Search

In order to enhance the chaotic degree of cellular neural network (CNN), the memristive characteristic is combined in CNN, and a five-dimensional memristive CNN hyperchaotic system is designed. A secure communication system based on chaos synchronization is constructed by the full-dimensional state observer theory. The sliding mode control method is used to control chaos synchronization between the sending end and the receiving end and improve the robustness to the parameter uncertainties and disturbances in the system. An improved sliding mode surface is designed, and its convergence time is analyzed. Image secure communication, as an example, is given to verify the effectiveness of the proposed method. Image encryption is implemented by the hyperchaotic system, and image decryption is implemented by chaos synchronization based on the sliding mode control. Simulation experiment results show that the proposed sliding mode control method can be used to achieve the chaos synchronization of the hyperchaotic systems with parameter uncertainties and disturbances, and the proposed memristive hyperchaotic system can be applied to the secure communication based on the chaos synchronization.

Weibull clutter is used as an example in this paper. Based on the serial parallel analysis of Zero-memory non-linear transformation's Weibull distributed clutter algorithm, fine-grained optimization is performed. The fine-grained part uses the cuBLAS library to optimize the performance of convolution calculations. Compared with CUDA shared memory convolution method and GPU parallel matrix multiplication convolution method, its computational performance can be significantly improved under a large amount of data. Simulation results show that the Zero-memory non-linear transformation's Weibull distributed clutter simulation method is optimized and accelerated. The real-time performance of clutter data is significantly improved and its acceleration effect will be better as the amount of clutter data increases. It turns out that through fine-grained optimization, the performance of convolution calculations with large amounts of data is improved.

To extend the application of the cognitive associative network and improve the performance of the target tracking method, a novel tracking method based on the new network is proposed, i.e. the cognitive associative network is transformed and used to perform the target tracking. The local hue histograms are used to model the target, and the circle matching criterion is used to locate the target, which can restrain the disturbance from the adjacent regions in the background. It can also adjust adaptively the size of the tracking target according to the local matching results. Therefore, the tracking method based on the cognitive associative network has good invariability to the scale, the rotation, and the partial occlusion. Simulation results show that the tracking method can perform the target tracking in the disturbance environment or in the scene of the complex motion. The tracking method can locate the target more accurately than the common tracking methods such as the Camshift method or the compressive tracking method.

In order to extract more texture and detail information from a given image, a memristor cell neural network is proposed by replacing the state resistances of the neurons as the memristors. The characteristic of memristor could be maintained in the whole information processing of the cell neural network so that the historical state information could be used and the information processing ability of the network could be enhanced. Furthermore, based on the fractional order calculus theory, a fractional order control template is designed for the memristor cell neural network to enhance the middle- and high-frequency information and retain more low-frequency texture information in the image edge extraction. The network could be used to extract edge information from various kinds of images. The simulation results show that the edge images extracted by this method have more complete and clear contour information and more abundant texture detail information. Both the average gradient and the information entropy of the edge images could be significantly improved.

In order to improve the noise reduction performance and the clarity of denoising images, a composite convolutional neural network composed of the convolutional autoencoder network and the feature reconstruction network is proposed. Multiple convolutional layers are added into the autoencoder to extract the image feature information and improve the denoising performance, and the feature reconstruction network is designed to recover the texture and detail information of the image. The cross-connected structure is used to fuse feature information in the convolutional autoencoder network into the feature reconstruction network. Experimental results show that the proposed method has better noise reduction performance than the existing methods for different noise intensity. More texture and detail information could be retained, and the clearer denoising images could be obtained.

The border regression is a key technique of the regional convolution neural network (CNN) to locate the target. However, it relies on the border label information of a large number of sample data. Therefore, it is inefficient to generate the training sample set, and the location of the target is also inaccurate. For this, a novel target detection method based on the CNN and the particle search is proposed. A small number of probe particles are generated to roughly locate the target. The CNN is used to extract the image features, determine the target probability, and recognize the pattern of the target. A large number of searching particles are placed near the region where the target features are detected by the probe particles. The nearest neighbor clustering algorithm is used to classify the particles, which are recognized as the same category into different target sets. The positions of the targets can be determined by the bounding rectangle of the searching particles in the same target set. The method can be used to recognize and locate various kinds of targets. Furthermore, the method need not label the borders of the targets in the training samples, which enhance the generation efficiency of the samples. The simulation results show that the correctness of the recognition can be slightly improved, and the accuracy of the location can be significantly improved.

Brain-computer interfaces are a new pathway for communication between human body and the external environment. High classification accuracy for motor imagery electroencephalogram (EEG) signals is desirable by improving the algorithm of feature extraction and classification. A novel algorithm (VLPSO-MFDF) based on the variable length particle swarm optimization (VLPSO) and multi-feature deep fusion (MFDF) is proposed. First, each layer of the deep forest is reconstructed into two same classification modules. Then, several different features are extracted for the motor imagery EEG signal to feed separately to the classification modules. The VLPSO is used to search for the optimal weights for the probability vectors output by each classification module, which can continuously optimize the classification performance. Experimental results demonstrate that the VLPSO-MFDF algorithm can achieve higher classification accuracy for four classifications of motor imagery EEG signals compared with the traditional deep forest algorithm. The proposed method fused multi-domain features and corrected the prediction difference. It was of great significance for improving the performance of the classifier.

Aiming at the problem that CV model cannot segment the image with uneven grayscale well, a method combining differential information with CV model is proposed. Sobel operator is used to detect the edge information of the target. It adds differential information to grayscale values. Sobel operator is combined with CV model. The energy functional is constructed and grayscale mean of that is replaced by the grayscale mean with differential properties. The curve is approached to the target contour until the energy functional converges to the minimum value. Finally, the complete target is segmented. The experimental results show that the improved method can improve the segmentation effect of images with uneven grayscale.

Motor imagery brain-computer interface (MI-BCI) provides a novel way for human-computer interaction. Traditional neural networks often use serial structure to extract spatial features when dealing with motor imagery EEG signal classification, ignoring temporal information and a large amount of available information in the middle layer, resulting in poor classification performance of MI-BCI. A neural network feature fusion algorithm is proposed by combining the convolutional neural network (CNN) and the long-short-term memory network (LSTM). Specifically, the CNN and LSTM are connected in parallel. The CNN extracts spatial features, the LSTM extracts temporal features and the flatten layer added after the convolutional layer extracts the middle layer features. Then all the features are merged in the fully connected layer to improve the accuracy of classification. The average accuracy and Kappa value of all subjects were 87.68% and 0.8245, respectively. The result shows that the feature fusion neural network proposed in this paper can effectively improve the accuracy of motor imagery EEG, and provides new ideas for the study of feature extraction and classification of motor imagery brain-computer interfaces.

In order to improve the control performance of the transport vehicle, the dynamic model, instead of the kinematic model, is established. The equation of state can be divided into two independent subsystems: the speed sub system and the attitude angle sub system. The sliding mode controls based on the dynamic model are designed for the two subsystems to reduce the complexity of the overall system, and the armature voltages of the motors are used as the control inputs of the system of the transport vehicle. The disturbances, for instance the fluctuations in the power supply, are considered, so the control law has robustness to the external disturbances. Simulation results show that the control method can control the transport vehicle to move according to the desired signals, and the validity of the method can be proved.

In order to improve the versatility of the control method of the four-rotor aircraft, a general mathematical model, the rectangular four-rotor aircraft, is modeled, and two special cases, square cross structure and square X structure, are deduced. Based on the conventional global sliding mode control, an improved global sliding mode control is proposed to control the position and the attitude of the four-rotor aircraft. The dynamic sliding mode surface of the improved global sliding mode control can evolve into the linear sliding mode surface in a limited time by changing the decay function of the dynamic sliding mode surface. In this way, the controlled system has not only the strong global robustness but also the quick response rate. Simulation results show that the position and the attitude of the four-rotor aircraft can be controlled by the improved global sliding mode control, and the control performances, for instance the response speed, can be improved.

In order to improve the physical realization of the chaotic secure communication, a chaos masking encoding method based on the synchronous control of chaotic neurons with different structure is proposed. The conventional chaotic neuron is set as the transmitting system, and the hysteretic chaotic neuron is set as the receiving system. Hysteretic parameters have creep characteristic which make the hysteretic chaotic neuron have uncertainty. Sliding mode control is used to perform the synchronous control, and fuzzy inference is used to reduce the chattering of the sliding mode control. Simulation results prove the effectiveness of the method.

Acoustic emission (AE) source localization is one of the important purposes of nondestructive testing. The localization accuracy reflects the degree of coincidence between the identified location and the actual damage location. However, the anisotropy of carbon fiber three-dimensional braided composites will have a great impact on the accuracy of AE source location. In order to solve this problem, the time-frequency domain characteristics of AE signals in a carbon fiber braided composite tensile test were analyzed by Hilbert&ndash, Huang transform (HHT), and the corresponding relationship between damage modes and AE signals was established. Then, according to the time-frequency characteristics of HHT of tensile acoustic emission signals, the two-step method was used to locate the damage source. In the first step, the sound velocity was compensated by combining the time-frequency analysis results with the anisotropy of the experimental specimens, and the four-point circular arc method was used to locate the initial position. In the second step, there is an improvement of the Drosophila optimization algorithm, using the ergodicity of the chaotic algorithm and congestion adjustment mechanism in the fish swarm algorithm. The smoothing parameters and function construction in the probabilistic neural network were optimized, the number of iterations was reduced, the location accuracy was improved, and the damage mode of composite materials was obtained. Then, the damage location was obtained to achieve the purpose of locating the damage source.

In order to improve the efficiency of industrial wastewater treatment system and the immature operation management experience, an IoT-based sewage treatment monitoring system was designed. The system consists of an application layer, a network layer, and a sensing layer. The sensing layer collects the parameters of the plant wastewater through a variety of sensors and equipment components. The collected data is uploaded to the network layer and connected by the GPRS technology through the gateway, and then the data is processed and stored by using the Baidu cloud platform, and finally applied to the actual, and can be viewed by the client at any time. It provides a complete solution for the operation and management of sewage treatment plants and has high reliability in practical applications.

In order to improve the effect of image stitching and eliminate the gap between images caused by the uneven local brightness of the image in the process of stitching, an improved gradual fusion algorithm is proposed. Compared with linear gradual fusion method, the improved gradual fusion method has a non-linear change in weight, so it achieve no stitching gap and smooth transition at the stitching point .Surf algorithm is used for feature point detection and Hessian matrix is used to obtain feature points. The feature points are matched by Euclidean distance, and the improved gradual fusion algorithm is used for image fusion to achieve the smooth transition. The experimental results show that the improved algorithm could improve the effect of image fusion significantly.

A hysteretic chaotic neural network with the uncertain term is constructed, and a novel control strategy based on fuzzy sliding mode is proposed to control its dynamic behaviors without changing its structure. The control law contains fuzzy inference term which can reduce the chattering on the sliding mode. The system stability is proved by the Lyapunov theory. The hysteretic chaotic neural network based on the fuzzy sliding mode control can be used to resolve the function optimization problem without destroying chaos generation mechanism and changing the structures of the neural network, which is different to the common annealing optimization strategy. Simulation results prove that the control strategy is a valid method to control the dynamic behaviors of the neural network and resolve the function optimization problem. A hysteretic chaotic neural network with the uncertain term is constructed.A novel control strategy based on fuzzy sliding mode is proposed.Fuzzy inference term can reduce the chattering on the sliding mode.Function optimization problem is resolved by the hysteretic chaotic neural network.

A chaotic neuron with hysteretic and creep characteristics is proposed based on the conventional chaotic neuron model, with which a neural network is constructed, and the synchronous control between the hysteretic creep chaotic neuron or neural network and the deterministic chaotic neuron or neural network is investigated by the sliding mode control. The hysteretic activation function of the neuron is constructed by shifting the sigmoid function. The hysteretic parameters have creep properties, which lead to the uncertain responses of the neurons. The equivalent sliding mode control law can be designed according to the error equation of the synchronous system, and the Lyapunov method is used to prove the stability of the system. Furthermore, fuzzy sliding mode control law is designed to restrain the chattering, to reduce the synchronous error, as well as to shorten the transition time. The validity of this method has been proved by simulation experiments.

In order to improve the engineering feasibility of the memristive cellular neural network, a new memristor model with the smooth characteristic curve is designed. Based on the new memristor model, a new four-dimensional chaotic memristive cellular neural network (CNN) system is constructed, and its chaotic dynamic behaviors are analyzed. It can be applied to the secure communication based on the chaos synchronization control. Because both the external disturbances and uncertainties of internal parameters are maybe in the practical secure communication system, sliding mode control is used to perform the chaos synchronization between the sender and receiver. A new terminal sliding mode surface is designed to make the error system converge to zero in a finite time. Simulation results show that the new terminal sliding mode control has good robustness to the external disturbances and uncertainties of internal parameters, and the new chaotic memristive CNN system can be used in the secure communication by the chaos synchronization based on sliding mode control.