Advanced Materials Research Vols. 989-994

Abstract: Heavy mechanical devices of complex industrial process produce soundly mechanical vibration and acoustical signals. Some difficult-to-measure key process parameters have direct relationship with these signals. A newly ensemble empirical mode decomposition (EEMD), Fast Fourier Transform (FFT), Mutual information (MI), and Kernel partial least squares (KPLS) based modeling approach is proposed to measure these process parameters. At first, different scale intrinsic mode functions (IMFs) of mechanical vibration and acoustical signals are obtained using EEMD technology. Then, FFT transforms these multi-scale IMFs into frequency domain, and MI based feature selection method selects interesting frequency spectral features. Finally, KPLS constructs the final soft sensor models using the selected features. Experimental results based on vibration and acoustical signals of ball mill demonstrate this approach is more effective than other exist multi-scale decomposition based methods.

Abstract: Self-adapt distance measure supervised locally linear embedding solves the problem that Euclidean distance measure can not apart from samples in content-based image retrieval. This method uses discriminative distance measure to construct k-NN and effectively keeps its topological structure in high dimension space, meanwhile it broadens interval of samples and strengthens the ability of classifying. Experiment results show the ADM-SLLE date-reducing-dimension method speeds up the image retrieval and acquires high accurate rate in retrieval.

Abstract: Extreme learning machine (ELM), a relatively novel machine learning algorithm for single hidden layer feed-forward neural networks (SLFNs), has been shown competitive performance in simple structure and superior training speed. To improve the effectiveness of ELM for dealing with noisy datasets, a deep structure of ELM, short for DS-ELM, is proposed in this paper. DS-ELM contains three level networks (actually contains three nets ): the first level network is trained by auto-associative neural network (AANN) aim to filter out noise as well as reduce dimension when necessary; the second level network is another AANN net aim to fix the input weights and bias of ELM; and the last level network is ELM. Experiments on four noisy datasets are carried out to examine the new proposed DS-ELM algorithm. And the results show that DS-ELM has higher performance than ELM when dealing with noisy data.

Abstract: Lightweight, flexible motion simulation is the demand of airborne laser communication optical transceive when apply to outside test. A new parallel 2_DOF platform that has the function of azimuth and pitching is put forword based on the analysis of airplane position-pose changes affect the performance airborne laster communication APT system, and the kinematics model is established by using closed-loop vector method. Kinematics model is right through the comparison of mathematical model and simulation results of ADAMS, which provides the reference and basis for the design of control system.

Abstract: The optimal group is an important problem of histogram algorithm, and how to confirm group number has not a quantitative rule. So the concept of the close degree is imported to make the close degree between the upper contour line of histogram and the PDF(probability density function) of parameter as the judging criteria of optimal group. With the unknown of the PDF of parameter, the improved kernel density estimation algorithm can pre-select and estimate the PDF. This improved kernel density estimation algorithm combine the selection of fixed window and variable window's width to achieve the window's width automatic adjustment value between the different estimation points based on the sample distribution. In the parameter analysis of radar emitter signal, the algorithm based on improved kernel density estimation and close degree is used to determine optimal group, and the result indicate that this method is effective and can search the optimal group automatically.

Abstract: Aiming at the problem of poor performance of suppressing the wide band jamming and fast following jamming in Frequency-Hopping (FH) communication system, a communication scheme by combination of the complementally transformed minimum variance with FH technique for arbitrary array (CTMV-FH) is proposed, based on the purpose to maximize the output Signal to Interference and Noise Ratio (SINR). Basic theory of CTMV is introduced. The algorithm is used to suppress the wide band jamming and fast following jamming. Simulation results show that the scheme can suppress the wide band jamming and fast following jamming effectively in FH communication system.

Abstract: Amount of data in collecting data of fabric image in the textile industry put forward a new challenge to sensor end. Compressed Sensing (CS) breaks limit of conventional Shannon’s sampling theorem, so we can reconstruct a signal in Sub-sampling rate. In addition, theoretical analysis tells us that collecting the fabric image data by CS method have a better advantage than collecting the general image data. Having reconstructed three fabric images and one general image by CS method, we can easily find that the former have a high quality of reconstruction.

Abstract: To further improve the performance of adaptive equalization in HF channel, a new variable step size algorithm is proposed based on the iterative gradient. The step size is set to a vector according to the weights of the equalizer, and it can vary with the iterative gradient during the equalization process. The algorithm overcomes the traditional variable step size with the constant scale for all the weights of the equalizer, thus it can obtain faster convergence rate and lower steady state error in HF channel. In the complex channel condition, the algorithm proposed in this paper can avoid the local minimum point of the objective function to obtain the global convergence performance. Simulation result shows the effectiveness in the Watterson channel model.

Abstract: A new method to estimate the Doppler and multipath time delay is presented in impulse noise environment. First, the Doppler is estimated by energy cumulation of multipath component based on fractional lower-order-frational correlation transform. Then, an order-reduced signal is reconstructed combining the Doppler with the prior knowledge of the transmitted signal, and the echoes signal is converted to many single-frequency signals. Finally, the multipath time delay is obtained by the fractional lower-order power spectrum method. The method is adapted for low SNR noise, can restrain the affection of the cross-items, and has a high time-delay estimation resolution. Some computer simulations are given in this paper and the results show that the new method is valid.

Abstract: This paper takes the-stable distribution as the noise model and works on the parameter estimation problem of bistatic Multiple-Input Multiple-Output (MIMO) radar system in the impulsive noise environment.This paper presents a signal model and a novel method for parameter estimation in bistatic MIMO radar system in the impulsive noise environment. Firstly, a signal array model is constructed based on the-stable distribution model. Secondly, Doppler parameters are jointly estimated by searching the optimal rotation angle to meet concentrated-energy of the FLOS-FC. Furthermore, two algorithms are presented for the estimation of DODs and DOAs, including based on FLOS-MUSIC algorithm and FLOS-ESPRIT algorithm. Simulation results are presented to verity the effectiveness of the proposed method.