Confluence point intelligent navigation function degradation judgment method and device

Abstract

The invention discloses a convergence point intelligent navigation function degradation judgment method, and relates to the technical field of intelligent driving, and the method comprises the steps: obtaining the current lane width at the current position of a target vehicle and the preview lane width at a preset preview distance in front of the target vehicle; determining width change information based on the current lane width and the preview lane width; acquiring a left side lane line variance and a right side lane line variance of the target vehicle in the current lane; determining variance information based on the left side lane line variance and the right side lane line variance; acquiring road edge information of a current lane; and when the width change information, the variance information and the road edge information meet a preset degradation condition, controlling an intelligent navigation function of the target vehicle to perform degradation operation. Accurate recognition and function degradation of the confluence point scene are realized, and the driving safety and the system robustness are improved.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to a method and device for determining the downgrade of intelligent navigation function at merging points. Background Technology

[0002] In the field of intelligent driving assistance systems, Intelligent Cruise Assist (ICA) integrates multi-source perception information such as lane lines, road edges, and traffic targets to achieve vehicle centering control within the lane, thereby improving driving safety and comfort. However, in real-world road scenarios, especially at merging points where lanes merge from ramps to main roads, existing technologies typically rely on merging point markers directly output by sensors as the trigger for function degradation. This means that when a merging point is detected, the driver is prompted to take over and disengage lateral control. However, open road testing reveals instability in sensor recognition of merging points, including numerous false detections and missed detections. This leads to frequent false degradation of the ICA function or failure to degrade in a timely manner when necessary, posing potential risks to driving safety. Therefore, a method for determining ICA degradation at merging points is urgently needed to address the aforementioned technical problems. Summary of the Invention

[0003] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solutions, nor is it intended to determine the scope of protection of the claimed technical solutions.

[0004] This application aims to address the technical problem of incorrect or non-incorrect lane detection in intelligent navigation functions at merging points, caused by sensors misdetecting lane lines. The method determines width change information by acquiring the current lane width and the target lane width, determines variance information based on the variance of the left and right lane lines, and combines this with roadside information for comprehensive judgment. When preset degradation conditions are met, the function is downgraded. This achieves accurate identification and function degradation at merging points, improving driving safety and system robustness.

[0005] Firstly, this application provides a method for determining the degradation of intelligent navigation function at merging points, including: Obtain the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; Based on the current lane width and the predicted lane width, determine the width change information; Obtain the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; Variance information is determined based on the variance of the left lane line and the variance of the right lane line; Obtain the curb information of the current lane; When the width change information, the variance information, and the curb information meet the preset downgrade conditions, the intelligent navigation function of the target vehicle is controlled to perform a downgrade operation.

[0006] In some embodiments, obtaining the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle includes: Obtain the left lane line equation and right lane line equation of the target vehicle in the current lane; Based on the left lane line equation and the right lane line equation, determine the current lane width at the current position of the target vehicle; Based on the left lane line equation and the right lane line equation, the width of the pre-aiming lane at a preset pre-aiming distance in front of the target vehicle is determined.

[0007] In some implementations, obtaining the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane includes: Based on the current position of the target vehicle, select multiple aiming points forward along the vehicle trajectory line; Based on the multiple aiming points, determine the left lateral distance sequence between each aiming point and the left lane line and the right lateral distance sequence between each aiming point and the right lane line; The variance of the left lane line is determined based on the left lateral distance sequence; The variance of the right lane line is determined based on the right-side lateral distance sequence.

[0008] In some implementations, the variance information includes variance weight difference and variance confidence level, and determining the variance information based on the variance of the left lane line and the variance of the right lane line includes: Based on the variance of the left lane line and the variance of the right lane line, determine the weight of the left variance and the weight of the right variance; The variance weight difference is determined based on the left-side variance weight and the right-side variance weight; Based on the variance of the left lane line and the variance of the right lane line, the variance confidence level of the lane line with the larger variance is determined.

[0009] In some implementations, before selecting multiple aiming points forward along the vehicle trajectory based on the current position of the target vehicle, the method further includes: Obtain the left longitudinal distance of the identified left lane line and the right longitudinal distance of the identified right lane line; Based on the left longitudinal distance and the right longitudinal distance, the lane line on the side with the longer longitudinal distance is determined as the reference side; Based on the lane line equations on the reference side and the vehicle trajectory lines, the lane lines on the side with shorter longitudinal distances are completed to generate the completed lane lines.

[0010] In some implementations, obtaining the curb information of the current lane includes: Detect whether there are curbs on both sides of the current lane, and determine whether curbs exist; When a curb is detected, the lateral distance between the curb and the lane line on the same side is determined; When the lateral distance is less than a preset distance threshold, the lateral variance weight difference of the roadside is determined based on the roadside and the lane line on the same side. The curb information includes curb presence information, and when the curb is detected, it also includes the lateral distance, and when the lateral distance is less than the preset distance threshold, it also includes the curb lateral variance weight difference.

[0011] In some implementations, the preset degradation condition includes a first sub-condition, a second sub-condition, and a third sub-condition, and the preset degradation condition is determined in the following manner: If the difference between the current lane width and the target lane width in the width change information is greater than a preset width threshold and the duration reaches a preset anti-shake time, it is determined as the first sub-condition. If the variance weight difference in the variance information is greater than a first preset weight threshold and the variance confidence is greater than a preset confidence threshold, it is determined as the second sub-condition. The third sub-condition is defined as follows: one side of the current lane has a road edge, the lateral distance between the road edge and the lane line on the same side is less than a preset distance threshold, the corresponding difference in road edge lateral variance weights is less than a second preset weight threshold, and the other side of the current lane does not have a road edge.

[0012] In some implementations, when the width change information, the variance information, and the curb information meet preset degradation conditions, controlling the intelligent navigation function of the target vehicle to perform a degradation operation includes: When the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information satisfies the third sub-condition, the intelligent navigation function of the target vehicle is controlled to exit the lateral control operation and maintain the longitudinal adaptive cruise operation.

[0013] In some implementations, it also includes: When the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information does not satisfy the third sub-condition, it is determined that a lane line false detection has occurred. Based on the lane line corresponding to the smaller variance between the left lane line variance and the right lane line variance, a single-lane control operation is performed on the target vehicle.

[0014] Secondly, this application proposes a device for determining the degradation of intelligent navigation function at merging points, comprising: The lane width acquisition unit is used to acquire the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; A width change determination unit is used to determine width change information based on the current lane width and the target lane width; The lane variance acquisition unit is used to acquire the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; The variance information determination unit is used to determine variance information based on the variance of the left lane line and the variance of the right lane line; A curb information acquisition unit is used to acquire curb information of the current lane; The function degradation judgment unit is used to control the intelligent navigation function of the target vehicle to perform a degradation operation when the width change information, the variance information and the curb information meet the preset degradation conditions.

[0015] In summary, the merging point intelligent navigation function degradation judgment method provided in this application monitors the dynamic changes in lane width by acquiring the current lane width at the target vehicle's current position and the predicted lane width at a preset predicted distance ahead, providing data support for identifying lane narrowing trends. Based on the width change information determined by the current lane width and the predicted lane width, it can effectively capture the physical characteristics of a reduction in the number of lanes ahead and perceive the existence of potential merging scenarios. By acquiring the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane, it quantitatively analyzes the regularity and fluctuation characteristics of the lane lines on both sides, providing a basis for judging the trend of lane alignment changes. Based on the variance weight difference determined by the variance of the left lane line and the variance of the right lane line, it characterizes the effect of the left and right lane lines on the overall vehicle... The differences in the contribution of lane shape changes are used to identify the dominant side of lane merging. By acquiring the current lane's curb information, road physical structure features are introduced as a verification dimension to provide an environmental reference for judging the authenticity of merging scenarios. When the width change information, variance weight difference, and curb information simultaneously meet the preset degradation conditions, the intelligent navigation function is controlled to perform a degradation operation. This means that the multi-dimensional joint judgment is made by integrating lane width narrowing trends, lane alignment change characteristics, and road physical structure information to accurately identify real merging scenarios, reduce the single dependence on sensor merging point markers, avoid functional degradation due to false detection of merging points, and ensure timely exit of lateral control in real merging scenarios, thereby improving the operational safety of the intelligent navigation function in complex road environments. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a method for determining the degradation of intelligent navigation function at merging points, provided in an embodiment of this application; Figure 2 This is a schematic diagram of a merging point intelligent navigation function degradation judgment device provided in an embodiment of this application. Detailed Implementation

[0017] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0018] This application relates to the field of intelligent driving technology, specifically to the assessment of functional degradation in vehicles equipped with intelligent cruise assist in merging scenarios. Merging scenarios mainly refer to road sections where the number of lanes changes, such as ramps merging into main roads or multiple lanes decreasing to fewer lanes. In such scenarios, due to lane merging ahead, if the vehicle does not have the ability to actively change lanes, it needs to promptly disengage lateral control to avoid collisions with guardrails or other road edge facilities. The technical solution of this application can be applied to various vehicle types, including passenger cars and commercial vehicles equipped with advanced driver assistance systems, and is particularly suitable for low- to mid-range vehicles that do not have high-cost urban intelligent cruise assist functions. Algorithm optimization improves functional safety and system robustness in merging scenarios.

[0019] To clearly describe the technical solution of this application, some of the features involved in this application are described below.

[0020] The intelligent navigation assistance described in this application refers to a driving assistance function that can simultaneously provide the driver with longitudinal speed control and lateral steering wheel control. It uses sensors to perceive information such as lane lines, road edges, and traffic targets to enable the vehicle to stay centered within the lane.

[0021] The lane line equation in this application refers to a polynomial equation obtained by mathematically modeling discrete lane line points identified by sensors through a curve fitting algorithm, which is used to describe the geometric shape and direction of the lane line in the vehicle coordinate system.

[0022] The merging point in this application refers to a road section where multiple lanes merge into fewer lanes, with typical scenarios including ramps merging into the main road.

[0023] The current lane width in this application refers to the lateral distance between the left and right lane lines of this lane at the current position of the target vehicle.

[0024] The aiming distance in this application refers to the longitudinal distance for predictive perception of the road ahead, used to determine the geometric change trend of the road ahead in advance.

[0025] The pre-aiming lane width in this application refers to the lateral distance between the lane lines on the left and right sides of this lane at a preset pre-aiming distance in front of the target vehicle.

[0026] The vehicle trajectory line in this application refers to the center line of the future driving path predicted based on the current motion state of the vehicle output by the sensors.

[0027] The lane line variance in this application refers to the variance of the sequence of lateral distance values ​​between the lane line and a certain side, measured at multiple preview points along the vehicle trajectory line, and is used to quantify the degree of morphological fluctuation of the lane line on that side.

[0028] The variance information in this application is a set of parameters used to characterize the trend of lane alignment changes, determined based on the variance of the left lane line and the variance of the right lane line.

[0029] The variance weight in this application refers to the proportion obtained after normalizing the variances of the lane lines on the left and right sides, which is used to characterize the contribution of changes in lane lines on one side to the overall lane shape changes.

[0030] The variance weight difference in this application refers to the absolute value of the difference between the variance weights of the left and right sides, which is used to determine whether lane line changes are dominated by one side.

[0031] The lane line confidence level in this application refers to the reliability assessment value of the sensor's recognition of the authenticity and accuracy of the lane lines it identifies.

[0032] In this application, "roadside" refers to the physical boundary structure on both sides of a road, including curbs on urban roads, guardrails on highways or urban expressways, crash barriers, and other road boundary facilities.

[0033] The curb information in this application refers to a set of data related to the curb, including whether the curb exists, the lateral distance between the curb and the lane line on the same side, and the variance weight difference calculated based on the relationship between the curb and the lane line.

[0034] The preset downgrade conditions in this application refer to a multi-dimensional combination of conditions pre-set to determine whether to trigger the downgrade of the intelligent navigation function.

[0035] The downgrade operation in this application refers to the intelligent navigation function disengaging from lateral control, retaining only longitudinal adaptive cruise control, and returning lateral control of the vehicle to the driver.

[0036] The single-lane control operation described in this application refers to an operation mode in which, when lane line misdetection is determined to have occurred, the intelligent navigation function temporarily performs lateral control based solely on the reliable lane line on the side with the smaller variance, provided that the width and variance change conditions are met but the roadside information does not meet the preset degradation conditions, until the dual-lane control conditions are met again. This operation mode aims to maintain basic lateral control capabilities in lane line misdetection scenarios and avoid functional degradation due to misdetection.

[0037] Please see Figure 1 The above is a flowchart illustrating a method for determining the downgrade of intelligent navigation at merging points, provided in an embodiment of this application. The method includes: S110, Obtain the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; For example, the left and right lane line information of the lane where the target vehicle is located is obtained through vehicle-mounted perception sensors. Based on the left and right lane line information, the lateral position of the left and right lane lines at the current position of the target vehicle is determined, and the current lane width is calculated. At the same time, the lateral position of the left and right lane lines at the preset aiming distance forward along the direction of travel of the target vehicle is calculated based on the geometric equation of the left and right lane lines, and the aiming lane width is calculated.

[0038] S120. Determine width change information based on the current lane width and the predicted lane width; For example, by comparing the current lane width at the current position of the target vehicle with the pre-aimed lane width at a preset pre-aimed distance ahead, the longitudinal trend of lane width change is analyzed to determine the width change information that can reflect whether the lane is narrowing. This information is used to characterize whether there is a potential merging point scenario on the road ahead.

[0039] S130. Obtain the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; For example, the variance of the left lane line is calculated by the lateral offset of the left lane line at multiple longitudinal positions, and the variance of the right lane line is calculated by the lateral offset of the right lane line at multiple longitudinal positions. The variances of the left and right lane lines are used to characterize the degree of morphological fluctuation of the left and right lane lines in the direction of vehicle travel, respectively.

[0040] S140. Determine variance information based on the variance of the left lane line and the variance of the right lane line; For example, by weighting the variances of the left lane line and the right lane line, the variance weights of the left and right lane lines are obtained, and the variance weight difference is determined based on the variance weights of the left and right lane lines. At the same time, based on the comparison results of the variances of the left and right lane lines, the lane line with the larger variance is determined, and the confidence level corresponding to the lane line on that side is obtained as the variance confidence level. The variance information includes the variance weight difference and the variance confidence level.

[0041] S150, Obtain the curb information of the current lane; For example, the presence of curbs on both sides of the lane where the target vehicle is located is detected by the vehicle-mounted perception sensor, and the presence status of the curbs, the lateral distance between the curbs and the lane lines on the same side, and the curb lateral variance weight difference calculated based on the curbs and the lane lines on the same side when the lateral distance is less than a preset distance threshold are obtained. The curb presence information, the lateral distance, and the curb lateral variance weight difference are used as the curb information of the current lane.

[0042] S160. When the width change information, variance information, and curb information meet the preset downgrade conditions, control the intelligent navigation function of the target vehicle to perform a downgrade operation.

[0043] For example, when the width change information, variance information and curb information obtained through the aforementioned steps meet the preset degradation conditions, it is determined that the current vehicle is in a real merging scenario and has exceeded the lateral control capability range. Then, the intelligent navigation function is controlled to perform a degradation operation, that is, to exit lateral control and retain only longitudinal adaptive cruise control, while reminding the driver to take over the steering wheel, thereby effectively avoiding driving safety risks in merging scenarios.

[0044] In summary, this application's embodiments monitor lane width changes by acquiring the current lane width at the target vehicle's current position and the predicted lane width at a preset forward aiming distance, providing a forward-looking data foundation for identifying lane narrowing trends. Based on the current lane width and the predicted lane width, width change information is determined, capturing the physical characteristics of a decrease in the number of lanes ahead and sensing the existence of potential merging scenarios. By acquiring the variance of the left and right lane lines of the target vehicle in the current lane, the regularity and fluctuation characteristics of the lane lines on both sides are quantitatively analyzed, providing a quantitative basis for judging lane alignment change trends. Based on the variance of the left and right lane lines, variance information is determined, which can characterize the difference in contribution of the left and right lane lines to the overall lane shape change, thereby identifying the dominant side of lane merging and avoiding reliance solely on a single lane. Misjudgments caused by lateral changes are addressed by acquiring the current lane's curb information and introducing the road's physical boundary as a verification dimension. This provides an environmental reference for judging the authenticity of merging scenarios, distinguishing between real lane merging and false changes caused by sensor misdetection of lane lines. When width change information, variance information, and curb information simultaneously meet preset degradation conditions, the intelligent navigation function is controlled to perform a degradation operation. This means that a multi-dimensional joint judgment is made by integrating lane width narrowing trends, lane alignment change characteristics, and road physical structure information. This can accurately identify real merging scenarios, reduce the reliance on sensor merging point markers, avoid functional degradation due to merging point misdetection, and ensure timely exit of lateral control in real merging scenarios. This improves the operational safety and system robustness of the intelligent navigation function in complex road environments.

[0045] In some instances, obtaining the current lane width at the target vehicle's current position and the pre-aimed lane width at a preset pre-aimed distance in front of the target vehicle includes: Obtain the left and right lane line equations of the target vehicle in the current lane; Determine the current lane width at the current position of the target vehicle based on the left lane line equation and the right lane line equation. Based on the left lane line equation and the right lane line equation, the width of the pre-aiming lane at a preset pre-aiming distance in front of the target vehicle is determined.

[0046] For example, the target vehicle uses onboard perception sensors to collect discrete point information of the lane lines on both the left and right sides of its lane in real time. Each discrete point includes the longitudinal distance and lateral offset relative to the vehicle coordinate system. To transform the discrete points into a continuous lane geometry description, a curve fitting algorithm is used to fit the discrete points on the left and right sides respectively, generating the left lane line equation and the right lane line equation. In the field of intelligent driving, a cubic polynomial is used as the mathematical model for the lane lines, and its general form is... y =C0+C1· x +C2· x2 +C3· x 3 ,in, x Indicates vertical distance. y This indicates lateral offset, with coefficient C0 being the intercept term, representing the lane line at... x At the lateral position of =0, C1 is the linear coefficient, representing the initial slope of the lane line (the linear part of the curvature); C2 is the quadratic coefficient, representing the rate of change of curvature (the second derivative), which determines the degree of curvature; C3 is the cubic coefficient, representing the higher-order curvature change, used to fit more complex curves (such as S-curves). The fitted left and right lane line equations characterize the direction and curvature of the two lane lines in the vehicle coordinate system.

[0047] Based on the lane line equations described above, the current lane width at the target vehicle's current position is determined by... x Substituting 0 into the equations for the left and right lane lines respectively, we obtain the lateral offset of the left lane line at the current position as: y left (0)=C 0,left The right lane line is offset laterally at its current position. y right (0)=C 0,right The sum of the two values ​​is the current lane width L. current = y left (0)+ y right (0). For the pre-aimed lane width at the preset pre-aimed distance ahead, substitute the preset pre-aimed distance value d (e.g., 30 meters) into the lane line equations on both sides to obtain the left lateral offset at the point d meters ahead. y left (d) and right lateral offset y right (d), the sum of which is the target lane width L preview = y left (30)+ y right (30). The preset aiming distance is a value pre-calibrated based on the vehicle dynamics response time and system control cycle, used to balance the lead time of the aiming and the stability of the lane line fitting. By continuously comparing the current lane width with the aiming lane width, the longitudinal trend of lane width changes in real time is perceived, providing a quantitative basis for identifying whether there is a merging scenario where the lane narrows ahead.

[0048] In summary, this embodiment of the application obtains the left and right lane line equations of the target vehicle in the current lane, transforming discrete lane line perception information into a continuous mathematical description, thereby achieving lane geometry modeling. Based on the lane line equations, it determines the current lane width at the target vehicle's current position and the pre-aimed lane width at a preset forward aiming distance. On the one hand, it monitors the lateral passage space at the vehicle's location in real time; on the other hand, it can perceive the changing trend of the lane width ahead in advance, achieving proactive identification of potential merging scenarios. By comparing the current lane width with the pre-aimed lane width, it quantifies the degree of lane width narrowing, providing input information for subsequent function degradation decisions, avoiding judgment lag caused by relying solely on instantaneous perception, and improving the response speed and judgment accuracy of the intelligent navigation function in merging point scenarios.

[0049] In some instances, the variance of the target vehicle's left lane line and right lane line in the current lane is obtained, including: Based on the target vehicle's current position, select multiple aiming points forward along the vehicle's trajectory line; Based on multiple preview points, determine the left lateral distance sequence between each preview point and the left lane line and the right lateral distance sequence between each preview point and the right lane line; Determine the variance of the left lane line based on the left lateral distance sequence; The variance of the right lane line is determined based on the right-side lateral distance sequence.

[0050] For example, the vehicle trajectory line is the centerline of the future driving path predicted by onboard sensors based on the vehicle's current motion state. It reflects the vehicle's expected driving trajectory over the next distance. To obtain the continuous longitudinal variation characteristics of the lane line, starting from the vehicle's current position, a preset number of pre-aiming points are selected at equal intervals along the vehicle trajectory line. This preset number can be calibrated based on computing resources, for example, selecting 10 pre-aiming points. Each pre-aiming point corresponds to a specific longitudinal distance, and its lateral position is determined by the coordinates of the vehicle trajectory line at that longitudinal distance. This constructs a set of discrete sampling positions in front of the vehicle, providing a spatial reference for subsequent measurements of the relative distance to the lane line.

[0051] For each preview point, extend the horizontal line to the left to intersect the left lane line, and calculate the lateral distance between the preview point and the intersection point. This distance represents the lateral offset of the left lane line relative to the vehicle's trajectory at that preview point. Similarly, extend the horizontal line to the right to intersect the right lane line, obtaining the lateral distance of the right lane line. If a lane line on one side is missing at a certain longitudinal position due to recognition length limitations, the lateral distance at the corresponding position is calculated using the lane line equation after completion processing, ensuring that each preview point obtains complete distance values ​​for both the left and right sides. Arrange the left lateral distances corresponding to all preview points in longitudinal order to form a left lateral distance sequence; arrange the right lateral distances corresponding to all preview points in longitudinal order to form a right lateral distance sequence. These two sequences respectively depict the positional change trends of the left and right lane lines relative to the vehicle's trajectory within a certain distance in front of the vehicle.

[0052] Variance is a statistical measure of the dispersion of a set of data. It is calculated by dividing the sum of the squares of the differences between each data point and the mean of the sequence by the number of data points. The variance of the left lane line is calculated by examining the left lateral distance sequence; the variance of the right lane line is calculated by examining the right lateral distance sequence. The magnitude of the left lane line variance reflects the longitudinal fluctuation of the left lane line: if the left lane line is straight and parallel to the vehicle trajectory, the differences between values ​​in the left lateral distance sequence are small, and the variance approaches zero; if the left lane line gradually curves towards the vehicle trajectory (e.g., on the side where the merging point is about to disappear), the left lateral distance sequence shows an increasing or decreasing trend, and the variance is larger. Similarly, the right lane line variance characterizes the morphological fluctuation of the right lane line. By calculating the variance, the geometric characteristics of the lane lines are quantified into comparable values, providing basic data for analyzing lane line change trends.

[0053] In summary, this embodiment selects multiple aiming points along the vehicle trajectory line, which is equivalent to setting observation positions at different distances in front of the vehicle to capture the continuous longitudinal changes of the lane lines. At each aiming point, the lateral distances of the left and right lane lines are measured, and these distance values ​​are recorded sequentially, forming two sets of distance sequences for the left and right sides. This allows the geometric shape of the lane lines to be transformed into specific numerical values. The variance of these two sets of distance sequences is calculated. The magnitude of the variance reflects the severity of lane line fluctuations: a smaller variance indicates a straighter and more stable lane line, while a larger variance indicates that the lane line is bending or converging. Through this series of processes, the originally complex lane line changes are refined into quantifiable data indicators, providing a basis for determining which side is the lane merging side and whether it is a real merging scenario. This makes the intelligent navigation function's degradation judgment at the merging point more accurate and reliable.

[0054] In some instances, variance information includes variance weighted differences and variance confidence scores. Variance information is determined based on the variances of the left and right lane lines, including: Based on the variances of the left and right lane lines, determine the weights of the left and right variances. The variance weight difference is determined based on the left-side variance weight and the right-side variance weight; Based on the variances of the left and right lane lines, determine the variance confidence level of the lane line with the larger variance.

[0055] For example, the variances of the left and right lane lines quantify the longitudinal fluctuations of the left and right lane lines, respectively. However, the absolute values ​​of these two variances cannot directly reflect the relative importance of the changes on both sides to the overall lane shape. Therefore, this embodiment normalizes the variances of the left and right sides and calculates the weights of the left and right variances. The formula for calculating the weight of the left variance is the left lane line variance divided by the sum of the variances of the left and right lane lines, while the weight of the right variance is the right lane line variance divided by the sum of the variances of both sides. Through normalization, the variances of both sides are transformed into relative proportional values ​​ranging from 0 to 1, and the sum of the weights of the left and right variances is 1. The larger the weight of the left variance, the stronger the dominant role of the fluctuations of the left lane line in the overall lane change; the larger the weight of the right variance, the more significant the change in the right lane line.

[0056] Based on the calculated left and right variance weights, the variance weight difference is determined. The variance weight difference is the absolute value of the difference between the left and right variance weights, and its value also ranges from 0 to 1. The variance weight difference quantifies the asymmetry in the contribution of the left and right sides to lane shape changes: when the variance weight difference approaches 0, it indicates that the fluctuation of the lane lines on both sides is roughly equal, and the lane width may remain essentially unchanged; when the variance weight difference is large (e.g., greater than a preset first weight threshold), it indicates that the fluctuation of the lane lines on one side is significantly greater than that on the other side, and this side is the dominant side where the lane is narrowing or merging. Through the variance weight difference, the direction of lane merging can be accurately identified, providing a lateral judgment basis for function degradation decisions.

[0057] While calculating the variance weight difference, it is also necessary to consider the quality and reliability of the lane lines used for judgment. Sensors typically output confidence values ​​when identifying lane lines, characterizing the authenticity and accuracy of the lane line recognition result. This embodiment determines the lane line with the larger variance based on a comparison of the variances of the left and right lane lines, as this side is the main contributor to lane width changes, and its recognition quality directly affects the accuracy of merging scenario judgment. Therefore, the confidence score of this side's lane line is obtained from the sensor as the variance confidence score. Only when this variance confidence score is higher than a preset confidence threshold (e.g., 0.3) is the lane line on the side with the larger variance considered reliable, and its change trend truly reflects the road geometry, thus avoiding misjudgments caused by lane line false detections or low-quality recognition. If the confidence score is low, it may mean that the change on this side is caused by false detection, such as misidentifying road stains, shadows, or guardrails as lane lines.

[0058] In summary, this embodiment eliminates the influence of the absolute difference in variance between the two sides by normalizing the variances of the left and right lane lines into variance weights, making the contribution of changes on both sides comparable. Further calculation of the variance weight difference can quantify the degree of asymmetry of fluctuations on the left and right sides, thereby identifying the dominant side of lane merging and providing lateral positioning basis for downgrade decision-making. At the same time, the confidence level of the lane line with the larger variance is introduced as the variance confidence level to verify the quality of the lane lines used for judgment, ensuring that the change characteristics are adopted only when the lane line recognition is reliable, effectively eliminating false merging signals introduced by lane line misdetection or low-quality perception.

[0059] In some instances, before selecting multiple aiming points forward along the vehicle's trajectory based on the target vehicle's current position, the process also includes: Obtain the left longitudinal distance of the identified left lane line and the right longitudinal distance of the identified right lane line; Based on the left longitudinal distance and the right longitudinal distance, the lane line on the side with the longer longitudinal distance is determined as the reference side; Based on the lane line equations and vehicle trajectory lines on the reference side, the lane lines on the side with shorter longitudinal distances are supplemented to generate the supplemented lane lines.

[0060] For example, before selecting multiple preview points along the vehicle's trajectory line based on the target vehicle's current position, the completeness of left and right lane line recognition is first evaluated. Because the recognition capabilities of onboard perception sensors for left and right lane lines may vary in complex road environments—for instance, a lane line disappearing near a merging point may only be stably recognized within a short longitudinal distance due to wear, curvature changes, or occlusion—while the lane line continuing on the other side may be recognized at a greater distance—the left longitudinal distance of the recognized left lane line and the right longitudinal distance of the recognized right lane line, output by the sensors, are obtained. The longitudinal distance refers to the distance extending forward along the lane line from the vehicle's current position to the last valid recognition point. By comparing the left and right longitudinal distances, the lane line with the longer longitudinal distance is designated as the baseline side, and the other side is the side to be supplemented. This step aims to ensure that subsequent calculations can be expanded based on more complete perception data.

[0061] After determining the reference side, the lane line equations of the reference side are used as a basis, combined with the vehicle trajectory line, to complete the lane line on the side with the shorter longitudinal distance. The vehicle trajectory line is the centerline of the future driving path predicted by onboard sensors based on the current vehicle motion state; it reflects the spatial position the vehicle is about to pass through. The specific completion method is as follows: for the longitudinal position where the lane line to be completed exceeds its actual recognition length, based on the direction of the reference side lane line at that longitudinal distance and the geometric constraints of the vehicle trajectory line, the lateral position of the lane line to be completed at that position is calculated, thereby generating a continuous theoretical lane line, so that the lane line to be completed extends to the same length as the reference side in the longitudinal range.

[0062] After imputation, both the left and right lane lines have complete lane line equations within the range from the vehicle's current position to the preset maximum aiming distance. This ensures that when selecting multiple aiming points along the vehicle's trajectory in subsequent steps, each aiming point can obtain the lateral distance value intersecting with the left and right lane lines. In this way, the left and right lateral distance sequences have the same number of data points and cover the same longitudinal range, providing a comparable basis for calculating the variance of the left and right lane lines. If the original recognition data is used directly without imputation, the shorter side will lack distance values ​​for several subsequent aiming points, causing the variance calculation results to fail to accurately reflect the overall fluctuation trend of the lane line on that side, and even making effective two-sided comparisons impossible, thus affecting the accurate judgment of merging point scenarios.

[0063] In summary, this application first uses longitudinal distance to filter out a more reliable reference side, avoiding the problem of insufficient recognition length on one side preventing a complete comparison. Then, it uses the lane line equation and vehicle trajectory line of the reference side to reasonably extend the shorter side virtually, so that the lane lines on both sides have a complete geometric description within the same longitudinal range. Based on the complete distance sequence after completion, the variance of the left and right lane lines is calculated, which can truly reflect the fluctuation characteristics of each lane line from near to far, thereby identifying the lane merging side and its changing trend. This process effectively eliminates the comparison bias caused by the inconsistency of sensor recognition range, improves the reliability of variance information, and enables subsequent merging point judgment based on variance weight difference and roadside information to be based on a more accurate data foundation, improving the robustness and accuracy of intelligent navigation function in downgraded decision-making in merging scenarios.

[0064] In some instances, the curb information for the current lane is obtained, including: Detect whether there are curbs on both sides of the current lane and confirm the presence of curbs; When a curb is detected, determine the lateral distance between the curb and the lane line on the same side; When the lateral distance is less than a preset distance threshold, the difference in lateral variance weights of the roadside is determined based on the roadside and the lane lines on the same side. The curb information includes curb presence information, as well as lateral distance when a curb is detected, and curb lateral variance weight difference when the lateral distance is less than a preset distance threshold.

[0065] For example, when acquiring curb information for the current lane, the vehicle's onboard perception sensors first detect the physical boundaries of the road on both sides of the lane where the target vehicle is located to determine whether a curb exists. A curb refers to the concrete or stone boundary structure at the edge of a road; in urban roads, it manifests as a curb curb, and in highways or urban expressways, it manifests as a guardrail or crash barrier. The detection process is based on the point cloud or image recognition results output by the sensors to determine whether there are targets on the left and right sides that match the geometric features and physical attributes of a curb. If a curb is detected on a certain side, the state of curb presence on that side is recorded as part of the curb presence information; if no curb is detected, the curb presence information only includes an indication of no curb, and no further curb-related calculations are performed on that side.

[0066] When a curb is detected on one side, the lateral distance between that curb and the lane line on the same side is further determined. This lateral distance refers to the vertical distance from the curb to the lane line on the same side in the current vehicle coordinate system, which can be calculated by fusing the curb position information and lane line equations output by the sensors. Specifically, based on the established left or right lane line equations, the lateral coordinates of the lane line on the same side at the current longitudinal position are obtained, and the lateral coordinates of the curb at the same longitudinal position are also obtained. The absolute value of the difference between the two is the lateral distance between the curb and the lane line on that side. This lateral distance is used to determine the proximity between the curb and the lane line. If they are close, it may mean that the lane line is adjacent to the physical boundary of the road, which is one of the typical characteristics of merging point scenarios. A preset distance threshold (e.g., 1 meter) is used to define the criteria for determining proximity. When the lateral distance is less than this threshold, the curb and the lane line are considered to be in a spatial proximity state, and further analysis of their relative changes is required.

[0067] If the lateral distance is less than a preset distance threshold, the difference in lateral variance weights between the roadside and the lane lines on the same side is calculated. This calculation involves selecting multiple aiming points along the vehicle's trajectory in front of the vehicle. Each aiming point corresponds to a longitudinal position. The lateral distance between the roadside and the vehicle's trajectory at each aiming point is measured, forming a sequence of roadside lateral distance values. Simultaneously, the lateral distance between the lane lines on the same side and the vehicle's trajectory at the same aiming point is measured, forming a sequence of lane line lateral distance values. The variances of these two distance value sequences are calculated to obtain the roadside lateral variance and the lane line lateral variance. These variances are then normalized, and the weights for the roadside lateral variance and the lane line lateral variance are calculated. The roadside lateral variance weight is calculated by dividing the roadside lateral variance by the sum of the two variances, and the lane line lateral variance weight is calculated similarly. The absolute value of the difference between the roadside lateral variance weight and the lane line lateral variance weight is determined as the roadside lateral variance weight difference. This difference quantifies the consistency of longitudinal changes between the curb and lane lines: if they maintain a relatively stable parallel relationship, the fluctuation trends of the curb lateral distance value series and the lane line lateral distance value series are similar, their variance weights are relatively close, and the difference is small; if the lane line is a false detection (e.g., misidentifying curb shadows or guardrails as lane lines), the lane line lateral distance value series may deviate significantly from the curb lateral distance value series, resulting in a larger difference in variance weights. Therefore, the difference in curb lateral variance weights can be used as a basis for judging whether lane lines truly reflect road boundaries.

[0068] In summary, this application's embodiments, through the aforementioned process of acquiring curb information, introduce the road's physical structure as a verification dimension to evaluate the lane line perception results. First, the presence of the curb is detected and recorded, providing environmental background features for determining merging point scenarios, as real merging points are often accompanied by guardrails or curbs on one side. Second, by measuring the lateral distance between the curb and the lane line on the same side, candidate sides of the lane line that are close to the physical boundary are selected. Spatial proximity is used as a prerequisite for further analysis, avoiding unnecessary calculations for lane lines far from the curb and improving processing efficiency. Third, based on spatial proximity, the lateral variance weight difference of the curb is calculated, and the longitudinal change consistency is used to test whether the relative relationship between the lane line and the curb is stable, thereby effectively distinguishing between real lane lines (extending parallel to the curb) and falsely detected lane lines (such as misidentifying guardrails as lane lines, leading to inconsistent change trends). This series of processing steps not only provides the physical background of the scene but also serves as a quantitative tool for verifying the authenticity of lane lines. It provides a reliable verification basis for merging width change information and variance information to make merging point downgrade judgments, which helps to eliminate false merging signals caused by sensor misdetection and improves the robustness and safety of intelligent navigation functions in complex road environments.

[0069] In some instances, the preset degradation conditions include a first sub-condition, a second sub-condition, and a third sub-condition. The preset degradation conditions are determined in the following way: If the difference between the current lane width and the target lane width in the width change information is greater than the preset width threshold and the duration reaches the preset anti-shake time, it is determined as the first sub-condition. If the variance weight difference in the variance information is greater than the first preset weight threshold and the variance confidence is greater than the preset confidence threshold, it is determined as the second sub-condition. The third sub-condition is defined as follows: if there is a roadside on one side of the current lane in the roadside information, and the lateral distance between the roadside and the lane line on the same side is less than a preset distance threshold, and the corresponding difference in the roadside lateral variance weight is less than a second preset weight threshold, and there is no roadside on the other side of the current lane.

[0070] For example, in the process of constructing the preset degradation conditions, the first sub-condition is judged based on the difference between the current lane width and the target lane width in the width change information. When the difference is greater than the preset width threshold and the duration reaches the preset anti-shake time, the first sub-condition is considered to be true. Here, the preset width threshold is used to quantify the degree of lane width narrowing. Only when the narrowing exceeds this threshold is it considered that there may be physical features of lane merging ahead, rather than slight width fluctuations caused by natural changes in road curvature. The preset anti-shake time is used to filter out accidental width fluctuations caused by sensor noise or instantaneous interference. It requires that the width narrowing state remain stable for a certain period of time to ensure that the detected lane narrowing trend has continuity and reliability.

[0071] The second sub-condition is based on variance information, requiring the variance weight difference to be greater than the first preset weight threshold, and the variance confidence score of the lane line on the side with the larger variance to be greater than the preset confidence score threshold. The magnitude of the variance weight difference directly reflects the asymmetry in the contribution of the left and right lane lines to the overall lane shape change. When this difference exceeds the first preset weight threshold, it indicates that the fluctuation of one lane line is significantly greater than that of the other, and this side is the dominant side of lane merging. The variance confidence score is used to verify the recognition quality of the dominant side lane line. Only when the confidence score of this side lane line is higher than the preset confidence score threshold can it be confirmed that its fluctuation characteristics truly reflect the road geometry changes, rather than false signals introduced by false detections or low-quality perception.

[0072] The third sub-condition integrates multiple features from the curb information, requiring that a curb exists on one side of the current lane, and the lateral distance between this curb and the lane line on the same side is less than a preset distance threshold. Simultaneously, the difference in lateral variance weights of the curb on this side is less than a second preset weight threshold, while the other side of the current lane has no curb. The spatial proximity of the curb and lane line vividly portrays the typical characteristic of lane lines being close to the physical boundary of the road in the merging point area. The fact that the difference in lateral variance weights of the curb is less than the second preset weight threshold further verifies that the longitudinal trends of the curb and lane line are consistent, meaning they maintain a relatively stable parallel relationship within a certain distance in front of the vehicle. This eliminates the possibility of misidentifying curb shadows or guardrails as lane lines. If a lane line is falsely detected, its fluctuation trend often differs significantly from that of the actual curb, resulting in a large difference in variance weights. The constraint of no curb on the other side emphasizes the characteristic that the merging point has a physical boundary on only one side, avoiding false triggering in sections with curbs in both directions.

[0073] In summary, the three sub-conditions described in this application embodiment constitute a judgment framework for preset degradation conditions from different dimensions. The first sub-condition captures the physical characteristics of the merging point through the width change trend, ensuring early detection of lane narrowing. The second sub-condition identifies the dominant merging side through lane alignment change analysis and eliminates false detection interference by using confidence level. The third sub-condition introduces the road physical structure as a verification benchmark, utilizing multiple checks such as roadside existence, spatial proximity, and consistency of change to further confirm the authenticity of the merging scenario. By comprehensively considering lane geometric changes, lane alignment features, and road environment information for multi-dimensional joint judgment, the reliance on a single sensor merging point marker is effectively reduced. This avoids functional degradation due to false lane line detection while ensuring timely exit from lateral control in real merging scenarios, improving the operational safety and system robustness of the intelligent navigation function in complex road environments.

[0074] In some instances, when the width variation information, variance information, and curb information meet preset degradation conditions, the intelligent navigation function of the target vehicle is controlled to perform a degradation operation, including: When the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information satisfies the third sub-condition, the intelligent navigation function of the target vehicle exits the lateral control operation and maintains the longitudinal adaptive cruise operation.

[0075] For example, when the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information satisfies the third sub-condition, it is determined that the current road scene is a real merging point and has exceeded the lateral control capability range of the intelligent navigation function.

[0076] Specifically, in the width change information, if the difference between the current lane width and the target lane width is greater than the preset width threshold and continues to reach the preset anti-shake time, it indicates that the lane ahead is showing a stable and significant narrowing trend, possessing the physical characteristics of lane merging. In the variance information, if the variance weight difference is greater than the first preset weight threshold, it indicates that there is a significant asymmetry in the fluctuation of the lane lines on the left and right sides. The change of one lane line dominates the narrowing of the lane width. At the same time, the variance confidence of the lane line on the side with larger variance is higher than the preset confidence threshold, confirming that the recognition quality of the lane line on that side is reliable, and its change trend truly reflects the road geometry. In the curb information, if there is a curb on one side of the current lane and the lateral distance between the curb and the lane line on the same side is less than the preset distance threshold, it indicates that the lane line is close to the physical boundary, which meets the typical environmental characteristics of a merging point. The variance weight difference of the curb side on that side is less than the second preset weight threshold, further indicating that the curb and the lane line maintain a stable parallel relationship in the longitudinal direction, eliminating the possibility of misdetecting the curb shadow or guardrail as a lane line. At the same time, the absence of a curb on the other side strengthens the scenario attribute of unilateral merging.

[0077] When the above three sub-conditions are met simultaneously, the system integrates lane width changes, lane alignment fluctuations, and road physical structure information to confirm that the vehicle is at a real merging point and about to enter the lane line disappearance area. Then, the intelligent navigation function is deactivated from lateral control, retaining only longitudinal adaptive cruise control. At the same time, the system reminds the driver to take over the steering wheel through the human-machine interface, returning lateral control to the driver in time before it exceeds the vehicle's control capabilities, thus avoiding the risk of the vehicle deviating from the lane or colliding with the guardrail due to lane line merging.

[0078] In some instances, it also includes: When the width change information meets the first sub-condition, the variance information meets the second sub-condition, and the curb information does not meet the third sub-condition, it is determined that a lane line false detection has occurred. The target vehicle is controlled by a single lane based on the lane line corresponding to the smaller variance between the left and right lane line variances.

[0079] For example, when the width change information satisfies the first sub-condition and the variance information satisfies the second sub-condition, but the curb information does not satisfy the third sub-condition, it is determined that a lane line false detection has occurred in the current scenario. The fulfillment of the first sub-condition indicates that the lane ahead has a narrowing trend, which is a physical characteristic of a merging scenario. The fulfillment of the second sub-condition indicates that there is a significant asymmetry in the fluctuation of the left and right lane lines, and the lane line on the side with greater fluctuation has a higher confidence level, meaning that the lane line change on that side dominates the width narrowing. However, the third sub-condition requires that there must be a curb on the merging side, and that the curb and lane line maintain spatial proximity and longitudinal consistency, while the other side has no curb, which is an environmental constraint of the actual merging point. When the curb information is not satisfied, it means that although the sensor detected lane narrowing and unilateral line changes, there is a lack of physical boundary evidence, or the relationship between the curb and lane line is abnormal, such as the curb not existing, the curb being too far from the lane line, or the curb lateral variance weight difference being too large, indicating that the lane line change and the curb change trend are inconsistent. These situations often arise because sensors misidentify non-lane line targets such as road shadows, guardrails, and stains as lane lines, creating a false impression of lane merging at the data level. If the function is downgraded directly in this situation, the false detections could lead to unnecessary driver intervention, impacting the user experience. If dual-lane control continues to be used, the falsely detected lane lines could mislead the vehicle into deviating from its actual lane, posing a safety risk.

[0080] To address the aforementioned false detection scenarios, a single-lane control strategy is employed. Specifically, upon confirming a lane line false detection, the variances of the left and right lane lines are compared, and the lane line with the smaller variance is selected as the reference for lateral control. A smaller variance indicates lower fluctuation in the lateral distance sequence corresponding to that lane line, suggesting that it remains relatively straight longitudinally and parallel to the vehicle's trajectory, resulting in more stable and reliable perception. In merging scenarios, the lane line with the smaller variance is typically a real, non-merging lane line, such as a lane line extending from the main road. The desired lateral control amount is calculated based on the equation of that lane line, controlling the vehicle to maintain a predetermined distance from it, thus achieving a single-lane tracking lateral control mode. This mode continues until the dual-lane control conditions are met again (e.g., roadside information is restored or the false detection disappears), at which point regular dual-lane centering control resumes. In this way, even in the event of lane line false detection, the vehicle can still maintain basic lateral guidance based on a relatively reliable single-lane line, avoiding complete loss of function or dangerous maneuvers due to false detection.

[0081] In summary, this application introduces variance comparison as a screening criterion to automatically identify the less volatile and more reliable lane line between the two sides. This ensures that a relatively reliable reference benchmark can still be selected when false detection occurs, avoiding control deviations caused by using falsely detected lane lines. The single-lane control mode retains lateral assist capability, allowing the vehicle to continue driving along the reliable lane line. This prevents unnecessarily degrading the vehicle due to false detection (i.e., unnecessarily disengaging lateral control) and ensures that the vehicle still has basic lane-keeping function when there is a lack of or anomalies in the curb information at the actual merging point, improving its adaptability to complex perception environments. When width changes and variance characteristics indicate a potential merging, but the curb verification fails, the system does not simply abandon the judgment but enters an emergency control state. Regular control is resumed after the environmental information becomes clear, achieving a balance between safety and usability and enhancing the overall robustness of the intelligent navigation function.

[0082] Please see Figure 2 The diagram below illustrates a merging point intelligent navigation function degradation judgment device provided in this application embodiment, comprising: Lane width acquisition unit 21 is used to acquire the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; Width change determination unit 22 is used to determine width change information based on the current lane width and the target lane width; Lane variance acquisition unit 23 is used to acquire the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; The variance information determination unit 24 is used to determine variance information based on the variance of the left lane line and the variance of the right lane line; The curb information acquisition unit 25 is used to acquire the curb information of the current lane; The function degradation judgment unit 26 is used to control the intelligent navigation function of the target vehicle to perform degradation operation when the width change information, variance information and curb information meet the preset degradation conditions.

[0083] The above 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.

[0084] Although preferred embodiments have been described in this specification, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications that fall outside the scope of this specification.

[0085] Obviously, those skilled in the art can make various modifications to this specification without departing from its spirit and scope. Therefore, this specification also intends to include any modifications that fall within the scope of the claims and their equivalents.

Claims

1. A method for determining the degradation of intelligent navigation function at merging points, characterized in that, include: Obtain the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; Based on the current lane width and the predicted lane width, determine the width change information; Obtain the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; Variance information is determined based on the variance of the left lane line and the variance of the right lane line; Obtain the curb information of the current lane; When the width change information, the variance information, and the curb information meet the preset downgrade conditions, the intelligent navigation function of the target vehicle is controlled to perform a downgrade operation.

2. The method according to claim 1, characterized in that, The step of obtaining the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle includes: Obtain the left lane line equation and right lane line equation of the target vehicle in the current lane; Based on the left lane line equation and the right lane line equation, determine the current lane width at the current position of the target vehicle; Based on the left lane line equation and the right lane line equation, the width of the pre-aiming lane at a preset pre-aiming distance in front of the target vehicle is determined.

3. The method according to claim 1, characterized in that, The step of obtaining the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane includes: Based on the current position of the target vehicle, select multiple aiming points forward along the vehicle trajectory line; Based on the multiple aiming points, determine the left lateral distance sequence between each aiming point and the left lane line and the right lateral distance sequence between each aiming point and the right lane line; The variance of the left lane line is determined based on the left lateral distance sequence; The variance of the right lane line is determined based on the right-side lateral distance sequence.

4. The method according to claim 3, characterized in that, The variance information includes variance weight difference and variance confidence level. Determining the variance information based on the variance of the left lane line and the variance of the right lane line includes: Based on the variance of the left lane line and the variance of the right lane line, determine the weight of the left variance and the weight of the right variance; The variance weight difference is determined based on the left-side variance weight and the right-side variance weight; Based on the variance of the left lane line and the variance of the right lane line, the variance confidence level of the lane line with the larger variance is determined.

5. The method according to claim 3, characterized in that, Before selecting multiple aiming points forward along the vehicle trajectory line based on the current position of the target vehicle, the method further includes: Obtain the left longitudinal distance of the identified left lane line and the right longitudinal distance of the identified right lane line; Based on the left longitudinal distance and the right longitudinal distance, the lane line on the side with the longer longitudinal distance is determined as the reference side; Based on the lane line equations on the reference side and the vehicle trajectory lines, the lane lines on the side with shorter longitudinal distances are completed to generate the completed lane lines.

6. The method according to claim 1, characterized in that, The step of obtaining the curb information of the current lane includes: Detect whether there are curbs on both sides of the current lane, and determine whether curbs exist; When a curb is detected, the lateral distance between the curb and the lane line on the same side is determined; When the lateral distance is less than a preset distance threshold, the lateral variance weight difference of the roadside is determined based on the roadside and the lane line on the same side. The curb information includes curb presence information, and when the curb is detected, it also includes the lateral distance, and when the lateral distance is less than the preset distance threshold, it also includes the curb lateral variance weight difference.

7. The method according to claim 4, characterized in that, The preset degradation conditions include a first sub-condition, a second sub-condition, and a third sub-condition, and the preset degradation conditions are determined in the following way: If the difference between the current lane width and the target lane width in the width change information is greater than a preset width threshold and the duration reaches a preset anti-shake time, it is determined as the first sub-condition. If the variance weight difference in the variance information is greater than a first preset weight threshold and the variance confidence is greater than a preset confidence threshold, it is determined as the second sub-condition. The third sub-condition is defined as follows: one side of the current lane has a road edge, the lateral distance between the road edge and the lane line on the same side is less than a preset distance threshold, the corresponding difference in road edge lateral variance weights is less than a second preset weight threshold, and the other side of the current lane does not have a road edge.

8. The method according to claim 7, characterized in that, When the width change information, the variance information, and the curb information meet preset downgrade conditions, the intelligent navigation function of the target vehicle is controlled to perform a downgrade operation, including: When the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information satisfies the third sub-condition, the intelligent navigation function of the target vehicle is controlled to exit the lateral control operation and maintain the longitudinal adaptive cruise operation.

9. The method according to claim 7, characterized in that, Also includes: When the width change information satisfies the first sub-condition, the variance information satisfies the second sub-condition, and the curb information does not satisfy the third sub-condition, it is determined that a lane line false detection has occurred. Based on the lane line corresponding to the smaller variance of the left lane line variance and the right lane line variance, a single-lane control operation is performed on the target vehicle.

10. A device for determining the degradation of intelligent navigation function at merging points, characterized in that, include: The lane width acquisition unit is used to acquire the current lane width at the current position of the target vehicle and the pre-aiming lane width at a preset pre-aiming distance in front of the target vehicle; A width change determination unit is used to determine width change information based on the current lane width and the target lane width; The lane variance acquisition unit is used to acquire the variance of the left lane line and the variance of the right lane line of the target vehicle in the current lane; The variance information determination unit is used to determine variance information based on the variance of the left lane line and the variance of the right lane line; A curb information acquisition unit is used to acquire curb information of the current lane; The function degradation judgment unit is used to control the intelligent navigation function of the target vehicle to perform a degradation operation when the width change information, the variance information and the curb information meet the preset degradation conditions.

Images