Shen Yajing

Adaptive PID controller based on online LSSVM identification

The paper proposes a proportional-integral-derivative (PID) controller based on least squares support vector machines (LSSVM) identifier (PID_LSSVMI). The PID parameters of the adaptive controller are adjusted by gradient information of LSSVM for online identification of nonlinear time-varying system. To illustrate the efficiency of the proposed controller, the simulation experiments are made to compare performance of three controllers, namely, PID_LSSVMI controller, classical PID controller, and PID controller based neural network. The results show that the proposed control method is effective and can achieve better control performance in control loop of nonlinear time-varying system.

Application of flexible tolerance genetic algorithm for optimum design of double-crank mechanism

A hybrid method, a flexible tolerance genetic algorithm (FTAGA), is applied in this paper to solve a complicated engineering problem concerning synthesis optimization of a double-crank mechanism. FTAGA is based on the combination of adaptive genetic algorithm (AGA) and flexible tolerance method (FTM) and exploits the advantages of both optimization algorithms. It can efficiently and reliably obtain more accurate global optima for complex, nonlinear, high-dimension, and multimodal optimization problems subject to nonlinear constraints. The successful use of FTAGA for the optimum design of a double-crank mechanism demonstrates that FTAGA is applicable to solve more real-world problems.