From: MRDPGA: a multiple restart dynamic population genetic algorithm for scheduling road traffic
Author(s) | Objective(s) | Tools/techniques | Results | Limitations |
---|---|---|---|---|
Dao et al. [4] | Global optimization | Adaptive restart and elite chromosomes in genetic algorithm | Performed better than the benchmark approaches | Population size modification is only at a restart; a new population may not guarantee the early achievement of global optima |
Potuzak [3] | Road traffic network load balancing | Distributed genetic algorithm computation and traffic division | Reduced computation time reported | Suffers from local optima challenge |
Potuzak [28] | Road traffic network load balancing | Distributed genetic algorithm and graph coarsening | Reduced computation time reported | Suffers from local optima challenges as well as spectral and cut guarantees challenges associated with graph coarsening |
Throughput: traffic flow enhancement | Standard genetic algorithm and cellular automata simulations | Achieved between 9.21 and 36.98% throughput improvement | Associated challenges of standard genetic algorithms | |
Basak [29] | Minimize the effect of a constant control operator | Adaptive mutation approach in genetic algorithm | Better performance reported | Focused on a single parameter of the genetic algorithm |
Muzid [30] | Global optimization | Fuzzy logic approach to determining crossover and mutation probability | Better results compared to the standard genetic algorithm technique | The choice of population size and boundaries remain a challenge. Non-adaptability of fuzzy logic is a problem |
Villalba-morales and RamĆrez-echeverry [31] | Steel trusses optimization in three-dimensional space | Multi-chromosome and self-adaptive parameters | 35% weight minimization achieved | Hamming cliffs, uninformed precision, and uneven schema importance are challenges associated with binary-coded genetic algorithms |
Mao et al. [15] | Traffic control optimization | Combined genetic algorithm and machine learning regression | Reported 43% minimization of average waiting time | Suffers from computational complexity and requires that the machine learning regression model be adequately trained |
Al-Madi and Hnaif [32] | Minimizing congestion and duration | Human community-based genetic algorithm | Achieved 83% and 13% in congestion and duration minimization | Possibility and impact of ineffective modeling of problem constraints |