Capacity estimation of torque converters with piston holes using the response surface method and an artificial neural network

Chun, Doo‐Man; Lee, J. C.; Yeo, Jun-Cheol; Ahn, Sung‐Hoon; Hong, Sungwha; Jang, Jiseon

doi:10.1007/s12239-011-0032-x

Cited by 2 publications

(1 citation statement)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Characterized by the high generalization ability and strong robustness, the artificial neural networks can approximate the nonlinear functions with a high precision and have an obvious advantage in solving abstract mathematical problems. 29 In view of this, the radial basis function (RBF) neural network will be adopted to fit the intrinsic mapping between the design variables and the multi-objective functions. The basic implementation steps are given as follows.…”

Section: Multi-objective Optimization Problemmentioning

confidence: 99%

Multi-objective design optimization of power-cycling hydrodynamic mechanical transmissions

Liu

Sun

Qin

et al. 2018

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

The power-cycling hydrodynamic mechanical transmissions have the advantages of continuously adjustable speed ratio and high efficiency compared with the traditional automatic transmissions, so they may be a good substitute for the prior art. For off-highway vehicles which frequently work in low speed and confront great resistance, the zero-speed-ratio torque ratio (TR) of the power-cycling hydrodynamic mechanical transmissions represents the power performance of the vehicles. Furthermore, the complexity of the transmissions is an indispensable consideration for industrial designers. The radial and axial dimensions of the power-cycling hydrodynamic mechanical transmissions are determined by the effective diameter of the torque converter’s circuit and the number of transmissions gears, respectively. In order to optimize the zero-speed-ratio TR and the complexity of the power-cycling hydrodynamic mechanical transmissions, a design methodology is proposed. Considering that there is no explicit mathematical relationship between the design variables and the multi-objective functions, the parametric design and numerical simulation for the torque converter are carried out. The intrinsic mapping between the design variables and the multi-objective functions is fitted by the radial basis function neural network. On this basis, the fast and elitist non-dominated sorting genetic algorithm (NSGA-ΙΙ) is used to solve the multi-objective optimization problem. The numerical simulation for one group of solution selected from the Pareto optimal solutions is conducted. The simulation results indicate that the design methodology proposed in this study is effective. The optimal results show that the zero-speed-ratio TR of the power-cycling hydrodynamic mechanical transmissions is heavily influenced by the radial and axial space of such transmissions. The design optimization helps to find the optimal solutions for the power-cycling HMTs, which are superior to the traditional automatic transmissions and match well the prime mover.

show abstract

Section: Multi-objective Optimization Problemmentioning

confidence: 99%