Driving risk level determination method, device, computer equipment and storage medium

By obtaining the time series of vehicle driving status, using variable point identification and classification models, accurately judging driving risk levels, the problem of difficulty in obtaining user driving behavior characteristics in the prior art is solved, and the rationality of auto insurance pricing is improved.

CN114241429BActive Publication Date: 2024-11-19PING AN TECH (SHENZHEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111535265.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-15
Publication Date
2024-11-19
Estimated Expiration
2041-12-15

AI Technical Summary

Technical Problem

The prior art is difficult to accurately obtain user driving behavior characteristics and determine driving risk levels, resulting in unreasonable pricing of auto insurance.

Method used

By obtaining the time series of the vehicle's driving state, using the variable point recognition model to determine the mutation time, determining the rules and classification models based on the preset sub-sequence, and determining the risk level corresponding to the time series, including behaviors such as acute acceleration, acute deceleration, and sharp turn.

Benefits of technology

It realizes the accurate processing and classification of vehicle driving status information, can effectively determine the driving risk level, and improves the rationality of auto insurance pricing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114241429B_ABST
    Figure CN114241429B_ABST
Patent Text Reader

Abstract

The present application relates to the field of user profiling, and provides a method, device, equipment and computer storage medium for determining a driving risk level, the method comprising: obtaining a time series reflecting the driving status information of a vehicle; determining the mutation moment in the time series based on a change point recognition model; determining the subsequence corresponding to the mutation moment in the time series based on a preset subsequence determination rule; determining whether the subsequence corresponds to a preset target state based on a preset classification model, the target state including at least one of rapid acceleration, rapid deceleration and sharp turn; determining the risk level corresponding to the time series based on whether the subsequence corresponds to the preset target state based on a preset risk level determination rule. The present application also relates to artificial intelligence, and the driving risk level determination method of the present application can be applied to cloud servers of cloud computing services of big data and artificial intelligence platforms.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of user profiling, and in particular to a method, device, computer equipment and storage medium for determining a driving risk level. Background Art

[0002] Usage Based Insurance (UBI) is a type of auto insurance that is priced based on the user's driving behavior. Unlike uniformly priced auto insurance, UBI determines the user's risk level based on the user's behavioral characteristics during driving, such as the number of sudden accelerations, sudden decelerations, and sharp turns per 100 kilometers, and then prices the insurance based on different risk levels. Compared with traditional uniformly priced auto insurance, UBI auto insurance can provide lower-priced auto insurance for users with good driving habits, making the pricing of auto insurance more reasonable. However, how to obtain the user's driving behavior characteristics and how to determine the user's driving behavior risk level based on these driving behavior characteristics urgently requires an accurate and easy method. Summary of the invention

[0003] The main purpose of this application is to provide a method, device, equipment and computer storage medium for determining a driving risk level, aiming to determine the corresponding driving risk level according to the driving status of a vehicle.

[0004] In a first aspect, the present application provides a method for determining a driving risk level, the method comprising the following steps:

[0005] Acquire a time series reflecting the driving status information of the vehicle;

[0006] Determine a mutation moment in the time series based on a change point recognition model, wherein the driving state information of the vehicle before and after the mutation moment is different;

[0007] Based on a preset subsequence determination rule, determining a subsequence corresponding to the mutation moment in the time series;

[0008] Based on a preset classification model, determining whether the subsequence corresponds to a preset target state, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn;

[0009] Based on a preset risk level determination rule, the risk level corresponding to the time series is determined according to whether the subsequence corresponds to a preset target state.

[0010] In a second aspect, the present application further provides a driving risk level determination device, the driving risk level determination device comprising:

[0011] A time series acquisition module, used to acquire a time series reflecting the driving state information of the vehicle;

[0012] a change point recognition module, used to determine the mutation time in the time series based on the change point recognition model, wherein the driving state information before and after the mutation time is different;

[0013] A subsequence determination module, configured to determine the subsequence corresponding to the mutation time in the time sequence based on a preset subsequence determination rule;

[0014] a target state determination module, configured to determine whether the subsequence corresponds to a preset target state based on a preset classification model, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn;

[0015] The risk level determination module is used to determine the risk level corresponding to the time series based on a preset risk level determination rule and according to whether the subsequence corresponds to a preset target state.

[0016] In a third aspect, the present application also provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the method for determining the driving risk level as described above is implemented.

[0017] In a fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the method for determining the driving risk level as described above is implemented.

[0018] The present application provides a driving risk level determination method, device, equipment and computer storage medium. The present application obtains a time series reflecting the driving state information of a vehicle; determines the mutation moment in the time series based on a change point recognition model, and the driving state information of the vehicle before and after the mutation moment is different; determines the subsequence corresponding to the mutation moment in the time series based on a preset subsequence determination rule; determines whether the subsequence corresponds to a preset target state based on a preset classification model, and the target state includes at least one of rapid acceleration, rapid deceleration and sharp turn; determines the risk level corresponding to the time series based on whether the subsequence corresponds to the preset target state based on a preset risk level determination rule. The time series reflecting the driving state information of the vehicle can be processed and classified, and the risk level corresponding to the time series can be determined. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flowchart of a method for determining a driving risk level provided in one embodiment of the present application;

[0021] Figure 2 A usage scenario diagram of a method for determining a driving risk level provided in an embodiment of the present application;

[0022] Figure 3 A usage scenario diagram of a method for determining a driving risk level provided in another embodiment of the present application;

[0023] Figure 4 A schematic block diagram of a driving risk level determination device provided in an embodiment of the present application;

[0024] Figure 5 The present invention is a block diagram showing the structure of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0026] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0027] Embodiments of the present application provide a method, apparatus, computer device, and computer-readable storage medium for determining a driving risk level.

[0028] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0029] Please refer to Figure 1 , Figure 1A flow chart of a method for determining a driving risk level provided for an embodiment of the present application. The method for determining a driving risk level can be used in a terminal or a server, for example, by obtaining a time series reflecting the driving status information of a vehicle through a terminal, and by processing the time series reflecting the driving status information of the vehicle through a server. Among them, the terminal can be an electronic device such as a mobile phone, a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device; the server can be an independent server, or a server cluster, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0030] Please refer to Figure 2 , Figure 2 This is a usage scenario diagram provided by an embodiment of the present application. Figure 2 As shown, the time series reflecting the driving status information of the vehicle can be obtained through a sensor carried by the mobile terminal.

[0031] Please refer to Figure 3 , Figure 3 This is a usage scenario diagram provided by another embodiment of the present application. Figure 3 As shown, the time series reflecting the driving state information of the vehicle can also be obtained by a vehicle-mounted sensor provided on the vehicle. Of course, the time series reflecting the driving state information of the vehicle can also be obtained by other methods, which are not limited here.

[0032] like Figure 1 As shown, the driving risk level determination method includes steps S101 to S105.

[0033] Step S101: Acquire a time series reflecting the driving state information of the vehicle.

[0034] Exemplarily, the time series reflecting the driving status information of the vehicle can be obtained through sensors carried by mobile terminals, such as sensors carried by mobile phones and tablet computers, or can be obtained through vehicle-mounted sensors installed on the vehicle, which is not limited here.

[0035] In some embodiments, obtaining a time series reflecting driving status information of a vehicle includes: obtaining the driving status information of the vehicle at preset time intervals, wherein the driving status information includes at least one of the following: angular velocity in at least one direction and / or acceleration in at least one direction.

[0036] Exemplarily, the preset time interval can be set according to actual needs. In particular, the preset time interval can also be set to 0, that is, the driving state information of the vehicle is obtained continuously.

[0037] Exemplarily, the angular velocity in at least one direction is obtained by a gyro sensor and / or the acceleration in at least one direction is obtained by an acceleration sensor. For example, the angular velocity in at least one direction of the X-axis, the Y-axis, and the Z-axis is obtained by a gyro sensor and / or the acceleration in at least one direction of the X-axis, the Y-axis, and the Z-axis is obtained by an acceleration sensor.

[0038] Step S102: determining a mutation moment in the time series based on a change point recognition model, wherein the driving state information of the vehicle before and after the mutation moment is different.

[0039] Exemplarily, the change point recognition model is used to determine the mutation moment in the time series so as to classify the subsequence corresponding to the mutation moment.

[0040] Exemplarily, the change point recognition model is implemented based on a mixed Gaussian model. Specifically, the change point recognition model estimates the probability density ratio at any moment, that is, the ratio of the prior probability density to the posterior probability density at any moment, through the mixed Gaussian model, and compares the probability density ratio with a preset threshold to determine the mutation moment.

[0041] In some embodiments, the determination of the mutation moment in the time series based on the change point recognition model includes: obtaining the probability density ratio of each moment in the time series, the probability density ratio being the ratio of the prior probability density to the posterior probability density of the moment; if the probability density ratio is greater than a preset threshold, determining the moment to be a mutation moment.

[0042] Exemplarily, since the mutation moment has the property that the prior probability density is greater than the posterior probability density, the probability density ratio of each moment is calculated and then compared with a preset threshold, thereby avoiding the direct calculation of the prior probability density and the posterior probability density and reducing the complexity of the calculation.

[0043] In some embodiments, obtaining the probability density ratio at any moment in the time series includes: based on a Gaussian mixture model, fitting the ratio of the prior probability density to the posterior probability density at the moment by an optimization method to obtain the probability density ratio at the moment.

[0044] For example, through the Gaussian model Estimate the probability density ratio, where I takes values ​​in [1, 2, 3..., N], is the Gaussian kernel function, σ is the Gaussian width, θ=[θ1,...,θn ] T are the preset model parameters, and the superscript T represents matrix transpose.

[0045] For example, the second-order error between the true density ratio function and the estimated density ratio function is minimized. where q α (x) = αp(x) + (1-α)p′(x), p(x) represents the prior probability density, and p′(x) represents the posterior probability density.

[0046] For example, q α Substitute (x) into J0(θ) to obtain the objective function Calculation by optimization method The minimum value of the probability density ratio r α The value of (x) is compared with the preset threshold to determine r α The moment when the value of (x) is greater than or equal to the preset threshold is the mutation moment.

[0047] Step S103: determining the subsequence corresponding to the mutation moment in the time series based on a preset subsequence determination rule.

[0048] Exemplarily, the subsequence corresponding to the mutation moment may be determined according to a preset duration, for example, a time sequence with a duration of 0.5 seconds at the mutation moment is determined as the subsequence corresponding to the mutation moment.

[0049] Exemplarily, the mutation time is located at the time midpoint of the subsequence.

[0050] Step S104: determining whether the subsequence corresponds to a preset target state based on a preset classification model, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn.

[0051] Exemplarily, the classification model includes an echo state network and an autoencoder.

[0052] In some embodiments, determining the target state corresponding to the subsequence based on a preset classification model includes: mapping the subsequence to a vector dimension based on an echo state network to obtain a description vector corresponding to the subsequence; based on an autoencoder, reducing the dimension of the description vector to obtain a low-dimensional vector corresponding to the description vector; and determining the target state corresponding to the low-dimensional vector based on a preset classification model.

[0053] For example, an echo state network (ESN) is a recursive neural network consisting of an input layer, a hidden layer, and an output layer. The hidden layer is a network consisting of sparsely connected neurons. Since the input weights of the input layer of the echo state network and the hidden weights of the hidden layer are randomly generated, training the echo state network only requires training the output weights from the hidden layer to the output layer, so only a simple linear regression is needed to complete the training of the echo state network.

[0054] Exemplarily, the echo state network can model the subsequence by the following set of equations:

[0055]

[0056] Where x(t)∈RD×1 is the hidden layer state vector of the echo state network at time t; h is the activation function of the neural network, which is generally a hyperbolic tangent function; v∈R D×1 is the input weight matrix, U∈R D×D is the hidden weight matrix, w∈R D×1 is the output weight matrix; y = (y(1), ..., y(T)), T represents the length of sequence y, that is, y∈R T×1 ; Represents the echo state network's estimate of y(t+1).

[0057] The solved hidden state vector x(t) is represented as a matrix X∈R T×D , represent y(t) as vector y, and obtain the objective function l(w)=||Xw-y|| 2 , plus the regularization term to get l(w)+μ 2 ||w|| 2 , the solution can be obtained to get the output layer weight w = (X T X+μ 2 I D ) -1 X T y. It can be seen that the output weight of the echo state network does not need to be solved by methods such as gradient descent, but only requires simple linear regression, which reduces the complexity of the calculation.

[0058] Exemplarily, a corresponding output weight is solved for each of the subsequences, and the output weight is represented as a description vector, and the description vector can describe the characteristics of the subsequence.

[0059] Exemplarily, the auto-encoder (AE) includes a recognition network (Encoder) and a generation network (Decoder). The recognition network is used to convert the input description vector into an internal representation, and the generation network can reconstruct the output according to the internal representation. Since the dimension of the internal representation is smaller than the dimension of the input, the recognition network of the auto-encoder can reduce the dimension of the description vector, that is, obtain the internal representation obtained by inputting the description vector into the auto-encoder as the low-dimensional vector.

[0060] In some embodiments, determining the target state corresponding to the low-dimensional vector based on a preset classification model includes: clustering the low-dimensional vector based on a clustering model, and determining the target state corresponding to the low-dimensional vector according to the clustering result.

[0061] Exemplarily, the low-dimensional vectors are clustered by a k-means clustering algorithm. Specifically, the low-dimensional vectors are input into the classification model to determine the classification parameter k, wherein the classification parameter k is used to indicate the number of the target states; k low-dimensional vectors are randomly determined as initial cluster centers; the distances from other low-dimensional vectors to the cluster centers are calculated, and the other low-dimensional vectors are divided into classes where the nearest cluster centers are located, thereby obtaining k classes, wherein the other low-dimensional vectors are low-dimensional vectors other than the cluster centers; k cluster centers are re-determined for the k classes; the calculation of the distances from other low-dimensional vectors to the cluster centers is re-executed, and the other low-dimensional vectors are divided into classes where the nearest cluster centers are located, thereby obtaining k classes and the re-determining of k cluster centers for the k classes, until the re-determined k cluster centers and no longer change.

[0062] Exemplarily, the target parameter k can be set to 3, which respectively represents three target states of rapid acceleration, rapid deceleration, and rapid turning, but is certainly not limited thereto.

[0063] Step S105: Based on a preset risk level determination rule, determine the risk level corresponding to the time series according to whether the subsequence corresponds to a preset target state.

[0064] In some implementations, the risk level corresponding to the driving state information of the vehicle is determined based on the number of subsequences in the time series corresponding to the preset target state, wherein the risk level is positively correlated with the number of target states.

[0065] Exemplarily, the risk level corresponding to the driving state information of the vehicle is determined according to the number of subsequences corresponding to the preset target state in the time series and the type of the preset target state.

[0066] For example, the risk level can be determined by the risk score obtained by weighting the corresponding target state type according to the number of target states corresponding to the subsequences in the time series. Specifically, a corresponding weight value can be set for each target state type. For example, the weight value of sudden acceleration is set to 1.5, the weight value of sudden deceleration is set to 1, and the weight value of sharp turn is set to 2. The sum of the product of the number of each target state and the corresponding weight value is calculated to obtain the risk score, and then the corresponding risk level is determined according to the preset risk score value. For example, if the risk score is greater than or equal to the first preset value, the corresponding first risk level is determined; if the risk score is less than the first preset value and greater than or equal to the second preset value, the corresponding second risk level is determined, and so on, wherein the first preset value is greater than the second preset value, and the driving risk of the first risk level is greater than the driving risk of the second risk level.

[0067] The driving risk level determination method provided in the above embodiment is to obtain a time series reflecting the driving state information of the vehicle; determine the mutation moment in the time series based on the change point recognition model, and the driving state information of the vehicle before and after the mutation moment is different; determine the subsequence corresponding to the mutation moment in the time series based on a preset subsequence determination rule; determine whether the subsequence corresponds to a preset target state based on a preset classification model, and the target state includes at least one of rapid acceleration, rapid deceleration, and sharp turn; determine the risk level corresponding to the time series based on whether the subsequence corresponds to the preset target state based on a preset risk level determination rule. The time series reflecting the driving state information of the vehicle can be processed and classified, and the risk level corresponding to the time series can be determined.

[0068] See also Figure 4 , Figure 4 It is a schematic diagram of a driving risk level determination device provided in an embodiment of the present application. The driving risk level determination device can be configured in a server or a terminal to execute the aforementioned driving risk level determination method.

[0069] like Figure 4 As shown, the driving risk level determination device includes: a time series acquisition module 110, a change point recognition module 120, a subsequence determination module 130, a target state determination module 140, and a risk level determination module 150.

[0070] The time series acquisition module 110 is used to acquire the time series reflecting the driving state information of the vehicle;

[0071] A change point identification module 120, for determining a mutation time in the time series based on a change point identification model, wherein the driving state information before and after the mutation time is different;

[0072] A subsequence determination module 130, configured to determine the subsequence corresponding to the mutation time in the time sequence based on a preset subsequence determination rule;

[0073] A target state determination module 140, configured to determine whether the subsequence corresponds to a preset target state based on a preset classification model, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn;

[0074] The risk level determination module 150 is used to determine the risk level corresponding to the time series based on a preset risk level determination rule and according to whether the subsequence corresponds to a preset target state.

[0075] Exemplarily, the target state determination module 140 includes a sequence mapping submodule, a vector dimension reduction submodule, and a vector classification submodule.

[0076] The sequence mapping submodule is used to map the subsequence to a vector dimension based on an echo state network to obtain a description vector corresponding to the subsequence.

[0077] The vector dimension reduction submodule is used to reduce the dimension of the description vector based on the autoencoder to obtain a low-dimensional vector corresponding to the description vector.

[0078] The vector classification submodule is used to determine the target state corresponding to the low-dimensional vector based on a preset classification model.

[0079] Exemplarily, the vector classification submodule includes a clustering model submodule.

[0080] The clustering model submodule is used to cluster the low-dimensional vectors based on the clustering model, and determine the target state corresponding to the low-dimensional vector according to the clustering result.

[0081] Exemplarily, the change point identification module 120 includes a probability density ratio determination submodule and a mutation time determination submodule.

[0082] The probability density ratio determination submodule is used to obtain the probability density ratio of each moment in the time series, where the probability density ratio is the ratio of the prior probability density to the posterior probability density at the moment.

[0083] The mutation moment determination submodule is used to determine that the moment is a mutation moment if the probability density ratio is greater than a preset threshold.

[0084] Exemplarily, the probability density ratio determination submodule includes a probability density ratio fitting submodule.

[0085] The probability density ratio fitting submodule is used to fit the ratio of the prior probability density and the posterior probability density at the moment through an optimization method based on a Gaussian mixture model to obtain the probability density ratio at the moment.

[0086] Exemplarily, the time series acquisition module 110 includes a time series acquisition submodule.

[0087] The time series acquisition submodule is used to acquire the driving state information of the vehicle at a preset time interval, wherein the driving state information includes at least one of the following: an angular velocity in at least one direction and / or an acceleration in at least one direction.

[0088] Exemplarily, the risk level determination module 150 includes a risk level determination submodule.

[0089] The risk level determination submodule is used to determine the risk level corresponding to the driving state information of the vehicle according to the target state type and the target state quantity corresponding to the subsequence in the time series, wherein the risk level is positively correlated with the target state quantity.

[0090] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and modules and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0091] The method and apparatus of the present application can be used in many general or special computing system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0092] Exemplarily, the above method and apparatus may be implemented in the form of a computer program. The computer program may be implemented in Figure 5 Runs on the computer device shown.

[0093] See also Figure 5 , Figure 5 A schematic block diagram of the structure of a computer device provided in an embodiment of the present application. The computer device may be a server or a terminal.

[0094] like Figure 5 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a storage medium and an internal memory.

[0095] The storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the driving risk level determination methods.

[0096] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0097] The internal memory provides an environment for the operation of the computer program in the storage medium. When the computer program is executed by the processor, the processor can execute any one of the driving risk level determination methods.

[0098] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0099] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0100] In one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:

[0101] Acquire a time series reflecting the driving status information of the vehicle;

[0102] Determine a mutation moment in the time series based on a change point recognition model, wherein the driving state information of the vehicle before and after the mutation moment is different;

[0103] Based on a preset subsequence determination rule, determining a subsequence corresponding to the mutation moment in the time series;

[0104] Based on a preset classification model, determining whether the subsequence corresponds to a preset target state, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn;

[0105] Based on a preset risk level determination rule, the risk level corresponding to the time series is determined according to whether the subsequence corresponds to a preset target state.

[0106] In one embodiment, when the processor determines the target state corresponding to the subsequence based on a preset classification model, it is used to implement:

[0107] Based on an echo state network, mapping the subsequence to a vector dimension to obtain a description vector corresponding to the subsequence;

[0108] Based on the autoencoder, the description vector is reduced in dimension to obtain a low-dimensional vector corresponding to the description vector;

[0109] Based on a preset classification model, a target state corresponding to the low-dimensional vector is determined.

[0110] In one embodiment, when the processor determines the target state corresponding to the low-dimensional vector based on a preset classification model, it is used to implement:

[0111] The low-dimensional vector is clustered based on a clustering model, and a target state corresponding to the low-dimensional vector is determined according to the clustering result.

[0112] In one embodiment, when the processor determines the mutation moment in the time series based on the change point recognition model, it is used to implement:

[0113] Obtaining a probability density ratio at each moment in the time series, wherein the probability density ratio is a ratio of a priori probability density to a posterior probability density at the moment;

[0114] If the probability density ratio is greater than a preset threshold, the moment is determined to be a mutation moment.

[0115] In one embodiment, when obtaining the probability density ratio at any moment in the time series, the processor is used to implement:

[0116] Based on the Gaussian mixture model, the ratio of the prior probability density to the posterior probability density at the moment is fitted by an optimization method to obtain the probability density ratio at the moment.

[0117] In one embodiment, when acquiring the time series reflecting the driving state information of the vehicle, the processor is used to implement:

[0118] At preset time intervals, driving state information of the vehicle is acquired, where the driving state information includes at least one of the following: an angular velocity in at least one direction and / or an acceleration in at least one direction.

[0119] In one embodiment, when the processor determines the risk level corresponding to the driving state information of the vehicle according to the target state corresponding to the subsequence based on the preset risk level determination rule, it is used to implement:

[0120] The risk level corresponding to the driving state information of the vehicle is determined according to the target state type and the target state quantity corresponding to the subsequence in the time series, wherein the risk level is positively correlated with the target state quantity.

[0121] It should be noted that technicians in the relevant field can clearly understand that for the convenience and brevity of description, the specific working process of determining the driving risk level described above can refer to the corresponding process in the aforementioned driving risk level determination control method embodiment, and will not be repeated here.

[0122] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions. The method implemented when the program instructions are executed can refer to the various embodiments of the driving risk level determination method of the present application.

[0123] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the computer device.

[0124] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0125] It should also be understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0126] The serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific implementation mode of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for determining a driving risk level, characterized in that: The method comprises: Acquire a time series reflecting the driving status information of the vehicle; Determine a mutation moment in the time series based on a change point recognition model, wherein the driving state information of the vehicle before and after the mutation moment is different; Based on a preset subsequence determination rule, determining a subsequence corresponding to the mutation moment in the time series; Based on a preset classification model, determining whether the subsequence corresponds to a preset target state, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn; Based on a preset risk level determination rule, determining the risk level corresponding to the time series according to whether the subsequence corresponds to a preset target state; Wherein, determining the mutation moment in the time series based on the change point recognition model includes: Obtaining a probability density ratio at each moment in the time series, wherein the probability density ratio is a ratio of a priori probability density to a posterior probability density at the moment; If the probability density ratio is greater than a preset threshold, determining the moment is a mutation moment; and determining the target state corresponding to the subsequence based on a preset classification model, including: Based on an echo state network, mapping the subsequence to a vector dimension to obtain a description vector corresponding to the subsequence; Based on the autoencoder, the description vector is reduced in dimension to obtain a low-dimensional vector corresponding to the description vector; Determining a target state corresponding to the low-dimensional vector based on a preset classification model; The obtaining of the probability density ratio of each moment in the time series includes: Based on the Gaussian mixture model, the ratio of the prior probability density to the posterior probability density at the moment is fitted by an optimization method to obtain the probability density ratio at the moment.

2. The method for determining the driving risk level according to claim 1, characterized in that: The determining, based on a preset classification model, a target state corresponding to the low-dimensional vector includes: The low-dimensional vector is clustered based on a clustering model, and a target state corresponding to the low-dimensional vector is determined according to the clustering result.

3. The method for determining the driving risk level according to any one of claims 1 to 2, characterized in that: The step of obtaining a time series reflecting the driving state information of the vehicle includes: At preset time intervals, driving state information of the vehicle is acquired, where the driving state information includes at least one of the following: an angular velocity in at least one direction and / or an acceleration in at least one direction.

4. The method for determining the driving risk level according to any one of claims 1 to 2, characterized in that: The step of determining the risk level corresponding to the driving state information of the vehicle based on a preset risk level determination rule and according to the target state corresponding to the subsequence includes: The risk level corresponding to the driving state information of the vehicle is determined according to the number of subsequences corresponding to the preset target state in the time series, wherein the risk level is positively correlated with the number of the target states.

5. A driving risk level determination device, characterized in that: The driving risk level determination device comprises: A time series acquisition module, used to acquire a time series reflecting the driving state information of the vehicle; a change point recognition module, used to determine the mutation time in the time series based on the change point recognition model, wherein the driving state information before and after the mutation time is different; A subsequence determination module, configured to determine the subsequence corresponding to the mutation time in the time sequence based on a preset subsequence determination rule; a target state determination module, configured to determine whether the subsequence corresponds to a preset target state based on a preset classification model, wherein the target state includes at least one of a sudden acceleration, a sudden deceleration, and a sharp turn; A risk level determination module, configured to determine the risk level corresponding to the time series based on a preset risk level determination rule and according to whether the subsequence corresponds to a preset target state; Wherein, the change point identification module is also used for: Obtaining a probability density ratio at each moment in the time series, wherein the probability density ratio is a ratio of a priori probability density to a posterior probability density at the moment; If the probability density ratio is greater than a preset threshold, the moment is determined to be a mutation moment; The target state determination module is also used for: Based on an echo state network, mapping the subsequence to a vector dimension to obtain a description vector corresponding to the subsequence; Based on the autoencoder, the description vector is reduced in dimension to obtain a low-dimensional vector corresponding to the description vector; Determining a target state corresponding to the low-dimensional vector based on a preset classification model; The change point identification module is also used for: Based on the Gaussian mixture model, the ratio of the prior probability density to the posterior probability density at the moment is fitted by an optimization method to obtain the probability density ratio at the moment.

6. A computer device, characterized in that: The computer device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the steps of the driving risk level determination method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the steps of the method for determining a driving risk level according to any one of claims 1 to 4 are implemented.