Search

In this article, the adaptive neural control problem is studied for the input-delayed stochastic nonlinear system that has both states and input quantization. The main difficulty is to deal with the discontinuity of the virtual control laws caused by states quantization, and to compensate the influence of the input quantization by building the relation between the control input and quantized input. Moreover, this paper considers a sector bounded quantizer and proposes a new adaptive neural network (NN) control algorithm to handle the problem of quantization errors cannot be automatically bounded. For the unknown function in stochastic strict-feedback systems, the radial basis function neural network (RBFNN) is utilized to approximate it. The Pade approximation technique is employed to tackle the issue of input delay. All the closed-loop signals can maintain semi-globally uniformly ultimately bounded (SGUUB) in probability under the constructed controller and quantizer. Lastly, the efficiency of the developed control method is evaluated by a simulation experiment.

This paper proposes combining an improved particle swarm optimization and Pareto archive algorithm to solve the multi-objective reactive power optimization problem. The idea of $\varepsilon $ -greedy strategy is adopted and designed to improve particle swarm optimization algorithm. It makes some particles have stronger global search capability, meanwhile, others have stronger local search capability during the whole iteration process. Henceforth, the strategy significantly explores the possibility of optimal solution in local space at the early stage of the iteration, in addition, it mitigates the tendency to fall into the local optimal solution at the later stage of the iteration. The Pareto optimal solution selection problem is solved by minimizing the sum of the difference between each objective function and its optimal solution. The proposed approach is tested on IEEE39-bus and IEEE118-bus system, and it is demonstrated that the proposed approach not only restores the nodes voltage to the normal range and achieves better value for each objective function, but also outperforms other algorithms including standard particle swarm optimization and non-dominated sorting genetic algorithm II(NSGA-II).

With the coming of artificial intelligence and the era of big data, convolutional neural network (CNN) has become one of the research hotspots in many scientific fields. However, there exist serious edge information loss problems in the information transmission process of CNN. Therefore, in order to suppress the loss of important information during the information transmission process, new methods are proposed which use the singular value decomposition (SVD) algorithm based on the phase space reconstruction to analyze the bearing vibration signal. Singular values are regarded as the features for evaluating the bearing's health condition. Then, the features radiate from the center around to form a spiral matrix as the input of the CNN, which can effectively resist information loss problems during the information delivery process. In order to verify the performance of the proposed methods better than the conventional ones, experiments are carried out. This paper shows that the proposed methods have excellent performances in the field of bearing fault diagnosis.