HDRA: An Hybrid Divine Religion Algorithm Integrating Ant Colony Optimization and Memetic Search for Sustainable Electric Vehicle Routing in Smart Cities

In the bustling heart of modern cities, a quiet revolution is taking place on the streets. Electric vehicles are replacing traditional gas-powered trucks and vans, promising a future with cleaner air and less noise. However, managing a fleet of these electric vehicles is a puzzle far more complex than driving a single car. Unlike a gasoline car that can refuel in minutes at any station, an electric vehicle must carefully plan its journey to ensure it has enough battery power to reach its destination and return to the depot. It must also decide when and where to stop to recharge, balancing the time spent charging against the cost of the electricity and the urgency of the deliveries. This challenge, known as the electric vehicle routing problem, is a massive mathematical headache for city planners and logistics companies. If the routes are not planned perfectly, vehicles might run out of power, deliveries will be late, and the promised environmental benefits will be lost to inefficiency. Finding the best possible route among millions of options is a task so difficult that even the most powerful computers struggle to solve it quickly.

To tackle this challenge, a team of researchers from universities in Iran has developed a new computer program designed to find the most efficient paths for electric vehicles. They call their creation the Hybrid Divine Religion Algorithm, a name that might sound unusual but describes a method that mimics how people in a community interact and learn from one another. The researchers combined this social learning model with two other proven strategies: one that simulates how ants find food by leaving scent trails, and another that acts like a local guide, refining a solution by making small, careful adjustments. By weaving these three approaches together, the team created a system that can explore a vast landscape of possible routes without getting stuck in dead ends, while also digging deep into the most promising areas to find the absolute best path.

The researchers tested their new system against three other well-known methods used to solve similar problems. They ran simulations on thirty-five different scenarios, ranging from small deliveries with just a few customers to large-scale operations involving dozens of vehicles and many stops. In every test, the new hybrid system consistently found routes that were cheaper and more efficient than the others. On average, it reduced the total cost of the routes by about 12 percent compared to one of the older methods, and by roughly 7 to 11 percent compared to the others. This might seem like a small number, but in the world of logistics, where companies move thousands of packages every day, saving even a small percentage of cost translates into significant savings in money, time, and energy. The system achieved these results by balancing two competing needs: exploring new, untried routes to see if they are better, and exploiting the best routes it has already found to make them even more efficient.

What makes this new approach particularly clever is how it handles the flow of information. Imagine a group of travelers trying to find the shortest path through a dense forest. Some travelers act like ants, leaving marks on the ground to show others which paths are good. Others act like members of a community, where the most successful travelers share their specific knowledge with the rest of the group, allowing everyone to improve their journey at the same time. The researchers' system uses this mix of sharing and searching to avoid the common trap where computer programs get stuck in a local optimum—a route that looks good but is actually far from the best possible solution. By constantly refreshing its knowledge and allowing different parts of the system to learn from each other, the algorithm keeps moving forward, finding solutions that are not just good, but remarkably close to perfect.

The study also looked at how stable and reliable the system is. When the researchers ran the same problem multiple times, the new algorithm produced very similar results each time, showing that it is not just lucky but consistently effective. It performed well whether the problem was small or large, though it faced some challenges with the very largest scenarios, which is expected given the sheer complexity of those tasks. The researchers noted that while other methods were sometimes faster, they often settled for routes that were not as efficient. In contrast, this new system took a bit more time to think but delivered a much better final plan. This trade-off is often worth it for businesses that need to minimize their environmental footprint and operational costs simultaneously.

Ultimately, the work presented in this paper offers a powerful new tool for building smarter, greener cities. By improving how electric vehicles are routed, cities can reduce the total distance traveled, which directly lowers energy consumption and emissions. The researchers suggest that their method could serve as a decision-support tool for city planners and logistics managers, helping them design transportation networks that are both efficient and sustainable. While the study was conducted through computer simulations rather than real-world road tests, the results are robust and statistically significant, providing a strong foundation for future applications. As electric vehicles become more common, algorithms like this one will be essential for ensuring that the transition to green transportation is not just a promise, but a practical reality that works smoothly every day.