An Improved Whale Optimization Algorithm via Angle Penalized Distance for Automatic Train Operation.

Area of Science:

• Optimization Algorithms
• Intelligent Transportation Systems
• Urban Rail Transit

Background:

• Automatic Train Operation (ATO) quality is crucial for urban rail transit efficiency.
• Existing optimization algorithms may not fully address multi-objective requirements for ATO.
• Developing advanced algorithms is necessary to improve energy saving, punctuality, accurate stopping, and comfort.

Purpose of the Study:

• To propose a novel Improved Whale Optimization Algorithm via Angle Penalized Distance (IWOA-APD).
• To establish a multi-objective optimization model for ATO target speed curves.
• To enhance the rationality and applicability of the evaluation strategy for ATO performance.

Main Methods:

• Developed a multi-objective optimization model for ATO using energy saving, punctuality, accurate stopping, and comfort as indexes.
• Proposed an angle-penalized distance as a comprehensive evaluation strategy.
• Introduced IWOA-APD with strategies including non-linear decreasing convergence factor, advanced out-of-bounds elimination, genetic evolution mechanisms, and elite maintenance.

Main Results:

• IWOA-APD demonstrated significant performance improvements over traditional MOWOA.
• Achieved a 69.1% reduction in IGD value, 12.5% decrease in energy consumption.
• Improved punctuality by 64.6%, parking accuracy by 76.5%, and comfort by 15.9% in a real-world scenario.

Conclusions:

• The proposed IWOA-APD effectively enhances ATO quality by optimizing target speed curves.
• The angle-penalized distance evaluation strategy improves assessment rationality.
• IWOA-APD offers a superior approach for multi-objective optimization in urban rail transit operations.