Monitoring method based on dynamic traffic flow
A traffic flow and dynamic technology, which is applied in the direction of traffic flow detection, traffic control system, neural learning method, etc., can solve the problems that the accuracy and accuracy need to be further improved, so as to improve resource integration, improve prediction accuracy, and improve The effect of precision and accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0052] A monitoring method based on dynamic traffic flow, predicting the downstream traffic flow Q of one direction of the crossroad at time t+m through the traffic condition of a certain intersection at time t and time t-m 预测(t+m) , so as to monitor the traffic flow at different levels, including the following steps:
[0053] S1. Determine the independent variable parameters of the training model as
[0054] Three upstream traffic flow q in this direction at this intersection at time t 1 ,q 2 ,q 3 , obtained by video surveillance,
[0055] At time t and time t-m, the downstream traffic flow Q of this crossroad in this direction (t) , Q (t-m) , obtained by video surveillance,
[0056] Traffic impact level L of vehicles downstream in this direction at the intersection at time t,
[0057] The dependent variable parameter is the downstream traffic flow Q in one direction of the intersection at time t+m 预测(t+m) ;
[0058] Traffic flow is a dynamically changing quantity, w...
Embodiment 2
[0074] Contain all steps of embodiment 1;
[0075] Further, between S3 and S4 also includes
[0076] S30, obtain the predicted RF value Q of downstream traffic flow in one direction of the intersection at time t+m through random forest model prediction rf预测(t+m) , specifically, the random forest traffic model is obtained by inputting the historical data of the respective variable parameters and dependent variable parameters into the random forest model;
[0077] At time t, the values of the respective variables are input into the random forest model to obtain the RF prediction value Q of downstream traffic flow at this crossroad in this direction at time t+m rf预测(t+m) .
[0078] Further, where, S4, Q 预测(t+m) = K 1 *Q bp预测(t+m) +K 2 *Q rf预测(t+m) .
[0079] further,
[0080] S4, K 1 、K 2 The initial condition is K 1 +K 2 = 1; and K 1 、K 2 Calculated by the optimal weighted combination method.
Embodiment 3
[0113] A dynamic traffic flow based monitoring system comprising,
[0114] The first storage unit is used to store the following independent variable parameters of the training model,
[0115] Three upstream traffic flow q in this direction at this intersection at time t 1 ,q 2 ,q 3 ,
[0116] At time t and time t-m, the downstream traffic flow Q of this crossroad in this direction (t) , Q (t-m) ,
[0117] Traffic impact level L of downstream vehicles in this direction at the intersection at time t;
[0118] The first calculation unit is used to calculate the BP prediction value Q of downstream traffic flow in one direction of the intersection at time t+m from the above independent variable parameters through the BP neural network model bp预测(t+m) ;
[0119] The second calculation unit is used to calculate the above independent variable parameters through the random forest model to obtain the RF prediction value Q of downstream traffic flow in one direction of the cross...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More