Method for Predicting the Collision Risk between an Electric Bicycle and a Vehicle Based on Multimodal Behaviors of the Electric Bicycle

Abstract

The present invention belongs to the field of road vehicle control systems and relates to a method for predicting the collision risk between an electric bicycle and a vehicle based on multi-modal behaviors of the electric bicycle. This method uses physical modeling, with the braking distance of the electric bicycle and the impacts on the vehicle caused by its lateral and longitudinal stability as the basic risks; further considers the risks to the vehicle caused by the environment, and comprehensively conducts multi-dimensional environmental risk assessment modeling in four dimensions of dynamic distance-speed, adjacent vehicle types, traffic flow density, and weather conditions (precipitation, temperature, wind speed, visibility); combines the risks to the vehicle due to the operating behaviors of the electric bicycle (acceleration changes, steering angles, and frequencies); uses an environmental sensitivity coefficient to capture the complex relationship between environmental changes and vehicle risks in real time. Finally, for the multi-modal riding behaviors of the electric bicycle, a comprehensive, accurate, and real-time multi-dimensional comprehensive risk assessment method is provided for the vehicle to reduce the collision risk in a complex traffic flow environment.

Description

Technical Field

[0001] The present invention belongs to the field of road vehicle control systems, and relates to the prediction of upcoming collision risks. Specifically, it relates to a method for predicting the collision risk between an electric bicycle and a vehicle based on multi-modal behaviors of the electric bicycle. Background Art

[0002] Currently, China is the country with the largest number of motor vehicles in the world. As of 2024, the national motor vehicle ownership reached 450 million vehicles, showing a continuous high-growth trend, with both the total quantity and the increment ranking first in the world. At the same time, China also has the world's largest electric bicycle market. As one of the most popular means of transportation in modern times, cars provide an efficient and flexible way of personal travel, greatly improving the convenience of life and the efficiency of time utilization. As one of the most original innovative products with Chinese characteristics since China's reform and opening up, electric bicycles are a national industry that is green, close to people's livelihood, and has a high degree of independent intellectual property rights. Both show a positive development trend in terms of market scale, industrial chain development, technological progress, policy support, and market prospects.

[0003] With the growth of the number of motor vehicles and the development of electric bicycles, the negative impacts they bring to urban road traffic have become increasingly prominent, and the number of motor vehicle accidents caused by electric bicycles has been increasing year by year. Currently, to solve the problem of conflicts between motor vehicles and non-motor vehicles, the main measure taken is to install a blind spot video monitoring system on the vehicle side. However, due to the flexible and changeable characteristics of electric bicycles, even if the blind spot monitoring system feeds back the traffic operation status in the blind spot to the vehicle driver, it is very difficult for the driver to make timely actions, and the accident rate cannot be effectively reduced.

[0004] In order to predict the risk of conflicts between motor vehicles and non-motor vehicles in advance, scholars have designed risk assessment systems for conflicts between motor vehicles and non-motor vehicles. However, the following problems generally exist in existing risk assessment systems for conflicts between motor vehicles and non-motor vehicles:

[0005] (1) Most rely on traffic simulation models (such as VISSIM, AIMSUN, etc.) to simulate multiple levels such as traffic flow, traffic lights, non-motor vehicle and pedestrian behaviors. The factors considered are broad but not precise, and the operating characteristics of electric bicycles (such as lateral and longitudinal stability, acceleration changes) are rarely considered, making it difficult to conduct a refined risk assessment of the risks caused by electric bicycles in a targeted manner. In addition, most existing conflict models between motor vehicles and motor vehicles, and between motor vehicles and pedestrians start from the analysis of motor vehicle driving behaviors (such as lane change, emergency braking), pedestrian behaviors (such as walking speed, forward direction), and do not specifically consider the operating characteristics of electric bicycles, and cannot be applied to risk assessment for electric bicycles. Coupled with the fact that existing risk assessment systems for electric bicycles are even rarer, the accuracy of risk identification of electric bicycles by vehicles in complex traffic scenarios is not high.

[0006] (2) Only focusing on a single factor. For example, TTC (Time to Collision) calculates the time when a collision may occur based on the relative distance and relative speed of two objects, so as to judge the magnitude of the collision risk. Since the selected indicators are too single and cannot handle complex non-linear behaviors, it is difficult to make a more accurate risk assessment for electric bicycles with high flexibility in complex scenarios for cars.

[0007] (3) Most existing studies assume good road conditions and ignore the impact of dynamic changes in many factors such as traffic flow density, adjacent vehicle types, and weather conditions on the risk assessment system. Some studies even use the road geometric structure captured by map point cloud information as the environmental condition, resulting in a narrow scope of application, weak generalization ability, and low accuracy of the assessment results of the risk assessment system.

[0008] Chinese Patent CN112581756A discloses "A Driving Risk Assessment Method Based on Mixed Traffic", which uniformly expresses the force relationships of motor vehicles, non-motor vehicles, and road environments through a social force model, establishes a risk model for the interaction and collision of motor vehicles and non-motor vehicles, and thus realizes the dynamic driving risk assessment in a mixed traffic scenario. However, this method simply abstracts the individual behavior results of any motor vehicle or non-motor vehicle in the traffic scenario into a combination of different forces, ignoring the potential impact of the differences in the types of microscopic traffic individuals and the changes in behavior characteristics (such as the transverse and longitudinal stability, steering angle and frequency of non-motor vehicles) on the driving risk. At the same time, the mechanical model it uses adopts simple linear superposition for different risk scenarios and is difficult to respond in a timely manner to sudden traffic changes, and it is difficult to fully meet the requirements of complex traffic environments in practical applications.

[0009] Chinese Patent CN112015842B discloses "A Risk Assessment Method and System for Autonomous Vehicles Based on Bicycle Trajectory Prediction", which combines the movement characteristics of cyclists and map point cloud information, and outputs the possible positions of cyclists through intention inference and trajectory prediction. However, when this method performs trajectory prediction, it relies on a 2D static map that can only reflect the road geometric structure to reflect environmental factors, ignoring the dynamically changing environmental conditions (such as weather conditions, traffic flow density), resulting in poor adaptability and robustness of the model in complex environments, insufficient generalization ability, and thus reducing the accuracy of the prediction results. At the same time, it needs to rely on a digital map for positioning before collecting information data, and being too dependent on accurate map information restricts its application scenarios relatively, and the data acquisition conditions are cumbersome.

[0010] Therefore, it is necessary to design a method to solve the traffic conflict problem between cars and electric bicycles, improve the traffic experience of travelers, and avoid the risk of road traffic accidents caused thereby. Summary of the Invention

[0011] In view of the above technical problems and deficiencies, the object of the present invention is to provide a method for predicting the collision risk between an electric bicycle and a vehicle based on multi-modal behavior. This method dynamically evaluates the potential risks in a mixed traffic environment by collecting real-time dynamic information of non-motor vehicles, combining external environmental factors, and using multi-dimensional modeling and intelligent algorithms. Such a method can not only accurately predict the collision risks caused by electric bicycles, timely capture high-risk behaviors, but also adjust the vehicle control strategy through environmental sensitivity, reduce traffic accidents, and improve traffic safety levels and driving experiences.

[0012] To achieve the above object, the present invention adopts the following technical solutions:

[0013] A method for predicting the collision risk between an electric bicycle and a vehicle based on multi-modal behavior, the method comprising the following steps:

[0014] Step 1. Decompose the overall vehicle risk of the car into the risk of the electric bicycle's inherent attributes to the car , the risk of the environment to the car and the risk of the electric bicycle's operation behavior to the car ;

[0015] Step 2. Evaluate the risk of the electric bicycle's inherent attributes to the car , the risk of the environment to the car and the risk of the electric bicycle's operation behavior to the car ;

[0016] Among them, the risk of the electric bicycle's inherent attributes to the car refers to the risk brought to the car by the electric bicycle's inherent attributes during the driving process of the car, including the risk brought to the car by the braking distance of the electric bicycle , the risk brought to the car by the stability of the electric bicycle ;

[0017] ；

[0018] Among them, is a proportionality coefficient, is a power exponent, is the rider's corrected reaction time, is the driving speed of the electric bicycle, is the acceleration of the electric bicycle;

[0019] ;

[0020] Among them, is a proportionality coefficient, is a power exponent, is the lateral stability index of the electric bicycle, is the longitudinal stability index of the electric bicycle, , is the weight coefficient, , , are the non-linear exponents;

[0021] Automobile risks caused by the environment refers to the risks caused to the automobile during driving due to changes in external environmental factors, including the risks caused by the dynamic distance-speed to the automobile , the risks caused by the types of adjacent vehicles to the automobile , the risks caused by the traffic flow density to the automobile , the risks caused by the weather conditions to the automobile ;

[0022] When evaluating the risks caused by the dynamic distance-speed to the automobile, the relative speed and distance between the automobile and the first vehicle behind are comprehensively considered; when evaluating the risks caused by the types of adjacent vehicles to the automobile, the risks caused by all vehicle types within the detection range to the automobile are calculated, and the highest risk is used for the risk prediction of the whole automobile; when evaluating the risks caused by the traffic flow density to the automobile, the form of the Gaussian distribution is used to construct the risk function of the traffic flow density to the automobile; when evaluating the risks caused by the weather conditions to the automobile, the risks caused by precipitation, temperature, wind speed and visibility to the automobile are comprehensively considered, that is: the risks caused by the weather conditions to the automobile include the risks caused by precipitation to the automobile , the risks caused by temperature to the automobile , the risks caused by wind speed to the automobile and the risks caused by visibility to the automobile ;

[0023] Risks of electric bicycle operation behavior to the automobile refers to the risks brought to the driving safety of the automobile by the driving behavior of the rider during the driving of the automobile, including the risks caused by the acceleration change frequency of the electric bicycle to the automobile , the risks caused by the steering angle and frequency of the electric bicycle to the automobile ;

[0024] ;

[0025] Among them, , , represent the weights of different acceleration change frequency bands, is the corrected acceleration change rate of the electric bicycle, ， is the acceleration change rate of the electric bicycle, is the speed of the electric bicycle, is the adjustment coefficient for controlling the amplification effect of the speed on the acceleration change frequency;

[0026] ;

[0027] Among them, is the steering angle of the electric bicycle, is used to adjust the steering angle is the constant coefficient for the degree of risk influence, is the speed amplification coefficient, which controls the exponential influence of the speed on the steering angle risk, is the distance between the car and the electric bicycle, represents the coefficient for the degree of risk influence of the steering frequency, represents time, that is, the time window for observing the steering frequency, is the unit time the number of steering operations within; represents the adjustment coefficient for the amplification effect of the speed on the acceleration change frequency, is the interaction risk term coefficient between the steering angle and the frequency;

[0028] Step 3. Predict the overall vehicle risk of the car based on the capital asset pricing model, and at the same time introduce the environmental sensitivity coefficient , evaluate the sensitivity of the environmental change to the overall vehicle risk of the car; in addition, add and the interaction term between to describe the superimposed effect of the electric bicycle operation behavior on the overall vehicle risk under different environmental conditions; add the high-order effect to reflect the non-linear trend of the environmental risk influence;

[0029] The prediction model of the overall vehicle risk of the car is:

[0030] ;

[0031] Among them, is the interaction risk coefficient, and the value range is [0.1, 1.0], which is used to reflect and the combined effect between; is the power of the high-order effect, which determines the influence speed and amplitude of the environmental risk change on the overall risk, value ; is the high-order environmental sensitivity coefficient, and the value range is [0.05, 0.5].

[0032] As an optimization of the present invention, the risk of the inherent attributes of the electric bicycle to the car The evaluation model for

[0033] ;

[0034] Among them, , , are non-linear exponents, has a value range of [0.5, 2.0], has a value range of [0.5, 2.0]; has a value range of [0.5, 3.0].

[0035] As a preference of the present invention, the risk of the vehicle caused by the environment The evaluation model for

[0036] ;

[0037] Among them, , , , represent the weight coefficients of each risk factor, has a value range of [0.25, 0.5], has a value range of [0.05, 0.2], has a value range of [0.15, 0.3], has a value range of [0.2, 0.4]; and finally satisfies ; represents the power of the contribution of each risk to the overall risk, has a value range of [1.5, 3.0], has a value range of [1.0, 2.0], has a value range of [1.2, 2.5], has a value range of [1.5, 3.0]; is the maximum value in

[0038] As a preference of the present invention, the risk of the vehicle caused by the operation behavior of the electric bicycle The evaluation model for

[0039] ;

[0040] Among them, , are weight coefficients, has a value range of [0.4, 0.7], has a value range of [0.3, 0.6]; , are non-linear exponents, The value range is [1.0, 2.0]; The value range is [1.5, 3.0]; The value range is [1.0, 2.0].

[0041] As an optimization of the present invention, the risk caused by the dynamic distance - speed to the vehicle evaluation model is:

[0042] ;

[0043] wherein, is the relative speed risk coefficient, is the critical value of the relative speed, , , are adjustment parameters, is the relative speed between the vehicle and the first vehicle behind, is the distance between the vehicle and the first vehicle behind, is the safe distance between the vehicle and the first vehicle behind, is a minimum value, taking 10 -6 ~10 -9 .

[0044] As an optimization of the present invention, the risk caused by the adjacent vehicle types to the vehicle evaluation model is:

[0045] ;

[0046] wherein, , , are weight coefficients, is the size of the vehicle type, is the mass of the vehicle type, is the blind area size.

[0047] As an optimization of the present invention, the risk caused by the traffic flow density to the vehicle evaluation model is:

[0048] ;

[0049] wherein, represents the traffic flow density, refers to the point with the highest impact on the vehicle driving risk under medium traffic flow density, is the risk sensitivity coefficient in the low - density area, is the risk sensitivity coefficient in the high - density area, is the risk change rate in the low - density area, is the risk change rate in the high - density area, is an adjustment parameter with a value range of [0.1, 0.5], which is used to regulate the overall risk level under low traffic flow density.

[0050] As an optimization of the present invention, the risk caused by weather conditions to vehicles has an evaluation model as follows:

[0051] ;

[0052] where , , , represent the weight coefficients of each factor, reflecting the contribution of different weather factors to the total risk; has a value range of [0.2, 0.4], has a value range of [0.1, 0.3], has a value range of [0.1, 0.3], has a value range of [0.2, 0.4]; and finally satisfies ; represents the power of the contribution of each risk to the overall risk, reflecting the non-linear influence of each factor, has a value range of [1.0, 2.0], has a value range of [1.5, 3.0], has a value range of [1.0, 2.5], has a value range of [0.5, 1.5]; where is the maximum value in

[0053] As an optimization of the present invention, the risk caused by precipitation to vehicles has an evaluation model as follows:

[0054] ;

[0055] where represents the minimum risk under dry conditions, which is a very small value, taking 10 -6 ~10 -9 ; is the coefficient of risk increase caused by the decrease of friction factor, is the magnitude of precipitation, is the friction coefficient of the dry road surface, is the friction coefficient under medium precipitation, is the lowest friction coefficient under heavy precipitation, , are the critical values of light, medium and heavy precipitation, is the exponential decay coefficient in the initial precipitation stage, indicating the amplitude of the initial friction coefficient decrease, is the logarithmic decline coefficient in the moderate precipitation stage, representing the impact of increased precipitation on the friction coefficient;

[0056] The risk caused by temperature to the vehicle The evaluation model is:

[0057] ;

[0058] Among them, represents the temperature; represents whether there is ice or snow on the road surface. If it exists, take 1; if it does not exist, take 0. represents the minimum risk under suitable temperature, which is a very small value, take 10 -6 ~10 -9 ; represents the coefficient of risk increase caused by temperature drop, indicating the degree of influence of temperature reduction on risk. is the friction coefficient of the dry road surface. represents the temperature at which the road surface begins to freeze. The critical temperature that controls the decline rate of the friction coefficient;

[0059] The risk caused by wind speed to the vehicle The evaluation model is:

[0060] ;

[0061] Among them, is the spatial interaction coefficient, reflecting the possibility that an electric bicycle invades the vehicle lane or collides with a vehicle after getting out of control, depending on the lane width and crosswind intensity; is the traffic flow coefficient, indicating the amplification effect of an electric bicycle getting out of control on the traffic flow composed of vehicles. represents the electric bicycle rollover risk under the condition of wind speed and the incident angle ;

[0062] The risk caused by visibility to the vehicle The evaluation model is:

[0063] ;

[0064] Among them, represents the current visibility. represents the basic risk value, that is, the lowest risk under good visibility, take 10 -6 ~10 -9 ; represents the risk adjustment parameter, used to control the influence range of visibility reduction on risk. represents the critical visibility.

[0065] As a preferred embodiment of the present invention, the environmental sensitivity coefficient The initial value range is [0.05, 0.5]. Subsequent iterative training is carried out based on the existing data to finally obtain the optimal value under different traffic scenarios. The calculation formula is:

[0066] ;

[0067] in, Risks for the entire vehicle Automotive risks to the environment The covariance of Automobile risks caused by the environment fluctuations; The risks of cars to the environment The variance of represents the fluctuation range of environmental risk.

[0068] Advantages and beneficial effects of the present invention:

[0069] (1) The present invention provides a method for predicting the risk of collision between an electric bicycle and a vehicle based on its multimodal behavior, which can effectively solve the problem that the traditional traffic risk assessment system has poor adaptability to complex mixed traffic scenes. The method comprehensively considers the impact of a series of complex factors on the driving of the vehicle, such as the lateral and longitudinal stability of the electric bicycle, acceleration changes, traffic flow density of the vehicle, and weather conditions, and can better handle the coupling effects and complex interdependencies between various factors; the method selects a variety of indicators that can adapt to nonlinear changes, and can make more accurate vehicle risk assessments for highly flexible electric bicycles and complex and changing environments.

[0070] (2) The present invention utilizes physical modeling, taking the braking distance of electric bicycles and the impact of their lateral and longitudinal stability on automobiles as the basic risk; further considering the automobile risks caused by the environment, the present invention integrates the dynamic distance-speed, adjacent vehicle types, traffic flow density, and weather conditions (precipitation, temperature, wind speed, and visibility) into four dimensions to conduct multi-dimensional environmental risk assessment modeling; the risk to the automobile caused by the operation behavior of electric bicycles (acceleration changes and steering angles and frequencies) is combined; the environmental sensitivity coefficient is used to capture the complex relationship between environmental changes and vehicle risks in real time, and finally, a comprehensive, accurate, and real-time multi-dimensional comprehensive risk assessment method is provided for the automobile based on the multi-modal riding behavior of electric bicycles. The method comprehensively considers various situations that the automobile may face during driving and has strong practicality; at the same time, the model performs nonlinear modeling for a variety of risk scenarios, thereby enhancing the generalization ability of the model, thereby effectively improving the driving safety of the vehicle and the road traffic efficiency.

[0071] (3)When the present invention evaluates the vehicle risks caused by the environment , the combination of independent power and differential weight is adopted, which can better consider the non-linear effects of different types of environmental risks on ; taking the maximum power as the reference power for calculation , while balancing the calculation efficiency and sensitivity, it ensures the accuracy of risk assessment; adopting adjustable parameters, it can be flexibly adjusted according to the complex environmental changes in different traffic scenarios, with strong adaptability; at the same time, compared with complex methods such as traditional neural networks, this method can be optimized with a small amount of data, especially having significant advantages in lightweight when applied in embedded systems, and can meet the need for accurately evaluating the vehicle risks caused by the environment in a short time.

[0072] (4)The method provided by the present invention has the advantages of real-time, dynamic and multi-dimensional comprehensive analysis, etc., can capture high-risk behaviors in time and provide a basis for subsequent warning or automatic control solutions for drivers, thereby effectively reducing the occurrence probability of traffic accidents, improving the driving safety and traffic efficiency in mixed traffic scenarios, and providing an innovative solution for intelligent traffic management in complex road environments.

[0073] (5)When the present invention evaluates the risks of vehicles caused by the environmental factor of traffic flow density, the method of combining piecewise Gaussian function and dynamic parameter adjustment is adopted, breaking through the limitations of traditional linear models and being able to accurately depict the non-linear risk characteristics in low, medium and high density areas. By introducing independent parameters ( , and , ), the risk sensitivity coefficients and change rates in different density areas are respectively regulated, and combined with the adjustment parameter to calibrate the potential risks in the low density area, solving the problems of risk underestimation in low-flow high-speed scenarios and large prediction deviations in the medium density peak area of traditional models; the model ensures the smooth transition of parameters at through continuity conditions, significantly improving the physical rationality and calculation stability of the risk curve; at the same time, the parameter fitting mechanism based on actual accident data (such as the values of , ) enhances the adaptability of the model to real traffic scenarios, provides a high-precision and high-robustness dynamic risk assessment tool for autonomous driving decision-making and traffic control, and can effectively reduce the collision risks in complex traffic flow environments.

[0074] (6)The present invention decomposes the overall vehicle risk into the risks of electric bicycles' inherent attributes to vehicles , the vehicle risks caused by the environment and the risks of electric bicycles' operating behaviors to vehicles Three independent modules support the independent optimization and dynamic adjustment of the parameters of each module, ensuring that the model can be quickly adapted to different scenarios, and significantly improving the dynamic response ability and robustness of risk prediction. For example, under the condition of sudden strong crosswind, the interaction coefficient between the electric bicycle and the car can be dynamically increased (such as ), and at the same time, the non-linear exponent of the steering angle and frequency of the electric bicycle is calibrated based on real-time sensor data (such as ), realizing the refined adaptation of risk prediction under extreme weather, thereby enhancing the anti-interference ability and real-time warning accuracy of the model to complex environmental disturbances. Brief Description of the Drawings

[0075] By referring to the following description in conjunction with the drawings, and with a more comprehensive understanding of the present invention, other objects and results of the present invention will become more obvious and easier to understand. In the drawings:

[0076] Figure 1 It is a flowchart for predicting the collision risk between an electric bicycle and a vehicle based on the multi-modal behavior of the electric bicycle provided by the present invention. Detailed Embodiment

[0077] To enable those skilled in the art to better understand the technical solutions and advantages of the present invention, the present application will be described in detail below with reference to the drawings, but it is not intended to limit the protection scope of the present invention.

[0078] This embodiment provides a method for predicting the collision risk between an electric bicycle and a vehicle based on the multi-modal behavior of the electric bicycle. From the perspective of the car, this method analyzes the safety impact on its driving caused by the multi-modal behavior of the electric bicycle and environmental factors, combines various sensing devices installed on the car (such as radar, camera) and vehicle networking technology (5G-V2X), and uses physical modeling. The basic risks are the impact of the braking distance, lateral and longitudinal stability of the electric bicycle on the car; further considering the risks of the car caused by the environment, a multi-dimensional environmental risk assessment model is built by comprehensively considering four dimensions: dynamic distance-speed, adjacent vehicle types, traffic flow density, and weather conditions (precipitation, temperature, wind speed, visibility); combining the risks of the car caused by the operation behavior of the electric bicycle (acceleration change, steering angle and frequency); using the environmental sensitivity coefficient to capture the complex relationship between environmental changes and vehicle risks in real time. Finally, for the multi-modal riding behavior of the electric bicycle, a comprehensive, accurate and real-time multi-dimensional comprehensive risk assessment method is provided for the car; this method details various situations that may be faced during the driving process of the car, and has strong practicability; at the same time, the model performs non-linear modeling for various risk scenarios, enhancing the generalization ability of the model, thereby effectively improving the driving safety of the vehicle and the road traffic efficiency.

[0079] Such as Figure 1As shown, a method for predicting the collision risk between an electric bicycle and a vehicle based on multi-modal behaviors of the electric bicycle includes the following steps:

[0080] Step 1. Decompose the overall vehicle risk of the car into the risk of the electric bicycle's inherent attributes to the car , the risk of the car caused by the environment and the risk of the electric bicycle's operation behavior to the car ;

[0081] Step 2. Evaluate the risk of the electric bicycle's inherent attributes to the car , the risk of the car caused by the environment and the risk of the electric bicycle's operation behavior to the car ;

[0082] Among them, the risk of the electric bicycle's inherent attributes to the car refers to the risk brought to the car by the electric bicycle's inherent attributes during the driving process of the car, which is a risk factor not affected by the external environment, including the risk brought to the car by the braking distance of the electric bicycle , the risk brought to the car by the stability of the electric bicycle ;

[0083] The risk of the car caused by the environment refers to the risk caused to the car by the change of external environmental factors during the driving process of the car, including the risk caused to the car by the dynamic distance - speed , the risk caused to the car by the type of adjacent vehicles , the risk caused to the car by the traffic flow density , the risk caused to the car by the weather condition ; Among them, the risk of the weather condition to the car includes the risk caused to the car by the precipitation , the risk caused to the car by the temperature , the risk caused to the car by the wind speed and the risk caused to the car by the visibility ;

[0084] The risk of the electric bicycle's operation behavior to the car refers to the risk brought to the driving safety of the car by the driving behavior of the rider during the driving process of the car, including the risk caused to the car by the change frequency of the electric bicycle's acceleration , the risk caused to the car by the steering angle and frequency of the electric bicycle ; Among them, the risk of the steering angle and frequency of the electric bicycle to the car includes the risk caused to the car by the steering angle of the electric bicycle , Risks posed by the turning frequency of electric bicycles to automobiles ;

[0085] Step 3. Predict the overall risk of the automobile based on the Capital Asset Pricing Model (CAPM), and at the same time introduce the environmental sensitivity coefficient , and evaluate the sensitivity of the overall risk of the automobile to environmental changes; in addition, considering that the environmental risk of the automobile and the behavioral characteristics of the electric bicycle will affect each other, add and The interaction term between them is used to describe the superimposed effect of the operation behavior of the electric bicycle on the overall risk of the automobile under different environmental conditions; moreover, in order to more accurately capture on the overall risk of the automobile A more detailed influence, add the high-order effect to reflect the non-linear trend of the environmental risk influence, which is applicable to the situation where the risk growth under harsh conditions is much greater than the linear expectation;

[0086] The overall risk of the automobile The prediction model is:

[0087]

[0088] Where ， is the interaction risk coefficient, and its value range is [0.1, 1.0], which is used to reflect and The combined effect between them. This interaction term can capture the amplification effect of the operation behavior of the electric bicycle on the automobile risk in complex driving scenarios, especially in high-risk environments such as bad weather or complex traffic; is the power of the high-order effect, which determines the speed and amplitude of the influence of environmental risk changes on the overall risk. Generally speaking The value ; under extreme environmental conditions, the risk is amplified in a high-order form, and the value can be greater than 2; is the high-order environmental sensitivity coefficient, and its value range is [0.05, 0.5], which is used to measure the amplification effect of larger environmental risk changes on the overall risk. This coefficient determines the increase in non-linear risk. For example, under extreme weather conditions, the risk increases faster.

[0089] Furthermore, in this embodiment, the risk posed by the braking distance of the electric bicycle to the automobile is evaluated as follows:

[0090] The braking distance refers to the distance required for the electric bicycle to come to a complete stop from the start of operating the brake pedal, which directly affects the driving risk of the automobile. In uniformly decelerated motion, the braking distance can be calculated by the kinematic formula:

[0091]

[0092] Wherein: is the final speed of the electric bicycle (when completely stopped ); is the driving speed of the electric bicycle, which can be obtained by Doppler speed measurement and visual displacement tracking with a radar installed on the vehicle; is the acceleration of the electric bicycle (negative value during braking, i.e., deceleration), which can be calculated based on the collected speed change rate; is the braking distance of the electric bicycle.

[0093] Substitute into the formula to obtain the braking distance of the electric bicycle as:

[0094]

[0095] Considering that in actual driving, the reaction time of the rider will significantly affect the braking distance. Before the rider reacts, the electric bicycle will continue to travel a certain distance at the initial speed , which can be expressed as:

[0096]

[0097] Wherein, is the reaction time of the rider, which can be taken as 0.3 - 1.0 seconds.

[0098] Considering that the traffic flow in front of the electric bicycle may trigger emergencies through unstable speed changes, which will increase the attention demand of the rider and thus affect the reaction time . To reflect the influence of the randomness of the environment on the reaction time of the rider, the modified reaction time can be defined as:

[0099]

[0100] Wherein: represents the sensitivity coefficient of the increase in the rider's attention when the electric bicycle approaches the vehicle, which can be taken as 0.5 - 2.0 according to the actual situation; represents the rate of change of the distance between the electric bicycle and the vehicle, that is, the ratio of the change in the distance between the electric bicycle and the vehicle to the change in time.

[0101] Combined with the reaction distance of the rider, the modified total braking distance can be expressed as:

[0102]

[0103] Among them ， is the distance value traveled within the corrected reaction time;

[0104] A positive correlation non-linear relationship between the risk of the vehicle and the braking distance of the electric bicycle is established, linking the braking distance with the risk. The expression is:

[0105] ;

[0106] Among them, is the risk brought by the braking distance of the electric bicycle to the vehicle, that is, considering the impact of the braking distance of the electric bicycle on the vehicle risk caused; is the proportionality coefficient, is the power exponent.

[0107] In this embodiment, historical accident data is used to optimize the parameters and in the model through the non-linear regression method. The value range of is [0.001, 0.1], The value range of is [0.5, 2.0]. Taking the mean square error as the objective function to measure the error between the model prediction and the observed data:

[0108]

[0109] Among them, is the number of observation samples, is the actual risk value of the th observation sample, is the risk of the th sample calculated according to the model.

[0110] When performing non-linear estimation, the initial values of and must be given first, which can be determined based on a simple linear regression method. The initial value of can be taken as 0.01, The initial value of can be taken as 1.0, and then the gradient descent method is used to update the parameters by taking the partial derivatives of the objective function:

[0111]

[0112]

[0113] Among them, is the updated value; is the value before update; is the updated value; is the value before update value; is the learning rate considering the impact of braking distance on vehicle risk, the step size for controlling parameter update, with a value range of [0.001, 0.1], represents MSE 1 the partial derivative of with respect to; represents MSE 1 the partial derivative of with respect to.

[0114] Furthermore, in this embodiment, the risk brought by the stability of the electric bicycle to the vehicle is evaluated as follows:

[0115] Stability reflects the ability of the electric bicycle to maintain balance and control under various operating conditions. An electric bicycle with poor stability is prone to losing control during operations such as steering, accelerating, and braking, thus posing a higher risk to the vehicle. In this embodiment, the stability in the transverse and longitudinal directions of the electric bicycle is mainly considered.

[0116] (1) Transverse stability:

[0117] The transverse stability of the electric bicycle is mainly reflected in the risk of rollover and is mainly affected by the following key physical factors:

[0118] Center of gravity position (height of the center of mass) : It can be obtained by judging the vehicle type through the visual recognition system installed on the vehicle, matching the default value in the vehicle design specifications, and estimating the weight of the rider. The value range is [0.8, 1.2] meters, and a conservative value can be taken. The center of gravity height of the electric bicycle has an important impact on its transverse rollover tendency during turning; the higher the center of gravity, the easier it is to roll over during high-speed turning.

[0119] Wheelbase : It refers to the horizontal distance between the center of the front wheel and the center of the rear wheel of the electric bicycle. It can be obtained by judging the vehicle type through the visual recognition system installed on the vehicle and matching the default value in the vehicle design specifications. The value range is [1.0, 1.5] meters, and a conservative value can be taken. The wheelbase has an important impact on its driving stability; at high speeds, the longer the wheelbase, the higher the stability of the electric bicycle.

[0120] Mass : It refers to the total mass of the electric bicycle and the rider. It can be estimated through the visual recognition system installed on the vehicle. The value range is [80, 150] kilograms, and a conservative value can be taken.

[0121] Travel speed : The current driving speed of the electric bicycle can be obtained by Doppler speed measurement and visual displacement tracking using the radar installed on the vehicle. Especially when turning, the higher the speed, the greater the centrifugal force and lateral acceleration the electric bicycle is subjected to.

[0122] Turning radius : The radius of curvature of the driving path of the electric bicycle when turning, which can be achieved by point cloud fitting path curvature using the radar sensor installed on the vehicle, with a value range of [2, 5] meters. The smaller the turning radius, the greater the centrifugal force of the electric bicycle when turning and the worse the stability.

[0123] Lateral stability index of electric bicycle Characterizes the risk of lateral rollover, mainly determined by the relationship between the centrifugal force and the stabilizing moment generated by gravity When the centrifugal force exceeds the stabilizing moment generated by gravity, the electric bicycle may roll over:

[0124]

[0125] Among them, is half of the wheelbase of the electric bicycle, that is, the distance from the center of mass of the electric bicycle to the steering axis; is the total gravity of the electric bicycle and the rider; is the centrifugal force of the electric bicycle and the rider, represents the acceleration due to gravity, with a value of 9.8 N / kg.

[0126] When , the stabilizing moment generated by gravity is greater than the centrifugal force, and the electric bicycle can maintain stability;

[0127] When , the centrifugal force exceeds the stabilizing moment generated by gravity, and the electric bicycle has a risk of rollover.

[0128] (2) Longitudinal stability:

[0129] The longitudinal stability of the electric bicycle mainly refers to its stability during acceleration and deceleration, especially the changes in load transfer and the grip of the front and rear wheels. Due to the relatively light weight, short wheelbase, and relatively high center of gravity of the electric bicycle, during dynamic driving, especially during rapid acceleration and braking, the grip distribution of the front and rear wheels is prone to significant changes, thereby affecting longitudinal stability. This has an impact on the driving safety of the vehicle.

[0130] Dynamic load transfer occurs during the acceleration or deceleration process of the electric bicycle, which is the core of longitudinal stability. The height of the vehicle's center of mass and the wheelbase determine the amount of load transfer between the front and rear wheels. Load transfer leads to changes in the normal forces of the front and rear wheels, thereby affecting the grip distribution.

[0131] When an electric bicycle accelerates (decelerates), the center of gravity moves backward (forward), the load on the front wheel (rear wheel) decreases, and the load on the rear wheel (front wheel) increases. The load transfer amount is defined as and can be calculated by the following formula:

[0132]

[0133] where is the acceleration component of the electric bicycle in the driving direction;

[0134] During acceleration, the load transfers from the front wheel to the rear wheel, the normal force on the front wheel decreases, resulting in insufficient grip of the front wheel and affecting stability. The greater the acceleration, the greater the load transfer amount and the more obvious the decrease in the grip of the front wheel. During emergency braking, the load on the rear wheel may decrease to nearly zero, causing the rear wheel to lose grip. Therefore, the load transfer during braking has a more significant impact on the longitudinal stability of the electric bicycle.

[0135] The load transfer directly affects the normal pressure of the front and rear tires on the ground, thus determining the grip of the vehicle. It is defined that the friction coefficients between the front and rear wheels and the ground are both the wheelbase of the front wheel (i.e., the horizontal distance from the front wheel axle to the center of mass) is the wheelbase of the rear wheel (i.e., the horizontal distance from the rear wheel axle to the center of mass) is the normal force on the front wheel is the normal force on the rear wheel is the grip of the front wheel is the grip of the rear wheel is then:

[0136]

[0137]

[0138]

[0139]

[0140] Taking into account the grip of the front and rear wheels comprehensively, the longitudinal stability can be defined as the relative proportion of the difference in the grip of the front and rear wheels. The following formula can be established to construct a risk model and define the longitudinal stability index of the electric bicycle characterizing the risk.

[0141]

[0142] where is a regulation coefficient used to reflect the impact of load transfer on risk, and its value range is [0.1, 1.0].

[0143] The lateral stability index of the electric bicycle and the longitudinal stability index Taking all factors into consideration, the stability of the electric bicycle is established In this embodiment, a non-linear combination method is used to comprehensively consider these two risk factors, and a non-linear index of the longitudinal and lateral stability risks on the stability of the electric bicycle is introduced to adjust the non-linear index of the overall combination form of the two risk terms and to adjust the weight coefficients of the lateral stability and longitudinal stability on the overall stability of the electric bicycle 、 as follows:

[0144]

[0145] where the parameter has a value range of [0.5, 2.0], has a value range of [0.5, 2.0], has a value range of [0.5, 3.0], has a value range of [0, 1], has a value range of [0, 1].

[0146] The stability of the electric bicycle is associated with the risk caused to the car Since the greater the stability of the electric bicycle, the lower the risk caused to the car, a negative correlation non-linear relationship between the car risk and the stability of the electric bicycle is established to link the stability and the risk, and the expression is:

[0147] ;

[0148] where is the risk brought by the stability of the electric bicycle to the car, that is, considering the impact of the stability of the electric bicycle on the car risk caused, is the proportionality coefficient, is the power exponent, the larger the smaller, and the higher the stability of the electric bicycle.

[0149] To better fit the above risk model, in this embodiment, using historical accident data, the parameters and in the model are optimized by the non-linear regression method has a value range of [0.001, 0.1], has a value range of [0.5, 2.0], with the mean square error is the objective function, measuring the error between the model prediction and the observed data:

[0150]

[0151] where, is the number of observed samples, is the actual risk value of the th observed sample, is the risk of the th sample calculated according to the model.

[0152] When performing nonlinear estimation, initial values of and should be given first, which can be determined based on a simple linear regression method. The initial value of can be taken as 0.01, and the initial value of

[0153]

[0154]

[0155] where, is the updated value; is the value before update; is the updated value; is the value before update; is the learning rate considering the impact of the electric bicycle stability on the vehicle risk, controlling the step size of parameter update, with a value range of [0.001, 0.1], represents the partial derivative of MSE2 with respect to ; represents the partial derivative of MSE2 with respect to ;

[0156] Furthermore, in this embodiment, considering that the risk brought by the braking distance of the electric bicycle to the vehicle may interact with the risk

[0157]

[0158] and and is a non-linear exponent, which is adjusted by an optimization algorithm according to historical data or experimental data ; The value range is [0.5, 2.0], The value range of is [0.5, 2.0], The value range of is [0.5, 3.0]. To minimize the error between the model prediction value and the actual value, so as to accurately fit the risk model. Under extreme driving conditions, can be appropriately reduced to enhance the effect of the non-linear combination, reflecting the situation where two risk terms may jointly amplify the vehicle risk in an emergency

[0159] Furthermore, in this embodiment, the risk caused by the dynamic distance-speed to the vehicle is evaluated as follows:

[0160] When the vehicle interacts with other vehicles in the driving environment, the driving risk of the vehicle is not only related to the distance between the vehicle and the adjacent vehicle, but also related to their relative speed

[0161] The closer the distance and the higher the relative speed, the greater the driving risk to the vehicle. Conversely, when the relative speed is low, even if the distance is close, the risk is relatively low First, it is necessary to determine the safety distance

[0162]

[0163] between the vehicle and other vehicles, which is defined as the minimum distance at which the first vehicle behind the vehicle can stop safely: is the current speed of the first vehicle behind; is the reaction time of the driver, which can be taken as 0.3 - 1.0 seconds; is the deceleration of the first vehicle behind;

[0164] According to the comparison between the distance between the vehicle and the first vehicle behind and the safety distance , the distance control risk can be defined; if is less than , the vehicle may be in a high-risk state; when is greater than , the risk is reduced; the specific quantification formula is as follows:

[0165]

[0166] where, , , It is a parameter for adjustment, used to regulate the non - linear degree of risk. Control the risk growth rate when the distance between two vehicles is less than the safe distance, which is usually related to the reaction time and braking ability of the following vehicle, and the value range is [0.1, 2.0]; Control the risk decay rate when the distance between two vehicles is greater than the safe distance, reflecting the risk decline when the distance between two vehicles exceeds the safe range, and the value range is [0.2, 1.5]; Control the risk decay degree when the distance exceeds the safe distance. This parameter can be adjusted to reflect different scenarios outside the safe distance, and the value range is [0.05, 0.5]. The determination of the three parameters can be based on fitting the existing data or obtained through simulation experiments.

[0167] Secondly, introduce the relative - speed risk coefficient to describe the additional impact of relative speed on risk:

[0168]

[0169] Among them, is the maximum value of the relative speed. Generally, 50 km / h can be taken on urban roads; represents the impact of relative speed on risk. As the speed increases, the risk rises significantly.

[0170] Finally, obtain the risk caused by the dynamic distance - speed to the vehicle :

[0171]

[0172] Among them, is the critical value of the relative speed, which can be determined according to factors such as traffic rules, driving behavior, vehicle characteristics, and data analysis. Generally, it can be set between ; is a minimum value, which can be taken as 10 -6 ~10 -9 , used to represent the risk that is almost zero.

[0173] Furthermore, in this embodiment, the evaluation method for the risk caused by adjacent vehicle types to the vehicle is as follows:

[0174] In complex environmental conditions, the types of adjacent vehicles in the same lane also affect the driving risk of a car. Especially when large vehicles (such as trucks and buses) and cars appear in the same lane, it will increase the attention burden of the car driver and amplify the accident risk; at the same time, there is also the possibility of causing a chain accident. The adjacent vehicle types are limited to the vehicles adjacent to the car in the same lane. Different types of vehicles pose different threats to the car, and thus the risks caused to the car are also different. Generally speaking, the larger the vehicle type, the heavier the mass, and the larger the blind spot, the higher the potential risk. According to these characteristics, different risk weights are assigned to different types of vehicles to reflect the collision threat they pose to the car.

[0175] For the detection range, it can be set to detect 50 - 150 meters in front of the car and 20 - 50 meters behind the car within the same lane. The specific values can be adjusted according to the recognition range of the camera and the requirements for risk calculation accuracy.

[0176] Regarding the calculation method of the vehicle type risk weight, let represent the vehicle type. Four vehicle types are considered in the model, and each type corresponds to a risk weight:

[0177] Non - motor vehicles: such as bicycles, tricycles, electric bicycles, etc.;

[0178] Small cars: usually refer to vehicle models with an engine displacement of about 1.0 - 1.3, such as sedans, electric cars, etc.;

[0179] Medium - sized cars: usually refer to vehicle models with an engine displacement of about 1.3 - 3.0, such as minivans, SUVs, etc.;

[0180] Large cars: usually refer to vehicle models with an engine displacement of more than 3.0, such as trucks, lorries, buses, etc.

[0181] Each variable defined according to physical characteristics can be integrated into a risk weight model through a linear combination. This model takes size, weight, and blind spot as independent risk factors, and uses weight coefficients 、 、 to control the influence degree of each factor. The formula for the risk weight model (the risk caused by adjacent vehicle types to the car) is as follows:

[0182]

[0183] Where: is the size of the vehicle type, measured by the vehicle volume, usually in cubic meters (m 3 ), which can be estimated by the camera for the external dimensions or accurately measured in combination with millimeter - wave radar; is the mass of the vehicle type, in tons (t). The vehicle appearance or type can be recognized through a camera to simply estimate the mass of the vehicle; is the blind spot size, in square meters (m²), representing the area that the driver cannot directly observe. Combining the vehicle type, it can be recognized through the camera which areas are blocked; 、 、 is the weight coefficient ( ), the value range is [0.2, 0.5], the value range is [0.3, 0.6], the value range is [0.1, 0.3], which is used to reflect the influence degree of each physical characteristic on the overall risk weight. The determination of the parameters needs to analyze the performance of different vehicle types in actual traffic accidents, fit the relationship between these three physical characteristics and the actual collision risk, and then estimate the contribution weight of each characteristic to the overall risk.

[0184] When calculating the risk caused by all vehicle types within the detection range to the vehicle, the vehicle with the highest risk is taken as the dominant risk source for predicting the overall vehicle risk.

[0185] Furthermore, in this embodiment, the risk caused by the traffic flow density to the vehicle is evaluated as follows:

[0186] The traffic flow density is a parameter describing the road traffic state at a certain moment, representing the number of vehicles passing through a certain traffic area per unit time, which will directly cause risks to the driving of the vehicle.

[0187] When the traffic flow density is very low, the distance between the vehicle and surrounding vehicles is large, the interaction between vehicles is less, and the driving space of the vehicle is sufficient, so the collision risk is low. However, if the traffic flow density is too low, the vehicle is likely to encounter other vehicles driving at high speed. In this case, the relative speed is large, which may instead increase the risks of both sides.

[0188] As the traffic flow density increases, the headway shrinks, the interaction frequency between the vehicle and other vehicles increases, and the complexity of driving operations improves. At this time, the risk also gradually increases. Usually, when the traffic flow density is medium, it is the moment when the vehicle risk is relatively large, because at this time the vehicles are all driving at a relatively high speed, but the distance between vehicles is not sufficient to provide enough reaction time.

[0189] In the case of high traffic flow density, the distance between vehicles becomes smaller, but usually the vehicle speed will also decrease accordingly. At this time, although the driving space between vehicles is compressed, the risk may not increase significantly because the vehicle driving speed is very slow and the consequences of collisions may not be so serious. Therefore, in the congested state, although the interaction frequency is high, the accident risk may not necessarily be higher.

[0190] To capture the non-linear relationship described above, a peak-shaped function is used to describe the relationship between traffic flow density and risk. It is assumed that the risk reaches its maximum value at and shows a normal distribution trend with the change of traffic flow. The risk function caused by traffic flow density to vehicles can be constructed in the form of a Gaussian distribution:

[0191]

[0192] Where: represents the traffic flow density, that is, the number of vehicles per unit length on the road at a certain moment, with the unit of ; represents the risk situation caused by traffic flow density to vehicles; refers to the point with the highest impact on vehicle driving risk under medium traffic flow density, and its value range is [20, 40] veh / km. At this time, the traffic flow density reaches a relatively high value, but it is not completely congested. The speeds of vehicles are relatively fast and the interactions are frequent, so collisions are likely to occur; it can be obtained by analyzing the accident peak based on medium traffic flow density, and find the traffic flow density value with the highest accident incidence rate as ; is the risk sensitivity coefficient in the low-density area, and its value range is [0.5, 1.5], is the risk sensitivity coefficient in the high-density area, and its value range is [0.2, 1.0]. According to the continuity condition, at at the point is equal to ; is the risk change rate in the low-density area, and its value range is [5, 10] veh / km, is the risk change rate in the high-density area, and its value range is [10, 20] veh / km. and The specific values can be determined by observing the accident incidence rates under low traffic flow density and high traffic flow density and then fitting; is an adjustment parameter, and its value range is [0.1, 0.5], which is used to regulate the overall risk level under low traffic flow density to reflect the potential risks existing under low traffic flow density in reality.

[0193] Furthermore, in this embodiment, the risk caused by the weather condition to vehicles is evaluated as follows:

[0194] To more comprehensively evaluate the risk situation caused by different weather conditions to vehicles, this embodiment mainly considers the risk impact of variable weather conditions on vehicles from four aspects: precipitation, temperature, wind speed, and visibility.

[0195] In terms of precipitation, the impact of the ground friction coefficient on risk is mainly considered. In the stage of small precipitation, due to the mixing of oil stains, dust on the road surface with a small amount of rainwater, a very slippery surface layer is formed, resulting in a sharp drop in the friction coefficient; in the stage of moderate precipitation, the oil stains on the road surface are gradually washed away by rainwater, and a water film is formed, and the friction coefficient continues to decrease, but more slowly than in the initial stage; in the stage of heavy precipitation, water accumulation may form on the road surface, and at this time the friction coefficient hardly changes anymore and tends to a stable low value. Based on this, a change model of the friction coefficient with precipitation is established:

[0196]

[0197] where is the amount of precipitation, with the unit of millimeter; is the road surface friction coefficient corresponding to different precipitation amounts; is the friction coefficient of the dry road surface, and the value range is [0.7, 0.9]; is the friction coefficient under moderate precipitation, and the value range is [0.4, 0.6]; is the lowest friction coefficient under heavy precipitation, and the value range is [0.15, 0.35]; and are the critical values of small, medium, and heavy precipitation, with the value range of [5, 15] mm / 24h, with the value range of [20, 30] mm / 24h; is the exponential decline coefficient in the initial precipitation stage, indicating the decline amplitude of the initial friction coefficient, and the value range is [0.1, 0.5] mm -1 ; is the logarithmic decline coefficient in the moderate precipitation stage, indicating the impact of the increase in precipitation on the friction coefficient, and the value range is [0.05, 0.2] mm -1 .

[0198] Once the friction factor changes with precipitation, we can construct a driving risk model through the relationship between the friction factor and the driving risk of the vehicle; the lower the friction factor, the higher the risk, and the risk is inversely proportional to the friction factor:

[0199]

[0200] where represents the impact of precipitation on the driving risk of the vehicle; represents the risk, and represents the minimum risk under dry conditions, which is a very small value and can take 10 -6 ~10 -9 ; The coefficient representing the increase in risk due to the decrease in the friction coefficient, indicating the degree of influence of the decrease in the friction coefficient on the risk, with a value range of [0.01, 0.5];

[0201] In summary, the piecewise function relationship between the precipitation and the vehicle risk is obtained as follows:

[0202] ;

[0203] Regarding temperature, the influence of ice and snow on the road surface under low-temperature conditions is mainly considered.

[0204] Under low-temperature conditions below zero, if there is snowfall, the road surface may freeze, resulting in a sharp drop in the friction coefficient, and at this time the driving risk of the vehicle increases:

[0205]

[0206] Among them: represents the temperature; represents whether there is ice and snow on the road surface. If it exists, take 1; if it does not exist, take 0, which can be obtained through camera recognition; represents the driving risk of the vehicle under low-temperature ice and snow conditions (the risk caused by temperature to the vehicle); represents the risk, which is the minimum risk under suitable temperature, and is a very small value, which can be taken as 10 -6 ~10 -9 ; represents the coefficient of risk increase due to temperature drop, indicating the degree of influence of temperature decrease on the risk, with a value range of [1.0, 5.0]; is the friction coefficient of the dry road surface, with a value range of [0.7, 0.9]; represents the temperature at which the road surface begins to freeze, with a value range of [-2, 0] degrees Celsius; The critical temperature for controlling the speed of friction coefficient decrease, with a value range of [2, 10] degrees Celsius. The lower the temperature, the faster the friction coefficient drops.

[0207] Regarding wind speed, crosswinds are likely to cause instability in the driving of vehicles with relatively small masses such as electric bicycles: when the crosswind speed is strong, the lateral torque of the electric bicycle will increase, and there is a risk of rollover or deviation from the lane, which will then affect the vehicle. First, a model for the risk caused by crosswinds to electric bicycles is established:

[0208]

[0209] Among them, represents the wind speed; represents the risk of electric bicycle rollover under the wind speed and the incident angle ; is the risk value, representing the minimum risk under windless conditions, and can take the value of 10 -6 ~10 -9 ; is the crosswind risk coefficient, with a value range of [0.5, 2.0], reflecting the influence degree of wind force on the rollover risk; is the lateral resistance coefficient, with a value range of [0.3, 1.2], determining the magnitude of the crosswind force on the electric bicycle; is the lateral windward area of the electric bicycle, affecting the lateral force generated by the wind on the electric bicycle; is the air density, usually taking the standard value of 1.225 kg / m³; is the height of the vehicle's center of mass. The higher the center of mass, the greater the rollover risk; represents the sine value of the crosswind incident angle, the closer it is to 90°, that is, the closer the wind is to being perpendicular to the driving direction of the electric bicycle, the greater the lateral force of the crosswind and the higher the rollover risk.

[0210] For a car, since the crosswind can cause the electric bicycle to lose control (roll over or deviate from the lane), thus affecting the driving safety of the car, a risk model transmitted to the car is constructed as:

[0211]

[0212] Among them, is the spatial interaction coefficient, with a value range of [1.0, 2.5], reflecting the possibility of the electric bicycle invading the car lane or colliding with the car after losing control, depending on the lane width and crosswind intensity; is the traffic flow coefficient, with a value range of [1.0, 3.0], indicating the amplification effect of the electric bicycle losing control on the traffic flow composed of cars.

[0213] In terms of visibility, an exponential decay model is constructed to depict the non-linear relationship between the car risk and visibility:

[0214]

[0215] Among them, represents the current visibility, with the unit of meter; represents the risk caused by visibility to the car; represents the basic risk value, that is, the minimum risk under good visibility, and can take the value of 10 -6 ~10 -9 ; represents the risk adjustment parameter, used to control the influence range of visibility reduction on the risk, with a value range of [0.5, 5.0]. The specific value can be determined by fitting the historical data analysis, reflecting the rising rate of risk when visibility decreases; the larger the value, the more significant the influence of visibility on the risk; Denote the critical visibility, with a value range of [50, 200] meters. When the visibility drops below this critical value, the risk will increase significantly, and it will be impossible to continue driving safely. The specific value can be estimated through historical visibility data and traffic accident data.

[0216] When the visibility is very high, such as on a sunny day or during the day, the risk function approaches the risk , indicating a low risk level; when the visibility drops to a medium level, the risk begins to increase rapidly, and the exponential term causes the risk to rise faster and faster; when the visibility continues to drop and approaches , due to the significant reduction in the driver's reaction time and predictability, the risk will increase significantly, and it is not recommended to continue driving at this time.

[0217] Integrate the four aspects of precipitation, temperature, wind speed, and visibility, considering that the four factors may interact with each other. To better handle the interaction and non-linear relationship between weather factors, in this embodiment, a weighted power combination method is used to construct a vehicle driving risk model under variable weather conditions:

[0218]

[0219] Among them, represents the risk caused by weather factors to the vehicle; , , , represent the weight coefficients of each factor, reflecting the contribution of different weather factors to the total risk; has a value range of [0.2, 0.4], has a value range of [0.1, 0.3], has a value range of [0.1, 0.3], has a value range of [0.2, 0.4]; and finally satisfies ; represents the power of the contribution of each risk to the overall risk, reflecting the non-linear influence of each factor, has a value range of [1.0, 2.0], has a value range of [1.5, 3.0], has a value range of [1.0, 2.5], has a value range of [0.5, 1.5]; among them, is the maximum value in.

[0220] In this embodiment, through the above calculations, the risk caused by the dynamic distance - speed to the vehicle, the risk 1. The risks posed by traffic flow density to vehicles 2. The risks posed by weather conditions to vehicles ; In order to better handle the interactions and non - linear relationships among various environmental factors, a weighted power combination method is adopted to construct a vehicle driving risk model in a complex environment. The expression is as follows:

[0221]

[0222] where , , , represent the weight coefficients of each risk factor, reflecting the contribution of different environmental factors to the total risk. The value range of is [0.25, 0.5], The value range of is [0.05, 0.2], ; represents the power of the contribution of each risk to the overall risk, reflecting the non - linear impact of each factor, The value range of is [1.5, 3.0], The value range of is [1.0, 2.0], is the maximum value in

[0223] Furthermore, in this embodiment, the risk of the electric bicycle operation behavior to the vehicle is evaluated as follows:

[0224] The risk of the electric bicycle operation behavior to the vehicle refers to the potential safety hazards to the vehicle caused by the instability and improper operations of the rider's driving behavior during the driving of the electric bicycle. It mainly includes the influence of key operation parameters such as acceleration change, steering angle and frequency. When the rider frequently performs operations such as accelerating, decelerating or making sharp turns, it will have a significant impact on the driving safety of the vehicle. Therefore, the risk of the electric bicycle operation behavior to the vehicle mainly evaluates the risk brought by the rider's driving behavior to the driving safety of the vehicle.

[0225] In this embodiment, the evaluation method of the risk caused by acceleration change to the vehicle is as follows:

[0226] Acceleration change is an important indicator to measure the stability of the operation behavior of electric bicycle riders, representing the frequency and degree of change of the acceleration or deceleration of the rider per unit time. Frequent and large acceleration changes usually mean that the rider's operation is relatively aggressive, which has a greater impact on the driving safety of the vehicle. By evaluating the acceleration change, potential driving risks can be effectively predicted.

[0227] To evaluate the impact of the acceleration change frequency of electric bicycles on vehicles, the acceleration change value of the electric bicycle, the time window, and the following distance are comprehensively considered, that is, the more drastic the acceleration change of the electric bicycle rider in a shorter time and the closer the following distance, the higher the operation risk.

[0228]

[0229] Among them, is the acceleration change rate; represents time, that is, the time window when the operation occurs, and the value range is [5, 30] seconds. Too short will introduce noise, and too long may ignore instantaneous risks; represents the time within which the number of acceleration or deceleration operations; is the acceleration change amount of each operation; is the time interval of each operation; is the following distance between the vehicle and the electric bicycle during each operation.

[0230] Considering that the impact of acceleration change on the stability of electric bicycles is different at different vehicle speeds, and thus the risks caused to vehicles are also different. Therefore, the speed of the electric bicycle is introduced to make the acceleration change frequency associated with the speed:

[0231]

[0232] Among them, is the corrected acceleration change rate; is the adjustment coefficient to control the amplification effect of speed on the acceleration change frequency, and the value range is [0.5, 1.5].

[0233] A high-order polynomial is used to establish the relationship between the acceleration change frequency and the vehicle risk :

[0234]

[0235] Among them, , , represent the weights of different acceleration change frequency bands, which can flexibly respond to the risk changes of the acceleration change frequency in different frequency bands, The value range is [0.1, 0.5]. The value range is [0.05, 0.3]. The value range is [0.01, 0.2].

[0236] In this embodiment, the method for evaluating the risk caused by the steering angle and frequency to the vehicle is as follows:

[0237] The steering angle reflects the deflection amplitude of the handlebar during the steering process of the electric bicycle, and the steering frequency reflects the handling stability of the rider during the vehicle driving process. A larger steering angle usually means a sharp turn, especially at high speeds, which will increase the driving risk of adjacent vehicles. Frequent steering operations may indicate that the driver frequently changes lanes or avoids in a complex traffic environment, increasing the complexity and uncertainty of the operation.

[0238] First, consider the impact of the steering angle of the electric bicycle on the vehicle risk. The steering angle of the electric bicycle on the vehicle risk is related to the square of the absolute value of the steering angle and increases exponentially with the driving speed of the electric bicycle and decreases with the increase of the distance between the vehicle and the electric bicycle. The expression is:

[0239]

[0240] where is the steering angle of the electric bicycle, with the unit of and can be obtained by using the object detection and tracking technology through the on-vehicle camera; is the constant coefficient for adjusting the influence degree of the steering angle on the risk, and the value range is [0.2, 0.8], which can be set by data fitting or experience; is the speed amplification coefficient, which controls the exponential influence of the speed on the steering angle risk, and the value range is [0.1, 0.5], which can be set by data fitting or experience; is the distance between the vehicle and the electric bicycle, which can be obtained by the on-vehicle radar.

[0241] Similarly, define the steering frequency as the number of times the rider turns the handlebar per unit time, and a relational expression between the steering frequency of the electric bicycle and the vehicle risk can be established:

[0242]

[0243] where represents time, that is, the time window for observing the steering frequency, and the value range is [5, 30] seconds. Too short will introduce noise, and too long may ignore the instantaneous risk; The coefficient representing the degree of influence of the steering frequency on the risk, with a value range of [0.1, 0.4], which can be set by data fitting or experience; is the number of steering operations within unit time; It should be noted that ; is the critical steering angle, with a value range of [10°, 30°]; Only the steering is counted. If , the corresponding steering is ignored; represents the adjustment coefficient of the amplification effect of speed on the acceleration change frequency, with a value range of [0.5, 1.5], which can be set by data fitting or experience; is the distance between the vehicle and the electric bicycle, which can be obtained through on-vehicle radar.

[0244] Considering the risk interaction between the steering angle and frequency, in the final vehicle risk model, the non-linear superposition form of the two is used to obtain the relationship between the steering angle and frequency and the vehicle risk as follows:

[0245]

[0246] where is the interaction risk term coefficient of the steering angle and frequency, with a value range of [0.05, 0.2], reflecting the influence degree of the superposition effect of the steering angle and steering frequency on the operation risk under a certain driving speed.

[0247] Finally, the risk of the electric bicycle's operation behavior on the vehicle is:

[0248]

[0249] where 、 is the weight coefficient, used to adjust the mutual influence between different risk terms, with a value range of [0.4, 0.7]; with a value range of [0.3, 0.6]; , are non-linear exponents, used to adjust the non-linear growth amplitude of each risk term under different operation conditions, with a value range of [1.0, 2.0]; with a value range of [1.5, 3.0]; with a value range of [1.0, 2.0].

[0250] In this embodiment, the environmental risk sensitivity coefficient is introduced to measure the vehicle risk caused by the environment to the overall vehicle risk It reflects the fluctuation of vehicle risk when external environmental conditions (dynamic distance-speed, adjacent vehicle type, traffic flow density, weather conditions) change. The contribution of environmental changes to the risk of the entire vehicle can be quantified, allowing the risk assessment model to more accurately capture the potential impact of the external environment on vehicle risks. The initial value ranges from [0.05, 0.5]. It can be determined by preliminarily evaluating the contribution of environmental risk to the risk of the entire vehicle based on actual conditions. Subsequently, it can be iterated and trained continuously based on existing data to eventually obtain the optimal value under different traffic scenarios.

[0251] Specifically, calculate the automotive risks caused by the environment The variance can reflect the volatility of environmental changes:

[0252]

[0253] in, It is Environmental risk value at a time point; is the average value of the automobile risk caused by the environment; is the number of time series of data.

[0254] Calculate vehicle risk Automotive risks to the environment The covariance of can reflect the joint volatility of overall risk and environmental risk:

[0255]

[0256] in, It is The risk value of the vehicle at a point in time; is the average value of the vehicle risk; from this, the environmental sensitivity coefficient can be obtained The calculation formula is:

[0257]

[0258] Among them, the covariance Indicates the risk of the entire vehicle Automobile risks caused by the environment If the covariance value is high, it means that environmental changes significantly affect the overall risk; It indicates the fluctuation range of environmental risk. The larger the variance, the more drastic the change in environmental conditions.

[0259] The present invention also provides an electronic device, including: one or more processors and a memory; wherein, the memory is used for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method for predicting the collision risk with a vehicle based on the multimodal behavior of an electric bicycle as described above.

[0260] The present invention also provides a computer-readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for predicting the collision risk with a vehicle based on the multimodal behavior of an electric bicycle as described above is implemented.

[0261] Those skilled in the art can understand that all or part of the functions of the above-mentioned various methods / modules can be implemented in a hardware manner or in a computer program manner. When all or part of the functions in the above-mentioned embodiments are implemented in a computer program manner, the program can be stored in a computer-readable storage medium, and the storage medium may include: read-only memory, random access memory, magnetic disk, optical disk, hard disk, etc. The above functions are implemented by a computer executing the program. For example, the program is stored in the memory of the device, and when the processor executes the program in the memory, the above-mentioned all or part of the functions can be implemented.

[0262] In addition, when all or part of the functions in the above-mentioned embodiments are implemented in a computer program manner, the program can also be stored in a storage medium such as a server, another computer, magnetic disk, optical disk, flash drive or mobile hard disk, and is saved to the memory of the local device by downloading or copying, or the system of the local device is updated. When the processor executes the program in the memory, all or part of the functions in the above-mentioned embodiments can be implemented.

[0263] The above uses specific examples to elaborate on the present invention, which is only used to help understand the present invention and is not intended to limit the present invention. For those skilled in the technical field to which the present invention belongs, based on the idea of the present invention, several simple deductions, deformations or substitutions can also be made. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for predicting the risk of collision between an electric bicycle and a vehicle based on its multimodal behavior, characterized in that: The method comprises the following steps: Step 1. Decompose the vehicle risk R into the risk R of the electric bicycle's inherent attributes on the vehicle f 、Automotive risks caused by the environment m and the risk of electric bicycle operation to cars v ; Step 2. Evaluate the risk R of the inherent attributes of electric bicycles to cars f 、Automotive risks caused by the environment m and the risk of electric bicycle operation to cars v ; Among them, the inherent attributes of electric bicycles pose a risk to cars. f It refers to the risk that the inherent properties of electric bicycles bring to the car during driving, including the risk R caused by the braking distance of electric bicycles. f-B , the risks that electric bicycle stability brings to cars f-S ; Among them, α1 is the proportional coefficient, β1 is the power exponent, t r ' is the corrected reaction time of the rider, u is the speed of the electric bicycle, and a is the acceleration of the electric bicycle; Among them, α2 is the proportional coefficient, β2 is the power exponent, S h is the lateral stability index of the electric bicycle, S z is the longitudinal stability index of the electric bicycle, ω1 and ω2 are weight coefficients, and γ1, γ2, and γ3 are nonlinear exponents; in, It is half of the wheelbase of the electric bicycle, that is, the distance from the center of mass of the electric bicycle to the steering axis; g represents the acceleration due to gravity, h cg is the center of gravity, u is the driving speed, and r is the turning radius; F grip,f =μ·F normal，f ; F grip,r =μ·F normal，r ; Among them, the normal force of the front wheel is F normal，f ; The normal force on the rear wheel is F normal，r ; The front wheel's grip is F grip,f ; The grip of the rear wheel is F grip,r ; The load transfer is defined as ΔF x ; k is an adjustment coefficient used to reflect the impact of load transfer on risk, with a value range of [0.1, 1.0]; a x is the acceleration component of the electric bicycle in the direction of travel; the friction coefficient between the front and rear wheels and the ground is μ; the front wheel wheelbase is l f ; The rear wheelbase is l r ; The total mass of the electric bicycle and the rider is m; Risks of electric bicycles to cars f The evaluation model is: Among them, θ1, θ2, and θ3 are nonlinear exponents. The value range of θ1 is [0.5, 2.0], the value range of θ2 is [0.5, 2.0], and the value range of θ3 is [0.5, 3.0]. Environmental risks to automobiles m It refers to the risk caused to the car by changes in external environmental factors during the driving process, including the risk R caused by dynamic distance-speed distance (v, d), the risk R caused by the adjacent vehicle type to the car vehicle (T v ), the risk R caused by traffic flow density to cars density (D) Risks of weather conditions to cars R wx ; When evaluating the risk posed by dynamic distance-speed to the car, the relative speed and distance between the car and the first car behind it are comprehensively considered; when evaluating the risk posed by adjacent vehicle types to the car, the risk posed by all vehicle types within the detection range to the car is calculated, and the highest risk is taken for the risk prediction of the whole car; when evaluating the risk posed by traffic flow density to the car, the risk function posed by traffic flow density to the car is constructed in the form of Gaussian distribution; when evaluating the risk posed by weather conditions to the car, the risk posed by precipitation, temperature, wind speed and visibility to the car is comprehensively considered, that is, the risk posed by weather conditions to the car includes the risk R posed by precipitation to the car rain (P), Risks of Temperature to Cars R ice (T) Risk of wind speed to cars R wind (v wind ,θ) and the risk R posed by visibility to the car vis (V); Risk of wind speed to cars R wind (v wind , the evaluation model of θ) is: R wind (v wind ,θ)=R wind-b (v wind ,will wind-interaction ·b wind-interaction ; Among them, α wind-interaction is the spatial interaction coefficient, reflecting the possibility of an electric bicycle occupying a car lane or colliding with a car after losing control, which depends on the lane width and crosswind intensity; β wind-interaction is the traffic flow coefficient, which indicates the amplification effect of the out-of-control electric bicycle on the traffic flow composed of cars. wind-b (v wind , θ) represents the wind speed v wind and the rollover risk of the electric bicycle under the condition of the incident angle θ; Among them, R w-o is the risk value, indicating the minimum risk under windless conditions, k side_wind Crosswind risk coefficient, value range [0.5, 2.0]; C d is the lateral drag coefficient, with a range of [0.3, 1.2]; A is the windward area of ​​the side of the electric bicycle, ρ is the air density, and sinθ represents the sine value of the side wind incident angle; Environmental risks to automobiles m The evaluation model is: in, represents the weight coefficient of each risk factor, The value range is [0.25, 0.5], The value range is [0.05, 0.2], The value range is [0.15, 0.3], The value range is [0.2, 0.4]; and finally satisfies j1, j2, j3, j4 represent the power of each risk's contribution to the overall risk. The value range of j1 is [1.5, 3.0], the value range of j2 is [1.0, 2.0], the value range of j3 is [1.2, 2.5], and the value range of j4 is [1.5, 3.0]. max(j1, j2, j3, j4) is the maximum value among j1, j2, j3, j4. Risks of electric bicycle operation to vehicles v It refers to the risk of the driver's driving behavior to the driving safety of the car during the driving process, including the risk of the acceleration change frequency of the electric bicycle to the car. v-a , Risks of electric bicycle steering angle and frequency to cars R v-θ ; Among them, δ1, δ2, and δ3 represent the weights of different acceleration change frequency bands. is the corrected acceleration rate of the electric bicycle, f a is the acceleration change rate of the electric bicycle, u is the speed of the electric bicycle, and γ is the adjustment coefficient of the control speed on the amplification effect of the acceleration change frequency; Among them, θ z is the steering angle of the electric bicycle, β θ-rad To adjust the steering angle θ z The constant coefficient of the risk impact, λ is the speed amplification factor, which controls the exponential impact of speed on the steering angle risk, d c-b is the distance between the car and the electric bicycle, β θ-f The coefficient representing the influence of the steering frequency on the risk, TN represents the time, that is, the time window for observing the steering frequency, n θ is the number of steering operations per unit time TN; γ a Indicates the adjustment coefficient of the speed on the acceleration change frequency amplification effect, β Ix is the interaction risk term coefficient between steering angle and frequency; Risks of electric bicycle operation to vehicles v The evaluation model is: Among them, α u1 , α u2 is the weight coefficient, α u1 The value range is [0.4, 0.7], α u2 The value range is [0.3, 0.6]; p1, p2, and p3 are nonlinear exponents, the value range of p1 is [1.0, 2.0]; the value range of p2 is [1.5, 3.0]; the value range of p3 is [1.0, 2.0]; Step 3: Predict the risk of the whole vehicle based on the capital asset pricing model, and introduce the environmental sensitivity coefficient β to evaluate the sensitivity of environmental changes to the risk of the whole vehicle; in addition, add R m and R v The interaction term between is used to describe the superimposed effect of electric bicycle operation behavior on vehicle risk under different environmental conditions; the high-order effect is added to reflect the nonlinear trend of environmental risk impact; The prediction model of vehicle risk R is: R=R f +β·(R m -R f )+b γ ·(R m -R f ) τ +b interaction ·(R m ·R v )+R v ; Among them, β interaction is the interaction risk coefficient, with a value range of [0.1, 1.0], which reflects the R m and R v The joint effect between them; τ is the power of the higher-order effect, which determines the speed and magnitude of the impact of environmental risk changes on the overall risk, and τ takes the value of [1, 2]; β γ is the high-order environmental sensitivity coefficient, and its value range is [0.05, 0.5].

2. A method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: Dynamic distance-speed risk to cars R distance The evaluation model of (v,d) is: Among them, R v (v) is the relative speed risk coefficient, v threshold is the critical value of relative speed, k1, k2, k3 are adjustment parameters, v is the relative speed between the car and the first car behind, d is the distance between the car and the first car behind, and d safe is the safe distance between the car and the first car behind it, ∈ is a minimum value, taking 10 -6 ~10 -9 .

3. The method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: Risk R posed to the car by the type of adjacent vehicles vehicle (T v ) is: R vehicle (T v )=χ1·d T +χ2·m T +χ3·b T ; Among them, χ1, χ2, χ3 are weight coefficients, d T is the size of the vehicle, m T is the mass of the vehicle model, b T is the blind area size.

4. The method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: The risk R posed by traffic density to cars density The evaluation model of (D) is: Where D represents the traffic flow density, D peak It refers to the point with the highest impact on vehicle driving risk under medium traffic density, k l is the risk sensitivity coefficient of the low-density area, k d is the risk sensitivity coefficient of high-density area, σ l is the risk change rate in the low-density area, σ d is the risk change rate in high-density areas, and Δ is an adjustment parameter with a value range of [0.1, 0.5], which is used to regulate the overall risk level under low traffic flow density.

5. The method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: Risks to cars from weather conditions wx The evaluation model is: Among them, λ1, λ2, λ3, and λ4 represent the weight coefficients of each factor, reflecting the contribution of different weather factors to the total risk; the value range of λ1 is [0.2, 0.4], the value range of λ2 is [0.1, 0.3], the value range of λ3 is [0.1, 0.3], and the value range of λ4 is [0.2, 0.4]; and finally λ1+λ2+λ3+λ4=1 is satisfied; i1, i2, i3, and i4 represent the power of each risk's contribution to the overall risk, reflecting the nonlinear influence of each factor, the value range of i1 is [1.0, 2.0], the value range of i2 is [1.5, 3.0], the value range of i3 is [1.0, 2.5], and the value range of i4 is [0.5, 1.5]; among them, max(i1, i2, i3, i4) is the maximum value among i1, i2, i3, and i4.

6. The method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: Risk of precipitation to cars rain The evaluation model of (P) is: Among them, R p-o Indicates the minimum risk under dry conditions, which is a minimum value of 10 -6 ~10 -9 ;k μ-risk The coefficient of risk increase caused by the decrease of friction factor, P is the amount of precipitation, μ dry is the friction coefficient of dry road surface, μ wet is the friction coefficient under moderate precipitation, μ low is the minimum friction coefficient under heavy precipitation, P1 and P2 are the critical values ​​of light, medium and heavy precipitation, s1 is the exponential decrease coefficient in the initial precipitation stage, indicating the magnitude of the decrease in the initial friction coefficient, and s2 is the logarithmic decrease coefficient in the medium precipitation stage, indicating the effect of increased precipitation on the friction coefficient; Risks of temperature to cars ice The evaluation model of (T) is: Where o represents the temperature; ψ represents whether there is ice or snow on the road, which is 1 if there is ice or 0 if there is no ice or snow. T-o It represents the minimum risk under suitable temperature, which is a minimum value, taking 10 -6 ~10 -9 ;k T-risk The coefficient indicating the increase in risk due to temperature drop, indicating the degree of impact of temperature drop on risk, μ dry is the friction coefficient of dry road surface, T freeze Indicates the temperature at which the road surface begins to freeze, T critical The critical temperature that controls the rate at which the friction coefficient decreases; Risk of visibility to cars vis The evaluation model of (V) is: Among them, V represents the current visibility, R vis-o It represents the basic risk value, that is, the minimum risk under good visibility, and is set to 10 -6 ~10 -9 ;k v represents the risk adjustment parameter, which is used to control the impact of reduced visibility on risk, V critical Indicates critical visibility.

7. The method for predicting the risk of collision between an electric bicycle and a vehicle based on the multimodal behavior of an electric bicycle according to claim 1, characterized in that: The initial value of the environmental sensitivity coefficient β ranges from [0.05, 0.5]. Subsequent iterative training is performed based on existing data to finally obtain the optimal value under different traffic scenarios. The calculation formula of the environmental sensitivity coefficient β is: Among them, Cov(R, R m ) is the vehicle risk R and the vehicle risk R caused by the environment m The covariance of the vehicle risk R and the vehicle risk R caused by the environment m Var(R m ) is the risk of automobiles to the environment m The variance of represents the fluctuation range of environmental risk.

Images