Huang Wei

Optimization of sliding mode controller for double inverted pendulum based on genetic algorithm

The double inverted pendulum system is a less-driven, multi-parameter, highly nonlinear system. The sliding mode controller is designed to make the double pendulum stable at the upright equilibrium position. But, it is difficult to determine the sliding surface in design, which is the key to the performance of the control system. The improved genetic strategy is applied to search the optimal sliding surface and other important parameters. The improved genetic strategy effectively improves the ability of the global convergence and local search capabilities, enhancing the diversity of the population and avoiding the premature convergence. Simulation experiment shows that the sliding mode controller optimized by the improved genetic algorithm stabilizes the double inverted pendulum with satisfied performance.

A dynamic allocation strategy of bandwidth of networked control systems with bandwidth constraints

This paper proposes a dynamic allocation strategy for networked control system with two constraints. The first constraint confines the bandwidth of each control loop in a bounded interval, and the second one makes the overall bandwidth allocation not beyond the network capacity. The scheme contains four steps. In the first step, the error of each control loop is estimated using the mean absolute error. In the second step, a desired bandwidth algorithm is designed by referring to the estimated error, which satisfies the first constraint. In the third step, dynamic weight is determined via quadratic mapping method. In the forth step, the ideal bandwidth allocation strategy is presented for ensuring the second constraint. Simulation results illustrate the effectiveness of the proposed allocation strategy.

Robust synchronisation of uncertain chaotic neural networks with time-varying delay via stochastic sampled-data controller

This paper addresses the robust synchronization problem for chaotic neural networks with time-varying delay and parameter uncertainties. For simplicity, assuming two different sampling intervals whose happening probabilities satisfy Bernoulli distribution and are known constants. The sampling system is changed into a continuous system by employing an input delay scheme. Sufficient conditions for the robust mean-square synchronization are proved by constructing a proper Lyapunov functional. In addition, it is easily to validated the obtained results for their feasibility via MATLAB and other tool boxes. One simulation example is applied to show the effectiveness of proposed results.

Design and research of dynamic partial reconfiguration for industrial Ethernet

Consumers have to purchase many spare parts to satisfy the maintenance and replacement of different network by reason that there is great deal of communication network in industrial fields nowadays. Aimed at solving this problem, a new method of dynamic partial reconfiguration for industrial Ethernet is proposed in this paper. Firstly, the reconfigurable network node is designed based on dynamic partial reconfiguration technology for two industrial Ethernet protocols (EtherCAT and Powerlink). Secondly, communication state between slave station and master station can be observed by building hardware platform base on FPGA and getting real-time waveform based on Signal Tab. The obtained real-time waveform and observations show that the proposed method achieves good performance.