Guan Shouping

An Adaptive Filter Model Based on Wavelet Transform

Due to the fact that it is not easy to filter out the overlap noise between noisy signal and noise using the traditional method of wavelet denoising, an adaptive filter model based on the wavelet transform is constructed. In this model, the adaptive filter is used to filter out noise secondary on the basis of first wavelet denoising on the original noisy signal. The experimental results show that the model can effectively remove the noise. And comparing with the original method of wavelet denosing, the SNR of the signal which uses the model to filter out noise is much higher, reaching a better filtering effect.

On the simulation application for teaching practice in the course of motion control system

Computer simulation technology, which with application of software platforms likes FLASH, MultiSim and MATLAB/SIMULINK, was introduced into teaching progress, considering the existing teaching situation and features of the motion control system. By combining the theory explaining and simulation results together, it become more easy to reveal and exposit the essence of the lecture, more specific and straightforward to make the teaching process become, moreover, it help students to understand abstract theory and key segments of the motion control system better and stimulate their learning positivity, thereby achieve a outstanding teaching result.

A new evaluation strategy-based interval optimization algorithm and its simulation analysis

This paper presents a new interval optimization algorithm (ESIA) combining interval algorithm with evolution strategy in bionics., to improve the search efficiency and make the accelerated tool constructed easier comparing with the traditional interval algorithm (IA), hence it can be applied to high dimensional optimization problems better. The ESIA employed the evaluation strategy to construct accelerated tool, which can be used to cut off the interval elements with low probability of including the global optimal solution, and a reliable upper bound is provided to prune intervals and the calculation of the algorithm is reduced. Meanwhile, a new splitting rule is proposed to make the reliable interval, which probably contains the global optimal solution, have more chance to split, so as to further improve the search efficiency. The numerical experiments on several typical test functions show that the ESIA is more efficient than traditional IA.

Structure and algorithm of interval RBF neural network

This paper presents a structure and learning algorithm for interval Radial Basis Function (RBF) neural network. The subtractive clustering algorithm combined with the BP algorithm is used to train the neural network, which can cluster properly based on the set of interval data, and leading to obtain both the parameters of radial basis function and the number of clustering center, and to improve the mapping capability of neural network. The simulation results show that the converging and the approximating ability of the interval RBF are both better then the interval BP neural network.

Multivariable inverse control for glutamic acid fermentation process

To improve the performance of the glutamic acid fermentation process, a multivariable adaptive inverse control method is proposed considering nonlinear and multivariable coupling of the fermentation process. According to the characteristics of the fermentation process, the kinetics model and the simplified model are given and the reversibility of the model is analyzed. An adaptive inverse control system of glutamic acid fermentation process is designed based on the research of multivariable adaptive inverse control, and a disturbance eliminator is designed in a practical engineering system to further improve control accuracy. Simulation results show that the proposed method achieved a very good control in glutamic acid fermentation process, the disturbance has been well suppressed, the system has strong robustness, and the structure is simple and easy to project implementation.

Condition division method for complex processes based on the Modified Fuzzy C-Means clustering algorithm

A kind of Modified Fuzzy C-Means (MFCM) clustering algorithm is presented to improve the problems of the conventional Fuzzy C-Means (FCM) clustering algorithm from three aspects: the way of clustering centre selection, application of the method of weighted dot density and the theory of information granularity. Then this new algorithm MFCM solves the problems suffer from FCM algorithm such as the sensitivity to initial value, the slow convergence speed, the possibility to fall into local optimal solution, the lost of best clustering number and equivalence partition and so on. Based on MFCM algorithm, a new condition division method for complex processes is proposed and applied to the glutamic acid fermentation process. The satisfactory simulation results are obtained and illustrated in the end of the paper.

Robust state and input simultaneous estimation for continuous linear time-varying uncertain systems

State and input simultaneous estimation for continuous linear time varying systems with norm-bounded parametric uncertainty are addressed in Hinfin setting. Using linear quadratic game formulation, sufficient solvable conditions for the problem are presented in terms of solution to two Riccati equations. One possible estimator is then presented with separation innovation structure, where innovation is used to update state observation tuned by a gain matrix and simultaneously to provide input estimation through a projector matrix. With state and input simultaneous estimation ability, the proposed estimator has a wide application in control, filtering, signal processing and fault diagnosis.