Jin Weidong

Novel Quantum Genetic Algorithm and Its Applications

By introducing strong parallelism of quantum computing into evolutionary algorithm, a novel quantum genetic algorithm (NQGA) is proposed. In NQGA, a novel approach for updating the rotation angles of quantum logic gates and a strategy for enhancing search capability and avoiding premature convergence are adopted. Several typical complex continuous functions are chosen to test the performance of NQGA. Also, NQGA is applied in selecting the best feature subset from a large number of features in radar emitter signal recognition. The testing and experimental results of feature selection show that NQGA presents good search capability, rapid convergence, short computing time, and ability to avoid premature convergence effectively.

A Novel Feature Extraction Approach for Radar Emitter Signals

Feature extraction is the crucial technology to deinterleave and recognize the new system radar emitter signals. In this paper, a novel time-frequency atom feature extraction approach is presented. Based on the over-complete multiscale dictionary of Gaussian Chirplet atoms, adopting match pursuit (MP) to decompose signals and the improved quantum genetic algorithm (IQGA) to reduce the search time for MP, the optimal Chirplet atoms to represent the feature information of the radar emitter signals can be obtained. The validity and feasibility of the approach was proved by using fewer Chirplet atoms to acquire more accurate feature information compared with Gabor atoms approach.

Statistical background subtraction with adaptive threshold

Detection of moving objects in surveillance video is the first relevant step of information extraction for many applications such as tracking and recognition. We present a new algorithm for the purpose of robust foreground detection using a statistical representation of the scene background. The weighted kernel density estimation is applied for each pixel by the analysis of temporal distribution in background initialization phase. Based on kernel density estimation, an adaptive threshold approach is demonstrated to estimate foreground threshold automatically. Significant improvements are shown on both synthetic and real video data. The incorporating adaptive threshold into the statistical background for background subtraction leads to an improved segmentation performance compared to the standard methods.

The application of multi-criterion satisfactory optimization in fuzzy controller design

This paper presents a new approach to the fuzzy controller design. After review of standard fuzzy system operation, a new optimization model called multi-criterion satisfactory optimization model is presented. The disadvantages of traditional membership functions and controller rule base optimization are introduced. The approach that uses multi-criterion satisfactory optimization model in fuzzy controller design is illustrated with several control systems. The results of simulation demonstrated our method is useful for the fuzzy controller design.

A study on intelligent computation of methods of optimization operation for train

Energy-saving train operations are significant both in theory and in applications, but computing the optimization of train operations is very difficult and complex. The optimization computation problem of energy-saving train operations on an undulating-slope line is discussed by means of an intelligent computation model in this paper. To generate the optimal train operation diagram, an intelligent computation model combining local optimization with global optimization is proposed. The local optimizations numerical functions are obtained from the simulation computation, and the construction of those data is realized by a neural network. The global optimization computation, using a genetic algorithm, generates the train operation diagram. Theoretical analysis and simulation experiments show that the result is satisfactory. Moreover, compared to other methods, not only is the energy saving greater but the computational efficiency is greatly improved too.

Feature Extraction of Radar Emitter Signal Based on Wavelet Packet and EMD

To research the problem of radar emitter signal (RES) recognition, and to further enhance emitter signal recognition capacity of the electronic warfare equipment, the empirical mode decomposition (EMD) theory and wavelet packet (WP) are introduced into feature extraction of radar emitter signal. A new feature extraction method of radar emitter signal is proposed based on wavelet packet and empirical mode decomposition theory. First, it uses wavelet packet to finish decomposition, de-noising and reconstruction of the RES. Then it will obtain the intrinsic mode function (IMF) through EMD method, which can incarnate the characteristics of RES. The energy of each IMF are calculated and normalized, which would be regarded as the feature vector. Finally it realizes the classification of radar emitter signal by constructing BP neural network classifier. Experiment results show that the feature extraction method based on wavelet packet and empirical mode decomposition is an effective feature extraction method for RES, which can achieve satisfying correct recognition rate in a larger signal to noise ratio. It would have certain reference value in follow-up in-depth study.

The application of multi-criterion satisfactory optimization in FIR digital filter design

The paper presents a new approach, which uses multi-criterion satisfactory optimization (MCSO) in digital filter design. We use the frequency-sampling method in finite impulse response (FIR) digital filter design, and use multi-criterion satisfactory optimization to optimize the unspecified frequency samples occurring in the transition bands. The results show that our method is useful for FIR digital filter design.

Feature Extraction of Radar Emitter Signal Based on Empirical Mode Decomposition

To enhance accurate recognition rate of radar emitter signal (RES), a novel feature extraction method of radar emitter signal is proposed based on empirical mode decomposition (EMD) theory. The EMD algorithm is used to decompose the radar emitter signal into a number of intrinsic mode functions (IMF) and a residue component, these IMFs can reflect characteristics of the radar emitter signal. After that, the energy of each IMF is calculated and normalized, which is regarded as an element of the feature vector. Finally, it realizes the recognition of radar emitter signal through BP neural networks (BPNN). Experiment results shows that EMD-based feature extraction method of radar emitter signal is an effective method, energy feature that extraction from EMD decomposition has a higher recognition rate. The main work of this paper is that it applied the EMD method to feature extraction of radar emitter signal for the first time.

A fast adaptive fuzzy median filter for removal of impulses from color images

Median filter has been widely used for restoration of noise image. In this paper, we present a novel adaptive fuzzy median filter to the restoration of salt & pepper impulse noise-corrupted image, which is particularly effective at removing highly impulsive noise. First a fuzzy estimation phase is based on the calculation of the sum of eight direction gradient values (for each color component), then an extensive window is selected for the best tradeoff between noise suppression and detail preservation. The proposed filter has demonstrated preferable performance in suppressing different noise ratio of color image noise from 10% to 90% noise ratio. Besides, the median computation method based on histogram and multilevel staged search is used to improve the speed of the filtering process.

The application of multicriterion satisfactory optimization in computer networks design

The computer networks design is a nonlinear combinatorial optimization problem with constraint set identical to that of the multiple choice multiconstraint knapsack problem, which is known to be NP-complete. We present a new approach in which a multicriterion satisfactory optimization is used in the computer networks design. The optimal computing model is proposed. The satisfactory rate function of the criteria, which represents the importance of performance specification, and the synthesis satisfactory rate function representing the optimization are designed. An improved genetic algorithm is used for optimization computing. Computational experience shows that this method is efficient and effective.