Vehicle collision risk dynamic assessment method based on multi-modal perception

Abstract

The invention relates to the technical field of intelligent driving, in particular to a vehicle collision risk dynamic assessment method based on multi-modal perception. In the vehicle driving process, driving information and environment characteristic information of a vehicle are collected through a sensor array, and vehicle increment movement prediction is carried out according to the environment characteristic information; acquiring a second driving speed of the vehicle, performing vehicle incremental movement identification, and acquiring an identified vehicle incremental movement distance; according to the first driving speed and the second driving speed, calculating to obtain a fusion driving speed, and carrying out weighted calculation on the predicted vehicle increment moving distance and the identified vehicle increment moving distance to obtain a vehicle increment moving distance; according to the first monitoring distance and the vehicle increment moving distance, a second monitoring distance is obtained through calculation, correction is conducted through the fused driving speed, a third monitoring distance is obtained, collision risk assessment is conducted, and an assessment result is obtained. The technical effect of accurately and quickly evaluating the collision risk is achieved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a vehicle collision risk dynamic assessment method based on multimodal perception. Background Art

[0002] Existing vehicle collision risk assessment technology primarily relies on a single distance threshold to determine vehicle collision risk, using only sensors to measure vehicle distance. It fails to consider the impact of environmental factors (such as slippery roads and weather conditions) and vehicle dynamics (such as displacement caused by skidding) on ​​collision risk. This leads to high rates of misjudgment and missed judgment in complex road conditions, and fails to meet the needs of intelligent driving for accurate collision risk assessment. During vehicle driving, it is impossible to effectively integrate multimodal data to dynamically and accurately assess collision risk, and it is unable to address the large assessment errors, lack of real-time performance, and insufficient reliability caused by dynamic changes in the environment and the vehicle itself. In summary, existing technologies suffer from the technical problem of being unable to accurately and quickly assess vehicle collision risk. Summary of the Invention

[0003] The present invention aims to solve the technical problem that the existing technology cannot accurately and quickly assess the risk of vehicle collision, and provides a dynamic assessment method for vehicle collision risk based on multimodal perception.

[0004] The technical solution of the present invention to solve the above technical problems is as follows: In a first aspect, the present invention provides a method for dynamic assessment of vehicle collision risk based on multimodal perception, including: a method for dynamic assessment of vehicle collision risk based on multimodal perception, the method including: during vehicle driving, collecting the vehicle's first driving speed, first monitoring distance and environmental feature information through a sensor array, performing vehicle incremental movement prediction based on the environmental feature information to obtain a predicted vehicle incremental movement distance; obtaining the vehicle's second driving speed, and performing vehicle incremental movement identification in combination with the first driving speed to obtain an identified vehicle incremental movement distance; calculating a fused driving speed based on the first driving speed and the second driving speed, configuring prediction weights and identification weights, performing weighted calculation on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance to obtain a vehicle incremental movement distance; calculating a second monitoring distance based on the first monitoring distance and the vehicle incremental movement distance, and using the fused driving speed for correction to obtain a third monitoring distance, performing collision risk assessment, and obtaining an assessment result.

[0005] Optionally, during vehicle driving, a sensor array is used to collect the vehicle's first driving speed, first monitoring distance, and environmental characteristic information, and a vehicle incremental movement prediction is performed based on the environmental characteristic information to obtain a predicted vehicle incremental movement distance, including: during vehicle driving, a sensor array is used to collect the vehicle's first driving speed, first monitoring distance, and environmental characteristic information, wherein the sensor array includes a ranging sensor, a speed sensor, and a visual sensor, and the environmental characteristic information includes an environmental visual image; a trained environmental incremental movement predictor is called; the environmental characteristic information is input into the environmental incremental movement predictor, and a prediction output is made to predict the vehicle incremental movement distance.

[0006] Among them, the training steps of the environment incremental movement predictor include: collecting a set of sample environment feature information based on the test data of vehicle incremental movement, and collecting the vehicle incremental movement distance under different sample environment feature information to obtain a set of sample predicted vehicle incremental movement distances; constructing an environment incremental movement predictor based on machine learning; using the sample environment feature information set and the sample predicted vehicle incremental movement distance set to supervise the environment incremental movement predictor, and completing the training after convergence.

[0007] Optionally, a second driving speed of the vehicle is obtained, and combined with the first driving speed, vehicle incremental movement recognition is performed to obtain the identified vehicle incremental movement distance, including: obtaining the second driving speed through the vehicle control system; inputting the first driving speed and the second driving speed into a speed incremental movement identifier, and identifying and outputting the identified vehicle incremental movement distance, wherein the speed incremental movement identifier is trained using a sample first driving speed set, a sample second driving speed set, and a sample identified vehicle incremental movement distance set.

[0008] Optionally, performing a weighted calculation on the predicted vehicle incremental moving distance and the identified vehicle incremental moving distance to obtain the vehicle incremental moving distance includes: using the prediction weight and the identification weight to perform a weighted calculation on the predicted vehicle incremental moving distance and the identified vehicle incremental moving distance to obtain the vehicle incremental moving distance.

[0009] Optionally, a second monitoring distance is calculated based on the first monitoring distance and the vehicle incremental moving distance, and the fused driving speed is used for correction to obtain a third monitoring distance, and a collision risk assessment is performed to obtain an assessment result, including: subtracting the vehicle incremental moving distance from the first monitoring distance to obtain the second monitoring distance; correcting and calculating the second monitoring distance based on the ratio of a preset average driving speed to the fused driving speed to obtain a third monitoring distance; calculating the ratio of the safe collision distance threshold to the third monitoring distance to obtain a collision risk coefficient as an assessment result.

[0010] By implementing the present invention, it is possible to collect the vehicle's first driving speed, first monitoring distance, and environmental feature information through a sensor array during vehicle travel, predict the vehicle's incremental movement based on the environmental feature information, and obtain a predicted incremental vehicle movement distance. Using information such as environmental visual images, potential skidding risks can be predicted in advance, thereby achieving forward-looking perception of vehicle travel risks. By implementing the present invention, it is possible to obtain the second driving speed of the vehicle, combine it with the first driving speed, perform vehicle incremental movement identification, obtain the identified vehicle incremental movement distance, quickly detect the vehicle's skidding state, and provide real-time data support for collision risk assessment; By implementing the present invention, it is possible to calculate a fused driving speed based on the first driving speed and the second driving speed, configure prediction weights and identification weights, perform weighted calculation on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance, and obtain the vehicle incremental movement distance. This effectively integrates the advantages of different information sources and improves the accuracy and robustness of the vehicle incremental movement distance calculation. By implementing the present invention, it is possible to calculate the second monitoring distance based on the first monitoring distance and the vehicle incremental movement distance, and use the fused driving speed for correction to obtain the third monitoring distance, perform collision risk assessment, and obtain an assessment result, thereby providing a reliable decision-making basis for the vehicle active safety system and enhancing the reliability and practicality of risk assessment.

[0011] In summary, by implementing the present invention, the technical effect of accurately and quickly assessing collision risks can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A schematic flow chart of a vehicle collision risk dynamic assessment method based on multimodal perception provided by the present invention; DETAILED DESCRIPTION The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0013] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Therefore, features specified as "first" or "second" may explicitly or implicitly include one or more features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0014] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0015] Example 1, as Figure 1 As shown, an embodiment of the present invention provides a vehicle collision risk dynamic assessment method based on multimodal perception, including: S100: During vehicle travel, a sensor array is used to collect a first vehicle travel speed, a first monitoring distance, and environmental characteristic information of the vehicle, and an incremental movement prediction of the vehicle is performed based on the environmental characteristic information to obtain a predicted incremental movement distance of the vehicle. S200: Acquire a second driving speed of the vehicle, and perform vehicle incremental movement identification in combination with the first driving speed to obtain an identified vehicle incremental movement distance; S300: Calculating a fused driving speed based on the first driving speed and the second driving speed, configuring a prediction weight and an identification weight, and performing a weighted calculation on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance to obtain a vehicle incremental movement distance; S400: Calculate a second monitoring distance based on the first monitoring distance and the vehicle incremental movement distance, and use the fused driving speed for correction to obtain a third monitoring distance, perform a collision risk assessment, and obtain an assessment result.

[0016] In step S100 of the embodiment of the present application, during the vehicle's travel, a first travel speed, a first monitoring distance, and environmental characteristic information of the vehicle are collected through a sensor array, and an incremental movement prediction of the vehicle is performed based on the environmental characteristic information to obtain a predicted incremental movement distance of the vehicle, including: During the driving of the vehicle, a first driving speed, a first monitoring distance, and environmental characteristic information of the vehicle are collected through a sensor array, wherein the sensor array includes a distance sensor, a speed sensor, and a visual sensor, and the environmental characteristic information includes an environmental visual image; Call the trained environment incremental movement predictor; The environmental feature information is input into the environmental incremental movement predictor, and a prediction output is made to predict the incremental movement distance of the vehicle.

[0017] In the embodiment of the present application, the first driving speed is the actual speed of the vehicle, that is, the moving speed relative to the ground; the first monitoring distance is the distance from the adjacent vehicle or obstacle monitored while the vehicle is driving; the environmental characteristic information is an environmental visual image, including road surface characteristics (such as wet, icy, dry), weather conditions (rainy days, foggy days) and other environmental characteristic information.

[0018] Among them, the first driving speed can be measured by a speed sensor (such as a GPS speed measurement module); the first monitoring distance can be measured by a ranging sensor (such as a lidar, a millimeter-wave radar); and the environmental feature information can be collected by a visual sensor (such as a camera, an infrared imaging system).

[0019] In this embodiment of the present application, the environmental incremental movement predictor is used to predict the incremental movement distance based on environmental characteristic information. A slippery road can cause a vehicle to skid during braking, resulting in incremental movement and, in turn, a collision. Therefore, the incremental movement distance is the distance that a vehicle may skid due to road characteristics.

[0020] The training step of the environment incremental movement predictor includes: Based on the test data of vehicle incremental movement, a set of sample environment feature information is collected, and the vehicle incremental movement distance under different sample environment feature information is collected to obtain a set of sample predicted vehicle incremental movement distances; Build an environment incremental movement predictor based on machine learning; The sample environment feature information set and the sample predicted vehicle incremental movement distance set are used to perform supervised training on the environment incremental movement predictor, and the training is completed after convergence.

[0021] In this embodiment of the present application, the training samples for the environmental incremental movement predictor (i.e., environmental visual images labeled with the vehicle's actual incremental movement distance) can be collected in a controlled test field, simulating different environments (e.g., wet and icy roads), or in natural scenes on actual roads (e.g., rainy days, nighttime), ensuring driving safety. Each frame of the image is then manually annotated with an environmental label (e.g., "wet and slippery road," "dry road," "icy road," etc.).

[0022] For each captured environmental visual image, the vehicle's actual incremental movement distance in the corresponding environment is calculated (using an IMU or high-precision positioning system) and used as a label for supervised learning (e.g., labeling the actual incremental movement distance as 10 meters). The vehicle's incremental movement distance is collected under different sample environmental characteristics to obtain a set of sample-predicted incremental movement distances. Ultimately, no fewer than 10,000 sets of sample data (including different environmental scenarios) are collected and divided into a training set (70%), a validation set (15%), and a test set (15%) for training the environmental incremental movement predictor.

[0023] Alternatively, the environment incremental motion predictor can be built using a convolutional neural network. The main structure of the environment incremental motion predictor from the input layer to the output layer is as follows: input layer, receiving the environment visual image, input color image, the resolution can be 224×224; convolution layer 1, the number of convolution kernels is 32, the kernel size is 3×3, the stride is 1, using same padding, and the activation function is ReLU; pooling layer 1, using maximum pooling, the pooling kernel size is 2×2, the stride is 2; convolution layer 2, the number of convolution kernels is 64, the kernel size is 3×3, the stride is 1, using same padding, and the activation function is ReLU; pooling layer 2, using maximum pooling, the pooling kernel size is 2×2, the stride is 2; flattening layer, used to expand the multi-dimensional feature map into a one-dimensional vector; fully connected layer 1, the number of neurons is 256, the activation function is ReLU, and the dropout rate is 0.2; output layer, the number of neurons is 1, and the linear activation function is used.

[0024] In training the incremental motion predictor, Adam is used as the optimizer; the learning rate is set to 0.001; L2 regularization is used with a coefficient of 0.001; the mean squared error (MSE) is used as the loss function; an early stopping strategy is implemented, monitoring the validation set loss and stopping if it does not decrease for 10 consecutive epochs; and the maximum number of training epochs is preset to 500. Combined with the early stopping strategy, convergence is typically achieved within 100-200 epochs. The convergence criterion for the incremental motion predictor can be set to a validation set MSE below 0.01.

[0025] The environment incremental movement predictor is trained using the above method until convergence is achieved. By inputting environmental feature information (environmental visual images) into the environment incremental movement predictor, the predicted vehicle incremental movement distance (e.g., 10 meters) is output.

[0026] In step S200 of the embodiment of the present application, the second driving speed of the vehicle is obtained, and the vehicle incremental movement identification is performed in combination with the first driving speed to obtain the identified vehicle incremental movement distance, including: obtaining a second driving speed through a control system of the vehicle; The first driving speed and the second driving speed are input into a speed incremental movement identifier, and the recognition output is used to obtain the identified vehicle incremental movement distance, wherein the speed incremental movement identifier is trained using a sample first driving speed set, a sample second driving speed set, and a sample identified vehicle incremental movement distance set.

[0027] In this embodiment of the present application, the second driving speed is the expected speed assuming no tire slippage on the ground, i.e., the theoretical driving speed, i.e., the rotational linear velocity of the outer edge of the tire. This can be calculated by obtaining vehicle information from the vehicle control system (e.g., engine ECU, transmission gear ratio). Specifically, the theoretical driving speed can be calculated by reading parameters such as engine speed, transmission gear ratio, and tire radius via the CAN bus or OBD-II interface. For example, the second driving speed can be: Second Driving Speed ​​= (Engine Speed ​​* Tire Circumference) / (Transmission Gear Ratio * Final Drive Ratio).

[0028] As for the speed incremental movement identifier, it is necessary to use the sample first driving speed set, the sample second driving speed set and the sample identification vehicle incremental movement distance set for training. The sample first driving speed and the sample second driving speed can be directly obtained by the aforementioned method, which will not be repeated here. The sample identification vehicle incremental movement distance is the actual measured vehicle incremental movement distance under the specific sample first driving speed and sample second driving speed. For example, in a closed field or simulation environment, through acceleration, deceleration, emergency braking or sharp turning tests under different road conditions (dry, wet, icy), the actual incremental movement distance can be obtained in real time using a high-precision positioning system (such as RTK-GPS) or IMU measurement, and multiple sample identification vehicle incremental movement distances can be collected to obtain the sample identification vehicle incremental movement distance set.

[0029] The above-mentioned sample first driving speed set, sample second driving speed set and sample identification vehicle incremental movement distance set are divided into training set, verification set and test set according to the ratio of 70%, 20% and 10% for training the speed incremental movement identifier.

[0030] In an embodiment of the present application, the speed increment movement identifier identifies the abnormal movement distance of the vehicle caused by factors such as slipping by analyzing the difference between two speeds (the first driving speed and the second driving speed), so it can be built using a gated recurrent unit (GRU) network.

[0031] Optionally, the speed incremental movement identifier includes the following parts: an input layer, which is used to receive speed pairs (first driving speed, second driving speed) within a sliding time window (such as 3 seconds, sampling frequency 30Hz, a total of 90 time steps), with a feature dimension of 90×2; a GRU layer, which is set with 64 neurons to learn the temporal features of speed differences; an attention mechanism layer, which is used to weight the temporal features output by the GRU and highlight key time points (such as the start and end moments of slipping); a fully connected layer, which is used to reduce the feature dimension to 1 dimension and output the recognized incremental movement distance.

[0032] The speed incremental motion recognizer is trained using the Adam optimizer with an initial learning rate of 0.001. The Huber loss function is used as the loss function. Dropout (0.2) is used for regularization to prevent overfitting. The batch size is 64, and the number of training epochs is 100. An early stopping strategy is implemented, stopping when the validation set loss does not decrease after five consecutive epochs. A layered training strategy is used, with only the fully connected layers trained in the first phase and the GRU layer weights frozen. In the second phase, the GRU layers are unfrozen and fine-tuned. If the validation set MSE does not show a significant decrease (e.g., a decrease of ≤0.05%) over 100 consecutive training epochs, convergence is considered, and the speed incremental motion recognizer is obtained. The first and second driving speeds are input into the speed incremental motion recognizer, and the output is the identified vehicle's incremental travel distance.

[0033] In step S300 of the embodiment of the present application, a fused driving speed is calculated based on the first driving speed and the second driving speed, and a prediction weight and a recognition weight are configured, including: Calculating an average according to the first driving speed and the second driving speed to obtain a fused driving speed; Calculating a ratio of the fused driving speed to a preset average driving speed, and performing correction calculation on a preset recognition weight to obtain a recognition weight; Based on the recognition weight, a prediction weight is calculated.

[0034] In this embodiment of the present application, the first driving speed is the actual driving speed directly measured by onboard sensors (e.g., 50 km / h). The second driving speed is the theoretical driving speed calculated based on engine output and transmission system parameters (e.g., 60 km / h). The combined driving speed is the average of the aforementioned two speeds (e.g., 55 km / h).

[0035] The preset average driving speed is a typical driving speed obtained based on a large amount of historical data (e.g., 30 km / h for urban roads and 80 km / h for highways). This preset average speed is used as a benchmark speed to evaluate the particularity of the current driving state.

[0036] In this embodiment of the present application, the preset recognition weight can be set to 0.5, indicating that environmental prediction and speed recognition are initially given equal importance. The correction formula can be: Recognition weight = preset recognition weight * (fused driving speed / preset average driving speed) α, where α is an adjustment coefficient (which can be 0.8) that controls the strength of the correction. When the fused driving speed exceeds the preset average driving speed (e.g., when driving at high speed), the risk of vehicle skidding increases, speed recognition becomes more important, and the recognition weight increases. When the fused driving speed is less than the preset average driving speed (e.g., in congested traffic), environmental factors (such as slippery roads) have a more significant impact, and the recognition weight decreases. Once the recognition weight is obtained, the prediction weight can be further calculated using the formula: prediction weight = 1 - recognition weight. This dynamic weight configuration method allows the system to adapt to different driving conditions, optimize the fusion of environmental prediction and speed recognition, and improve the accuracy of collision risk assessment.

[0037] In step S300 of the embodiment of the present application, weighted calculation is performed on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance to obtain the vehicle incremental movement distance, including: The predicted vehicle incremental movement distance and the identified vehicle incremental movement distance are weightedly calculated using the prediction weight and the identification weight to obtain the vehicle incremental movement distance.

[0038] In the vehicle collision risk assessment in the embodiment of the present application, weighted fusion of the output of the environment incremental movement predictor (predicting the vehicle incremental movement distance) and the output of the speed incremental movement identifier (identifying the vehicle incremental movement distance) is a key step to improve the prediction accuracy.

[0039] Alternatively, the weighting method can be: vehicle incremental movement distance = prediction weight * vehicle incremental movement distance + identification weight * identification vehicle incremental movement distance. For example, when the prediction weight = 0.4, the identification weight = 0.6, the predicted vehicle incremental movement distance = 1.2m, and the identified vehicle incremental movement distance = 0.8m, then the vehicle incremental movement distance = 0.4*1.2m+0.6*0.8=0.96m. The fused vehicle incremental movement distance will be used to correct the actual distance between the vehicle and the obstacle. In step S400 of the embodiment of the present application, a second monitoring distance is calculated based on the first monitoring distance and the vehicle incremental movement distance, and the fused driving speed is used for correction to obtain a third monitoring distance. A collision risk assessment is performed to obtain an assessment result, including: Subtracting the vehicle incremental movement distance from the first monitoring distance to obtain a second monitoring distance; Correcting and calculating the second monitoring distance according to a ratio of a preset average driving speed to a fused driving speed to obtain a third monitoring distance; The ratio of the safe collision distance threshold to the third monitoring distance is calculated to obtain a collision risk coefficient as an evaluation result.

[0040] In this embodiment of the present application, the first monitoring distance is the real-time distance between the vehicle and adjacent vehicles / obstacles, as directly measured by a sensor (e.g., a lidar or millimeter-wave radar). The vehicle's incremental movement distance is calculated using the method described in S300. The second monitoring distance is the remaining safe distance after correcting for slippage, calculated as follows: second monitoring distance = first monitoring distance - vehicle incremental movement distance.

[0041] In this embodiment of the present application, the higher the driving speed, the longer the vehicle's braking distance, and the greater the safety distance required. Therefore, the second monitoring distance needs to be dynamically adjusted based on the current speed. Specifically, this adjustment is performed by using a preset average driving speed as the benchmark speed (e.g., 30 km / h for urban roads and 80 km / h for highways) and calibrating it with the fused driving speed (the average of the first and second driving speeds, reflecting the current actual speed).

[0042] A specific correction method may be: third monitoring distance = second monitoring distance * (preset average driving speed / fused driving speed).

[0043] Furthermore, it is necessary to calculate the ratio of the safe collision distance threshold to the third monitoring distance to obtain a collision risk coefficient as an evaluation result. The safe collision distance threshold is the minimum safe distance calculated according to the vehicle dynamics model (such as the braking distance formula). For example, under certain environmental conditions and vehicle speeds, the value can be 50m. The method for obtaining the safe collision distance threshold is the existing technology and will not be repeated here. The calculation method of the collision risk coefficient can be: collision risk coefficient = safe collision distance threshold / third monitoring distance. Furthermore, it can be defined that when the collision risk coefficient is ≤1, it means that it is safe and there is no risk; when the collision risk coefficient is >1, it means that there is a collision risk and the driver needs to be reminded to avoid it in time. It should be noted that in the above embodiments, the description of each embodiment has its own focus. For the part that is not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0044] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0045] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0046] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0047] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0048] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A dynamic assessment method for vehicle collision risk based on multimodal perception, characterized in that: The method comprises: During the vehicle's travel, the sensor array collects a first travel speed, a first monitoring distance, and environmental characteristic information of the vehicle, and performs a vehicle incremental movement prediction based on the environmental characteristic information to obtain a predicted vehicle incremental movement distance; Acquire a second driving speed of the vehicle, and perform vehicle incremental movement identification in combination with the first driving speed to obtain an identified vehicle incremental movement distance; Calculating a fused driving speed based on the first driving speed and the second driving speed, configuring a prediction weight and an identification weight, and performing a weighted calculation on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance to obtain a vehicle incremental movement distance; A second monitoring distance is calculated based on the first monitoring distance and the vehicle incremental movement distance, and the fused driving speed is used for correction to obtain a third monitoring distance, and a collision risk assessment is performed to obtain an assessment result.

2. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 1 is characterized in that: During vehicle travel, a first driving speed, a first monitoring distance, and environmental characteristic information of the vehicle are collected through a sensor array, and an incremental movement prediction of the vehicle is performed based on the environmental characteristic information to obtain a predicted incremental movement distance of the vehicle, including: During the driving of the vehicle, a first driving speed, a first monitoring distance, and environmental characteristic information of the vehicle are collected through a sensor array, wherein the sensor array includes a distance sensor, a speed sensor, and a visual sensor, and the environmental characteristic information includes an environmental visual image; Call the trained environment incremental movement predictor; The environmental feature information is input into the environmental incremental movement predictor, and a prediction output is made to predict the incremental movement distance of the vehicle.

3. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 2 is characterized in that: The training steps of the environment incremental movement predictor include: Based on the test data of vehicle incremental movement, a set of sample environment feature information is collected, and the vehicle incremental movement distance under different sample environment feature information is collected to obtain a set of sample predicted vehicle incremental movement distances; Build an environment incremental movement predictor based on machine learning; The sample environment feature information set and the sample predicted vehicle incremental movement distance set are used to perform supervised training on the environment incremental movement predictor, and the training is completed after convergence.

4. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 1 is characterized in that: Acquiring a second driving speed of the vehicle, and performing vehicle incremental movement identification in combination with the first driving speed to obtain an identified vehicle incremental movement distance, including: obtaining a second driving speed through a control system of the vehicle; The first driving speed and the second driving speed are input into a speed incremental movement identifier, and the recognition output is used to obtain the identified vehicle incremental movement distance, wherein the speed incremental movement identifier is trained using a sample first driving speed set, a sample second driving speed set, and a sample identified vehicle incremental movement distance set.

5. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 1 is characterized in that: Calculating a fused driving speed based on the first driving speed and the second driving speed, and configuring a prediction weight and a recognition weight include: Calculating an average according to the first driving speed and the second driving speed to obtain a fused driving speed; Calculating a ratio of the fused driving speed to a preset average driving speed, and performing correction calculation on a preset recognition weight to obtain a recognition weight; Based on the recognition weight, a prediction weight is calculated.

6. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 1 is characterized in that: Performing weighted calculation on the predicted vehicle incremental movement distance and the identified vehicle incremental movement distance to obtain the vehicle incremental movement distance includes: The predicted vehicle incremental movement distance and the identified vehicle incremental movement distance are weightedly calculated using the prediction weight and the identification weight to obtain the vehicle incremental movement distance.

7. The vehicle collision risk dynamic assessment method based on multimodal perception according to claim 1 is characterized in that: A second monitoring distance is calculated based on the first monitoring distance and the incremental movement distance of the vehicle, and the fused driving speed is used for correction to obtain a third monitoring distance. A collision risk assessment is performed to obtain an assessment result, including: Subtracting the vehicle incremental movement distance from the first monitoring distance to obtain a second monitoring distance; Correcting and calculating the second monitoring distance according to a ratio of a preset average driving speed to a fused driving speed to obtain a third monitoring distance; The ratio of the safe collision distance threshold to the third monitoring distance is calculated to obtain a collision risk coefficient as an evaluation result.

Images