Automated Optimization of Intersections Using a Genetic Algorithm

Abstract

Traffic jams in large cities, in addition to having a very high economic cost, cause an increase in emissions generated by vehicles over the same route being driven under normal conditions. In recent years, there has been a rapid evolution in the technologies applied to the field of autonomous vehicles. There are currently commercial solutions for assisted driving and semi-autonomous driving systems, with very favorable forecasts for reaching a completely autonomous vehicle scenario in the coming decades. This new environment generates opportunities and challenges to reduce congestion in scenarios with autonomous or semi-autonomous vehicles. This paper focuses on the automatic optimization of the passage of vehicles through intersections. The intersections are one of the most conflict-generating elements in a traffic network. This type of conflicts arises because the intersections must manage multiple traffic flows with different priorities and preferences, often leading to traffic jams. The problem has been addressed by proposing three mechanisms to model any type of intersection, to calculate the roads with fewer points of conflict between their inputs and outputs, and to optimize the arrival rate of vehicles using a Genetic Algorithm to achieve the maximum performance of the intersection. To validate this solution, a cellular automata simulator has been developed, which can be adapted to both autonomous and conventional vehicle scenarios and can provide realistic results when certain conditions are met. The results obtained have been compared with other traditional solutions (priority and traffic lights) using microscopic traffic simulations, and with those obtained in other studies showing the advantages of the proposed system. The proposed systems achieve a throughput improvement between 9.21% and 36.98% compared with the traditional solutions.

RETRACTED: MuTraff: A Smart-City Multi-Map Traffic Routing Framework

“…We assume that 50% of the vehicles are using TWM. [25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] Our TWM now contains three maps: (1) µ F for firemen routing inside their emergency corridor; (2) µ P for policemen routing inside their emergency corridor; and (3) for the rest of the vehicles to discourage using the emergency corridors. µ F and µ P scale down original edge weights in the corridors, whilst µ V adds weight penalty to these edges.…”

A Node Selecting Approach for Traffic Network Based on Artificial Slime Mold

“…In figure 5, it is apparent that the traffic network topologies formed each time are slightly different because of the propagation capability or spreading ability of different nodes to be ranked, but the artificial system can find the optimal solution with the same redundancy rate 1.9444 from formula (22) Summing up the results in figure 4 and figure 5, the proposed artificial slime mold is verified to be able to connect all the nodes in Mexican highway system after the continuous learning and optimization of the myxamoebas, and get similar results as [11] but with shorter calculating time(70 hours is cost to get a solution in reference [11]). According to the optimized traffic networks with the same redundancy rate, the artificial slime mold can help us rank these nodes by the propagation capability or spreading ability.…”

“…But it is more difficult to efficiently evaluate the propagation capabilities by network data, i.e., [22] described an automated optimization of intersections using a genetic algorithm. The researches above applied a lot of complex artificial intelligence algorithms, such as deep learning [3], [4], spectral learning [19], genetic algorithm(GA) [22]. In 2010, a novel slime mold method was used for traffic network optimization [9]- [11].…”

“…Different from most living things using a brain or neurons to learn about the environment, the slime mold can only learn by a single-cell structure without any brain or neurons. The above learning algorithms [3], [4], [19], [22] are different from the single-cell slime mold, and cannot illustrate its operation mechanism. More importantly, the described slime mold in our paper is an artificial intelligence algorithm to solve the planning problems of traffic networks by simulating the foraging behavior of a real slime mold, rather than a true slime mold in [9]- [11], [23]- [25].…”

“…The whole algorithm flow is shown in Figure 2. The proposed selecting method is different from traditional selecting algorithms [1]- [8], [12]- [22] or other artificial intelligence algorithms [4], [19], [22]. First, it comprises the main components of a slime mold, such as external food sources, plasmodium, myxamoebas, and nutrients, etc., as shown in figure 1.…”