A Simulation-Based Multi-Objective Optimization Design Method for Pump-Driven Electro-Hydrostatic Actuators

• Main Authors: Longxian Xue, Shuai Wu, Yuanzhi Xu, Dongli Ma
• Format: Article
• Language: English
• Published: MDPI AG 2019-05-01
• Series: Processes
• Subjects: electro-hydrostatic actuator; multi-objective optimization; weight; energy consumption; rise time; dynamic stiffness
• Online Access: https://www.mdpi.com/2227-9717/7/5/274

A pump-driven actuator, which usually called an electro-hydrostatic actuator (EHA), is widely used in aerospace and industrial applications. It is interesting to optimize both its static and dynamic performances, such as weight, energy consumption, rise time, and dynamic stiffness, in the design phase. It is difficult to decide the parameters, due to the high number of objectives to be taken into consideration simultaneously. This paper proposes a simulation-based multi-objective optimization (MOO) design method for EHA with AMESim and a python script The model of an EHA driving a flight control surface is carried out by AMESim. The python script generates design parameters by using an intelligent search method and transfers them to the AMESim model. Then, the script can run a simulation of the AMESim model with a pre-set motion and load scenario of the control surface. The python script can also obtain the results when the simulation is finished, which can then be used to evaluate performance as the objective of optimization. There are four objectives considered in the present study, which are weight, energy consumption, rise time, and dynamic stiffness. The weight is predicted by the scaling law, based on the design parameters. The performances of dynamic response energy efficiency and dynamic stiffness are obtained by the simulation model. A multi-objective particle swarm optimization (MOPSO) algorithm is applied to search for the parameter solutions at the Pareto-front of the desired objectives. The optimization results of an EHA, based on the proposed methodology, are demonstrated. The results are very useful for engineers, to help determine the design parameters of the actuator in the design phase. The proposed method and platform are valuable in system design and optimization.