In the rapidly evolving landscape of smart cities, where urban systems are becoming increasingly complex and interconnected, researchers are turning to multi-objective optimization (MOO) as a critical tool for planning, sustainability, and real-time decision-making. A comprehensive systematic literature review, led by YiFan Chen of the Jiaxing Key Laboratory of Industrial Intelligence and Digital Twin at Jiaxing Vocational and Technical College in China, sheds light on the state-of-the-art algorithms, persistent challenges, and future directions in this field. The study, published in PeerJ Computer Science (translated as “PeerJ Computer Science”), offers valuable insights for the energy sector and beyond.
The review, which analyzed 117 peer-reviewed studies published between 2015 and 2025, categorizes existing MOO techniques into four main families: bio-inspired, mathematical theory-driven, physics-inspired, and machine-learning-enhanced. Among these, established methods like Non-dominated Sorting Genetic Algorithm II (NSGA-II) and Multiobjective Evolutionary Algorithm based on Decomposition (MOED/D) have remained popular. However, the study highlights that hybrid frameworks, which combine deep learning with evolutionary search, are showing superior adaptability in high-dimensional, dynamic environments.
“Hybrid frameworks are particularly promising for the energy sector,” says Chen. “They can handle the complexity and uncertainty of energy systems, optimizing for multiple objectives such as cost, reliability, and sustainability simultaneously.”
The research identifies several persistent challenges in the field, including limited cross-domain generalizability, inadequate handling of uncertainty, and low interpretability of AI-assisted models. To address these issues, the study outlines a roadmap towards scalable, interpretable, and resilient optimization frameworks. It also provides a ready-to-use benchmarking toolkit and a deployment-oriented algorithm-selection matrix to guide researchers, engineers, and policymakers in real-world smart-city applications.
One of the key findings of the review is the need for privacy-preserving optimization techniques. As cities become smarter and more connected, the amount of data being generated and used for optimization purposes is increasing exponentially. Ensuring the privacy and security of this data is paramount, especially in the energy sector where sensitive information is often involved.
The study also emphasizes the importance of integrating MOO techniques with emerging technologies like digital twins, large language models, and neuromorphic computing. These integrations could lead to more accurate, efficient, and adaptive optimization solutions for smart cities.
As the world continues to urbanize, the demand for smart, sustainable, and efficient cities will only grow. This research provides a valuable roadmap for the future of urban optimization, offering insights and tools that can help shape the development of smarter, more resilient cities. For the energy sector, the findings could lead to more efficient and sustainable energy systems, ultimately benefiting both businesses and consumers.
In the words of Chen, “The future of smart cities lies in our ability to optimize for multiple objectives simultaneously. By addressing the challenges and leveraging the opportunities outlined in this review, we can pave the way for more sustainable, efficient, and resilient urban systems.”