Search

This paper explores the influence of different turbocharging modes and multi-parameter coordinated control on the performance of compression-ignition (CI) aviation piston engine (APE). Firstly, based on a constructed one-dimensional thermodynamic model of a CI APE, the study investigates the effects of different turbocharging modes and combinations of high and low-pressure stage variable geometry turbines (VGT) on the operational performance of the engine. Subsequently, a novel stepwise approximate multi-objective optimization algorithm is proposed, combining backpropagation neural networks and the non-dominated sorting genetic algorithm II. This algorithm evaluates the influence of multiple control parameters on a two-stage turbocharged engine's performance, achieving a balance between fuel economy and reliability. The research shows that equipping CI APEs with twin VGT for supercharging can notably enhance engine performance, enabling the achievement of favorable power recovery objectives at an altitude of 8000 m. The proposed optimization algorithm exhibits strong predictability and reliability, substantially accelerating the computation speed and reducing the data volume by approximately 95%. At an engine speed of 3887 rpm, compared to the unoptimized conditions, the brake specific fuel consumption of the best scenario at altitudes of 2000 m, 4000 m, 6000 m, and 8000 m is reduced by 7.1%, 7.2%, 8.6%, and 6.9%, respectively. • Influences of different turbocharging modes on the performance of a CI APE are investigated. • A novel stepwise approximate multi-objective optimization method based on a hybrid model is proposed. • Explored the altitude-varying control strategy under coordinated multi-parameter optimization. • Increased the altitude ceiling for the CI APE, reaching 82% power recovery target at 8000 m. [ABSTRACT FROM AUTHOR]

The automated berthing of vessels constitutes a high-risk operation. Equipping ships with azimuth and lateral thrusters enhances manoeuvrability, although it poses challenges in achieving optimal thrust allocation (TA) efficiency. This study introduces an improved adaptive non-dominated sorting genetic algorithm III (ANSGAIII) specifically designed to address the TA problem in autonomous ship berthing. The algorithm enhances the objective function by accounting for the nonlinear relationship between actual power consumption and thruster speed, along with considering the treatment of forbidden sectors while maintaining a high level of fidelity. Additionally, the algorithm considers equality constraints, actuator rate and position constraints, and other operational constraints. A modification to the non-dominated sorting method is implemented to improve the processing speed of the optimization algorithm. Significantly, this algorithm surpasses NSGAII, NSGAIII, and CMOEA\D in meeting ship safety and stability requirements while concurrently reducing energy consumption. • An ANSGAIII TA scheme is proposed to enhance berthing safety and energy efficiency. • Cost function and forbidden sector strategy are improved. • The non-dominated sorting method is refined to expedite the TA process. [ABSTRACT FROM AUTHOR]

Cloud computing, a technology providing flexible and scalable computing resources, faces a critical challenge in task scheduling, directly impacting system performance and customer satisfaction. The task scheduling problem's NP-completeness makes finding solutions difficult. To address this, researchers propose a hybrid algorithm combining the Grey Wolf Optimization Algorithm (GWO) and the Genetic Algorithm (GA). The hybrid GWO-GA algorithm targets multi-objective task scheduling in cloud computing, aiming to minimize makespan, energy consumption, and cost. Enhancements to the proposed algorithm involve leveraging the genetic algorithm's crossover and mutation operator. Furthermore, the GA-based GWO algorithm's faster convergence in large scheduling problems presents an advantage. Evaluation using the Cloudsim toolkit demonstrates the proposed algorithm's efficiency compared to existing methods. We have used both synthetic and real world data set. The results are verified using the Analysis of Variance (ANOVA) statistical tool. Experimental results showcase its effectiveness in minimizing makespan, energy consumption, and computational cost. Particularly, the proposed algorithm outperforms traditional GWO, GA, and PSO algorithms in terms of makespan, cost, and energy consumption, achieving reductions of 19%, 21%, and 15%, respectively, when compared to each approach. Additionally, it yields energy savings of 17%, 19%, and 23% compared to GWO, GA, and PSO, respectively, while reducing total scheduling costs by 13%, 17%, and 22%. These findings demonstrate the efficacy of the proposed algorithm in addressing the task scheduling problem in cloud computing environments. • A hybrid task scheduling strategy based on GA and GWO is proposed. • Introducing mutation and crossover to preserve wolf diversity and boost local search in the algorithm. • A novel fitness function is proposed considering multiple objectives simultaneously. • We have synthetic and real world data set for the simulation. • Results with CloudSim show the enhanced performance of proposed algorithm. [ABSTRACT FROM AUTHOR]