Steering wheel hand-off state and driving intention recognition method and control system based on three-dimensional force perception

Abstract

The invention relates to a three-dimensional force perception-based steering wheel off-hand state and driving intention identification method, and belongs to the technical field of intelligent driving. The method comprises the following steps: acquiring a three-dimensional force signal acting on the rim of the steering wheel in real time; processing the three-dimensional force signal, and judging and generating driver state information based on the amplitude, duration and dynamic change characteristics of each component of the three-dimensional force signal; according to the driver state information, intelligent driving control instructions are output, and the control instructions comprise authority management, early warning control, control right handover and system coordination instructions; and the intelligent driving control instruction is executed, and safe driving and permission handover of the vehicle are achieved through cooperative control of a steering system, a braking system, an alarm system and a man-machine interface system. According to the method, the binary judgment limitation of traditional hand-leaving detection is broken through, accurate perception from whether the hand leaves or not to how to shake hands is achieved, and a safer, smoother and more intelligent man-machine interaction and authority management basis is provided for high-order automatic driving.

Description

Technical Field

[0001] This invention belongs to the field of automotive electronics and control technology, and relates to a method and control system for recognizing the steering wheel off-hand state and driving intention based on three-dimensional force perception. Background Technology

[0002] With the continuous improvement of automotive intelligence, Advanced Driver Assistance Systems (ADAS) and autonomous driving technologies have been widely applied. In these systems, accurately determining whether the driver is effectively controlling the steering wheel and recognizing driving intentions have become key technologies for ensuring driving safety. Hands-off detection (HOD), as one of the core technologies of human-machine co-driving, directly affects the safety and reliability of intelligent driving systems.

[0003] Currently, the main steering wheel hands-off detection technologies include capacitive sensing, pressure sensing, and torque sensing solutions. Capacitive sensing detects hands by placing capacitive sensors on the steering wheel surface and using the capacitance changes caused by hand proximity. However, this solution only detects the contact state between the hand and the steering wheel, failing to distinguish between active control and accidental contact. It is also susceptible to environmental temperature and humidity fluctuations, exhibiting poor stability. More importantly, it cannot acquire crucial information reflecting driving intentions, such as grip strength and operational force. Secondly, pressure sensor-based detection solutions use thin-film or dot-matrix pressure sensors to measure the pressure distribution of the hand on the steering wheel. While this provides some force information, traditional pressure sensors typically only detect normal forces perpendicular to the steering wheel surface, failing to sense tangential friction and torsional torque. This makes it difficult for the system to accurately identify important driving behavior characteristics such as steering intentions and emergency maneuvers. Furthermore, torque sensor-based solutions determine the operating state by monitoring the steering torque applied by the driver. However, this method cannot distinguish between active and passive driver input, and its sensitivity is insufficient under low-torque conditions such as straight-line driving.

[0004] To improve detection performance, multi-dimensional force sensing technology has been applied in steering wheel hands-off detection systems in recent years. However, with the development of intelligent driving technology and the increasing demands for functionality, existing research shows that simple hands-off detection is no longer sufficient for high-level intelligent driving systems. These systems need a deeper understanding of the driver's intentions, including the strength of steering intent, the urgency of the operation, fatigue levels, and other multi-dimensional information. While current technologies can acquire three-dimensional force information, they still have limitations in feature extraction and intention understanding. They cannot fully extract key information reflecting driving behavior characteristics from the mechanical signals, making it difficult to build accurate driving intention recognition models. This technological bottleneck severely restricts the improvement of human-machine interaction and safety in intelligent driving systems.

[0005] Therefore, to address the problems of misjudgment in hands-off detection, difficulty in recognizing driving intentions, and the impact of sensor integration on steering wheel operation feel in existing technologies, there is an urgent need to develop a three-dimensional force perception system that can accurately determine hands-off status and continuously recognize driving intentions without altering the original structure and operating characteristics of the steering wheel. This system requires collaborative innovation across multiple levels, including sensor structure design, multi-dimensional signal processing, and hierarchical recognition algorithms, to achieve comprehensive and accurate perception of the driver's state, representing a key technological breakthrough for improving the safety and naturalness of intelligent driving interaction. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a method and control system for recognizing the steering wheel off-hand state and driving intention based on three-dimensional force perception. This method achieves refined and hierarchical recognition of the driver's state by decoupling and analyzing the three-dimensional mechanical signals on the steering wheel, and provides a decision basis for intelligent driving systems.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for recognizing steering wheel hands-off state and driving intention based on three-dimensional force perception, the method specifically includes the following steps: S1. Real-time acquisition of three-dimensional force signals acting on the steering wheel rim, wherein the three-dimensional force signals include at least radial force, tangential force and axial force; S2. Process the three-dimensional force signal and, based on the amplitude, duration and dynamic change characteristics of each component, determine and generate driver state information, which includes discrete hands-off state signals and continuous control intention signals. S3. Based on the driver status information, output intelligent driving control commands, including access control, early warning control, control handover, and system coordination commands; S4. Execute intelligent driving control commands to achieve safe driving and authority transfer through the coordinated control of the steering system, braking system, alarm system and human-machine interface system.

[0008] Furthermore, step S1 specifically includes: the sensing units are distributed in a multi-point manner inside the steering wheel rim, with sensing units set at the 3 o'clock, 6 o'clock, 9 o'clock, and 12 o'clock positions to form a complete force sensing loop, thereby acquiring the three-dimensional force signal acting on the steering wheel rim in real time. The three-dimensional force signal includes at least radial force. Tangential force and axial force .

[0009] Furthermore, in step S2, driver status information is generated, which includes two levels: Discrete state layer: Based on the amplitude and duration of the force signal, determine the "confirmed release from hand" or "hand in loop" state; Continuous Intent Layer: In the "hand on ring" state, further analyze the dynamic change patterns of each component force to generate continuous signals such as "steering intent intensity" and "grip strength level".

[0010] Furthermore, step S2 specifically includes: S21. Signal preprocessing: Low-pass filtering (cutoff frequency 10Hz) is performed on the original three-dimensional force signal to remove high-frequency noise, and coordinate normalization is performed to eliminate signal differences caused by different hand positions. S22. Calculate the resultant force amplitude; calculate the resultant force amplitude of the instantaneous three-dimensional forces. ; S23. Discrete state judgment and processing: 1) "Confirm release" judgment, judgment Is it consistently below the threshold? F L (e.g., 0.5N) and the time reaches T Hold (e.g., 3 seconds); if so, it is determined to be in the "confirmed release" state, and the release timer is started or incremented; 2) Parallel "hand in loop" judgment, judging any force component ( , , Does the instantaneous value of ) exceed the threshold? F H (e.g., 2N); if so, immediately determine the state as "hand on ring" and reset the off-hand timer; S24. After determining that "hand is on the ring", perform continuous intent recognition: 1) Steering intent recognition and continuous monitoring of tangential force Calculate its rate of change and steady-state value; when rate of change If the value exceeds a positive threshold (e.g., 5 N / s) and the steady-state value continues to increase, then a value similar to... The proportional "steering intention intensity" signal is normalized to 0-100%; simultaneously, the radial force is verified. (Compression) or axial force Whether a pull / push pattern emerges that corresponds to the steering direction can improve confidence; for example, a left turn intention typically corresponds to a positive turn. (Define direction), increase (Right hand squeeze) or negative (Pull with left hand). If the patterns are consistent, increase the "confidence" of the intent signal; if they are inconsistent, reduce the weight of the intent signal or mark it as potential noise. 2) Grip strength level identification, based on axial force. The filtered amplitude is mapped to a "grip strength level" signal, which can be used to assess the driver's level of tension or fatigue. S25. Summarize and output the driver status information, package the discrete status (hands off / on loop) and continuous intention signals (steering intention intensity, direction, confidence level; grip strength level) and hands-off timer value into a driver status information package, and output it to the decision module.

[0011] Furthermore, step S3 includes: When a "confirmed hands-off" status signal is received and the duration exceeds the safety threshold, a driver takeover request alarm is triggered. When a "hands-on-loop" status signal is received in autonomous driving mode, but the control intention signal is below the intervention threshold, the current autonomous driving level is maintained. When a control intention signal higher than the intervention threshold is received in autonomous driving mode, a smooth transfer of driving control from the autonomous driving system to the driver is initiated.

[0012] Furthermore, in step S4, the steering control system performs the following: inputting a steering intention signal from the decision-making layer, including intention intensity (0-100%), steering direction, and confidence level; the system calculates the target assist torque curve; when the intention intensity exceeds a 30% threshold, the system enters a handover preparation state and preloads control parameters; when the intention intensity exceeds 60%, the actual torque transfer begins; during the handover process, a micro-vibration motor on the steering wheel generates progressive vibration, with the vibration intensity increasing linearly with the handover progress, simulating the "clutch engagement" sensation and providing the driver with intuitive tactile feedback; when the assist torque drops to a preset minimum value and the driver's torque is detected to be stably maintained for more than 0.5 seconds, the system confirms that control has been completely transferred and updates the driving mode to manual driving.

[0013] The present invention also provides an intelligent driving control system, the system comprising: A three-dimensional force sensing module, integrated into the steering wheel rim, is used to detect three-dimensional force signals; The signal processing module is electrically connected to the three-dimensional force sensing module and is used to run the steps of the method described above and output driver status information. The vehicle control module is communicatively connected to the signal processing module and is used to receive the driver status information, make intelligent driving control decisions based on the preset control strategy, and generate corresponding control commands. The execution interface module is communicatively connected to the vehicle control decision module and is used to convert the control commands into specific execution signals to drive the vehicle actuators to achieve vehicle control and alarm prompts.

[0014] The present invention also provides a vehicle comprising the intelligent driving control system described above.

[0015] The beneficial effects of this invention are as follows: (1) Upgraded perception dimension: From single-dimensional torque or capacitance signals to three-dimensional force collaborative analysis, a richer information source is provided for detecting driver operation information. This can not only determine whether the hands are in the loop, but also distinguish different types of operation intentions, providing unprecedentedly refined information input for intelligent driving systems.

[0016] (2) More accurate state judgment: The system uses both the amplitude and duration of the force to determine "confirmed release of hands", which effectively avoids misjudgment caused by instantaneous interference such as road bumps and improves the reliability of release detection. At the same time, the system uses an instantaneous threshold to quickly determine "hands on the loop", which ensures the system's rapid response to the driver's return. This hierarchical judgment mechanism makes state recognition both highly accurate and robust.

[0017] (3) Quantification of continuous driving intention recognition: For the first time in the "hands on the loop" state, the quantitative recognition and output of continuous intention signals such as the intensity of the driver's steering intention and the degree of grip are realized. This enables the intelligent driving system to perceive the driver's slight operating intention in advance, and shifts from passive "reactive" control to forward-looking "predictive" assistance, providing the intelligent driving system with key predictive capabilities.

[0018] (4) Enhanced safety: By verifying the consistency between multi-dimensional force signals (such as the coordination mode of tangential force and radial / axial force during steering), the system can effectively filter noise signals caused by unintentional touches, handheld objects, or external impacts, greatly improving the confidence level of intent recognition. Decision-making based on high-confidence information makes the control commands of the intelligent driving system more reliable, fundamentally improving driving safety.

[0019] (5) More natural human-computer interaction experience: The system’s ability to understand the driver’s continuous intentions makes the handover of driving rights smooth and gradual based on the intensity of intentions, which is highly consistent with the human driver’s operational expectations, greatly reduces the sense of conflict and tension during the takeover process, and significantly improves the user’s trust and acceptance of the intelligent driving system.

[0020] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0021] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is the overall flowchart of the system; Figure 2 Flowchart for generating driver status information; Figure 3 Flowchart for turning intention recognition; Figure 4 This is a flowchart of the intelligent driving control decision-making process. Figure 5 This is a timing diagram for monitoring normal driving conditions. Figure 6 This is a sequence diagram for identifying steering intent and handover of control. Figure 7 This is the timing diagram for detecting the off-hand state; Figure 8 Sequence diagram for emergency response; Figure 9 This is a block diagram of the overall system structure. Detailed Implementation

[0022] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0023] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0024] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0025] This invention provides the following technical solution: In a first aspect, the present invention provides a method for recognizing the steering wheel off-hand state and driving intention based on three-dimensional force perception, including: S1: A multi-point distributed sensor unit is installed inside the steering wheel rim at the 3 o'clock, 6 o'clock, 9 o'clock, and 12 o'clock positions to form a complete force sensing loop, acquiring the three-dimensional force signal acting on the steering wheel rim in real time. The three-dimensional force signal includes at least radial force (…). ), tangential force ( ) and axial force ( ).

[0026] S2: Process the three-dimensional force signal to generate driver state information. This information contains two levels: Discrete state layer: Based on the amplitude and duration of the force signal, determine the "confirmed release from hand" or "hand in loop" state.

[0027] Continuous Intent Layer: In the "hand on ring" state, further analyze the dynamic change patterns of each component force to generate continuous signals such as "steering intent intensity" and "grip strength level".

[0028] S3: Based on the driver status information, output intelligent driving control commands, such as triggering an alarm, maintaining autonomous driving, or initiating a handover of permissions.

[0029] S4: Executes intelligent driving control commands, and achieves safe driving and authority handover through the coordinated control of the steering system, braking system, alarm system and human-machine interface system.

[0030] Secondly, the present invention provides an intelligent driving control system for implementing the above method, comprising: A three-dimensional force sensing module is used to integrate on the rim of the steering wheel to detect three-dimensional force signals applied by the driver's hands. The three-dimensional force signals include at least radial force, tangential force, and axial force. The signal processing module is communicatively connected to the three-dimensional force sensing module and is used to receive and process the three-dimensional force signal. By running the hands-off state and driving intention recognition algorithm, it generates driver state information containing discrete state layer and continuous intention layer. The vehicle control decision module is communicatively connected to the signal processing module and is used to receive the driver status information, make intelligent driving control decisions according to the preset control strategy, and generate corresponding control commands. The control execution interface module is communicatively connected to the vehicle control decision module and is used to convert the control commands into specific execution signals to drive the vehicle actuators to achieve vehicle control and alarm prompts. The discrete state layer includes a "confirmed release" state based on the force signal amplitude and duration, and a "hand in loop" state based on the instantaneous threshold of the force signal; the continuous intention layer includes steering intention intensity and grip strength level signals generated based on the dynamic change pattern of each component force in the "hand in loop" state.

[0031] The three-dimensional force sensing module includes a three-dimensional force sensor array distributed at multiple locations on the rim of the steering wheel, and each sensing unit can simultaneously detect force components in three orthogonal directions.

[0032] The signal processing module includes a microprocessor and a storage unit. The microprocessor is configured to perform the following processes: preprocessing the three-dimensional force signal, including filtering, noise reduction, and coordinate transformation; calculating the resultant force amplitude of the three-dimensional force; determining the discrete state based on the resultant force amplitude and its duration; and recognizing continuous driving intentions based on the dynamic change patterns of each component force.

[0033] The vehicle control decision module is configured to execute at least one of the following control strategies based on the driver's state information: triggering a graded alarm when the hands-off state continues to time out; initiating a transfer of control when the steering intention intensity exceeds the intervention threshold; and assessing the driver's state based on the grip strength level and adjusting system parameters accordingly.

[0034] The control execution interface module includes: a steering control interface for implementing progressive authority handover based on steering intention intensity; a braking control interface for executing a graded braking strategy based on hands-off status; an alarm execution interface for driving a multimodal alarm system; and a human-machine interaction interface for providing driving status display and takeover guidance. Figure 9 This is a block diagram of the overall system structure.

[0035] Example 1: Driver State Recognition and Control System Based on Three-Dimensional Force Sensing refer to Figure 1 The overall process of this method begins at step S101.

[0036] S101: The system uses a three-dimensional force sensor array embedded in the steering wheel rim to collect the radial force applied by the driver's hands in real time. ), tangential force ( ) and axial force ( The original three-dimensional force signal () , , These signals, after being converted by an ADC, are sent to the signal processing module. These signals can originate from the nested magnetorheological sensor described in claim 7, or from other sensors capable of providing three-dimensional force output.

[0037] The next step is the core step S102: driver status information generation. Details of this step are as follows... Figure 2 As shown.

[0038] S201: Signal preprocessing, low-pass filtering (cutoff frequency 10Hz) is performed on the original three-dimensional force signal to remove high-frequency noise, and coordinate normalization is performed to eliminate signal differences caused by different hand positions.

[0039] S202: Calculate the resultant force amplitude, calculate the resultant force amplitude of instantaneous three-dimensional forces. .

[0040] S203: Discrete State Judgment (Parallel Processing) S203a: "Confirm Release" judgment, judgment Is it consistently below the threshold? F L (e.g., 0.5N) and the time reaches T Hold (For example, 3 seconds). If so, proceed to S204, determine the "confirmed release" state, and start or increment the release timer.

[0041] S203b: Parallel "hand in loop" judgment, judging any force component ( , , Does the instantaneous value of ) exceed the threshold? F H (e.g., 2N). If so, proceed to S205, immediately determine the "hand on ring" state, and reset the off-hand timer.

[0042] S206: After determining that "hand is on the ring", perform continuous intent recognition: S206a: Steering intention recognition, continuous monitoring of tangential force Calculate its rate of change and steady-state value. When rate of change If the value exceeds a positive threshold (e.g., 5 N / s) and the steady-state value continues to increase, then a value similar to... The proportional "steering intention intensity" signal is normalized to 0-100%. Simultaneously, the radial force is verified. (Compression) or axial force Whether a pull / push interaction pattern emerges that corresponds to the steering direction can increase confidence. For example, a left turn intention typically corresponds to a positive turn. (Define direction), increase (Right hand squeeze) or negative (Pull with left hand). If the patterns are consistent, increase the "confidence" of the intent signal; if they are inconsistent, reduce the weight of the intent signal or mark it as potential noise.

[0043] S206b: Grip strength level identification. Axial force is taken. The filtered amplitude is mapped to a "grip strength level" signal, which can be used to assess the driver's level of tension or fatigue.

[0044] S207: Summarize and output driver status information, package discrete states (hands off / on loop) and continuous intention signals (steering intention strength, direction, confidence level; grip strength level) and hands-off timer values ​​into a driver status information package, and output it to the decision module.

[0045] exist Figure 1 The S103 intelligent driving control strategy has the following logic: Figure 3 As shown.

[0046] S301: The system is in automatic driving mode.

[0047] S302: Check if a "confirm release" signal has been received and if the release timer has expired (e.g., 15 seconds). If so, proceed to S303 to trigger a tiered alarm (e.g., sound, light, touch).

[0048] S304: Check if a "hands-on-loop" signal is received. If so, proceed to S305 to further check if the "steering intent strength" exceeds the intervention threshold.

[0049] S306: If the intensity of intent is lower than the intervention threshold (e.g., 30%), the system determines that the driver has no clear intention to take over and maintains automatic driving.

[0050] S307: If the steering intention intensity exceeds the intervention threshold (e.g., 30%) and the confidence level is higher than a certain level, the system determines that the driver has a clear intention to take over. The system initiates the control transfer process, gradually reduces the auxiliary steering torque, and smoothly returns control to the driver.

[0051] exist Figure 1 S104: Intelligent Vehicle Execution System.

[0052] Control commands are sent to each actuator via the CAN bus to achieve coordinated control.

[0053] S401: Steering control system execution. Input steering intention signal from the decision-making layer, including intention intensity (0-100%), steering direction, and confidence level. The system calculates the target assist torque curve. When the intention intensity exceeds the 30% threshold, the system enters the handover preparation state and preloads control parameters. When the intention intensity exceeds 60%, actual torque transfer begins. During the handover process, a micro-vibration motor in the steering wheel generates progressive vibration, with the vibration intensity increasing linearly with the handover progress, simulating the "clutch engagement" sensation and providing the driver with intuitive tactile feedback. When the assist torque drops to the preset minimum value and the driver's torque is detected to be stable for more than 0.5 seconds, the system confirms complete transfer of control and updates the driving mode to manual driving. Figure 4 This is a flowchart of the intelligent driving control decision-making process. Figure 5 This is a timing diagram for monitoring normal driving conditions. Figure 6 This is a sequence diagram for identifying steering intent and handing over control.

[0054] S402: Braking system activation, input hands-off status information (hands-off duration), vehicle status (vehicle speed, acceleration) and environmental information (distance to the vehicle in front, road type).

[0055] Determine the target deceleration based on a predefined risk mapping table: 5-15 seconds after release: low risk, with a comfortable deceleration of 0.05-0.1g.

[0056] 15-30 seconds after release: medium risk, with a noticeable deceleration of 0.1-0.2g.

[0057] If hands are off for more than 30 seconds or an emergency road condition is detected: High risk, execute a forced deceleration of 0.2-0.3g.

[0058] Collision approach: Execute 0.6-0.8g emergency braking.

[0059] A gradient control algorithm is employed to limit the rate of deceleration change to within 0.05 g / s, ensuring smooth operation. Pre-braking is performed to eliminate braking gaps, and the distribution of braking force to each wheel is dynamically optimized based on real-time vehicle conditions (load, coefficient of adhesion). Simultaneously, real-time communication with the steering system is maintained; when the system detects a steering intention from the driver, the braking force is appropriately reduced to maintain steering agility, and during braking, the sensitivity of the steering system is limited to prevent loss of control. Figure 7 This is the timing diagram for detecting the off-hand state.

[0060] S403: Multi-level alarm system execution, based on a multi-modal collaborative alarm strategy of state risk assessment, dynamically adjusts alarm intensity and mode according to driving scenario (highway, city, night).

[0061] Alarm level mapping: Level 1 (Reminder): A gentle reminder in the initial stage of letting go.

[0062] Level 2 (Warning): Timeout or distraction detected.

[0063] Level 3 (Severe): Continued absence of contact with the environment and increased environmental risk.

[0064] Emergency (Immediate Takeover): System limits or emergency situation. Figure 8 A sequence diagram for emergency response.

[0065] S404: Human-machine interface system execution, multi-channel interactive feedback strategy based on driving status.

[0066] Visual interface: The instrument panel and head-up display system dynamically display the driving mode (e.g., blue - automatic driving, yellow - in transition, green - manual driving), permission handover progress bar, and surrounding environmental risk indicators.

[0067] Haptic interface: Provides contextualized vibration cues of different intensities and modes during permission handover, operation confirmation, and system boundary reminders.

[0068] Voice interface: Performs tiered voice guidance, from "Autonomous driving is activated" to "Please prepare to take over", and then to "Take over vehicle control immediately" in an emergency.

[0069] Personalization: The system learns driver preferences, allows for customized interaction styles, and dynamically adjusts the intensity of feedback based on driver fatigue levels.

[0070] S405: System coordination and fault-tolerant control execution. A central gateway ensures time synchronization of instructions from all actuators, with system-level coordination errors controlled within a threshold time (e.g., 15ms). An execution status feedback loop is established to verify the effectiveness of instruction execution in real time. Complete control instructions, execution status, and abnormal events are recorded. In extreme fault conditions, a "lowest risk state" strategy is executed, progressively decelerating to a safe stop, automatically engaging P gear, engaging the electronic parking brake, unlocking the doors, and activating the emergency call.

[0071] Example 2: An Optimized Example Based on Deep Learning Intent Recognition In another embodiment, the continuous intent recognition in step S206 can be optimized using a deep learning model. Specifically, the preprocessed three-dimensional force signal time-series data ( , , The input is a pre-trained 1D convolutional neural network or LSTM network. This network is trained on a large amount of real driving data (including various steering and grip behavior labels), and can directly extract high-level features from the raw signal, outputting more refined intent recognition results, such as categories and probabilities like "intent to steering," "intent to hold," and "minor adjustment." This method can better handle complex and personalized driver operation patterns, further improving the accuracy and generalization ability of recognition.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception, characterized in that, The method specifically includes the following steps: S1. Real-time acquisition of three-dimensional force signals acting on the steering wheel rim, wherein the three-dimensional force signals include at least radial force, tangential force and axial force; S2. Process the three-dimensional force signal and, based on the amplitude, duration and dynamic change characteristics of each component, determine and generate driver state information, which includes discrete hands-off state signals and continuous control intention signals. S3. Based on the driver status information, output intelligent driving control commands, including access control, early warning control, control handover, and system coordination commands; S4. Execute intelligent driving control commands to achieve safe driving and authority transfer through the coordinated control of the steering system, braking system, alarm system and human-machine interface system.

2. The method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception according to claim 1, characterized in that, Step S1 specifically includes: the sensing units are distributed in a multi-point manner inside the steering wheel rim, with sensing units set at the 3 o'clock, 6 o'clock, 9 o'clock, and 12 o'clock positions to form a complete force sensing loop, thereby acquiring the three-dimensional force signal acting on the steering wheel rim in real time. The three-dimensional force signal includes at least radial force. Tangential force and axial force .

3. The method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception according to claim 2, characterized in that, In step S2, driver status information is generated, which includes two levels: Discrete state layer: Based on the amplitude and duration of the force signal, determine the "confirmed release from hand" or "hand in loop" state; Continuous Intent Layer: In the "hand on ring" state, further analyze the dynamic change patterns of each component force to generate continuous signals such as "steering intent intensity" and "grip strength level".

4. The method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception according to claim 3, characterized in that, Step S2 specifically includes: S21. Signal preprocessing: Low-pass filtering is performed on the original three-dimensional force signal to remove high-frequency noise, and coordinate normalization is performed to eliminate signal differences caused by different hand positions. S22. Calculate the resultant force amplitude; calculate the resultant force amplitude of the instantaneous three-dimensional forces. ; S23. Discrete state judgment and processing: 1) "Confirm release" judgment, judgment Is it consistently below the threshold? F L And the time reached T Hold If so, it is determined to be in the "confirmed release" state, and the release timer is started or incremented; 2) Parallel "hand in loop" judgment, judging any force component ( , , Does the instantaneous value of ) exceed the threshold? F H If so, it will be immediately determined as a "hand on ring" state, and the off-hand timer will be reset; S24. After determining that "hand is on the ring", perform continuous intent recognition: 1) Steering intent recognition and continuous monitoring of tangential force Calculate its rate of change and steady-state value; when rate of change If the positive threshold is exceeded and the steady-state value continues to increase, then a value similar to... The proportional "steering intent intensity" signal is normalized to 0-100%; simultaneously, the radial force is verified. or axial force If a matching pattern appears that corresponds to the steering direction, the confidence level is increased; if the pattern matches, the confidence level of the intention signal is increased; if the pattern does not match, the intention signal is downweighted or marked as potential noise. 2) Grip strength level identification, based on axial force. The filtered amplitude is mapped to a "grip strength level" signal, which can be used to assess the driver's level of tension or fatigue. S25. Summarize and output the driver status information, package the discrete status and continuous intention signals and the hands-off timer value into a driver status information package, and output it to the decision module.

5. The method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception according to claim 4, characterized in that, Step S3 includes: When a "confirm hands-off" status signal is received and the duration exceeds the safety threshold, a driver takeover request alarm is triggered. When a "hands-on-loop" status signal is received in autonomous driving mode, but the control intention signal is below the intervention threshold, the current autonomous driving level is maintained. When a control intention signal higher than the intervention threshold is received in autonomous driving mode, a smooth transfer of driving control from the autonomous driving system to the driver is initiated.

6. The method for recognizing steering wheel off-hand state and driving intention based on three-dimensional force perception according to claim 5, characterized in that, In step S4, the steering control system performs the following actions: inputting steering intention signals from the decision-making layer, including intention intensity, steering direction, and confidence level; the system calculates the target assist torque curve; when the intention intensity exceeds a 30% threshold, the system enters a handover preparation state and preloads control parameters; when the intention intensity exceeds 60%, the actual torque transfer begins; during the handover process, a micro-vibration motor on the steering wheel generates progressive vibration, with the vibration intensity increasing linearly with the handover progress, simulating the "clutch engagement" sensation and providing the driver with intuitive tactile feedback; when the assist torque drops to a preset minimum value and the driver's torque is detected to be stably maintained for more than 0.5 seconds, the system confirms that control has been completely transferred and updates the driving mode to manual driving.

7. An intelligent driving control system for implementing the method as described in any one of claims 1-6, characterized in that, The system includes: A three-dimensional force sensing module, integrated into the steering wheel rim, is used to detect three-dimensional force signals; The signal processing module is electrically connected to the three-dimensional force sensing module and is used to run the steps of the method described in any one of claims 1-6 and output driver status information; The vehicle control module is communicatively connected to the signal processing module and is used to receive the driver status information, make intelligent driving control decisions based on the preset control strategy, and generate corresponding control commands. The execution interface module is communicatively connected to the vehicle control decision module and is used to convert the control commands into specific execution signals to drive the vehicle actuators to achieve vehicle control and alarm prompts.

8. A vehicle, characterized in that, It includes the intelligent driving control system as described in claim 7.

Images