Driving risk estimation method and device

Abstract

The embodiment of the present application provides a driving risk estimation method and device. The driving risk estimation method includes: obtaining various actions that may occur when the driving risk estimation device is working, and obtaining the driving risk state corresponding to each driving risk type, if the action change of the driving risk estimation device causes the driving risk state change, then determining the corresponding state transition probability, so as to estimate the risk coefficient in the driving process according to the state transition probability, and then determining the total benefit corresponding to the action performed by the driving risk estimation device within the preset time period according to all the risk coefficients estimated within the preset time period; finally, determining the action of the driving risk estimation device after this estimation according to the risk coefficient and the total benefit, so as to perform the risk estimation work according to the action, while ensuring the accuracy of the risk estimation, improving the benefit of the estimated action, and reducing the resource consumption of the driving risk estimation device.

Description

Technical Field

[0001] The present application relates to the field of computer and communication technology, and in particular to a driving risk estimation method and device. Background Art

[0002] In the process of driving risk estimation, various resources are consumed, such as power resources, data storage resources, etc., and high accuracy requirements are also required. If the risk estimation device maintains the estimation state for a long time, the estimation accuracy can be improved, but it will also lead to higher resource consumption. For example, the accuracy of three consecutive estimates is higher than that of one estimate. The higher the number, the more accurate it is, and the greater the benefit obtained. However, multiple estimates will also bring higher costs. Therefore, in the process of risk estimation, there is a problem of low resource consumption and inaccurate estimation results. Summary of the invention

[0003] The embodiments of the present application provide a driving risk estimation method and device, thereby improving the benefits of the estimation action and reducing the resource consumption of the driving risk estimation device while at least ensuring the accuracy of the risk estimation.

[0004] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0005] According to one aspect of an embodiment of the present application, a driving risk estimation method is provided, comprising: obtaining various actions that may occur when a driving risk estimation device is working, and obtaining a driving risk state corresponding to each driving risk type; if a change in the action of the driving risk estimation device causes a change in the driving risk state, determining a corresponding state transition probability, and estimating a risk coefficient during driving based on the state transition probability; determining a total benefit corresponding to the actions performed by the driving risk estimation device within a preset time period based on all risk coefficients estimated within the preset time period; and determining an action of the driving risk estimation device after this estimation based on the risk coefficient and the total benefit.

[0006] According to one aspect of an embodiment of the present application, a driving risk estimation device is provided, comprising: a first acquisition unit, used to acquire various actions that may occur in the driving risk estimation device when it is working, and to acquire the driving risk state corresponding to each driving risk type; a first estimation unit, used to determine the corresponding state transition probability if the action change of the driving risk estimation device causes the driving risk state change, and estimate the risk coefficient of the driving process according to the state transition probability; a benefit unit, used to determine the total benefit corresponding to the actions performed by the driving risk estimation device within a preset time period according to all risk coefficients estimated within the preset time period; and an action unit, used to determine the action of the driving risk estimation device after this estimation according to the risk coefficient and the total benefit.

[0007] In some embodiments of the present application, based on the aforementioned scheme, the state corresponding to the driving risk type includes a hidden state and an explicit state; the first acquisition unit includes: a second acquisition unit, used to obtain the driving risk type and the estimation process corresponding to each driving risk type; a state decomposition unit, used to decompose the state of the historical data corresponding to the driving risk type according to the estimation process, to obtain the hidden state and the explicit state corresponding to the driving risk type, the data corresponding to the hidden state includes the sampled risk data, and the data corresponding to the explicit state includes the risk coefficient.

[0008] In some embodiments of the present application, based on the aforementioned scheme, the first estimation unit includes: a first determination unit, used to determine the hidden state value corresponding to the adjacent moment according to the state transition probability and the risk data corresponding to the hidden state collected at the current moment; a second determination unit, used to determine the explicit state value corresponding to the hidden state value according to the hidden state value corresponding to the adjacent moment, as the risk coefficient.

[0009] In some embodiments of the present application, based on the aforementioned scheme, the first estimation unit includes: a first correction unit, used to correct the data processing operation error in the risk estimation process according to the state transition probability; an updating unit, used to update the state transition probability according to the corrected data processing operation error; and a second estimation unit, used to estimate the risk coefficient in the driving process according to the updated state transition probability.

[0010] In some embodiments of the present application, based on the aforementioned scheme, the first correction unit includes: a third determination unit, used to determine the first sample variance corresponding to the state transition probability at the current moment and the second sample variance corresponding to the state transition probability at the adjacent moment; a fourth determination unit, used to determine the correction method of the data processing operation error based on the difference between the first sample variance and the second sample variance and a preset variance difference threshold; and a second correction unit, used to correct the data processing operation error according to the correction method.

[0011] In some embodiments of the present application, based on the aforementioned scheme, the state transition probability includes a hidden state transition probability and an explicit state transition probability; the first correction unit includes: a third correction unit, used to correct the data sampling error in the risk estimation process according to the hidden state transition probability; a fourth correction unit, used to correct the data transmission error in the risk estimation process according to the explicit state transition probability.

[0012] In some embodiments of the present application, based on the aforementioned scheme, the profit unit includes: a second acquisition unit, used to acquire all risk coefficients within a preset time period; a fifth determination unit, used to determine the instantaneous profit corresponding to the risk coefficient from historical profit data; a first calculation unit, used to calculate the expectation of all instantaneous profits within the preset time period, and obtain the total profit based on the expectation of all instantaneous profits within the preset time period.

[0013] In some embodiments of the present application, based on the aforementioned scheme, the first calculation unit includes: a factor unit, used to determine the discount factor corresponding to each risk coefficient according to the estimated moment corresponding to each risk coefficient; a second calculation unit, used to calculate the expectation of all the instantaneous returns based on the discount factor corresponding to each risk coefficient and the instantaneous return.

[0014] In some embodiments of the present application, based on the aforementioned scheme, the driving risk estimation device further includes: an action sequence unit, used to determine an action sequence after this estimation based on the risk coefficient and the total benefit, and the action sequence includes the estimated actions to be executed and the execution timing of each of the estimated actions to be executed.

[0015] According to one aspect of an embodiment of the present application, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the driving risk estimation method as described in the above embodiment is implemented.

[0016] According to one aspect of an embodiment of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the driving risk estimation method as described in the above embodiments.

[0017] In the technical solutions provided in some embodiments of the present application, by obtaining various actions that may occur when a driving risk estimation device is working, and obtaining the driving risk state corresponding to each driving risk type, if the action change of the driving risk estimation device causes the driving risk state to change, the corresponding state transition probability is determined, so as to estimate the risk coefficient of the driving process according to the state transition probability, and then determine the total benefit corresponding to the action performed by the driving risk estimation device within the preset time period according to all the risk coefficients estimated within the preset time period; finally, determine the action of the driving risk estimation device after this estimation according to the risk coefficient and the total benefit, so as to perform the risk estimation work according to the action, while ensuring the accuracy of the risk estimation, improving the benefit of the estimated action and reducing the resource consumption of the driving risk estimation device.

[0018] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0020] Figure 1 A schematic diagram showing an exemplary system architecture to which the technical solution of the embodiments of the present application can be applied;

[0021] Figure 2 A schematic diagram showing an exemplary system architecture to which the technical solution of the embodiments of the present application can be applied;

[0022] Figure 3 A flowchart of a driving risk estimation method according to an embodiment of the present application is schematically shown;

[0023] Figure 4 A schematic diagram of an application scenario of a driving risk estimation method according to an embodiment of the present application is schematically shown;

[0024] Figure 5 A schematic diagram schematically shows a process of estimating driving risk according to an embodiment of the present application;

[0025] Figure 6 A flowchart for estimating a risk factor in a driving process according to a state transition probability according to an embodiment of the present application is schematically shown;

[0026] Figure 7 A flowchart schematically illustrates a method for determining a total benefit corresponding to an action performed by a driving risk estimation device within a preset period of time according to an embodiment of the present application;

[0027] Figure 8 A flowchart for calculating the expectation of all instantaneous benefits within the preset time period according to an embodiment of the present application is schematically shown;

[0028] Fig. 9 A schematic diagram of a driving risk estimation device according to an embodiment of the present application is schematically shown;

[0029] Fig.10 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown. DETAILED DESCRIPTION

[0030] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concept of the example embodiments to those skilled in the art.

[0031] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0032] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0033] The flowcharts shown in the accompanying drawings are only exemplary 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 can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0034] Figure 1 A schematic diagram of an exemplary system architecture to which the technical solution of the embodiments of the present application can be applied is shown.

[0035] like Figure 1 As shown, the system architecture may include a driving risk estimation device 101, a network 102, and (eg Figure 1 The network 102 is used as a medium for providing a communication link between the driving risk estimation device and the terminal device. The network 102 may include various connection types, such as a wired communication link, a wireless communication link, etc.

[0036] It should be understood that Figure 1The number of driving risk estimation devices, networks, and terminal devices is only illustrative. Depending on the implementation requirements, there may be any number of driving risk estimation devices, networks, and terminal devices. When devices with the same function are working at the same time, one may be selected and used, or both may be used. For example, the terminal device in this embodiment may be a vehicle-mounted terminal 103 installed on a vehicle, or a mobile terminal of a driver having a terminal device function, etc., and may not be limited here.

[0037] The driving risk estimation device 101 in this embodiment can be installed on a vehicle to estimate the risk of the vehicle during driving. It can also be a device external to the vehicle to estimate the risk of one or more vehicles at the same time, and then send the estimated risk results to a terminal device.

[0038] The user can use the driving risk estimation device 101 to interact with the terminal device through the network 102 to receive or send messages, etc. The terminal device can be a terminal device that provides various information displays and voice instructions. For example, the driving risk estimation device 101 obtains various actions that may occur when the driving risk estimation device is working, and obtains the driving risk state corresponding to each driving risk type. If the action change of the driving risk estimation device causes the driving risk state to change, the corresponding state transition probability is determined to estimate the risk coefficient in the driving process according to the state transition probability, and then the total benefit corresponding to the action performed by the driving risk estimation device within the preset time period is determined according to all the risk coefficients estimated within the preset time period; finally, the action of the driving risk estimation device after this estimation is determined according to the risk coefficient and the total benefit, so as to perform the risk estimation work according to the action, while ensuring the accuracy of the risk estimation, the benefit of the estimated action is improved, and the resource consumption of the driving risk estimation device is reduced.

[0039] It should be noted that the driving risk estimation method provided in the embodiment of the present application is generally executed by a driving risk estimation device. Accordingly, the driving risk estimation device is generally arranged in a computer or a terminal device. Therefore, in other embodiments of the present application, the terminal device may also have similar functions as the driving risk estimation device, so as to execute the driving risk estimation method provided in the embodiment of the present application.

[0040] Figure 2 A schematic diagram of an exemplary system architecture to which the technical solution of the embodiments of the present application can be applied is shown.

[0041] like Figure 2As shown, the system architecture may include a driving risk estimation device 201, a vehicle control device 202, and a vehicle 203. The driving risk estimation device 201 obtains the driving risk state corresponding to each driving risk type. If the action change of the driving risk estimation device 201 causes the driving risk state change, the corresponding state transition probability is determined to estimate the risk coefficient in the driving process according to the state transition probability. Then, according to all the risk coefficients estimated in the preset period, the total benefit corresponding to the action performed by the driving risk estimation device 201 in the preset period is determined; finally, the action of the driving risk estimation device 201 after this estimation is determined according to the risk coefficient and the total benefit, so as to perform the risk estimation work according to the action, obtain the risk estimation result, and send the risk estimation result to the vehicle control device 202, so that the vehicle control device 202 generates a control instruction based on the risk estimation result, and sends the control instruction to the vehicle 203 to control the vehicle to perform the action corresponding to the control instruction. When there is driving in the estimation result of the driving risk estimation device 201, the vehicle control device 202 can trigger the control instruction in time to control the vehicle 303 to drive automatically and safely.

[0042] At the same time, in the case of human driving, the driving risk estimation device 201 can also have a display function of a terminal device, so as to replace the terminal device installed on the vehicle, estimate the vehicle risk, and display the estimation result on the screen. Further, the driving risk estimation device 201 can also have a voice broadcast function, which is used to remind the driver to drive carefully by voice broadcast when a risk is detected.

[0043] The implementation details of the technical solution of the embodiment of the present application are described in detail below:

[0044] Figure 3 A flowchart of a driving risk estimation method according to an embodiment of the present application is shown. The driving risk estimation method can be performed by a driving risk estimation device. The driving risk estimation device can be Figure 1 The driving risk estimation device shown in . Figure 3 As shown, the driving risk estimation method includes at least steps S310 to S340, which are described in detail as follows:

[0045] In step S310, various actions that may occur when the driving risk estimation device is working are obtained, and the driving risk state corresponding to each driving risk type is obtained.

[0046] like Figure 4As shown, the method of this embodiment can be used for safety assisted driving, vehicle networking, vehicle-road collaboration, and autonomous driving products on the product side, especially safety assisted driving, vehicle networking, vehicle-road collaboration, and autonomous driving products that need to mine driver risky driving behaviors. For example, when autonomous driving becomes more and more common, most of the vehicles on the road may be autonomous driving vehicles. In this case, the accuracy of risk estimation needs to be further increased to ensure the safety of autonomous driving.

[0047] The application scenario of this embodiment needs to meet one or more of the following conditions: 1) Partial information on vehicle driving risk is observable; 2) The condition of the vehicle will not change drastically in a short period of time; 3) The instantaneous gain or loss after each driving risk estimation is obtainable and quantifiable; 4) The differences in risks of different Internet of Vehicles are identifiable; 5) The driver's condition will not change drastically in a short period of time. For example, the driver will not have unexpected abnormalities, and his physical and mental conditions will not change drastically in a short period of time; 6) There is a difference between the importance of previous indicators and the importance of future indicators; 7) On the highway, vehicles within a certain speed range should maintain a certain distance. For example, if the speed exceeds 100 kilometers per hour, the safe distance between vehicles in the same lane should be maintained at more than 100 meters; if the speed is less than 100 kilometers, the safe distance between vehicles in the same lane should be maintained at more than 50 meters.

[0048] In one embodiment of the present application, various actions that may occur when the driving risk estimation device is working include two possible actions: continuing to estimate Δ 1 and stop estimating Δ 2 , and thus the action set of the driving risk estimation device during operation is obtained as A = {Δ 1 ,Δ 2}.

[0049] Optionally, for each type of driving risk to be estimated, the estimated time within a preset period is obtained. Within the preset period, the number of estimated driving risks to be estimated is limited, and all estimated times within the time period are obtained. The set of estimated times is recorded as T = {1, 2, ..., N}, where 1, 2, ..., N is used to represent the time identifier of the estimated time, that is, the first time, the second time, and the Nth time, etc.

[0050] In one embodiment of the present application, the state corresponding to the driving risk type in step S310 includes a hidden state and an explicit state, and the process of obtaining the driving risk state corresponding to each driving risk type includes the following steps:

[0051] Obtaining driving risk types and an estimation process corresponding to each driving risk type;

[0052] According to the estimation process, the state of the historical data corresponding to the driving risk type is decomposed to obtain a hidden state and an explicit state corresponding to the driving risk type, wherein the data corresponding to the hidden state includes the sampled risk data, and the data corresponding to the explicit state includes the risk coefficient.

[0053] In one embodiment of the present application, various driving risk types are pre-set in the present embodiment, wherein the driving risk types may include but are not limited to vehicle collision risk, road narrowing risk, and roadside sharp turn risk.

[0054] like Figure 5 As shown, in one embodiment of the present application, during the risk estimation process performed by the driving risk estimation device, each risk estimation result obtained will correspond to an estimation process. The estimation process can be, by obtaining the environmental data during the current driving process, that is, the driving risk true value 501, screening out the risk data related to the current possible risk, that is, the data corresponding to the hidden state 502, and analyzing based on these data, obtaining the risk type and the risk coefficient corresponding to the risk type, that is, the data corresponding to the visible state 503, and finally performing a driving risk warning 504 on the terminal device according to the visible state. In summary, the process can be described as the driving risk true value first becoming a hidden state, and then becoming a visible state, to assist in driving risk warning.

[0055] For example, in the process of identifying the risk of road narrowing, when the environmental data of the current driving process is obtained, it is identified that the current risk type may be the risk of road narrowing, and the remaining data not associated with the road data, such as the surrounding vehicle data, can be deleted to obtain the road data as the data corresponding to the hidden state. Then, data analysis is performed based on the road data to obtain the risk type and risk coefficient, that is, the data corresponding to the visible state of the observation value, and the risk type and risk coefficient are displayed on the terminal interface.

[0056] Optionally, for each type of driving risk to be estimated, the hidden state and the explicit state correspond to the hidden state. The set of hidden states is denoted as S = {s t |t∈T}, let the set of explicit states be Z={z t |t∈T}, where s t represents the hidden state detected at time t that cannot be accessed by the terminal interface, z t It indicates the display status detected at time t, which the user can directly obtain through the terminal interface.

[0057] In one embodiment of the present application, in a specific risk estimation process, the collision risk of any two vehicles is calculated and outputted by collecting information such as the number of vehicles, vehicle type, vehicle speed, vehicle position, vehicle acceleration, vehicle direction, gender of the vehicle driver, age of the vehicle driver, driving experience of the vehicle driver, type of vehicle driver, health status of the vehicle driver, and reaction ability of the vehicle driver. The driving risk estimation device brings the collected data into the gravitational field theory model in the field of physics to obtain the driving risk brought to the vehicle by non-moving objects, brings it into the spring potential energy model to calculate the driving risk brought to the vehicle by road signs or other similar objects, brings it into the gravitational field model and the Doppler effect model to calculate the driving risk sampling value of moving objects, and the sum of these three types of driving risks is the sampling value of the driving risk of the driving risk estimation device, that is, the hidden state, and finally obtains the observation value based on the hidden state, that is, the visible state.

[0058] In one embodiment of the present application, the driving risk estimation device in this embodiment samples and analyzes the road conditions to obtain a driving risk sampling value, but the driving risk sampling value is invisible to the driver, and the driving risk estimation device needs to display it in some way (for example, transmit it to the vehicle-side display screen for display) so that the driver can see it. The displayed value that the driver can see is the explicit state, which can correspond to the risk coefficient displayed on the terminal interface, and the driving risk sampling value of the driving risk estimation device that the driver cannot see is the hidden state, that is, the corresponding driving risk sampling value. In order to more intuitively illustrate the meaning of the hidden state and the explicit state, the hidden state value and the explicit state value are used in this embodiment to represent the size of the hidden state and the explicit state respectively, please refer to Table 1 and Table 2 as shown:

[0059]

[0060] Table 1 Hidden state values ​​of the mutual collision risk of four vehicles

[0061]

[0062] Table 2 The apparent state values ​​of the mutual collision risk of four vehicles

[0063] As shown in Table 1 and Table 2, Table 1 and Table 2 respectively list the hidden state values ​​and explicit state values ​​of the mutual collision risks of the four vehicles. Due to the delay of data acquisition, transmission and processing, there will be certain experiments during the acquisition of data to obtain the hidden state and then the hidden state is processed to obtain the explicit state, which may cause the driving environment to change. Therefore, for the same risk type, the corresponding hidden state and explicit state may change. In this embodiment, step S610 is used to correct the data processing operation error in the risk estimation process according to the state transition probability, so as to minimize the difference between the hidden state and the explicit state and improve the accuracy of the estimation.

[0064] In step S320, if the action change of the driving risk estimation device causes a change in the driving risk state, a corresponding state transition probability is determined, and the risk coefficient in the driving process is estimated according to the state transition probability.

[0065] In one embodiment of the present application, when the action of the driving risk estimation device changes, it is very likely to affect the driving risk state. In this case, if the change in the action of the driving risk estimation device causes a change in the driving risk state, the corresponding state transition probability is determined.

[0066] In one embodiment of the present application, the state transition probability in this embodiment includes a hidden state transition probability and an explicit state transition probability. The hidden state transition probability is used to indicate the probability that the hidden state changes after the action of the driving risk estimation device changes; the explicit state transition probability is used to indicate the probability that the explicit state changes after the action of the driving risk estimation device changes.

[0067] Specifically, when determining the hidden state transition probability, first determine whether the hidden state is a discrete state. If so, obtain its state transition probability density; if not, obtain its state transition probability law. The hidden state transition probability density or state transition law is denoted as Indicates that the current hidden state is s t , take action a t After that, the hidden state at the next moment is s t+1 The probability of A represents the action set.

[0068] Specifically, when determining the explicit state transition probability, first determine whether the explicit state is a discrete state. If so, obtain its state transition probability density; if not, obtain its state transition probability law. The state transition probability density or state transition probability law is denoted as φ(z t+1 |s t+1 ,a t ), indicating that action a is being taken t The hidden state is transferred to s t+1 The post-display state is z t+1 probability.

[0069] In one embodiment of the present application, the process of estimating the risk factor in the driving process according to the state transition probability in step S320 includes the following steps:

[0070] Determine the hidden state value corresponding to the adjacent time according to the state transition probability and the risk data corresponding to the hidden state collected at the current time;

[0071] According to the hidden state values ​​corresponding to the adjacent moments, the explicit state value corresponding to the hidden state value is determined as the risk coefficient.

[0072] In one embodiment of the present application, the hidden state value and the visible state value of the next moment after the current moment do not exist at the current moment, so there is no existing hidden state value and visible state value of the next moment at the current moment. Therefore, the role of the transition probability is to predict the hidden state value of the next moment based on the hidden state value of the current moment, and then predict the visible state value of the future moment, so as to predict the risk coefficient of the next moment.

[0073] For example, based on the current driving risk sampling value (i.e., the hidden state value s t ) and hidden state transition probability Among them, a t The value of Δ indicates continued sampling. 1 , we can get the driving risk sampling value s at the next moment t+1 ; Then based on the driving risk sampling value s at the next moment t+1 and the explicit state transition probability φ(z t |s t ,a t-1 ) can get the driving risk observation value z at the next moment t+1 (i.e. risk factor).

[0074] In one embodiment of the present application, Figure 6 As shown, the process of estimating the risk coefficient in the driving process according to the state transition probability in step S320 includes steps S610 to S630, which are specifically described as follows:

[0075] In step S610, a data processing operation error in the risk estimation process is corrected according to the state transition probability.

[0076] In one embodiment of the present application, in the process of estimating risk by the driving risk estimation device, there are often many data processing operation errors. For example, in the process of acquiring data, sampling data, and generating estimation results, errors are often generated due to various factors such as delay and interference. Therefore, this embodiment improves the accuracy of risk estimation by correcting the data processing operation errors in the risk estimation process according to the state transition probability.

[0077] In one embodiment of the present application, the process of correcting the data processing operation error in the risk estimation process according to the state transition probability in step S610 includes the following steps:

[0078] Determine a first sample variance corresponding to the state transition probability at the current moment and a second sample variance corresponding to the state transition probability at an adjacent moment;

[0079] Determining a correction method for the data processing operation error according to a difference between the first sample variance and the second sample variance and a preset variance difference threshold;

[0080] The data processing operation error is corrected according to the correction method.

[0081] Specifically, according to the state transition probability at the current moment, the first sample variance corresponding to the state transition probability is determined, and the second sample variance corresponding to the adjacent moment is determined. The specific calculation method is to calculate the sample variance corresponding to the current probability according to the state transition probability of the normal distribution. After the first sample variance and the second sample variance are calculated, the difference between the first sample variance and the second sample variance is calculated, and the difference between the two is compared with a preset variance difference threshold, and the correction method of the data processing operation error is determined according to the comparison result, so as to correct the data processing operation error according to the correction method.

[0082] Furthermore, we can also obtain multiple driving risk observation values ​​after this moment, namely, the risk coefficient z 1 , z 2 ,···,z t , z t+1 , and use its sample variance as the risk estimation error at time t+1 to correct the data processing operation error.

[0083] For example, according to the driving risk observation value at the current moment (i.e., the display state value z t ) and the explicit state transition probability φ(z t |s t ,a t-1 ) corrects the data transmission error of the driving risk estimation device during the driving risk estimation process; according to the driving risk sampling value (i.e., the hidden state value s t ) and hidden state transition probability Correct the data sampling error in the driving risk estimation process.

[0084] In step S620, the state transition probability is updated according to the corrected data processing operation error.

[0085] In the above steps, after the data processing operation error is corrected, the state transition probability is updated according to the corrected data processing operation error. The specific updating method is to update action a t The action details of the specific estimation steps included in the estimation, such as the action details of data sampling, are used to re-determine the state transition probabilities corresponding to different actions and states according to the updated action details.

[0086] In step S630, the risk factor in the driving process is estimated according to the updated state transition probability.

[0087] After the state transition probability is updated, a new state transition probability is obtained, and the risk coefficient in the driving process is re-estimated based on the updated state transition probability and the risk coefficient estimation process in step S320.

[0088] In this embodiment, a correction method of the state transition probability is determined based on the state transition probability at the current moment and the next moment, so as to obtain a more accurate explicit state value through the implicit state value, thereby improving the accuracy of risk estimation.

[0089] Continue to refer to Figure 3 As shown, in step S330, based on all risk coefficients estimated within a preset period of time, the total benefit corresponding to the actions performed by the driving risk estimation device within the preset period of time is determined.

[0090] In one embodiment of the present application, the risk coefficient obtained by the estimated actions within a period of time can correspond to the total benefit or total loss obtained by the actions within the period of time. In this embodiment, the total benefit or total loss is used to measure the resource consumption within the period of time.

[0091] In one embodiment of the present application, Figure 7 As shown, the process of determining the total benefit corresponding to the action performed by the driving risk estimation device within the preset time period according to all risk coefficients estimated within the preset time period in step S330 includes steps S710 to S730, which are described as follows:

[0092] In step S710, all risk factors within a preset time period are obtained.

[0093] In one embodiment of the present application, there will be multiple risk estimates within a preset time period, and each risk estimate will consume a portion of resources to obtain a risk coefficient. Therefore, in this step, the risk coefficient obtained from each estimate within the preset time period is first obtained.

[0094] In step S720, the instantaneous return corresponding to the risk factor is determined from the historical return data.

[0095] In one embodiment of the present application, the instantaneous gain or loss after each estimation is obtained. The instantaneous gain and loss respectively represent the positive contribution or negative impact of a certain estimation result of the driving risk estimation device on the driving risk judgment. For example, the positive impact may refer to the improvement of the accuracy of the driving risk estimation, which is beneficial to the driving risk judgment, and the negative effect may refer to the consumed vehicle-road collaborative resources.

[0096] In one embodiment of the present application, the total benefit is the difference between the benefit and the loss, and the instantaneous benefit and loss are recorded as R and C respectively. Exemplarily, R can be obtained by statistical historical data. For example, the estimation error of driving risk will mislead the vehicle, cause traffic accidents, and bring economic losses, but each estimation can reduce the estimation error, thereby helping to reduce traffic accidents and thus reduce economic losses. Therefore, R can be obtained through statistical historical data. For example, the economic loss reduction amount for each 1% increase in the driving risk estimation accuracy, R can be taken as this economic reduction amount. C can be taken as the cost of the driving risk estimation device, such as energy consumption, such as maintenance costs, and can also be obtained through historical records and voltage and current tests. The power consumption can be calculated by voltage and current. Optionally, the above data is determined in advance and can be stored in the cloud, so R and C can be obtained from the cloud.

[0097] In step S730, the expectation of all instantaneous benefits within the preset period is calculated, and the total benefit is obtained according to the expectation of all instantaneous benefits within the preset period.

[0098] In one embodiment of the present application, each time the driving risk estimation device executes a step of continuing the estimation action or stopping the estimation action, it estimates the difference between the instantaneous gain R and the loss C as the total gain or total loss based on the observation value at the current moment, that is, the risk coefficient, and uses this to measure the resource consumption of the estimation action.

[0099] In one embodiment of the present application, the total gain is the difference between the gain and the loss, and the instantaneous gain and loss are recorded as R and C respectively. Because the instantaneous gain R and loss C are random, in this embodiment, the expectation of the instantaneous gain R and loss C corresponding to multiple estimated actions within a preset period is calculated to obtain E(R|s t ,z t ,a t ) and E(C|s t ,z t ,a t ), and the difference between these two expectations is taken as the total benefit.

[0100] In one embodiment of the present application, Figure 8 As shown, the process of calculating the expectation of all instantaneous benefits within the preset time period in step S730 includes steps S810 to S820, which are described as follows:

[0101] In step S810, a discount factor corresponding to each risk coefficient is determined according to the estimation time corresponding to each risk coefficient.

[0102] In one embodiment of the present application, when each risk coefficient is estimated, there is a corresponding estimation time. In this embodiment, the discount factor corresponding to each risk coefficient is determined according to the estimation time corresponding to the risk coefficient.

[0103] Specifically, the time factor in this embodiment is calculated according to the estimated time 1, 2, ..., N in step S310, and is specifically equal to the difference of the reciprocal of the estimated time 1, that is, the discount factor is determined by the following formula:

[0104]

[0105] In this embodiment, γ represents the difference between the importance of the current indicator and the importance of the future indicator, because as the working time of the current vehicle driving risk estimation device increases, the importance of its related indicators will become weaker.

[0106] In step S820, the expectation of all the instantaneous returns is calculated according to the discount factor corresponding to each of the risk coefficients and the instantaneous return.

[0107] In one embodiment of the present application, when calculating the expectation of all instantaneous returns within a preset time, the discount factor corresponding to each risk coefficient is used as the weight of its corresponding instantaneous return to calculate the expectation of all instantaneous returns within the preset time period.

[0108] In one embodiment of the present application, after calculating the expectation of all instantaneous benefits within a preset period, the total benefit is obtained according to the expectation of all instantaneous benefits within the preset period. The specific way to obtain the total benefit can be to multiply the expectation by a preset total benefit calculation coefficient to obtain the total benefit; or to perform corresponding calculation processing on the expectation to obtain the total benefit, which is not limited here.

[0109] In step S340, the action of the driving risk estimation device after the current estimation is determined according to the risk coefficient and the total benefit.

[0110] In one embodiment of the present application, after the total benefit is calculated, the work of the driving risk estimation device after this estimation is determined based on the risk coefficient and the total benefit. Among them, when the total benefit has reached the maximum during this period, the driving risk estimation device stops estimating the driving risk, otherwise, it continues to estimate the driving risk. Optionally, if the total loss is calculated, because the benefit can be regarded as a negative loss, when the total loss has reached the minimum, the driving risk estimation device stops estimating the driving risk, otherwise, it continues to estimate the driving risk. In this way, the total benefit of this risk estimation within this period is the highest, or the total loss is the lowest, so as to achieve the purpose of reducing resource consumption in the estimation process.

[0111] In one embodiment of the present application, the process of determining the action of the driving risk estimation device after the current estimation according to the risk coefficient and the total benefit in step S340 includes:

[0112] An action sequence after the current estimation is determined according to the risk coefficient and the total benefit, wherein the action sequence includes estimation actions to be executed and an execution timing of each estimation action to be executed.

[0113] In one embodiment of the present application, the action sequence is the action that the driving risk estimation device should take at each estimation moment, that is, a sequence of estimating driving risk or stopping estimating driving risk arranged in chronological order. After obtaining the action sequence, the estimation work is performed according to the estimated actions and their execution sequence.

[0114] For example, if the actions that the driving risk estimation device should perform at the first three moments are "estimate driving risk", "estimate driving risk", and "stop estimating driving risk", then the action sequence of the driving risk estimation device at the first three moments is a1, a2, a 3 , where a1, a2, a 3 They respectively stand for "estimate driving risk", "estimate driving risk", and "stop estimating driving risk".

[0115] Optionally, in this embodiment, a partially observable Markov decision process of a driving risk estimation device is obtained based on an action set, a decision time set, a hidden state set, an explicit state set, a hidden state transition probability density or a state transition probability, an explicit state transition probability density or a state transition probability, and a gain or loss, and the action sequence of the optimal strategy is obtained by using a partially observable Markov decision method:

[0116]

[0117] Among them, x is used to indicate the identifier of the estimated time, a x =Δ 1 Used to indicate that the action taken at the estimation time of x is to continue estimating Δ 1 , a x =Δ 2 Used to indicate that the action taken at the estimation time of x is to stop estimating Δ 2 Optionally, the specific optimal strategy can be calculated by first converting it into a linear programming problem, then solving it using a simplex or interior point method, and then determining the action sequence of the driving risk estimation device according to the optimal control strategy to maximize the total benefit or minimize the total loss.

[0118] The following describes an embodiment of the device of the present application, which can be used to execute the driving risk estimation method in the above embodiment of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the driving risk estimation method in the above embodiment of the present application.

[0119] Fig. 9A block diagram of a control device of a driving risk estimation device according to an embodiment of the present application is shown.

[0120] Reference Fig. 9 As shown, a control device 900 of a driving risk estimation device according to an embodiment of the present application includes:

[0121] A first acquisition unit 901 is used to acquire various actions that may occur when the driving risk estimation device is working, and to acquire a driving risk state corresponding to each driving risk type;

[0122] A first estimation unit 902 is configured to determine a corresponding state transition probability if the action change of the driving risk estimation device causes a driving risk state change, and estimate a risk coefficient in the driving process according to the state transition probability;

[0123] The benefit unit 903 is used to determine the total benefit corresponding to the actions performed by the driving risk estimation device within the preset time period according to all risk coefficients estimated within the preset time period;

[0124] The action unit 904 is used to determine the action of the driving risk estimation device after the current estimation according to the risk coefficient and the total benefit.

[0125] In one embodiment of the present application, the state corresponding to the driving risk type includes a hidden state and an explicit state; the first acquisition unit 901 includes:

[0126] A second acquisition unit, used to acquire a driving risk type and an estimation process corresponding to each driving risk type;

[0127] A state decomposition unit is used to decompose the state of the historical data corresponding to the driving risk type according to the estimation process to obtain a hidden state and an explicit state corresponding to the driving risk type, wherein the data corresponding to the hidden state includes the sampled risk data, and the data corresponding to the explicit state includes the risk coefficient.

[0128] In one embodiment of the present application, the first estimation unit 902 includes:

[0129] A first determining unit, configured to determine a hidden state value corresponding to an adjacent time point according to the state transition probability and risk data corresponding to the hidden state collected at the current time point;

[0130] The second determining unit is used to determine the explicit state value corresponding to the hidden state value according to the hidden state value corresponding to the adjacent time moments as the risk coefficient.

[0131] In one embodiment of the present application, the first estimation unit 902 includes:

[0132] A first correction unit, used for correcting a data processing operation error in a risk estimation process according to the state transition probability;

[0133] An updating unit, configured to update the state transition probability according to the corrected data processing operation error;

[0134] The second estimation unit is used to estimate the risk coefficient in the driving process according to the updated state transition probability.

[0135] In one embodiment of the present application, the first correction unit includes:

[0136] A third determining unit, used to determine a first sample variance corresponding to the state transition probability at a current moment and a second sample variance corresponding to the state transition probability at an adjacent moment;

[0137] a fourth determining unit, configured to determine a correction method for the data processing operation error according to a difference between the first sample variance and the second sample variance and a preset variance difference threshold;

[0138] The second correction unit is used to correct the data processing operation error according to the correction method.

[0139] In one embodiment of the present application, the state transition probability includes a hidden state transition probability and an explicit state transition probability; the first correction unit includes:

[0140] A third correction unit, used to correct the data sampling error in the risk estimation process according to the hidden state transition probability;

[0141] The fourth correction unit is used to correct the data transmission error in the risk estimation process according to the explicit state transition probability.

[0142] In one embodiment of the present application, the revenue unit 903 includes:

[0143] A second acquisition unit is used to acquire all risk factors within a preset period of time;

[0144] A fifth determining unit, used to determine the instantaneous return corresponding to the risk coefficient from the historical return data;

[0145] The first calculation unit is used to calculate the expectation of all instantaneous benefits within the preset time period, and obtain the total benefit according to the expectation of all instantaneous benefits within the preset time period.

[0146] In one embodiment of the present application, the first computing unit includes:

[0147] A factor unit, used for determining a discount factor corresponding to each risk coefficient according to the estimation moment corresponding to each risk coefficient;

[0148] The second calculation unit is used to calculate the expectation of all the instantaneous returns according to the discount factor corresponding to each of the risk coefficients and the instantaneous returns.

[0149] In one embodiment of the present application, the driving risk estimation device also includes: an action sequence unit, which is used to determine an action sequence after this estimation based on the risk coefficient and the total benefit, and the action sequence includes the estimation actions to be executed and the execution timing of each of the estimation actions to be executed.

[0150] Fig.10 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown.

[0151] It should be noted that Fig.10 The computer system 1000 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0152] like Fig.10 As shown, the computer system 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1002 or the program loaded from the storage part 1008 to the random access memory (RAM) 1003, such as executing the method described in the above embodiment. In the RAM 1003, various programs and data required for system operation are also stored. The CPU 1001, the ROM 1002 and the RAM 1003 are connected to each other through the bus 1004. The input / output (I / O) interface 1005 is also connected to the bus 1004.

[0153] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1010 as needed so that a computer program read therefrom is installed into the storage section 1008 as needed.

[0154] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 1009, and / or installed from a removable medium 1011. When the computer program is executed by a central processing unit (CPU) 1001, various functions defined in the system of the present application are executed.

[0155] It should be noted that the computer-readable medium shown in the embodiment of the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by an instruction execution system, device or device or used in combination with it. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0156] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0157] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. The names of these units do not constitute limitations on the units themselves in some cases.

[0158] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment.

[0159] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0160] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the implementation methods of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation methods of the present application.

[0161] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0162] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A driving risk estimation method, characterized in that: include: Acquire various actions that may occur when the driving risk estimation device is working, and acquire the driving risk state corresponding to each driving risk type; If the change in the action of the driving risk estimation device causes a change in the driving risk state, determining a corresponding state transition probability, and estimating a risk coefficient during driving according to the state transition probability; Determining, based on all risk factors estimated within a preset period of time, a total benefit corresponding to an action performed by the driving risk estimation device within the preset period of time; The total benefit is used to characterize the resource consumption of the driving risk estimation device in performing driving risk estimation within the preset time period; The action of the driving risk estimation device after the current estimation is determined according to the risk coefficient and the total benefit; wherein the types of the action include the driving risk estimation device continuing the driving risk estimation and the driving risk estimation device stopping the driving risk estimation.

2. The method according to claim 1, characterized in that The states corresponding to the driving risk types include hidden states and visible states; Get the status corresponding to each driving risk type, including: Obtaining driving risk types and an estimation process corresponding to each driving risk type; According to the estimation process, the state of the historical data corresponding to the driving risk type is decomposed to obtain a hidden state and an explicit state corresponding to the driving risk type, wherein the data corresponding to the hidden state includes the sampled risk data, and the data corresponding to the explicit state includes the risk coefficient.

3. The method according to claim 2, characterized in that The risk factor in the driving process is estimated according to the state transition probability, including: Determine the hidden state value corresponding to the adjacent time according to the state transition probability and the risk data corresponding to the hidden state collected at the current time; According to the hidden state values ​​corresponding to the adjacent moments, the explicit state value corresponding to the hidden state value is determined as the risk coefficient.

4. The method according to claim 1, characterized in that: The risk factor in the driving process is estimated according to the state transition probability, including: Correcting data processing operation errors in the risk estimation process according to the state transition probability; updating the state transition probability according to the corrected data processing operation error; The risk factor during driving is estimated based on the updated state transition probability.

5. The method according to claim 4, characterized in that Correcting data processing operation errors in the risk estimation process according to the state transition probability includes: Determine a first sample variance corresponding to the state transition probability at the current moment and a second sample variance corresponding to the state transition probability at an adjacent moment; Determining a method for correcting the data processing operation error according to a difference between the first sample variance and the second sample variance and a preset variance difference threshold; The data processing operation error is corrected according to the correction method.

6. The method according to claim 4, characterized in that The state transition probability includes a hidden state transition probability and an explicit state transition probability; Correcting data processing operation errors in the risk estimation process according to the state transition probability includes: Correcting data sampling errors in the risk estimation process according to the hidden state transition probability; The data transmission error in the risk estimation process is corrected according to the explicit state transition probability.

7. The method according to claim 1, characterized in that Determining the total benefit corresponding to the action performed by the driving risk estimation device within the preset time period according to all risk coefficients estimated within the preset time period includes: Obtain all risk factors within a preset period; Determine the instantaneous return corresponding to the risk factor from the historical return data; The expectation of all instantaneous benefits within the preset time period is calculated, and the total benefit is obtained according to the expectation of all instantaneous benefits within the preset time period.

8. The method according to claim 7, characterized in that Calculating the expectation of all instantaneous returns within the preset period, including: Determine a discount factor corresponding to each risk factor according to the estimated time corresponding to each risk factor; According to the discount factor corresponding to each of the risk coefficients and the instantaneous return, the expectation of all the instantaneous returns is calculated.

9. The method according to any one of claims 1 to 8, characterized in that: Determining actions after this estimation based on the risk factor and the total benefit includes: An action sequence after the current estimation is determined according to the risk coefficient and the total benefit, wherein the action sequence includes estimation actions to be executed and an execution timing of each estimation action to be executed.

10. A driving risk estimation device, characterized in that: include: A first acquisition unit, used to acquire various actions that may occur when the driving risk estimation device is working, and to acquire a driving risk state corresponding to each driving risk type; a first estimation unit, configured to determine a corresponding state transition probability if a change in the operation of the driving risk estimation device causes a change in the driving risk state, and estimate a risk coefficient in the driving process according to the state transition probability; A benefit unit, configured to determine a total benefit corresponding to an action performed by the driving risk estimation device within a preset period of time according to all risk coefficients estimated within the preset period of time; The total benefit is used to characterize the resource consumption of the driving risk estimation device in performing driving risk estimation within the preset time period; An action unit is used to determine the action of the driving risk estimation device after the current estimation according to the risk coefficient and the total benefit; wherein the types of the action include the driving risk estimation device continuing to perform driving risk estimation and the driving risk estimation device stopping driving risk estimation.

11. A computer readable medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor, the driving risk estimation method according to any one of claims 1 to 9 is implemented.

12. An electronic device, characterized in that: include: at least one processor; A storage device for storing at least one program, when the at least one program is executed by the at least one processor, enables the at least one processor to implement the driving risk estimation method according to any one of claims 1 to 9.

13. A computer program product, characterized in that It includes a computer program carried on a computer-readable medium, and when the computer program is executed, the driving risk estimation method according to any one of claims 1 to 9 is implemented.

Images