Human-machine sharing based intelligent electric vehicle lateral control system and method

Abstract

The application provides a human-machine shared intelligent electric automobile lateral control system and method, relates to the technical field of intelligent automobile control, and comprises the following steps: obtaining a current steering wheel torque and its differential, a longitudinal vehicle speed, a yaw angular velocity and a desired front wheel turning angle; determining a driver intervention intention level according to the absolute value of the torque and determining an initial weight coefficient; constructing a steering quality coefficient by analyzing the time domain fluctuation characteristics of the torque and its differential, and simultaneously correcting the first target front wheel turning angle of automatic driving based on the vehicle speed and the desired turning angle; performing adaptive smoothing filtering on the torque according to the quality coefficient to generate a driver demand turning angle, and then weighting and fusing the first target turning angle and the driver demand turning angle according to the initial weight to obtain a second target front wheel turning angle of shared control and output execution, so that the comfort, smoothness and safety of human-machine shared driving are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent vehicle control technology, specifically to an intelligent electric vehicle lateral control system and method based on human-machine sharing. Background Technology

[0002] With the rapid development of intelligent electric vehicle technology, advanced driver assistance systems (ADAS) and autonomous driving systems are increasingly being used in vehicle lateral control. However, limited by current technology and the complex and ever-changing traffic environment, fully autonomous driving is unlikely to be achieved in the short term. Human-machine co-driving mode will become the mainstream technology path for a considerable period of time. In human-machine co-driving mode, the driver and the autonomous driving system jointly participate in the lateral motion control of the vehicle. How to dynamically and smoothly allocate the control weights between the two based on the driver's real-time operation status and intentions, while ensuring the driver's driving sovereignty and sense of participation, and ensuring the safety of vehicle driving, is a core problem that urgently needs to be solved in the field of human-machine co-driving technology. Current human-machine co-driving control methods often only switch responsibilities based on the presence or magnitude of the driver's torque, failing to fully consider the refined characteristics of the driver's operation. This can easily lead to abruptness in the human-machine control switching process, and even cause human-machine conflict, affecting driving comfort and safety.

[0003] In the prior art, a method for allocating lateral driving rights in human-machine co-driving that considers driver skills is disclosed in CN108819951A. This method establishes a driver driving skill evaluation model and comprehensively considers indicators such as driving time, number of times the vehicle leaves the lane line, number of times it brakes suddenly, and lateral standard deviation to evaluate the driver's driving skills offline. Based on the driver's skill evaluation value, the vehicle's distance from the lane departure point, and the deviation between the driver's expected turning angle and the lane keeping controller's expected turning angle, the driving right weight coefficient is calculated online.

[0004] While the above method incorporates driver skill as a basis for weight allocation, its driver skill evaluation relies on offline test data in specific scenarios, making it difficult to reflect the dynamic changes in the driver's operational state during real-time driving. Furthermore, this method only focuses on the deviation between the driver's desired steering angle and the system's desired steering angle, without analyzing or processing the smoothness and vibration of the driver's operation. When the driver's operation becomes abrupt or shaky due to tension or fatigue, the system still allocates weights according to the original deviation and responds directly, easily transmitting the driver's unstable operation to the vehicle's actuator, causing lateral swaying and affecting driving comfort and the stability of path tracking.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a lateral control system and method for intelligent electric vehicles based on human-machine sharing, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a lateral control method for intelligent electric vehicles based on human-machine sharing, specifically including:

[0008] Step 1: Obtain the driver's steering wheel control parameters, vehicle motion state parameters, and the desired front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel control parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate.

[0009] Step 2: Analyze the driver's lateral intervention intention based on the steering wheel torque to obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level.

[0010] Step 3: Analyze the temporal fluctuation characteristics of the driver's steering wheel operation parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's operation; based on the desired front wheel angle of the autonomous driving system and the vehicle motion state parameters, calculate the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment.

[0011] Step 4: Based on the handling quality coefficient, the driver's steering wheel torque is smoothed and filtered to generate a smoothed driver-demanded steering angle. Combined with the driver's initial weight allocation coefficient and the first target front wheel steering angle, the second target front wheel steering angle under the shared control of the driver and the autonomous driving system at the current sampling time is analyzed and obtained. The vehicle is then laterally controlled based on the second target front wheel steering angle.

[0012] Furthermore, the steering wheel torque refers to the torque value applied by the driver to the steering wheel, which is obtained through a torque angle sensor installed on the steering column, and the current sampling time is recorded. The steering wheel torque is denoted as The unit is The rule is that turning left is positive and turning right is negative;

[0013] The differential of the steering wheel torque is obtained by performing a first-order numerical differential calculation on the steering wheel torque, with units of . The calculation formula is as follows:

[0014]

[0015] In the formula, The current sampling time The differential torque of the steering wheel; The steering wheel torque at the previous moment; Indicates the sampling time interval; This is the index of the sampling time. ;

[0016] The vehicle's current sampling time is obtained through wheel speed sensors. The longitudinal speed of the vehicle is denoted as The unit is ;

[0017] The vehicle's position at the current sampling time is obtained through the inertial measurement unit. The angular velocity of rotation about the vertical axis, i.e., the yaw rate, is denoted as ω0. The unit is The rule is that counterclockwise is positive;

[0018] Obtain the current sampling time through the autonomous driving system. The expected front wheel steering angle is denoted as . The unit is The desired front wheel steering angle refers to the angle the front wheels should turn to when fully controlled by the autonomous driving system.

[0019] Furthermore, based on the current sampling time Steering wheel torque Combined with preset intervention trigger thresholds and strong intervention threshold The specific logic used to determine the level of a driver's lateral intervention intent is as follows:

[0020] like Determine the current sampling time The driver's lateral intervention intent level is 0, meaning there is no intervention intent;

[0021] like Determine the current sampling time The driver's lateral intervention intent level is Level 1, which is weak intervention intent;

[0022] like Determine the current sampling time The driver's lateral intervention intent level is level 2, which is a strong intervention intent;

[0023] Among them, the intervention trigger threshold and strong intervention threshold satisfy .

[0024] Furthermore, the initial weight allocation coefficient of the driver at the current sampling time is determined based on the lateral intervention intention level. The formula it is based on is as follows:

[0025]

[0026] In the formula, , , These represent the levels of lateral intervention intent as 0, 1, and 2, respectively.

[0027] Furthermore, based on the current sampling time Build length is Extract the steering wheel torque sequence within a sliding time window. and steering wheel torque differential sequence ;when At that time, historical data was insufficient. There are 1 sampling point, at which point only the existing ones are utilized. Historical data are used in the calculation, and the sampling time is reached. Then, the full sliding time window is used for calculation;

[0028] The fluctuation characteristic index of steering wheel torque is calculated, which includes torque range and torque root mean square; wherein, torque range refers to the difference between the maximum value and the minimum value of steering wheel torque within the sliding window; torque root mean square refers to the square root of the sum of the squares of the differential values ​​of steering wheel torque at each time point within the sliding window, divided by the window length.

[0029] Based on the torque range and the root mean square of the torque derivative, the driver's handling quality coefficient is determined according to the following formula:

[0030]

[0031] In the formula, The current sampling time The following is the control quality coefficient; Indicates the current sampling time The torque difference is extremely large; Indicates the current sampling time The root mean square of the differential torque; , The preset weighting coefficients satisfy... .

[0032] Furthermore, based on the current sampling time Given the vehicle's longitudinal speed and desired front wheel steering angle, the desired yaw rate of the vehicle under ideal conditions is calculated using the following formula:

[0033]

[0034] In the formula, Indicates the current sampling time The expected yaw rate; This refers to the vehicle's wheelbase. The preset stability coefficient;

[0035] Subtract the expected yaw rate at the current sampling moment from the actual yaw rate collected to obtain the value at that moment. Yaw angular velocity deviation value ;

[0036] A proportional-integral-derivative control law is used, based on a backward calculation from the current sampling time. Calculate the yaw rate deviation at each sampling time, and calculate the current sampling time. Front wheel steering angle correction :

[0037]

[0038] In the formula, , , These are the proportional, integral, and derivative control gains, respectively, all of which are preset normal values; To backtrack from the current sampling time The cumulative sum of deviations within each sampling time point, when hour, ;when hour, , ; This is an index for the sampling time.

[0039] Adding the desired front wheel steering angle to the front wheel steering angle correction, we obtain the first target front wheel steering angle required for full control by the autonomous driving system at the current sampling moment. .

[0040] Furthermore, based on the current sampling time The following is the handling quality coefficient Determine the smoothing filter factor ,in ;

[0041] A first-order low-pass filter is used to smooth the driver's steering wheel torque, generating a smoothed driver torque. :

[0042]

[0043] in, ;

[0044] Convert the smoothed driver torque into a smoothed driver-demand steering angle. :

[0045]

[0046] In the formula, This refers to the gear ratio gain of the steering system, in units of These are fixed parameters for the vehicle, obtained from the steering system calibration data.

[0047] Furthermore, based on the current sampling time The smoothed-down driver's required turning angle Driver initial weight allocation coefficient and the first target front wheel steering angle The weighted fusion method is used to calculate the front wheel steering angle of the second target under human-machine shared control. The formula used is as follows:

[0048]

[0049] In the formula, The current sampling time The second target is the front wheel steering angle, in units of ;

[0050] The second target front wheel angle The maximum permissible steering angle defined by the preset vehicle lateral stability envelope constraint. and minimum allowable turning angle Compare the results and adjust the limiting according to the following formula:

[0051]

[0052] This is the final target front wheel steering angle after correction;

[0053] Lateral control of the vehicle is performed based on the revised and finalized second target front wheel steering angle.

[0054] The present invention also provides a human-machine sharing-based intelligent electric vehicle lateral control system, which is used to execute the above-described human-machine sharing-based intelligent electric vehicle lateral control method, including:

[0055] The data acquisition module is used to acquire the driver's steering wheel operation parameters, vehicle motion state parameters, and the expected front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel operation parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate.

[0056] The driver lateral intervention intention analysis module is used to analyze the driver's lateral intervention intention based on the steering wheel torque, obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level.

[0057] The driver handling quality evaluation module is used to analyze the time-domain fluctuation characteristics of the driver's steering wheel handling parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's handling; based on the expected front wheel angle of the autonomous driving system and the vehicle motion state parameters, it calculates the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment.

[0058] The human-machine shared lateral control output module is used to smooth and filter the driver's steering wheel torque based on the operation quality coefficient, generate a smoothed driver demand angle, and analyze the second target front wheel angle under the shared control of the driver and the autonomous driving system at the current sampling time by combining the driver's initial weight allocation coefficient and the first target front wheel angle. The vehicle is then laterally controlled based on the second target front wheel angle.

[0059] Compared with the prior art, the beneficial effects of the present invention are:

[0060] This invention constructs a driver operation quality coefficient to quantify the temporal fluctuation characteristics of the driver's steering wheel operation. It can evaluate the smoothness and stability of the driver's operation in real time, and adaptively adjust the filtering intensity of the driver's input signal based on the quality coefficient. When the driver's operation is smooth, the original intention is preserved, and when the driver's operation is rapid and shaky, it is softened. This effectively suppresses the transmission of unstable operation caused by factors such as driver tension and fatigue to the vehicle's execution end, and avoids lateral swaying of the vehicle.

[0061] Meanwhile, the present invention adopts a hierarchical control architecture. First, it identifies the driver's lateral intervention intention based on the steering wheel torque and determines the initial weight allocation coefficient. Then, it smooths and filters the driver's input through the manipulation quality coefficient. Finally, it weights and fuses the smoothed driver's required steering angle with the first target front wheel steering angle of the autonomous driving system. This achieves the organic integration of the driver's intention and the system's intention, which respects the driver's driving sovereignty and ensures the smoothness of human-machine collaborative control.

[0062] Furthermore, this invention introduces a vehicle lateral stability envelope constraint for safety verification before the final output turning angle, ensuring that the vehicle's motion is limited to a stable region under any operating condition, thus fundamentally guaranteeing driving safety; and significantly improving the comfort, smoothness, and safety of human-machine shared driving. Attached Figure Description

[0063] Figure 1This is a schematic diagram of the overall method flow of the present invention;

[0064] Figure 2 A dual Y-axis image of torque range, torque root mean square derivative, and handling quality coefficient;

[0065] Figure 3 This is a schematic diagram of the overall system modules of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0067] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0068] Example:

[0069] Please see Figure 1-2 This invention provides a lateral control method for intelligent electric vehicles based on human-machine sharing, specifically including:

[0070] Step 1: Obtain the driver's steering wheel control parameters, vehicle motion state parameters, and the desired front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel control parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate.

[0071] In this embodiment, the steering wheel torque refers to the torque value applied by the driver to the steering wheel, which is obtained by a torque angle sensor installed on the steering column, and the current sampling time is recorded. The steering wheel torque is denoted as The unit is The rule is that turning left is positive and turning right is negative;

[0072] The differential of the steering wheel torque is obtained by performing a first-order numerical differential calculation on the steering wheel torque, with units of . The calculation formula is as follows:

[0073]

[0074] In the formula, The current sampling time The differential torque of the steering wheel; The steering wheel torque at the previous moment; Indicates the sampling time interval; This is the index of the sampling time. ;

[0075] The vehicle's current sampling time is obtained through wheel speed sensors. The longitudinal speed of the vehicle is denoted as The unit is ;

[0076] The vehicle's position at the current sampling time is obtained through the inertial measurement unit. The angular velocity of rotation about the vertical axis, i.e., the yaw rate, is denoted as ω0. The unit is The rule is that counterclockwise is positive;

[0077] Obtain the current sampling time through the autonomous driving system. The expected front wheel steering angle is denoted as . The unit is The desired front wheel steering angle refers to the angle that the front wheels should turn to when fully controlled by the autonomous driving system, that is, the target angle that the front wheels should reach relative to the longitudinal centerline of the vehicle.

[0078] Step 1 obtains a series of parameters, including the driver's steering wheel torque and its derivative, longitudinal vehicle speed, yaw rate, and desired front wheel angle, providing basic data support for subsequent driver intent recognition, handling quality assessment, and human-machine control allocation. Among them, the steering wheel torque and its derivative are used to determine the driver's intervention intention and operation smoothness, the longitudinal vehicle speed and yaw rate are used to describe the real-time motion state of the vehicle to calculate the control requirements of the autonomous driving system, and the desired front wheel angle represents the control target of the autonomous driving system for the lateral movement of the vehicle, providing the system-side data foundation for the final human-machine shared control angle fusion.

[0079] Step 2: Analyze the driver's lateral intervention intention based on the steering wheel torque to obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level.

[0080] In this embodiment, based on the current sampling time Steering wheel torque Combined with preset intervention trigger thresholds and strong intervention threshold The specific logic used to determine the level of a driver's lateral intervention intent is as follows:

[0081] like Determine the current sampling time The driver's lateral intervention intent level is 0, meaning there is no intervention intent;

[0082] like Determine the current sampling time The driver's lateral intervention intent level is Level 1, which is weak intervention intent;

[0083] like Determine the current sampling time The driver's lateral intervention intent level is level 2, which is a strong intervention intent;

[0084] Among them, the intervention trigger threshold and strong intervention threshold satisfy ; and The parameters were determined through real-vehicle calibration experiments based on steering system free play, sensor noise levels, and ergonomic data. Set to a minimum torque value slightly higher than the steering wheel free play and sensor noise. Set as the lower limit of torque when the driver has a clear intention to steer strongly.

[0085] The driver's initial weight allocation coefficient at the current sampling time is determined based on the level of lateral intervention intent. The formula it is based on is as follows:

[0086]

[0087] In the formula, , , These represent the levels of lateral intervention intent as 0, 1, and 2, respectively.

[0088] when When this occurs, it indicates that the controller is completely handed over to the autonomous driving system;

[0089] when At that time, the driver's initial weight allocation coefficient varies. The increase linearly rises from 0 to 1;

[0090] when When this is the case, it means that the controller is completely handed over to the driver.

[0091] The initial weight allocation coefficient of the driver at the current sampling time is determined based on the driver's lateral intervention intention level. , The range of values ​​is This is used to characterize the level of control weight the driver should have at the current moment; specifically, within the weak intervention intention range, the driver's initial weight allocation coefficient... Calculated based on the absolute value of the torque at the current sampling time, its value increases as the absolute value of the torque increases, and when the absolute value of the torque approaches the strong intervention threshold... hour, Approaching 1; this design allows the system to calculate the corresponding weighting coefficients in real time when the driver applies different absolute values ​​of torque at different times, realizing the dynamic distribution of control between the autonomous driving system and the driver; as the absolute value of the driver's torque gradually increases over time, the driver's initial weighting coefficient... This weighting gradually increases, thus achieving a smooth transition of control; it should be noted that the driver's initial weight allocation coefficient... It is only valid at the current sampling moment, and its value depends entirely on the absolute value of the torque at that moment. It does not require the driver's torque to change continuously in a linear manner. Even if the absolute value of the torque changes abruptly, the weighting coefficient will be updated in real time accordingly. This avoids the sudden change in control commands caused by fixed thresholds in traditional responsibility switching, allowing the driver to feel the system's response to his operating intentions in an instant. It effectively suppresses the lateral vibration of the vehicle at the moment of human-machine control handover, improves driving comfort and the smoothness of path tracking, and fully reflects the collaborative rather than adversarial control concept in human-machine co-driving mode.

[0092] Step 3: Analyze the temporal fluctuation characteristics of the driver's steering wheel operation parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's operation; based on the desired front wheel angle of the autonomous driving system and the vehicle motion state parameters, calculate the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment.

[0093] In this embodiment, based on the current sampling time Build length is The sliding time window, when At that time, extract the steering wheel torque sequence within that time window. and steering wheel torque differential sequence ;when At that time, historical data was insufficient. There are 1 sampling point, at which point only the existing ones are utilized. Historical data are used in the calculation, and the sampling time is reached. Then, the full sliding time window is used for calculation;

[0094] The fluctuation characteristic index of steering wheel torque is calculated, which includes torque range and torque root mean square; wherein, torque range refers to the difference between the maximum value and the minimum value of steering wheel torque within the sliding window; torque root mean square refers to the square root of the sum of the squares of the differential values ​​of steering wheel torque at each time point within the sliding window, divided by the window length.

[0095] The formula for calculating the torque range is as follows:

[0096]

[0097] In the formula, Indicates the current sampling time The torque difference is extremely large; Sampling time The steering wheel torque is reduced; For the index of the sampling time, when hour, ;when hour, , ;

[0098] The formula for calculating the root mean square of the differential torque is as follows:

[0099]

[0100] In the formula, Indicates the current sampling time The root mean square of the differential torque; Sampling time The differential torque of the steering wheel; The actual number of sampling points used in the calculation, when hour, ;when hour, , , ;

[0101] By constructing a sliding time window and calculating two fluctuation characteristic indicators, namely torque range and torque root mean square, the smoothness of the driver's operation over a recent period can be effectively quantified. Torque range reflects the range of fluctuation in the amplitude of the torque applied by the driver; the larger the torque range, the more drastic the change in the amplitude of the driver's operation. Torque root mean square represents the swiftness of the change in the driver's operation; the larger its value, the more frequent the driver's operation jitters. Combining the two as inputs to the handling quality coefficient allows for the evaluation of the smoothness of the driver's operation from both amplitude fluctuation and frequency fluctuation aspects.

[0102] Based on the torque range and the root mean square of the torque derivative, the driver's handling quality coefficient is determined according to the following formula:

[0103]

[0104] In the formula, The current sampling time The following is the control quality coefficient; , The preset weighting coefficients satisfy... During driving operations, the root mean square of the torque derivative reflects the urgency and frequency of the driver's operation. This high-frequency vibration component has a more direct and significant impact on the smoothness of human-machine collaborative control. Minor vibrations caused by driver tension or fatigue often manifest first as an abnormal increase in the rate of torque change, rather than a drastic change in the absolute value of torque. Therefore, setting... Value greater than This means that the system is more sensitive to the urgency of the operation and can capture abnormal vibrations in the driver's operation more promptly.

[0105] For this formula, the dependent variable manipulates the quality coefficient. Used to characterize the smoothness and stability of the driver's operation at the current moment; The larger the value, the smoother and more stable the driver's operation, the smaller the torque fluctuation and the lower the rate of change. In this case, the system should retain the driver's original operating intention with high confidence. The smaller the value, the more abrupt and shaky the driver's operation, and the more drastic and rapid the torque fluctuations. In this case, the system needs to enhance the filtering strength and soften the driver's input to suppress the impact of unstable operation on the vehicle's lateral movement.

[0106] Torque range reflects the fluctuation range of the driver's operating amplitude within the sliding time window. The larger the range, the more drastic the changes in the driver's operating amplitude, such as frequent large back-and-forth corrections under tension. Such large fluctuations will reduce the handling quality. Torque derivative root mean square reflects the urgency and frequency of the driver's operation. The larger the root mean square value, the faster the driver's operation changes and the more frequent the vibration, such as hand tremors due to fatigue or rapid corrections under tension. Both torque range and frequency fluctuation jointly characterize the smoothness of the driver's operation. An increase in either indicator will lead to a decrease in the handling quality coefficient.

[0107] This formula structure guarantees the handling quality coefficient. Always in Within this range, when the driver's operation is relatively smooth, and both the torque range and the root mean square of the torque derivative are close to 0... A value approaching 1 indicates optimal handling quality; as the torque range and the root mean square of the torque derivative increase, the denominator increases. The corresponding decrease, and the two volatility characteristic indicators are related to The contribution is determined by the weighting coefficient. , The adjustments reflect the differences in the degree to which different fluctuation characteristics affect the quality of operation.

[0108] Table 1: Statistical Table of Operational Quality Coefficients

[0109]

[0110] Based on the 15 sets of data shown in Table 1, combined with Figure 2 It can be seen that as the torque range and the root mean square of the torque derivative increase, the handling quality coefficient shows a gradual decreasing trend. When both the torque range and the root mean square of the torque derivative are at relatively small values, such as when the sampling time index is 1, the torque range is 0.5 and the root mean square of the torque derivative is 0.2. At this time, the handling quality coefficient is 0.9091, and the driver's operation is stable. When the torque range and the root mean square of the torque derivative gradually increase to a medium level, the handling quality coefficient drops to the range of 0.5-0.7, indicating that the driver's operation begins to fluctuate to a certain extent. When the torque range and the root mean square of the torque derivative further increase to a higher level, the handling quality coefficient drops below 0.3, at which point the driver's operation is abrupt and the shaking is obvious.

[0111] The above data variation patterns verify that the handling quality coefficient constructed in this invention can effectively quantify the smoothness of driver operation, providing a reliable basis for subsequent adaptive smoothing filtering. When the handling quality coefficient is high, the system retains the driver's original intention; when the handling quality coefficient is low, the system enhances the filtering strength and softens the driver's input, thereby realizing differentiated responses to driving operations of different quality and effectively suppressing the transmission of unstable operations caused by driver tension or fatigue to the vehicle execution end.

[0112] Based on the current sampling time Given the vehicle's longitudinal speed and desired front wheel steering angle, the desired yaw rate of the vehicle under ideal conditions is calculated using the following formula:

[0113]

[0114] In the formula, Indicates the current sampling time The expected yaw rate; This refers to the vehicle's wheelbase. The preset stability coefficient;

[0115] This formula is derived from the steady-state steering characteristics in a linear two-degree-of-freedom vehicle dynamics model. It describes the ideal yaw rate that a vehicle should produce given a longitudinal vehicle speed and front wheel steering angle input. The numerator of the formula... This demonstrates the fundamental contributions of vehicle speed and front wheel steering angle to yaw rate: at low speeds, yaw rate is approximately proportional to the product of vehicle speed and front wheel steering angle; the denominator... This introduces a stability coefficient to correct the steering characteristics, where It is determined by the front and rear wheel lateral stiffness, wheelbase, and mass distribution of the vehicle; when At high speeds, the vehicle exhibits understeer characteristics. As speed increases, the yaw rate gain from the same front wheel steering angle decreases. This is because at high speeds, the tire slip angle increases, requiring a larger front wheel steering angle to produce the same yaw response. This characteristic makes the vehicle more stable and controllable at high speeds. When the steering is neutral, the yaw rate is linearly related to the vehicle speed; when During oversteering, the yaw rate gain increases abnormally at high speeds, easily leading to vehicle instability. This formula allows the autonomous driving system to calculate the desired front wheel steering angle. This is converted into a desired yaw rate that conforms to the inherent steering characteristics of the vehicle, providing a theoretical reference benchmark that conforms to the laws of vehicle dynamics for subsequent calculation and correction of yaw rate deviation.

[0116] based on and Calculate the current sampling time The actual yaw rate deviation value collected below :

[0117]

[0118] A proportional-integral-derivative control law is used, based on a backward calculation from the current sampling time. Calculate the yaw rate deviation at each sampling time, and calculate the current sampling time. Front wheel steering angle correction :

[0119]

[0120] In the formula, , , These are the proportional, integral, and derivative control gains, respectively, all of which are preset normal values; To backtrack from the current sampling time The cumulative sum of deviations within each sampling time point, when hour, ;when hour, , ; This is an index for the sampling time.

[0121] Adding the desired front wheel steering angle to the front wheel steering angle correction, we obtain the first target front wheel steering angle required for full control by the autonomous driving system at the current sampling moment. .

[0122] The proportional term in the formula The system provides an immediate response to the current yaw rate deviation; the larger the deviation, the larger the correction, ensuring the system can quickly eliminate the current error. (Integral term) Regarding the past The deviations within each sampling time point are accumulated to eliminate steady-state errors caused by factors such as road disturbances, crosswinds, or tire nonlinearity, ensuring that the vehicle can accurately track the desired path during long-term driving; differential term The error development trend is predicted by calculating the rate of change of the deviation, increasing system damping, suppressing overshoot and oscillation, and improving control stability; the front wheel steering angle correction is obtained by superimposing these three factors. Adding this to the desired front wheel steering angle planned by the autonomous driving system constitutes a feedforward + feedback composite control structure. The feedforward term provides the main control benchmark based on path planning, while the feedback term corrects deviations caused by various uncertainties. This structure ensures both rapid response in path tracking and enhances the system's robustness against external disturbances, ultimately leading to the obtained first target front wheel steering angle. It can more accurately bring the vehicle's actual yaw rate closer to the desired value.

[0123] Step 4: Based on the handling quality coefficient, the driver's steering wheel torque is smoothed and filtered to generate a smoothed driver's required steering angle. Combined with the driver's initial weight allocation coefficient and the first target front wheel steering angle, the second target front wheel steering angle under the shared control of the driver and the autonomous driving system at the current sampling time is analyzed and obtained. The vehicle is then laterally controlled based on the second target front wheel steering angle.

[0124] In this embodiment, based on the current sampling time The following is the handling quality coefficient Determine the smoothing filter factor ,in ;

[0125] A first-order low-pass filter is used to smooth the driver's steering wheel torque, generating a smoothed driver torque. :

[0126]

[0127] in, ;

[0128] Smoothed driver torque This value characterizes the driver torque actually used for subsequent steering angle conversion at the current sampling moment after filtering. It is the result of smoothing and softening the original driver input torque. The magnitude of this value directly reflects the driver's operating intensity retained after filtering: when A larger value indicates that the torque applied by the driver remains at a large amplitude after smoothing, corresponding to a strong steering intention from the driver; when When the value is small, it indicates that the absolute value of the torque applied by the driver is small or has been significantly softened after filtering, corresponding to a weaker steering intention or effective suppression of driving vibration. Through this filtering process, high-frequency vibration components can be adaptively filtered out according to the driving quality while preserving the original intention direction of the driver, making the torque signal used for the final angle conversion smoother and more stable.

[0129] Manipulation quality coefficient The range of values ​​is The larger the value, the smoother the driver's operation; the smaller the value, the more rapid and shaky the driver's operation. The mapping, when the driver operates smoothly, When it approaches 1, Approaching 0, the output of the first-order low-pass filter depends almost entirely on the original torque input at the current moment, and the system retains the driver's original operating intention with high confidence; when the driver's operation is significantly shaky, When it decreases, As the output of the filter increases accordingly, it becomes more dependent on the smoothing value of the previous moment, thus effectively softening the current input. This design allows the filter strength to be adaptively adjusted according to the driver's real-time operation quality. The worse the quality, the stronger the filter, thereby effectively suppressing the transmission of jittery operation caused by tension or fatigue to the vehicle's actuators without changing the driver's control.

[0130] Convert the smoothed driver torque into a smoothed driver-demand steering angle. :

[0131]

[0132] In the formula, This refers to the gear ratio gain of the steering system, in units of These are fixed parameters for the vehicle, obtained from the steering system calibration data; the vehicle's steer-by-wire system has a defined torque-angle mapping relationship. It reflects the mapping relationship between the mechanical transmission ratio, torque and steering wheel angle of the steering system. It is an inherent parameter obtained through calibration when the vehicle leaves the factory. After the torque applied by the driver is smoothed and filtered, it is multiplied by this gain to obtain the target angle that the front wheels should reach to achieve the torque input under the current steering system characteristics.

[0133] Based on the current sampling time The smoothed-down driver's required turning angle Driver initial weight allocation coefficient and the first target front wheel steering angle The weighted fusion method is used to calculate the front wheel steering angle of the second target under human-machine shared control. The formula used is as follows:

[0134]

[0135] In the formula, The current sampling time The second target is the front wheel steering angle, in units of ;

[0136] The second target front wheel angle The maximum permissible steering angle defined by the preset vehicle lateral stability envelope constraint. and minimum allowable turning angle Compare the results and adjust the limiting according to the following formula:

[0137]

[0138] This is the final target front wheel steering angle after correction;

[0139] Lateral control of the vehicle is performed based on the revised and finalized second target front wheel steering angle.

[0140] Here, the driver's initial weight allocation coefficient is first used as the dynamic weight. The driver's desired steering angle, after smoothing and filtering, is weighted and fused with the first target front wheel steering angle required by the autonomous driving system. When the driver has no intention to intervene, the system is completely controlled by the autonomous driving system. When the driver strongly intervenes, the system is completely driven by the driver. When the driver weakly intervenes, the driver's desired steering angle output to the driver and the first target front wheel steering angle output by the autonomous driving system are used to calculate the shared lateral control output through weighted summation. It should be noted that the real-time changes in the driver's input torque are uncertain. It may fluctuate over time rather than change continuously and smoothly; accordingly, the fused result... It may also exhibit fluctuating characteristics; however, due to the changing demands of drivers participating in the integration... Having undergone adaptive smoothing filtering to manipulate the quality coefficients, its high-frequency jitter components have been effectively suppressed, therefore even... Even in the event of a sudden change, the fusion result remains relatively stable, avoiding drastic fluctuations in control commands caused by sudden weight changes. This design respects the driver's autonomy while ensuring the smoothness of human-machine collaboration. Subsequently, the fused second target front wheel steering angle is compared with the maximum and minimum permissible steering angles defined by the preset vehicle lateral stability envelope constraints. By first taking the smaller value between the fused steering angle and the maximum permissible steering angle, and then taking the larger value between the result and the minimum permissible steering angle, a limiting correction is performed to ensure that the final output steering angle command is always constrained within the safe boundary for stable vehicle driving. This design ensures that any steering angle command exceeding the vehicle's physical limits is automatically truncated, fundamentally guaranteeing driving safety and reflecting the reasonable design concept of prioritizing intent and providing a safety safety net in human-machine co-driving control.

[0141] It should be noted that the above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0142] Please see Figure 3 A human-machine sharing-based intelligent electric vehicle lateral control system includes:

[0143] The data acquisition module is used to acquire the driver's steering wheel operation parameters, vehicle motion state parameters, and the expected front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel operation parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate.

[0144] The driver lateral intervention intention analysis module is used to analyze the driver's lateral intervention intention based on the steering wheel torque, obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level.

[0145] The driver handling quality evaluation module is used to analyze the time-domain fluctuation characteristics of the driver's steering wheel handling parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's handling; based on the expected front wheel angle of the autonomous driving system and the vehicle motion state parameters, it calculates the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment.

[0146] The human-machine shared lateral control output module is used to smooth and filter the driver's steering wheel torque based on the operation quality coefficient, generate a smoothed driver demand angle, and analyze the second target front wheel angle under the shared control of the driver and the autonomous driving system at the current sampling time by combining the driver's initial weight allocation coefficient and the first target front wheel angle. The vehicle is then laterally controlled based on the second target front wheel angle.

[0147] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0148] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0149] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A lateral control method for intelligent electric vehicles based on human-machine sharing, characterized in that, Specifically, it includes: Step 1: Obtain the driver's steering wheel control parameters, vehicle motion state parameters, and the desired front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel control parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate. Step 2: Analyze the driver's lateral intervention intention based on the steering wheel torque to obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level. Step 3: Analyze the temporal fluctuation characteristics of the driver's steering wheel operation parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's operation; based on the desired front wheel angle of the autonomous driving system and the vehicle motion state parameters, calculate the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment. Step 4: Based on the handling quality coefficient, the driver's steering wheel torque is smoothed and filtered to generate a smoothed driver-demanded steering angle. Combined with the driver's initial weight allocation coefficient and the first target front wheel steering angle, the second target front wheel steering angle under the shared control of the driver and the autonomous driving system at the current sampling time is analyzed and obtained. The vehicle is then laterally controlled based on the second target front wheel steering angle.

2. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 1, characterized in that: The steering wheel torque refers to the torque value applied by the driver to the steering wheel, which is obtained by a torque angle sensor installed on the steering column, and the current sampling time is recorded. The steering wheel torque is denoted as The unit is The rule is that turning left is positive and turning right is negative; The differential of the steering wheel torque is obtained by performing a first-order numerical differential calculation on the steering wheel torque, with units of . The calculation formula is as follows: In the formula, The current sampling time The differential torque of the steering wheel; The steering wheel torque at the previous moment; Indicates the sampling time interval; For the index of the sampling time, ; The vehicle's current sampling time is obtained through wheel speed sensors. The longitudinal speed of the vehicle is denoted as The unit is ; The vehicle's position at the current sampling time is obtained through the inertial measurement unit. The angular velocity of rotation about the vertical axis, i.e., the yaw rate, is denoted as ω0. The unit is The rule is that counterclockwise is positive; Obtain the current sampling time through the autonomous driving system. The expected front wheel steering angle is denoted as . The unit is The desired front wheel steering angle refers to the angle the front wheels should turn to when fully controlled by the autonomous driving system.

3. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 2, characterized in that: Based on the current sampling time Steering wheel torque Combined with preset intervention trigger thresholds and strong intervention threshold The specific logic used to determine the level of a driver's lateral intervention intent is as follows: like Determine the current sampling time The driver's lateral intervention intent level is 0, meaning there is no intervention intent; like Determine the current sampling time The driver's lateral intervention intent level is Level 1, which is weak intervention intent; like Determine the current sampling time The driver's lateral intervention intent level is level 2, which is a strong intervention intent; Among them, the intervention trigger threshold and strong intervention threshold satisfy .

4. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 3, characterized in that: The driver's initial weight allocation coefficient at the current sampling time is determined based on the level of lateral intervention intent. The formula it is based on is as follows: In the formula, , , These represent the levels of lateral intervention intent as 0, 1, and 2, respectively.

5. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 2, characterized in that: Based on the current sampling time Build length is Extract the steering wheel torque sequence within a sliding time window. and steering wheel torque differential sequence ;when At that time, historical data was insufficient. There are 1 sampling point, at which point only the existing ones are utilized. Historical data are used in the calculation, and the sampling time is reached. Then, the full sliding time window is used for calculation; The fluctuation characteristic index of steering wheel torque is calculated, which includes torque range and torque root mean square; wherein, torque range refers to the difference between the maximum value and the minimum value of steering wheel torque within the sliding window; torque root mean square refers to the square root of the sum of the squares of the differential values ​​of steering wheel torque at each time point within the sliding window, divided by the window length. Based on the torque range and the root mean square of the torque derivative, the driver's handling quality coefficient is determined according to the following formula: In the formula, The current sampling time The following is the control quality coefficient; Indicates the current sampling time The torque difference is extremely large; Indicates the current sampling time The root mean square of the differential torque; , The preset weighting coefficients satisfy... .

6. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 2, characterized in that: Based on the current sampling time Given the vehicle's longitudinal speed and desired front wheel steering angle, the desired yaw rate of the vehicle under ideal conditions is calculated using the following formula: In the formula, Indicates the current sampling time The expected yaw rate; This refers to the vehicle's wheelbase. The preset stability coefficient; Subtract the expected yaw rate at the current sampling moment from the actual yaw rate collected to obtain the current sampling moment. Yaw angular velocity deviation value .

7. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 6, characterized in that: A proportional-integral-derivative control law is used, based on a backward calculation from the current sampling time. Calculate the yaw rate deviation at each sampling time, and calculate the current sampling time. Front wheel steering angle correction : In the formula, , , These are the proportional, integral, and derivative control gains, respectively, all of which are preset normal values; To backtrack from the current sampling time The cumulative sum of deviations within each sampling time point, when hour, ;when hour, , ; This is an index for the sampling time. Adding the desired front wheel steering angle to the front wheel steering angle correction, we obtain the first target front wheel steering angle required for full control by the autonomous driving system at the current sampling moment. .

8. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 5, characterized in that: Based on the current sampling time The following is the handling quality coefficient Determine the smoothing filter factor ,in ; A first-order low-pass filter is used to smooth the driver's steering wheel torque, generating a smoothed driver torque. : in, ; Convert the smoothed driver torque into a smoothed driver-demand steering angle. : In the formula, This refers to the gear ratio gain of the steering system, in units of These are fixed parameters for the vehicle, obtained from the steering system calibration data.

9. The lateral control method for intelligent electric vehicles based on human-machine sharing according to claim 8, characterized in that: Based on the current sampling time The smoothed-down driver's required turning angle Driver initial weight allocation coefficient and the first target front wheel steering angle The weighted fusion method is used to calculate the front wheel steering angle of the second target under human-machine shared control. The formula used is as follows: In the formula, The current sampling time The second target is the front wheel steering angle, in units of ; The second target front wheel angle The maximum permissible steering angle defined by the preset vehicle lateral stability envelope constraint. and minimum allowable turning angle Compare the results and adjust the limiting according to the following formula: This is the final target front wheel steering angle after correction; Lateral control of the vehicle is performed based on the revised and finalized second target front wheel steering angle.

10. A lateral control system for intelligent electric vehicles based on human-machine sharing, characterized in that: The human-machine sharing-based intelligent electric vehicle lateral control system is used to execute the human-machine sharing-based intelligent electric vehicle lateral control method according to any one of claims 1-9, comprising: The data acquisition module is used to acquire the driver's steering wheel operation parameters, vehicle motion state parameters, and the expected front wheel steering angle of the autonomous driving system at the current sampling time; the driver's steering wheel operation parameters include steering wheel torque and steering wheel torque derivative, and the vehicle motion state parameters include longitudinal vehicle speed and yaw rate. The driver lateral intervention intention analysis module is used to analyze the driver's lateral intervention intention based on the steering wheel torque, obtain the lateral intervention intention level, and determine the driver's initial weight allocation coefficient at the current sampling time based on the lateral intervention intention level. The driver handling quality evaluation module is used to analyze the time-domain fluctuation characteristics of the driver's steering wheel handling parameters and calculate the handling quality coefficient to characterize the smoothness of the driver's handling; based on the expected front wheel angle of the autonomous driving system and the vehicle motion state parameters, it calculates the first target front wheel angle required for full control by the autonomous driving system at the current sampling moment. The human-machine shared lateral control output module is used to smooth and filter the driver's steering wheel torque based on the operation quality coefficient, generate a smoothed driver demand angle, and analyze the second target front wheel angle under the shared control of the driver and the autonomous driving system at the current sampling time by combining the driver's initial weight allocation coefficient and the first target front wheel angle. The vehicle is then laterally controlled based on the second target front wheel angle.

Images