CN114331019A - Urban traffic safety risk real-time assessment method and device based on risk factor - Google Patents
Urban traffic safety risk real-time assessment method and device based on risk factor Download PDFInfo
- Publication number
- CN114331019A CN114331019A CN202111451889.1A CN202111451889A CN114331019A CN 114331019 A CN114331019 A CN 114331019A CN 202111451889 A CN202111451889 A CN 202111451889A CN 114331019 A CN114331019 A CN 114331019A
- Authority
- CN
- China
- Prior art keywords
- risk
- real
- factors
- time
- traffic safety
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000013210 evaluation model Methods 0.000 claims abstract description 50
- 238000011156 evaluation Methods 0.000 claims abstract description 37
- 238000012502 risk assessment Methods 0.000 claims abstract description 26
- 206010039203 Road traffic accident Diseases 0.000 claims description 51
- 238000003062 neural network model Methods 0.000 claims description 24
- 238000013528 artificial neural network Methods 0.000 claims description 18
- 230000003068 static effect Effects 0.000 claims description 14
- YBJHBAHKTGYVGT-ZKWXMUAHSA-N (+)-Biotin Chemical compound N1C(=O)N[C@@H]2[C@H](CCCCC(=O)O)SC[C@@H]21 YBJHBAHKTGYVGT-ZKWXMUAHSA-N 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 13
- FEPMHVLSLDOMQC-UHFFFAOYSA-N virginiamycin-S1 Natural products CC1OC(=O)C(C=2C=CC=CC=2)NC(=O)C2CC(=O)CCN2C(=O)C(CC=2C=CC=CC=2)N(C)C(=O)C2CCCN2C(=O)C(CC)NC(=O)C1NC(=O)C1=NC=CC=C1O FEPMHVLSLDOMQC-UHFFFAOYSA-N 0.000 claims description 13
- 238000010606 normalization Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 4
- 238000007477 logistic regression Methods 0.000 claims description 4
- 239000004576 sand Substances 0.000 claims description 4
- 208000012260 Accidental injury Diseases 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 239000008186 active pharmaceutical agent Substances 0.000 claims description 3
- 230000006378 damage Effects 0.000 claims description 3
- 230000034994 death Effects 0.000 claims description 3
- 231100000517 death Toxicity 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 11
- 238000004458 analytical method Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000013135 deep learning Methods 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000002265 prevention Effects 0.000 description 3
- 238000012954 risk control Methods 0.000 description 3
- 238000003786 synthesis reaction Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention provides a real-time urban traffic safety risk assessment method and device based on risk factors, wherein the method comprises the steps of firstly, identifying the risk factors to obtain a risk factor list containing all risk factors obtained through identification; then, calculating risk values of risk factors, and calculating key indexes of the risk factors in the risk factor list by analyzing historical accident data to obtain the risk values of the risk factors; then, real-time risk values of individual risk sources of people, vehicles, roads and enterprises are evaluated respectively by constructing an individual risk source risk value evaluation model, then group risk value grade evaluation is carried out, and finally the urban comprehensive traffic safety risk value is obtained by constructing an urban comprehensive traffic safety risk value evaluation model.
Description
Technical Field
The invention relates to the field of traffic safety management, in particular to a real-time urban traffic safety risk assessment method and device based on risk causing factors.
Background
With the development and progress of social economy, the automobile holding capacity is continuously increased, and the corresponding traffic safety accidents are increased, so that the loss of personnel and property is caused on one hand, and traffic jam is caused on the other hand, so that the evaluation and prediction of traffic safety are necessary.
In the prior art, solutions for traffic safety risk assessment can be roughly classified into three categories.
(1) An accident risk assessment method based on accident link analysis. The method establishes an accident risk link of a risk source and a risk type through a text mining technology, a data correlation mining technology, a multi-layer network theory and the like, and evaluates accident risks in the transportation process;
(2) a traffic safety risk assessment method based on machine learning algorithms such as deep learning and neural network. Some convolutional neural network algorithms are used for predicting traffic safety risks in the next moment of the urban area according to the traffic flow and the environmental data at the current moment; some of the vehicle journey risk value training methods are based on a full-connection neural network model, and take vehicle types, journey data, environment data, the number of various alarms in the journey and the like as input, and output the journey risk value of the vehicle through a training neural network.
(3) A single risk source or city comprehensive traffic safety risk assessment method based on key index analysis. The method carries out safety risk assessment on a single risk source by analyzing key indexes of the single risk source in the transportation field of road sections, drivers, environments and cities.
In the process of implementing the present invention, the inventors of the present application find that the methods in the prior art have at least the following technical problems:
although the technical scheme of accident risk assessment based on accident link analysis can establish the association between a risk source and a risk type to a certain extent, and can trace the source and manage and control traffic safety risk events in time, quantitative analysis is difficult to be carried out on the risk per se, and the risk cannot be hierarchically classified, so that the accuracy of traffic safety risk management and control cannot be guaranteed. In the existing traffic safety risk assessment technical scheme based on machine learning algorithms such as deep learning and neural network, because the input is a global index, the output risk prediction result is a coarse-grained risk value, and a specific strategy cannot be formulated according to the prediction result to realize the effect of pre-prevention. The technical scheme of single risk source or urban comprehensive traffic safety risk assessment based on key index analysis only considers risk factors of the single risk source, and is not beneficial to carrying out safety risk control of the system.
Disclosure of Invention
The invention provides a real-time urban traffic safety risk assessment method and device based on risk causing factors, which are used for solving or at least partially solving the technical problem that the assessment result in the prior art is not accurate enough.
In order to solve the technical problem, a first aspect of the present invention provides a real-time urban traffic safety risk assessment method based on risk factors, including:
s1: identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
s2: calculating key indexes of risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, wherein the risk values of all the risk factors form a risk value library of the risk factors;
s3: constructing an individual risk source risk value evaluation model, wherein the individual risk source risk value evaluation model comprises a human, vehicle, road and enterprise individual risk source risk value evaluation model and is used for respectively taking real-time risk factors of human, vehicle, road and enterprise individuals as characteristic input to respectively obtain real-time risk values of human, vehicle, road and enterprise individual risk sources, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
s4: real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group are evaluated in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
s5: and constructing an urban comprehensive traffic safety risk value evaluation model, wherein the urban comprehensive traffic safety risk value evaluation model is used for taking real-time traffic safety risk values of a driver group, a vehicle group, a road network group and a vehicle-enterprise group as input to obtain an urban comprehensive traffic safety risk value, and is a fully-connected MLP neural network model.
In one embodiment, the classification of the risk factors of people, vehicles, roads and enterprises in step S1 includes multiple levels, wherein the first level classification is divided into static factors and dynamic factors, the static factors represent static information that does not change in a short period, and the dynamic factors represent real-time dynamically changing information.
In one embodiment, the risk value of the risk factor is a risk value of each risk factor S as a leading factor, which causes a traffic accident, and S2 includes:
s2.1: calculating the grade of the possibility of the traffic accident caused by the risk factors;
s2.2: calculating the severity grade of the accident caused by risk factors;
s2.3: and combining the grade of the possibility of the occurrence of the traffic accident caused by the risk factors and the grade of the severity of the accident caused by the risk factors to obtain a risk value of the risk factors.
In one embodiment, S2.1 comprises:
s2.1.1: calculating the probability of the occurrence of the traffic accident caused by the risk factor S, wherein the frequency of the occurrence of the road traffic accident caused by the risk factor S as the main cause of the damage is fSProbability of PS，
Wherein n isSRepresenting the times of road traffic accidents caused by the risk factor S as a main cause in a period of time; n is the total number of all traffic accidents occurring within a period of time. According to Bernoulli's theorem in probability theory, when n is large, the probability that the frequency of accident has large deviation with probability is small, and f can be usedSProbability of occurrence of traffic accident P as risk factorS；
S2.1.2 calculates the level of likelihood that the risk factor will cause the traffic accident,
wherein < > represents rounding.
In one embodiment, S2.2 comprises:
s2.2.1: calculating the severity of the consequences of the traffic accident caused by the risk factors according to the measures of the severity of the traffic accident, wherein the measures of the severity of the traffic accident comprise the number of dead people in the accident, the number of injured people in the accident and the economic loss of the accident:
wherein, VSTo the extent that risk factors lead to the severity of the consequences of an accident, DSFor accident-related deaths due to risk factors, SSFor the number of accident injuries caused by risk factors ESThe accident economic loss is caused by risk factors;
s2.2.2: for the severity degree V of traffic accident consequences caused by risk factorsSThe normalization treatment is carried out, and the normalization treatment is carried out,
wherein,
min (V) for results after normalizationS) Max (V) which is the minimum value of the severity of the accident consequence caused by the risk factorS) The maximum value of the severity of the accident consequence caused by risk factors;
s2.2.3: and performing multiple value taking on the result of the normalization processing to obtain the severity grade of the traffic accident consequence caused by risk factors:
wherein,<>means that the rounding is carried out to round,
the grade of the severity degree of the traffic accident consequence caused by the risk factors is an integer of (0, 10).
In one embodiment, S2.3 multiplies the risk factor-causing traffic accident occurrence probability level and the risk factor-causing accident severity level to synthesize a risk value of the risk factor:
RSthe risk value representing the risk factor is a number of (0, 100).
In one embodiment, in step S3, the risk value assessment model of individual risk source comprises an input layer, a hidden layer and an output layer,
wherein, the input layer is represented by a vector X, the output of the hidden layer is S (W)1X+b1) Wherein W is1Is the weight of the input layer to the hidden layer, b1Is the bias from the input layer to the hidden layer, the function S represents the sigmoid function of the activation function, which is expressed as:
sigmoid(a)＝1/(1+e-a)
where a represents the input to the sigmod function;
the output of the hidden layer is represented by X1, hidden layer to inputOut-of-layer representation of a multi-class logistic regression, i.e. softmax regression, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer; the output of the output layer is represented as:
Y＝G(b2+W2(S(b1+W1X)))
x denotes the input of the neural network, Y denotes the output of the neural network, G denotes the softmax function, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer;
when the individual risk value evaluation is carried out through the individual risk source risk value evaluation model, the online prediction stage is adopted, and the input data set is the identification information of the individual risk source and the risk value of risk factors occurring in real time of the individual risk source and is expressed as a vector RSAnd outputting a real-time traffic safety risk value R which is an individual risk source and is expressed as:
R＝G(b2 *+W2 *(S(b1 *+W1 *RS)))
wherein R is a real-time traffic safety risk value of a human individual risk source, a real-time traffic safety risk value of a vehicle individual risk source, a real-time traffic safety risk value of a road individual risk source, and a real-time traffic safety risk value of an enterprise individual risk source.
In one embodiment, in S4, the real-time traffic safety risk value of the group is obtained by weighted calculation according to the real-time traffic safety risk value R of the individual risk source:
Rg＝∑Ri*ri
wherein R isiRepresenting a certain safety risk value, r, in a group of people or vehicles or roads or enterprises at the same timeiRepresenting the number of people or vehicles or roads or a certain security risk value in a group of enterprises at the same time.
In one embodiment, the urban integrated traffic safety risk value assessment model in step S5 includes an input layer, a hidden layer, and an output layer, and is obtained by offline training.
Based on the same inventive concept, the second aspect of the invention discloses an urban traffic safety risk real-time assessment device based on risk factors, which comprises:
the risk factor identification module is used for identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
the risk value calculation module of the risk factors is used for calculating key indexes of the risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, and the risk values of all the risk factors form a risk value library of the risk factors;
the individual risk source risk value evaluation module is used for constructing an individual risk source risk value evaluation model, the individual risk source risk value evaluation model comprises individual risk source risk value evaluation models of people, vehicles, roads and enterprises, the individual risk source risk value evaluation model is used for respectively taking real-time risk factors of the people, the vehicles, the roads and the enterprises as characteristic input to respectively obtain real-time risk values of the individual risk sources of the people, the vehicles, the roads and the enterprises, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
the group risk value grade evaluation module is used for evaluating real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
the urban comprehensive traffic safety risk value evaluation module is used for constructing an urban comprehensive traffic safety risk value evaluation model and obtaining an urban comprehensive traffic safety risk value by taking the real-time traffic safety risk values of a driver group, a vehicle group, a road network group and a vehicle-enterprise group as input, wherein the urban comprehensive traffic safety risk value evaluation model is a fully-connected MLP neural network model
One or more technical solutions in the embodiments of the present application have at least one or more of the following technical effects:
the urban traffic safety risk real-time assessment method based on risk factors, provided by the invention, comprises the steps of firstly, identifying the risk factors to obtain a risk factor list containing all risk factors obtained by identification; then, calculating risk values of risk factors, and calculating key indexes of the risk factors in the risk factor list by analyzing historical accident data to obtain the risk values of the risk factors; then, real-time risk values of individual risk sources of people, vehicles, roads and enterprises are evaluated respectively by constructing an individual risk source risk value evaluation model, then group risk value grade evaluation is carried out, and finally the urban comprehensive traffic safety risk value is obtained by constructing an urban comprehensive traffic safety risk value evaluation model.
The individual risk source risk value evaluation model and the urban comprehensive traffic safety risk value evaluation model adopt a fully-connected MLP neural network model, so that the method is an urban traffic safety risk real-time evaluation method based on an MLP neural network algorithm, and compared with other safety risk evaluation methods, the method has the characteristics of high-precision real-time prediction and more comprehensive and finer granularity of evaluation objects, and the input risk factors are classified in a grading manner, real-time refined indexes of people, vehicles, roads and enterprises are input, and the real-time dynamic safety risk values of individuals, groups and urban comprehensive real-time dynamic safety risk values of people, vehicles, roads and enterprises are output; meanwhile, the models and the corresponding parameters can be flexibly adjusted and combined according to actual needs, so that a timely management and control strategy can be made more favorably according to a prediction result and specific risk factors, and real-time traffic safety risk prevention and control can be realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of a real-time risk assessment method for urban traffic safety based on risk factors in an embodiment of the invention;
FIG. 2 is a schematic structural diagram of an MLP neural network model for evaluating individual risk source dynamic security risk values according to an embodiment of the present disclosure;
FIG. 3 is a flow chart of an MLP neural network model construction according to an embodiment of the present invention;
FIG. 4 is an overall framework diagram of the real-time risk assessment of urban traffic safety in an embodiment of the invention;
FIG. 5 is a convergence curve of the training phase of the MLP neural network model in an embodiment of the present invention.
Detailed Description
The inventor of the application finds out through a great deal of research and practice that:
(1) although the technical scheme of accident risk assessment based on accident link analysis can establish the association between a risk source and a risk type to a certain extent and perform timely tracing and control on traffic safety risk events, quantitative analysis is difficult to perform on risks per se, and the risks cannot be hierarchically classified, so that the accuracy of traffic safety risk control cannot be ensured;
(2) in the existing traffic safety risk assessment technical scheme based on machine learning algorithms such as deep learning and neural network, a neural network model is established by the existing deep learning method, but the input and the output of the neural network model are different from the application, the purpose of the neural network model is different from the application, the risk factors are not classified in a grading way by the existing method, the input of the neural network model is a global index, and the input of the neural network model is not real to real-time refined indexes of people, vehicles, roads and enterprises. Compared with the prior art, the method has the advantages that based on the fully-connected neural network model, a vehicle traffic safety risk evaluation network, a driver traffic safety risk evaluation network, a road section traffic safety risk evaluation network, an enterprise traffic safety risk evaluation network and an urban comprehensive traffic safety risk evaluation network model are constructed, static and dynamic real-time risk factors of vehicle-person-road-enterprise are input, vehicle-person-road-enterprise individual and group traffic safety risk levels and urban comprehensive traffic safety risk levels are output through the trained multilayer fully-connected neural network, and real-time early warning and pushing are carried out. Classifying the input risk factors in a grading way, inputting real-time refined indexes of people, vehicles, roads and enterprises, and outputting real-time dynamic security risk values of individuals and groups of the people, the vehicles, the roads and the enterprises and urban comprehensive real-time dynamic security risk values; meanwhile, the models and corresponding parameters can be flexibly adjusted and combined according to actual needs, so that a timely management and control strategy can be formulated according to a prediction result and specific risk factors, and real-time traffic safety risk prevention and control can be realized;
(3) the technical scheme of single risk source or urban comprehensive traffic safety risk assessment based on key index analysis only considers risk factors of the single risk source, does not comprehensively analyze direct risk factors, and is not beneficial to safety risk control of the system.
Compared with other safety risk evaluation methods, the method can evaluate the risk values of vehicle-person-road-enterprise individuals and groups in the urban traffic transportation process in real time, has the characteristics of high-precision real-time prediction and more comprehensive and finer granularity of evaluation objects, and is favorable for providing better data decision support for traffic safety control of related departments.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example one
The embodiment of the invention provides a real-time urban traffic safety risk assessment method based on risk factors, which comprises the following steps:
s1: identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
s2: calculating key indexes of risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, wherein the risk values of all the risk factors form a risk value library of the risk factors;
s3: constructing an individual risk source risk value evaluation model, wherein the individual risk source risk value evaluation model comprises a human, vehicle, road and enterprise individual risk source risk value evaluation model and is used for respectively taking real-time risk factors of human, vehicle, road and enterprise individuals as characteristic input to respectively obtain real-time risk values of human, vehicle, road and enterprise individual risk sources, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
s4: real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group are evaluated in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
s5: and constructing an urban comprehensive traffic safety risk value evaluation model, wherein the urban comprehensive traffic safety risk value evaluation model is used for taking real-time traffic safety risk values of a driver group, a vehicle group, a road network group and a vehicle-enterprise group as input to obtain an urban comprehensive traffic safety risk value, and is a fully-connected MLP neural network model.
Please refer to fig. 1, which is a flowchart of a real-time urban traffic safety risk assessment method based on risk factors in an embodiment of the present invention.
Step S1, identifying and combing risk factors to generate a risk factor list, step S2, calculating risk values of the risk factors, and obtaining the risk values of the risk factors by calculating key indexes in combination with historical accident data; s3 is individual risk source risk value evaluation, real-time risk factors of people, vehicles, roads and enterprises are used as characteristic input to respectively obtain real-time risk values of the people, vehicles, roads and enterprises, and S4 is group risk value grade evaluation and is used for evaluating traffic safety risk values of driver groups, vehicle groups, road networks and vehicle-enterprise groups in real time; s5 is the urban comprehensive traffic safety risk value evaluation, the people-vehicle-road-enterprise risk evaluation network is comprehensively analyzed, the urban traffic safety risk evaluation network is constructed, and the dynamic real-time risk value of the city is obtained.
Compared with the prior art, the invention has the advantages that:
(1) collecting and identifying risk factors: aiming at a person-vehicle-road-enterprise-loop in a vehicle transportation link, the invention collects and arranges an insurance factor list, and collects real-time information such as driver behavior, vehicle running state and the like in addition to the acquisition of basic information of a driver and basic information of a vehicle; according to different risk sources and static dynamic attributes, hierarchical management is carried out on the list, and comprehensive fine-grained dynamic analysis is facilitated.
(2) Dynamically evaluating the human-vehicle-road-enterprise traffic safety risk in real time: the method is based on an MLP neural network algorithm, a risk evaluation network of a driver, a vehicle, a road section, an enterprise and a city is respectively constructed, and real-time and dynamic traffic safety risk values of risk source individuals, groups and the city whole are obtained through off-line training and on-line prediction.
(3) And (3) real-time risk early warning and pushing: the method and the system make a traffic safety risk grade judgment table to obtain the traffic safety risk grade, carry out real-time early warning and pushing on the high-risk source and risk factors with higher risk grade values, and are convenient for carrying out timely decision and traffic safety management and control.
The risk assessment technical scheme is beneficial to reducing the safety accident rate, building the industry safety environment, has better application prospect, can be applied to other business systems in the same industry, can also be popularized to other industries, and has better practical significance.
In one embodiment, the classification of the risk factors of people, vehicles, roads and enterprises in step S1 includes multiple levels, wherein the first level classification is divided into static factors and dynamic factors, the static factors represent static information that does not change in a short period, and the dynamic factors represent real-time dynamically changing information.
In the specific implementation process, the main risk factor analysis is performed around 4 risk sources of people, vehicles, roads and enterprises in the embodiment.
The first-level classification of the risk factors of the person-vehicle-road-enterprise risk source can be divided into static factors and dynamic factors, wherein the static factors refer to static information which does not change in a short period, such as the age of a driver, the driving age, the vehicle age, the road type, the enterprise qualification and the like, and the dynamic factors refer to real-time dynamically-changing information, such as driving behaviors, vehicle speed, road flow, enterprise safety management and the like; in addition, in the actual transportation process, environmental risk factors all affect the safety risk assessment of drivers/vehicles/roads/enterprises, so when the safety risk level of people/vehicles/roads/enterprises is assessed, the risk factors of people/vehicles/roads/enterprises and the environmental risk factors need to be considered; the concrete expression is in the model: the input of the human/vehicle/road/enterprise risk evaluation network needs the risk value of the risk factor of the environment where the user is located at the moment. Environmental risk factors including time, weather, temperature, etc., may improve the accuracy of the prediction.
According to the first-level classification of the risk factors, the second-level classification and the specific risk factors are sorted, and a risk factor list is obtained by referring to relevant industry standards and specifications and is shown in table 1.
TABLE 1 list of risk factors
In one embodiment, the risk value of the risk factor is a risk value of each risk factor S as a leading factor, which causes a traffic accident, and S2 includes:
s2.1: calculating the grade of the possibility of the traffic accident caused by the risk factors;
s2.2: calculating the severity grade of the accident caused by risk factors;
s2.3: and combining the grade of the possibility of the occurrence of the traffic accident caused by the risk factors and the grade of the severity of the accident caused by the risk factors to obtain a risk value of the risk factors.
Specifically, the risk value of the risk factor defined by the present invention is a risk value of each risk factor S as a leading factor, which causes a traffic accident. Calculating the risk value of the risk factor, and calculating the grade of the possibility of the occurrence of the traffic accident caused by the risk factor
And risk factors leading to accident severity ratings
Combining the two according to a certain synthesis rule to obtain a risk value R of the risk factorS，
In one embodiment, S2.1 comprises:
s2.1.1: calculating the probability of the occurrence of the traffic accident caused by the risk factor S, wherein the frequency of the occurrence of the road traffic accident caused by the risk factor S as the main cause of the damage is fSProbability of PS，
Wherein n isSRepresenting the times of road traffic accidents caused by the risk factor S as a main cause in a period of time; n is the total number of all traffic accidents occurring within a period of time. According to Bernoulli's theorem in probability theory, when n is large, the probability that the frequency of accident has large deviation with probability is small, and f can be usedSProbability of occurrence of traffic accident P as risk factorS；
S2.1.2 calculates the level of likelihood that the risk factor will cause the traffic accident,
wherein < > represents rounding.
Specifically, the bernoulli theorem in probability theory yields:
when n is large, there is little possibility that the frequency of occurrence of the accident has a large deviation from the probability, and therefore f is setSThe probability of occurrence of a traffic accident as a risk factor is as follows:
and (3) the accident occurrence probability caused by the risk factors is the decimal of (0, 1), and the result is subjected to multiple rounding to obtain the accident occurrence probability grade caused by the risk factors.
In one embodiment, S2.2 comprises:
s2.2.1: calculating the severity of the consequences of the traffic accident caused by the risk factors according to the measures of the severity of the traffic accident, wherein the measures of the severity of the traffic accident comprise the number of dead people in the accident, the number of injured people in the accident and the economic loss of the accident:
wherein, VSTo the extent that risk factors lead to the severity of the consequences of an accident, DSFor accident-related deaths due to risk factors, SSFor the number of accident injuries caused by risk factors ESThe accident economic loss is caused by risk factors;
s2.2.2: for the severity degree V of traffic accident consequences caused by risk factorsSThe normalization treatment is carried out, and the normalization treatment is carried out,
wherein,
min (V) for results after normalizationS) Max (V) which is the minimum value of the severity of the accident consequence caused by the risk factorS) The maximum value of the severity of the accident consequence caused by risk factors;
s2.2.3: and performing multiple value taking on the result of the normalization processing to obtain the severity grade of the traffic accident consequence caused by risk factors:
wherein,<>means that the rounding is carried out to round,
the grade of the severity degree of the traffic accident consequence caused by the risk factors is an integer of (0, 10).
In one embodiment, S2.3 multiplies the risk factor-causing traffic accident occurrence probability level and the risk factor-causing accident severity level to synthesize a risk value of the risk factor:
RSthe risk value representing the risk factor is a number of (0, 100).
Risk factor the risk value is
And
obtained according to a certain synthesis rule because
The probability index is represented by a number of words,
indicating a severity index, the inventive synthesis rule can be multiplied accordingly.
Through the method, the traffic safety risk values of all risk factors in the risk list can be calculated, and the risk value library of the risk factors can be obtained through sorting.
In one embodiment, in step S3, the risk value assessment model of individual risk source comprises an input layer, a hidden layer and an output layer,
wherein, the input layer is represented by a vector X, the output of the hidden layer is S (W)1X+b1) Wherein W is1Is the weight of the input layer to the hidden layer, b1Is the bias from the input layer to the hidden layer, the function S represents the sigmoid function of the activation function, which is expressed as:
sigmoid(a)＝1/(1+e-a)
where a represents the input to the sigmod function;
the output of the hidden layer is represented by X1, and the hidden layer to the output layer represents a multi-class logistic regression, i.e. softmax regression, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer;
the output of the output layer is represented as:
Y＝G(b2+W2(S(b1+W1X)))
x denotes the input of the neural network, Y denotes the output of the neural network, G denotes the softmax function, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer;
when the individual risk value evaluation is carried out through the individual risk source risk value evaluation model, the online prediction stage is adopted, and the input data set is the identification information of the individual risk source and the risk value of risk factors occurring in real time of the individual risk source and is expressed as a vector RSAnd outputting a real-time traffic safety risk value R which is an individual risk source and is expressed as:
R＝G(b2 *+W2 *(S(b1 *+W1 *RS)))
wherein R is a real-time traffic safety risk value of a human individual risk source, a real-time traffic safety risk value of a vehicle individual risk source, a real-time traffic safety risk value of a road individual risk source, and a real-time traffic safety risk value of an enterprise individual risk source.
Specifically, the individual risk source risk values defined by the invention refer to traffic safety risk values of individuals respectively evaluating people, vehicles, roads and enterprises. The method adopts an MLP multilayer perception neural network algorithm to carry out dynamic security risk value evaluation on individual risk sources, and obtains corresponding weights of risk factors of the risk sources, wherein an algorithm model is shown in figure 2, and the algorithm steps are as follows:
the method comprises the following steps: inputting a data set
The input data set includes individual risk source identification information and real-time risk factor risk values. The identification information of the driver is an identification number, the identification information of the vehicle is a license plate number, the identification information of the road section is a road section number, and the identification information of the operation enterprise is an enterprise number; and the real-time risk value of the risk factors is the risk factors of the risk source in the actual transportation process, and the risk values of the risk factors are obtained by associating the risk factors with a risk database.
Step two: training phase
1) The input layer is represented by vector X, the output of the hidden layer is S (W)1X+b1)，
2) The output of the hidden layer is represented by X1, and the hidden layer to the output layer can be regarded as a multi-class logistic regression, namely, softmax regression, so that the output of the output layer can be represented as softmax (W)2X1+b2)。
3) The output of the output layer can be expressed as:
Y＝G(b2+W2(S(b1+W1X)))
where X represents the input of the neural network, Y represents the output of the neural network, and G represents the softmax function. The output value of the training stage can take the artificial set risk value experience value as the output, and iterative training is carried out by using a gradient descent method (SGD), so that the error is small enough to obtain the parameter W of the trained neural network1 *，b1 *，W2 *，b2 *，W1 *Is the weight of the input layer to the hidden layer, b1 *Is the bias of the input layer to the hidden layer, W2 *Is a weight of the hidden layer to the output, b2 *Is the biasing of the hidden layer to the output layer.
The trained model is saved and is called by the back end of the application program, and the model building process is shown in fig. 3.
Step three: on-line prediction phase
The input data set in the online prediction stage is represented as a vector R for identifying information of the individual risk source and risk values of risk factors of the individual risk source occurring in real timeSAnd outputting the real-time traffic safety risk value which is the individual risk source.
Step four: traffic safety risk early warning
And further mapping the obtained risk value of the individual risk source of the person, the vehicle, the road and the enterprise into a risk level according to a traffic safety risk level judgment table. The system pushes the high-risk individual risk source early warning and risk factors causing the high-risk individual risk source early warning to be judged as high risk according to a certain rule, and subsequent decision and scheduling are guided conveniently. The traffic safety risk level determination table is shown in table 2.
Table 2 traffic safety risk level decision table
| Serial number | Range of risk rating values | Risk rating |
| 1 | [0,10] | Low risk |
| 2 | (10,50] | Middle risk |
| 3 | (50,80] | High risk |
| 4 | (80,100] | Extremely high risk |
It should be noted that the risk level values in the above table include individual risk level values, group risk level values and city comprehensive risk level values. The role of the table is to map risk level values to risk levels, with risk level values falling within a range of corresponding risk level values being determined as corresponding risk levels. The risk grade is mainly used for risk early warning pushing.
In one embodiment, in S4, the real-time traffic safety risk value of the group is obtained by weighted calculation according to the real-time traffic safety risk value R of the individual risk source:
Rg＝∑Ri*ri
wherein R isiRepresenting a certain safety risk value, r, in a group of people or vehicles or roads or enterprises at the same timeiRepresenting the number of people or vehicles or roads or a certain security risk value in a group of enterprises at the same time.
In particular, the group risk values defined herein, with RgThe expression refers to the evaluation of the traffic safety risk values of people, vehicles, roads and enterprise risk source groups respectively. The group safety risk value is obtained by weighting according to the real-time traffic safety risk value R of the individual risk source, and the quantity of a certain safety risk value in a person, a vehicle, a road or an enterprise group at the same time is in proportion to RiThe calculation is performed according to the following formula:
wherein, count (R)i) Representing a safety risk value R in a group of persons or vehicles or roads or enterprisesiCount (r) represents the number of all security risk values in a person or vehicle or road or enterprise group.
In one embodiment, the urban integrated traffic safety risk value assessment model in step S5 includes an input layer, a hidden layer, and an output layer, and is obtained by offline training.
The urban comprehensive traffic safety risk value is obtained by comprehensive analysis of people, vehicles, roads and enterprises, model construction is carried out by adopting an MLP algorithm, and an integral algorithm model diagram is shown in FIG. 4.
Wherein, the output values of the vehicle, person, road and enterprise individual risk evaluation network are vehicle, person, road and enterprise individual real-time security risk values respectively; and calculating to obtain real-time safety risk values of vehicle, person, road and enterprise groups according to a group risk value evaluation model method, using the real-time safety risk values as input values of the urban traffic safety risk evaluation network, repeating training and prediction steps similar to the individual risk source risk value evaluation model, and obtaining real-time urban comprehensive traffic safety risk values through offline training and online verification processes.
The method provided by the present invention is illustrated by specific experimental data.
And selecting 50 thousands of accident data, and verifying the algorithm model provided by the invention. For people, vehicles, roads and enterprises, the risk value of the risk factor of top10 is taken as input, the input belongs to a general simple data set, and in order to prevent overfitting, a hidden layer is set to be 1; in general, since the hidden layer node number is (input number + output number) × 2/3, the hidden layer node number is 8, and the iteration number is 100, and an error convergence curve of the training model is obtained, as shown in fig. 5. The Mean Absolute Percentage Error (MAPE) is used as an evaluation index to carry out online prediction, and verification shows that the mean absolute percentage error of the algorithm model can be stabilized to about 5.1 percent, and the actual application requirements can be met. Through the steps, the comprehensive real-time traffic safety risk value of individuals, groups and cities can be obtained, high-risk early warning and corresponding risk factors are pushed according to the safety risk level, and traffic safety control is purposefully performed according to the high-risk early warning and the corresponding risk factors, so that the follow-up traffic safety control is facilitated.
The method is combined with the MLP neural network model to carry out real-time traffic safety risk assessment, on one hand, the method can assess the risk values of individual risk sources of people, vehicles, roads and enterprises in a fine-grained manner, and is more favorable for related departments to carry out more accurate and more comprehensive traffic safety management and control; on the other hand, the neural network can have a plurality of nonlinear layers and parameters, so that the method for adjusting the parameters is more flexible and the errors are more easily converged under the conditions of large data volume and high calculation complexity.
Example two
Based on the same inventive concept, the embodiment provides a real-time urban traffic safety risk assessment device based on risk factors, which comprises:
the risk factor identification module is used for identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
the risk value calculation module of the risk factors is used for calculating key indexes of the risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, and the risk values of all the risk factors form a risk value library of the risk factors;
the individual risk source risk value evaluation module is used for constructing an individual risk source risk value evaluation model, the individual risk source risk value evaluation model comprises individual risk source risk value evaluation models of people, vehicles, roads and enterprises, the individual risk source risk value evaluation model is used for respectively taking real-time risk factors of the people, the vehicles, the roads and the enterprises as characteristic input to respectively obtain real-time risk values of the individual risk sources of the people, the vehicles, the roads and the enterprises, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
the group risk value grade evaluation module is used for evaluating real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
and the urban comprehensive traffic safety risk value evaluation module is used for constructing an urban comprehensive traffic safety risk value evaluation model and obtaining an urban comprehensive traffic safety risk value by taking the real-time traffic safety risk values of the driver group, the vehicle group, the road network group and the vehicle-enterprise group as input, and the urban comprehensive traffic safety risk value evaluation model is a fully-connected MLP neural network model.
Since the device introduced in the second embodiment of the present invention is a device used for implementing the method for real-time assessment of urban traffic safety risk based on risk causing factors in the first embodiment of the present invention, a person skilled in the art can understand the specific structure and deformation of the device based on the method introduced in the first embodiment of the present invention, and thus, details are not described herein again. All the devices adopted in the method of the first embodiment of the present invention belong to the protection scope of the present invention.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A real-time urban traffic safety risk assessment method based on risk factors is characterized by comprising the following steps:
s1: identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
s2: calculating key indexes of risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, wherein the risk values of all the risk factors form a risk value library of the risk factors;
s3: constructing an individual risk source risk value evaluation model, wherein the individual risk source risk value evaluation model comprises a human, vehicle, road and enterprise individual risk source risk value evaluation model and is used for respectively taking real-time risk factors of human, vehicle, road and enterprise individuals as characteristic input to respectively obtain real-time risk values of human, vehicle, road and enterprise individual risk sources, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
s4: real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group are evaluated in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
s5: and constructing an urban comprehensive traffic safety risk value evaluation model, wherein the urban comprehensive traffic safety risk value evaluation model is used for taking real-time traffic safety risk values of a driver group, a vehicle group, a road network group and a vehicle-enterprise group as input to obtain an urban comprehensive traffic safety risk value, and is a fully-connected MLP neural network model.
2. The method as claimed in claim 1, wherein the classification of the risk factors of people, vehicles, roads and enterprises in step S1 includes multiple stages, wherein the classification of the risk factors of people, vehicles, roads and enterprises includes a static factor and a dynamic factor, the static factor represents static information that does not change in a short period, and the dynamic factor represents real-time dynamically changing information.
3. The method for real-time assessment of urban traffic safety risk based on risk factors according to claim 1, wherein the risk value of risk factors is a risk value of occurrence of a traffic accident with each risk factor S as a leading factor, and S2 comprises:
s2.1: calculating the grade of the possibility of the traffic accident caused by the risk factors;
s2.2: calculating the severity grade of the accident caused by risk factors;
s2.3: and combining the grade of the possibility of the occurrence of the traffic accident caused by the risk factors and the grade of the severity of the accident caused by the risk factors to obtain a risk value of the risk factors.
4. The risk real-time assessment method for urban traffic safety based on risk-causing factors according to claim 3, characterized in that S2.1 comprises:
s2.1.1: calculating the probability of the occurrence of the traffic accident caused by the risk factor S, wherein the frequency of the occurrence of the road traffic accident caused by the risk factor S as the main cause of the damage is fSProbability of PS，
Wherein n isSRepresenting the times of road traffic accidents caused by the risk factor S as a main cause in a period of time; n is the total number of the traffic accidents in a period of time; according to Bernoulli's theorem in probability theory, when n is large, the probability that the frequency of accident has large deviation with probability is small, and f isSProbability of occurrence of traffic accident P as risk factorS；
S2.1.2 calculates the level of likelihood that the risk factor will cause the traffic accident,
wherein < > represents rounding.
5. The risk real-time assessment method for urban traffic safety based on risk-causing factors according to claim 4, wherein S2.2 comprises:
s2.2.1: calculating the severity of the consequences of the traffic accident caused by the risk factors according to the measures of the severity of the traffic accident, wherein the measures of the severity of the traffic accident comprise the number of dead people in the accident, the number of injured people in the accident and the economic loss of the accident:
wherein, VSTo the extent that risk factors lead to the severity of the consequences of an accident, DSFor accident-related deaths due to risk factors, SSFor the number of accident injuries caused by risk factors ESThe accident economic loss is caused by risk factors;
s2.2.2: for the severity degree V of traffic accident consequences caused by risk factorsSThe normalization treatment is carried out, and the normalization treatment is carried out,
wherein,
min (V) for results after normalizationS) Max (V) which is the minimum value of the severity of the accident consequence caused by the risk factorS) The maximum value of the severity of the accident consequence caused by risk factors;
s2.2.3: and performing multiple value taking on the result of the normalization processing to obtain the severity grade of the traffic accident consequence caused by risk factors:
wherein,<>means that the rounding is carried out to round,
the grade of the severity degree of the traffic accident consequence caused by the risk factors is an integer of (0, 10).
6. The real-time urban traffic safety risk assessment method based on risk factors as claimed in claim 5, wherein S2.3 synthesizes the risk level of the risk factors causing traffic accidents and the severity level of the accidents caused by the risk factors according to multiplication to obtain the risk value of the risk factors:
RSthe risk value representing the risk factor is a number of (0, 100).
7. The risk real-time assessment method for urban traffic safety based on risk-causing factors according to claim 1, wherein in step S3, the risk value assessment model of individual risk source comprises an input layer, a hidden layer and an output layer,
wherein, the input layer is represented by a vector X, the output of the hidden layer is S (W)1X+b1) Wherein W is1Is the weight of the input layer to the hidden layer, b1Is the bias from the input layer to the hidden layer, the function S represents the sigmoid function of the activation function, which is expressed as:
sigmoid(a)＝1/(1+e-a)
where a represents the input to the sigmod function;
the output of the hidden layer is represented by X1, and the hidden layer to the output layer represents a multi-class logistic regression, i.e. softmax regression, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer;
the output of the output layer is represented as:
Y＝G(b2+W2(S(b1+W1X)))
x denotes the input of the neural network, Y denotes the output of the neural network, G denotes the softmax function, W2Is a weight of the hidden layer to the output, b2Is the bias of the hidden layer to the output layer;
when the individual risk value evaluation is carried out through the individual risk source risk value evaluation model, the online prediction stage is adopted, and the input data set is the identification information of the individual risk source and the risk value of risk factors occurring in real time of the individual risk source and is expressed as a vector RSAnd outputting a real-time traffic safety risk value R which is an individual risk source and is expressed as:
R＝G(b2 *+W2 *(S(b1 *+W1 *RS)))
wherein R is a real-time traffic safety risk value of a driver individual risk source, a real-time traffic safety risk value of a vehicle individual risk source, a real-time traffic safety risk value of a road individual risk source, and a real-time traffic safety risk value of an enterprise individual risk source.
8. The risk real-time assessment method for urban traffic safety based on risk-causing factors according to claim 1, wherein in S4, the real-time traffic safety risk value of the group is obtained by weighted calculation according to the real-time traffic safety risk value R of the individual risk source:
Rg＝∑Ri*ri
wherein R isiRepresenting a certain safety risk value, r, in a group of people or vehicles or roads or enterprises at the same timeiRepresenting the number of people or vehicles or roads or a certain security risk value in a group of enterprises at the same time.
9. The risk real-time assessment method for urban traffic safety based on risk-causing factors according to claim 1, wherein the urban integrated traffic safety risk value assessment model in step S5 comprises an input layer, a hidden layer and an output layer, and is obtained by offline training.
10. A real-time urban traffic safety risk assessment device based on risk factors is characterized by comprising the following components:
the risk factor identification module is used for identifying risk factors by combining human, vehicle, road and enterprise risk sources to generate a risk factor list, wherein the risk factor list comprises all risk factors obtained through identification;
the risk value calculation module of the risk factors is used for calculating key indexes of the risk factors in the risk factor list by analyzing historical accident data to obtain risk values of the risk factors, and the risk values of all the risk factors form a risk value library of the risk factors;
the individual risk source risk value evaluation module is used for constructing an individual risk source risk value evaluation model, the individual risk source risk value evaluation model comprises individual risk source risk value evaluation models of people, vehicles, roads and enterprises, the individual risk source risk value evaluation model is used for respectively taking real-time risk factors of the people, the vehicles, the roads and the enterprises as characteristic input to respectively obtain real-time risk values of the individual risk sources of the people, the vehicles, the roads and the enterprises, and each individual risk source risk value evaluation model is a fully-connected MLP neural network model;
the group risk value grade evaluation module is used for evaluating real-time traffic safety risk values of a driver group, a vehicle group, a road network and a vehicle-enterprise group in real time according to the obtained real-time risk values of individual risk sources of people, vehicles, roads and enterprises;
and the urban comprehensive traffic safety risk value evaluation module is used for constructing an urban comprehensive traffic safety risk value evaluation model and obtaining an urban comprehensive traffic safety risk value by taking the real-time traffic safety risk values of the driver group, the vehicle group, the road network group and the vehicle-enterprise group as input, and the urban comprehensive traffic safety risk value evaluation model is a fully-connected MLP neural network model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111451889.1A CN114331019A (en) | 2021-12-01 | 2021-12-01 | Urban traffic safety risk real-time assessment method and device based on risk factor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111451889.1A CN114331019A (en) | 2021-12-01 | 2021-12-01 | Urban traffic safety risk real-time assessment method and device based on risk factor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114331019A true CN114331019A (en) | 2022-04-12 |
Family
ID=81048075
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111451889.1A Pending CN114331019A (en) | 2021-12-01 | 2021-12-01 | Urban traffic safety risk real-time assessment method and device based on risk factor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114331019A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522097A (en) * | 2023-11-03 | 2024-02-06 | 中铁一局集团市政环保工程有限公司 | Construction method and system based on three-dimensional GIS and BIM integration |
| CN117952413A (en) * | 2023-11-16 | 2024-04-30 | 北京智城联合科技发展有限公司 | Urban security risk assessment method and system |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103177302A (en) * | 2013-03-28 | 2013-06-26 | 深圳市环境科学研究院 | Risk source assessment method for urban area reservoir water source |
| CN106228499A (en) * | 2016-07-06 | 2016-12-14 | 东南大学 | A kind of cargo security evaluation model based on people's bus or train route goods multi-risk System source |
| CN106897826A (en) * | 2017-02-23 | 2017-06-27 | 吉林大学 | A kind of street accidents risks appraisal procedure and system |
| CN109034549A (en) * | 2018-06-29 | 2018-12-18 | 广东电网有限责任公司 | Power operation risk evaluation modeling method |
| CN109409683A (en) * | 2018-09-28 | 2019-03-01 | 北京图安世纪科技股份有限公司 | Quickly dynamic urban safety risk class appraisal procedure |
| WO2019213980A1 (en) * | 2018-05-08 | 2019-11-14 | 清华大学 | Intelligent vehicle safety decision-making method employing driving safety field |
| CN110826905A (en) * | 2019-11-06 | 2020-02-21 | 苏交科集团股份有限公司 | Long and large bridge operation safety risk assessment method |
| CN111144772A (en) * | 2019-12-30 | 2020-05-12 | 交通运输部公路科学研究所 | Road transportation safety risk real-time assessment method based on data mining |
| CN112232254A (en) * | 2020-10-26 | 2021-01-15 | 清华大学 | Pedestrian risk assessment method considering pedestrian acceleration rate |
| CN112801541A (en) * | 2021-02-25 | 2021-05-14 | 交通运输部天津水运工程科学研究所 | Dangerous chemical road transportation risk dynamic assessment and risk navigation method |
| CN112991743A (en) * | 2021-04-22 | 2021-06-18 | 泰瑞数创科技(北京)有限公司 | Real-time traffic risk AI prediction method based on driving path and system thereof |
| CN113160593A (en) * | 2021-01-18 | 2021-07-23 | 重庆交通大学 | Mountain road driving safety early warning method based on edge cloud cooperation |
| CN113361665A (en) * | 2021-08-11 | 2021-09-07 | 成都理工大学 | Highland mountain tourism safety risk early warning method based on reinforcement learning |
| CN113487863A (en) * | 2021-06-30 | 2021-10-08 | 中国科学技术大学先进技术研究院 | Multi-factor traffic risk real-time quantitative evaluation system and method |
-
2021
- 2021-12-01 CN CN202111451889.1A patent/CN114331019A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103177302A (en) * | 2013-03-28 | 2013-06-26 | 深圳市环境科学研究院 | Risk source assessment method for urban area reservoir water source |
| CN106228499A (en) * | 2016-07-06 | 2016-12-14 | 东南大学 | A kind of cargo security evaluation model based on people's bus or train route goods multi-risk System source |
| CN106897826A (en) * | 2017-02-23 | 2017-06-27 | 吉林大学 | A kind of street accidents risks appraisal procedure and system |
| WO2019213980A1 (en) * | 2018-05-08 | 2019-11-14 | 清华大学 | Intelligent vehicle safety decision-making method employing driving safety field |
| CN109034549A (en) * | 2018-06-29 | 2018-12-18 | 广东电网有限责任公司 | Power operation risk evaluation modeling method |
| CN109409683A (en) * | 2018-09-28 | 2019-03-01 | 北京图安世纪科技股份有限公司 | Quickly dynamic urban safety risk class appraisal procedure |
| CN110826905A (en) * | 2019-11-06 | 2020-02-21 | 苏交科集团股份有限公司 | Long and large bridge operation safety risk assessment method |
| CN111144772A (en) * | 2019-12-30 | 2020-05-12 | 交通运输部公路科学研究所 | Road transportation safety risk real-time assessment method based on data mining |
| CN112232254A (en) * | 2020-10-26 | 2021-01-15 | 清华大学 | Pedestrian risk assessment method considering pedestrian acceleration rate |
| CN113160593A (en) * | 2021-01-18 | 2021-07-23 | 重庆交通大学 | Mountain road driving safety early warning method based on edge cloud cooperation |
| CN112801541A (en) * | 2021-02-25 | 2021-05-14 | 交通运输部天津水运工程科学研究所 | Dangerous chemical road transportation risk dynamic assessment and risk navigation method |
| CN112991743A (en) * | 2021-04-22 | 2021-06-18 | 泰瑞数创科技(北京)有限公司 | Real-time traffic risk AI prediction method based on driving path and system thereof |
| CN113487863A (en) * | 2021-06-30 | 2021-10-08 | 中国科学技术大学先进技术研究院 | Multi-factor traffic risk real-time quantitative evaluation system and method |
| CN113361665A (en) * | 2021-08-11 | 2021-09-07 | 成都理工大学 | Highland mountain tourism safety risk early warning method based on reinforcement learning |
Non-Patent Citations (3)
| Title |
|---|
| 殷志浩主编: "《机器学习原理及应用》", 31 August 2021, 上海财经大学出版社, pages: 129 - 134 * |
| 胡立伟;凌浩晗;杨锦青;赵雪亭;尹宇;田海龙;: "基于模糊小波神经网络的营运客车运行风险评估模型研究", 安全与环境学报, no. 03, 25 June 2020 (2020-06-25), pages 862 - 870 * |
| 赵学刚;魏朗;: "道路交通安全风险预警模糊评判模型", 太原理工大学学报, no. 04, 15 July 2009 (2009-07-15), pages 365 - 368 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522097A (en) * | 2023-11-03 | 2024-02-06 | 中铁一局集团市政环保工程有限公司 | Construction method and system based on three-dimensional GIS and BIM integration |
| CN117952413A (en) * | 2023-11-16 | 2024-04-30 | 北京智城联合科技发展有限公司 | Urban security risk assessment method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Traffic Accident Prediction Based on LSTM‐GBRT Model | |
| Lu et al. | Short-term demand forecasting for online car-hailing using ConvLSTM networks | |
| CN109947898B (en) | Equipment fault testing method based on intellectualization | |
| Takale et al. | Road accident prediction model using data mining techniques | |
| Dimitriou et al. | Fuzzy modeling of freeway accident duration with rainfall and traffic flow interactions | |
| Csikós et al. | Traffic speed prediction method for urban networks—An ANN approach | |
| Ai et al. | Short-term road speed forecasting based on hybrid RBF neural network with the aid of fuzzy system-based techniques in urban traffic flow | |
| CN104376716A (en) | Method for dynamically generating bus timetables on basis of Bayesian network models | |
| CN114331019A (en) | Urban traffic safety risk real-time assessment method and device based on risk factor | |
| Sarkar et al. | An integrated approach using rough set theory, ANFIS, and Z-number in occupational risk prediction | |
| CN116663742B (en) | Regional capacity prediction method based on multi-factor and model fusion | |
| Li et al. | New model of travel-time prediction considering weather conditions: case study of urban expressway | |
| JABBAR et al. | Predictive intelligence: A neural network learning system for traffic condition prediction and monitoring on freeways | |
| Zheng et al. | Modeling stochastic service time for complex on-demand food delivery | |
| CN113344257B (en) | Prediction method for layer analysis response time in homeland space cloud platform | |
| Zhao et al. | A mining algorithm to improve LSTM for predicting customer churn in railway freight traffic | |
| Gürbüz et al. | Classification rule discovery for the aviation incidents resulted in fatality | |
| CN117275220B (en) | Mountain expressway real-time accident risk prediction method based on incomplete data | |
| Wang et al. | Using deep neural network to analyze travel mode choice with interpretable economic information: An empirical example | |
| Zhang et al. | Spatial-temporal dynamic semantic graph neural network | |
| Гавриленко et al. | Тhe task of analyzing publications to build a forecast for changes in cryptocurrency rates | |
| CN114492967A (en) | Urban rail station passenger flow prediction method and medium based on CEEMDAN and BLSTM combined model | |
| CN112232607B (en) | Subway passenger flow volume prediction method and device | |
| Borisov et al. | The System of Fuzzy Cognitive Analysis and Modeling of System Dynamics | |
| CN113723660A (en) | Specific behavior type prediction method and system based on DNN-LSTM fusion model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |