Data processing method based on artificial intelligence

Abstract

The invention provides a data processing method based on artificial intelligence, and belongs to the technical field of data processing. Comprising the following steps: step 1, collecting and storing historical traffic flow data, vehicle type data and real-time traffic data from an ith crossroad, and constructing a traffic data set; 2, constructing a traffic flow prediction model according to the traffic data set; 3, constructing a real-time traffic dynamic data set; step 4, constructing a fuel vehicle proportion coefficient and a congestion coefficient of the jth lane; and step 5, associating the fuel vehicle proportion coefficient with the congestion coefficient of the j-th lane, constructing a carbon emission risk coefficient of the j-th lane, optimizing road vehicle driving in real time, and achieving the goal of carbon emission reduction. According to the method, the extension from traffic flow prediction to carbon emission risk prediction is realized, and the potential carbon emission standard exceeding risk can be identified in advance in the road running process.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically, to a data processing method based on artificial intelligence. Background Technology

[0002] With the acceleration of urbanization, urban road traffic flow continues to increase, especially at road intersections, where vehicles are queuing for traffic lights, resulting in vehicles idling for long periods of time, which in turn generates a large amount of exhaust carbon emissions, adversely affecting urban air quality and public health.

[0003] Existing traffic management systems mostly rely on static signal control, which is insufficient to respond to the dynamic changes in real-time traffic flow and cannot effectively identify the impact of carbon emissions from fuel vehicles. For example, the correlation between lane congestion and carbon emissions has not been quantitatively analyzed, which means that traffic optimization schemes cannot directly serve carbon reduction targets. Furthermore, existing systems usually lack the ability to deeply analyze real-time traffic data, making it difficult to achieve a comprehensive assessment and dynamic control of lane traffic efficiency and carbon emission risks.

[0004] Therefore, an artificial intelligence-based data processing method is proposed to solve this problem. Summary of the Invention

[0005] To overcome the above deficiencies, the present invention provides an artificial intelligence-based data processing method that overcomes or at least partially solves the above technical problems.

[0006] This invention is implemented as follows: This invention provides a data processing method based on artificial intelligence, comprising: Step 1: Set up several intersections with traffic lights on a road. Establish a traffic data management node at the i-th intersection to collect and store historical traffic flow data, vehicle type data and real-time traffic data from the i-th intersection, and build a traffic dataset. Step 2: Preprocess the historical traffic flow data, vehicle type data and real-time traffic data in the traffic dataset, and build a traffic flow prediction model based on the traffic dataset; Step 3: After constructing the traffic flow prediction model, collect data on the vehicle flow, number of fuel vehicles, and lane occupancy rate of the i-th intersection, and construct a real-time traffic dynamic dataset. Step four involves inputting the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, predicting the vehicle data results for the i-th intersection, and constructing a fuel vehicle proportion coefficient based on the real-time traffic dynamics dataset. and the congestion coefficient of lane j ; Step 5: Adjust the percentage of gasoline-powered vehicles. and the congestion coefficient of lane j Correlation is used to construct the carbon emission risk coefficient for the j-th lane. It optimizes road vehicle operation in real time to achieve carbon emission reduction targets.

[0007] In a preferred embodiment, step one includes: S11. Select a road in the city and set up a first intersection, a second intersection, a third intersection... an nth intersection with traffic lights on the road. Set up a traffic data management node at the i-th intersection and use the microcontroller in the traffic data management node to store the historical traffic flow data, vehicle type data and real-time traffic data of the i-th intersection, and build a traffic dataset.

[0008] In a preferred embodiment, step two includes: S21. Perform outlier processing on historical traffic flow data, time-series feature processing on vehicle type data, and data smoothing on real-time traffic data based on traffic datasets. The method for handling outliers in the historical traffic flow data is as follows: By setting a maximum and minimum value for historical traffic flow, the minimum value is 0, and the maximum value is 1.5 times the historical maximum traffic flow of the road segment. Values ​​outside the range are replaced with the maximum and minimum values. When the historical traffic flow is less than the minimum value, it is ignored because the flow value is small. When the historical traffic flow exceeds the maximum value, it is considered abnormal because the flow value is large; For example, the historical maximum traffic flow at the i-th intersection is 200 vehicles per hour; At this point, the maximum value is 200 * 1.5 = 300; the minimum value is 0. When the actual historical traffic flow is 350, it is considered an outlier and will be adjusted to 300.

[0009] In a preferred embodiment, the vehicle type data is processed using the following temporal feature method: Statistical analysis of aggregated features over a past period, including; The average number of electric vehicles over the past hour reflects the overall usage level of electric vehicles. Standard deviation, calculated by the fluctuation in the number of fuel vehicles over the past 30 minutes, is used to measure the stability of fuel vehicle traffic flow. The maximum / minimum values ​​are used to calculate the peak number of electric vehicles over the past 2 hours to capture extreme traffic conditions. The specific method for smoothing the real-time traffic data is as follows: By smoothing real-time traffic data, random fluctuations and noise interference are reduced, highlighting the overall trend of traffic flow. For example, a moving average method can be used to average the real-time traffic over the past 5 or 15 minutes to smooth out instantaneous fluctuations.

[0010] In a preferred embodiment, step two further includes; S22. Based on the processed traffic dataset, construct a traffic flow prediction model, train and test the traffic flow prediction model using the processed traffic dataset, and use the trained traffic flow prediction model as a traffic data test and evaluation model. At the same time, use the intermediate layer output of the traffic flow prediction model run by the device as a feature vector to identify feature information, and use the trained traffic flow prediction model as data for prediction.

[0011] In a preferred embodiment, step three includes: S31. By deploying high-definition cameras at the i-th intersection and monitoring vehicles traveling in one direction within 100 meters of the stop line at the i-th intersection in real time, the number of vehicles is obtained. It uses high-definition cameras to identify vehicle license plate numbers and transmits the data wirelessly to the backend for vehicle attribute identification, thus counting the number of gasoline-powered vehicles. Furthermore, based on real-time monitoring of the number of vehicles in the j-th lane per unit time using high-definition cameras, the occupancy rate of the j-th lane is obtained by using the ratio of the number of vehicles in the j-th lane to the number of vehicles per unit time. And construct a real-time traffic dynamics dataset.

[0012] In a preferred embodiment, step four includes: S41. Input the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, and output the vehicle data prediction results for the i-th intersection. S42. Based on real-time traffic dynamics datasets, the proportion coefficient of fuel-powered vehicles is calculated using the following method. ; ; S43, Preset threshold for the proportion of fuel-powered vehicles And set a threshold for the proportion of fuel-powered vehicles. Coefficient of proportion of fuel vehicles A comparison is made to generate an assessment instruction for the proportion of fuel-powered vehicles, including: when > When the number of fuel vehicles at the i-th intersection is abnormal, a fuel vehicle ratio optimization instruction is generated, which includes adding a tidal flow lane to increase traffic capacity by 20%-25%, extending the green light duration by 10%-14%, and activating the bus lane to increase vehicle traffic efficiency by 35%-40%. when ≤ When the current number of fuel-powered vehicles at the i-th intersection is normal, the current traffic plan will be maintained to allow vehicles to pass, and monitoring will continue.

[0013] In a preferred embodiment, step four further includes; S44. Based on real-time traffic dynamics dataset, the congestion coefficient of lane j is calculated using the following method. ; S45, Preset congestion threshold for lane j. And set the congestion threshold for lane j. Congestion coefficient of lane j The comparison is performed to generate lane congestion assessment instructions, including: when > When the signal is received, it indicates that the j-th lane in the i-th intersection is congested abnormally. A lane congestion optimization instruction is generated, which includes extending the green light time of the j-th lane in the i-th intersection by 9%-13%, extending the green wave time by 8%-12%, and activating the tidal flow lane to improve the traffic efficiency of the j-th lane by 27%-32%. when ≤ When the current traffic situation is normal, it indicates that the current traffic plan will continue to be implemented and monitoring will continue.

[0014] In a preferred embodiment, step five includes: S51, Adjusting the proportion of fuel-powered vehicles Congestion coefficient of lane j Related, and normalized; The normalized proportion coefficient of fuel vehicles was obtained respectively. and the congestion coefficient of lane j ; The carbon emission risk coefficient of lane j is calculated using the following formula. .

[0015] In a preferred embodiment, step five further includes; S51, Preset carbon emission risk threshold for lane j. And set the carbon emission risk threshold for lane j. Carbon emission risk coefficient of lane j By comparing the results, a carbon emission risk assessment instruction for the j-th lane is generated, including: when > When the current carbon emission of lane j in the i-th intersection is abnormal, the green light duration of lane j in the direction of travel needs to be extended by 10% to 15% to improve traffic capacity and reduce idling emissions. Temporarily add tidal flow lanes to improve traffic capacity by 20% to 25%. In conjunction with green wave optimization, extend the green wave time by 8% to 12% to reduce carbon emissions caused by continuous stops at intersections. when ≤ When the value is zero, it indicates that the carbon emissions of the j-th lane in the i-th intersection are normal. The current traffic control plan should be maintained without optimization, and monitoring should continue.

[0016] The present invention provides a data processing method based on artificial intelligence, the beneficial effects of which include: By establishing a traffic data management node at the i-th intersection, historical traffic flow data, vehicle type data, and real-time traffic data can be collected simultaneously to construct a complete traffic dataset, providing a data foundation for subsequent modeling and prediction. By preprocessing the traffic dataset and constructing a traffic flow prediction model, it is possible to predict vehicle flow, the number of fuel vehicles, and lane occupancy rates, and to construct a carbon emission risk coefficient. This extends the prediction from traffic flow to carbon emission risk, enabling the early identification of potential carbon emission exceedance risks during road operation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1, referring to Figure 1 The present invention provides a technical solution: a data processing method based on artificial intelligence, comprising; Step 1: Set up several intersections with traffic lights on a road. Establish a traffic data management node at the i-th intersection to collect and store historical traffic flow data, vehicle type data and real-time traffic data from the i-th intersection, and build a traffic dataset. Step 2: Preprocess the historical traffic flow data, vehicle type data and real-time traffic data in the traffic dataset, and build a traffic flow prediction model based on the traffic dataset; Step 3: After constructing the traffic flow prediction model, collect data on the vehicle flow, number of fuel vehicles, and lane occupancy rate of the i-th intersection, and construct a real-time traffic dynamic dataset. Step four involves inputting the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, predicting the vehicle data results for the i-th intersection, and constructing a fuel vehicle proportion coefficient based on the real-time traffic dynamics dataset. and the congestion coefficient of lane j ; Step 5: Adjust the percentage of gasoline-powered vehicles. and the congestion coefficient of lane j Correlation is used to construct the carbon emission risk coefficient for the j-th lane. It optimizes road vehicle operation in real time to achieve carbon emission reduction targets.

[0021] In this embodiment, by establishing a traffic data management node at the i-th intersection, historical traffic flow data, vehicle type data, and real-time traffic data can be collected simultaneously to construct a complete traffic dataset, providing a data foundation for subsequent modeling and prediction. By preprocessing the traffic dataset and constructing a traffic flow prediction model, it is possible to predict vehicle flow, the number of fuel vehicles, and lane occupancy.

[0022] By collecting real-time traffic dynamic data and constructing a fuel vehicle proportion coefficient and a lane congestion coefficient, a refined characterization of the operating status of different lanes is achieved. It can reflect the proportion of fuel vehicles and the level of traffic congestion in real time. By correlating the fuel vehicle proportion coefficient with the lane congestion coefficient, a carbon emission risk coefficient is constructed, realizing the extension from traffic flow prediction to carbon emission risk prediction. It can identify potential carbon emission exceedance risks in advance during road operation.

[0023] Example 2 is an explanation of Example 1; please refer to it. Figure 1 Specifically, step one includes: S11. Select a road in the city and set up a first intersection, a second intersection, a third intersection... an nth intersection with traffic lights on the road. Set up a traffic data management node at the i-th intersection and use the microcontroller in the traffic data management node to store the historical traffic flow data, vehicle type data and real-time traffic data of the i-th intersection, and build a traffic dataset.

[0024] In this embodiment, by deploying traffic data management nodes at multiple intersections on urban roads and using the microcontrollers in the nodes to collect and store historical traffic flow data, vehicle type data, and real-time traffic data, global monitoring and localized management of road operation status can be achieved. The traffic dataset built on this basis has the characteristics of wide data sources, high timeliness, and diverse types, which not only ensures the data integrity and reliability of subsequent traffic flow prediction models, but also provides high-precision data support for traffic operation analysis and carbon emission assessment.

[0025] Example 3 is an explanation of Example 1; please refer to the provided text. Figure 1 Specifically, step two includes: S21. Perform outlier processing on historical traffic flow data, time-series feature processing on vehicle type data, and data smoothing on real-time traffic data based on traffic datasets. The method for handling outliers in the historical traffic flow data is as follows: By setting a maximum and minimum value for historical traffic flow, the minimum value is 0, and the maximum value is 1.5 times the historical maximum traffic flow of the road segment. Values ​​outside the range are replaced with the maximum and minimum values. When the historical traffic flow is less than the minimum value, it is ignored because the flow value is small. When the historical traffic flow exceeds the maximum value, it is considered abnormal because the flow value is large; For example, the historical maximum traffic flow at the i-th intersection is 200 vehicles per hour; At this point, the maximum value is 200 * 1.5 = 300; the minimum value is 0. When the actual historical traffic flow is 350, it is considered an outlier and will be adjusted to 300.

[0026] In this embodiment, by setting reasonable maximum and minimum value ranges for historical traffic flow data and replacing or removing values ​​that exceed the range, abnormal data points can be effectively filtered out.

[0027] Example 4 is an explanation of Example 1; please refer to the provided text. Figure 1 Specifically, the method for performing time-series feature processing on the vehicle type data is as follows: Statistical analysis of aggregated features over a past period, including; The average number of electric vehicles over the past hour reflects the overall usage level of electric vehicles. Standard deviation, calculated by the fluctuation in the number of fuel vehicles over the past 30 minutes, is used to measure the stability of fuel vehicle traffic flow. The maximum / minimum values ​​are used to calculate the peak number of electric vehicles over the past 2 hours to capture extreme traffic conditions. The specific method for smoothing the real-time traffic data is as follows: By smoothing real-time traffic data, random fluctuations and noise interference are reduced, highlighting the overall trend of traffic flow. For example, a moving average method can be used to average the real-time traffic over the past 5 or 15 minutes to smooth out instantaneous fluctuations.

[0028] In this embodiment, by performing time-series feature processing on vehicle type data, aggregated features such as mean, standard deviation, and maximum / minimum values ​​over a past period can be extracted. By smoothing real-time traffic data, the interference of random fluctuations and sensor noise can be reduced, highlighting the overall trend of traffic flow changes. This processing method effectively reduces the interference of short-term abnormal peaks on the model input, making the prediction results more stable and reliable, while also improving the model's ability to continuously express the road operating state.

[0029] Example 5 is an explanation of Example 1; please refer to it. Figure 1 Specifically, step two also includes; S22. Based on the processed traffic dataset, construct a traffic flow prediction model, train and test the traffic flow prediction model using the processed traffic dataset, and use the trained traffic flow prediction model as a traffic data test and evaluation model. At the same time, use the intermediate layer output of the traffic flow prediction model run by the device as a feature vector to identify feature information, and use the trained traffic flow prediction model as data for prediction.

[0030] In this embodiment, by constructing a traffic flow prediction model based on the processed traffic dataset and using the dataset to train and test the model, the prediction accuracy and stability of the model can be significantly improved. At the same time, using the trained traffic flow prediction model as a traffic data test and evaluation model can effectively verify the input data and improve the robustness of the system.

[0031] Example 6 is an explanation of Example 1; please refer to the provided text. Figure 1 Specifically, step three includes: S31. By deploying high-definition cameras at the i-th intersection and monitoring vehicles traveling in one direction within 100 meters of the stop line at the i-th intersection in real time, the number of vehicles is obtained. It uses high-definition cameras to identify vehicle license plate numbers and transmits the data wirelessly to the backend for vehicle attribute identification, thus counting the number of gasoline-powered vehicles. Furthermore, based on real-time monitoring of the number of vehicles in the j-th lane per unit time using high-definition cameras, the occupancy rate of the j-th lane is obtained by using the ratio of the number of vehicles in the j-th lane to the maximum capacity of the j-th lane. And construct a real-time traffic dynamics dataset.

[0032] In this embodiment, by deploying a high-definition camera at the i-th intersection, passing vehicles can be detected in real time, the number of vehicles can be accurately obtained, and by combining license plate recognition and background vehicle attribute recognition, the number of fuel vehicles can be automatically counted, thereby improving the accuracy and intelligence of data collection.

[0033] Furthermore, by using high-definition cameras to monitor the number of vehicles in lane j within a unit of time and calculating the lane occupancy rate based on the ratio of vehicle count to time, the dynamic usage and congestion level of the lane can be accurately reflected. The real-time traffic dynamic dataset built upon this foundation not only improves the real-time performance and completeness of the data but also provides highly timely and accurate input data support for subsequent traffic flow prediction models, thus ensuring the reliability of the prediction results and laying a solid foundation for carbon emission risk assessment and traffic control optimization. Example 7 is an explanation of Example 1; please refer to it. Figure 1 Specifically, step four includes: S41. Input the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, and output the vehicle data prediction results for the i-th intersection. S42. Based on real-time traffic dynamics datasets, the proportion coefficient of fuel-powered vehicles is calculated using the following method. ; ; The following is the percentage of gasoline-powered vehicles. The sample data example table is shown below;

[0034] S43, Preset threshold for the proportion of fuel-powered vehicles ; Threshold for the proportion of fuel-powered vehicles Used to determine whether the proportion of fuel-powered vehicles at the current i-th intersection is abnormal; fuel-powered vehicle proportion threshold. This can be obtained based on historical traffic data statistics, specifically by setting the average and standard deviation of the historical proportion of fuel-powered vehicles, thereby establishing the threshold for the proportion of fuel-powered vehicles. Set to 0.59; And set a threshold for the proportion of fuel-powered vehicles Coefficient of proportion of fuel vehicles A comparison is made to generate an assessment instruction for the proportion of fuel-powered vehicles, including: when > When the number of fuel vehicles at the i-th intersection is abnormal, a fuel vehicle ratio optimization instruction is generated, which includes adding a tidal flow lane to increase traffic capacity by 20%-25%, extending the green light duration by 10%-14%, and activating the bus lane to increase vehicle traffic efficiency by 35%-40%. when ≤ When the current number of fuel-powered vehicles at the i-th intersection is normal, the current traffic plan will be maintained to allow vehicles to pass, and monitoring will continue.

[0035] Based on the proportion of fuel vehicles Using sample data, we will continue to construct a threshold for the proportion of gasoline-powered vehicles. Coefficient of proportion of fuel vehicles A sample table of comparison data is shown below;

[0036] In this embodiment, by inputting real-time traffic dynamic datasets into the traffic flow prediction model, the vehicle data prediction results for the i-th intersection can be output, enabling real-time understanding and early warning of traffic operation trends.

[0037] Based on real-time traffic dynamic datasets, the proportion of fuel vehicles is further calculated and compared with a preset threshold. This enables dynamic monitoring and risk identification of the proportion of fuel vehicles. When the proportion of fuel vehicles exceeds the threshold, the system can automatically generate a fuel vehicle proportion optimization instruction.

[0038] Example 8 is an explanation of Example 1; please refer to it. Figure 1 Specifically, step four also includes; S44. Based on real-time traffic dynamics dataset, the congestion coefficient of lane j is calculated using the following method. ; First, calculate the occupancy rate of lane j. ; In the formula, Let j represent the maximum number of vehicles that the j-th lane can carry. Based on historical data, a standard straight lane can accommodate approximately 20 vehicles within 60 seconds of a green light. Then calculate the congestion coefficient of lane j. ; ; In the formula, This represents the maximum occupancy rate of the j-th lane; The following is the congestion coefficient for lane j. The sample data example table is shown below;

[0039] S45, Preset congestion threshold for lane j. ; Preset congestion threshold for lane j This is used to determine the congestion status of the j-th lane in the i-th intersection, where the j-th lane has a congestion threshold. This can be obtained based on historical traffic data statistics, specifically by setting the average and standard deviation of historical lane occupancy rates. The congestion threshold for lane j can then be derived from historical data. The value is 0.50; And the congestion threshold of the j-th lane Congestion coefficient of lane j The comparison is performed to generate lane congestion assessment instructions, including: when > When the signal is received, it indicates that the j-th lane in the i-th intersection is congested abnormally. A lane congestion optimization instruction is generated, which includes extending the green light time of the j-th lane in the i-th intersection by 9%-13%, extending the green wave time by 8%-12%, and activating the tidal flow lane to improve the traffic efficiency of the j-th lane by 27%-32%. when ≤ When the current traffic situation is normal, it indicates that the current traffic plan will continue to be implemented and monitoring will continue.

[0040] Based on the congestion coefficient of lane j Using sample data, we can further construct the congestion threshold for the j-th lane. Congestion coefficient of lane j The comparison data example table is shown below;

[0041] In this embodiment, the first [item] is obtained by calculating based on a real-time traffic dynamic dataset. jThe system calculates the congestion coefficient of each lane and compares it with a preset congestion threshold, enabling quantitative monitoring and dynamic evaluation of lane operation status. When the congestion coefficient exceeds the threshold, the system automatically generates lane congestion optimization instructions; when the congestion coefficient is less than or equal to the threshold, the system maintains the current traffic plan and only performs continuous monitoring to ensure the stability and rationality of traffic control.

[0042] Example 9, this example is an explanation of Example 1, please refer to it. Figure 1 Specifically, step five includes: S51, Adjusting the proportion of fuel-powered vehicles Congestion coefficient of lane j Related, and normalized; The normalized proportion coefficient of fuel vehicles was obtained respectively. and the congestion coefficient of lane j ; The carbon emission risk coefficient of lane j is calculated using the following formula. ; .

[0043] The following is the carbon emission risk coefficient for lane j. The sample data example table is shown below;

[0044] In this embodiment, by associating the fuel vehicle proportion coefficient with the congestion coefficient of the j-th lane and normalizing the correlation, the dimensional differences between different indicators can be eliminated, ensuring the comparability and consistency of the data. Furthermore, the carbon emission risk coefficient of the j-th lane is calculated using a formula, achieving a coupled analysis and quantitative expression of the fuel vehicle proportion and lane congestion level.

[0045] Example 10: This example is an explanation of Example 1. Please refer to the provided text. Figure 1 Specifically, step five also includes; S51, Preset carbon emission risk threshold for lane j. ; A preset carbon emission risk threshold for lane j is used to determine the carbon emission status of lane j at the current i-th intersection. This threshold is dynamically adjusted based on real-time traffic flow, the proportion of fuel-powered vehicles, and congestion levels. The carbon emission risk threshold for lane j can be derived from historical data. And take the value 0.60; The carbon emission risk threshold for lane j Carbon emission risk coefficient of lane j By comparing the results, a carbon emission risk assessment instruction for the j-th lane is generated, including: when > When the current carbon emission of lane j in the i-th intersection is abnormal, the green light duration of lane j in the direction of travel needs to be extended by 10% to 15% to improve traffic capacity and reduce idling emissions. Temporarily add tidal flow lanes to improve traffic capacity by 20% to 25%. In conjunction with green wave optimization, extend the green wave time by 8% to 12% to reduce carbon emissions caused by continuous stops at intersections. when ≤ When the value is zero, it indicates that the carbon emissions of the j-th lane in the i-th intersection are normal. The current traffic control plan should be maintained without optimization, and monitoring should continue.

[0046] Based on the carbon emission risk coefficient of lane j Using sample data, we can further construct the carbon emission risk threshold for the j-th lane. Carbon emission risk coefficient of lane j The comparison data example table is shown below;

[0047] In this embodiment, by preset a carbon emission risk threshold for the j-th lane and comparing it with a carbon emission risk coefficient, real-time early warning and dynamic assessment of lane exhaust emission risk can be achieved. When the carbon emission risk coefficient exceeds the threshold, the system can automatically generate optimization instructions.

[0048] By employing a multi-strategy approach, the risk of localized high emissions can be effectively mitigated, achieving the dual goals of carbon reduction and improved traffic efficiency. When the carbon emission risk coefficient is below or equal to the threshold, the system maintains the current traffic control plan and only performs continuous monitoring, thereby ensuring the stability and reliability of traffic regulation.

[0049] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0050] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A data processing method based on artificial intelligence, characterized in that, include; Step 1: Set up several intersections with traffic lights on a road. Establish a traffic data management node at the i-th intersection to collect and store historical traffic flow data, vehicle type data and real-time traffic data from the i-th intersection, and build a traffic dataset. Step 2: Preprocess the historical traffic flow data, vehicle type data and real-time traffic data in the traffic dataset, and build a traffic flow prediction model based on the traffic dataset; Step 3: After constructing the traffic flow prediction model, collect data on the vehicle flow, number of fuel vehicles, and lane occupancy rate of the i-th intersection, and construct a real-time traffic dynamic dataset. Step four involves inputting the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, predicting the vehicle data results for the i-th intersection, and constructing a fuel vehicle proportion coefficient based on the real-time traffic dynamics dataset. and the congestion coefficient of lane j ; Step 5: Adjust the percentage of gasoline-powered vehicles. and the congestion coefficient of lane j Correlation is used to construct the carbon emission risk coefficient for the j-th lane. It optimizes road vehicle operation in real time to achieve carbon emission reduction targets.

2. The data processing method based on artificial intelligence according to claim 1, characterized in that, Step one includes: S11. Select a road in the city and set up a first intersection, a second intersection, a third intersection... an nth intersection with traffic lights on the road. Set up a traffic data management node at the i-th intersection and use the microcontroller in the traffic data management node to store the historical traffic flow data, vehicle type data and real-time traffic data of the i-th intersection, and build a traffic dataset.

3. The data processing method based on artificial intelligence according to claim 2, characterized in that, Step two includes: S21. Perform outlier processing on historical traffic flow data, time-series feature processing on vehicle type data, and data smoothing on real-time traffic data based on traffic datasets. The method for handling outliers in the historical traffic flow data is as follows: By setting a maximum and minimum value for historical traffic flow, the minimum value is 0, and the maximum value is 1.5 times the historical maximum traffic flow of the road segment. Values ​​outside the range are replaced with the maximum and minimum values. When the historical traffic flow is less than the minimum value, it is ignored because the flow value is small. When the historical traffic flow exceeds the maximum value, it is considered abnormal because the flow value is large.

4. The data processing method based on artificial intelligence according to claim 3, characterized in that, The method for processing the vehicle type data using time-series features is as follows: Statistical analysis of aggregated features over a past period, including; The average number of electric vehicles over the past hour reflects the overall usage level of electric vehicles. Standard deviation, calculated by the fluctuation in the number of fuel vehicles over the past 30 minutes, is used to measure the stability of fuel vehicle traffic flow. The maximum / minimum values ​​are used to calculate the peak number of electric vehicles over the past 2 hours to capture extreme traffic conditions. The specific method for smoothing the real-time traffic data is as follows: By smoothing real-time traffic data, random fluctuations and noise interference are reduced, highlighting the overall trend of traffic flow.

5. The data processing method based on artificial intelligence according to claim 4, characterized in that, Step two also includes; S22. Based on the processed traffic dataset, construct a traffic flow prediction model, train and test the traffic flow prediction model using the processed traffic dataset, and use the trained traffic flow prediction model as a traffic data test and evaluation model. At the same time, use the intermediate layer output of the traffic flow prediction model run by the device as a feature vector to identify feature information, and use the trained traffic flow prediction model as data for prediction.

6. The data processing method based on artificial intelligence according to claim 5, characterized in that, Step three includes: S31. By deploying high-definition cameras at the i-th intersection and monitoring vehicles traveling in one direction within 100 meters of the stop line at the i-th intersection in real time, the number of vehicles is obtained. It uses high-definition cameras to identify vehicle license plate numbers and transmits the data wirelessly to the backend for vehicle attribute identification, thus counting the number of gasoline-powered vehicles. Furthermore, based on real-time monitoring of the number of vehicles in the j-th lane per unit time using high-definition cameras, the occupancy rate of the j-th lane is obtained by using the ratio of the number of vehicles in the j-th lane to the number of vehicles per unit time. And construct a real-time traffic dynamics dataset.

7. The data processing method based on artificial intelligence according to claim 6, characterized in that, Step four includes: S41. Input the real-time traffic dynamics dataset into the traffic flow prediction model for analysis, and output the vehicle data prediction results for the i-th intersection. S42. Based on real-time traffic dynamics datasets, the proportion coefficient of fuel-powered vehicles is calculated using the following method. ; ； S43, Preset threshold for the proportion of fuel-powered vehicles And set a threshold for the proportion of fuel-powered vehicles. Coefficient of proportion of fuel vehicles A comparison is made to generate an assessment instruction for the proportion of fuel-powered vehicles, including: when > When the number of fuel vehicles at the i-th intersection is abnormal, a fuel vehicle ratio optimization instruction is generated, which includes adding a tidal flow lane to increase traffic capacity by 20%-25%, extending the green light duration by 10%-14%, and activating the bus lane to increase vehicle traffic efficiency by 35%-40%. when ≤ When the current number of fuel-powered vehicles at the i-th intersection is normal, the current traffic plan will be maintained to allow vehicles to pass, and monitoring will continue.

8. The data processing method based on artificial intelligence according to claim 7, characterized in that, Step four also includes; S44. Based on real-time traffic dynamics dataset, the congestion coefficient of lane j is calculated using the following method. ; S45, Preset congestion threshold for lane j. And set the congestion threshold for lane j. Congestion coefficient of lane j The comparison is performed to generate lane congestion assessment instructions, including: when > When the signal is received, it indicates that the j-th lane in the i-th intersection is congested abnormally. A lane congestion optimization instruction is generated, which includes extending the green light time of the j-th lane in the i-th intersection by 9%-13%, extending the green wave time by 8%-12%, and activating the tidal flow lane to improve the traffic efficiency of the j-th lane by 27%-32%. when ≤ When the current traffic situation is normal, it indicates that the current traffic plan will continue to be implemented and monitoring will continue.

9. A data processing method based on artificial intelligence according to claim 8, characterized in that, Step five includes: S51, Adjusting the proportion of fuel-powered vehicles Congestion coefficient of lane j Related, and normalized; The normalized proportion coefficient of fuel vehicles was obtained respectively. and the congestion coefficient of lane j ; The carbon emission risk coefficient of lane j is calculated using the following formula. .

10. A data processing method based on artificial intelligence according to claim 9, characterized in that, Step five also includes; S51, Preset carbon emission risk threshold for lane j. And set the carbon emission risk threshold for lane j. Carbon emission risk coefficient of lane j By comparing the results, a carbon emission risk assessment instruction for the j-th lane is generated, including: when > When the current carbon emission of lane j in the i-th intersection is abnormal, the green light duration of lane j in the direction of travel needs to be extended by 10% to 15% to improve traffic capacity and reduce idling emissions. Temporarily add tidal flow lanes to improve traffic capacity by 20% to 25%. In conjunction with green wave optimization, extend the green wave time by 8% to 12% to reduce carbon emissions caused by continuous stops at intersections. when ≤ When the value is zero, it indicates that the carbon emissions of the j-th lane in the i-th intersection are normal. The current traffic control plan should be maintained without optimization, and monitoring should continue.

Images