are collected. As for other implementation methods, the implementer can set the frequency according to the actual situation.
[0057] Thus, acceleration signals, angular velocity signals, high-frequency acceleration signals, high-frequency angular velocity signals, and tilt angles of the six-degree-of-freedom motion platform at different times and the elongation of each telescopic rod at different times are obtained for multiple control cycles.
[0058] Step 2: Extract feature vectors reflecting motion volatility and average level from acceleration and angular velocity signals, and cluster all control cycles in the test dataset; calculate the motion evaluation value of each cluster based on the difference between the elongation of the telescopic rod and the elongation limit within each cluster and the motion volatility; evaluate the energy differences between high-frequency acceleration signals and acceleration signals, and between high-frequency angular velocity signals and angular velocity signals in the frequency domain for each control cycle, calculate the low-frequency residual for each control cycle, and combine the motion evaluation value to obtain the over-limit risk degree of each cluster.
[0059] In simulated driving, the driving behaviors corresponding to smooth cruising, continuous cornering, and emergency obstacle avoidance differ significantly. If the filtering algorithm uses a fixed second-order cutoff frequency, it is difficult to simultaneously meet the requirements of realistic simulation and platform motion boundary constraints. Under smooth driving conditions, if the second-order cutoff frequency of the high-pass filter is high, it will over-filter low-frequency acceleration information, resulting in the loss of motion sensation during vehicle acceleration and deceleration, thus weakening the driving realism. In contrast, under severe conditions such as continuous cornering and emergency obstacle avoidance, if the second-order cutoff frequency of the high-pass filter is low, low-frequency motion components cannot be filtered out in time, which can easily lead to the continuous accumulation of platform displacement or attitude angles, exceeding the platform's motion boundaries, significantly damaging the immersive experience and potentially affecting the driver's personal safety. Therefore, to improve the realism and safety of motion simulation, it is necessary to dynamically adjust the second-order cutoff frequency according to different driving behaviors to optimize the simulation effect and ensure safety.
[0060] Secondly, because the dynamic behaviors of vehicles, such as acceleration, deceleration, and steering, differ significantly under different driving conditions—for example, the changes in acceleration and angular velocity are relatively small during stable straight-line driving, while they change drastically during emergency obstacle avoidance—driving behaviors in different control cycles are classified based on the severity of changes in acceleration and angular velocity signals. Specifically:
[0061] For each control cycle in the experimental dataset, the degree of dispersion of the acceleration at all times in the acceleration signal of each control cycle is denoted as the first degree of dispersion.
[0062] The degree of dispersion of the angular velocity at all times in the angular velocity signal of each control cycle is denoted as the second degree of dispersion.
[0063] In this embodiment, the degree of dispersion is measured by calculating the coefficient of variation of acceleration at all times in the acceleration signal of each control cycle and the coefficient of variation of angular velocity at all times in the angular velocity signal of each control cycle. The calculation of the coefficient of variation is a well-known technique and will not be described in detail here.
[0064] The sum of the first and second discrete values is used as the motion discrete value for each control cycle.
[0065] The average value of the acceleration at all times in the acceleration signal of each control cycle is recorded as the first average value;
[0066] The average angular velocity at all times in the angular velocity signal of each control cycle is denoted as the second average.
[0067] It should be noted that the larger the motion discrete value, the more intense the vehicle motion is within the control cycle, reflecting more frequent acceleration, deceleration and steering operations during driving, and more complex driving behavior; the larger the first mean value, the greater the average acceleration of the vehicle, and the more acceleration operations are performed during driving; the larger the second mean value, the greater the average steering angular velocity of the vehicle, and the more steering operations are performed during driving, reflecting the possibility of continuous cornering or emergency obstacle avoidance.
[0068] The first mean, the second mean, and the discrete values of motion are used to form a feature vector;
[0069] Cluster the feature vectors of all control cycles in the experimental dataset to obtain multiple clusters;
[0070] In this embodiment, the DBSCAN clustering algorithm is used for clustering. The DBSCAN clustering algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers may use other methods of existing technology, such as the k-means clustering algorithm, the DPC clustering algorithm, etc. This embodiment does not impose any special restrictions on this.
[0071] It should be noted that the feature vector reflects the average motion state of the vehicle and the intensity of the vehicle's dynamic motion during the control cycle, and one cluster corresponds to one driving condition.
[0072] Furthermore, motion simulation is performed using a fixed second-order cutoff frequency to differentiate between various driving conditions. The suitability of the fixed second-order cutoff frequency for different driving conditions is evaluated, and the second-order cutoff frequencies of the clusters corresponding to different driving conditions are then adjusted. First, it is assessed whether the six-degree-of-freedom motion platform might experience motion over-limit risks under the driving conditions corresponding to each cluster, i.e., whether the platform might exceed its motion boundaries due to excessive displacement or deflection. Therefore, the deviation of the elongation of each telescopic rod at different times during each control cycle within the cluster, as well as the average level of the motion discrete values, are analyzed to calculate the motion evaluation value, specifically:
[0073] Obtain the ultimate elongation of the telescopic rod on a six-degree-of-freedom motion platform;
[0074] It should be noted that the limit elongation is obtained by consulting the user manual of the telescopic pole equipment, which represents the maximum elongation length that the telescopic pole can physically achieve.
[0075] The maximum elongation of all telescopic rods at all times under all control cycles within each cluster is selected and recorded as the maximum extension value.
[0076] The difference between the limiting elongation and the maximum extension value is used as the limiting deviation of each cluster.
[0077] It should be noted that the smaller the limit deviation, the closer the elongation of the telescopic rod in the six-degree-of-freedom motion platform is to the limit value under the driving conditions corresponding to the cluster, and the greater the risk of exceeding the limit. Therefore, it is more necessary to increase the second-order cutoff frequency to accelerate the washing out of low-frequency motion components and suppress displacement accumulation.
[0078] Calculate the mean of the motion discrete values of all control cycles within each cluster, and use it as the motion volatility of each cluster;
[0079] The ratio of motion fluctuation to limit deviation is used as the motion evaluation value for each cluster.
[0080] It should be noted that, when calculating the ratio, to avoid the denominator being 0, a preset value greater than 0 is added to the denominator. In this embodiment, the preset value greater than 0 is 0.1. In other implementation methods, the implementer can set it according to the actual situation. Secondly, the greater the motion fluctuation, the more intense the vehicle motion is under the driving conditions corresponding to the cluster, and the greater the motion response requirements of the six-degree-of-freedom motion platform. The second-order cutoff frequency needs to be increased to avoid motion exceeding limits. The greater the motion evaluation value, the more intense the driving behavior is, and the higher the risk of the platform exceeding limits. The second-order cutoff frequency of the shuffling algorithm should be increased.
[0081] Furthermore, the washing algorithm uses a high-pass filter to filter out low-frequency signals in order to prevent the six-degree-of-freedom motion platform from exceeding the limit due to the accumulation of low-frequency displacement. Therefore, by analyzing the proportion of low-frequency components in the acceleration and angular velocity signals input into the washing algorithm, the low-frequency residual is calculated to further evaluate whether the second-order cutoff frequency of the washing algorithm needs to be adjusted.
[0082] Frequency domain analysis was performed on the acceleration signal, angular velocity signal, high-frequency acceleration signal, and high-frequency angular velocity signal for each control cycle, and the spectrum diagrams were obtained respectively. The sum of the energies of all frequency components in the spectrum diagrams was recorded as the total energy.
[0083] In this embodiment, the Fourier transform algorithm is used for frequency domain analysis. The Fourier transform algorithm is a well-known technology and will not be described in detail here.
[0084] Calculate the ratio of the high-frequency acceleration signal to the total energy of the acceleration signal in each control cycle, and denote it as the first ratio.
[0085] Calculate the ratio of the total energy between the high-frequency angular velocity signal and the angular velocity signal in each control cycle, and denote it as the second ratio.
[0086] The average of the first ratio and the second ratio is used as the low-frequency residual value for each control cycle.
[0087] It should be noted that the larger the first or second ratio, the more the energy of the high-frequency signal is, and the more the high-frequency signal is almost the same as the original signal. This indicates that the high-frequency filter has filtered out most of the signal energy, including a large amount of low-frequency energy. This means that less low-frequency signal has been filtered out, and the platform may exceed the motion boundary due to the accumulation of low-frequency displacement, increasing the risk of exceeding limits. The larger the low-frequency residue, the more low-frequency signal remains after high-pass filtering, resulting in a high risk of the platform exceeding limits and requiring an increase in the second-order cutoff frequency.
[0088] Furthermore, based on the low-frequency residuals of all control cycles within each cluster, combined with motion assessment values, the risk of exceeding limits is determined, specifically as follows:
[0089] The product of the mean of the low-frequency residuals of all control cycles within each cluster and the motion assessment value is used as the excess risk degree of each cluster.
[0090] It should be noted that the over-limit risk level reflects the risk of the six-degree-of-freedom motion platform exceeding limits under the driving conditions corresponding to that cluster due to an unreasonable setting of the fixed second-order cutoff frequency. The higher the value, the more intense the vehicle motion and the less low-frequency energy is filtered. Consequently, the extension of the telescopic rod on the six-degree-of-freedom motion platform is closer to the limit value, and the greater the possibility of over-limit risk. Therefore, it is more necessary to increase the second-order cutoff frequency of the rinsing algorithm to speed up the rinsing process, avoid over-limit risk, and improve simulation safety.
[0091] Thus, the out-of-limit risk level of each cluster is obtained.
[0092] Step 3: Calculate the response difference degree of each cluster by considering the differences in elongation and tilt angle of the same telescopic rod between different control cycles within each cluster. Combined with the risk degree of exceeding limits, the clusters are classified into different categories. The second-order cutoff frequency of the shuffling algorithm corresponding to each category of clusters is adjusted using the response difference degree. The feature vector of the current control cycle is clustered in real time, and the shuffling algorithm is run at the adjusted second-order cutoff frequency corresponding to its respective cluster to control the six-degree-of-freedom motion platform for motion simulation.
[0093] Furthermore, it is necessary not only to consider the characteristics of driving behavior, but also whether the actual motion response of the platform is consistent. If the motion feedback of the six-degree-of-freedom motion platform under the same driving condition is significantly different, it may mean that although the driving behavior characteristics are similar, the motion response of the motion platform is more likely to be inconsistent because the second-order cutoff frequency of the high-pass filter is not set reasonably, resulting in poor washing effect or filtering out more effective low-frequency information. In this case, it is more necessary to adjust the second-order cutoff frequency to improve the consistency of the platform's motion feedback.
[0094] Based on the above analysis, the response difference degree is calculated by analyzing the differences in the elongation and tilt angle of the same telescopic rod between different control cycles within the cluster. The flowchart of the method for obtaining the response difference degree of each cluster provided in this embodiment is shown in Figure 2, and specifically includes:
[0095] The elongation of each telescopic rod at all times within each control cycle is used to form an elongation sequence; for each cluster, the difference in the elongation sequence of the same telescopic rod between any two control cycles is calculated and recorded as the difference value.
[0096] In this embodiment, the Euclidean distance of the elongation sequence of the same telescopic rod between any two control cycles is calculated and denoted as the difference value. The calculation of the Euclidean distance is a well-known technique and will not be described in detail here.
[0097] Calculate the average of the differences between all telescopic rods between any two control cycles, and use this as the first difference between any two control cycles;
[0098] The tilt angles of the six-degree-of-freedom motion platform at all times within each control cycle are compiled into an angle sequence; the difference between the angle sequences between any two control cycles is calculated and denoted as the second difference.
[0099] In this embodiment, the Euclidean distance of the angle sequence between any two control cycles is calculated and denoted as the second difference.
[0100] The sum of the first difference and the second difference is taken as the relative difference between any two control cycles.
[0101] The mean of the relative differences between any two control periods within each cluster is taken as the response difference of each cluster.
[0102] It should be noted that the larger the difference value, the larger the first difference, indicating that under the same driving condition represented by the cluster, the more significant the inconsistency in the motion trajectory of the same telescopic rod between two control cycles, reflecting a large difference in the platform's motion response between different control cycles; the larger the second difference, indicating that the overall spatial attitude of the motion platform is very inconsistent between different control cycles under the same driving condition, reflecting a large difference in the platform's overall attitude between different control cycles; the larger the relative difference, indicating that the platform's overall motion response is significantly different in these two control cycles, and the greater the difference in response, the larger the difference in motion response, indicating a large difference in the platform's motion response and attitude changes under the driving condition represented by the cluster, requiring adjustment of the second-order cutoff frequency to improve the consistency of motion feedback.
[0103] Secondly, based on the response difference and the risk of exceeding limits, the discriminant coefficient is determined as follows:
[0104] The product of the response difference and the risk of exceeding the limit is used as the discriminant coefficient for each cluster.
[0105] It should be noted that the larger the discrimination coefficient, the higher the risk of the fixed second-order cutoff frequency being used, and the more significant the inconsistency in the platform's motion response under the same operating conditions. This reflects that the fixed second-order cutoff frequency is poorly matched with the driving conditions represented by the cluster, and therefore the second-order cutoff frequency needs to be adjusted.
[0106] Furthermore, based on the discriminant coefficient, the clusters are divided, and the second-order cutoff frequency of the shuffling algorithm is adjusted using the response difference, specifically as follows:
[0107] Calculate the mean and standard deviation of the discriminant coefficients of all clusters, and obtain the first threshold interval according to the Laida criterion; clusters whose discriminant coefficients are not distributed in the first threshold interval are recorded as out-of-limit clusters;
[0108] Since the problem of unreasonable second-order cutoff frequency settings is not limited to over-limit clusters, there may also be cases where the second-order cutoff frequency is set too high, resulting in the loss of low-frequency features and affecting the realism of the simulation. In such cases, the second-order cutoff frequency needs to be lowered. Therefore, there are clusters in the clusters that have a second-order cutoff frequency that needs to be lowered.
[0109] Calculate the mean and standard deviation of the response variability of all clusters except the overlimit clusters, and obtain the second threshold interval according to the Laida criterion. The clusters whose response variability is not distributed in the second threshold interval are denoted as downregulated clusters. All clusters except the overlimit clusters and downregulated clusters are denoted as other clusters.
[0110] It should be noted that the Laida criterion, also known as the 3σ criterion, is a well-known technique. For ease of understanding, it is assumed that the mean and standard deviation of the discriminant coefficients of all clusters are respectively... and Then the first threshold interval is Assume that the mean and standard deviation of the response variability of all clusters except the overlimit cluster are respectively and Then the second threshold interval is .
[0111] The adjusted second-order cutoff frequency for each cluster is calculated using the following formula:
[0112]
[0113] in, For the first Each cluster corresponds to an adjusted second-order cutoff frequency. For the first Each cluster corresponds to the second-order cutoff frequency before adjustment. For the first The response difference corresponding to each cluster For normalization function, Let be the set of all transfinite clusters. For the set of all downregulated clusters, For the set of all other clusters, Indicates belonging to, among which, , The universal set represents the set of all clusters. This indicates finding the union of sets. In this embodiment, the tanh function is used for normalization. The tanh function is a well-known technique and will not be described in detail here. As for other implementation methods, implementers may use other methods of the prior art, such as the sigmoid function. This embodiment does not impose any special restrictions on this.
[0114] In this embodiment, the second-order cutoff frequency of all clusters before adjustment is 2.2. As for other implementation methods, implementers can set it according to the actual situation.
[0115] It should be noted that for clusters exceeding the limits, their second-order cutoff frequency needs to be increased to avoid platform over-limit situations and improve simulation safety, while for clusters decreasing the limits, their second-order cutoff frequency needs to be decreased to retain more low-frequency motion information and improve the simulation realism of the platform.
[0116] Based on the acceleration and angular velocity signals of the current control cycle, obtain the feature vector of the current control cycle;
[0117] Clustering is performed on the feature vector of the current control cycle and the feature vectors of all control cycles in the experimental dataset. The adjusted second-order cutoff frequency corresponding to the cluster to which the current control cycle belongs is used as the second-order cutoff frequency of the shuffling algorithm corresponding to the current control cycle. The shuffling algorithm is used to obtain the target posture that the six-degree-of-freedom motion platform needs to reproduce. The inverse kinematics solution is performed to solve the target elongation of each telescopic rod on the six-degree-of-freedom motion platform. The target elongation is transmitted to the servo controller. The servo controller drives the servo motor to control the elongation of each telescopic rod in the six-degree-of-freedom motion platform, and finally drives the six-degree-of-freedom motion platform to reproduce the target motion, thereby improving the safety and realism of vehicle motion simulation.
[0118] Based on the same inventive concept as the above method, this application embodiment also provides a vehicle six-degree-of-freedom motion simulation device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described vehicle six-degree-of-freedom motion simulation methods.
[0119] It should be understood that although the steps in the flowchart of Figure 1 are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in Figure 1 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.
[0120] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0121] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.
Claims
1. A method for simulating six-degree-of-freedom motion of a vehicle, characterized in that, The method includes the following steps: Acceleration and angular velocity signals are collected during multiple vehicle driving simulation tests at different control cycles. High-frequency acceleration and angular velocity signals are obtained using a pre-set second-order cutoff frequency extraction algorithm. The tilt angle of the six-degree-of-freedom motion platform and the elongation of each telescopic rod are also collected at each moment within each control cycle to form a test dataset. Feature vectors reflecting motion volatility and average level are extracted from the acceleration and angular velocity signals, and all control cycles within the test dataset are clustered. Based on the difference in the elongation of the telescopic rods from their elongation limits within each cluster and the aforementioned motion volatility, the motion evaluation value of each cluster is calculated. The high-frequency acceleration of each control cycle is evaluated. The energy differences between velocity and acceleration signals, and between high-frequency angular velocity and angular velocity signals in the frequency domain are used to calculate the low-frequency residual of each control cycle. Combined with the motion evaluation value, the over-limit risk degree of each cluster is obtained. By analyzing the differences in the elongation and tilt angle of the same telescopic rod between different control cycles within each cluster, the response difference degree of each cluster is calculated. Combined with the over-limit risk degree, the clusters are classified into different classes, and the second-order cutoff frequency of the corresponding washout algorithm for each cluster is adjusted using the response difference degree. The feature vector of the current control cycle is clustered in real time, and the washout algorithm is run at the adjusted second-order cutoff frequency corresponding to its respective cluster to control the six-degree-of-freedom motion platform for motion simulation.
2. The vehicle six-degree-of-freedom motion simulation method as described in claim 1, characterized in that, The feature vector construction process is as follows: For each control cycle in the experimental dataset, calculate the dispersion and mean of the acceleration at all times in the acceleration signal of each control cycle, and denot them as the first dispersion and the first mean, respectively; calculate the dispersion and mean of the angular velocity at all times in the angular velocity signal of each control cycle, and denot them as the second dispersion and the second mean, respectively; take the sum of the first dispersion and the second dispersion as the motion discrete value of each control cycle; and use the first mean, the second mean, and the motion discrete value to form the feature vector of each control cycle.
3. The vehicle six-degree-of-freedom motion simulation method as described in claim 2, characterized in that, The calculation of the motion evaluation value for each cluster includes: obtaining the limit elongation of the telescopic rod on the six-degree-of-freedom motion platform; selecting the maximum value of the elongation of all telescopic rods at all times under all control cycles within each cluster, and recording it as the maximum extension value; taking the difference between the limit elongation and the maximum extension value as the limit deviation of each cluster; calculating the mean of the motion discrete values for all control cycles within each cluster, and taking it as the motion variability of each cluster; and taking the ratio of the motion variability to the limit deviation as the motion evaluation value of each cluster.
4. The method for simulating six-degree-of-freedom motion of a vehicle as described in claim 1, characterized in that, The calculation of the low-frequency residual for each control cycle includes: performing frequency domain analysis on the acceleration signal, angular velocity signal, high-frequency acceleration signal, and high-frequency angular velocity signal of each control cycle, obtaining the spectrum diagrams for each, and recording the sum of the energies of all frequency components in the spectrum diagrams as the total energy; calculating the ratio of the total energy between the high-frequency acceleration signal and the acceleration signal in each control cycle, and recording it as the first ratio; calculating the ratio of the total energy between the high-frequency angular velocity signal and the angular velocity signal in each control cycle, and recording it as the second ratio; the low-frequency residual is the average of the first ratio and the second ratio.
5. The vehicle six-degree-of-freedom motion simulation method as described in claim 1, characterized in that, The risk level exceeding the limit is the product of the mean of the low-frequency residuals of all control cycles within each cluster and the motion assessment value.
6. The vehicle six-degree-of-freedom motion simulation method as described in claim 1, characterized in that, The calculation of the response difference for each cluster includes: forming an elongation sequence from the elongation of each telescopic rod at all times within each control cycle; calculating the difference in the elongation sequence of the same telescopic rod between any two control cycles for each cluster, and recording it as a difference value; calculating the average of the difference values of all telescopic rods between any two control cycles, and using it as the first difference between the two control cycles; forming an angle sequence from the tilt angles of the six-degree-of-freedom motion platform at all times within each control cycle; calculating the difference in the angle sequence between any two control cycles, and recording it as a second difference; and using the sum of the first difference and the second difference as the relative difference between the two control cycles; the response difference is the average of the relative differences between any two control cycles within each cluster.
7. The vehicle six-degree-of-freedom motion simulation method as described in claim 1, characterized in that, The process of classifying clusters into different categories includes: using the product of response variability and out-of-limit risk as the discriminant coefficient for each cluster; calculating the mean and standard deviation of the discriminant coefficients for all clusters and obtaining a first threshold interval based on the Laida criterion; designating clusters whose discriminant coefficients are not distributed within the first threshold interval as out-of-limit clusters; calculating the mean and standard deviation of the response variability for all clusters other than out-of-limit clusters and obtaining a second threshold interval based on the Laida criterion; designating clusters whose response variability is not distributed within the second threshold interval as down-regulated clusters; and designating all clusters other than out-of-limit clusters and down-regulated clusters as other clusters.
8. The vehicle six-degree-of-freedom motion simulation method as described in claim 7, characterized in that, No. Each cluster corresponds to an adjusted second-order cutoff frequency. The calculation formula is: ,in, For the first Each cluster corresponds to the second-order cutoff frequency before adjustment. For the first The response difference corresponding to each cluster For normalization function, Let be the set of all transfinite clusters. For the set of all downregulated clusters, For the set of all other clusters, Indicates belonging to, among which, ， The universal set represents the set of all clusters. This indicates finding the union of sets.
9. The method for simulating six-degree-of-freedom motion of a vehicle as described in claim 1, characterized in that, The motion simulation of the six-degree-of-freedom motion platform includes: obtaining the feature vector of the current control cycle based on the acceleration and angular velocity signals of the current control cycle; clustering the feature vector of the current control cycle and the feature vectors of all control cycles in the experimental dataset; using the adjusted second-order cutoff frequency corresponding to the cluster of the current control cycle as the second-order cutoff frequency of the washout algorithm corresponding to the current control cycle; using the washout algorithm to obtain the target posture that the six-degree-of-freedom motion platform needs to reproduce; performing inverse kinematics to solve the elongation of each telescopic rod on the six-degree-of-freedom motion platform; and using the elongation to drive the servo motor to control the six-degree-of-freedom motion platform to perform motion simulation.
10. A vehicle six-degree-of-freedom motion simulation device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the vehicle six-degree-of-freedom motion simulation method as described in any one of claims 1-9.