Automatic driving track planning method for actively relieving motion sickness of automobile passenger

By introducing discrete point reference path smoothing and occupant motion sickness cost indicators in the trajectory planning algorithm of autonomous vehicles, combining spiral curve and Bezier curve planning, the six-degree of freedom subjective vertical conflict model is used to solve the occupants in the existing technology The problem of poor slowing motion sickness has achieved more accurate driving decisions and a comfortable riding experience.

CN118977737AInactive Publication Date: 2024-11-19JILIN UNIVERSITY

Patent Information

Application Number
CN202411473334.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-22
Publication Date
2024-11-19
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

There is still room for improvement in the accuracy of standard in terms of slowing occupants' motion sickness in existing autonomous vehicles. Traditional methods mainly rely on planning control algorithms, and the quality of perception and positioning cannot directly determine the vehicle. Sports quality.

Method used

Based on the conventional trajectory planning algorithm based on state space sampling, a performance evaluation cost function of reference path smoothing and occupant motion sickness cost indicators is introduced based on discrete points. The path planning of spiral curves and Bezier curves is combined with six freedoms The subjective vertical conflict model is used to calculate the incidence of occupants' motion sickness and optimize the trajectory to slow motion sickness.

Benefits of technology

In high-speed scenarios, the trajectory planning algorithm effectively slows down occupants' motion sickness, and the motion sickness index (MSI) has dropped by 4.22%, providing more accurate and comfortable driving decisions and improving the riding experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118977737A_ABST
    Figure CN118977737A_ABST
Patent Text Reader

Abstract

The invention discloses an automatic driving track planning method for actively relieving motion sickness of an automobile passenger in the technical field of automatic driving, which comprises the following steps of: smoothing a reference path; a reference path with smaller curvature and oscillation frequency is generated, and trajectory planning is decoupled into path planning based on a spiral curve and speed planning based on a Bezier curve; a path planning result and a corresponding speed planning result are combined pairwise to obtain a track cluster, and a passenger motion sickness occurrence rate calculation model under the condition of no person replacement sensor is established based on a six-degree-of-freedom subjective vertical conflict model; according to the method, a passenger motion degree cost evaluation function is obtained through calculation, cost evaluation is carried out on each track in the obtained track cluster, track efficiency, comfort and driving behavior reasonability are evaluated, the track with the minimum cost is obtained through integration, namely the optimal track, and more comfortable riding experience can be provided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an autonomous driving trajectory planning method for actively alleviating motion sickness of car occupants. Background Art

[0002] During the ride, passengers may experience symptoms such as sweating, vomiting, stomach discomfort, pale complexion, and lack of concentration. These symptoms are collectively referred to as motion sickness, commonly known as motion sickness. The susceptibility to motion sickness varies from person to person, and almost everyone has experienced this discomfort. Previous studies have shown that when the vehicle movement expected by the passenger is inconsistent with the actual movement, motion sickness may occur, and low-frequency vibrations between 0.1Hz and 0.5Hz are also prone to induce motion sickness. Therefore, the ergonomics and chassis design optimization of the vehicle have become a common way to solve this problem. At present, the motion sickness evaluation model for occupant motion sickness has developed rapidly, and how to improve and apply these models has become a hot topic in current motion sickness-related research. By incorporating these models into the autonomous driving algorithm, the system can combine the vehicle's real-time driving information and the occupant's physiological state information to make more accurate and comfortable driving decisions, thereby providing a more comfortable riding experience.

[0003] Considering that the application of self-driving cars will cause the driver to change his role to that of a passenger, the passengers' primary and secondary tasks and riding behavior will change accordingly. As we can see from the above, passengers are more prone to motion sickness than drivers.

[0004] The current methods for alleviating motion sickness in traditional automobiles are mainly based on the design of the vehicle body, cockpit, and chassis. The methods directly improve the ability of passengers to predict the future movement of the vehicle, reduce the conflict between vestibular and visual inputs, and reduce the occurrence of harmful motion stimuli to alleviate motion sickness symptoms. Improving the passenger's field of vision can reduce the occurrence of motion sickness. The field of vision can be improved simply and conveniently by increasing the window surface area and lighting openings of the vehicle. It can also be done by minimizing the obstacles of the A, B, and C pillars during automobile design, or by using a low waistline, low shoulder line shape design, and appropriate seat height to avoid poor vision. To reduce the impact of unstable posture, the posture control of passengers can be improved by designing rotating seats or seats that allow people to lie on their backs. In addition, seats with fixed headrests can be used to limit head movement. The impact of in-car odor on motion sickness can be improved by using air purifiers or fragrance systems. For example, a 20% concentration of ginger flower odor can be used to conduct a motion sickness reduction experiment for passengers, or anti-motion sickness gas generators integrated in air conditioning equipment or wearable respiratory devices can be used to release anti-motion sickness gas to reduce motion sickness. Motion sickness can also be treated directly with the help of medication. The use of motion sickness medicine and motion sickness patches can effectively reduce the severity of motion sickness.

[0005] By investigating the research results of many current research institutions and autonomous driving manufacturers, it can also be found that the method of reducing motion sickness of autonomous driving vehicle occupants focuses on decision-making planning and motion control algorithms. In comparison, the quality of perception and positioning cannot directly determine the quality of vehicle motion. The existing methods of reducing motion sickness of autonomous driving vehicle occupants are mainly based on planning control algorithms, with optimal control as the main solution. Most of them use MSDV as a motion sickness evaluation index, use the MPC algorithm for obstacle avoidance, and use the linear variable parameter (LPV) method to design an LQR controller for path tracking. There is still room for improvement in algorithm diversity and the accuracy of motion sickness evaluation indicators. Summary of the invention

[0006] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0007] Therefore, the purpose of the present invention is to provide an automatic driving trajectory planning method for actively alleviating motion sickness of car occupants. On the basis of the conventional trajectory planning algorithm based on state space sampling, a reference path smoothing based on discrete points and a performance evaluation cost function considering the occupant motion sickness cost index are introduced to improve the conventional trajectory planning algorithm against motion sickness, so that the system can combine the vehicle's real-time driving information and the occupant's physiological state information to make more accurate and comfortable driving decisions, thereby providing a more comfortable riding experience.

[0008] To solve the above technical problems, according to one aspect of the present invention, the present invention provides the following technical solutions: An automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants, characterized by comprising: S1. Smoothing the reference path to generate a reference path with smaller curvature and oscillation frequency; S2. Decouple the trajectory planning into path planning based on spiral curves and speed planning based on Bezier curves, and combine the path planning results and the corresponding speed planning results in pairs to obtain trajectory clusters; S3. Establish a model for calculating the incidence of motion sickness of passengers without replacement sensors based on the six-degree-of-freedom subjective vertical conflict model; S4. In the occupant motion sickness incidence calculation model under the condition of no replacement sensor, the occupant motion cost evaluation function is calculated; S5. For each trajectory in the trajectory cluster obtained in step S2, a cost evaluation is performed to assess the trajectory efficiency, comfort, and rationality of the driving behavior. The trajectory with the lowest comprehensive cost is the optimal trajectory.

[0009] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants described in the present invention, in step S1, the step of smoothing the reference path is as follows: Construct the optimization objective function and then use the conjugate gradient method to solve the components of the reference path Cartesian discrete point coordinates ,……, ,……, The conjugate gradient method continuously updates the optimal result in each iteration and ends the iteration process after the convergence condition is met.

[0010] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants described in the present invention, in step S2, the steps of path planning based on the spiral curve are as follows: Use a cubic spiral curve to connect the first and last points, as shown in formula 1; one Assuming the initial state , target state , is the curve length, substitute into formula 1, we can get formula 2: ; Use the trapezoidal integration method, as shown in formula 3: ; ; Through continuous iterative calculation of Formula 2 and Formula 3, the discrete path point set of the curve is obtained.

[0011] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants described in the present invention, the steps of velocity planning based on Bezier curve are as follows: The Bezier curve expression is formula 4: ; Where: is the proportionality factor The 2D coordinates of the lower curve, For the The 2D coordinates of the control points, is the order of the Bezier curve, is the Bernstein basis function; ; The proportional term in the Bezier curve expression Converted into a time-related function, we get the parameter expression of the Bezier curve in the time domain: ; Where t is time, the sampling initial time t=0, and the sampling end time t=T; Therefore, the velocity, acceleration and path length functions are Equation 6, Equation 7 and Equation 8 respectively: ; ; ; Among them, t f It represents the end time of the motion process, A represents the starting speed v0 in the speed formula, B represents the starting acceleration a0 in the acceleration formula, C represents the coefficient of change of speed and acceleration in the middle process, and D represents the speed v at the end point f With acceleration a f The endpoint value of Given starting speed , initial acceleration and terminal velocity , terminal acceleration , and the length of the path Get formula nine: .

[0012] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants described in the present invention, in step S3, the six-degree-of-freedom subjective vertical conflict model includes a perception module A, a feedback gain module B and an internal model module C; The input of the six-DOF subjective vertical conflict model is the reference gravity-inertia acceleration and angular velocity of the occupant's head, which produces the perceived gravity acceleration, inertia acceleration, and angular velocity; A second-order integrator was used to map the one-dimensional index output by the Hill function of the six-DOF subjective vertical conflict model to the motion sickness incidence MSI, and the experimental results were fitted through parameter calibration to obtain Equation 10: ; Where: is the maximum percentage of people who suffer from motion sickness, is the time constant, and s represents the exposure time, that is, the time the occupant is exposed to acceleration or vibration stimulation in the vehicle.

[0013] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of automobile occupants according to the present invention, in step S3, a calculation model for the incidence of motion sickness of occupants without replacement sensors is established based on the six-degree-of-freedom subjective vertical conflict model: a conversion module for obtaining occupant head motion information through vehicle motion information is added to the six-degree-of-freedom subjective vertical conflict model to obtain a calculation model for the incidence of motion sickness of occupants without replacement sensors, specifically including: The origin of the occupant's head coordinate system is transferred to the center of gravity of the car, and the acceleration generated when the vehicle rotates is ignored so that the angular relationship between the car and the occupant's head is expressed by the car's angle vector and the linear head motion factor describing body motion After multiplying element by element and rotating, we get: ; Where: ——Hadamard product, which represents the product of corresponding elements of the matrix, t represents time, It represents the head angle changing with time; - rotation matrix, ,in , , ; Linear head motion factor is defined as the linear relationship between vehicle motion and occupant head motion, defining ; After obtaining the head angle, the corresponding head angular velocity can be obtained by taking the derivative, as shown in Formula 12: ; The linear acceleration of the occupant's head and the gravity-inertia acceleration do not require the conversion of the linear head motion factor. and the acceleration due to gravity Based on the direct rotation transformation, we can get it as shown in Equation 13 and Equation 14: ; ; The angular velocity and acceleration of the occupant's head can be obtained from the vehicle's angular velocity and acceleration through the conversion module, which is then connected to the six-degree-of-freedom subjective vertical conflict model to form a motion sickness incidence calculation model under sensorless conditions, and the MSI value can be calculated.

[0014] As a preferred solution of the automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants according to the present invention, in step S4, the occupant motion cost evaluation function is specifically as shown in Formula 15: ; ; ; in, , , , , is the weight coefficient, is the improved total time cost function, is the total space cost function.

[0015] Compared with the prior art, the present invention has the following beneficial effects: on the basis of the conventional trajectory planning algorithm based on state space sampling, the present invention introduces a reference path smoothing based on discrete points and a performance evaluation cost function considering the occupant motion sickness cost index, and improves the conventional trajectory planning algorithm to prevent motion sickness. According to simulation results, the trajectory planning algorithm can effectively alleviate occupant motion sickness in high-speed scenarios, and the MSI decreases by 4.22%, achieving the effect of alleviating motion sickness, so that the system can combine the vehicle's real-time driving information and the occupant's physiological state information to make more accurate and comfortable driving decisions, thereby providing a more comfortable riding experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below in combination with the accompanying drawings and detailed embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them: Figure 1 This is a flow chart of an automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants according to the present invention; Figure 2 A system block diagram of an automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants according to the present invention; Figure 3 A six-degree-of-freedom subjective vertical conflict model provided by the present invention; Figure 4 A calculation model for the incidence of motion sickness of passengers provided by the present invention under the condition of no replacement sensor; Figure 5 A six-degree-of-freedom subjective vertical conflict model diagram introducing conflict conversion quantity provided by the present invention; Figure 6A schematic diagram of a road module provided by the present invention; Figure 7 The driving path provided by the present invention is likely to cause motion sickness to passengers; Figure 8 A schematic diagram of a trajectory tracking result provided by the present invention; Fig. 9 and Fig.10 A schematic diagram of data comparison between the conventional trajectory and the anti-sickness trajectory (a)-(h) provided by the present invention; Fig.11 A schematic diagram showing the comparison of equivalent acceleration between the conventional trajectory and the anti-sickness trajectory provided by the present invention; Fig.12 A schematic diagram of MSI comparison between the conventional trajectory and the anti-sickness trajectory provided by the present invention. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0018] Secondly, the present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0019] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0020] See also Figure 1 and Figure 2 The steps of an automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants of the present invention are as follows: 1. Reference path smoothing The most commonly used reference path smoothing algorithm is the optimization algorithm based on discrete points. The optimization variables are the components of the reference path. Cartesian discrete point coordinates ,……, ,……, The present invention first constructs an optimization objective function and then uses the conjugate gradient method to solve the The optimal solution set of discrete points.

[0021] The conjugate gradient method will continuously update the optimal result in each iteration, and end the iteration process after the convergence condition is met. Finally, a balance is achieved between several objective functions to achieve the purpose of path smoothing.

[0022] 2. Path planning based on spiral curves The cubic spiral curve expression is a polynomial function of curvature with respect to the vehicle's travel distance. When the curvature function of a curve is given, the geometric shape of the curve can be completely determined. Compared with polynomial spline curves, spiral curves can easily meet the approximate curvature constraints required in path planning problems, ensure curvature continuity, avoid excessive lateral acceleration caused by sharp turns, and also meet position constraints and tangent conditions at boundary points.

[0023] To ensure the feasibility of the path planning results, the vehicle driving curve needs to satisfy the kinematic equation. Therefore, after determining the current target state of the vehicle, the model needs to be used to solve the continuous path from the starting point to the end point. The path planning problem is transformed into a two-point boundary value problem. In solving this problem, this paper uses a cubic spiral curve to connect the first and last points, as shown in Equation 1.

[0024] one Where s is the arc length.

[0025] Assuming the initial state , target state , is the curve length, substituting it into formula 1, we can get: two The curvature polynomial of the curve is calculated, and in theory a comfortable and smooth path is obtained, but the specific path point parameters are still unknown. Therefore, in practical applications, it is necessary to sample to obtain discrete path points before subsequent trajectory evaluation can be performed. Because the final generated point is a path cluster point set, the more efficient trapezoidal integration method is used here, as shown in Formula 3: ; three By continuously iterating and calculating Formula 2 and Formula 3, the discrete path point set of the curve can be obtained for subsequent operations.

[0026] 3. Speed ​​planning based on Bezier curve The Bezier curve can naturally generate a smooth curve to ensure the continuity and smoothness of the speed and avoid sudden changes. The Bezier curve expression is as follows: Four Where: ——Proportional coefficient The 2D coordinates of the lower curve; ——No. The two-dimensional coordinates of the control points; ——the order of the Bezier curve; ——Bernstein basis functions, ; The tangent vector and second-order derivative vector of any point on the curve with respect to time can correspond to the speed and acceleration of the vehicle during driving, so as to consider the vehicle motion constraints in the trajectory planning link. However, the standard Bezier curve equation does not have a parameter related to time. Therefore, the proportional term in the Bezier curve Converted into a time-related function, the parameter expression of the Bezier curve in the time domain can be obtained as Equation 5: five Therefore, the velocity, acceleration and path length functions can be designed as: six seven eight Among them, t f It represents the end time of the motion process, A represents the starting speed v0 in the speed formula, B represents the starting acceleration a0 in the acceleration formula, C represents the coefficient of change of speed and acceleration in the middle process, and D represents the speed v at the end point f With acceleration a f The endpoint value of Given starting speed , initial acceleration and terminal velocity , terminal acceleration , and the length of the path .

[0027] Nine 4. Conventional trajectory selection The path planning results and the corresponding speed planning results are combined in pairs to obtain trajectory clusters. In order to select the best solution, a cost evaluation is required for each trajectory to evaluate the trajectory efficiency, comfort and rationality of driving behavior. The trajectory with the lowest cost is the best choice.

[0028] The total cost function is shown in Formula 10, which includes space cost and time cost. Dependent on the path geometry, time cost Related to speed, acceleration, etc. in the time dimension.

[0029] ten Where: ——No. candidate trajectories; ——Total number of candidate trajectories; , ——Weight coefficient.

[0030] The space cost and time cost are introduced below respectively.

[0031] (1) Space cost Track length cost

[0032] eleven Where: ——Candidate trajectories Length of trajectory along the Frenet coordinate system; ——The trajectory length threshold along the Frenet coordinate system.

[0033] The cost term represents the preference for longer trajectories, which avoids possible overreaction of the vehicle due to trajectories that are too short.

[0034] Curvature Cost

[0035] twelve Where: ——Candidate trajectories Curvature value; ——Track curvature threshold.

[0036] The cost term represents the smoothness of the trajectory, and the comfort of lateral motion can be improved by introducing this term.

[0037] Cost of deviation from reference path

[0038] Thirteen Where: ——Candidate trajectories Length of lateral deviation from the reference path; ——The reference path lateral deviation length threshold.

[0039] The cost term represents the preference of the trajectory to the reference path to ensure that the vehicle can follow the guidance of the reference path.

[0040] Trajectory continuity costs

[0041] fourteen Where: ——Candidate trajectories The terminal lateral offset; ——The lateral offset of the terminal that selected the optimal trajectory in the previous frame.

[0042] The cost term represents the preference for trajectory continuity to prevent the inconsistency between adjacent replannings from causing oscillations, overshoots, or even instability in the control system.

[0043] In summary, the space cost is shown in Table 1, and each cost item is standardized as .

[0044] Table 1 Space cost

[0045] So the total space cost function As in Formula 15: fifteen Where: , , , ——Weight coefficient.

[0046] (2) Time cost Trajectory time cost

[0047] sixteen Where: ——Candidate trajectories The length of the trajectory; ——Trace time length threshold.

[0048] The cost term represents the preference of the trajectory for shorter travel time. The larger the cost term, the more maneuverable the vehicle is.

[0049] Trajectory longitudinal acceleration cost

[0050] XVII Where: - longitudinal acceleration of the candidate trajectory; ——Trace longitudinal acceleration threshold.

[0051] The cost item represents the trajectory's preference for comfort. Excessive acceleration can be avoided by setting this cost item.

[0052] Track vertical jerk cost

[0053] eighteen Where: ——jerk of candidate trajectories; - Trajectory jerk threshold.

[0054] The cost item also represents the trajectory’s preference for comfort. The smaller the indicator is, the smoother the acceleration is and the better the riding experience will be.

[0055] In summary, the time cost is shown in Table 2, and each cost item is standardized as .

[0056] Table 2 Space cost

[0057] Total time cost function As in formula 19: nineteen Where: , ——Weight coefficient.

[0058] 5. Consider motion sickness and choose the best trajectory 1. Six-DOF Subjective Vertical Conflict Motion Sickness Evaluation Model Considering the application of this article in the trajectory planning algorithm, the complex activities of the occupants during the ride can be temporarily ignored. Passive motion is sufficient to briefly characterize the motion state of the occupants when riding in an autonomous vehicle. Therefore, this article uses the existing 6DOF-SVC six-degree-of-freedom subjective vertical conflict motion sickness evaluation model that includes vision, prediction and other information.

[0059] The 6DOF-SVC model is established in the human head reference system and consists of three main parts: perception module A, feedback gain module B and internal model module C, such as Figure 3 , the input is the linear acceleration of the human head in actual motion and the angular velocity .

[0060] The sensory module A represents the sensory dynamics of the vestibular system, where The module represents the otolith organ used to sense linear acceleration. The module represents the semicircular canal organ used to sense the angular velocity of rotation. Based on the study of the characteristics of the dynamic response of the two, the transfer function of the otolith is generally set to the unit matrix, and the transfer function of the semicircular canal is set to Formula 20: twenty Where: ——The angular velocity of rotation actually sensed by the semicircular canals; ——Time constant, usually set to 7s; ——Time constant, usually set to 190s.

[0061] When the otolith receives linear acceleration, according to Einstein's equivalence principle, the combined force of gravity and inertia will act on the otolith at the same time, as shown in Formula 21: Twenty-one Where: ——The gravity-inertial acceleration actually sensed by the otoliths; ——Acceleration due to gravity.

[0062] The model input is the reference gravity-inertial acceleration and angular velocity of the occupant's head, which will produce the perceived gravity acceleration, inertial acceleration and angular velocity. Mayne (1974) believed that in order to separate the gravity component and the inertial acceleration component, it can only be based on a distinguishing feature, that is, one considers constancy and the other considers transientness. For normal human and animal movements, the acceleration encountered is always short-lived, or periodic, so it is assumed that the gravity acceleration is recognized by the system as the constant part of the signal and the inertial acceleration is recognized as the transient part of the signal, so the central nervous system will pre-process the gravity-inertial acceleration in the form of frequency separation. If only linear acceleration is involved, a low-pass filter can be used in the earth reference frame to filter out the gravity transmitted by the otoliths. For more complex motion stimuli involving angular velocity, the gravity acceleration can be separated from the linear acceleration through the interaction between the otoliths and the semicircular canal vestibular organs. This process is The module can be calculated by Mayne equation, as shown in Formula 22: XXII Where: ——The gravitational acceleration actually perceived by the human body, that is, the perceived vertical; —— Time constant, usually set to 5s; After processing by this module, gravity-inertia acceleration can be processed to perceive vertical and perceived linear acceleration .

[0063] The internal model module is a hypothetical human central nervous system that can make an estimate of the current self-motion based on previous motion experience, representing the expected motion perception, with the purpose of tracking the perceived gravitational acceleration, inertial acceleration and angular velocity. Bos (1998) believes that it is the tracking error caused by the phase lag of the frequency that causes motion sickness. In the motion perception experiment, the subject can perceive and report the estimation of the relevant information. Its input is the linear acceleration and rotational angular velocity considered by the central nervous system, so the relevant sensory information will undergo certain processing, such as Equation 23 and Equation 24: XXIII XXIV Where: ——The estimated rotational angular velocity of the body; ——Linear acceleration after body estimation; , ——The accuracy of the estimated rotational angular velocity and linear acceleration of the body.

[0064] This module includes the internal model of the sensory organs and the internal model of the interactions between them. The internal model of the sensory dynamics of the vestibular system is similar to the previous one. The module represents the internal model of the otolith, and its transfer function is the same as in the perception module and can be described using the identity matrix, The module represents the internal model of the semicircular canal, and its transfer function is described by Equation 25 according to its dynamic response characteristics: XXV Where: ——The subjective expected angular velocity of rotation obtained by the internal model; ——Time constant, which is the same as the time constant in the semicircular canal transfer function in sensory module A.

[0065] The module still represents the interaction between the internal model of the otoliths and the internal model of the semicircular canals, and its mechanistic role is similar to that of the perception module. The same is specifically described as Formula 26: XXVI Where: - the gravitational inertial acceleration acting on the internal otolith model after being estimated and updated by the body; ——The subjective expected gravitational acceleration obtained by the internal model based on previous movement experience, that is, subjective vertical; ——time constant, and Same in module.

[0066] After processing, the subjective expected result can be obtained. The difference between the result of the perception module includes the error between the perceived rotation angular velocity and the subjective rotation angular velocity. , the error between perceived vertical and subjective vertical , the error between perceived inertial acceleration and subjective inertial acceleration These values ​​will be processed by the feedback gain module B and used as another part of the input of the internal model C. This is why the human body will gradually adapt to motion sickness after receiving long-term motion stimulation. Feedback is given to the internal model of the semicircular canal and the weight is updated as , Feedback is given to the internal model of the otolith and the weight is updated as , Feedback is given to the internal model of the otolith and the weight is updated as .

[0067] According to the subjective vertical conflict theory, This is the ultimate cause of motion sickness. In order to convert the three-dimensional vertical conflict into a one-dimensional index, The Hill function is processed so that the incidence of motion sickness will grow logarithmically when the sensory conflict is small, and exponentially when it is large, as shown in Formula 27: XXVII Where: ——The amount of sensory conflict conversion.

[0068] The final motion sickness rate MSI is the sensory conflict conversion amount Accumulation over time. However, considering that sensory conflict does not accumulate infinitely over time, the maximum is 100%, and the severity of motion sickness gradually increases until it reaches the maximum value, and returns to zero when the sensory conflict stops or disappears. At the same time, in order to reproduce McCauley's experimental results, a second-order integrator is finally used to map the one-dimensional index output by the Hill function to the motion sickness incidence MSI, and the fitting of the experimental results is achieved through parameter calibration, as shown in Formula 28: XXVIII Where: - the maximum percentage of people suffering from motion sickness; ——Time constant, usually set to 720s.

[0069] The optimization results of relevant parameters of the 6DOF-SVC model are shown in Table 3.

[0070] Table 3 Optimized parameters of the motion sickness model for vehicle occupants

[0071] Calculation of motion sickness incidence based on vehicle data The researchers also developed a new numerical model for motion sickness, MSDV, based on previous relevant data, as shown in Equation 29: XXIX Where: - exposure time (20 minutes to 6 hours); ——frequency-weighted acceleration in vertical direction, , ——effective weighted acceleration, ,in Indicates that after considering the acceleration direction weight The acceleration value of each sample, Indicates the sample size.

[0072] Based on the results of MSDV, the MSI indicator can be expressed as: thirty Where: .

[0073] In addition, this indicator can also be extended to the Illness Rate (IR), as shown in the following formula: Thirty-one Where: .

[0074] The illness rating is divided into four levels, with 0 meaning "feeling fine", 1 meaning "mildly unwell", 2 meaning "very unwell" and 3 meaning "very bad".

[0075] The ISO standard divides the calculation of this model into two steps. In the first step, the vertical acceleration is weighted and filtered, and in the second step, it is integrated over time. The result is the cumulative measurement value of motion sickness.

[0076] The transfer functions of the high-pass and low-pass filters can be described as Equation 32 and Equation 33: XXXII XXXIII After acceleration filtering, weighting is performed, and its transfer function can be described by equation 34 and equation 35: XXXIV XXXV Where: ——resonance quality factor; ——weighting factor, , represents the corner frequency.

[0077] The product of represents the band-limited transfer function, The product represents the actual weighted transfer function for a particular application, and the overall frequency weighting function is given by Equation 36: Thirty-six.

[0078] Calculation of occupant motion sickness According to the previous introduction, the 6DOF-SVC model uses the linear acceleration and rotational angular velocity generated by the human head in actual motion as input. However, considering that the relevant human head motion information is unknown when performing trajectory planning in this article, it is necessary to add a conversion module to obtain the occupant's head motion information through the vehicle motion information in the model, such as Figure 4 As shown in the figure, the motion sickness incidence calculation without human body sensors first assumes that the occupant's body has no damping effect without human body sensors, transfers the origin of the head coordinate system to the center of gravity of the car, and ignores the acceleration generated when the vehicle rotates. In this way, the angular relationship between the car and the occupant's head can be obtained by the car's angle vector and the linear head motion factor describing body motion After multiplying element by element and rotating, we get: XXXVII Where: ——Hadamard product, which represents the product of corresponding elements of the matrix; - rotation matrix, ,in , , .

[0079] Linear head motion factor is defined as the linear relationship between the vehicle motion and the occupant's head motion. , then the roll angle of the head is considered to be opposite to the roll angle of the vehicle, the yaw angle of the head corresponds to the yaw angle of the vehicle, and the pitch angle of the head is always 0. This paper defines , which assumes that the occupant's body movement is synchronized with the vehicle's movement, that is, the occupant passively accepts the motion stimulus without deliberately violating the vehicle's motion stimulus, which is equivalent to the car occupant looking straight ahead ( ), the occupants do not have additional motion compensation. In real life, considering a series of physiological and psychological factors of the occupants, this assumption will have a certain error, but if it is only used as a part of the motion sickness incidence calculation to be considered in trajectory planning, compared with directly ignoring the motion information conversion, this error is acceptable.

[0080] After obtaining the head angle, the corresponding head angular velocity can be obtained by derivative: XXXVIII The linear acceleration of the occupant's head and the gravity-inertia acceleration do not require the conversion of the linear head motion factor and can be calculated at the vehicle acceleration. and the acceleration due to gravity Based on the direct rotation transformation, as shown in Equation 39 and Equation 40: XXXIX forty Through the above-mentioned human-vehicle conversion process, the angular velocity and acceleration of the occupant's head can be obtained from the vehicle's angular velocity and acceleration, which are connected with the 6DOF-SVC model to form a motion sickness incidence calculation model under sensorless conditions, and the MSI value can be calculated.

[0081] Improvement of the cost function by introducing motion sickness assessment As mentioned above, the subjective vertical conflict theory states that the error between the vestibular system's perceived verticality and the subjective verticality of the internal model The MSI calculated in the 6DOF-SVC model is the sum of the three-dimensional subjective vertical conflict The Hill function is used to convert the sensory conflict into a one-dimensional sensory conflict. After that, the MSI value with a time constant of 12 minutes is accumulated to evaluate the severity of motion sickness. However, the duration of each re-planned trajectory in trajectory planning is generally only a few seconds, so the MSI value with a large time constant is not suitable for direct use as an evaluation indicator for motion sickness. yes The further processing of the value directly represents the size of the subjective vertical conflict of motion sickness, and it is normalized to between 0 and 1, so it is very suitable for representing the change in the severity of motion sickness in trajectory selection. The sensory conflict conversion quantity is introduced to characterize the motion sickness comfort, such as Figure 5 shown.

[0082] The current trajectory related parameters include position , heading angle , vehicle center of mass speed , longitudinal acceleration and curvature , while the input of the 6DOF-SVC model includes three linear accelerations and three rotational angular velocities, relatively speaking, there are too few known parameters of the current trajectory. In order to calculate a more accurate sensory conflict conversion , the trajectory also needs to introduce other linear acceleration values ​​and rotational angular velocity values.

[0083] Assuming that the vehicle is in steady-state steering and has a small tire slip angle, the vehicle dynamics equations are further derived based on the bicycle model. The yaw rate caused by wheel steering is and lateral acceleration for: XXI Chapter 42 Where: ——Gross vehicle mass (kg); ——vehicle center of mass speed (m / s); , ——Cornering stiffness of the front and rear wheels respectively (N / rad).

[0084] , ——The distance from the center of mass of the vehicle to the front axle and the rear axle respectively (m). Wheelbase

[0085] ——Trajectory curvature.

[0086] Through the above calculation formula, we can get the numerical values ​​of the two vehicle motion parameters and substitute them into Figure 4 The motion sickness incidence calculation model under sensorless conditions is used, and the final calculated MSI value can be used to evaluate the severity of occupant motion sickness in simulation and experiments.

[0087] In the trajectory planning stage, the sensory conflict is converted into As the final result of the model output, the motion sickness severity cost term is the cumulative value of the subjective vertical conflict per unit length of the trajectory, as shown in Formula 43: Chapter 43 Where: ——Candidate trajectories The amount of sensory conflict transformation along the trajectory. The cost term is introduced into the conventional time cost.

[0088] The time cost to obtain the current improvement is shown in Table 4.

[0089] Table 4 Improved time cost

[0090] Therefore, the improved As in Formula 44: Chapter 44 Where: , , , ——Weight coefficient.

[0091] The spatial cost setting is consistent with the conventional trajectory optimization, so the total cost of the final trajectory performance evaluation is As in Formula 45: Chapter 45 Where: , ——Weight coefficient.

[0092] In order to verify the effectiveness of the method of the present invention, a motion sickness-prone path was constructed in PreScan. The motion sickness-prone path or scene features have the following characteristics: First, there are often more than 5m / s 2 The first is the sudden lateral acceleration movement, the second is the high-intensity low-frequency (0.1Hz-0.2Hz) acceleration during the movement, the third is the lateral acceleration generated by the vehicle is synchronized with the yaw oscillation, the fourth is the sudden or discontinuous changes in speed and steering input, and the fifth is the continuous opposite turns on the path.

[0093] Therefore, combined with the motion sickness-prone path characteristics proposed above, the motion sickness-prone scene is designed to be composed of a road module consisting of a short straight road and a large curvature curve, such as Figure 6 , the length of the straight section is set to 100m, and the radius of the curve is set to 45m. After several superpositions are used as simulation experiment scenarios, the following can be designed: Figure 7 The driving path shown is likely to cause motion sickness in the occupants.

[0094] The simulated vehicle starts from a standstill, with a target cruising speed of 20 m / s. The simulation lasts for about 1200 seconds, and the performance of the conventional trajectory and the anti-sickness trajectory are compared. Because the simulation path is long, only the following part is captured when analyzing the simulation results. Figure 8 The data analysis was performed on the path between 500m and 1300m of the middle horizontal coordinate, and the conventional trajectory and anti-sickness trajectory mentioned in the legend and in the illustration below represent the trajectory planning performed under the motion control algorithm described above.

[0095] The vehicle trajectory tracking results under this condition are as follows: Figure 8 As shown in the simulation results, it can be found that most of the time, both can maintain a low road centerline offset of about 0.1m. In the fierce curve part, the conventional track will produce an offset of about 0.25m, and the anti-sickness track will produce an offset of about 0.4m. Figure 8 and Fig. 9 As can be seen in (a), when entering or exiting a curve, the conventional trajectory attempts to cut the curve to increase the turning radius, thereby reducing lateral acceleration and improving comfort. The anti-sickness trajectory has a relatively small change in steering angle, with a maximum value of 4.58 degrees, which is smaller than the 4.81 degrees under the conventional trajectory. Fig.10 (e) This results in a smoother movement when entering and exiting a curve, so the anti-sickness track does not perform a more aggressive cornering operation to reduce lateral acceleration as expected, which ultimately results in a small difference in lateral acceleration between the two, with a maximum value of 3.4m / s 2 about.

[0096] Then from Fig. 9 As can be seen in (c), the maximum speed of the conventional trajectory is 20m / s, while the maximum speed of the anti-sickness trajectory is 19.8m / s. The anti-sickness trajectory does not reach the cruising speed value like the conventional trajectory. This is because the subjective vertical conflict added to the anti-sickness trajectory is essentially related to acceleration, angular velocity, etc., so the algorithm considers this point of view and limits the longitudinal acceleration and yaw angular velocity of the vehicle, such as Fig. 9 (b) Fig. 9 (d) and Fig.10 As shown in (h), the maximum longitudinal acceleration of the conventional trajectory and the anti-sickness trajectory are 4.61m / s 2 and 3.82m / s 2 , down 17.1%, the maximum yaw rate was 16.04deg / s and 15.42deg / s, down 3.75%, and the maximum subjective vertical conflict was 0.48 and 0.47, down 2.08%. Because the straight section of the simulation scene is relatively short, the vehicle failed to accelerate to the cruising speed under the guidance of the anti-sickness trajectory.

[0097] Other vehicle parameters, such as the vehicle roll angle, the peak values ​​of the conventional trajectory and the anti-sickness trajectory were 2.21 degrees and 2.11 degrees, respectively, a decrease of 4.52%, and the vehicle pitch peak values ​​were 1.26 degrees and 1.18 degrees, respectively, a large decrease of 6.35%. It can be seen that the peak values ​​of these indicators of the anti-sickness trajectory are smaller than those of the conventional trajectory, indicating that the anti-sickness trajectory is more stable during driving.

[0098] In the analysis of speed and acceleration data, we also compared the difference in subjective vertical conflict between the two trajectories, which indicates the change in subjective verticality and indirectly represents the magnitude of motion sickness MSI. Because the entire path is similar, the final result is as follows Fig.11 and Fig.12 As shown. After 1200s of simulation, the MSI index of the conventional algorithm is 8.06%, and the MSI index of the anti-sickness trajectory is 7.52%. The MSI of the anti-sickness trajectory is reduced by 9.89% compared with the conventional trajectory. However, because the vehicle speed is slower under the anti-sickness trajectory, the anti-sickness trajectory does not reach the position where the conventional trajectory is completed. Therefore, under the same driving distance, the MSI index of the anti-sickness trajectory is reduced to 7.72%, a decrease of 4.22%. In general, it can be considered that the trajectory planning algorithm can reduce motion sickness.

[0099] Although the present invention has been described above with reference to the embodiments, various modifications may be made thereto and parts thereof may be replaced by equivalents without departing from the scope of the present invention. In particular, as long as there is no structural conflict, the various features in the embodiments disclosed in the present invention may be used in combination with each other in any manner, and the fact that these combinations are not exhaustively described in this specification is only for the sake of omitting space and saving resources. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. An automatic driving trajectory planning method for actively alleviating motion sickness of vehicle occupants, characterized in that: include: S1. Smoothing the reference path to generate a reference path with smaller curvature and oscillation frequency; S2. Decouple trajectory planning into path planning based on spiral curves and speed planning based on Bezier curves, and combine the path planning results and the corresponding speed planning results in pairs to obtain trajectory clusters; S3. Establish a model for calculating the incidence of motion sickness of passengers without replacement sensors based on the six-degree-of-freedom subjective vertical conflict model; S4. In the occupant motion sickness incidence calculation model under the condition of no replacement sensor, the occupant motion cost evaluation function is calculated; S5. For each trajectory in the trajectory cluster obtained in step S2, a cost evaluation is performed to assess the trajectory efficiency, comfort, and rationality of the driving behavior. The trajectory with the lowest comprehensive cost is the optimal trajectory.

2. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: In step S1, the steps of smoothing the reference path are as follows: Construct the optimization objective function and then use the conjugate gradient method to solve the components of the reference path Cartesian discrete point coordinates ,……, ,……, The conjugate gradient method continuously updates the optimal result in each iteration and ends the iteration process after the convergence condition is met.

3. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: In step S2, the steps of path planning based on the spiral curve are as follows: Use a cubic spiral curve to connect the first and last points, as shown in formula 1; ; Where s is the arc length; Assuming the initial state , target state , is the curve length, substitute into formula 1, we can get formula 2: ; Use the trapezoidal integration method, as shown in formula 3: ; ; Through continuous iterative calculation of Formula 2 and Formula 3, the discrete path point set of the curve is obtained.

4. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: The steps of velocity planning based on Bezier curve are as follows: The Bezier curve expression is formula 4: ; Where: is the proportionality factor The 2D coordinates of the lower curve, For the The 2D coordinates of the control points, is the order of the Bezier curve, is the Bernstein basis function; ; The proportional term in the Bezier curve expression Converted into a time-dependent function, the parameter expression of the Bezier curve in the time domain is obtained, as shown in Formula 5: ; Where t is time, the sampling initial time t=0, and the sampling end time t=T; Therefore, the velocity, acceleration and path length functions are Equation 6, Equation 7 and Equation 8 respectively: ; ; ; Among them, t f It represents the end time of the motion process, A represents the starting speed v0 in the speed formula, B represents the starting acceleration a0 in the acceleration formula, C represents the coefficient of change of speed and acceleration in the middle process, and D represents the speed v at the end point f With acceleration a f The endpoint value of Given starting speed , initial acceleration and terminal velocity , terminal acceleration , and the length of the path Get formula nine: 。 5. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: In step S3, the six-degree-of-freedom subjective vertical conflict model includes a perception module A, a feedback gain module B and an internal model module C; The input of the six-DOF subjective vertical conflict model is the reference gravity-inertia acceleration and angular velocity of the occupant's head, which produces the perceived gravity acceleration, inertia acceleration, and angular velocity; A second-order integrator was used to map the one-dimensional index output by the Hill function of the six-DOF subjective vertical conflict model to the motion sickness incidence MSI, and the experimental results were fitted through parameter calibration to obtain Equation 10: ; Where: is the maximum percentage of people who suffer from motion sickness, is the time constant, h is the sensory conflict conversion amount, and s represents the exposure time, that is, the time the occupant is exposed to acceleration or vibration stimulation in the vehicle.

6. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: In step S3, a calculation model for the incidence of occupant motion sickness without a replacement sensor is established based on the six-degree-of-freedom subjective vertical conflict model: a conversion module for obtaining occupant head motion information through vehicle motion information is added to the six-degree-of-freedom subjective vertical conflict model to obtain a calculation model for the incidence of occupant motion sickness without a replacement sensor, specifically including: The origin of the occupant's head coordinate system is transferred to the center of gravity of the car, and the acceleration generated when the vehicle rotates is ignored so that the angular relationship between the car and the occupant's head is expressed by the car's angle vector and the linear head motion factor describing body motion After multiplying element by element and rotating, we get: ; Where: ——Hadamard product, which represents the product of corresponding elements of the matrix, t represents time, It represents the head angle changing with time; - rotation matrix, ,in, , , ; Linear head motion factor is defined as the linear relationship between vehicle motion and occupant head motion, defining ; After obtaining the head angle, the corresponding head angular velocity can be obtained by derivative, as shown in Formula 12: ; The linear acceleration of the occupant's head and the gravity-inertia acceleration do not require the conversion of the linear head motion factor. and the acceleration due to gravity Based on the direct rotation transformation, we can get it as shown in Equation 13 and Equation 14: ; ; The angular velocity and acceleration of the occupant's head can be obtained from the vehicle's angular velocity and acceleration through the conversion module, which is then connected to the six-degree-of-freedom subjective vertical conflict model to form a motion sickness incidence calculation model under sensorless conditions, and the MSI value can be calculated.

7. The method for autonomous driving trajectory planning for actively alleviating motion sickness of vehicle occupants according to claim 1, characterized in that: In step S4, the occupant motion cost evaluation function is specifically as shown in Formula 15: ; ; ; in, , , , , is the weight coefficient, is the improved total time cost function, is the total space cost function, w1: the value is between 0.1 and 0.3, w2 the value is between 0.3 and 0.6, w3 the value is between 0.4 and 0.7, w4 the value is between 0.5 and 0.8, The value is between 0.2 and 0.

5. The value is between 0.5−0.8.

Citation Information

Patent Citations

  • Track planning method and device, electronic equipment and storage medium

    CN116224997A

Cited By

  • Intelligent electric vehicle motion sickness resistance cruise planning method based on subjective and objective comprehensive indexes

    CN119943359A

  • Intelligent Electric Vehicle Anti-Motion Sickness Cruise Planning Method Based on Comprehensive Subjective and Objective Indicators

    CN119943359B

  • Interaction method for inducing and relieving motion sickness of passengers through driving behaviors of driver

    CN120422878A

  • A driver driving behavior inducing interaction method for relieving passenger motion sickness

    CN120422878B

  • Automatic driving control method based on motion sickness perception and comfort of passengers

    CN121106313A