CN-121981578-A - Ultra-large-scale traffic control scheme optimization system and method based on parallel computing

CN121981578ACN 121981578 ACN121981578 ACN 121981578ACN-121981578-A

Abstract

The present disclosure provides a system and a method for optimizing a super-large-scale traffic control scheme based on parallel computation, wherein the method for optimizing the super-large-scale traffic control scheme comprises the following steps of generating a plurality of candidate control schemes based on road network actual traffic conditions; the method comprises the steps of carrying out batch parallel simulation evaluation on a plurality of candidate management and control schemes by using a Master-workbench parallel architecture and a two-step simulation workflow, analyzing and comparing the plurality of candidate management and control schemes subjected to simulation evaluation from a plurality of dimensions, and carrying out recommended sequencing on the plurality of candidate management and control schemes by using a comprehensive scoring model and combining road network actual conditions.

Inventors

LIU NAN
YI QIAN
SUN LING
WANG LIN
Yang Fengman
NIU JIAPENG

Assignees

交通运输部公路科学研究所

Dates

Publication Date: 20260505
Application Date: 20260228

Claims (10)

1. The ultra-large scale traffic control scheme optimizing method based on parallel computing is characterized by comprising the following steps of: generating a plurality of candidate management and control schemes based on the road network actual traffic conditions; Performing batch parallel simulation evaluation on the plurality of candidate management and control schemes by adopting a Master-workbench parallel architecture and a two-step simulation workflow; analyzing and comparing the multiple candidate control schemes subjected to simulation evaluation from multiple dimensions; And recommending and sequencing the plurality of candidate control schemes by adopting a comprehensive scoring model and combining the actual condition of the road network.
2. The optimization method of a traffic control scheme based on parallel computing according to claim 1, wherein the generating a plurality of candidate control schemes based on road network actual traffic conditions comprises the steps of: Collecting and analyzing the actual traffic condition of the road network; adapting a plurality of management and control strategies based on the actual condition of the road network to generate a plurality of initial candidate management and control schemes; Performing management and control parameter calculation of a plurality of initial candidate management and control schemes based on traffic flow theory; generating scheme metadata and outputting the plurality of candidate management schemes.
3. The method for optimizing a traffic control scheme on a very large scale based on parallel computing according to claim 2, wherein the plurality of control strategies comprises: variable speed limit VSS, speed limit of the upstream of the congestion road section is regulated; an inlet flow control TEC for controlling the flow of vehicles per unit time at the inlet of the main road, and Dynamic hard shoulder DHS, hard shoulder traffic of dynamic open congested road segments, The plurality of management policies are configured to be implemented alone or in combination.
4. The optimization method of a traffic control scheme based on parallel computing according to claim 2, wherein the control parameters include: VSS parameter limit_speed = reference speed x severity coefficient; TEC parameter inflow _limit= (road network capacity-current flow) x safety factor; DHS parameters open_condition=flow capability > flow threshold AND speed < speed threshold, AND Timing coordination and/or spatial coordination of at least two of variable speed limit VSS, ingress flow control TEC, dynamic hard shoulder DHS.
5. The optimization method of the ultra-large-scale traffic control scheme based on parallel computing according to claim 1, wherein the batch parallel simulation evaluation of the candidate control schemes by using a Master-workbench parallel architecture and a two-step simulation workflow comprises the following steps: Executing simulation preparation stages of a plurality of candidate management and control schemes in parallel; And executing simulation execution stages of a plurality of candidate management and control schemes in parallel.
6. The optimization method of the ultra-large-scale traffic control scheme based on parallel computing according to claim 1, wherein the analyzing and comparing the plurality of candidate control schemes subjected to simulation evaluation from a plurality of dimensions comprises the following steps: performing aggregation statistical analysis on simulation results of a plurality of candidate management and control schemes; Calculating improvement effect parameters of a plurality of candidate management and control schemes; And generating multi-dimensional comparison results of a plurality of candidate management and control schemes.
7. The optimization method of a parallel computing-based ultra-large-scale traffic control scheme according to claim 6, wherein the generating multi-dimensional comparison results of a plurality of candidate control schemes comprises the steps of: Generating a multi-dimensional comparison table of a plurality of candidate management and control schemes; Generating peak curve comparison diagrams of a plurality of candidate management and control schemes; And performing visual display.
8. The optimization method of the ultra-large-scale traffic control scheme based on parallel computing according to claim 1, wherein the recommendation ordering of the plurality of candidate control schemes by combining the actual condition of the road network by adopting the comprehensive scoring model comprises the following steps: constructing a comprehensive scoring model; Parameters of the comprehensive scoring model are adjusted according to the suitability of the road network actual conditions; Executing scoring operations of a plurality of candidate management and control schemes and performing preferential ranking according to scoring results; a decision suggestion report is generated.
9. The optimization method of a parallel computing-based traffic control scheme of claim 8, wherein the comprehensive scoring model performs scoring using the following formula: total_score = a1×score_speed+ a2×score_delay+ a3×score_capacity+ a4×score_feasibility score_speed represents a speed up score, its weight is a1, score_delay represents a delay reduction score, its weight is a2, score_capability represents a traffic capacity score, its weight is a3, score_feassibility represents an implementation feasibility score, its weight is a4.
10. The ultra-large scale traffic control scheme optimizing system based on parallel computing is characterized by comprising: The scheme generation module is configured to generate a plurality of candidate management and control schemes based on the road network actual traffic conditions; The scheme evaluation module is configured to perform batch parallel simulation evaluation on the plurality of candidate management and control schemes by adopting a Master-workbench parallel architecture and a two-step simulation workflow; The comparison analysis module is configured to analyze and compare the multiple candidate management and control schemes subjected to simulation evaluation from multiple dimensions; And the optimal decision module is configured to adopt a comprehensive scoring model to carry out recommended sorting on the plurality of candidate management and control schemes in combination with the road network actual conditions.

Description

Ultra-large-scale traffic control scheme optimization system and method based on parallel computing Technical Field The invention belongs to the technical field of intelligent traffic control simulation optimization, and particularly relates to a system and a method for optimizing a super-large-scale traffic control scheme based on parallel calculation, which are suitable for multi-strategy control scheme generation, batch simulation evaluation, multidimensional comparison analysis and optimal scheme selection of super-large-scale road networks such as expressways, urban expressways and the like. Background In the actual traffic management, certain road sections or areas can periodically present similar traffic jam modes, such as normalized jam of a specific road section in the early peak period of the expressway, repeated jam of the urban expressway and the late peak, regular jam of a specific entrance and exit during holidays and the like. Traffic control has important functions of balancing traffic network load, relieving congestion, improving transportation efficiency, reducing safety accidents and the like. How to scientifically and reasonably formulate an optimal traffic control scheme is a problem to be solved in traffic management. Disclosure of Invention On the one hand, the method aims at realizing efficient generation, batch evaluation, multidimensional comparison and scientific preference of a traffic control scheme of the ultra-large-scale road network so as to rapidly select a comprehensive control scheme which is most suitable for the actual situation of the road network. Some embodiments of the present disclosure provide a method for optimizing a traffic control scheme on a very large scale based on parallel computing, the method for optimizing the traffic control scheme on the very large scale comprising the steps of: generating a plurality of candidate management and control schemes based on the road network actual traffic conditions; Performing batch parallel simulation evaluation on the plurality of candidate management and control schemes by adopting a Master-workbench parallel architecture and a two-step simulation workflow; analyzing and comparing the multiple candidate control schemes subjected to simulation evaluation from multiple dimensions; And recommending and sequencing the plurality of candidate control schemes by adopting a comprehensive scoring model and combining the actual condition of the road network. In some embodiments, the generating a plurality of candidate management and control schemes based on the road network actual traffic conditions includes the steps of: Collecting and analyzing the actual traffic condition of the road network; adapting a plurality of management and control strategies based on the actual condition of the road network to generate a plurality of initial candidate management and control schemes; Performing management and control parameter calculation of a plurality of initial candidate management and control schemes based on traffic flow theory; generating scheme metadata and outputting the plurality of candidate management schemes. In some embodiments, the plurality of management strategies includes: variable speed limit VSS, speed limit of the upstream of the congestion road section is regulated; an inlet flow control TEC for controlling the flow of vehicles per unit time at the inlet of the main road, and Dynamic hard shoulder DHS, hard shoulder traffic of dynamic open congested road segments, The plurality of management policies are configured to be implemented alone or in combination. In some embodiments, the management parameters include: VSS parameter limit_speed = reference speed x severity coefficient; TEC parameter inflow _limit= (road network capacity-current flow) x safety factor; DHS parameters open_condition=flow capability > flow threshold AND speed < speed threshold, AND Timing coordination and/or spatial coordination of at least two of variable speed limit VSS, ingress flow control TEC, dynamic hard shoulder DHS. In some embodiments, the batch parallel simulation evaluation of the plurality of candidate management and control schemes by using a Master-workbench parallel architecture and a two-step simulation workflow comprises the following steps: Executing simulation preparation stages of a plurality of candidate management and control schemes in parallel; And executing simulation execution stages of a plurality of candidate management and control schemes in parallel. In some embodiments, the analyzing and comparing the plurality of candidate management and control schemes subjected to simulation evaluation from the plurality of dimensions comprises the following steps: performing aggregation statistical analysis on simulation results of a plurality of candidate management and control schemes; Calculating improvement effect parameters of a plurality of candidate management and control schemes; And generating multi-dimensional co