CN112561175A - Traffic accident influence factor prediction method, device, equipment and storage medium - Google Patents
Traffic accident influence factor prediction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN112561175A CN112561175A CN202011506865.7A CN202011506865A CN112561175A CN 112561175 A CN112561175 A CN 112561175A CN 202011506865 A CN202011506865 A CN 202011506865A CN 112561175 A CN112561175 A CN 112561175A
- Authority
- CN
- China
- Prior art keywords
- traffic accident
- influence
- prediction
- training model
- factor
- 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
- 206010039203 Road traffic accident Diseases 0.000 title claims abstract description 232
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 158
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 21
- 238000000605 extraction Methods 0.000 claims abstract description 21
- 238000012216 screening Methods 0.000 claims description 31
- 230000011218 segmentation Effects 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 6
- 238000000556 factor analysis Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 8
- 231100000817 safety factor Toxicity 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000013461 design Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 238000000546 chi-square test Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Probability & Statistics with Applications (AREA)
- Educational Administration (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to the field of big data and discloses a traffic accident influence factor prediction method, a device, equipment and a storage medium. The method comprises the following steps: extracting a plurality of first influence factors from a plurality of pieces of historical traffic accident data by adopting a preset text information extraction algorithm; respectively inputting the first influence factor into a preset first pre-training model and a preset second pre-training model for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result; calculating the likelihood ratio of the first prediction result and the second prediction result to adjust random parameters in the second pre-training model to obtain a prediction model; and acquiring second influence factors of the current traffic accident, and predicting the influence of each second influence factor on the current traffic accident by adopting a prediction model so as to determine the influence factors of the current traffic accident. The invention also relates to a blockchain technique, the traffic accident data being stored in a blockchain. The invention improves the heterogeneity of pedestrian traffic safety factor analysis.
Description
Technical Field
The invention relates to the field of big data, in particular to a traffic accident influence factor prediction method, a device, equipment and a storage medium.
Background
With the increase of highway mileage and the improvement of infrastructure, the accident rate of unit mileage and the death rate of hundred million kilometers tend to decrease, but the total number of accidents still remains high, the severity of the accidents increases year by year, the safety situation is still severe, and the safety of people's lives and properties is seriously influenced. Therefore, it is very meaningful to analyze the cause of the road traffic accident, analyze the occurrence mechanism of the traffic accident, and provide corresponding safety improvement measures to improve the traffic safety situation from the source.
The traditional pedestrian traffic safety factor influence analysis method is mainly realized by methods such as descriptive data comparison and trend analysis. At present, traffic management departments, scientific research institutions and the like have adopted various statistical methods to analyze the influence factors of traffic accidents and the influence degrees of the influence factors, wherein common statistical models include polynomial logarithms, ordered probabilities, Bayesian hierarchical models, random parameter models and the like. However, the standard-of-the-upper-class, fixed-parameter, ordered models are not sufficient to capture the heterogeneity not observed in accident model data, for example, for elderly collisions, it is not appropriate in the fixed-parameter model to assume that the estimated parameters remain the same in all observations, in order to take into account the health, road traffic characteristics, driver behavior, etc. of the elderly. If an important consideration is addressed and the parameters actually vary among the observations, limiting the parameters to be fixed may result in inconsistent parameter estimates, which in turn affects the inefficiency and bias of the model. Therefore, the problem of insufficient heterogeneity exists in the prior art for analyzing the pedestrian traffic safety factors through the ordered model with fixed parameters.
Disclosure of Invention
The invention mainly aims to solve the technical problem that the heterogeneity of pedestrian traffic safety factors is insufficient through the analysis of an ordered model with fixed parameters in the prior art.
The invention provides a traffic accident influence factor prediction method in a first aspect, which comprises the following steps:
acquiring a plurality of pieces of historical traffic accident data, and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
Optionally, in a first implementation manner of the first aspect of the present invention, the extracting, by using a preset text information extraction algorithm, a plurality of first impact factors in a traffic accident from each piece of traffic accident data includes:
performing word segmentation on the corresponding content of each traffic accident data to obtain a plurality of words, and extracting influence words in the traffic accidents from each word by adopting preset prior knowledge;
counting the occurrence frequency of each influence word, and screening a preset number of influence words from the influence words as a first influence factor according to the fact that the occurrence frequency is from high to low.
Optionally, in a second implementation manner of the first aspect of the present invention, after the screening, from the high occurrence frequency to the low occurrence frequency, a preset number of influence tokens from the influence tokens as a first influence factor, the method further includes:
calculating a first significance score of each first influence factor by adopting a preset significance test algorithm;
judging whether the significance score is lower than a preset score threshold value or not;
if so, retaining the first impact factor below the score threshold.
Optionally, in a third implementation manner of the first aspect of the present invention, after the acquiring the plurality of pieces of historical traffic accident data, the method further includes:
extracting identification information, attribute information and data structure information in the traffic accident data;
checking the traffic accident data by adopting a preset rule according to the identification information, the attribute information and the data structure information to determine invalid data in the traffic accident data;
and screening out invalid data in the traffic accident data.
Optionally, in a fourth implementation manner of the first aspect of the present invention, the calculating a likelihood ratio between the first prediction result and the second prediction result includes:
determining a fixed parameter of the first influence factor according to the first prediction result, and determining a random parameter of the first influence factor according to the second prediction result;
calculating a first logarithm likelihood value after the first pre-training model is trained according to the fixed parameters, and calculating a second logarithm likelihood value after the second pre-training model is trained according to the random parameters; and calculating the difference value between the second log likelihood value and the first log likelihood value to obtain the likelihood ratio of the first prediction result and the second prediction result.
Optionally, in a fifth implementation manner of the first aspect of the present invention, the traffic accident data is further stored in a block chain.
A second aspect of the present invention provides a traffic accident influence factor prediction apparatus, including:
the extraction module is used for acquiring a plurality of pieces of historical traffic accident data and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
the device comprises an acquisition module, a processing module and a control module, wherein the acquisition module is used for acquiring a first pre-training model and a second pre-training model, and the second pre-training model is added with preset random parameters relative to the first pre-training model;
the training module is used for inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
the adjusting module is used for calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting the random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
the prediction module is used for acquiring second influence factors in the current traffic accident and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
and the screening module is used for screening the second influence factor which accords with the preset condition as the influence factor of the current traffic accident according to the influence degree.
Optionally, in a first implementation manner of the second aspect of the present invention, the extracting module includes:
the word segmentation unit is used for carrying out word segmentation on the corresponding content of each traffic accident data to obtain a plurality of words, and extracting influence words in the traffic accidents from each word segmentation by adopting preset prior knowledge;
and the screening unit is used for counting the occurrence frequency of each influence word, and screening a preset number of influence words from the influence words as first influence factors according to the fact that the occurrence frequency is from high to low.
Optionally, in a second implementation manner of the second aspect of the present invention, after the screening unit, the method further includes:
the first calculating unit is used for calculating a first significance score of each first influence factor by adopting a preset significance test algorithm;
the judging unit is used for judging whether the significance score is lower than a preset score threshold value or not; if so, retaining the first impact factor below the score threshold.
Optionally, in a third implementation manner of the second aspect of the present invention, after the extracting module, the method further includes:
the inspection module is used for extracting the identification information, the attribute information and the data structure information in the traffic accident data; checking the traffic accident data by adopting a preset rule according to the identification information, the attribute information and the data structure information to determine invalid data in the traffic accident data;
and the screening module is used for screening invalid data in the traffic accident data.
Optionally, in a fourth implementation manner of the second aspect of the present invention, the adjusting module includes:
the determining unit is used for determining a fixed parameter of the first influence factor according to the first prediction result and determining a random parameter of the first influence factor according to the second prediction result;
the second calculation unit is used for calculating a first log likelihood value after the first pre-training model is trained according to the fixed parameters and calculating a second log likelihood value after the second pre-training model is trained according to the random parameters;
and the third calculating unit is used for calculating the difference value between the second logarithm likelihood value and the first logarithm likelihood value to obtain the likelihood ratio of the first prediction result and the second prediction result.
Optionally, in a fifth implementation manner of the second aspect of the present invention, the traffic accident data is further stored in a block chain.
A third aspect of the present invention provides a traffic accident influence factor prediction apparatus comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the traffic accident impact factor prediction device to perform the traffic accident impact factor prediction method described above.
A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which, when run on a computer, cause the computer to execute the above-mentioned traffic accident impact factor prediction method.
According to the technical scheme provided by the invention, a plurality of first influence factors in traffic accident data are extracted and respectively input into a first pre-training model and a second pre-training model for training so as to respectively predict the influence degree of the first influence factors on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result; then, calculating the likelihood ratio of the output results of the two models through the first prediction result and the second prediction result so as to compare the advantages and disadvantages of the two models; adjusting random parameters in the second pre-training model through the likelihood ratio to optimize the second pre-training model until the optimal second pre-training model is obtained, namely the optimal second pre-training model is used as a corresponding prediction model; and finally, inputting the second influence factors of the current traffic accident into the prediction model to predict the influence degree of the second influence factors on the traffic accident, and increasing random parameters to screen the random influence factors to improve the heterogeneity of the prediction model for analyzing pedestrian traffic safety factors.
Drawings
FIG. 1 is a schematic diagram of a first embodiment of the traffic accident impact factor prediction method of the present invention;
FIG. 2 is a diagram of a traffic accident impact factor prediction method according to a second embodiment of the present invention;
FIG. 3 is a schematic diagram of a traffic accident influence factor prediction method according to a third embodiment of the present invention;
FIG. 4 is a diagram illustrating a fourth embodiment of the method for predicting an impact factor of a traffic accident according to the present invention;
FIG. 5 is a schematic diagram of an embodiment of the traffic accident influence factor prediction apparatus according to the present invention;
FIG. 6 is a schematic diagram of another embodiment of the traffic accident influence factor prediction apparatus according to the present invention;
fig. 7 is a schematic diagram of an embodiment of the traffic accident influence factor prediction apparatus according to the present invention.
Detailed Description
The embodiment of the invention provides a traffic accident influence factor prediction method, a device, equipment and a storage medium, wherein a preset text information extraction algorithm is used for extracting a plurality of first influence factors from a plurality of pieces of historical traffic accident data; respectively inputting the first influence factor into a preset first pre-training model and a preset second pre-training model for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result; calculating the likelihood ratio of the first prediction result and the second prediction result to adjust random parameters in the second pre-training model to obtain a prediction model; and acquiring second influence factors of the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting a prediction model so as to determine the influence factors of the current traffic accident. The invention improves the heterogeneity of the prediction model for analyzing the pedestrian traffic safety factors.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
For convenience of understanding, a detailed flow of an embodiment of the present invention is described below, and referring to fig. 1, a first embodiment of a traffic accident impact factor prediction method according to an embodiment of the present invention includes:
101. acquiring a plurality of pieces of historical traffic accident data, and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
it is to be understood that the implementation subject of the present invention may be a traffic accident influence factor prediction device, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject. It is emphasized that the traffic accident data may also be stored in a node of a block chain in order to further ensure privacy and security of the traffic accident data.
In the embodiment, traffic accident data of urban roads can be collected from a traffic management department, classified statistics is carried out according to the traffic accident data of the urban roads, the accident occurrence place, the accident occurrence time, the accident type, the accident form, the accident reason, the accident consequence and the like, and after the identification is carried out by using a retrieval field, the data are stored in a database; here, the corresponding historical traffic accident data can be obtained according to the type of the traffic accident, for example, if the cause of the traffic accident at night is to be analyzed, the traffic accident data in the time period of 18:00-6:00 can be selected.
In this embodiment, the first influencing factor includes accident time, intersection geometric design, intersection control type, speed limit, vehicle type, vehicle registration type, driver gender, whether the driver is responsible or not, signal control intersection, one-way road section, speed limit exceeding 50km/h, and the like. If the traffic accident of the elderly pedestrian is analyzed, the age of the elderly pedestrian, the motion state of the elderly person and the like may be included.
102. Acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
in this embodiment, the first pre-training model includes fixed parameters, and the second pre-training model introduces random parameters based on the first pre-training model, where the first pre-training model is preferably a Probit model (a model that obeys normal distribution), and specifically, the following functions may be referred to for setting of the fixed parameters and the random parameters:
βi'＝βi+αi；
wherein, betai' random parameter, beta, being the ith first influence factoriA fixed parameter being the ith first influence factor, alphaiThe vector is randomly distributed, and the random distribution can be normal distribution, Poisson distribution and the like; in the first pre-training model, a first impact factor corresponds to a fixed parameter, and in the second pre-training model, a second impact factor corresponds to a random parameter.
For example, the fixed influence factors include accident time, intersection geometric design, intersection control types, speed limit, vehicle types, vehicle registration types, driver gender, whether the driver is responsible or not, age of the elderly pedestrians and the like, and the random influence factors include signal control intersections, one-way road sections, speed limit exceeding 50km/h, motion states of the elderly pedestrians and the like; setting fixed parameters for accident time, intersection geometric design, intersection control type, speed limit, vehicle type, vehicle registration type, driver gender, whether the driver is responsible or not, age of the elderly pedestrians, signal control intersection, one-way road section, speed limit exceeding 50km/h and motion state of the elderly pedestrians in a first pre-training model; in the second pre-training model, fixed parameters are set for accident time, intersection geometric design, intersection control types, speed limit, vehicle types, vehicle registration types, driver gender, whether the driver is responsible or not and the age of the elderly pedestrians, and random parameters are set for the signal control intersection, the one-way road section, the speed limit exceeding 50km/h and the motion state of the elderly pedestrians.
103. Inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
in this embodiment, the first pre-training model and the second pre-training model need to determine the key points of training, and if the Probit model is adopted, the key points include response frequency, total number of measured values, factors, and covariates; the two models are used for predicting the influence degree of the first influence factor on the traffic accidents, so that the first influence factor can be used as a covariate, and the number of the traffic accidents corresponding to the first influence factor is used as a response frequency; the total number of the occurred traffic accidents is taken as the total number of the actual measurement values, and only the influence factors of the single type of traffic accidents are considered, so the factor range is [1,1 ]; and finally, setting a model output result to be represented in a probability form. And training can be carried out after the key points of the first pre-training model and the second pre-training model are set.
In this example, the degree of influence of the first influence factors on the traffic accident is represented in the form of probability, that is, the probability value of the type of traffic accident caused by each first influence factor, wherein the first pre-trained model outputs the first prediction result, and the second pre-trained model outputs the second prediction result.
104. Calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
in the embodiment, the likelihood ratio of the first prediction result and the second prediction result is calculated, and then the relative merits of the first pre-training model and the second pre-training model are tested through the likelihood ratio; when the first pre-training model is better, it shows that corresponding random parameters do not need to be introduced into the second pre-training model aiming at partial random influence factors, and the random parameters need to be removed and then trained again; when the second pre-training model is better, the second pre-training model can be continuously updated in an iterative manner; and stopping until the model converges again, thus obtaining the corresponding prediction model.
105. Acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
106. and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
In this embodiment, the type of the current traffic accident needs to be the same as the type of the traffic accident corresponding to the traffic accident data training, for example, if the prediction model training is trained by relying on the traffic accident data related to "elderly pedestrians" and "night", the prediction model is only applicable to predicting the influence degree of each second influence factor in the traffic accidents with "elderly pedestrians" and "night". The second influence factor refers to influence factors which may cause a current traffic accident, and the influence factors include accident time, intersection geometric design, intersection control types, speed limit, vehicle types, vehicle registration types, driver gender, whether the driver is responsible or not, signal control intersections, one-way road sections, speed limit exceeding 50km/h and the like.
Specifically, a typical scene, such as a traffic accident occurring in an "old community" or "at night", can be selected, the influence degree of each second influence factor on the traffic accident of the "old pedestrian" is analyzed, and the illumination degree of a signal lamp intersection and road facilities and the like are found to belong to random influence factors, but the influence on the traffic accident of the old pedestrian is large. The discovery is helpful for the road traffic management department to strengthen the traffic safety management aiming at the old community, carry out special management and supervision aiming at signal lamp intersections and community lighting systems, and provide and implement traffic safety improvement measures.
In the embodiment of the invention, a plurality of first influence factors in traffic accident data are extracted and respectively input into a first pre-training model and a second pre-training model for training so as to respectively predict the influence degree of the first influence factors on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result; then, calculating the likelihood ratio of the output results of the two models through the first prediction result and the second prediction result so as to compare the advantages and disadvantages of the two models; adjusting random parameters in the second pre-training model through the likelihood ratio to optimize the second pre-training model until the optimal second pre-training model is obtained, namely the optimal second pre-training model is used as a corresponding prediction model; and finally, inputting the second influence factors of the current traffic accident into the prediction model to predict the influence degree of the second influence factors on the traffic accident, and increasing random parameters to screen the random influence factors to improve the heterogeneity of the prediction model for analyzing pedestrian traffic safety factors.
Referring to fig. 2, a second embodiment of the method for predicting a traffic accident influence factor according to the present invention includes:
201. acquiring a plurality of pieces of historical traffic accident data;
202. performing word segmentation on the corresponding content of each traffic accident data to obtain a plurality of words, and extracting influence words in the traffic accidents from each word by adopting preset prior knowledge;
in this embodiment, a word segmentation tool is used to segment the content corresponding to each traffic accident data, and the word segmentation tool includes: HMM (Hidden Markov Model), CRF (Conditional Random Field Model), ME ((Maximum Entropy Model), N-gram (N-gram Model), SVM (support vector machines), and the like.
In this embodiment, a special data processing model is established, the model integrates data processing experience of traffic accident influence factors as prior knowledge, and each participle is sequentially input into the data processing model, so that whether the participle is an influence participle in a traffic accident or not can be identified.
203. Counting the occurrence frequency of each influence word, and screening a preset number of influence words from the influence words as a first influence factor according to the fact that the occurrence frequency is from high to low;
in the embodiment, the occurrence frequency of each influence word is counted and a possible influence factor library is automatically output according to the statistical analysis result of the big data, wherein each influence word in the influence factor library is sorted according to the occurrence frequency; for the influence participles with low occurrence frequency, the influence on the traffic accident can be considered to be small, so that the influence is not considered, and only a preset number of influence participles are screened as the first influence factor.
204. Calculating a first significance score of each first influence factor by adopting a preset significance test algorithm;
205. judging whether the significance score is lower than a preset score threshold value or not, if so, keeping a first influence factor lower than the score threshold value;
in this embodiment, a special significance test algorithm is established, a Probability statistical method is used to calculate P-value (significance value) of each first influence factor, and the P-value is compared with a preset score threshold of the P-value, and only the first influence factor with the P-value lower than the score threshold enters the next training of the first pre-trained model and the second pre-trained model.
206. Acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
207. inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
208. calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
209. acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
210. and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
In the embodiment of the invention, the extraction process of the first influence factor in the traffic accident data is introduced in detail, influence participles in the traffic accident data are extracted through priori knowledge, the occurrence frequency of each influence participle is counted, the influence participles with higher occurrence frequency in preset number are preliminarily screened to serve as the first influence factor, the first influence factor with the first significance score lower than the preset score threshold is reserved through a preset significance test algorithm, and the comprehensiveness of the finally obtained first influence factor on the traffic accident is ensured.
Referring to fig. 3, a third embodiment of the method for predicting a traffic accident influence factor according to the present invention includes:
301. acquiring a plurality of pieces of historical traffic accident data;
302. extracting identification information, attribute information and data structure information in the traffic accident data;
in this embodiment, a standard data field, an attribute field, a data structure field, and the like may be defined in a database storing traffic accident data, where the standard data field is used as identification information of the traffic accident data, the attribute field is used as attribute information thereof, and the data structure field is used as data structure information thereof; then writing rules of data consistency check, data missing check, data duplication check and the like to form a database compiling language.
303. Checking the traffic accident data by adopting a preset rule according to the identification information, the attribute information and the data structure information to determine invalid data in the traffic accident data;
304. screening out invalid data in the traffic accident data;
in this embodiment, the traffic accident data is grouped and sorted according to the identification information, the attribute information, and the data structure information, and then the traffic accident data is checked according to various corresponding rules to check invalid data therein.
Specifically, for example, the traffic accident data is subjected to data duplication checking, grouping and sorting are performed on identification fields, such as accident time, license plates of vehicles involved in an accident, and the like, then a count (×) function is used for counting, wherein the records without duplication are 1, and all the feedback results greater than 1 are the duplicate records. For completely repeated traffic accident data, deletion can be performed through delete; for partially repeated data, it may be merged or complemented.
305. Extracting a plurality of first influence factors in the traffic accident from the traffic accident data by adopting a preset text information extraction algorithm;
306. acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
307. inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
308. calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
309. acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
310. and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
In the embodiment of the invention, the method specifically introduces how to preliminarily screen out invalid data in traffic accident data, and determines and screens out traffic accident data which repeatedly appears, part of data is missing and the data structure is wrong according to identification information, attribute information and data structure information contained in a traffic accident so as to reduce noise data in the traffic accident data and increase the efficiency of model training.
Referring to fig. 4, a fourth embodiment of the method for predicting a traffic accident influence factor according to the present invention includes:
401. acquiring a plurality of pieces of historical traffic accident data, and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
402. acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
403. inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
404. determining a fixed parameter of the first influence factor according to the first prediction result, and determining a random parameter of the first influence factor according to the second prediction result;
in the present embodiment, the parameter β is assumediFixed parameter, beta, representing a certain factor ii' is a random parameter of the factor after considering the influence of the random factor,
βi＝β+μi
wherein muiIs a randomly distributed vector.
405. Calculating a first logarithm likelihood value after the first pre-training model is trained according to the fixed parameters, and calculating a second logarithm likelihood value after the second pre-training model is trained according to the random parameters;
in this embodiment, according to different first pre-training models, the random distribution rules of each fixed parameter and each random parameter are also different, and corresponding formulas are required to calculate log likelihood values of the first pre-training model and the second pre-training model. Specifically, if the first pre-training function adopts the Probit model, each fixed parameter and each random parameter obey normal distribution, assuming that Xi ═ βi' or Xi ═ betaiFollowing a normal distribution N (μ, σ 2), if i is greater than or equal to 1 and less than or equal to N, then the first log-likelihood value or the second log-likelihood value is calculated as follows:
wherein, when x ═ β', InL (μ, σ)2(ii) a β') is a first log likelihood value, InL (μ, σ) when x ═ β2(ii) a β) is the second log likelihood value.
406. Calculating a difference value between the second log likelihood value and the first log likelihood value to obtain a likelihood ratio of the first prediction result and the second prediction result;
in this embodiment, the likelihood ratio calculation formula of the first prediction result and the second prediction result is as follows:
LR＝-2[InL(μ,σ2；β')-InL(μ,σ2；β)]；
and if the LR value is greater than the chi-square test value, the second pre-training model introducing the random parameters is not superior to the first pre-training model only having fixed parameters, namely the random parameters do not need to be introduced for a part of first influence factors.
407. According to the likelihood ratio, adjusting random parameters in the second pre-training model to obtain a corresponding prediction model;
408. acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
409. and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
In the embodiment of the invention, the pre-training models of fixed parameters and random parameters are introduced in detail, the likelihood ratio calculation process of the prediction result is realized by respectively calculating the first logarithm likelihood value of the fixed parameters in the first pre-training model and the second logarithm likelihood value of the random parameters in the second pre-training model, then the difference value of the first logarithm likelihood value and the second logarithm likelihood value is calculated to determine the likelihood ratio of the first prediction result and the second prediction result, the advantages and the disadvantages of the random parameters can be determined through the likelihood ratio, and the value of the random parameters is adjusted to improve the goodness of the final prediction model.
The above description of the method for predicting a traffic accident influence factor according to the embodiment of the present invention, and the following description of the device for predicting a traffic accident influence factor according to the embodiment of the present invention, refer to fig. 5, where an embodiment of the device for predicting a traffic accident influence factor according to the embodiment of the present invention includes:
the extraction module 501 is configured to obtain a plurality of pieces of historical traffic accident data, and extract a plurality of first influence factors in a traffic accident from each piece of traffic accident data by using a preset text information extraction algorithm;
an obtaining module 502, configured to obtain a first pre-training model and a second pre-training model, where the second pre-training model is added with a preset random parameter relative to the first pre-training model;
a training module 503, configured to input the first impact factor into the first pre-training model and the second pre-training model respectively for training, so as to predict an impact degree of the first impact factor on the traffic accident, and obtain a corresponding first prediction result and a corresponding second prediction result;
an adjusting module 504, configured to calculate a likelihood ratio between the first prediction result and the second prediction result, and adjust a random parameter in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
the prediction module 505 is configured to obtain second influence factors in the currently occurring traffic accident, and predict, by using the prediction model, the degree of influence of each second influence factor on the currently occurring traffic accident;
and the screening module 506 is configured to screen a second influence factor meeting a preset condition according to the influence degree, as an influence factor of the currently occurring traffic accident.
In the embodiment of the invention, a plurality of first influence factors in traffic accident data are extracted and respectively input into a first pre-training model and a second pre-training model for training so as to respectively predict the influence degree of the first influence factors on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result; then, calculating the likelihood ratio of the output results of the two models through the first prediction result and the second prediction result so as to compare the advantages and disadvantages of the two models; adjusting random parameters in the second pre-training model through the likelihood ratio to optimize the second pre-training model until the optimal second pre-training model is obtained, namely the optimal second pre-training model is used as a corresponding prediction model; and finally, inputting the second influence factors of the current traffic accident into the prediction model to predict the influence degree of the second influence factors on the traffic accident, and increasing random parameters to screen the random influence factors to improve the heterogeneity of the prediction model for analyzing pedestrian traffic safety factors.
Referring to fig. 6, another embodiment of the traffic accident influence factor predicting apparatus according to the embodiment of the present invention includes:
the extraction module 501 is configured to obtain a plurality of pieces of historical traffic accident data, and extract a plurality of first influence factors in a traffic accident from each piece of traffic accident data by using a preset text information extraction algorithm;
an obtaining module 502, configured to obtain a first pre-training model and a second pre-training model, where the second pre-training model is added with a preset random parameter relative to the first pre-training model;
a training module 503, configured to input the first impact factor into the first pre-training model and the second pre-training model respectively for training, so as to predict an impact degree of the first impact factor on the traffic accident, and obtain a corresponding first prediction result and a corresponding second prediction result;
an adjusting module 504, configured to calculate a likelihood ratio between the first prediction result and the second prediction result, and adjust a random parameter in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
the prediction module 505 is configured to obtain second influence factors in the currently occurring traffic accident, and predict, by using the prediction model, the degree of influence of each second influence factor on the currently occurring traffic accident;
and the screening module 506 is configured to screen a second influence factor meeting a preset condition according to the influence degree, as an influence factor of the currently occurring traffic accident.
Specifically, the extraction module includes:
the word segmentation unit 5011 is configured to perform word segmentation on the content corresponding to each traffic accident data to obtain a plurality of words, and extract influence words in the traffic accidents from the words by using preset priori knowledge;
the screening unit 5012 is configured to count occurrence frequencies of the influence segments, and screen a preset number of influence segments from the influence segments as a first influence factor according to the occurrence frequencies from high to low.
Specifically, after the screening unit, the method further includes:
the first calculating unit 5013 is configured to calculate a first significance score of each first influence factor by using a preset significance testing algorithm;
a determination unit 5014, configured to determine whether the saliency score is lower than a preset score threshold; if so, retaining the first impact factor below the score threshold.
Specifically, after the extracting module, the method further includes:
the inspection module 507 is used for extracting identification information, attribute information and data structure information in the traffic accident data; checking the traffic accident data by adopting a preset rule according to the identification information, the attribute information and the data structure information to determine invalid data in the traffic accident data;
a screening module 508 for screening invalid data in the traffic accident data.
Specifically, the adjusting module includes:
a determining unit 5041, configured to determine a fixed parameter of the first impact factor according to the first prediction result, and determine a random parameter of the first impact factor according to the second prediction result;
a second calculating unit 5042, configured to calculate a first log likelihood value after the first pre-training model is trained according to the fixed parameter, and calculate a second log likelihood value after the second pre-training model is trained according to the random parameter;
a third calculating unit 5043, configured to calculate a difference between the second log likelihood and the first log likelihood, and obtain a likelihood ratio between the first prediction result and the second prediction result.
Specifically, the traffic accident data is also stored in a block chain.
In the embodiment of the invention, the extraction process of the first influence factor in the traffic accident data is introduced in detail, and the first influence factor with high occurrence frequency and the first significance score lower than the preset score threshold is determined and reserved through the priori knowledge and the preset significance test algorithm, so that the comprehensiveness of the finally obtained first influence factor on the traffic accident is ensured; specifically, how to preliminarily screen invalid data in the traffic accident data is introduced so as to reduce noise data in the traffic accident data and increase the efficiency of model training. (ii) a The pre-training model of the fixed parameters and the random parameters is introduced in detail, the likelihood ratio calculation process of the prediction results can determine the advantages and disadvantages of the random parameters through the likelihood ratio of the first prediction result and the second prediction result, and the values of the random parameters are adjusted to improve the goodness of the final prediction model.
Fig. 5 and 6 describe the traffic accident influence factor prediction apparatus in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the traffic accident influence factor prediction apparatus in the embodiment of the present invention is described in detail from the perspective of hardware processing.
Fig. 7 is a schematic structural diagram of a traffic accident influence factor prediction apparatus 700 according to an embodiment of the present invention, where the traffic accident influence factor prediction apparatus 700 may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 710 (e.g., one or more processors) and a memory 720, one or more storage media 730 (e.g., one or more mass storage devices) storing an application 733 or data 732. Memory 720 and storage medium 730 may be, among other things, transient storage or persistent storage. The program stored in the storage medium 730 may include one or more modules (not shown), each of which may include a series of instructions operating on the traffic accident impact factor prediction device 700. Still further, the processor 710 may be configured to communicate with the storage medium 730 to execute a series of instruction operations in the storage medium 730 on the traffic accident impact factor prediction device 700.
The traffic accident impact factor prediction device 700 may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input-output interfaces 760, and/or one or more operating systems 731, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, and the like. Those skilled in the art will appreciate that the configuration of the traffic accident impact factor prediction device illustrated in fig. 7 does not constitute a limitation of the traffic accident impact factor prediction device and may include more or fewer components than those illustrated, or some components may be combined, or a different arrangement of components.
The invention also provides a traffic accident impact factor prediction device, which comprises a memory and a processor, wherein the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, cause the processor to execute the steps of the traffic accident impact factor prediction method in the above embodiments.
The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when run on a computer, cause the computer to perform the steps of the traffic accident impact factor prediction method.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; 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 traffic accident influence factor prediction method is characterized by comprising the following steps:
acquiring a plurality of pieces of historical traffic accident data, and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
acquiring a first pre-training model and a second pre-training model, wherein the second pre-training model is added with preset random parameters relative to the first pre-training model;
inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
acquiring second influence factors in the current traffic accident, and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
and screening a second influence factor which accords with preset conditions as the influence factor of the current traffic accident according to the influence degree.
2. The traffic accident influence factor prediction method according to claim 1, wherein the extracting a plurality of first influence factors in a traffic accident from the respective traffic accident data using a preset text information extraction algorithm comprises:
performing word segmentation on the corresponding content of each traffic accident data to obtain a plurality of words, and extracting influence words in the traffic accidents from each word by adopting preset prior knowledge;
counting the occurrence frequency of each influence word, and screening a preset number of influence words from the influence words as a first influence factor according to the fact that the occurrence frequency is from high to low.
3. The traffic accident influence factor prediction method according to claim 2, further comprising, after the screening a preset number of influence tokens from the influence tokens as a first influence factor from high to low according to the frequency of occurrence:
calculating a first significance score of each first influence factor by adopting a preset significance test algorithm;
judging whether the significance score is lower than a preset score threshold value or not;
if so, retaining the first impact factor below the score threshold.
4. The traffic accident impact factor prediction method of claim 1, wherein the calculating the likelihood ratio of the first prediction result and the second prediction result comprises:
determining a fixed parameter of the first influence factor according to the first prediction result, and determining a random parameter of the first influence factor according to the second prediction result;
calculating a first logarithm likelihood value after the first pre-training model is trained according to the fixed parameters, and calculating a second logarithm likelihood value after the second pre-training model is trained according to the random parameters;
and calculating the difference value between the second log likelihood value and the first log likelihood value to obtain the likelihood ratio of the first prediction result and the second prediction result.
5. The traffic accident impact factor prediction method of any one of claims 1-4, further comprising, after the obtaining the plurality of pieces of historical traffic accident data:
extracting identification information, attribute information and data structure information in the traffic accident data;
checking the traffic accident data by adopting a preset rule according to the identification information, the attribute information and the data structure information to determine invalid data in the traffic accident data;
and screening out invalid data in the traffic accident data.
6. The traffic accident impact factor prediction method of claim 1, wherein the traffic accident data is also stored in a block chain.
7. A traffic accident influence factor prediction apparatus, characterized by comprising:
the extraction module is used for acquiring a plurality of pieces of historical traffic accident data and extracting a plurality of first influence factors in the traffic accidents from the traffic accident data by adopting a preset text information extraction algorithm;
the device comprises an acquisition module, a processing module and a control module, wherein the acquisition module is used for acquiring a first pre-training model and a second pre-training model, and the second pre-training model is added with preset random parameters relative to the first pre-training model;
the training module is used for inputting the first influence factor into the first pre-training model and the second pre-training model respectively for training so as to predict the influence degree of the first influence factor on the traffic accident and obtain a corresponding first prediction result and a corresponding second prediction result;
the adjusting module is used for calculating the likelihood ratio of the first prediction result and the second prediction result, and adjusting the random parameters in the second pre-training model according to the likelihood ratio to obtain a corresponding prediction model;
the prediction module is used for acquiring second influence factors in the current traffic accident and predicting the influence degree of each second influence factor on the current traffic accident by adopting the prediction model;
and the screening module is used for screening the second influence factor which accords with the preset condition as the influence factor of the current traffic accident according to the influence degree.
8. The traffic accident impact factor prediction device of claim 7, wherein the extraction module comprises:
the word segmentation unit is used for carrying out word segmentation on the corresponding content of each traffic accident data to obtain a plurality of words, and extracting influence words in the traffic accidents from each word segmentation by adopting preset prior knowledge;
and the screening unit is used for counting the occurrence frequency of each influence word, and screening a preset number of influence words from the influence words as first influence factors according to the fact that the occurrence frequency is from high to low.
9. A traffic accident influence factor prediction apparatus, characterized by comprising: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the traffic accident impact factor prediction device to perform the traffic accident impact factor prediction method of any of claims 1-6.
10. A computer readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement a traffic accident impact factor prediction method according to any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011506865.7A CN112561175A (en) | 2020-12-18 | 2020-12-18 | Traffic accident influence factor prediction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011506865.7A CN112561175A (en) | 2020-12-18 | 2020-12-18 | Traffic accident influence factor prediction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112561175A true CN112561175A (en) | 2021-03-26 |
Family
ID=75030642
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011506865.7A Pending CN112561175A (en) | 2020-12-18 | 2020-12-18 | Traffic accident influence factor prediction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112561175A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113609760A (en) * | 2021-07-19 | 2021-11-05 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and device for estimating service life of integrated circuit, electronic equipment and storage medium |
| CN114333317A (en) * | 2021-12-31 | 2022-04-12 | 杭州海康威视数字技术股份有限公司 | Traffic event processing method and device, electronic equipment and storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106530720A (en) * | 2016-12-28 | 2017-03-22 | 吉林大学 | Highway road traffic safety dark-spot road recognition and early-warning method |
| CN106897826A (en) * | 2017-02-23 | 2017-06-27 | 吉林大学 | A kind of street accidents risks appraisal procedure and system |
| KR20180078029A (en) * | 2016-12-29 | 2018-07-09 | (주) 더아이엠씨 | Analysis Method for Risk of Traffic Accident and Storage Medium Having the Same |
| CN108665093A (en) * | 2018-04-19 | 2018-10-16 | 东南大学 | Traffic accidents Severity forecasting method based on deep learning |
| CN109598929A (en) * | 2018-11-26 | 2019-04-09 | 北京交通大学 | A kind of multi-class the number of traffic accidents prediction technique |
| CN109914334A (en) * | 2019-03-28 | 2019-06-21 | 交通运输部水运科学研究所 | A kind of classification mechanism based method analysis of fine generalization waterway engineering eco-environmental impact |
| CN110458244A (en) * | 2019-08-20 | 2019-11-15 | 合肥工业大学 | A kind of traffic accident Severity forecasting method applied to Regional Road Network |
| CN111126868A (en) * | 2019-12-30 | 2020-05-08 | 中南大学 | Road traffic accident occurrence risk determination method and system |
| CN111291916A (en) * | 2018-12-10 | 2020-06-16 | 北京嘀嘀无限科技发展有限公司 | Driving behavior safety prediction method and device, electronic equipment and storage medium |
| CN111489008A (en) * | 2019-08-24 | 2020-08-04 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | Traffic accident influence factor spatial effect analysis method and application thereof |
| CN111950603A (en) * | 2020-07-24 | 2020-11-17 | 中南大学 | Road section traffic accident rate prediction method and device and computer storage medium |
| CN112560714A (en) * | 2020-12-18 | 2021-03-26 | 深圳赛安特技术服务有限公司 | Drunk driving detection method and device based on artificial intelligence, server and storage medium |
-
2020
- 2020-12-18 CN CN202011506865.7A patent/CN112561175A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106530720A (en) * | 2016-12-28 | 2017-03-22 | 吉林大学 | Highway road traffic safety dark-spot road recognition and early-warning method |
| KR20180078029A (en) * | 2016-12-29 | 2018-07-09 | (주) 더아이엠씨 | Analysis Method for Risk of Traffic Accident and Storage Medium Having the Same |
| CN106897826A (en) * | 2017-02-23 | 2017-06-27 | 吉林大学 | A kind of street accidents risks appraisal procedure and system |
| CN108665093A (en) * | 2018-04-19 | 2018-10-16 | 东南大学 | Traffic accidents Severity forecasting method based on deep learning |
| CN109598929A (en) * | 2018-11-26 | 2019-04-09 | 北京交通大学 | A kind of multi-class the number of traffic accidents prediction technique |
| CN111291916A (en) * | 2018-12-10 | 2020-06-16 | 北京嘀嘀无限科技发展有限公司 | Driving behavior safety prediction method and device, electronic equipment and storage medium |
| CN109914334A (en) * | 2019-03-28 | 2019-06-21 | 交通运输部水运科学研究所 | A kind of classification mechanism based method analysis of fine generalization waterway engineering eco-environmental impact |
| CN110458244A (en) * | 2019-08-20 | 2019-11-15 | 合肥工业大学 | A kind of traffic accident Severity forecasting method applied to Regional Road Network |
| CN111489008A (en) * | 2019-08-24 | 2020-08-04 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | Traffic accident influence factor spatial effect analysis method and application thereof |
| CN111126868A (en) * | 2019-12-30 | 2020-05-08 | 中南大学 | Road traffic accident occurrence risk determination method and system |
| CN111950603A (en) * | 2020-07-24 | 2020-11-17 | 中南大学 | Road section traffic accident rate prediction method and device and computer storage medium |
| CN112560714A (en) * | 2020-12-18 | 2021-03-26 | 深圳赛安特技术服务有限公司 | Drunk driving detection method and device based on artificial intelligence, server and storage medium |
Non-Patent Citations (8)
| Title |
|---|
| CHU, HC ET AL: "RISK factors for the severity of injury incurred in crashes ivolving on-duty police cars", 《TRAFFIC IN JURY PREVENTION》, pages 495 - 501 * |
| KIM, J ET AL: "Proposal for an inumdation Hazard index of road Links for safer Routing Services in Car Navigation Systems", 《ETRL JOURNAL》, pages 287 - 294 * |
| 何璐;张霖;: "交通事故影响因素的逐步回归分析", 伊犁师范学院学报(自然科学版), no. 04, pages 20 - 24 * |
| 李洪奇;周倩倩;辛盈;刘艳芳;赵艳红;: "城市道路交通事故预警系统", 计算机工程与设计, no. 06, pages 287 - 294 * |
| 王成;谢波;: "土地利用视角下城市交通事故驱动机理的研究进展", 地理科学进展, no. 09, pages 175 - 184 * |
| 王涛;黎文皓;李文勇;: "电动自行车交通事故严重程度影响因素分析", 广西大学学报(自然科学版), no. 06, pages 104 - 112 * |
| 王立晓;左志;杨晔;刘江水;: "基于Scobit的交通事故受伤严重程度模型", 重庆交通大学学报(自然科学版), no. 04, pages 5 - 9 * |
| 王鹏;卢小钊;闫章存;张敏捷;: "基于Ordered Probit模型的追尾事故严重性影响因素分析", 公路交通科技, no. 04, pages 106 - 111 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113609760A (en) * | 2021-07-19 | 2021-11-05 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and device for estimating service life of integrated circuit, electronic equipment and storage medium |
| CN113609760B (en) * | 2021-07-19 | 2023-12-05 | 北京圣涛平试验工程技术研究院有限责任公司 | Integrated circuit life prediction method and device, electronic equipment and storage medium |
| CN114333317A (en) * | 2021-12-31 | 2022-04-12 | 杭州海康威视数字技术股份有限公司 | Traffic event processing method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109410588B (en) | Traffic accident evolution analysis method based on traffic big data | |
| CN109547423A (en) | A kind of WEB malicious requests depth detection system and method based on machine learning | |
| CN112561175A (en) | Traffic accident influence factor prediction method, device, equipment and storage medium | |
| KR101598343B1 (en) | System for automatically Identifying spatio-temporal congestion patterns and method thereof | |
| Meng et al. | Expressway crash prediction based on traffic big data | |
| CN114299742A (en) | Dynamic recognition and updating recommendation method for speed limit information of expressway | |
| CN116186759A (en) | Sensitive data identification and desensitization method for privacy calculation | |
| CN116824868B (en) | Method, device, equipment and medium for identifying illegal parking points and predicting congestion of vehicles | |
| Beshah et al. | Mining pattern from road accident data: Role of road user’s behaviour and implications for improving road safety | |
| Nour et al. | Road traffic accidents injury data analytics | |
| CN109447306B (en) | Subway accident delay time prediction method based on maximum likelihood regression tree | |
| CN115794798A (en) | Market supervision informationized standard management and dynamic maintenance system and method | |
| CN115565373A (en) | Real-time risk prediction method, device, equipment and medium for highway tunnel accident | |
| CN111860661B (en) | Data analysis method and device based on user behaviors, electronic equipment and medium | |
| CN111221704B (en) | Method and system for determining running state of office management application system | |
| CN110866624A (en) | Chemical accident prediction method and system | |
| Sridevi et al. | Road accident analysis using machine learning | |
| Gong | The road traffic safety risk projection based on improved random forest. | |
| CN107346335B (en) | Webpage theme block identification method based on combination characteristics | |
| Johnson et al. | Study on road accidents using data mining technology | |
| CN111724048A (en) | Characteristic extraction method for finished product library scheduling system performance data based on characteristic engineering | |
| CN110851450A (en) | Accompanying vehicle instant discovery method based on incremental calculation | |
| CN113064597B (en) | Redundant code identification method, device and equipment | |
| CN111582622B (en) | Danger analysis method of air traffic management system based on natural language processing method | |
| Shahsitha Siddique et al. | Analysing Road Accident Criticality using Data mining |
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 |