The A Type Trumpet Interchange Alignment Optimization Based On Genetic Algorithm

The set of the interchanges has great influence on traffic and local economic, the quality of design of ramp alignment also have an directly effects on ride comfort, safety and others of the interchange. Recent years have seen studies on highway intelligent alignment selection be just unfolding, the highway alignment selection become more and more intelligent, rational, while the optimization of interchange alignment rarely being analyzed. This paper aims to study the interchange ramp alignment optimization with genetic algorithm based on researches on highway alignment intelligent selection.Through analyzing the basic components of the trumpet interchange and actual situation of domestic construction, data shows that A type trumpet interchanges are in the majority in the setting of three road interchanges. So based on the basic principle of genetic algorithm, operating procedures and the key problems of applying need to be solved, the paper studies the alignment optimization of the A and B ramps(two left turn ramps)in the A type trumpet interchanges.The paper takes curvature chart and "BET algorithm" as the base of the interchange ramp alignment model, identifies design variables according to the A and B ramp alignment elements, identifies the final form of interchange ramp plane alignment model according to the calculation methods of plane alignment. It is also been analyzed about the mathematical expression form of longitudinal section model and constraint conditions, selecting five evaluation indexes and quantifying them, putting forward simplified objective function of interchange ramp alignment optimization based on membership functions. Referencing the method of basic genetic algorithm genetic coding, the paper determines the value range of alignment elements and the coding according to the demands on line length, transition curve length, circular curve radius and length in "Guidelines for Design of Highway Grade-separated Intersections". The paper further discusses the algorithm paralleling, the longitudinal section distance elevation models, objective function and restriction condition transformation. Genetic operator steps and processes are designed according to the general steps of the basic genetic algorithm.After confirming the engineering boundary conditions, the designed GA is operated as simulate and verification. The verification results show that, in the simplified engineering conditions, optimal interchange alignment with GA can solve the problem of parameters selecting and schemes evaluation more efficiently than the traditional method in design process.