Explore the vast range of titles available.

This study addresses the standardization of Singular Distributed Parameter Systems (SDPSs). It focuses on classifying and simplifying first- and second-order linear SDPSs using characteristic matrix theory. First, the study classifies first-order linear SDPSs into three canonical forms based on characteristic curve theory, with an example illustrating the standardization process for parabolic SDPSs. Second, under regular conditions, first-order SDPSs can be decomposed into fast and slow subsystems, where the fast subsystem reduces to an Ordinary Differential Equation (ODE) system, while the slow subsystem retains the spatiotemporal characteristics of the original system. Third, the standardization and classification of second-order SDPSs is proposed using three reversible transformations that achieve structural equivalence. Finally, an illustrative example of a building temperature control is built with SDPSs. The simulation results show the importance of system standardization in real-world applications. This research provides a theoretical foundation for SDPS standardization and offers insights into the practical implementation of distributed temperature systems.

According to the well-posed problem of nonlinear singular distributed parameter systems, first of all, the nonlinear GE-semigroup induced by a continuous （possibly nonlinear） operator is introduced in Banach space, which is a generalization of GE-semigroup （i.e., generalized operator semigroup）, and the properties of nonlinear GE-semigroup are discussed; and then the existence, uniqueness and constructive expression for the strong solution of nonlinear singular distributed parameter system are discussed by nonlinear GE-semigroup; at last, the exponential stability of nonlinear singular distributed parameter system is studied by using nonlinear GE-semigroup, functional analysis and operator theory in Banach space.

In this paper, the problem of dissipativity filtering for discrete-time nonlinear singular systems with distributed delays is addressed. An interval type-2 fuzzy model is employed to represent the nonlinear singular systems of which the parameter uncertainties are captured by interval type-2 membership functions characterized by lower and upper membership functions. Based on event-triggering scheme, an interval type-2 filter is proposed. By adopting the idea of input delay method, the filtering error systems is reformulated as a new event-triggering interval type-2 fuzzy singular systems with mixed time-delays. By using an improved reciprocally convex combination approach and some new techniques on matrix convexification to bound the forward difference of the double and the triple summation terms in the Lyapunov function, two less conservatism conditions have been derived. The event-triggered dissipative filter gains and the event-triggering parameters are obtained to determine the filter error singular systems admissible and dissipative. A numerical example is given to illustrate the effectiveness of this proposed method.

This paper deals with the imaging problem from data collected by means of a microwave photonics-based distributed radar network. The radar network is leveraged on a centralized architecture, which is composed of one central unit (CU) and two transmitting and receiving dual-band remote radar peripherals (RPs), it is capable of collecting monostatic and multistatic phase-coherent data. The imaging is herein formulated as a linear inverse scattering problem and solved in a regularized way through the truncated singular value decomposition inversion scheme. Specifically, two different imaging schemes based on an incoherent fusion of the tomographic images or a fully coherent data processing are herein developed and compared. Experimental tests carried out in a port scenario for imaging both a stationary and a moving target are reported to validate the imaging approach.

In practice, the consensus performance of a multi-UAV system can degrade significantly due to the presence of measurement noise and disturbances. However, simultaneously rejecting the noise and disturbances to achieve high-precision consensus tracking control is rather challenging. In this paper, to address this issue, we propose a novel distributed consensus tracking control framework consisting of a distributed observer and a local dual-estimator-based tracking controller. Each UAV’s distributed observer estimates the leader’s states and generates the local reference, functioning even under a switching communication topology. In the local tracking controller design, we reveal that classic uncertainty and disturbance estimator (UDE)-based control can magnify the noise. By combining the measurement error estimator (MEE) with UDE, a local robust tracking controller is designed to reject noise and disturbances simultaneously. The parameter tuning of MEE and UDE is unified into a single parameter, and the monotonic relationship between this parameter and system performance is revealed by the singular perturbation theorem. Finally, the validity of the proposed control framework is verified by both simulation and comparative real-world experiments.

Long short-term memory (LSTM) is a powerful deep learning technique that has been widely used in many real-world data-mining applications such as language modeling and machine translation. In this paper, we aim to minimize the latency of LSTM inference on cloud systems without losing accuracy. If an LSTM model does not fit in cache, the latency due to data movement will likely be greater than that due to computation. In this case, we reduce model parameters. If, as in most applications we consider, the LSTM models are able to fit the cache of cloud server processors, we focus on reducing the number of floating point operations, which has a corresponding linear impact on the latency of the inference calculation. Thus, in our system, we dynamically reduce model parameters or flops depending on which most impacts latency. Our inference system is based on singular value decomposition and canonical polyadic decomposition. Our system is accurate and low latency. We evaluate our system based on models from a series of real-world applications like language modeling, computer vision, question answering, and sentiment analysis. Users of our system can use either pre-trained models or start from scratch. Our system achieves 15× average speedup for six real-world applications without losing accuracy in inference. We also design and implement a distributed optimization system with dynamic decomposition, which can significantly reduce the energy cost and accelerate the training process.