Enhanced initialization of NSGA-II for multi-objective optimization of air turbo-rocket engine performance

This study establishes a multi-objective optimization framework for air turbo rocket (ATR) engines using an initial-value-enhanced NSGA-II algorithm to concurrently satisfy thrust requirements across flight regimes and maximize specific impulse. The model optimizes thrust and specific impulse subject to collaborative thermodynamic constraints. Pareto front analysis confirms that the method achieves a 97% gain in computational efficiency over exhaustive search while delivering superior convergence and diversity compared to both standard and surrogate-assisted NSGA-II. At design point, turbine expansion ratio dictates peak specific impulse. Increasing specific thrust shifts dominant constraints to compressor pressure ratio and exhaust gas temperature. Under off-design point conditions at lower inlet total temperatures, a Pareto inflection point emerges: above this threshold, thrust decays rapidly while specific impulse rises marginally; below it, thrust gains are offset by sharp specific impulse deterioration. Elevated inlet temperatures establish an inverse linear relationship between thrust and specific impulse.