This article focuses on the low-conservative stability criteria of delayed neural networks (DNNs). To achieve this goal, new techniques are developed to effectively utilize more system-related information. To use the time-varying delay information, some delay-product terms are introduced into the Lyapunov-Krasovskii functional (LKF), and an extended matrix-injection-based transformation method, which introduces delay-derivative-dependent slack matrices while obtaining the negative definite condition, is proposed. With respect to the use of activation function information, the terms related to the activation function are fully augmented in the LKF. In particular, by considering the sector-constraint information of the activation function, a new nonlinear-function-dependent functional term is established, and a sector-constraint-dependent matrix-separation-based inequality is developed. By applying the above techniques, several improved stability criteria are derived, and two typical examples are provided to illustrate the advantages of the proposed methods.

This paper addresses the stability and stabilization issues of Takagi-Sugeno (T-S) fuzzy systems under sampled-data control. In this paper, efforts are dedicated to developing a stability criterion and control strategy with high accuracy and low complexity, leveraging an improved looped functional and a fuzzy-related matrix injection method. First, an improved looped function containing quadratic information about the sampling sawtooth characteristic is employed to include as much system state information as possible in a simple form. In response to the nonlinear terms encountered in the functional derivatives, a fuzzy-related matrix injection method is proposed on the basis of linearization transformation. This method enables the unified treatment of reciprocally convex and quadratic terms, which is conducive to reducing estimation gaps. Based on the aforementioned techniques, stability and stabilization criteria for T-S fuzzy systems are derived. The efficacy and superiority of the proposed method are substantiated through rigorous theoretical analysis and illustrative instances.

Load frequency control (LFC) is crucial for ensuring the frequency stability of power system, and its control performance is often affected by nonlinearities in devices and multisource-induced hybrid delays in control signal transmission networks. This article investigates the LFC performance analysis and improvement of power systems under the nonlinearities and hybrid delays. First, the nonlinear part is approximated by a Takagi–Sugeno (T–S) fuzzy model and the hybrid delay is divided into two regions, thus establishing a T–S fuzzy LFC model with hybrid delays. Then, based on switched system theory, a delay-dependent weighted$H_\infty$performance analysis criterion for LFC system is proposed, giving the necessary conditions to achieve a certain robustness of the system against disturbances. On this basis, a controller design method is proposed to ensure the system have the optimal level of disturbance rejection under the designed controller. Finally, case studies based on single-area and three-area LFC demonstrate that, compared with previous methods, the proposed method can obtain a larger allowable upper bound of delays and better$H_\infty$performance estimation. Based on this, the designed controller enhances the robust performance of LFC, thereby ensuring the stable operation of power systems under nonlinearities and hybrid delays.

Haze removal typically works on a physical model to estimate how light is transmitted and lost due to absorption and scattering through the atmosphere. In this paper, a region detection network is proposed to learn the relationship between the hazy image and the medium transmission map in a patchwise manner; the transmission map is then used to remove haze via an atmospheric scattering model and enhance the detail of de-hazed images. To this end, we design a simple yet powerful deep convolutional neural network, which mainly consists of two types of network units and can be trained in an end-to-end manner. One network unit is a module with the residual structure that facilitates the learning process of the deep network. The other is a novel module with a cascaded cross channel pool, which fuses multi-level haze-relevant features and boosts the abstraction ability of the model on a nonlinear manifold. Moreover, an evolutionary-based enhancement method is developed to improve the level of detail of over-smoothed results. Several comparative experiments have been conducted on synthetic and real images, through which we conclude that the proposed method achieves state-of-the-art haze removal results, qualitatively and quantitatively. Supplementary experiments further indicate that our method works better against other adverse effects on vision quality (e.g., the mist formed by heavy rain and the veil met underwater). Moreover, we present a lightweight version of the proposed network, which achieves an impressive haze removal performance even on low-power devices.