Intelligent alarm method and system based on scene recognition

Abstract

The invention discloses an intelligent alarm method and system based on scene recognition, and relates to the technical field of alarm, and the method comprises the following steps: recognizing a target management and control road section and a cooperative road section according to historical traffic event data; analyzing the time correlation between the target management and control road section and the cooperative road section according to the historical traffic event data, and performing judgment according to the time correlation coefficient to obtain a condition judgment result; wherein the target management and control road section and the cooperative road section are in a road network communication state, and the condition judgment result comprises a space-time synchronization condition or a non-space-time synchronization condition; analyzing according to the historical traffic event data to obtain spatial and temporal distribution characteristics of alarm events in the target controlled road section, and obtaining alarm weight ratios of different alarm sections according to the spatial and temporal distribution characteristics; the effect is that accurate alarm information can be timely issued to traffic participants.

Description

Technical Field

[0001] This invention relates to the field of alarm technology, and more specifically, to an intelligent alarm method and system based on scene recognition. Background Technology

[0002] In the field of traffic management, timely and accurate alerts are crucial for ensuring smooth and safe traffic flow. Traditional traffic alert methods often focus on localized events on a single road segment, failing to adequately consider the interrelationships between different road segments within the network. Furthermore, traditional methods struggle to identify these relationships in advance, resulting in delayed and limited alerts. This hinders effective spatiotemporal coordination, impacting the efficiency of traffic emergency response, potentially causing the spread of traffic congestion, inconveniencing road users, and ultimately increasing the difficulty of traffic management. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of this invention is to provide an intelligent alarm method and system based on scene recognition.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A scene recognition-based intelligent alarm method, comprising the following steps:

[0006] Identify target control sections and collaborative sections based on historical traffic incident data;

[0007] The temporal correlation between the target controlled road segment and the cooperating road segment is analyzed based on historical traffic event data, and the condition judgment result is obtained based on the temporal correlation coefficient; wherein, the target controlled road segment and the cooperating road segment are in a road network connected state, and the condition judgment result includes spatiotemporal synchronization condition or non-spatiotemporal synchronization condition.

[0008] Based on the analysis of historical traffic incident data, the spatiotemporal distribution characteristics of alarm events in the target controlled road section are obtained, and the alarm weight ratio of different alarm sections is obtained based on the spatiotemporal distribution characteristics.

[0009] Based on the condition judgment results, alarm weight ratio, historical alarm frequency, event response time and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained.

[0010] Real-time acquisition of current traffic event data for target controlled road sections and collaborative road sections; generation of dynamic adjustment parameters for emergency digital broadcasting scheme after calling the corresponding comprehensive alarm correlation coefficient based on the spatiotemporal correlation conditions of the current traffic event data; and output of spatiotemporal linkage alarm scheme based on the dynamic adjustment parameters.

[0011] Preferably, the temporal correlation between the target controlled road segment and the cooperating road segment is analyzed based on historical traffic incident data, specifically including the following steps:

[0012] Extract road segment event information from historical traffic incident data for target controlled road segments and cooperating road segments; wherein, the road segment event information includes event attribute tags, event duration, event impact coverage, and event occurrence time characteristics;

[0013] An event association graph is constructed by taking the occurrence time of event information of the target controlled road segment as the horizontal dimension and taking the occurrence time of event information of the collaborative road segment as the vertical dimension.

[0014] The temporal correlation coefficient between the target controlled road segment and the cooperative road segment is obtained by performing a sliding window weighted calculation on the event correlation graph within the dynamic time correlation interval.

[0015] Preferably, the conditional judgment result is obtained based on the time correlation coefficient, specifically including the following steps:

[0016] The time correlation coefficient is compared with a preset time correlation threshold;

[0017] If the time correlation coefficient is greater than or equal to the preset time correlation threshold, it is determined that the target controlled road segment and the cooperating road segment meet the spatiotemporal synchronization conditions.

[0018] If the time correlation coefficient is less than the preset time correlation threshold, it is determined that the target controlled road segment and the cooperating road segment meet the non-temporal synchronization condition.

[0019] Preferably, the spatiotemporal distribution characteristics of alarm events in the target controlled road segment are obtained by analyzing historical traffic incident data, specifically including the following steps:

[0020] Extract full information of alarm events for the target controlled road segment from historical traffic incident data; wherein, the full information of alarm events includes the location and nature of the alarm event;

[0021] A spatiotemporal feature matrix of alarm events is constructed by combining the physical structure characteristics and traffic flow operation characteristics of the target control road section. The occurrence location of the alarm event and the traffic flow operation characteristics are used as the core feature dimensions, and the nature of the event and the physical structure characteristics of the road section are used as auxiliary feature dimensions.

[0022] Based on the distribution patterns and impact of alarm events in historical traffic incident data, and combined with the traffic management needs of the target control road sections, a multi-dimensional judgment standard for section division is set, including alarm event density within a unit time period, frequency of overlapping alarm event impact ranges, percentage of duration of reduced traffic efficiency caused by the event, and number of consecutive occurrences of similar alarm events. Among them, alarm event density within a unit time period is used to initially judge the event density of the section, frequency of overlapping alarm event impact ranges is used to assess the clustering of event distribution, percentage of duration of reduced traffic efficiency caused by the event is used to measure the actual interference intensity of the event on the operation of the road section, and number of consecutive occurrences of similar alarm events is used to supplement the judgment of the correlation of events in the section.

[0023] By comparing the spatiotemporal feature matrix of alarm events with multi-dimensional judgment criteria, the concentrated alarm sections and dispersed alarm sections of alarm events in the target controlled road section are obtained.

[0024] Preferably, the alarm weight ratio for different alarm segments is obtained based on spatiotemporal distribution characteristics, specifically including the following steps:

[0025] Extract the historical comprehensive alarm frequency of the target control road section affected by spatiotemporal factors from historical traffic incident data;

[0026] The centralized alarm weight ratio is obtained by comparing the alarm frequency of the centralized alarm section and the comprehensive alarm frequency in the historical traffic incident data.

[0027] The weighting ratio of dispersed alarms is obtained by comparing the alarm frequency of dispersed alarm sections with the overall alarm frequency in historical traffic incident data.

[0028] The combination of the centralized alarm weight ratio and the decentralized alarm weight ratio is called the alarm weight ratio.

[0029] Preferably, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained based on the condition judgment result, alarm weight ratio, historical alarm frequency, event response time, and impact range. This specifically includes the following steps:

[0030] The first critical node is obtained by marking high-frequency congestion areas and high-frequency accident areas in the target control road section as risk nodes;

[0031] The second key node is obtained by marking high-frequency congestion areas and high-frequency accident areas in the target control road section with traffic diversion nodes.

[0032] Select the node to be triggered from the first critical node and the second critical node, and obtain the first alarm correlation coefficient by statistically analyzing the alarm response time, coverage area and historical comprehensive alarm frequency of the node to be triggered based on the spatiotemporal synchronization conditions.

[0033] Under non-spatiotemporal synchronization conditions, the event time interval between the target controlled road segment and the cooperating road segment is determined. Based on the event time interval, the historical comprehensive alarm frequency and the event impact range between the target controlled road segment and the cooperating road segment are processed to obtain the second alarm correlation coefficient and the third alarm correlation coefficient.

[0034] The combination of the first alarm correlation coefficient, the second alarm correlation coefficient, and the third alarm correlation coefficient constitutes the comprehensive alarm correlation coefficient.

[0035] Preferably, the first alarm correlation coefficient is obtained by statistically analyzing the alarm response time, coverage area, alarm weight ratio, and historical comprehensive alarm frequency of the node to be triggered based on the spatiotemporal synchronization conditions. This specifically includes the following steps:

[0036] Based on the spatiotemporal synchronization conditions, the first response time is obtained by statistically analyzing the alarm response time of the target controlled road segment that is greater than or equal to the event impact threshold in the nodes to be triggered.

[0037] The first historical alarm coverage range is obtained by statistically analyzing the target control road segment and the alarm coverage range of the node to be triggered that is greater than or equal to the event impact threshold.

[0038] The first historical alarm coverage ratio is obtained by comparing the coverage area of ​​the first historical alarm with the total traffic range of the target controlled road section.

[0039] The first historical alarm coverage ratio and alarm weight ratio are summed to obtain the first alarm feature data;

[0040] Extract the historical comprehensive alarm frequency of the target controlled road section from historical traffic incident data;

[0041] The correlation coefficient of the first alarm is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the first response time, and the first alarm characteristic data.

[0042] Preferably, under non-spatiotemporal synchronization conditions, the event time interval between the target controlled road segment and the cooperating road segment is determined. Based on the event time interval, the historical comprehensive alarm frequency and the event impact range between the target controlled road segment and the cooperating road segment are processed to obtain the second alarm correlation coefficient and the third alarm correlation coefficient. Specifically, the following steps are included:

[0043] Under non-spatiotemporal synchronization conditions, the time interval between historical events between the target controlled road segment and the cooperating road segment is statistically analyzed;

[0044] If the historical event time interval is greater than or equal to the preset event time interval threshold, the second response time is obtained by statistically analyzing the alarm response time of the target controlled road segment that is greater than or equal to the event impact threshold in the node to be triggered, based on the non-spatiotemporal synchronization condition.

[0045] The second historical alarm coverage range is obtained by statistically analyzing the target control road segment and the alarm coverage range of the node to be triggered that is greater than or equal to the event impact threshold.

[0046] The second historical alarm coverage ratio is obtained by comparing the coverage area of ​​the second historical alarm with the total traffic range of the target controlled road section.

[0047] The second historical alarm coverage ratio and alarm weight ratio are summed to obtain the second alarm feature data;

[0048] The second alarm correlation coefficient is obtained based on the correlation characteristics between historical comprehensive alarm frequency, second response duration, and second alarm characteristic data.

[0049] If the historical event time interval is less than the preset event time interval threshold, the difference in the event impact range between the target controlled road segment and the cooperative road segment is calculated to obtain the impact range difference value; the third alarm correlation coefficient is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the impact range difference value and the historical event time interval.

[0050] An intelligent alarm system based on scene recognition includes:

[0051] Identification module: Identifies target controlled road sections and cooperative road sections based on historical traffic incident data;

[0052] Judgment Module: Analyzes the temporal correlation between the target controlled road segment and the cooperating road segment based on historical traffic event data, and makes a judgment based on the temporal correlation coefficient to obtain the condition judgment result; wherein, the target controlled road segment and the cooperating road segment are in a road network connected state, and the condition judgment result includes spatiotemporal synchronization conditions or non-spatiotemporal synchronization conditions;

[0053] Analysis module: Based on historical traffic incident data, the spatiotemporal distribution characteristics of alarm events in the target controlled road section are obtained, and the alarm weight ratio of different alarm sections is obtained based on the spatiotemporal distribution characteristics;

[0054] Processing module: Based on the condition judgment results, alarm weight ratio, historical alarm frequency, event response time and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained;

[0055] Output module: Real-time acquisition of current traffic event data for target controlled road sections and collaborative road sections; based on the spatiotemporal correlation conditions of the current traffic event data, calling the corresponding comprehensive alarm correlation coefficient to generate dynamic adjustment parameters for the emergency digital broadcasting scheme; and outputting the spatiotemporal linkage alarm scheme based on the dynamic adjustment parameters.

[0056] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement an intelligent alarm method based on scene recognition.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] This invention identifies target control road sections and cooperating road sections using historical traffic event data, pinpointing road areas requiring key attention, making traffic management more targeted and avoiding resource waste from indiscriminate monitoring. It analyzes the temporal correlation between target control road sections and cooperating road sections and determines conditions, distinguishing between spatiotemporally synchronized and non-spatially synchronized situations. This allows alarms to adapt to different temporal correlation scenarios, improving accuracy and timeliness. For example, in spatiotemporally synchronized situations, alarms are quickly linked to relevant road sections, while in non-spatially synchronized situations, alarm strategies can be rationally adjusted based on factors such as time intervals. Based on historical data, the spatiotemporal distribution characteristics of alarm events on target control road sections and the alarm weight ratios of different sections are obtained, improving the efficiency of traffic event handling and reducing traffic congestion caused by untimely or inadequate alarms in key areas. Based on multiple factors, a comprehensive alarm correlation coefficient applicable to spatiotemporally correlated scenarios is obtained. By comprehensively considering condition judgment results, weight ratios, historical frequency, response time, and impact range, the alarm correlation coefficient is more comprehensive and accurate, reflecting the degree of correlation of traffic events and providing a strong basis for the development of emergency intelligent broadcasting solutions. It can acquire real-time traffic event data and dynamically generate spatiotemporal linkage alarm schemes, realize dynamic adjustment of alarms, quickly respond to real-time changes in traffic events, and promptly release accurate alarm information to traffic participants, thereby improving the smoothness and safety of overall traffic operation. Attached Figure Description

[0059] Figure 1 This is a schematic diagram illustrating the steps of an intelligent alarm method based on scene recognition proposed in this invention.

[0060] Figure 2 This invention presents a schematic diagram of a scene recognition-based intelligent alarm system.

[0061] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.

[0062] 610. Processor; 620. Communication interface; 630. Memory; 640. Communication bus. Detailed Implementation

[0063] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0064] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0065] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0066] Reference Figures 1-3 As shown.

[0067] The embodiments further illustrate the intelligent alarm method and system based on scene recognition proposed in this invention.

[0068] A scene recognition-based intelligent alarm method, comprising the following steps:

[0069] Identify target control sections and collaborative sections based on historical traffic incident data;

[0070] Based on historical traffic event data analysis, the temporal correlation between the target controlled road segment and the cooperating road segment is determined, and the condition judgment result is obtained based on the temporal correlation coefficient. Among them, the target controlled road segment and the cooperating road segment are in a road network connected state, and the condition judgment result includes spatiotemporal synchronization condition or non-spatiotemporal synchronization condition.

[0071] Based on the analysis of historical traffic incident data, the spatiotemporal distribution characteristics of alarm events in the target controlled road section are obtained, and the alarm weight ratio of different alarm sections is obtained based on the spatiotemporal distribution characteristics.

[0072] Based on the condition judgment results, alarm weight ratio, historical alarm frequency, event response time and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained.

[0073] Real-time acquisition of current traffic event data for target controlled road sections and collaborative road sections; generation of dynamic adjustment parameters for emergency digital broadcasting scheme after calling the corresponding comprehensive alarm correlation coefficient based on the spatiotemporal correlation conditions of the current traffic event data; and output of spatiotemporal linkage alarm scheme based on the dynamic adjustment parameters.

[0074] First, historical traffic incident data is used to identify target control road sections requiring key attention, and simultaneously, connecting and related cooperating road sections within the road network are determined. Next, the historical traffic incident data is analyzed to explore the temporal correlation between the target control road sections and cooperating road sections, determining whether they are under spatiotemporally synchronized or not by calculating a temporal correlation coefficient. The spatiotemporal distribution characteristics of alarm events on the target control road sections in the historical traffic incident data are assessed, and alarm weight ratios for different alarm sections are determined based on these characteristics. Based on this, and combining the condition judgment results, alarm weight ratios, and factors such as historical alarm frequency, event response time, and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporally correlated scenarios is calculated. Real-time traffic incident data for the target control road sections and cooperating road sections is collected. Based on the spatiotemporal correlation conditions corresponding to the current data, the appropriate comprehensive alarm correlation coefficient is invoked to generate dynamic adjustment parameters for the emergency intelligent broadcasting scheme. The spatiotemporally linked alarm scheme is then output according to these parameters. When IoT sensing modules such as cameras, radars, and sensors on the road detect traffic events, they transmit the event information to the speaker columns. The speaker columns then directly retrieve and play pre-stored digital broadcast content that matches the event. Alternatively, the digital broadcasting center's platform can issue event program points, allowing the speaker columns to play the corresponding digital broadcast content, thereby promptly issuing warning information to traffic participants and helping to improve the efficiency and safety of traffic management.

[0075] Analyzing the temporal correlation between target control road sections and cooperating road sections based on historical traffic incident data includes the following steps:

[0076] Extract road segment event information from historical traffic incident data for target controlled road segments and cooperating road segments; the road segment event information includes event attribute tags, event duration, event impact coverage, and event occurrence time characteristics;

[0077] An event association graph is constructed by taking the occurrence time of event information of the target controlled road segment as the horizontal dimension and taking the occurrence time of event information of the collaborative road segment as the vertical dimension.

[0078] The temporal correlation coefficient between the target controlled road segment and the cooperative road segment is obtained by performing a sliding window weighted calculation on the event correlation graph within the dynamic time correlation interval.

[0079] First, road segment event information for the target controlled road segment and its cooperating road segments is extracted from historical traffic event data. This information includes event attribute labels, event duration, event impact coverage, and event occurrence time characteristics. An event correlation graph is constructed by using the occurrence time sequence of event information for the target controlled road segment as the horizontal dimension and the occurrence time sequence of event information for the cooperating road segment as the vertical dimension. A sliding window weighted calculation is then performed on the event correlation graph within a dynamic time correlation interval to obtain the time correlation coefficient between the target controlled road segment and the cooperating road segment.

[0080] Once a strong temporal correlation is established between a target controlled road segment and a cooperating road segment, if the sensing devices of the cameras on the road detect a road flooding event caused by heavy rain on the target controlled road segment, this event information will be transmitted to nearby loudspeakers. The loudspeakers will then play corresponding intelligent broadcast content, informing drivers that traffic is slow ahead due to flooding and suggesting that they choose to travel on the cooperating road segment based on the road segment correlation reflected by the temporal correlation coefficient. This achieves timely and effective traffic warnings, ensuring road traffic efficiency and safety.

[0081] The judgment result is obtained based on the time correlation coefficient, which includes the following steps:

[0082] The time correlation coefficient is compared with a preset time correlation threshold;

[0083] If the time correlation coefficient is greater than or equal to the preset time correlation threshold, it is determined that the target controlled road segment and the cooperating road segment meet the spatiotemporal synchronization conditions.

[0084] If the time correlation coefficient is less than the preset time correlation threshold, it is determined that the target controlled road segment and the cooperating road segment meet the non-temporal synchronization condition.

[0085] First, the time correlation coefficient is compared with a preset time correlation threshold. If the time correlation coefficient is greater than or equal to the preset time correlation threshold, the target controlled road segment and the cooperating road segment are determined to meet the spatiotemporal synchronization condition; if the time correlation coefficient is less than the preset time correlation threshold, the target controlled road segment and the cooperating road segment are determined to meet the non-spatiotemporal synchronization condition.

[0086] Assuming a temporal correlation coefficient between a target controlled road segment and its cooperating road segments, this coefficient is compared with a preset threshold to determine if they meet the spatiotemporal synchronization condition. If a serious traffic accident occurs on the target controlled road segment, the road sensing equipment detects the event and transmits the information to the audio column. The audio column then plays corresponding intelligent broadcast content, informing drivers on the target controlled road segment of the accident and congestion information, and also reminding drivers on the cooperating road segment to plan their routes in advance based on the spatiotemporal synchronization condition, avoiding entering the affected area. This achieves accurate and timely traffic warnings, effectively managing traffic and improving road safety and smoothness.

[0087] Based on historical traffic incident data analysis, the spatiotemporal distribution characteristics of alarm events in the target controlled road section are obtained, specifically including the following steps:

[0088] Extract full information of alarm events for the target controlled road segment from historical traffic incident data; the full information of alarm events includes the location and nature of the alarm event.

[0089] A spatiotemporal feature matrix of alarm events is constructed by combining the physical structure characteristics and traffic flow operation characteristics of the target control road section. The occurrence location of the alarm event and the traffic flow operation characteristics are used as the core feature dimensions, and the nature of the event and the physical structure characteristics of the road section are used as auxiliary feature dimensions.

[0090] Based on the distribution patterns and impact of alarm events in historical traffic incident data, and combined with the traffic management needs of the target control road sections, a multi-dimensional judgment standard for section division is set, including alarm event density within a unit time period, frequency of overlapping alarm event impact ranges, percentage of duration of reduced traffic efficiency caused by the event, and number of consecutive occurrences of similar alarm events. Among them, alarm event density within a unit time period is used to initially judge the event density of the section, frequency of overlapping alarm event impact ranges is used to assess the clustering of event distribution, percentage of duration of reduced traffic efficiency caused by the event is used to measure the actual interference intensity of the event on the operation of the road section, and number of consecutive occurrences of similar alarm events is used to supplement the judgment of the correlation of events in the section.

[0091] By comparing the spatiotemporal feature matrix of alarm events with multi-dimensional judgment criteria, the concentrated alarm sections and dispersed alarm sections of alarm events in the target controlled road section are obtained.

[0092] First, all alarm event information for the target controlled road segment is extracted from historical traffic incident data. This information includes the location and nature of the alarm events. Then, a spatiotemporal feature matrix of alarm events is constructed by combining the physical structure and traffic flow characteristics of the target controlled road segment. The location and traffic flow characteristics are used as the core feature dimensions, while the event nature and the physical structure characteristics of the road segment are used as auxiliary feature dimensions. Next, based on the distribution patterns and impact of alarm events in historical traffic incident data, and combined with the traffic management needs of the target controlled road segment, multi-dimensional criteria for segment division are established. These criteria include alarm event density within a unit time period, frequency of overlapping alarm event impact areas, percentage of duration of reduced traffic efficiency caused by the event, and number of consecutive occurrences of similar alarm events. Specifically, alarm event density within a unit time period is used to initially determine the event density of the segment; frequency of overlapping alarm event impact areas is used to assess the clustering of event distribution; percentage of duration of reduced traffic efficiency caused by the event is used to measure the actual interference intensity of the event on the road segment's operation; and number of consecutive occurrences of similar alarm events is used to supplement the determination of the correlation between events in the segment. Then, the spatiotemporal feature matrix of alarm events is compared with the multi-dimensional judgment criteria to obtain the concentrated alarm sections and dispersed alarm sections of alarm events in the target controlled road section.

[0093] If the above steps determine that a specific controlled road segment has a concentrated alarm zone, where accident-related alarm events occur frequently and their impact overlaps within a given time period, then when another traffic accident occurs in this concentrated alarm zone, the road sensing equipment will detect the event and transmit the information to the loudspeakers. The loudspeakers will then broadcast corresponding intelligent information, clearly informing drivers that an accident has occurred in the concentrated alarm zone, resulting in severe congestion, and suggesting that they detour through dispersed alarm zones or other alternative routes. This achieves precise alarms for different sections, effectively guiding traffic flow and improving the overall traffic efficiency of the road segment.

[0094] The alarm weight ratio for different alarm segments is obtained based on the spatiotemporal distribution characteristics, specifically including the following steps:

[0095] Extract the historical comprehensive alarm frequency of the target control road section affected by spatiotemporal factors from historical traffic incident data;

[0096] The centralized alarm weight ratio is obtained by comparing the alarm frequency of the centralized alarm section and the comprehensive alarm frequency in the historical traffic incident data.

[0097] The weighting ratio of dispersed alarms is obtained by comparing the alarm frequency of dispersed alarm sections with the overall alarm frequency in historical traffic incident data.

[0098] The alarm weight ratio is the combination of the centralized alarm weight ratio and the decentralized alarm weight ratio.

[0099] First, the historical comprehensive alarm frequency of the target control road segment affected by spatiotemporal factors is extracted from historical traffic incident data. Then, the alarm frequency of concentrated alarm segments and the comprehensive alarm frequency in the historical traffic incident data are compared to obtain the concentrated alarm weight ratio. At the same time, the alarm frequency of dispersed alarm segments and the comprehensive alarm frequency in the historical traffic incident data are compared to obtain the dispersed alarm weight ratio. The combined concentrated alarm weight ratio and the dispersed alarm weight ratio is the alarm weight ratio.

[0100] Assuming a target traffic control section has a historical total alarm frequency of 100 times, with 70 alarms occurring in concentrated alarm areas (70 times) and 30 alarms occurring in dispersed alarm areas (30 times) (30% weighting), when a traffic congestion event occurs in a concentrated alarm area, the road sensing equipment detects it and transmits the information to the loudspeakers. Based on the 70% concentrated alarm weighting, the loudspeakers prioritize and broadcast congestion information for that concentrated area more frequently, clearly informing drivers that the area is severely congested and suggesting alternative routes. For alarms in dispersed alarm areas, the broadcast frequency and intensity are relatively weaker. This achieves differentiated and precise alarms based on weighting, effectively improving the targeting and efficiency of traffic management.

[0101] Based on the condition judgment results, alarm weight ratio, historical alarm frequency, event response time, and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained, specifically including the following steps:

[0102] The first critical node is obtained by marking high-frequency congestion areas and high-frequency accident areas in the target control road section as risk nodes;

[0103] The second key node is obtained by marking high-frequency congestion areas and high-frequency accident areas in the target control road section with traffic diversion nodes.

[0104] Select the node to be triggered from the first critical node and the second critical node, and obtain the first alarm correlation coefficient by statistically analyzing the alarm response time, coverage area and historical comprehensive alarm frequency of the node to be triggered based on the spatiotemporal synchronization conditions.

[0105] Under non-spatiotemporal synchronization conditions, the event time interval between the target controlled road segment and the cooperating road segment is determined. Based on the event time interval, the historical comprehensive alarm frequency and the event impact range between the target controlled road segment and the cooperating road segment are processed to obtain the second alarm correlation coefficient and the third alarm correlation coefficient.

[0106] The combination of the first alarm correlation coefficient, the second alarm correlation coefficient, and the third alarm correlation coefficient constitutes the comprehensive alarm correlation coefficient.

[0107] First, high-frequency congestion areas and high-frequency accident areas within the target controlled road segment are marked as risk nodes, i.e., the first critical nodes; simultaneously, these areas are marked as traffic management nodes, i.e., the second critical nodes. Then, triggerable nodes are selected from the first and second critical nodes. Under spatiotemporal synchronization conditions, the alarm response duration, coverage area, and historical comprehensive alarm frequency of the triggerable nodes are statistically analyzed to obtain the first alarm correlation coefficient. Under non-spatiotemporal synchronization conditions, the event time interval between the target controlled road segment and the cooperating road segment is determined. Based on this time interval, the historical comprehensive alarm frequency and the event impact range between the target controlled road segment and the cooperating road segment are processed to obtain the second and third alarm correlation coefficients. The combination of the first, second, and third alarm correlation coefficients constitutes the comprehensive alarm correlation coefficient.

[0108] For example, high-frequency congestion areas on a target controlled road segment are marked as the first and second critical nodes. Nodes selected for triggering are identified based on their short alarm response time, wide coverage, and high historical alarm frequency under spatiotemporal synchronization conditions, yielding a corresponding first alarm correlation coefficient. When a congestion event occurs at this node, the road sensing device detects it and transmits the information to the audio column. The audio column then plays corresponding intelligent broadcast content based on the first alarm correlation coefficient, clearly informing drivers of the congestion situation in the area and suggesting that they be diverted to traffic control nodes as soon as possible. If the conditions are not spatiotemporally synchronized, and factors such as the event time interval between the target controlled road segment and the cooperating road segment are processed to obtain corresponding second or third alarm correlation coefficients, the audio column will also play intelligent broadcast content suitable for the scenario based on these correlation coefficients, such as reminding drivers to pay attention to the potential impact on the cooperating road segment. This achieves accurate alarms based on different correlation coefficients, improving the timeliness and effectiveness of traffic incident handling.

[0109] Based on the spatiotemporal synchronization conditions, the alarm response time, coverage area, alarm weight ratio, and historical comprehensive alarm frequency of the node to be triggered are statistically analyzed to obtain the first alarm correlation coefficient. The specific steps include:

[0110] Based on the spatiotemporal synchronization conditions, the first response time is obtained by statistically analyzing the alarm response time of the target controlled road segment that is greater than or equal to the event impact threshold in the nodes to be triggered.

[0111] The first historical alarm coverage range is obtained by statistically analyzing the target control road segment and the alarm coverage range of the node to be triggered that is greater than or equal to the event impact threshold.

[0112] The first historical alarm coverage ratio is obtained by comparing the coverage area of ​​the first historical alarm with the total traffic range of the target controlled road section.

[0113] The first historical alarm coverage ratio and alarm weight ratio are summed to obtain the first alarm feature data;

[0114] Extract the historical comprehensive alarm frequency of the target controlled road section from historical traffic incident data;

[0115] The correlation coefficient of the first alarm is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the first response time, and the first alarm characteristic data.

[0116] For example, under spatiotemporal synchronization, the alarm response duration exceeding the event impact threshold for a target controlled road segment is statistically determined as the first response duration, and the corresponding alarm coverage area is determined as the first historical alarm coverage area. This first historical alarm coverage ratio is obtained by comparing it to the total traffic area, and then summing this ratio with the alarm weight ratio to obtain the first alarm feature data. Simultaneously, the historical comprehensive alarm frequency of this road segment is extracted, and the first alarm correlation coefficient is obtained based on this data. When a traffic event occurs at this target node, the road sensing device detects it and transmits the information to the audio column. The audio column then plays corresponding intelligent broadcast content based on the first alarm correlation coefficient, such as detailed information about the event's impact duration, coverage area, and the weight information of road segments requiring attention. This allows drivers to clearly understand the road conditions and make reasonable travel decisions, thereby achieving accurate traffic alerts and improving the efficiency and safety of traffic management.

[0117] Under non-spatiotemporal synchronization conditions, the event time interval between the target controlled road segment and the cooperating road segment is determined. Based on the event time interval, the historical comprehensive alarm frequency and the event impact range between the target controlled road segment and the cooperating road segment are processed to obtain the second alarm correlation coefficient and the third alarm correlation coefficient. The specific steps include:

[0118] Under non-spatiotemporal synchronization conditions, the time interval between historical events between the target controlled road segment and the cooperating road segment is statistically analyzed;

[0119] If the historical event time interval is greater than or equal to the preset event time interval threshold, the second response time is obtained by statistically analyzing the alarm response time of the target controlled road segment that is greater than or equal to the event impact threshold in the node to be triggered, based on the non-spatiotemporal synchronization condition.

[0120] The second historical alarm coverage range is obtained by statistically analyzing the target control road segment and the alarm coverage range of the node to be triggered that is greater than or equal to the event impact threshold.

[0121] The second historical alarm coverage ratio is obtained by comparing the coverage area of ​​the second historical alarm with the total traffic range of the target controlled road section.

[0122] The second historical alarm coverage ratio and alarm weight ratio are summed to obtain the second alarm feature data;

[0123] The second alarm correlation coefficient is obtained based on the correlation characteristics between historical comprehensive alarm frequency, second response duration, and second alarm characteristic data.

[0124] If the historical event time interval is less than the preset event time interval threshold, the difference in the event impact range between the target controlled road segment and the cooperative road segment is calculated to obtain the impact range difference value; the third alarm correlation coefficient is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the impact range difference value and the historical event time interval.

[0125] Under non-spatiotemporal synchronization conditions, historical event time intervals between the target controlled road segment and the cooperating road segment are statistically analyzed. If the historical event time interval is greater than or equal to a preset event time interval threshold, the alarm response time of the target controlled road segment in the node to be triggered that is greater than or equal to the event impact threshold is statistically analyzed to obtain a second response time. Based on the second response time, the alarm coverage area of ​​the target controlled road segment in the node to be triggered that is greater than or equal to the event impact threshold is statistically analyzed to obtain a second historical alarm coverage area. Then, the ratio of the second historical alarm coverage area to the total traffic area of ​​the target controlled road segment is calculated to obtain a second historical alarm coverage ratio. The second historical alarm coverage ratio and the alarm weight ratio are summed to obtain second alarm feature data. Finally, the second alarm correlation coefficient is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the second response time, and the second alarm feature data.

[0126] If the historical event time interval is less than the preset event time interval threshold, the difference between the event impact range of the target controlled road segment and the cooperative road segment is calculated to obtain the impact range difference value; the third alarm correlation coefficient is obtained based on the correlation characteristics between the historical comprehensive alarm frequency, the impact range difference value and the historical event time interval.

[0127] For example, if the target controlled road segment and the cooperating road segment are not in a spatiotemporally synchronized state, and the time interval between historical events is greater than a preset threshold, the second alarm correlation coefficient is obtained through the above steps. If a traffic event occurs at the trigger node of the target controlled road segment, the road sensing device will detect it and transmit the information to the speaker. The speaker will then play corresponding intelligent broadcast content based on the second alarm correlation coefficient, informing the driver of the event's response time, coverage area, and other information, allowing the driver to be aware of road conditions and plan their trip accordingly. If the time interval between historical events is less than the preset threshold, and a third alarm correlation coefficient is obtained, the speaker will play intelligent broadcast content combining the differences in the event's impact range and the time interval. For example, it will remind the driver to pay attention to the differences in the impact of events on the target controlled road segment and the cooperating road segment, allowing them to prepare in advance. This achieves accurate alarms in different spatiotemporally synchronized scenarios, improving the effectiveness of traffic management.

[0128] An intelligent alarm system based on scene recognition includes:

[0129] Identification module: Identifies target controlled road sections and cooperative road sections based on historical traffic incident data;

[0130] Judgment Module: Analyzes the temporal correlation between the target controlled road segment and the cooperating road segment based on historical traffic event data, and makes judgments based on the temporal correlation coefficient to obtain the condition judgment result; wherein, the target controlled road segment and the cooperating road segment are in a road network connected state, and the condition judgment result includes spatiotemporal synchronization condition or non-spatiotemporal synchronization condition;

[0131] Analysis module: Based on historical traffic incident data, the spatiotemporal distribution characteristics of alarm events in the target controlled road section are obtained, and the alarm weight ratio of different alarm sections is obtained based on the spatiotemporal distribution characteristics;

[0132] Processing module: Based on the condition judgment results, alarm weight ratio, historical alarm frequency, event response time and impact range, a comprehensive alarm correlation coefficient applicable to spatiotemporal correlation scenarios is obtained;

[0133] Output module: Real-time acquisition of current traffic event data for target controlled road sections and collaborative road sections; based on the spatiotemporal correlation conditions of the current traffic event data, calling the corresponding comprehensive alarm correlation coefficient to generate dynamic adjustment parameters for the emergency digital broadcasting scheme; and outputting the spatiotemporal linkage alarm scheme based on the dynamic adjustment parameters.

[0134] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements an intelligent alarm method based on scene recognition.

[0135] like Figure 3 As shown, the electronic device may include a processor 610, a communication interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communication interface 620, and the memory 630 communicate with each other through the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute an intelligent alarm method based on scene recognition.

[0136] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0137] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program that can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to execute an intelligent alarm method based on scene recognition.

[0138] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform an intelligent alarm method based on scene recognition.

[0139] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0140] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A scene recognition based intelligent alarming method, characterized in that, The method comprises the following steps: According to the historical traffic event data, the target control section and the cooperative section are identified; According to the historical traffic event data, the time correlation of the target control section and the cooperative section is analyzed, and the condition judgment result is obtained according to the time correlation coefficient; wherein the target control section and the cooperative section are the connected state of the road network, and the condition judgment result includes the space-time synchronization condition or the non-space-time synchronization condition; According to the historical traffic event data, the space-time distribution characteristics of the warning events in the target control section are analyzed, and the warning weight ratio of different warning sections is obtained according to the space-time distribution characteristics; Based on the condition judgment result, the warning weight ratio, the historical warning frequency, the event response time and the influence range, the comprehensive warning correlation coefficient suitable for the space-time correlation scene is obtained; Real-time acquisition of the current traffic event data of the target control section and the cooperative section, calling the corresponding comprehensive warning correlation coefficient according to the space-time correlation condition of the current traffic event data to generate the dynamic adjustment parameter of the emergency intelligent broadcast scheme, and outputting the space-time linkage warning scheme according to the dynamic adjustment parameter. 2.The intelligent alarm method based on scene recognition of claim 1, wherein, According to the historical traffic event data, the time correlation of the target control section and the cooperative section is analyzed, and the condition judgment result is obtained according to the time correlation coefficient; wherein the target control section and the cooperative section are the connected state of the road network, and the condition judgment result includes the space-time synchronization condition or the non-space-time synchronization condition. Extract the road section event information of the target control section and the cooperative section in the historical traffic event data; wherein the road section event information contains event attribute label, event duration period, event influence coverage range and event occurrence time period characteristics; After taking the occurrence time sequence of the road section event information of the target control section as the horizontal dimension and the occurrence time sequence of the road section event information of the cooperative section as the vertical dimension, an event correlation graph is built; In the dynamic time correlation interval, the event correlation graph is subjected to sliding window weighted operation to obtain the time correlation coefficient of the target control section and the cooperative section. 3.The intelligent alarm method based on scene recognition of claim 1, wherein, According to the time correlation coefficient, the condition judgment result is obtained by judging, which comprises the following steps: Compare the time correlation coefficient with the preset time correlation threshold value; If the time correlation coefficient is greater than or equal to the preset time correlation threshold value, it is determined that the target control section and the cooperative section meet the space-time synchronization condition; If the time correlation coefficient is less than the preset time correlation threshold value, it is determined that the target control section and the cooperative section meet the non-space-time synchronization condition.

4. The intelligent alerting method based on scene recognition according to claim 1, characterized in that, According to the historical traffic event data, the space-time distribution characteristics of the warning events in the target control section are analyzed, and the warning weight ratio of different warning sections is obtained according to the space-time distribution characteristics; comprising the following steps: Extract the full information of the warning events in the target control section in the historical traffic event data; wherein the full information of the warning events contains the occurrence point of the warning events and the event nature; Combined with the physical structure characteristics and traffic flow running characteristics of the target control section, a warning event space-time feature matrix is built, the occurrence point of the warning events and the traffic flow running characteristics are taken as the core feature dimension, and the event nature and the road section physical structure characteristics are taken as the auxiliary feature dimension; The multi-dimensional judgment criteria for section division are set according to the distribution law and influence degree of the alarm events in the historical traffic event data and the traffic management demand of the target control section, including alarm event occurrence density in a unit period, alarm event influence range superposition frequency, event caused traffic efficiency decline duration proportion, and continuous occurrence number of the same alarm event; wherein the alarm event occurrence density in a unit period is used to preliminarily judge the section event density, the alarm event influence range superposition frequency is used to evaluate the event distribution aggregation, the event caused traffic efficiency decline duration proportion is used to measure the actual interference strength of the event on the section operation, and the continuous occurrence number of the same alarm event is used to supplement the correlation of the section event. The concentrated alarm section and the dispersed alarm section of the alarm events in the target control section are obtained by comparing the alarm event space-time feature matrix with the multi-dimensional judgment criteria.

5. The intelligent alerting method based on scene recognition according to claim 4, characterized in that, The alarm weight ratio of different alarm sections is obtained according to the space-time distribution characteristics, specifically including the following steps: The historical comprehensive alarm frequency of the target control section affected by the space-time factors is extracted from the historical traffic event data; The concentrated alarm weight ratio is obtained by comparing the alarm frequency of the concentrated alarm section and the comprehensive alarm frequency in the historical traffic event data; The dispersed alarm weight ratio is obtained by comparing the alarm frequency of the dispersed alarm section and the comprehensive alarm frequency in the historical traffic event data; The concentrated alarm weight ratio and the dispersed alarm weight ratio are combined as the alarm weight ratio.

6. The intelligent alerting method based on scene recognition according to claim 1, characterized in that, The comprehensive alarm correlation coefficient suitable for the space-time correlation scenario is obtained based on the conditional judgment result, the alarm weight ratio, the historical alarm frequency, the event response duration and the influence range, specifically including the following steps: The first key node is obtained by marking the risk nodes of the high-frequency congestion area and the high-frequency accident area in the target control section; The second key node is obtained by marking the dredging nodes of the high-frequency congestion area and the high-frequency accident area in the target control section; The first alarm correlation coefficient is obtained by selecting the to-be-triggered node from the first key node and the second key node, and according to the space-time synchronization condition, the alarm response duration, the coverage range and the historical comprehensive alarm frequency of the to-be-triggered node are counted; The event time interval between the target control section and the cooperative section is judged under the non-space-time synchronization condition, and the historical comprehensive alarm frequency and the event influence range between the target control section and the cooperative section are processed based on the event time interval to obtain the second alarm correlation coefficient and the third alarm correlation coefficient; The first alarm correlation coefficient, the second alarm correlation coefficient and the third alarm correlation coefficient are combined as the comprehensive alarm correlation coefficient.

7. The intelligent alarm method based on scene recognition according to claim 6, characterized in that, The first alarm correlation coefficient is obtained by counting the alarm response duration, the coverage range, the alarm weight ratio and the historical comprehensive alarm frequency of the to-be-triggered node according to the space-time synchronization condition, specifically including the following steps: The first response duration is counted according to the space-time synchronization condition, that is, the alarm response duration of the target control section in the to-be-triggered node which is greater than or equal to the event influence threshold value is counted; The first historical alarm coverage range is obtained by counting the alarm coverage range of the target control section in the to-be-triggered node which is greater than or equal to the event influence threshold value according to the first response duration; The first alarm correlation coefficient is obtained by counting the alarm response duration, the coverage range, the alarm weight ratio and the historical comprehensive alarm frequency of the to-be-triggered node according to the space-time synchronization condition, specifically including the following steps: The first historical alarm coverage ratio is obtained by performing ratio processing on the total passing range of the target management road section and the first historical alarm coverage range. The first alarm feature data is obtained by performing sum processing on the first historical alarm coverage ratio and the alarm weight ratio. The historical comprehensive alarm frequency of the target management road section is extracted from the historical traffic event data. The first alarm correlation coefficient is obtained according to the correlation characteristics among the historical comprehensive alarm frequency, the first response time length and the first alarm feature data. 8.The intelligent alarm method based on scene recognition of claim 7, wherein, The event time interval between the target management road section and the collaborative road section is judged under the non-time-space synchronization condition, and the historical comprehensive alarm frequency and the event influence range between the target management road section and the collaborative road section are processed based on the event time interval to obtain the second alarm correlation coefficient and the third alarm correlation coefficient, which specifically include the following steps: The historical event time interval between the target management road section and the collaborative road section is counted under the non-time-space synchronization condition. If the historical event time interval is greater than or equal to the preset event time interval threshold, the second response time length is counted according to the non-time-space synchronization condition, which is greater than or equal to the event influence threshold belonging to the alarm response time length of the target management road section in the to-be-triggered node. The second historical alarm coverage range is obtained according to the second response time length, which is greater than or equal to the event influence threshold belonging to the alarm coverage range of the target management road section in the to-be-triggered node. The second historical alarm coverage ratio is obtained by performing ratio processing on the total passing range of the target management road section and the second historical alarm coverage range. The second alarm feature data is obtained by performing sum processing on the second historical alarm coverage ratio and the alarm weight ratio. The second alarm correlation coefficient is obtained according to the correlation characteristics among the historical comprehensive alarm frequency, the second response time length and the second alarm feature data. If the historical event time interval is less than the preset event time interval threshold, the influence range difference value is obtained by performing difference processing on the event influence range between the target management road section and the collaborative road section, and the third alarm correlation coefficient is obtained according to the correlation characteristics among the historical comprehensive alarm frequency, the influence range difference value and the historical event time interval.

9. An intelligent alarm system based on scene recognition, applied to the intelligent alarm method based on scene recognition according to any one of claims 1 to 8, characterized in that, It includes: An identification module identifies the target management road section and the collaborative road section according to historical traffic event data; A judgment module analyzes the time correlation of the target management road section and the collaborative road section according to historical traffic event data, and obtains a condition judgment result according to a time correlation coefficient; wherein the target management road section and the collaborative road section are in a road network connected state, and the condition judgment result includes a time-space synchronization condition or a non-time-space synchronization condition; An analysis module analyzes the time-space distribution characteristics of alarm events in the target management road section according to historical traffic event data, and obtains an alarm weight ratio of different alarm sections according to the time-space distribution characteristics; A processing module obtains a comprehensive alarm correlation coefficient suitable for a time-space correlation scenario based on the condition judgment result, the alarm weight ratio, the historical alarm frequency, the event response time length and the influence range; The output module: real-time acquisition of the current traffic event data of the target control section and the cooperative section, calling the corresponding comprehensive alarm correlation according to the space-time correlation condition of the current traffic event data, generating the dynamic adjustment parameter of the emergency intelligent broadcast scheme, and outputting the space-time linkage alarm scheme according to the dynamic adjustment parameter.

10. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the scene recognition-based intelligent alarm method according to any one of claims 1 to 8 when executing the program.

Images