Research On Multiple-objective Detector Optimization Location Based On OD Estimation Under Cost Constraint

Optimization of traffic detector location is significant in reducing investment cost, improvement of traffic information collection, guidance of traffic management and control. Optimization of traffic detector location based on OD estimation is to locate a certain quantity of detectors in a proper road section to calculate OD matrix of road network with traffic volume of partial road sections. There are two aspects of the location, density of location what is quantity, and road section of location. But, there are still some deficiencies in this field.This paper was based on basic principles and models of optimization of traffic detector location, focused on the problem that location cost and road section grade were not considered in the existing model, and founded a improved location model which containing maximum OD information with more extensive application range, minimum location cost, maximum truncation flow and minimum contained section. On this base, founded multi-objective detector optimization location model taking location cost and OD coverage principle as constraints, and expanded it to mixed location of detectors and the situation that detectors have been located in road sections.Based on the importance of sub-targets, considering model feature and practical application, tolerance analytic hierarchy process was used to solve multi-objective detector optimization location model. To sub-targets in every layer, optimal solution was solved by using genetic algorithm, and a multiple target satisfied detector location program was achieved finally.At last, took Nguyen-Dupuis network as example, the multi-objective model was solved and the optimized location program was achieved. Reallocated computative OD in road network to analyze MRE of traffic volume in road network and all networks. Analyze MRE of computative OD and traffic volume, and compared with traditional optimization model to prove its higher estimated accuracy. Therefore, based on cost constraint, the multi-objective detector optimization location model and its arithmetic was easy and practical with wider application and higher accurate, and could offer solving ideas and methods for detector location.