A steering wheel hands-off detection method based on torque signal analysis and related device

Abstract

A steering wheel hand-off detection method based on torque signal analysis and related equipment relate to the field of automobile auxiliary driving and safety technology. By implementing the method, the detection system acquires steering torque sensor signals, motor working information and vehicle state information, accurately estimates the non-driver torque component generated by the motor counterforce and system noise with the help of the driver torque estimation model, effectively removes irrelevant interference factors, and makes the separated driver torque component (pure driver torque estimation value) more consistent with the real driving scene. The detection system maps the pure driver torque estimation value to the current hand-off probability score, inputs it to the probability accumulator for accumulation operation, obtains the cumulative probability value, compares it with the threshold value to determine the steering wheel state and confidence, and improves the accuracy of the steering wheel hand-off detection.

Description

Technical Field

[0001] This application relates to the field of automotive driver assistance and safety technology, and in particular to a steering wheel off-hand detection method and related equipment based on torque signal analysis. Background Technology

[0002] With the continuous improvement of automotive automation, intelligent driving assistance systems are being used more and more widely in vehicles. Steering assist, as an important component of intelligent driving assistance systems, can provide steering assistance to the driver in scenarios such as lane departure and lane keeping. In practical applications, to ensure driving safety, it is necessary to accurately determine whether the driver's hands have left the steering wheel, thereby promptly alerting the driver or taking appropriate measures.

[0003] Currently, the commonly used method for detecting whether a driver is holding the steering wheel is to use capacitive sensing technology. That is, a capacitive sensor is installed on the surface of the steering wheel. When the driver's hand touches the steering wheel, the capacitance of the sensor changes due to the influence of the human body's capacitance. The driver can then determine whether he is holding the steering wheel based on the change in the capacitance.

[0004] However, the steering wheel off-hand detection method based on capacitive sensing is easily affected by environmental factors. For example, in wet weather or when the driver's hands are sweaty, the capacitance measurement will produce a large error, resulting in low accuracy of steering wheel off-hand detection. Summary of the Invention

[0005] This application provides a steering wheel off-hand detection method and related equipment based on torque signal analysis, which can improve the accuracy of steering wheel off-hand detection.

[0006] In a first aspect, this application provides a steering wheel off-hand detection method based on torque signal analysis, applied to a detection system, the method comprising:

[0007] The system acquires steering torque sensor signals, motor operating information, and vehicle status information. The steering torque sensor signals are used to characterize the actual value of the steering wheel torque. The motor operating information includes the motor current and speed, and the vehicle status information includes vehicle speed and steering angle.

[0008] The steering torque sensor signal, motor operating information and vehicle status information are input into the driver torque estimation model to obtain the pure driver torque estimate. The driver torque estimation model estimates the non-driver torque component generated by the motor reverse force and system noise based on the motor operating information and vehicle status information, and calculates the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal.

[0009] Based on a pre-calibrated torque-probability mapping relationship, the pure driver torque estimate is converted into a current hands-off probability score;

[0010] Input the current probability score of leaving the hand into the preset probability accumulator for accumulation calculation to obtain the cumulative probability value;

[0011] Based on the comparison between the cumulative probability value and the preset probability threshold, the steering wheel status and confidence level are determined and output. The confidence level is calculated based on the cumulative probability value.

[0012] By employing the above technical solution, the detection system acquires steering torque sensor signals, motor operating information, and vehicle status information. Utilizing a driver torque estimation model, it accurately estimates the non-driver torque components generated by the motor's reverse force and system noise, effectively eliminating irrelevant interference factors and making the separated driver torque component (pure driver torque estimate) more closely resemble real-world driving scenarios. The detection system maps the pure driver torque estimate to a current hand-off probability score, inputs it to a probability accumulator for accumulation, obtains a cumulative probability value, and combines it with a threshold comparison to determine the steering wheel status and confidence level. This steering wheel hand-off detection method based on torque signal analysis avoids the shortcomings of traditional capacitive sensing detection, which is easily affected by environmental factors. It does not produce significant errors due to humid weather or sweaty hands, thus improving the accuracy of steering wheel hand-off detection.

[0013] In conjunction with some embodiments of the first aspect, in some embodiments, the steering torque sensor signal, motor operating information, and vehicle status information are input into the driver torque estimation model to obtain a pure driver torque estimate. The driver torque estimation model estimates the non-driver torque component generated by the motor's reverse force and system noise based on the motor operating information and vehicle status information, and calculates the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal, specifically including:

[0014] A motor model is established based on the motor's operating information, and the components of the motor's reverse force are calculated.

[0015] A system noise model is established based on vehicle status information, and the system noise components are calculated.

[0016] Subtracting the motor's reverse force component and the system noise component from the actual steering wheel torque value yields the pure driver torque estimate.

[0017] By employing the above technical solution, the detection system establishes a motor model and a system noise model to estimate the non-driver torque components (motor reverse force component and system noise component). Then, the non-driver torque components are subtracted from the actual steering wheel torque value to obtain the pure driver torque estimate. This modeling approach considers the main factors affecting steering torque, making torque separation more accurate: the motor model can accurately estimate the impact of the motor reverse force on steering torque, and the system noise model can effectively filter out interference factors such as road feedback, thus making the estimation of each component more precise and providing reliable basic data for subsequent hands-off state judgment.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, a motor model is established based on the motor operating information, and the motor reverse force component is calculated, specifically including:

[0019] Calculate the reverse force of the base motor based on the current, motor torque constant, and transmission efficiency;

[0020] Calculate the dynamic speed compensation value based on the rotational speed;

[0021] The motor's reverse force component is obtained by adding the basic motor's reverse force and the speed dynamic compensation value.

[0022] By adopting the above technical solution, the detection system considers both the basic motor reverse force and the dynamic speed compensation value during the calculation of the motor's reverse force component. The basic motor reverse force reflects the motor's fundamental characteristics, while the dynamic speed compensation value considers the dynamic impact of the motor's speed. This multi-dimensional compensation mechanism makes the estimation of the motor's reverse force component more comprehensive and accurate, effectively improving the separation accuracy of the non-driver torque component, thereby further enhancing the reliability of steering wheel off-hand detection.

[0023] In conjunction with some embodiments of the first aspect, in some embodiments, the current probability score of leaving the hand is input into a preset probability accumulator for accumulation calculation to obtain a cumulative probability value, specifically including:

[0024] Get the current accumulated value of the probability accumulator;

[0025] Add the current exit probability score to the current accumulated value to obtain the initial accumulated probability value;

[0026] The initial cumulative probability value is subjected to a limiting process to obtain a cumulative probability value, so that the cumulative probability value is kept within a preset range.

[0027] By employing the above technical solution, the detection system obtains the current accumulated value of the probability accumulator, adds it to the new current hands-off probability score, and then performs a limiting process to obtain the final accumulated probability value. The accumulation mechanism effectively smooths out instantaneous fluctuations, avoiding misjudgments of hands-off states due to brief torque changes. The limiting process ensures that the accumulated probability value remains within a reasonable range, preventing it from increasing indefinitely or decreasing excessively. This method guarantees both detection sensitivity and result stability, effectively addressing torque variation characteristics under various driving scenarios and improving the robustness and reliability of the detection system.

[0028] In conjunction with some embodiments of the first aspect, in some embodiments, the steering wheel state and confidence level are determined and output based on the comparison result of the cumulative probability value and a preset probability threshold. The confidence level is calculated based on the cumulative probability value, specifically including:

[0029] If the cumulative probability value is greater than the first probability threshold, the steering wheel is determined to be in the off-hand state and the off-hand state confidence is calculated. The off-hand state confidence is equal to the difference between the cumulative probability value and the first probability threshold divided by the difference between the preset maximum probability value and the first probability threshold.

[0030] If the cumulative probability value is between the first probability threshold and the second probability threshold, the steering wheel state is determined to be a state hold and the state hold confidence is calculated. The state hold confidence is equal to the difference between the cumulative probability value and the second probability threshold divided by the difference between the first probability threshold and the second probability threshold.

[0031] If the cumulative probability value is less than the second probability threshold, the steering wheel is determined to be in hand and the confidence level of the in hand state is calculated. The confidence level of the in hand state is equal to the difference between the second probability threshold and the cumulative probability value divided by the difference between the second probability threshold and the preset minimum probability value.

[0032] By adopting the above technical solution, the detection system sets two probability thresholds to classify the steering wheel status into three types: off-hand, held, and in-hand. A corresponding confidence calculation formula is provided for each steering wheel status. This hierarchical judgment mechanism adds intermediate transition states, avoids rigid binary judgments, and better reflects actual driving scenarios. It not only provides more nuanced status judgments but also offers more decision-making basis for subsequent warning strategies, effectively improving the practicality and reliability of the detection system.

[0033] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of inputting the steering torque sensor signal, motor operating information, and vehicle state information into the driver torque estimation model to obtain a pure driver torque estimate, and the driver torque estimation model estimating the non-driver torque component generated by the motor reverse force and system noise based on the motor operating information and vehicle state information, to calculate the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal, the method further includes:

[0034] The steering torque sensor signal, motor operating information, and vehicle status information are preprocessed by low-pass filtering to obtain the preprocessed steering torque sensor signal, motor operating information, and vehicle status information.

[0035] By adopting the above technical solution, the detection system preprocesses various input signals through a low-pass filter before signal processing, which improves the quality of subsequent signal processing, reduces interference components, lays a good data foundation for torque estimation and state judgment, and improves the anti-interference capability and accuracy of the entire detection system.

[0036] In conjunction with some embodiments of the first aspect, in some embodiments, after determining and outputting the steering wheel state and confidence level based on the comparison result of the cumulative probability value and a preset probability threshold, wherein the confidence level is calculated based on the cumulative probability value, the method further includes:

[0037] If the steering wheel is in an off-hand state and the confidence level of the off-hand state exceeds the preset confidence threshold, obtain the current vehicle driving scenario information;

[0038] Determine the scenario risk level based on the current vehicle driving scenario information;

[0039] Match the warning strategy corresponding to the scenario risk level from the warning strategy table.

[0040] By employing the above technical solution, when the detection system detects a hands-off state and the confidence level of this state exceeds a preset confidence threshold, it further analyzes the current vehicle driving scenario information and determines the scenario risk level to match an appropriate warning strategy. By combining hands-off detection with scenario risk analysis, the detection system can more accurately assess the degree of danger of the current situation, avoiding over-warning or under-warning. This intelligent warning strategy not only ensures driving safety but also improves the practicality of the detection system and reduces unnecessary disturbance to the driver.

[0041] In a second aspect, embodiments of this application provide a detection system comprising: one or more processors and a memory; the memory is coupled to the one or more processors and is used to store computer program code, the computer program code including computer instructions, wherein the one or more processors invoke the computer instructions to cause the detection system to perform the method described in the first aspect and any possible implementation thereof.

[0042] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a detection system, cause the detection system to perform the method described in the first aspect and any possible implementation thereof.

[0043] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a detection system, cause the detection system to perform the method described in the first aspect and any possible implementation thereof.

[0044] Understandably, the detection system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0045] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0046] 1. By adopting the above technical solution, the detection system acquires steering torque sensor signals, motor operating information, and vehicle status information. Utilizing a driver torque estimation model, it accurately estimates the non-driver torque components generated by the motor's reverse force and system noise, effectively eliminating irrelevant interference factors and making the separated driver torque component (pure driver torque estimate) more closely resemble real-world driving scenarios. The detection system maps the pure driver torque estimate to a current hand-off probability score, inputs it to a probability accumulator for accumulation, obtains a cumulative probability value, and combines it with a threshold comparison to determine the steering wheel status and confidence level. This steering wheel hand-off detection method based on torque signal analysis avoids the shortcomings of traditional capacitive sensing detection, which is easily affected by environmental factors. It does not produce significant errors due to humid weather or sweaty hands, thus improving the accuracy of steering wheel hand-off detection.

[0047] 2. By adopting the above technical solution, the detection system establishes a motor model and a system noise model to estimate the non-driver torque components (motor reverse force component and system noise component). Then, the non-driver torque components are subtracted from the actual steering wheel torque value to obtain the pure driver torque estimate. This modeling method considers the main factors affecting steering torque, making torque separation more accurate: the motor model can accurately estimate the impact of the motor reverse force on steering torque, and the system noise model can effectively filter out interference factors such as road feedback, thus making the estimation of each component more accurate and providing reliable basic data for subsequent hands-off state judgment.

[0048] 3. By adopting the above technical solution, when the detection system detects a hands-off state and the confidence level of the hands-off state exceeds a preset confidence threshold, it will further analyze the current vehicle driving scenario information and determine the scenario risk level to match the corresponding warning strategy. By combining hands-off detection with scenario risk analysis, the detection system can more accurately assess the degree of danger of the current situation, avoiding the problems of over-warning or under-warning. This intelligent warning strategy not only ensures driving safety but also improves the practicality of the detection system and reduces unnecessary disturbance to the driver. Attached Figure Description

[0049] Figure 1 This is a flowchart illustrating a steering wheel off-hand detection method based on torque signal analysis in an embodiment of this application.

[0050] Figure 2 This is another flowchart illustrating the steering wheel off-hand detection method based on torque signal analysis in this application embodiment;

[0051] Figure 3 This is a schematic diagram of the physical device structure of the detection system in the embodiments of this application. Detailed Implementation

[0052] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0053] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0054] The following describes the process of the method provided in this implementation. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a steering wheel off-hand detection method based on torque signal analysis in an embodiment of this application.

[0055] S101. Acquire steering torque sensor signal, motor operating information and vehicle status information. Steering torque sensor signal is used to characterize the actual value of steering wheel torque. Motor operating information includes motor current and speed. Vehicle status information includes vehicle speed and steering angle.

[0056] Among them, the steering torque sensor signal represents the actual torque value of the steering wheel measured by the steering torque sensor in the steering system, in Newton-meters (N·m); the motor operating information refers to the operating status parameters of the motor that assists steering, including the motor current (unit: ampere-A) and speed (unit: revolutions per minute rpm); the vehicle status information is used to represent the vehicle's motion status, including vehicle speed (unit: kilometers per hour (km / h)) and steering angle (unit: degrees (°)).

[0057] Specifically, the detection system collects steering torque sensor signals, motor operating information, and vehicle status information from the steering torque sensor, motor controller, and vehicle condition monitoring system via CAN bus or other communication interfaces. The sampling frequency is typically above 100Hz to ensure real-time and continuous data transmission. The detection system performs preliminary validity verification on the collected raw signals, eliminating obvious outliers.

[0058] S102. Input the steering torque sensor signal, motor operating information and vehicle status information into the driver torque estimation model to obtain the pure driver torque estimate. The driver torque estimation model estimates the non-driver torque component generated by the motor reverse force and system noise based on the motor operating information and vehicle status information, so as to calculate the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal.

[0059] Among them, the driver torque estimation model represents a model for separating the driver torque component, which can be obtained through machine learning or physical modeling; the pure driver torque estimate represents the driver torque component, which refers to the steering torque applied only by the driver, without including the influence of motor reverse force and system noise; the non-driver torque component refers to the interference torque generated by motor reverse force and system noise.

[0060] Specifically, on a test bench or in a real vehicle, a series of known and accurate driver torques are applied to the steering wheel using equipment. The detection system simultaneously collects steering torque sensor signals, motor operating information, and vehicle status information. A deep learning network (such as LSTM, CNN, or a fully connected network) is used to train the model on this data. The inputs are driver torque, steering torque sensor signals, motor operating information, and vehicle status information; the output is the model parameters. The goal is to minimize the error between the pure driver torque estimate calculated by the model and the known driver torque. For example, Mean Squared Error (MSE) can be used as the loss function to optimize the model parameters. The training dataset should broadly cover different vehicle speeds, steering angles, road conditions, and motor assist modes to ensure the model's generalization ability. Through training, the model learns how to isolate the influence of non-driver torque components generated by the motor's reaction force and system noise from the steering torque sensor signals, thereby accurately estimating the pure driver torque.

[0061] The detection system inputs the steering torque sensor signal, motor operating information and vehicle status information obtained in step S101 into the driver torque estimation model to obtain the pure driver torque estimate.

[0062] Optionally, in general, the steering torque sensor signal, motor operating information, and vehicle status information are input into the driver torque estimation model to obtain a pure driver torque estimate. The driver torque estimation model estimates the non-driver torque components generated by the motor's reaction force and system noise based on the motor operating information and vehicle status information. The pure driver torque estimate can be calculated based on the non-driver torque components and the steering torque sensor signal in the following ways, which are not limited here: establish a motor model based on the motor operating information and calculate the motor reaction force component; establish a system noise model based on the vehicle status information and calculate the system noise component; subtract the motor reaction force component and system noise component from the actual steering wheel torque value to obtain the pure driver torque estimate.

[0063] Among them, the motor model represents a mathematical model describing the characteristics of the motor generating reverse force, including current-torque characteristics and speed-damping characteristics; the motor reverse force component refers to the reverse torque generated by the motor on the steering system during operation, the magnitude of which is related to the motor's current and speed; the system noise model represents a mathematical model describing interference factors such as road feedback force and friction; the system noise component represents the interference torque generated by factors such as road feedback and mechanical friction; the actual steering wheel torque value refers to the total torque value actually measured by the steering torque sensor; and the pure driver torque estimate represents the estimation result of the steering torque applied only by the driver.

[0064] Specifically, the detection system builds a motor model based on the motor's operating information. The motor model includes the following sub-modules: a current-based torque calculation module, which multiplies the current by a torque constant to obtain the basic motor torque; and a speed-based dynamic damping module, which calculates the dynamic damping torque proportional to the speed. The outputs of these two modules are added together to obtain the motor's reverse force component.

[0065] The detection system establishes a system noise model based on vehicle speed and steering angle. The system noise model considers the road feedback force related to vehicle speed and the mechanical friction force related to steering angle, and obtains the system noise components by looking up tables or calculating functions.

[0066] The detection system subtracts the calculated motor reaction force component and system noise component from the actual steering wheel torque value to obtain a pure driver torque estimate. The entire process can be updated at a frequency of 100Hz to ensure real-time performance. The detection system also performs a plausibility check on the calculation results; if an anomaly is detected, a fault handling mechanism is triggered.

[0067] Optionally, under normal circumstances, the motor model can be established based on the motor's operating information, and the motor's reverse force component can be calculated in the following ways, which are not limited here: calculate the basic motor's reverse force based on the current, motor torque constant, and transmission efficiency; calculate the speed dynamic compensation value based on the speed; and add the basic motor's reverse force and the speed dynamic compensation value to obtain the motor's reverse force component.

[0068] Among them, the motor torque constant represents the linear proportionality coefficient between the motor current and the output torque, with the unit being N·m / A; the transmission efficiency refers to the energy conversion efficiency of the mechanical transmission chain of the steering system, which is usually less than 1; the basic motor reverse force represents the ideal motor torque considering only the current effect; the speed dynamic compensation value refers to the dynamic damping torque caused by the change of motor speed, which is proportional to the speed; and the motor reverse force component represents the actual motor reverse torque after considering all influencing factors.

[0069] Specifically, firstly, the detection system calculates the reverse force of the base motor using the formula: T_base=Kt×I×η, where T_base is the reverse force of the base motor, Kt is the motor torque constant (e.g., 0.1 N·m / A), I is the current, and η is the transmission efficiency (e.g., 0.85).

[0070] Next, the detection system calculates the dynamic speed compensation value. The calculation formula is: T_speed=Kd×ω, where T_speed is the dynamic speed compensation value, Kd is the speed damping coefficient (e.g., 0.001N·m·s / rad), and ω is the speed.

[0071] The detection system adds the two components together to obtain the complete motor reverse force component that takes into account various influencing factors.

[0072] S103. Based on the pre-calibrated torque-probability mapping relationship, the pure driver torque estimate is converted into the current hand-off probability score;

[0073] The torque-probability mapping relationship represents the correspondence between the pure driver torque estimate and the current hand-off probability score, which is obtained through experimental data calibration. The current hand-off probability score is a quantitative value used to evaluate the steering wheel status at the current moment. A positive score indicates a hand-off trend, and a negative score indicates a hand-on trend. Its value range is obtained through experimental data calibration, for example, set in the range of [-0.5, 0.5].

[0074] Specifically, the experimenters conducted numerous real-person experiments, recording the pure driver torque estimates calculated using the aforementioned driver torque estimation model, given the hand position (in hand / out). The experimenters then used deep learning or mathematical statistical analysis to determine torque segment intervals and the corresponding hand-off probability score for each torque segment interval.

[0075] Calibration principles:

[0076] (1) When the absolute value of the pure driver torque estimate is close to zero (e.g., |pure driver torque estimate| < 0.1 Nm), it corresponds to a higher positive score, indicating a strong tendency to disengage.

[0077] (2) When the absolute value of the pure driver torque estimate is within the range of typical hands on the steering wheel but without active force (e.g. 0.1 Nm < |pure driver torque estimate| < 0.5 Nm), it corresponds to a lower positive score or zero score.

[0078] (3) When the absolute value of the pure driver torque estimate is large, it indicates that the driver is actively steering (e.g., |pure driver torque estimate|>0.5Nm), which corresponds to a negative score, indicating a strong tendency to be in hand, which is used to offset the previous accumulation of off-hand.

[0079] The detection system uses a torque-probability mapping relationship pre-calibrated through a large amount of experimental data to map the pure driver torque estimate to the current hand-off probability score.

[0080] S104. Input the current probability score of leaving the hand into the preset probability accumulator for accumulation calculation to obtain the cumulative probability value;

[0081] The probability accumulator is a digital filter used to accumulate the off-hand probability scores over multiple consecutive time points. The accumulated probability value represents the final probability score after accumulation, which reflects the overall trend of the steering wheel off-hand status over a period of time.

[0082] Specifically, the detection system reads the current accumulated value from the probability accumulator, and then adds the current drop probability score to the current accumulated value. After the accumulation is complete, the detection system limits the resulting cumulative probability value to ensure it remains within a preset range, for example, ensuring it does not exceed the range [-1, 1]. For instance, when the cumulative probability value is greater than 1, it is limited to 1; when the cumulative probability value is less than -1, it is limited to -1. This accumulation mechanism can effectively filter out instantaneous interference caused by factors such as road bumps.

[0083] Optionally, under normal circumstances, the current exit probability score is input into a preset probability accumulator for accumulation to obtain the cumulative probability value. This can be achieved in the following ways, without limitation: obtain the current accumulated value of the probability accumulator; add the current exit probability score to the current accumulated value to obtain the initial cumulative probability value; and perform a limit processing on the initial cumulative probability value to obtain the cumulative probability value, so that the cumulative probability value is kept within a preset range.

[0084] The current accumulated value of the probability accumulator refers to the accumulated value recorded by the probability accumulator at the previous moment or in the previous cycle; it can also be described as the accumulated value stored by the probability accumulator at the current moment. The following explains in detail: The probability accumulator (denoted as P) is initialized to a fixed value when the detection system starts, usually 0, but other default values ​​are also possible. Each time the detection system acquires and processes a new signal (steering torque sensor signal, motor operating information, and vehicle status information), that is, after each complete cycle, it calculates the current hand-off probability score (denoted as S). The calculation process for the current hand-off probability score is described in steps S101-S103 and will not be repeated here. The detection system adds the newly obtained current hand-off probability score (S) to the accumulated value of the probability accumulator from the previous cycle (P_(n-1), i.e., the current accumulated value) to obtain a new accumulated probability value (P_n).

[0085] The formula is: P_n = P_(n-1) + S; where:

[0086] P_n: The cumulative probability value of the probability accumulator after this calculation;

[0087] P_(n-1): The accumulated value of the probability accumulator after the last calculation (i.e., the current accumulated value);

[0088] S: The current probability score of leaving the hand obtained in this calculation.

[0089] For example, assuming initial P=0, a preset interval range of [-10, 10], and the current exit probability scores calculated for three consecutive periods are 0.4, 0.5, and -0.2 respectively, the cumulative process is as follows:

[0090] After the first cycle: P_1 = P_0 + S_1 = 0 + 0.4 = 0.4; (within [-10, 10], no amplitude limiting is required)

[0091] After the second cycle: P_2=P_1+S_2=0.4+0.5=0.9; (within [-10, 10], no amplitude limiting is required)

[0092] After the 3rd cycle: P_3 = P_2 + S_3 = 0.9 + (-0.2) = 0.7; (within [-10, 10], no amplitude limiting is required)

[0093] Limiting refers to setting an upper or lower limit range to prevent the accumulated probability value from increasing or decreasing indefinitely.

[0094] Suppose that after multiple cycles of accumulation, the detection system reads the current accumulated value in the probability accumulator as 9.8, and the current probability score of leaving the hand is 0.5. The two are added together: 9.8 + 0.5 = 10.3. Since this exceeds the upper limit of 10, a limit is applied, and the final output accumulated probability value is 10.

[0095] S105. Based on the comparison between the cumulative probability value and the preset probability threshold, determine and output the steering wheel status and confidence level. The confidence level is calculated based on the cumulative probability value.

[0096] Specifically, the detection system employs a dual-threshold comparison method for state determination: when the cumulative probability value is greater than the first probability threshold, it is determined to be in a hands-off state, and a confidence score for this state is calculated; when the cumulative probability value is less than the second probability threshold, it is determined to be in hands-off state, and a confidence score for this state is calculated; when the cumulative probability value is between the two probability thresholds, the current state remains unchanged, and a confidence score for state retention is calculated. Finally, the detection system outputs the determined steering wheel state and the corresponding confidence score to the vehicle control system for subsequent warning or intervention control. This dual-threshold-based determination mechanism avoids frequent state switching and improves the stability of the detection.

[0097] Optionally, under normal circumstances, the steering wheel state and confidence level are determined and output based on the comparison between the cumulative probability value and the preset probability threshold. The confidence level, calculated based on the cumulative probability value, can be achieved in the following ways, without limitation: If the cumulative probability value is greater than the first probability threshold, the steering wheel state is determined to be off-hand state and the off-hand state confidence level is calculated. The off-hand state confidence level is equal to the difference between the cumulative probability value and the first probability threshold divided by the difference between the preset maximum probability value and the first probability threshold. If the cumulative probability value is between the first probability threshold and the second probability threshold, the steering wheel state is determined to be state held and the state held confidence level is calculated. The state held confidence level is equal to the difference between the cumulative probability value and the second probability threshold divided by the difference between the first probability threshold and the second probability threshold. If the cumulative probability value is less than the second probability threshold, the steering wheel state is determined to be on-hand state and the on-hand state confidence level is calculated. The on-hand state confidence level is equal to the difference between the second probability threshold and the cumulative probability value divided by the difference between the second probability threshold and the preset minimum probability value.

[0098] By employing the above technical solution, the detection system acquires steering torque sensor signals, motor operating information, and vehicle status information. Utilizing a driver torque estimation model, it accurately estimates the non-driver torque components generated by the motor's reverse force and system noise, effectively eliminating irrelevant interference factors and making the separated driver torque component (pure driver torque estimate) more closely resemble real-world driving scenarios. The detection system maps the pure driver torque estimate to a current hand-off probability score, inputs it to a probability accumulator for accumulation, obtains a cumulative probability value, and combines it with a threshold comparison to determine the steering wheel status and confidence level. This steering wheel hand-off detection method based on torque signal analysis avoids the shortcomings of traditional capacitive sensing detection, which is easily affected by environmental factors. It does not produce significant errors due to humid weather or sweaty hands, thus improving the accuracy of steering wheel hand-off detection.

[0099] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is another flowchart illustrating the steering wheel off-hand detection method based on torque signal analysis in this application embodiment.

[0100] S201. Acquire steering torque sensor signal, motor operating information and vehicle status information. The steering torque sensor signal is used to characterize the actual value of steering wheel torque. The motor operating information includes the motor current and speed. The vehicle status information includes vehicle speed and steering angle.

[0101] For details, please refer to step S101, which will not be repeated here.

[0102] S202. Perform low-pass filtering preprocessing on the steering torque sensor signal, motor operating information and vehicle status information to obtain the preprocessed steering torque sensor signal, motor operating information and vehicle status information.

[0103] Low-pass filtering preprocessing refers to a digital signal processing method that filters high-frequency noise from a signal to remove high-frequency interference components; the preprocessed signal represents the smoothed signal obtained after low-pass filtering.

[0104] Specifically, the detection system can employ a Butterworth low-pass filter with a cutoff frequency set to 10Hz to effectively filter out high-frequency noise introduced by factors such as mechanical vibration and electrical interference. For the steering torque sensor signal, the low-pass filter can smooth out instantaneous disturbances caused by road bumps; for motor operating information, the low-pass filter can suppress high-frequency fluctuations in the current and speed signals; for vehicle status information, the low-pass filter can smooth out sudden changes in vehicle speed and steering angle. The filtered signal is more stable and can better reflect the actual changing trends of various physical quantities, providing more reliable input data for subsequent torque estimation.

[0105] S203. Input the steering torque sensor signal, motor operating information and vehicle status information into the driver torque estimation model to obtain the pure driver torque estimate. The driver torque estimation model estimates the non-driver torque component generated by the motor reverse force and system noise based on the motor operating information and vehicle status information, so as to calculate the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal.

[0106] For details, please refer to step S102, which will not be repeated here.

[0107] S204. Based on the pre-calibrated torque-probability mapping relationship, the pure driver torque estimate is converted into the current hand-off probability score.

[0108] For details, please refer to step S103, which will not be repeated here.

[0109] S205. Input the current probability score of leaving the hand into the preset probability accumulator for accumulation calculation to obtain the cumulative probability value.

[0110] For details, please refer to step S104, which will not be repeated here.

[0111] S206. Based on the comparison between the cumulative probability value and the preset probability threshold, determine and output the steering wheel status and confidence level. The confidence level is calculated based on the cumulative probability value.

[0112] For details, please refer to step S105, which will not be repeated here.

[0113] S207. If the steering wheel is in an off-hand state and the confidence level of the off-hand state exceeds the preset confidence threshold, obtain the current vehicle driving scenario information.

[0114] The preset confidence threshold represents the minimum confidence requirement for triggering scenario analysis, and is usually set to 0.8 or higher. The current vehicle driving scenario information refers to comprehensive information describing the current driving environment and state of the vehicle, including but not limited to vehicle speed, acceleration, yaw rate, lane type, weather conditions, visibility, etc.

[0115] Specifically, when the detection system determines that the steering wheel is in a hands-off state, and the calculated confidence level of the hands-off state exceeds a preset confidence threshold (e.g., 0.9), the detection system begins collecting information about the current vehicle driving scenario. The detection system collects vehicle dynamic information through an onboard sensor network, including vehicle speed from the speed sensor, longitudinal and lateral acceleration from the accelerometer, and yaw rate from the gyroscope; it identifies lane line type (solid / dashed), road type (highway / urban road), and curve radius through a front-facing camera; and it obtains weather conditions (sunny / rainy / snowy), visibility level, and road surface adhesion coefficient through environmental sensors.

[0116] S208. Determine the scenario risk level based on the current vehicle driving scenario information.

[0117] The scenario risk level represents the graded assessment of the degree of danger in the current driving scenario, and is usually divided into three levels: low risk, medium risk, and high risk.

[0118] Specifically, the detection system employs a multi-dimensional weighted scoring method to quantitatively assess various scenario information: 3 points are awarded for speeds exceeding 80 km / h, 2 points for curve radii less than 200m, 2 points for solid lane lines, 2 points for rainy or snowy weather, and 2 points for visibility less than 200m, etc. The detection system adds up the scores of each item to obtain a total score, and determines the scenario risk level based on the range of the total score.

[0119] S209. Match the early warning strategy corresponding to the scenario risk level from the early warning strategy table.

[0120] The early warning strategy table refers to a pre-configured table of early warning strategies corresponding to different scenario risk levels; the early warning strategy refers to the specific early warning measures taken for a specific scenario risk level, including the audio-visual prompting method, prompting frequency, and prompting intensity.

[0121] Specifically, the detection system searches for a matching warning strategy in the warning strategy table based on the risk level of the scenario. For low-risk scenarios, a mild warning strategy is used: a hands-off warning icon is displayed on the dashboard, and a warning sound is emitted every 30 seconds; for medium-risk scenarios, a moderate warning strategy is used: a hands-off warning icon flashes on the dashboard, a warning sound is emitted every 10 seconds, and the steering wheel vibrates slightly; for high-risk scenarios, a strong warning strategy is used: a red hands-off warning icon flashes on the dashboard, a continuous warning sound is emitted, the steering wheel vibrates strongly, and the vehicle speed is reduced if necessary. The warning strategy continues to be executed until the detection system determines that the driver has regained control of the steering wheel.

[0122] The detection system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 3 This is a schematic diagram of the physical device structure of the detection system in this application embodiment.

[0123] It should be noted that, Figure 3 The structure of the detection system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0124] like Figure 3 As shown, the detection system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in the read-only memory ROM 302 or a program loaded from the storage section 308 into the random access memory RAM 303, such as performing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An I / O interface 305 is also connected to the bus 304.

[0125] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0126] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0127] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0128] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0129] Specifically, the detection system in this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the steering wheel off-hand detection method based on torque signal analysis provided in the above embodiment.

[0130] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the detection system described in the above embodiments; or it may exist independently and not assembled into the detection system. The storage medium carries one or more computer programs that, when executed by a processor of the detection system, cause the detection system to implement the steering wheel off-hand detection method based on torque signal analysis provided in the above embodiments.

[0131] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0132] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0133] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for detecting steering wheel removal based on torque signal analysis, characterized in that, The method, applied to a detection system, includes: acquiring steering torque sensor signals, motor operating information, and vehicle status information; the steering torque sensor signals characterizing the actual steering wheel torque value; the motor operating information including motor current and speed; and the vehicle status information including vehicle speed and steering angle; inputting the steering torque sensor signals, motor operating information, and vehicle status information into a driver torque estimation model to obtain a pure driver torque estimate; the driver torque estimation model estimating non-driver torque components generated by motor reaction force and system noise based on the motor operating information and vehicle status information, and calculating the pure driver torque estimate based on the non-driver torque components and the steering torque sensor signals; converting the pure driver torque estimate into a current hands-off probability score based on a pre-calibrated torque-probability mapping relationship; inputting the current hands-off probability score into a preset probability accumulator for accumulation to obtain a cumulative probability value; and determining and outputting the steering wheel status and confidence level based on a comparison between the cumulative probability value and a preset probability threshold, wherein the confidence level is calculated based on the cumulative probability value. The process involves inputting the steering torque sensor signal, the motor operating information, and the vehicle status information into a driver torque estimation model to obtain a pure driver torque estimate. The driver torque estimation model estimates the non-driver torque components generated by the motor's reaction force and system noise based on the motor operating information and the vehicle status information. The pure driver torque estimate is then calculated based on the non-driver torque components and the steering torque sensor signal. Specifically, this includes: establishing a motor model based on the motor operating information and calculating the motor's reaction force component; establishing a system noise model based on the vehicle status information and calculating the system noise component; and subtracting the motor's reaction force component and the system noise component from the actual steering wheel torque value to obtain the pure driver torque estimate. The step of establishing a motor model based on the motor operating information and calculating the motor reverse force component specifically includes: calculating the basic motor reverse force based on the current, motor torque constant, and transmission efficiency; calculating the speed dynamic compensation value based on the speed; and adding the basic motor reverse force and the speed dynamic compensation value to obtain the motor reverse force component.

2. The method according to claim 1, characterized in that, The step of inputting the current exit probability score into a preset probability accumulator for accumulation to obtain a cumulative probability value specifically includes: obtaining the current accumulated value of the probability accumulator; adding the current exit probability score to the current accumulated value to obtain an initial cumulative probability value; and performing a limiting process on the initial cumulative probability value to obtain the cumulative probability value, so that the cumulative probability value remains within a preset range.

3. The method according to claim 1, characterized in that, The step of determining and outputting the steering wheel state and confidence level based on the comparison result between the cumulative probability value and the preset probability threshold, wherein the confidence level is calculated based on the cumulative probability value, specifically includes: if the cumulative probability value is greater than the first probability threshold, determining that the steering wheel state is off-hand and calculating the off-hand state confidence level, wherein the off-hand state confidence level is equal to the difference between the cumulative probability value and the first probability threshold divided by the difference between the preset maximum probability value and the first probability threshold; if the cumulative probability value is between the first probability threshold and the second probability threshold, determining that the steering wheel state is in a state of hold and calculating the state of hold confidence level, wherein the state of hold confidence level is equal to the difference between the cumulative probability value and the second probability threshold divided by the difference between the first probability threshold and the second probability threshold; if the cumulative probability value is less than the second probability threshold, determining that the steering wheel state is in a state of hands and calculating the in-hand state confidence level, wherein the in-hand state confidence level is equal to the difference between the second probability threshold and the cumulative probability value divided by the difference between the second probability threshold and the preset minimum probability value.

4. The method according to claim 1, characterized in that, Before the step of inputting the steering torque sensor signal, the motor operating information, and the vehicle status information into the driver torque estimation model to obtain a pure driver torque estimate, and the driver torque estimation model estimating the non-driver torque component generated by the motor reverse force and system noise based on the motor operating information and the vehicle status information, and calculating the pure driver torque estimate based on the non-driver torque component and the steering torque sensor signal, the method further includes: performing low-pass filtering preprocessing on the steering torque sensor signal, the motor operating information, and the vehicle status information to obtain preprocessed steering torque sensor signal, motor operating information, and vehicle status information.

5. The method according to claim 3, characterized in that, After determining and outputting the steering wheel state and confidence level based on the comparison result of the cumulative probability value and the preset probability threshold, wherein the confidence level is calculated based on the cumulative probability value, the method further includes: if the steering wheel state is a hands-off state and the hands-off state confidence level exceeds a preset confidence threshold, obtaining current vehicle driving scenario information; determining the scenario risk level based on the current vehicle driving scenario information; and matching the warning strategy corresponding to the scenario risk level from the warning strategy table.

6. A detection system, characterized in that, The detection system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the detection system to perform the method as described in any one of claims 1 to 5.

7. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the detection system, the detection system performs the method as described in any one of claims 1 to 5.

8. A computer program product, characterized in that, When the computer program product is run on the detection system, the detection system performs the method as described in any one of claims 1 to 5.

Images