Optimizing Marshall Test Parameters on Asphalt Concrete Using Hybrid Neural Network - Genetic Algorithm Approach

The design of the street should be applied the knowledge of the engineer principles for the density of traffic flow and rapidity in order to reduce the accident. A dilapidated mix-aggregate estimation will cause the reducing the street's quality. Marshall test is technique to test and discover out the level aggregate in mix-construction of asphalt. Both Marshall Stability and Marshall Flow are resulting of the tested to discover how maximum of load will be used by the asphalt. However, it needs a guarantee by the accuracy of the values test of marshall with computing method such as Neural Network. This means to solve the issue of accuracy toward some various data's and it is not linear. An optimization Artificial Neural Network tested to produce the exact values, to apply the Genetic Algorithm. It purposes to rise the exact being generated by Artificial Neural Network. This experiment has been done to get the optimization of the architecture and to produce the exact more high. The best model can be standardized as initialization stages of design software application based mobile application system.