Ammonia/hydrogen dual-fuel engines present a promising zero-carbon solution but face challenges regarding high NOx emissions, long ignition delays, and suboptimal power output. This study investigates the effects of diesel injection timing and quantity on the combustion and emission characteristics of an engine fueled with a 9:1 ammonia/hydrogen mixture. Results indicate that advancing injection timing and increasing diesel quantity enhance peak in-cylinder pressure and premixed heat release. Specifically, advancing timing from -6 degrees CA to -18 degrees CA at a 10% diesel energy fraction increased peak pressure by 6.6% (1.6 MPa), but extended ignition delay by 1.1 degrees CA and advanced CA50 by 3 degrees CA. Conversely, increasing diesel quantity shortened ignition delay while significantly boosting late-stage heat release. Regarding emissions, retarding injection timing to -6 degrees CA reduced NOx by approximately 55% compared to -18 degrees CA, though it increased N2O by 9%, CO2 by 4.1%, and delayed the soot peak by 14.3 degrees CA. Increasing diesel quantity effectively reduced N2O but elevated CO2, NOx, and soot emissions. Evaluating overall performance, retarding injection timing to -8 degrees CA yielded the highest indicated mean effective pressure (IMEP) and peak power, representing a 1.3% increase over the -18 degrees CA baseline. Consequently, the combination of -8 degrees CA injection timing and a 10% diesel energy fraction is identified as the optimal operating condition, successfully balancing power output and emission reduction. These findings provide critical data for developing high-pressure common-rail engines utilizing zero-carbon ammonia/hydrogen blends.
This paper investigates safe formation control and path planning for multiple unmanned surface vessels (USVs) operating in restricted and obstacle-constrained waters. A cooperative control framework is proposed by integrating nonlinear model predictive control (NMPC), an adaptive virtual-obstacle replanning strategy, and a trajectory-prioritised formation-transformation strategy. When the reference trajectory repeatedly loses progress near a narrow obstacle gap, the virtual-obstacle diameter and rollback distance are increased iteratively until a collision-free replanned trajectory is obtained or the prescribed iteration limit is reached. For narrow-waterway navigation, the formation is reconfigured from a lateral arrangement to a longitudinal or compact sequential arrangement before entering the restricted region and is restored after passage. A first-order Markov process is used to generate time-varying generalised disturbances, and the resulting trajectory and tracking errors are compared with the nominal condition. MATLAB simulations evaluate local-difficulty escape, narrow-waterway formation transformation, moving-obstacle avoidance, and extensions to a four-USV formation. Real-USV experiments further demonstrate the implementation of the formation-transformation procedure. The results support the feasibility of the proposed framework under the tested constrained-water conditions.
With the rapid growth of marine engineering activities in navigable waters, collision risk identification has become increasingly complex due to the coexistence of operational vessels and passing traffic. To address this challenge, this paper proposes a collision risk identification method based on the Especial Cautious Navigation Angle (ECNA) model for waters with ongoing marine operations. Operational scenarios are classified according to the maneuvering characteristics of operational vessels, and corresponding safety boundaries are defined. Based on these characteristics, an extended ECNA model is developed to describe the collision risk relationship between navigating vessels and operational areas. Furthermore, a collision risk scoring algorithm is introduced to quantify collision risk levels and analyse their temporal evolution. To validate the effectiveness of the proposed approach, multiple simulation experiments are conducted under different encounter conditions. The results demonstrate that the proposed ECNA model provides improved visualisation capability in complex navigational environments. The proposed method enables intuitive collision risk interpretation, supports multi-target prioritisation, and offers a practical framework for real-time collision risk assessment in waters with ongoing marine operations.
