An improved quantum genetic algorithm with adaptive dynamic rotation for urban facility spatial optimization.

Area of Science:

• Urban Planning and Design
• Computational Intelligence
• Geospatial Optimization

Background:

• Facility allocation is critical for urban planning but presents complex optimization challenges.
• Existing genetic algorithms struggle with convergence and diversity in large-scale spatial problems.
• Urban environments pose constraints like geographical correlation and heterogeneity.

Purpose of the Study:

• To propose an improved quantum genetic algorithm with adaptive dynamic rotation angle (IQGA-ADA) for spatial facility configuration.
• To address premature convergence and limited search diversity in existing heuristic methods.
• To enhance the effectiveness and fairness of urban facility planning schemes.

Main Methods:

• Developed IQGA-ADA with an adaptive dynamic rotation angle mechanism for population renewal.
• Integrated S-transform to process multi-modal spatial data (population, traffic, facility locations).
• Extracted characteristic quantities (root mean square, mean value) for pattern identification.

Main Results:

• IQGA-ADA demonstrated a 68% improvement in fitness value compared to the real-coded genetic algorithm (RCGA).
• Achieved faster convergence (approx. 5 iterations) versus RCGA (125 iterations).
• Showcased enhanced global search capabilities and performance in high-dimensional spatial optimization.

Conclusions:

• IQGA-ADA effectively overcomes limitations of traditional algorithms for spatial facility optimization.
• The algorithm improves fairness in facility repositioning, as shown in a first aid facility case study.
• Quantum evolution mechanisms show significant potential for solving complex geospatial optimization problems.