Ground-air collision early warning method and system based on multi-dimensional risk assessment and storage medium

Abstract

The invention discloses a ground-air collision early warning method and system based on multi-dimensional risk assessment, and a storage medium. The method comprises the steps of obtaining a vehicle envelope box size parameter, a driving state vector, an air object landing state vector and a position covariance matrix; calculating the three-dimensional space-time envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector, and calculating the position probability distribution of the air object according to the landing state vector and the position covariance matrix; determining a multi-dimensional collision risk quantitative index according to the three-dimensional space-time envelope volume distribution and the position probability distribution; and calculating a ground-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantitative index so as to carry out ground-air collision early warning prompt. According to the invention, through the multi-dimensional collision risk quantitative index, the complexity of collision calculation is reduced, the risk value is discretized to generate the progressive early warning instruction, and more accurate and reliable safety early warning information is provided for the ground vehicle in the ground-air heterogeneous scene.

Description

Technical Field

[0001] The present invention relates to the technical field of ground-air collision warning, and particularly relates to a ground-air collision warning method, system and storage medium based on multi-dimensional risk assessment. Background Art

[0002] With the rapid development of the low-altitude economy, the application of aircraft such as unmanned aerial vehicles and urban air vehicles in the fields of urban logistics, emergency rescue and personal travel has become increasingly popular. However, due to factors such as aircraft equipment failure, operation error or environmental interference, the accidental fall of airborne objects poses a serious safety threat to ground vehicles.

[0003] Existing research mainly focuses on collision risk assessment for homogeneous targets (i.e., "ground-ground" or "air-air"). However, ground collision warning is limited by two-dimensional plane risk assessment and cannot cope with the three-dimensional threat caused by the fall of airborne objects; while air collision warning, although using three-dimensional route planning, relies on the symmetry assumption of homogeneous targets, and its collision probability model needs to perform convolution calculation on the probability density function of the two moving bodies, resulting in a relatively high computational complexity of O(n 2 ) and does not conform to the planar motion characteristics of ground vehicles if the target has an omnidirectional avoidance ability. Therefore, how to provide more accurate and reliable safety warning information for vehicles in the complex environment of ground-air traffic integration to improve the driving safety and active protection ability of vehicles has become an urgent problem to be solved.

[0004] The above content is only used to assist in understanding the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of the present invention is to provide a ground-air collision warning method, system and storage medium based on multi-dimensional risk assessment, aiming to solve the technical problem of how to provide more accurate and reliable safety warning information for vehicles in the complex environment of ground-air traffic integration to improve the driving safety and active protection ability of vehicles.

[0006] To achieve the above purpose, the present invention provides a ground-air collision warning method based on multi-dimensional risk assessment, and the ground-air collision warning method based on multi-dimensional risk assessment includes:

[0007] Obtain the envelope box size parameters and driving state vectors of the vehicle within the prediction time domain, and the landing state vector and position covariance matrix of the airborne object;

[0008] Calculate the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameters and the driving state vectors, and calculate the position probability distribution of the airborne object according to the landing state vector and the position covariance matrix;

[0009] Determine a multi-dimensional collision risk quantification index based on the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object;

[0010] Based on the multi-dimensional collision risk quantification index, calculate the ground-air collision risk evaluation value through a multi-dimensional risk evaluation function, and give a ground-air collision warning prompt for the vehicle according to the collision risk evaluation value.

[0011] Optionally, the step of calculating the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector includes:

[0012] Determine the angular velocity of the vehicle according to the driving state vector;

[0013] Determine the vehicle position coordinates, heading angle and linear velocity according to the angular velocity through the vehicle state transition equation;

[0014] Construct a vehicle position vector according to the vehicle position coordinates, the heading angle and the linear velocity;

[0015] Calculate the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the vehicle position vector and the envelope box size parameter.

[0016] Optionally, the step of calculating the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix includes:

[0017] Perform position prediction on the aerial object through an unscented Kalman filter based on the landing state vector to obtain the position mean vector of the aerial object;

[0018] Determine the random position vector of the aerial object;

[0019] Construct a position probability density function in the form of a multi-dimensional normal distribution according to the position mean vector, the random position vector and the position covariance matrix;

[0020] Calculate the position probability distribution of the aerial object through the position probability density function.

[0021] Optionally, the step of determining a multi-dimensional collision risk quantification index based on the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object includes:

[0022] Determine the target ground-air collision probability according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object, and calculate the minimum relative distance between the aerial object and the vehicle within the prediction time domain and the minimum relative time corresponding to the minimum relative distance;

[0023] Determine a multi - dimensional collision risk quantification index based on the target ground - air collision probability, the minimum relative distance, and the minimum relative time.

[0024] Optionally, the step of determining the target ground - air collision probability according to the three - dimensional spatio - temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object includes:

[0025] Integrate the three - dimensional spatio - temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object to obtain multiple ground - air collision probabilities within the prediction time domain;

[0026] Select the maximum ground - air collision probability from the multiple ground - air collision probabilities, and use the maximum ground - air collision probability as the target ground - air collision probability.

[0027] Optionally, the step of calculating the ground - air collision risk evaluation value through a multi - dimensional risk evaluation function based on the multi - dimensional collision risk quantification index includes:

[0028] Obtain the current linear velocity and the static safety distance of the vehicle;

[0029] Construct a multi - dimensional risk evaluation function according to the multi - dimensional collision risk quantification index, the current linear velocity, and the static safety distance, and calculate the ground - air collision risk evaluation value through the multi - dimensional risk evaluation function.

[0030] Optionally, the step of giving a ground - air collision warning prompt to the vehicle according to the collision risk evaluation value includes:

[0031] Map the collision risk evaluation value to discrete risk levels to determine the current collision risk level;

[0032] Determine a warning prompt strategy according to the current collision risk level, and give a ground - air collision warning prompt to the vehicle based on the warning prompt strategy.

[0033] In addition, to achieve the above - mentioned purpose, the present invention also proposes a ground - air collision warning system based on multi - dimensional risk assessment. The ground - air collision warning system based on multi - dimensional risk assessment includes:

[0034] An acquisition module, configured to acquire the envelope box size parameters and the driving state vector of the vehicle within the prediction time domain, and the landing state vector and the position covariance matrix of the aerial object;

[0035] A calculation module, configured to calculate the three - dimensional spatio - temporal envelope volume distribution of the vehicle according to the envelope box size parameters and the driving state vector, and calculate the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix;

[0036] The calculation module is further configured to determine a multi-dimensional collision risk quantification index according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object;

[0037] The warning module is configured to calculate a ground-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and perform a ground-air collision warning prompt on the vehicle according to the collision risk evaluation value.

[0038] In addition, to achieve the above object, the present invention further provides a ground-air collision warning device based on multi-dimensional risk assessment, the device includes: a memory, a processor, and a ground-air collision warning program based on multi-dimensional risk assessment stored on the memory and operable on the processor, the ground-air collision warning program based on multi-dimensional risk assessment is configured to implement the steps of the ground-air collision warning method based on multi-dimensional risk assessment as described above.

[0039] In addition, to achieve the above object, the present invention further provides a storage medium, on which a ground-air collision warning program based on multi-dimensional risk assessment is stored, and when the ground-air collision warning program based on multi-dimensional risk assessment is executed by a processor, it implements the steps of the ground-air collision warning method based on multi-dimensional risk assessment as described above.

[0040] The present invention first obtains the envelope box size parameters and driving state vectors of the vehicle within the prediction time domain, and the landing state vector and position covariance matrix of the aerial object, then calculates the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameters and driving state vectors, and calculates the position probability distribution of the aerial object according to the landing state vector and position covariance matrix. After that, a multi-dimensional collision risk quantification index is determined according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object. Finally, a ground-air collision risk evaluation value is calculated through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and a ground-air collision warning prompt is performed on the vehicle according to the collision risk evaluation value. The present invention reduces the computational complexity from O(n 2 ) to O(n) through the single integral of the three-dimensional spatio-temporal envelope volume distribution of the ground vehicle and the position probability distribution of the aerial object, constructs a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, solves the defect of insufficient dynamic adaptability of the existing method, and finally discretizes the risk value to generate a progressive warning instruction to support the vehicle hierarchical response strategy, providing more accurate and reliable safety warning information for ground vehicles in ground-air heterogeneous scenarios. Description of the Drawings

[0041] Figure 1 is a schematic structural diagram of a ground-air collision warning device based on multi-dimensional risk assessment in the hardware operating environment of the embodiment solution of the present invention;

[0042] Figure 2 Schematic flowchart of the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention;

[0043] Figure 3 Schematic diagram of the force analysis of the flying object in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention;

[0044] Figure 4 Schematic diagram of the density function of the position probability distribution of the aerial object in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention;

[0045] Figure 5 Schematic diagram of the calculation of the ground-air collision probability in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention;

[0046] Figure 6 Block diagram of the structure of the first embodiment of the ground-air collision warning system based on multi-dimensional risk assessment of the present invention.

[0047] The realization, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0048] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0049] Refer to Figure 1 , Figure 1 Schematic diagram of the structure of the ground-air collision warning device based on multi-dimensional risk assessment in the hardware operating environment involved in the embodiment solution of the present invention.

[0050] As Figure 1As shown in the figure, the ground-air collision warning device based on multi-dimensional risk assessment may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard). Optionally, the user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed Random Access Memory (RAM) or a stable Non-Volatile Memory (NVM), such as a disk memory. Optionally, the memory 1005 may also be a storage system independent of the aforementioned processor 1001.

[0051] Those skilled in the art can understand that Figure 1 the structure shown in does not constitute a limitation on the ground-air collision warning device based on multi-dimensional risk assessment, and may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0052] As Figure 1 shown, in the memory 1005 as a storage medium, there may be included an operating system, a network communication module, a user interface module, and a ground-air collision warning program based on multi-dimensional risk assessment.

[0053] In Figure 1 the ground-air collision warning device based on multi-dimensional risk assessment shown in the figure, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the ground-air collision warning device of the present invention may be arranged in the ground-air collision warning device based on multi-dimensional risk assessment. The ground-air collision warning device based on multi-dimensional risk assessment calls the ground-air collision warning program stored in the memory 1005 through the processor 1001 and executes the ground-air collision warning method provided by the embodiments of the present invention.

[0054] The embodiments of the present invention provide a ground-air collision warning method based on multi-dimensional risk assessment. Referring to Figure 2 , Figure 2 is a schematic flowchart of the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention.

[0055] In this embodiment, the ground-air collision warning method based on multi-dimensional risk assessment includes the following steps:

[0056] Step S10: Obtain the envelope box size parameters and driving state vector of the vehicle within the prediction time domain, as well as the landing state vector and position covariance matrix of the aerial object.

[0057] It is easy to understand that the execution subject of this embodiment can be a ground-air collision warning system based on multi-dimensional risk assessment with functions such as data processing, network communication, and program operation, or other computer devices with similar functions. This embodiment does not impose any restrictions.

[0058] It should be noted that the driving state vector x(t0) of the vehicle at the current moment is obtained through vehicle-mounted sensors (such as GPS, IMU, etc.):

[0059] x(t0) = [X(t0) Y(t0) ψ(t0) v(t0) ω(t0)] T

[0060] In the formula, x(t0) is the state vector at the current moment t0, X(t0) is the X coordinate of the position of the vehicle's center of gravity in the global coordinate system at the current moment t0, Y(t0) is the Y coordinate of the position of the vehicle's center of gravity in the global coordinate system at the current moment t0, ψ(t0) is the heading angle at the current moment t0, v(t0) is the linear velocity at the current moment t0, and ω(t0) is the angular velocity at the current moment t0.

[0061] The envelope box size parameters of the vehicle include length L, width W, and height H.

[0062] It also should be noted that assuming the current time t0 corresponds to time step k, the state estimation of the aerial object is performed through the Unscented Kalman Filter (UKF) to obtain the landing state vector and covariance matrix

[0063]

[0064] where, x g,k , y g,k , z g,k are the global coordinate system position coordinates of the aerial object at time step k, v xg,k , v yg,k , v zg,k are the velocity components, and δ k is the parameter related to air resistance.

[0065] Step S20: Calculate the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector, and calculate the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix.

[0066] Further, the method for calculating the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector is to determine the angular velocity of the vehicle according to the driving state vector; determine the vehicle position coordinates, heading angle and linear velocity through the vehicle state transition equation according to the angular velocity; construct a vehicle position vector according to the vehicle position coordinates, heading angle and linear velocity; calculate the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the vehicle position vector and the envelope box size parameter.

[0067] In a specific implementation, based on the Constant Turn Rate and Velocity (CTRV) model, assuming that the vehicle maintains a constant turn rate and velocity in a short time domain, the vehicle state transition equation is divided into two types according to different angular velocities ω(t k ): (Let Δt be the step size)

[0068] When ω(t k )≠0:

[0069]

[0070] v(t k +Δt) = v(t k )

[0071] ω(t k +Δt) = ω(t k )

[0072] ψ(t k +Δt) = ψ(t k ) + ω(t k )Δt

[0073] When ω(t k ) = 0:

[0074] X(t k +Δt) = X(t k ) + v(t k )cosψ(t k )Δt

[0075] Y(t k +Δt) = Y(t k ) + v(t k )sinψ(t k )Δt

[0076] v(t k+Δt) = v(t k )

[0077] ω(t k +Δt) = 0

[0078] ψ(t k +Δt) = ψ(t k )

[0079] At any time t within the prediction time domain [t0, t h , the three-dimensional spatio-temporal envelope volume distribution V(t k ) of the vehicle is expressed as: k ) is expressed as:

[0080]

[0081] where the position vector P(α, β, κ) is:

[0082]

[0083] where α, β, and κ are parameters corresponding to the longitudinal, lateral, and height direction components in the vehicle coordinate system, respectively.

[0084] Furthermore, the method for calculating the position probability distribution of the aerial object based on the landing state vector and the position covariance matrix is to perform position prediction on the aerial object through an unscented Kalman filter based on the landing state vector to obtain the position mean vector of the aerial object; determine the random position vector of the aerial object; construct a position probability density function in the form of a multi-dimensional normal distribution according to the position mean vector, the random position vector, and the position covariance matrix; and calculate the position probability distribution of the aerial object through the position probability density function.

[0085] In this embodiment, referring to Figure 3 , Figure 3 is the schematic diagram of the force analysis of the flying object in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention, and its corresponding state transition equation of the aerial object is:

[0086] x k+1 = f(x k ) + w k

[0087]

[0088] where v k = [v xg,k , v yg,k , v zg,k T is the velocity vector, g is the acceleration due to gravity, and w k is the process noise. ​

[0089] Use the Unscented Kalman Filter (UKF) for multi-step position prediction. The following steps are performed for each step of position prediction:

[0090] 1) Generate sigma points

[0091] Based on the current state estimate and calculate the sigma point set χ i :

[0092]

[0093] where n = 7 is the state vector dimension and λ is the scaling parameter.

[0094] 2) Propagate the sigma points

[0095] Use the state transition function to propagate each sigma point:

[0096]

[0097] 3) Predict the state mean

[0098] Predict the state mean at the next time step:

[0099]

[0100] where W i (m) is the mean weight.

[0101] 4) Calculate the covariance matrix

[0102] Predict the covariance matrix at the next time step:

[0103]

[0104] where W i (c) is the covariance weight and Q is the process noise covariance matrix.

[0105] 5) Update the state

[0106] Update the state estimate and the covariance matrix:

[0107]

[0108] After completing the UKF prediction, at any time t within the prediction time domain k , the position mean vector μ(t k ) of the airborne object is given by the state estimate vector (i.e., the landing state vector) The first three components represent:

[0109]

[0110] Among them, represents the first three components of

[0111] The position covariance matrix P r (t k ) is represented by the upper left 3×3 submatrix of Pk:

[0112] P r (t k ) = P k,1:3,1:3

[0113] Among them, P k,1:3,1:3 represents the upper left 3×3 submatrix of Pk.

[0114] At time t k below, the reference Figure 4 , Figure 4 is a schematic diagram of the density function of the position probability distribution of the airborne object in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention, and its position probability density function is constructed in the form of a multi-dimensional normal distribution:

[0115]

[0116] Among them, p(r,t k ) is the position probability distribution of the airborne object, and r = [x, y, z] T is a random position vector.

[0117] Step S30: Determine a multi-dimensional collision risk quantification index according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the airborne object.

[0118] Furthermore, the processing method for determining the multi-dimensional collision risk quantification index according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the airborne object is to determine the target ground-air collision probability according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the airborne object, and calculate the minimum relative distance between the airborne object and the vehicle within the prediction time domain and the minimum relative time corresponding to the minimum relative distance; determine the multi-dimensional collision risk quantification index according to the target ground-air collision probability, the minimum relative distance and the minimum relative time.

[0119] It should also be noted that the multi-dimensional collision risk quantification index includes the target ground-air collision probability, the minimum relative distance and the minimum relative time.

[0120] The method for determining the target ground-air collision probability based on the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the airborne object is to integrate the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the airborne object to obtain multiple ground-air collision probabilities within the prediction time domain; select the maximum ground-air collision probability from the multiple ground-air collision probabilities, and use the maximum ground-air collision probability as the target ground-air collision probability.

[0121] In this embodiment, referring to Figure 5 , Figure 5 is a schematic diagram of calculating the ground-air collision probability in the first embodiment of the ground-air collision warning method based on multi-dimensional risk assessment of the present invention. At time t within the prediction time domain [t0, t h , the position probability distribution of the airborne object is integrated with the three-dimensional spatio-temporal envelope volume distribution of the vehicle to calculate the collision probability: k

[0122]

[0123] Select the maximum value from the collision probabilities at each time as the target collision probability

[0124]

[0125] The closest distance is the minimum Euclidean distance (i.e., the minimum relative distance d min ) between the airborne object and the ground vehicle within the prediction time domain: )

[0126]

[0127] where μ(t k ) and x(t k ) respectively represent the predicted position vectors of the airborne object and the ground vehicle at time t k .

[0128] The time to closest point of approach (TCPA) represents the time required for the airborne object and the ground vehicle to reach the minimum relative distance (i.e., the minimum relative time TCPA):

[0129]

[0130] Step S40: Calculate the ground-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and perform a ground-air collision warning prompt for the vehicle according to the collision risk evaluation value.

[0131] Furthermore, the processing method of calculating the ground-air collision risk evaluation value through the multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index is to obtain the current linear velocity of the vehicle and the static safety distance; construct a multi-dimensional risk evaluation function according to the multi-dimensional collision risk quantification index, the current linear velocity and the static safety distance, and calculate the ground-air collision risk evaluation value through the multi-dimensional risk evaluation function.

[0132] In this embodiment, considering factors such as collision probability, time dimension, and space safety requirements, a multi-dimensional risk evaluation function is established to achieve probability-time-space joint risk assessment. The function form is as follows:

[0133]

[0134] Among them, λ1 to λ4 are weight parameters, is the collision probability value, TCPA is the time to the closest point of the vehicle and the trajectory of the aerial object, d0(v(t0)) represents the adaptive safety distance, t0 is the current moment, and d min is the shortest distance between the vehicle and the trajectory of the aerial object, is the sigmoid function.

[0135] Collision probability It is calculated by integrating the position probability density function of the aerial object and the spatio-temporal envelope of the vehicle, directly reflecting the collision possibility, and serving as the basic risk source in the evaluation function. The larger its value, the higher the overall risk level.

[0136] TCPA characterizes the urgency of the collision threat, and dynamically adjusts the weight of the probability term through the exponential decay term exp(-λ2TCPA). When TCPA is small (collision approaching), the original contribution of high collision probability is retained; when TCPA is large (collision far away), the weight of low-urgency risks is suppressed. The parameters λ1 and λ2 respectively control the amplitude and sensitivity of time correction.

[0137] To reflect the influence of the vehicle's driving speed on the safety distance, the safety distance is designed to be linearly related to the vehicle speed v:

[0138] d0(v(t0)) = d s + kv(t0)

[0139] Among them, d s represents the static basic safety distance, and k is used to adjust the influence of vehicle speed on the safety distance. The higher the speed, the larger the safety distance is correspondingly amplified. When d min < d0(v(t0)), the closest distance between the vehicle and the trajectory of the aerial object is less than the expected safety threshold, and it is determined that there is an additional collision risk, and the risk value needs to be quickly amplified; when d minIn the case of ≥d0(v(t0)), the additional risk is relatively controllable. The sigmoid function is used to smoothly map this distance difference to describe the collision threat when the safety distance is insufficient.

[0140] For hyperparameters such as λ1 to λ4, a grid search combined with a multi-objective optimization strategy is adopted. Based on historical data or simulation scenarios, with the risk assessment accuracy rate, false alarm rate, and calculation delay as evaluation indicators, a Pareto optimal parameter combination is selected to ensure the adaptability of the function to different ground-air interaction scenarios.

[0141] Furthermore, the processing method for giving a ground-air collision warning prompt to the vehicle according to the collision risk evaluation value is to map the collision risk evaluation value to discrete risk levels to determine the current collision risk level; determine the warning prompt strategy according to the current collision risk level, and give a ground-air collision warning prompt to the vehicle based on the warning prompt strategy.

[0142] In the specific implementation, based on the normalized collision risk evaluation value R, it is mapped to discrete risk levels, and a hierarchical response mechanism is established to implement a differential processing strategy for vehicles with different risk levels.

[0143] Map the normalized risk evaluation value R to discrete risk levels:

[0144]

[0145] where α1 and α2 are risk level division thresholds.

[0146] In the embodiment, the risk level division thresholds α1 and α2 directly affect the sensitivity and reliability of the warning system, and their values are determined by statistical analysis methods. A large number of risk assessment values R are collected in the simulation environment, covering low, medium, and high-risk events. According to the corresponding relationship between historical collision events and risk values, the threshold range is initially set. Continuously record the matching degree between the warning results and real events during actual operation. If a low-risk event is misjudged as a medium-risk event, then appropriately reduce α1, and vice versa, increase α1; make a similar adjustment to α2 until the false alarm rate and missed alarm rate of the system meet the requirements.

[0147] For different risk levels, a progressive response strategy (i.e., warning prompt strategy) is adopted:

[0148] 1) Low-risk state (Low):

[0149] Maintain the original motion plan and keep continuous monitoring and risk assessment of the aerial object.

[0150] 2) Medium-risk state (Mid):

[0151] Activate the preventive obstacle avoidance mechanism, and optimize the trajectory planning to increase the safety distance from the aerial object while maintaining the driving stability of the vehicle.

[0152] 3) High-risk state (High):

[0153] Activate the emergency avoidance mode. On the premise of ensuring basic safety, appropriately reduce the stability constraint boundary, allow the drive system to output power exceeding the rated power for a short time, and maximize the avoidance ability.

[0154] To avoid system jitter caused by frequent switching of response strategies, introduce a state transition hysteresis mechanism, which responds immediately when switching to a higher risk level and requires meeting a duration requirement when switching to a lower risk level.

[0155] In this embodiment, first, obtain the envelope box size parameters and driving state vectors of the vehicle within the prediction time domain, as well as the landing state vector and position covariance matrix of the aerial object. Then, calculate the three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameters and driving state vectors, and calculate the position probability distribution of the aerial object according to the landing state vector and position covariance matrix. After that, determine the multi-dimensional collision risk quantification index according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object. Finally, calculate the ground-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and give a ground-air collision warning prompt for the vehicle according to the collision risk evaluation value. In this embodiment, through the single integral of the three-dimensional spatio-temporal envelope volume distribution of the ground vehicle and the position probability distribution of the aerial object, the computational complexity is reduced from O(n 2 ) to O(n). A multi-dimensional risk evaluation function is constructed based on the multi-dimensional collision risk quantification index to solve the defect of insufficient dynamic adaptability of the existing method. Finally, the risk value is discretized to generate a progressive warning instruction to support the hierarchical response strategy of the vehicle, and provide more accurate and reliable safety warning information for the ground vehicle in the ground-air heterogeneous scenario.

[0156] Refer to Figure 6 , Figure 6 which is the structural block diagram of the first embodiment of the ground-air collision warning system based on multi-dimensional risk assessment of the present invention.

[0157] As Figure 6 shown, the ground-air collision warning system based on multi-dimensional risk assessment proposed in the embodiment of the present invention includes:

[0158] An acquisition module 6001, configured to acquire the envelope box size parameters and driving state vectors of the vehicle within the prediction time domain, as well as the landing state vector and position covariance matrix of the aerial object;

[0159] A calculation module 6002, configured to calculate a three-dimensional spatio-temporal envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector, and calculate a position probability distribution of the aerial object according to the landing state vector and the position covariance matrix;

[0160] The calculation module 6002 is further configured to determine a multi-dimensional collision risk quantification index according to the three-dimensional spatio-temporal envelope volume distribution of the vehicle and the position probability distribution of the aerial object;

[0161] An early warning module 6003, configured to calculate a ground-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and perform a ground-air collision early warning prompt on the vehicle according to the collision risk evaluation value.

[0162] Other embodiments or specific implementation manners of the ground-air collision early warning system based on multi-dimensional risk assessment of the present invention may refer to the above method embodiments, and will not be elaborated here.

[0163] It should be noted that in this article, the terms "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or system. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or system including the element.

[0164] The above serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0165] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation manner. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as a read-only memory / random access memory, magnetic disk, optical disk), and includes several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0166] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A ground-to-air collision warning method based on multi-dimensional risk assessment, characterized in that: The ground-to-air collision warning method based on multi-dimensional risk assessment comprises the following steps: Obtain the envelope box size parameters and driving state vector of the vehicle within the prediction time domain, and the landing state vector and position covariance matrix of the aerial object; Calculating the three-dimensional space-time envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector, and calculating the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix; Determining a multi-dimensional collision risk quantification index according to the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object; Based on the multi-dimensional collision risk quantification index, a ground-to-air collision risk evaluation value is calculated through a multi-dimensional risk evaluation function, and a ground-to-air collision warning prompt is given to the vehicle according to the collision risk evaluation value.

2. The method according to claim 1, characterized in that The step of calculating the three-dimensional space-time envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector comprises: determining an angular velocity of the vehicle according to the driving state vector; Determine the vehicle position coordinates, heading angle and linear velocity through a vehicle state transfer equation according to the angular velocity; Constructing a vehicle position vector according to the vehicle position coordinates, the heading angle and the linear speed; The three-dimensional space-time envelope volume distribution of the vehicle is calculated according to the vehicle position vector and the envelope box size parameter.

3. The method according to claim 1, characterized in that The step of calculating the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix comprises: Based on the landing state vector, the position of the aerial object is predicted by using an unscented Kalman filter to obtain a position mean vector of the aerial object; Determining a random position vector of the aerial object; Constructing a position probability density function in a multidimensional normal distribution form according to the position mean vector, the random position vector and the position covariance matrix; The position probability distribution of the aerial object is calculated using the position probability density function.

4. The method according to any one of claims 1 to 3, characterized in that: The step of determining a multi-dimensional collision risk quantification index according to the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object comprises: Determine the target ground-to-air collision probability according to the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object, and calculate the minimum relative distance between the aerial object and the vehicle within the predicted time domain and the minimum relative time corresponding to the minimum relative distance; A multi-dimensional collision risk quantification index is determined according to the target ground-to-air collision probability, the minimum relative distance and the minimum relative time.

5. The method according to claim 4, characterized in that The step of determining the target ground-to-air collision probability according to the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object comprises: Integrating the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object to obtain a plurality of ground-to-air collision probabilities within the predicted time domain; A maximum ground-to-air collision probability is selected from a plurality of ground-to-air collision probabilities, and the maximum ground-to-air collision probability is used as a target ground-to-air collision probability.

6. The method according to claim 5, characterized in that The step of calculating the ground-to-air collision risk evaluation value through a multidimensional risk evaluation function based on the multidimensional collision risk quantification index comprises: Obtaining the current linear speed and static safety distance of the vehicle; A multidimensional risk assessment function is constructed according to the multidimensional collision risk quantification index, the current linear speed and the static safety distance, and a ground-to-air collision risk assessment value is calculated through the multidimensional risk assessment function.

7. The method according to claim 6, characterized in that The step of providing the vehicle with a ground-to-air collision warning prompt according to the collision risk evaluation value comprises: Mapping the collision risk assessment value to discrete risk levels to determine a current collision risk level; A warning prompt strategy is determined according to the current collision risk level, and a ground-to-air collision warning prompt is provided to the vehicle based on the warning prompt strategy.

8. A ground-to-air collision warning system based on multi-dimensional risk assessment, characterized in that: The ground-to-air collision warning system based on multi-dimensional risk assessment includes: An acquisition module is used to obtain the envelope box size parameters and driving state vector of the vehicle within the prediction time domain, and the landing state vector and position covariance matrix of the aerial object; A calculation module, used to calculate the three-dimensional space-time envelope volume distribution of the vehicle according to the envelope box size parameter and the driving state vector, and calculate the position probability distribution of the aerial object according to the landing state vector and the position covariance matrix; The calculation module is further used to determine a multi-dimensional collision risk quantification index according to the three-dimensional space-time envelope volume distribution of the vehicle and the position probability distribution of the aerial object; The early warning module is used to calculate the ground-to-air collision risk evaluation value through a multi-dimensional risk evaluation function based on the multi-dimensional collision risk quantification index, and to provide the vehicle with a ground-to-air collision early warning prompt according to the collision risk evaluation value.

9. A storage medium, characterized in that: The storage medium stores a ground-to-air collision warning program based on multidimensional risk assessment. When the ground-to-air collision warning program based on multidimensional risk assessment is executed by the processor, the steps of the ground-to-air collision warning method based on multidimensional risk assessment as described in any one of claims 1 to 7 are implemented.

Images