WO2022257201A1 - 基于人机混合增强智能的城市交通安全预警方法及*** - Google Patents
基于人机混合增强智能的城市交通安全预警方法及*** Download PDFInfo
- Publication number
- WO2022257201A1 WO2022257201A1 PCT/CN2021/103202 CN2021103202W WO2022257201A1 WO 2022257201 A1 WO2022257201 A1 WO 2022257201A1 CN 2021103202 W CN2021103202 W CN 2021103202W WO 2022257201 A1 WO2022257201 A1 WO 2022257201A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- human
- traffic
- knowledge base
- early warning
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 76
- 238000011156 evaluation Methods 0.000 claims abstract description 58
- 238000012545 processing Methods 0.000 claims abstract description 42
- 239000011159 matrix material Substances 0.000 claims abstract description 32
- 206010039203 Road traffic accident Diseases 0.000 claims abstract description 28
- 238000003062 neural network model Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000013480 data collection Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 24
- 238000004458 analytical method Methods 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 238000007726 management method Methods 0.000 claims description 12
- 230000008439 repair process Effects 0.000 claims description 12
- 238000011160 research Methods 0.000 claims description 11
- 230000006399 behavior Effects 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 9
- 231100000279 safety data Toxicity 0.000 claims description 9
- 238000010801 machine learning Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 8
- 238000013499 data model Methods 0.000 claims description 5
- 230000003190 augmentative effect Effects 0.000 claims 1
- 230000002787 reinforcement Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 11
- 241000282414 Homo sapiens Species 0.000 description 9
- 238000004590 computer program Methods 0.000 description 8
- 230000001149 cognitive effect Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000011217 control strategy Methods 0.000 description 3
- 238000013210 evaluation model Methods 0.000 description 3
- 238000013277 forecasting method Methods 0.000 description 3
- 230000008447 perception Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000003416 augmentation Effects 0.000 description 1
- 238000009411 base construction Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000013209 evaluation strategy Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 210000002364 input neuron Anatomy 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000006996 mental state Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
- G08G1/0141—Measuring and analyzing of parameters relative to traffic conditions for specific applications for traffic information dissemination
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
- Y02A30/60—Planning or developing urban green infrastructure
Definitions
- the present disclosure relates to the technical field of traffic safety early warning, in particular to an urban traffic safety early warning method and system based on man-machine hybrid enhanced intelligence.
- the current traffic safety early warning system is mainly aimed at actively taking measures to regulate when people or vehicles in the traffic environment are in a dangerous state. For example, when the driver's physical and mental state is not good, the road weather condition is bad, the vehicle condition is not good, the road surface condition is not good, etc., it will be changed to a safe situation through monitoring, or through monitoring, it can be predicted in advance to reduce losses and avoid The occurrence of traffic accidents.
- the processing and analysis of human-computer mixed data makes the final early warning system less accurate.
- the present disclosure provides a method and system for early warning of urban traffic safety based on human-computer hybrid enhanced intelligence, combining the collection, processing and analysis of human-computer hybrid data to improve the efficiency and accuracy of safety early warning .
- the first aspect of the present disclosure provides an early warning method for urban traffic safety based on human-computer hybrid enhanced intelligence.
- a method for early warning of urban traffic safety based on human-computer hybrid enhanced intelligence including the following processes:
- Preprocessing the acquired data including at least data patching and prediction;
- Classify the preprocessed data establish a traffic accident cause attribute table according to the cause of the accident, select the traffic accidents of multiple road sections in the urban road as the sample group, and obtain the importance of different feature attribute index sets to the accident, and use different characteristics
- the attribute table of is analogous to the total set, and the importance of the characteristic attributes of different feature sets is obtained;
- the accident cause evaluation is carried out.
- the evaluation value of road traffic safety is obtained by weighting each accident factor by using the human-machine hybrid enhanced weight matrix
- An early warning instruction is generated and/or issued based on a comparison of the evaluation value with a preset threshold.
- the preset neural network is constructed according to the entropy weight method and the expert judgment method, the entropy weight matrix is obtained according to the entropy weight method, the fuzzy evaluation expert weight matrix is constructed according to the expert judgment method, and the entropy weight weight matrix and the fuzzy evaluation expert weight matrix
- the mean of the sum is the human-machine hybrid augmentation weight matrix.
- manually collected data includes social factor data, driver behavior status data, and traffic management department safety control experience data
- machine collected data includes at least road data, traffic accident data, natural environment data, and vehicle data.
- preprocessing the acquired data includes the following process:
- Crawl data of different types and lengths of time When the capture fails or when there is a problem with data processing, it sends a signal to request expert assistance, automatically records the data model and the expert's solution strategy, and adds a new data processing plan to the original There is a knowledge base, and the basic knowledge base is constantly updated;
- the second aspect of the present disclosure provides an urban traffic safety early warning system based on human-computer hybrid enhanced intelligence.
- An urban traffic safety early warning system based on human-computer hybrid enhanced intelligence including:
- the human-computer hybrid enhanced data collection sub-module is configured to: obtain data related to urban traffic early warning, at least including manual data collection and machine data collection;
- the human-computer hybrid enhanced data processing sub-module is configured to: preprocess the acquired data, at least including data patching and prediction;
- the human-computer hybrid enhanced accident cause analysis sub-module is configured to: classify and process the preprocessed data, establish a traffic accident cause attribute table according to the cause of the accident, and select traffic accidents on multiple road sections in urban roads as the sample group, and get The importance of different feature attribute index sets to the accident is obtained by analogy with the attribute tables of different features, and the importance of feature attributes of different feature sets is obtained; according to the importance of the acquired feature attributes, the accident cause evaluation is performed.
- human-machine hybrid enhanced safety data research and judgment and early warning sub-module which is configured to: obtain the human-machine hybrid enhanced weight matrix according to the classified data and the preset neural network model, and use the human-machine hybrid enhanced weight matrix to Each accident factor is weighted to obtain the evaluation value of road traffic safety, and an early warning instruction is generated and/or issued according to the comparison between the evaluation value and the preset threshold.
- human-computer hybrid enhanced data processing sub-module is also configured to: use human-computer hybrid enhanced learning to construct knowledge base, update knowledge base and enhance intelligence, including the following processes:
- Crawl data of different types and lengths of time When the capture fails or when there is a problem with data processing, it sends a signal to request expert assistance, automatically records the data model and the expert's solution strategy, and adds a new data processing plan to the original There is a knowledge base, and the basic knowledge base is constantly updated;
- the third aspect of the present disclosure provides a computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, the urban traffic safety early warning method based on human-computer hybrid enhanced intelligence as described in the first aspect of the present disclosure is realized. in the steps.
- the fourth aspect of the present disclosure provides an electronic device, including a memory, a processor, and a program stored in the memory and operable on the processor, and the processor implements the program described in the first aspect of the present disclosure when executing the program.
- the method, system, medium or electronic device described in this disclosure performs automatic and repetitive traffic data collection behaviors through roadside traffic equipment, and traffic participants and traffic managers input their artificial perception and management experience into data in digital form
- the collection sub-module extracts the required traffic data through simple manual decision-making operations; human-machine hybrid is used to enhance intelligence in the whole process, and machines are used to process repeated regular operations; labor consumption is reduced, data collection efficiency is improved and data collection is guaranteed. accuracy and completeness.
- the method, system, medium or electronic device described in this disclosure distinguishes and repairs the missing, wrong and abnormal data in the collected traffic data, and utilizes the historical data of different road sections to construct spatio-temporal fusion human-machine hybrid enhanced data prediction
- the model performs data prediction; combining the super computing power of the machine with the perceptual reasoning ability of human beings, realizes human-computer interaction and solves data processing problems quickly and efficiently.
- the methods, systems, media or electronic devices described in this disclosure perform statistical analysis on accidents on urban roads, use big data technology and rough set theory to conduct data mining on the causes of accidents, and calculate the importance of accident causes to determine the main factors. cause of the accident.
- the method, system, medium or electronic device described in this disclosure uses information entropy and expert analysis for fuzzy comprehensive evaluation to construct a human-computer hybrid enhanced urban safety early warning evaluation model to provide safety assessment for urban roads; enhance perception through human-computer hybrid Algorithms are used to provide traffic managers with road safety data and analysis results, assist traffic managers in making decisions, and improve the accuracy of analysis.
- FIG. 1 is a schematic structural diagram of an urban traffic safety early warning system based on human-computer hybrid enhanced intelligence provided by Embodiment 1 of the present disclosure.
- FIG. 2 is a schematic workflow diagram of the man-machine hybrid enhanced signal data collection sub-module provided in Embodiment 1 of the present disclosure.
- FIG. 3 is a schematic workflow diagram of the human-computer hybrid enhanced data processing sub-module provided in Embodiment 1 of the present disclosure.
- Fig. 4 is a schematic workflow diagram of the human-machine hybrid enhanced accident cause analysis sub-module provided in Embodiment 1 of the present disclosure.
- FIG. 5 is a schematic workflow diagram of the man-machine hybrid enhanced safety data research and judgment and early warning sub-module provided by Embodiment 1 of the present disclosure.
- Embodiment 1 of the present disclosure provides an urban traffic safety early warning system based on human-machine hybrid enhanced intelligence, which performs repeated data collection work through roadside road monitoring facilities, and uses human-computer hybrid enhanced intelligence Integrate the traffic control experience of traffic managers into the data collection process; after determining the type and type of data needed, screen the data, repair missing, wrong, and abnormal data, and use the professional knowledge of transportation industry experts to build a data prediction model to predict traffic traffic flow and other basic traffic parameters; use rough set theory to build an accident cause analysis model to determine the main cause of the accident; use the entropy method and neural network as the core of the algorithm to build a human-machine hybrid enhanced urban safety early warning research and judgment model, and analyze the urban road network.
- the urban traffic safety early warning platform based on human-computer hybrid enhanced intelligence combines the experience wisdom of traffic participants with machine learning and comprehensively considers many factors affecting traffic safety, and quickly and efficiently finds out the existing problems in the road traffic system through data mining and analysis of the causes of accidents. problems, and take targeted improvement measures to achieve the purpose of reducing traffic accidents and improving traffic safety.
- human-machine hybrid enhanced data collection sub-module human-machine hybrid enhanced data processing sub-module
- human-machine hybrid enhanced accident cause analysis sub-module human-machine hybrid enhanced safety data research and judgment and early warning sub-module.
- the data acquisition sub-module is the basic component system of the urban traffic safety early warning platform based on human-computer hybrid enhanced intelligence.
- the research and judgment of safety early warning data in urban traffic needs accurate road traffic data as the basis. Therefore, collecting reasonable and accurate traffic data is the prerequisite for the efficient operation of the entire early warning platform.
- the data required for early warning of urban road traffic safety can be divided into two categories, one is manual collection data of traffic participants, and the other is machine collection data collected by road detectors.
- Manually collected data can be divided into social factor data, driver behavior status data, and traffic management department safety control experience data
- machine collected data can be divided into road data, traffic accident data, natural environment data, and vehicle data.
- Social factor data refers to social event data such as holidays, epidemics, parades and rallies.
- Driver behavior state data refers to the state data of the driver such as the driver's psychological state, driving age, dangerous driving behavior records, and physical state.
- the safety experience data of the traffic management department can be divided into traffic control experience such as traffic safety regulations, illegal handling experience of the traffic management department, and signal control strategy of the signal control system.
- Road facility data can be divided into road comprehensive data such as road infrastructure conditions, road alignment, road topography, and road surface conditions.
- Traffic accident data can be divided into accident liability, accident rate per 10,000 vehicles, property damage, casualties and other accident impact data.
- Traffic environment data can be divided into weather data (rain, snow, fog, etc.), visual environment (road lighting, road visibility conditions, visibility, etc.), traffic flow environment (traffic flow, queue length, car-following distance, etc.).
- Vehicle data can be divided into vehicle data such as model, body color, and vehicle speed.
- the traffic sensing equipment and traffic participants on the road network collect the data of the vehicle and the surrounding environment at the same time.
- the road network of the entire city is divided into several different sub-regions for distributed data collection.
- the interconnection of data in different sub-regions constitutes a regional transmission network.
- the data collected by different collection methods is divided into four different characteristic data and uploaded to the data collection cloud of the traffic management department to form a traffic safety database.
- the human-machine hybrid enhanced intelligent data acquisition sub-module combines artificial intelligence and machine intelligence to collect the data required for road traffic safety research and judgment from the two, which improves the accuracy of the data and also strengthens the robustness of the safety early warning platform.
- the system classifies and collects the influencing factors of road traffic safety accidents from the four characteristic directions of people, vehicles, roads and the environment.
- Human characteristic data include the driver's safety awareness, psychological state, physical state (such as fatigue and sobriety), driver's driving age and illegal record, signal control strategy of the traffic control system, etc.
- the characteristic data of the vehicle include the model of the vehicle, the driving speed, the position of the vehicle, the degree and type of vehicle accident loss, etc.
- the characteristic data of roads include road infrastructure data, accident rate per 10,000 vehicles, technical grade of roads, incidence of road safety accidents, topographical conditions of roads, locations of areas where accidents frequently occur, etc.
- the characteristic data of the environment include natural weather data, road visual environment data, road traffic environment data, road traffic flow environment data, land use around the road, etc.
- the safety accident data collection sub-module enhanced by man-machine hybrid adopts the distributed data collection method to enhance the robustness of the system.
- the human-machine hybrid enhanced intelligence mutually verifies the same kind of data collected by different collection methods to obtain more real and accurate data.
- the human-computer hybrid enhanced data processing sub-module is composed of a data judgment repair module and a human-computer hybrid enhanced learning module.
- the data judgment and repair module is to judge the error and blank data and repair and predict the data.
- the data within a certain period of time are usually statistically stored. If some data is not entered within a certain period of time, it will be judged as blank data. If some data fluctuates more than the general category, it is judged as wrong data. To distinguish these blank data and erroneous data, the judgment rules are as follows:
- the data collected by the data acquisition sub-module is X
- the maximum value of a certain type of traffic parameter is X max
- the minimum value is X min
- the second step is to use the data before and after the road section or compare and repair the data of the adjacent road section at the same time, and fully consider various unexpected situations to perform data repair processing on the data to ensure accuracy and regularity.
- the time when the data has blank data or wrong data is the time period t.
- the data is X k (t)
- the historical data of the previous day is X k-1 (t)
- so on combined with the historical data of the previous month.
- the human-computer hybrid enhanced learning module has the functions of knowledge base construction, knowledge base update and intelligent enhancement. Merge existing databases into one multisource database.
- the system uses the experience of experts and senior practitioners in the transportation industry to collide and reorganize different databases to build a basic knowledge base to process data that meets the preset conditions, and to process data of different types and time lengths according to the needs of system users. crawl.
- the machine captures the required data through intelligent algorithms. When the capture fails or when there is a problem with the data processing, the machine will send a signal to request expert assistance.
- the human-computer hybrid enhanced learning module will automatically record the data model (that is, cognitive input conditions), and expert solution strategies (including data processing schemes and precautions, etc.) as cognitive strategies, and add new data processing schemes
- the original knowledge base is constantly updated with the basic knowledge base.
- machine learning algorithms are used to convert various heterogeneous data into same-dimensional data and input it into the knowledge base.
- the professional field knowledge contained in the knowledge base and various reasoning algorithms and solutions provide empirical ideas for system users to make decisions.
- the confidence of the obtained data is judged, and if the confidence of the data is high, the data is provided to the data user. If the data confidence is low, re-enter the knowledge base training until the data with high confidence is obtained. Randomly add data processing records to the knowledge base and update the knowledge base.
- the core of this function is human-machine fusion, which uses machines to convert human cognitive strategies into machine cognitive input, builds a knowledge base and continuously updates it, so as to continuously improve decision-making levels with the help of machine learning.
- the data processing sub-module based on human-computer hybrid enhancement combines the manager's management experience with the expert's professional knowledge through artificial intelligence and machine learning algorithms to produce good performance in data screening, cleaning, prediction, etc., providing a security warning platform Provides a solid data foundation.
- the accident cause analysis sub-module uses rough set theory to analyze the accident causes in traffic accidents and excavate the main causes of accidents.
- the attribute table B of the cause of the traffic accident is established, and the influence of different attributes on the occurrence of the traffic accident and the severity of the accident is studied.
- the number of fatalities ⁇ 1 , number of injured ⁇ 2 , direct property loss ⁇ 3 , accident type ⁇ 4 , and accident form ⁇ 5 are decision attributes.
- On-site management by traffic police Signal Control System Driver's age Dangerous driving and illegal behavior of drivers Traffic traffic density Ratio of large vehicles Accident rate per 10,000 vehicles road alignment road condition road infrastructure Cross Section Settings Marking line setting weather environment lighting conditions Visibility condition As the condition attribute, construct the cause condition attribute table of human, vehicle, road and environment accidents.
- R * is the dependence degree of feature attribute set *, is the number of elements in the *positive domain of B, and the number of elements in
- the dependence degree of the attribute sets of the four accident cause condition attribute tables respectively represent the impact of the four types of factor sets of "people, vehicles, roads, and environment" on the severity of traffic accidents.
- the attribute table of different characteristics is used as the total set and analogized:
- This sub-module judges the severity of the accident based on the decision-making attributes of the accident and mines the cause of the accident to evaluate and analyze the importance of four aspects: people, vehicles, roads, and the environment.
- people mainly analyzes the impact of drivers' driving behavior and traffic control measures of traffic managers on road safety accidents;
- vehicles it mainly analyzes the impact of traffic parameters such as traffic volume, vehicle type ratio, and accident rate per 10,000 vehicles on road safety accidents.
- traffic parameters such as traffic volume, vehicle type ratio, and accident rate per 10,000 vehicles on road safety accidents.
- For roads it mainly analyzes the impact of basic road environment such as road alignment, infrastructure, and road surface conditions on road safety accidents
- for the environment it mainly analyzes the impact of weather environment and lighting conditions on road safety accidents.
- Platform users pay different degrees of attention to the causes of road safety accidents according to their importance, so that the traffic management work is more targeted.
- Table 1 The importance of condition attributes to the severity of traffic accidents:
- the occurrence of road traffic accidents is mainly caused by the breakdown of the coordinated balance among people, roads, environment and vehicles. Therefore, a neural network safety evaluation model based on the entropy weight method and expert judgment is constructed based on the four main factors of people, roads, environment, and vehicles. The mixed weights of data entropy and traffic safety experts' scores are used as the weight matrix for constructing the neural network model.
- x ij represents the i-th index value of the road in the j-th evaluated area.
- the second step is to use fuzzy comprehensive evaluation to analyze the safety of road traffic, and compare the relative advantages and disadvantages of different indicators.
- the fuzzy evaluation expert weight matrix is shown in Table 2, and the human-computer hybrid enhanced weight matrix is constructed in this way
- the calculation formula is:
- the input layer of the neural network model is constructed from four aspects of human factors, vehicle factors, road factors, environmental factors and traffic flow factors to evaluate road traffic safety.
- the 16 evaluation indicators of 4 different risks are used as the input neurons of the neural network. Quantify these 16 road traffic safety evaluation indicators and evaluation results, and collect historical data of these indicators for a period of time as neural network training learning samples (x 1 ,...,x n ) and network output y.
- the training learning samples (x 1 ,...,x n ) and the network output y are fuzzy processed, and then the processed input and output are used as the training samples of the neural network.
- the samples are trained, and the error between the actual output value of the network and the output value of the neural network is calculated. If the error value is within the error range, the neural network learning ends, otherwise modify the number of hidden layer nodes and network weights again. Retrain the samples until the error reaches an acceptable range, and the neural network learning ends.
- the safety evaluation model inputs different road traffic safety evaluation index data to run the risk evaluation index value, and obtains the network evaluation result. Defuzzify the network evaluation results, and then use the human-machine hybrid to enhance the weight matrix
- the evaluation value P safety of road traffic safety is obtained by weighting each index.
- vehicle-road linkage is carried out with the early warning facilities around the road to build a road early warning network that integrates vehicles and roads.
- the vehicles and road facilities near the vehicle will get the vehicle safety early warning data in a timely manner, and seek help from the traffic manager, forming a road early warning environment in which people, vehicles, and roads respond to each other and intelligently cooperate, which can be good handling of emergencies.
- Human-machine hybrid enhanced safety data research and judgment and early warning sub-module provides traffic participants with road safety data through vehicle sensing equipment and road infrastructure, evaluates road safety, helps traffic participants better understand the degree of road safety, and reduces traffic Accidents happen.
- new traffic technologies such as the Internet of Things, cloud computing, and big data
- the urban traffic safety early warning system has comprehensive and thorough perception, overall system control, and quick and precise response, and the construction of public safety governance subjects can share data, interconnect, and form an integrated system.
- Embodiment 2 of the present disclosure provides an urban traffic safety early warning method based on man-machine hybrid enhanced intelligence, including the following process:
- S1 Obtain data related to urban traffic early warning, at least including manual data collection and machine data collection;
- S2 Preprocessing the acquired data, at least including data patching and prediction;
- S3 Classify the preprocessed data, establish a traffic accident cause attribute table according to the cause of the accident, select traffic accidents in multiple sections of urban roads as a sample group, and obtain the importance of different feature attribute index sets for accidents, and then The attribute tables of different features are analogous to the total set, and the importance of the feature attributes of different feature sets is obtained; according to the importance of the acquired feature attributes, the accident cause evaluation is performed.
- S4 According to the classified data and the preset neural network model, obtain the human-machine hybrid enhancement weight matrix; use the human-machine hybrid enhancement weight matrix to weight each accident factor to obtain the evaluation value of road traffic safety; according to the evaluation value and the preset The comparison of the thresholds generates and/or issues an early warning command.
- S1, S2, S3 and S4 are the same as the man-machine hybrid enhanced data acquisition sub-module, human-machine hybrid enhanced data processing sub-module, human-machine hybrid enhanced accident cause analysis sub-module, human-machine hybrid enhanced safety in embodiment 1
- the working methods of the data research and judgment and early warning sub-modules are corresponding, and will not be repeated here.
- Embodiment 3 of the present disclosure provides a computer-readable storage medium on which a program is stored.
- the program is executed by a processor, the urban traffic safety early warning method based on human-computer hybrid enhanced intelligence as described in Embodiment 2 of the present disclosure is implemented. in the steps.
- the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Accordingly, the present disclosure may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) having computer-usable program code embodied therein.
- a computer-usable storage media including but not limited to disk storage, optical storage, etc.
- These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions
- the device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.
- the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random AccessMemory, RAM), etc.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Analytical Chemistry (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Traffic Control Systems (AREA)
Abstract
一种基于人机混合增强智能的城市交通安全预警方法及***,该方法包括:获取城市交通预警相关数据;对获取的数据进行预处理;对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;根据获取的特征属性的重要度,进行事故成因评价;根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵;利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值;根据评价数值与预设阈值的比较生成和/或发出预警指令;该方法提高了城市交通安全的预警准确度。
Description
本公开涉及交通安全预警技术领域,特别涉及一种基于人机混合增强智能的城市交通安全预警方法及***。
本部分的陈述仅仅是提供了与本公开相关的背景技术,并不必然构成现有技术。
城市的基础交通设施的不足、驾驶员交通安全意识的浅薄和交通管控策略的不合理导致了交通事故的发生率持续升高。当前的交通安全预警***主要是针对交通环境中的人或车处于危险状态时,积极采取措施进行调控。例如当驾驶员身心理状态不佳、道路天气状况恶劣、车辆状况不佳、路面状况不佳等情况发生时,经过监测将其向安全的态势转变,或者通过监测,提前预知从而减少损失,避免交通事故的发生。
发明人发现,现有的交通安全预警***在数据采集策略、数据处理策略、事故成因评价策略和安全预警策略中大多只依靠物理数据的采集、处理和分析,没有考虑人机混合数据的采集和人机混合数据的处理分析,使得最终的预警***的准确度较低。
发明内容
为了解决现有技术的不足,本公开提供了一种基于人机混合增强智能的城市交通安全预警方法及***,结合人机混合数据的采集、处理和分析,提高了安全预警的效率和准确度。
为了实现上述目的,本公开采用如下技术方案:
本公开第一方面提供了一种基于人机混合增强智能的城市交通安全预警方法。
一种基于人机混合增强智能的城市交通安全预警方法,包括以下过程:
获取城市交通预警相关数据,至少包括人工采集数据和机器采集数据;
对获取的数据进行预处理,至少包括数据的修补和预测;
对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;
根据获取的特征属性的重要度,进行事故成因评价。
进一步的,根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵;
利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值;
根据评价数值与预设阈值的比较生成和/或发出预警指令。
更进一步的,预设神经网络根据熵权法和专家判断法构建,根据熵权法得到熵权权重矩阵,根据专家判断法构建模糊评价专家权重矩阵,以熵权权重矩阵和模糊评价专家权重矩阵之和的均值为人机混合增强权重矩阵。
进一步的,人工采集数据包括社会因素数据、驾驶员行为状态数据和交通管理部门安全管控经验数据;机器采集数据至少包括道路数据、交通事故数据、自然环境数据和车辆数据。
进一步的,对获取的数据进行预处理,包括以下过程:
当获取的城市交通预警相关数据中某一类型交通数据为零或者不在预设范围内时,判定获取的数据为空白数据或错误数据;
利用该路段前后时段的数据或将相邻路段同时段的数据进行比较修复,利用预设神经网络算法进行预测,得到交通数据预测结果,赋予不同权重进行加权融合,得到融合预测值。
更进一步的,利用人机混合增强学习进行知识库构建、知识库更新和智能增强,包括以下过程:
将现有的数据库合并成一个多源数据库,让不同数据库碰撞重组构建基础知识库对符合预设条件的数据进行处理;
对不同类型、时间长度的数据进行抓取,在抓取失败时或处理数据发生问题时,发出请求专家协助的信号,自动记录数据模型以及专家的解决策略,并将新的数据处理方案加入原有知识库,不断更新基础知识库;
利用机器学习算法将各种异构数据化为同维数据输入知识库,随后对所得到的数据进行置信性判断,若数据的置信性大于预设阈值,则数据用于提供给数据使用者,若数据置信性小于预设阈值,则重新进入知识库训练直至得到高置信性的数据为止;
随机将数据处理记录加入知识库,更新知识库。
本公开第二方面提供了一种基于人机混合增强智能的城市交通安全预警***。
一种基于人机混合增强智能的城市交通安全预警***,包括:
人机混合增强数据采集子模块,被配置为:获取城市交通预警相关数据,至少包括人工采集数据和机器采集数据;
人机混合增强数据处理子模块,被配置为:对获取的数据进行预处理,至少包括数据的修补和预测;
人机混合增强事故成因分析子模块,被配置为:对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到 不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;根据获取的特征属性的重要度,进行事故成因评价。
进一步的,还包括:人机混合增强安全数据研判及预警子模块,被配置为:根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵,利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值,根据评价数值与预设阈值的比较生成和/或发出预警指令。
进一步的,人机混合增强数据处理子模块,还被配置为:利用人机混合增强学习进行知识库构建、知识库更新和智能增强,包括以下过程:
将现有的数据库合并成一个多源数据库,让不同数据库碰撞重组构建基础知识库对符合预设条件的数据进行处理;
对不同类型、时间长度的数据进行抓取,在抓取失败时或处理数据发生问题时,发出请求专家协助的信号,自动记录数据模型以及专家的解决策略,并将新的数据处理方案加入原有知识库,不断更新基础知识库;
利用机器学习算法将各种异构数据化为同维数据输入知识库,随后对所得到的数据进行置信性判断,若数据的置信性大于预设阈值,则数据用于提供给数据使用者,若数据置信性小于预设阈值,则重新进入知识库训练直至得到高置信性的数据为止;
随机将数据处理记录加入知识库,更新知识库。
本公开第三方面提供了一种计算机可读存储介质,其上存储有程序,该程序被处理器执行时实现如本公开第一方面所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
本公开第四方面提供了一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的程序,所述处理器执行所述程序时实现如本公开第一方面所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
与现有技术相比,本公开的有益效果是:
1、本公开所述的方法、***、介质或电子设备,通过路侧交通设备执行自动化重复的交通数据采集行为,交通参与者和交通管理者将自己的人工感知和管理经验以数字化形式输入数据采集子模块,通过简单的人工决策操作提取需要的交通数据;整个流程中采用人机混合增强智能,利用机器处理重复的规律性操作;减少了人工的消耗,提高了数据采集效率也保证了数据的精确性和完整性。
2、本公开所述的方法、***、介质或电子设备,对采集的交通数据中的缺失、错误、异常数据进行辨别和修复处理,利用不同路段的历史数据构建时空融合人机混合增强数据预测 模型进行数据预测;将机器的超强计算能力与人类的感知推理能力相结合,实现人机交互,快速高效的解决数据处理问题。
3、本公开所述的方法、***、介质或电子设备,对城市道路发生的事故进行统计分析,利用大数据技术和粗糙集理论,对事故的成因进行数据挖掘,计算事故成因重要度确定主要事故成因。
4、本公开所述的方法、***、介质或电子设备,运用信息熵和专家分析进行模糊综合评价构建人机混合增强城市安全预警评价模型,为城市道路提供安全评估;通过人机混合增强感知算法来为交通管理者提供道路安全数据与分析结果,辅助交通管理者做出决策,提高了分析精度。
构成本公开的一部分的说明书附图用来提供对本公开的进一步理解,本公开的示意性实施例及其说明用于解释本公开,并不构成对本公开的不当限定。
图1为本公开实施例1提供的基于人机混合增强智能的城市交通安全预警***的结构示意图。
图2为本公开实施例1提供的人机混合增强信数据采集子模块的工作流程示意图。
图3为本公开实施例1提供的人机混合增强数据处理子模块的工作流程示意图。
图4为本公开实施例1提供的人机混合增强事故成因分析子模块的工作流程示意图。
图5为本公开实施例1提供的人机混合增强安全数据研判及预警子模块的工作流程示意图。
下面结合附图与实施例对本公开作进一步说明。
应该指出,以下详细说明都是例示性的,旨在对本公开提供进一步的说明。除非另有指明,本文使用的所有技术和科学术语具有与本公开所属技术领域的普通技术人员通常理解的相同含义。
需要注意的是,这里所使用的术语仅是为了描述具体实施方式,而非意图限制根据本公开的示例性实施方式。如在这里所使用的,除非上下文另外明确指出,否则单数形式也意图包括复数形式,此外,还应当理解的是,当在本说明书中使用术语“包含”和/或“包括”时,其指明存在特征、步骤、操作、器件、组件和/或它们的组合。
在不冲突的情况下,本公开中的实施例及实施例中的特征可以相互组合。
实施例1:
如图1-5所示,本公开实施例1提供了一种基于人机混合增强智能的城市交通安全预警***,通过路侧的道路监控设施进行重复的数据采集工作,运用人机混合增强智能将交通管理者交通管控经验融入数据采集过程;确定所需要的数据类型及种类之后对数据进行筛选,缺失、错误、异常数据修复处理,并利用交通运输行业专家的专业知识构建数据预测模型预测交通流量等基础交通参数;通过粗糙集理论构建事故成因分析模型,确定事故的主要成因;以熵值法和神经网络为算法核心融合构建人机混合增强的城市安全预警研判模型,对城市路网上的道路进行道路安全评价,同时对不同道路进行安全评级为驾驶员提供实时的道路安全预警。基于人机混合增强智能的城市交通安全预警平台将交通参与者的经验智慧和机器学习相结合综合考虑交通安全的诸多影响因素,通过对事故成因进行数据挖掘分析快速高效找出道路交通***中存在的问题,并采取具有针对性的改善措施,达到减少交通事故、改善交通安全的目的。
具体的,包括:人机混合增强数据采集子模块、人机混合增强数据处理子模块、人机混合增强事故成因分析子模块、人机混合增强安全数据研判及预警子模块。
(1)人机混合增强数据采集子模块
数据采集子模块是基于人机混合增强智能的城市交通安全预警平台的基础组成***。城市交通中安全预警数据的研判需要精确的道路交通数据为基础,因此,采集合理精确地采集交通数据是整个预警平台高效运转的前提。
按采集方式将城市道路交通安全预警所需要的数据分为两大类,一类为交通参与者的人工采集数据,另一类为道路检测器采集的机器采集数据。
人工采集数据可分为社会因素数据、驾驶员行为状态数据、交通管理部门安全管控经验数据,机器采集数据可分为道路数据、交通事故数据、自然环境数据、车辆数据。
社会因素数据是指受节假日、疫情、游行集会等社会事件数据。驾驶员行为状态数据是指驾驶员的心理状态、驾龄、危险驾驶行为记录、身体状态等驾驶员的状态数据。交通管理部门安全经验数据可分为交通安全法规、交管部门违法处理经验和信号控制***的信号控制策略等交通管控经验。
道路设施数据可分为道路基础设施状况、道路线形、道路地形、道路路面状况等道路综合数据。交通事故数据可分为事故责任、万车事故率、财产损失、伤亡情况等事故影响数据。交通坏境数据可分为天气数据(雨、雪、雾等)、视觉环境(道路照明、道路通视条件、能见度等)、车流环境(交通流量、排队长度、跟驰距离等)。车辆数据可分为车型、车身颜色、车速等车辆数据。
当车辆行驶进入城市路网时,路网上的交通传感设备和交通参与者同时采集车辆及周边环境数据。将整个城市的路网分为若干个不同的子区域,进行分布式采集数据。不同的子区域实行数据的互联构成一个区域传输网络。之后把不同采集方式采集来的数据分为四种不同的特征数据上传至交通管理部门的数据采集云端组成交通安全数据库。
人机混合增强智能数据采集子模块将人工智能与机器智能相结合从两个采集道路交通安全研判所需要的数据,提高了数据的准确性同时也加强了安全预警平台的鲁棒性。该***从人、车、路、环境四个特征方向分类收集道路交通安全事故影响因素。人的特征数据包括驾驶员的安全意识、心理状态、身体状态(如疲劳程度和清醒程度)、驾驶员驾龄和违法记录、交通控制***的信号控制策略等。车的特征数据包括车辆的车型、行驶速度、车辆位置、车辆事故损失程度及类型等。道路的特征数据包括道路基础设施数据、万车事故率、道路的技术等级、道路安全事故发生率、道路的地形条件、安全事故频发地段的位置等。环境的特征数据包括自然天气数据、道路的视觉环境数据、道路交通环境数据、道路车流环境数据、道路周边土地利用等。人机混合增强的安全事故数据采集子模块采用分布式数据采集方法增强了***的鲁棒性。而人机混合增强智能对不同采集方式采集来的同类数据进行相互验证得到更加真实准确的数据。
(2)人机混合增强数据处理子模块
在数据的采集过程中,因为传感器出现故障或者行人的失误可能会导致部分数据出现错误,需要对这些错误数据进行筛选并修复。此外安全预警模型的构建中需要大量的基础交通数据,因此不仅需要数据采集子模块采集的实时监测数据和各个路段的历史数据还需要对路段的交通数据进行预测。人机混合增强数据处理子模块可利用路侧交通设备的传感装置感知环境,记录认知条件,通过网络平台与人进行交互获得认知策略。该子模块的核心是引入人的智慧经验,通过人的智慧帮助机器解决问题,而机器又可借助人的策略进化增强。
人机混合增强数据处理子模块由数据判断修补模块、人机混合增强学习模块组成。
数据判断修补模块即判定错误和空白数据并对数据进行修复和预测。
第一步,采集数据时,通常会对某一段时间内的数据进行统计储存,若在一定时段内部分数据未被录入,则判定为空白数据。某些数据波动较大超过一般的范畴,则判定为错误数据。对这些空白数据和错误数据进行辨别,判定规则如下:
假设数据采集子模块采集的数据为X,某一类型交通参数最大值为X
max,最小值为X
min,大多数情况下空白数据为零,当X=0或
则判定获得数据为空白数据或错误 数据,此时需要对这两类数据进行修复处理。
第二步,利用该路段前后时段的数据或将相邻路段同时段的数据进行比较修复,充分考虑各种意外情况给数据进行数据修复处理保证精确性和规律性。假设数据出现空白数据或错误数据的时间是t时段该数据为X
k(t),前一天的历史数据是X
k-1(t),以此类推结合前一个月的历史数据对该数据进行修复得到时间修复数据:
同时根据错误数据所在路段相邻路段的交通参数进行数据修复,设相邻路段的历史数据为X
k(t-1)或X
k(t+1),以此为基础得到空间修复数据:
之后将时间修复数据与空间修复数据相结合得到修复数据X
k(t):
充分的考虑数据的时间特性和空间特性利用路段的历史数据和相邻路段交通参数综合分析得到了较为精确的修复数据。之后指数平滑预测法、神经网络预测法等适合的预测方法进行预测得到高精度的预测交通数据赋予不同权重进行加权融合,得到融合预测值X
融合(k)。
人机混合增强学习模块具有知识库构建、知识库更新和智能增强等功能。将现有的数据库合并成一个多源数据库。在数据统计阶段***,根据专家和交通行业资深从业人员的经验让不同数据库碰撞重组构建基础知识库对符合预设条件的数据进行处理,依据***使用者的需求对不同类型、时间长度的数据进行抓取。数据处理阶段时,机器通过智能算法抓取需要的数据,在抓取失败时或处理数据发生问题时,机器会发出信号请求专家协助。同时人机混合增强学习模块会自动记录将数据模型(即认知输入条件),以及专家的解决策略(包括数据处理方案及注意事项等)称为认知策略,并将新的数据处理方案加入原有知识库,不断更新基础知识库。
增强智能阶段,利用机器学***。
基于人机混合增强的数据处理子模块通过人工智能与机器学***台提供了坚实的数据基础。
(3)人机混合增强事故成因分析子模块
事故成因分析子模块利用粗糙集理论对交通事故中的事故原因进行分析挖掘事故主要成因。
首先根据事故的成因建立交通事故成因属性表B,研究不同属性对交通事故发生与否以及事故严重程度的影响。以事故的死亡人数α
1、受伤人数α
2、直接财产损失α
3、事故类型α
4、事故形态α
5为决策属性。以交警现场管理
信号控制***
驾驶员驾龄
驾驶员危险驾驶及违法行为
交通量
车流密度
大型车比例
万车事故率
道路线形
路面状况
道路基础设施
横断面设置
标志标线设置
天气环境
照明条件
通视条件
为条件属性,构建人、车、路、环境事故成因条件属性表。
为了高精确度的结果计算先样本群各个属性依赖度的平均值、方差进行T检验得到各属性的显著性,在城市道路中选取N条路段的交通事故为样本群,计算得到不同特征属性指标集对事故的重要度:
R
*为特征属性集*的依赖度,
为B的*正域中元素的个数,|U|论域U中元素的个数。四个事故成因条件属性表属性集的依赖度分别表示“人、车、路、环境”这四类因素集对交通事故严重程度的影响。同时根据上述公式,以不同特征的属性表为总集类推:
该子模块依据事故的决策属性判断事故严重程度挖掘事故成因从人、车、路、环境四个方面开展重要度评估及分析。对于人主要分析驾驶员的驾驶行为和交通管理者的交通管控措施对道路安全事故的影响;对于车主要分析道路上的车流量、车型比例、万车事故率等交通参数对于道路安全事故的影响;对于道路主要分析道路线形、基础设施、路面状况等基本道路环境对于道路安全事故的影响;对环境主要分析天气环境和照明条件对于道路安全事故的影响。平台使用者对造成道路安全事故的事故成因根据其重要度给予不同程度的关注,从而使交通管理工作更加具有针对性。
表1:条件属性对交通事故严重程度的重要度:
(4)人机混合增强安全数据研判及预警子模块
道路交通事故的发生主要是由于人、道路、环境、车辆协调平衡被打破所造成的。因此以人、道路、环境、车辆四大主要因素构建基于熵权法和专家判断的神经网络安全评价模型。利用数据熵和交通安全专家的评分混合权重作为构建神经网络模型的权重矩阵。
第一步利用数据处理子模块得到各类评价指标的历史数据构建熵权多指标评价矩阵X=(x
ij)
m×n,城市道路按分为n个不同区域特征的道路。则预设该矩阵有m个评价指标,n个评价区域道路,则:
其中,x
ij表示第j个被评价区域道路的第i个指标值。对矩阵X=(x
ij)
m×n进行以数据标准化公式:
maxx
ij为评价指标中的最大值,minx
ij评价指标的最小值处理得到R=(r
ij)
m×n,其中r
ij∈[0,1]。根据数据熵的定义及熵权计算公式得到不同道路安全评价指标的熵权。
评价指标的数据熵的计算如下所示:
根据熵权的计算公式:
第二步,利用模糊综合评价分析道路交通的安全性,比较不同指标的相对优劣程度。在 不同区域道路或同一条区域道路的不同路段,计算出各条道路的安全程度和不同因素对道路安全的影响情况。通过专家评价法获得不同评价因素的权重并进行归一化处理,推导出的不同评价指标的单层次模糊评判矩阵R=(r
ij)
m×n其中r
ij表示关于第i个指标的第j个被评价区域道路的隶属度,得到权重向量集W=(ω
ij)
m×n,通过隶属度r
ij和权向量集合ω
ij计算出模糊评价权重γ
ij,根据模糊评价权重计算公式γ
ij=r
ij×ω
ij,构建模糊评价专家权重矩阵
利用交警现场管理数据、信号控制***数据、驾驶员驾龄数据、驾驶员危险驾驶行为及违法数据构建人为因素评价指标,万车事故率、交通流量、车流密度、大型车比例构建车辆因素评价指标,道路基础设施、道路路面状况、横断面设置、道路线形、道路标志数据构建道路因素评价指标,天气环境数据(雨、雪、雾等不良天气)、照明条件、通视条件等道路环境数据构建道路因素评价指标。选用人为因素、车辆因素、道路因素、环境因素和交通流因素风险4个方面构建神经网络模型的输入层,对道路交通安全进行评价。
以4类不同风险的16个评价指标作为神经网络的输入神经元。对这16个道路交通安全度评价指标和评价结果进行量化,采集这些指标的一段时间的历史数据作为神经网络训练学习样本(x
1,...,x
n)和网络输出y。
对训练学习样本(x
1,...,x
n)和网络输出y模糊化处理,之后将经过处理的输入量和输出量作为神经网络的训练样本。对样本进行训练,计算网络的实际输出值与神经网络输出值的误差。如果误差值在误差范围内时,神经网络学习结束,否则再次修改隐含层节点数和网络权值。重新对样本进行训练直至误差达到可接受范围,神经网络学习结束。最后安全评价模型输入不同的道路交通安全评价指标数据运行风险评价指标值,得出网络评价结果。对网络评价结果进行去模糊化处理,再利用人机混合增强权重矩阵
对各项指标进行加权处理得到道路交通安全的评价数值P
安全。
以不同的数据传递方式将道路安全数据传递至交通参与者,向安全评价较高的路段靠近的行人或驾驶员及时做出预警。当某路段道路安全评价较低时,不仅要对行人和驾驶员进行预警。同时交通管理部门需要介入该路段采取相应措施排除安全隐患。若安全评价极高可以 采取网络数据发布的模式对行人和进行预警。
建立安全预警平台时要预先对行人和驾驶员分发安全预警手册,对安全预警研判子模块构建的评价指标进行普及,方便行人和驾驶员理解不同的安全预警等级。以车载的导航***以及驾驶员的手机App为核心预警方式,向驾驶员及时通报道路安全情况以及道路的交通状况。当导航与手机APP无法取得预期效果时,可利用交通广播、路侧基础设施等道路上的预警设备。为交通参与者提供即使有效的道路安全状况,以便驾驶员及时做出新的交通规划。利用车辆之间的数据互联,将道路上的车辆为节点实行车辆之间的预警数据传递。同时与道路周边的预警设施进行车路联动,构建一个车路融合的道路预警网络。当路段上某一车辆出现问题时,车辆附近的车辆和道路设施及时的得到车辆安全预警数据,并向交通管理者寻求帮助,形成人、车、路相互响应智能协作的道路预警环境,能够良好的处理应急突发事故。
人机混合增强安全数据研判及预警子模块通过车辆传感设备和道路基础设施为交通参与者提供道路安全数据,对道路进行安全评价,帮助交通参与者更好地了解道路安全程度,减少了交通事故的发生。借助物联网、云计算、大数据等全新的交通技术,城市交通安全预警体系具有全面透彻的感知、***整体的掌控和迅捷精确的响应,构建公共安全治理主体能够数据共享、互联互通,形成一体化的预警防控体系。
表2:模糊评价专家权重矩阵数据列表
实施例2:
本公开实施例2提供了一种基于人机混合增强智能的城市交通安全预警方法,包括以下过程:
S1:获取城市交通预警相关数据,至少包括人工采集数据和机器采集数据;
S2:对获取的数据进行预处理,至少包括数据的修补和预测;
S3:对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;根据获取的特征属性的重要度,进行事故成因评价。
S4:根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵;利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值;根据评价数值与预设阈值的比较生成和/或发出预警指令。
S1、S2、S3和S4的具体工作方法与实施例1中的人机混合增强数据采集子模块、人机混合增强数据处理子模块、人机混合增强事故成因分析子模块、人机混合增强安全数据研判及预警子模块的工作方法相对应,这里不再赘述。
实施例3:
本公开实施例3提供了一种计算机可读存储介质,其上存储有程序,该程序被处理器执行时实现如本公开实施例2所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
实施例4:
本公开实施例4提供了一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的程序,所述处理器执行所述程序时实现如本公开实施例1所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
本领域内的技术人员应明白,本公开的实施例可提供为方法、***、或计算机程序产品。因此,本公开可采用硬件实施例、软件实施例、或结合软件和硬件方面的实施例的形式。而且,本公开可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
本公开是参照根据本公开实施例的方法、设备(***)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器,使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存储记忆体(Random AccessMemory,RAM)等。
以上所述仅为本公开的优选实施例而已,并不用于限制本公开,对于本领域的技术人员来说,本公开可以有各种更改和变化。凡在本公开的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本公开的保护范围之内。
Claims (10)
- 一种基于人机混合增强智能的城市交通安全预警方法,其特征在于:包括以下过程:获取城市交通预警相关数据,至少包括人工采集数据和机器采集数据;对获取的数据进行预处理,至少包括数据的修补和预测;对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;根据获取的特征属性的重要度,进行事故成因评价。
- 如权利要求1所述的基于人机混合增强智能的城市交通安全预警方法,其特征在于:根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵;利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值;根据评价数值与预设阈值的比较生成和/或发出预警指令。
- 如权利要求2所述的基于人机混合增强智能的城市交通安全预警方法,其特征在于:预设神经网络根据熵权法和专家判断法构建,根据熵权法得到熵权权重矩阵,根据专家判断法构建模糊评价专家权重矩阵,以熵权权重矩阵和模糊评价专家权重矩阵之和的均值为人机混合增强权重矩阵。
- 如权利要求1所述的基于人机混合增强智能的城市交通安全预警方法,其特征在于:人工采集数据包括社会因素数据、驾驶员行为状态数据和交通管理部门安全管控经验数据;机器采集数据至少包括道路数据、交通事故数据、自然环境数据和车辆数据。
- 如权利要求1所述的基于人机混合增强智能的城市交通安全预警方法,其特征在于:对获取的数据进行预处理,包括以下过程:当获取的城市交通预警相关数据中某一类型交通数据为零或者不在预设范围内时,判定获取的数据为空白数据或错误数据;利用该路段前后时段的数据或将相邻路段同时段的数据进行比较修复,利用预设神经网络算法进行预测,得到交通数据预测结果,赋予不同权重进行加权融合,得到融合预测值。
- 如权利要求5所述的基于人机混合增强智能的城市交通安全预警方法,其特征在于:利用人机混合增强学习进行知识库构建、知识库更新和智能增强,包括以下过程:将现有的数据库合并成一个多源数据库,让不同数据库碰撞重组构建基础知识库对符合预设条件的数据进行处理;对不同类型、时间长度的数据进行抓取,在抓取失败时或处理数据发生问题时,发出请求专家协助的信号,自动记录数据模型以及专家的解决策略,并将新的数据处理方案加入原 有知识库,不断更新基础知识库;利用机器学习算法将各种异构数据化为同维数据输入知识库,随后对所得到的数据进行置信性判断,若数据的置信性大于预设阈值,则数据用于提供给数据使用者,若数据置信性小于预设阈值,则重新进入知识库训练直至得到高置信性的数据为止;随机将数据处理记录加入知识库,更新知识库。
- 一种基于人机混合增强智能的城市交通安全预警***,其特征在于:包括:人机混合增强数据采集子模块,被配置为:获取城市交通预警相关数据,至少包括人工采集数据和机器采集数据;人机混合增强数据处理子模块,被配置为:对获取的数据进行预处理,至少包括数据的修补和预测;人机混合增强事故成因分析子模块,被配置为:对预处理后的数据进行分类处理,根据事故的成因建立交通事故成因属性表,选取城市道路中多条路段的交通事故为样本群,得到不同特征属性指标集对事故的重要度,以不同特征的属性表为总集类推,得到不同特征集特征属性的重要度;根据获取的特征属性的重要度,进行事故成因评价。
- 如权利要求7所述的基于人机混合增强智能的城市交通安全预警***,其特征在于:还包括:人机混合增强安全数据研判及预警子模块,被配置为:根据分类后的数据和预设神经网络模型,得到人机混合增强权重矩阵,利用人机混合增强权重矩阵对各事故因素进行加权处理得到道路交通安全的评价数值,根据评价数值与预设阈值的比较生成和/或发出预警指令;或者,人机混合增强数据处理子模块,还被配置为:利用人机混合增强学习进行知识库构建、知识库更新和智能增强,包括以下过程:将现有的数据库合并成一个多源数据库,让不同数据库碰撞重组构建基础知识库对符合预设条件的数据进行处理;对不同类型、时间长度的数据进行抓取,在抓取失败时或处理数据发生问题时,发出请求专家协助的信号,自动记录数据模型以及专家的解决策略,并将新的数据处理方案加入原有知识库,不断更新基础知识库;利用机器学习算法将各种异构数据化为同维数据输入知识库,随后对所得到的数据进行置信性判断,若数据的置信性大于预设阈值,则数据用于提供给数据使用者,若数据置信性小于预设阈值,则重新进入知识库训练直至得到高置信性的数据为止;随机将数据处理记录加入知识库,更新知识库。
- 一种计算机可读存储介质,其上存储有程序,其特征在于,该程序被处理器执行时实现如权利要求1-/6任一项所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
- 一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的程序,其特征在于,所述处理器执行所述程序时实现如权利要求1-6任一项所述的基于人机混合增强智能的城市交通安全预警方法中的步骤。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110644617.7A CN113380033B (zh) | 2021-06-09 | 2021-06-09 | 基于人机混合增强智能的城市交通安全预警方法及*** |
CN202110644617.7 | 2021-06-09 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022257201A1 true WO2022257201A1 (zh) | 2022-12-15 |
Family
ID=77573318
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2021/103202 WO2022257201A1 (zh) | 2021-06-09 | 2021-06-29 | 基于人机混合增强智能的城市交通安全预警方法及*** |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN113380033B (zh) |
WO (1) | WO2022257201A1 (zh) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115782905A (zh) * | 2023-01-31 | 2023-03-14 | 北京航空航天大学 | 一种自动驾驶车辆驾驶安全度量化*** |
CN116665453A (zh) * | 2023-06-25 | 2023-08-29 | 河南大学 | 基于二级模糊综合评价的高速公路事故严重程度预测方法 |
CN117744908A (zh) * | 2024-02-19 | 2024-03-22 | 深圳市深河环保水务有限公司 | 一种基于机器视觉的城市排水设施巡检方法及*** |
CN118036337A (zh) * | 2024-04-01 | 2024-05-14 | 山东凤麟工业装备有限公司 | 一种基于数字仿真的风电运输倾倒预警方法及*** |
CN118115341A (zh) * | 2024-04-30 | 2024-05-31 | 中亿丰数字科技集团股份有限公司 | 一种基于时空大数据模型的城市规建管服的方法及*** |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114548836A (zh) * | 2022-04-25 | 2022-05-27 | 杭州玳数科技有限公司 | 一种基于疫情的多因素交通枢纽运营方法和*** |
CN116403403B (zh) * | 2023-04-12 | 2024-02-02 | 西藏金采科技股份有限公司 | 一种基于大数据分析的交通预警方法、***、设备及介质 |
CN116758763B (zh) * | 2023-05-06 | 2024-02-20 | 西藏金采科技股份有限公司 | 一种基于车联网的交通数据处理***及方法 |
CN117787699A (zh) * | 2023-12-26 | 2024-03-29 | 公安部道路交通安全研究中心 | 一种道路风险预测方法、装置、计算机设备及存储介质 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104978853A (zh) * | 2014-04-01 | 2015-10-14 | ***通信集团公司 | 一种道路交通安全评估方法及*** |
CN106228499A (zh) * | 2016-07-06 | 2016-12-14 | 东南大学 | 一种基于人‑车‑路‑货多风险源的货运安全评价模型 |
US20180107935A1 (en) * | 2016-10-18 | 2018-04-19 | Uber Technologies, Inc. | Predicting safety incidents using machine learning |
CN108182522A (zh) * | 2017-12-25 | 2018-06-19 | 中国电子科技集团公司第二十八研究所 | 一种基于ahp-熵权法的航道交通安全风险评估方法 |
CN109711691A (zh) * | 2018-12-17 | 2019-05-03 | 长安大学 | 一种基于熵权模糊综合评价模型的驾驶风格评价方法 |
CN111583639A (zh) * | 2020-04-30 | 2020-08-25 | 山东交通学院 | 一种道路交通拥堵预警方法及*** |
CN112836509A (zh) * | 2021-02-22 | 2021-05-25 | 西安交通大学 | 一种专家***知识库构建方法及*** |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101480449B1 (ko) * | 2007-10-16 | 2015-01-12 | 엘지전자 주식회사 | 멀티미디어 기반 교통 및 여행 정보 메시지를 이용한 상세정보 제공방법 및 이를 수행하기 위한 단말기 |
CN101465059B (zh) * | 2008-12-31 | 2010-06-09 | 公安部交通管理科学研究所 | 城市道路交通安全态势鉴判预警*** |
CN105243255A (zh) * | 2015-08-11 | 2016-01-13 | 北华航天工业学院 | 一种软基处理方案的评价方法 |
CN106056308A (zh) * | 2016-06-13 | 2016-10-26 | 宁波工程学院 | 一种公路隧道运行环境安全风险自动判定方法 |
CN107719376B (zh) * | 2017-09-18 | 2019-10-29 | 清华大学 | 人机混合增强智能驾驶***及电动汽车 |
CN109145170B (zh) * | 2018-06-27 | 2022-12-20 | 公安部道路交通安全研究中心 | 一种道路交通事故数据挖掘服务器、方法和*** |
CN111105153A (zh) * | 2019-12-13 | 2020-05-05 | 西安交通大学 | 基于ahp-熵权法的卫星健康状态多级模糊评价方法 |
CN110996366B (zh) * | 2019-12-13 | 2021-10-22 | 哈尔滨工业大学 | 一种异构专用网络垂直切换中权重确定方法 |
CN111898842A (zh) * | 2020-04-20 | 2020-11-06 | 国网上海市电力公司 | 一种基于模糊熵权的黑启动方案评估方法 |
CN111784873A (zh) * | 2020-07-01 | 2020-10-16 | 上海城市交通设计院有限公司 | 基于机器学习原理的高速路安全车载智能***及工作流程 |
CN112069662A (zh) * | 2020-08-20 | 2020-12-11 | 北京仿真中心 | 一种基于人机混合增强的复杂产品自主构建方法和模块 |
CN112183978A (zh) * | 2020-09-19 | 2021-01-05 | 西安石油大学 | 一种基于修正熵权法的油气管道土壤腐蚀分级评价方法 |
CN112347698A (zh) * | 2020-11-12 | 2021-02-09 | 东北大学 | 一种基于飞机结构件的人机混合增强智能设计方法 |
CN112507840B (zh) * | 2020-12-02 | 2024-06-18 | 中国船舶集团有限公司第七一六研究所 | 一种人机混合增强的小目标检测和跟踪方法及*** |
-
2021
- 2021-06-09 CN CN202110644617.7A patent/CN113380033B/zh active Active
- 2021-06-29 WO PCT/CN2021/103202 patent/WO2022257201A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104978853A (zh) * | 2014-04-01 | 2015-10-14 | ***通信集团公司 | 一种道路交通安全评估方法及*** |
CN106228499A (zh) * | 2016-07-06 | 2016-12-14 | 东南大学 | 一种基于人‑车‑路‑货多风险源的货运安全评价模型 |
US20180107935A1 (en) * | 2016-10-18 | 2018-04-19 | Uber Technologies, Inc. | Predicting safety incidents using machine learning |
CN108182522A (zh) * | 2017-12-25 | 2018-06-19 | 中国电子科技集团公司第二十八研究所 | 一种基于ahp-熵权法的航道交通安全风险评估方法 |
CN109711691A (zh) * | 2018-12-17 | 2019-05-03 | 长安大学 | 一种基于熵权模糊综合评价模型的驾驶风格评价方法 |
CN111583639A (zh) * | 2020-04-30 | 2020-08-25 | 山东交通学院 | 一种道路交通拥堵预警方法及*** |
CN112836509A (zh) * | 2021-02-22 | 2021-05-25 | 西安交通大学 | 一种专家***知识库构建方法及*** |
Non-Patent Citations (1)
Title |
---|
ZHANG ZE-YANG;MA JIAN-WEN;SUN YING-YU;ZHANG AN-XI;LI GUANG-ZHENG: "Establishment of Evaluation Model for the Seafarers Fatigue Based on Improved Fuzzy Comprehensive Evaluation", JOURNAL OF QINGDAO OCEAN SHIPPING MARINERS COLLEGE, vol. 40, no. 2, 30 June 2019 (2019-06-30), pages 64 - 68, XP093012330, ISSN: 2095-3747 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115782905A (zh) * | 2023-01-31 | 2023-03-14 | 北京航空航天大学 | 一种自动驾驶车辆驾驶安全度量化*** |
CN116665453A (zh) * | 2023-06-25 | 2023-08-29 | 河南大学 | 基于二级模糊综合评价的高速公路事故严重程度预测方法 |
CN117744908A (zh) * | 2024-02-19 | 2024-03-22 | 深圳市深河环保水务有限公司 | 一种基于机器视觉的城市排水设施巡检方法及*** |
CN117744908B (zh) * | 2024-02-19 | 2024-04-19 | 深圳市深河环保水务有限公司 | 一种基于机器视觉的城市排水设施巡检方法及*** |
CN118036337A (zh) * | 2024-04-01 | 2024-05-14 | 山东凤麟工业装备有限公司 | 一种基于数字仿真的风电运输倾倒预警方法及*** |
CN118115341A (zh) * | 2024-04-30 | 2024-05-31 | 中亿丰数字科技集团股份有限公司 | 一种基于时空大数据模型的城市规建管服的方法及*** |
Also Published As
Publication number | Publication date |
---|---|
CN113380033B (zh) | 2022-08-16 |
CN113380033A (zh) | 2021-09-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022257201A1 (zh) | 基于人机混合增强智能的城市交通安全预警方法及*** | |
WO2021218235A1 (zh) | 一种道路交通拥堵预警方法及*** | |
Jiang et al. | Analysis of motorcycle accidents using association rule mining-based framework with parameter optimization and GIS technology | |
WO2021012504A1 (zh) | 基于智慧交通的道路信息提示方法、装置、服务器及介质 | |
CN104050809B (zh) | 交通状况自动分析预测方法 | |
CN113905215B (zh) | 一种公交安全驾驶监控*** | |
CN112382086A (zh) | 自动驾驶汽车开放道路测试评价***的使用方法 | |
CN115662113B (zh) | 一种信号交叉口人车博弈冲突风险评估与预警方法 | |
CN113033840B (zh) | 一种公路养护的判断方法和装置 | |
Kravets et al. | The traffic safety management system in urban conditions based on the C4. 5 algorithm | |
CN113887872A (zh) | 一种基于态势评估技术的交通风险分级方法 | |
CN105719510A (zh) | 车联网环境下道路交通事故链阻断***的效率评价方法 | |
Nasrollahi et al. | Identification and prioritization of connected vehicle technologies for sustainable development in Iran | |
Kim et al. | A comparative analysis of incident service time on urban freeways | |
Driss et al. | A fuzzy logic model for identifying spatial degrees of exposure to the risk of road accidents (Case study of the Wilaya of Mascara, Northwest of Algeria) | |
Sivabalaselvamani et al. | Accident identification using fuzzy cognitive maps with adaptive non-linear Hebbian learning algorithm | |
Zheng et al. | An approach to predict road accident frequencies: application of fuzzy neural network | |
CN105205312A (zh) | 道路事故热点成因分析和破坏度评估方法 | |
Ghoushchi et al. | Development of a robust hybrid framework for evaluating and ranking smartification measures for sustainable mobility: a case study of Sicilian roadways, Southern Italy | |
Abberley et al. | Modelling road congestion using a fuzzy system and real-world data for connected and autonomous vehicles | |
Yang et al. | Rapid analysis and detection algorithm and prevention countermeasures of urban traffic accidents under artificial intelligence | |
CN114897374A (zh) | 公路车辆行驶安全风险预测方法及*** | |
Hu et al. | Development of the real-time evaluation and decision support system for incident management | |
Michelaraki et al. | Modelling the Safety Tolerance Zone: Recommendations from the i-DREAMS project | |
Zenina et al. | Transport travel demand simulation model development for mixed-use building |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21944707 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21944707 Country of ref document: EP Kind code of ref document: A1 |