This paper investigates adaptive neural control methods for robotic manipulators, subject to uncertain plant dynamics and constraints on the joint position. The barrier Lyapunov function is employed guarantee that are not violated, in which Moore-Penrose pseudo-inverse term used design. To handle unmodeled dynamics, network (NN) adopted approximate dynamics. NN based full-state feedback robots proposed when all states of closed loop known. Subsequently, only robot measurable practice; output...

In this paper, a data-driven scheme of key performance indicator (KPI) prediction and diagnosis is developed for complex industrial processes. For static processes, KPI approach proposed in order to improve the performance. comparison with standard partial least squares (PLS) method, alternative significantly simplifies computation procedure. By means realization so-called left coprime factorization (LCF) process, efficient prediction, algorithms are dynamic respectively, without measurable...

Accurate prediction of remaining useful life (RUL) is great significance to the safety and reliability lithium-ion batteries, which able provide reference information for maintenance. Particle filter (PF)-based prognostic methods have been widely used in RUL batteries. However, due degeneracy particles, accuracy traditional PF not high. In this article, a novel framework based on conditional variational autoencoder (CVAE) reweighting strategy proposed predict First, CVAE algorithm described...

This paper proposes a framework for quality-based fault detection and diagnosis nonlinear batch processes with multimode operating environment. The seeks to address 1) the mode partition problem using kernel fuzzy C-clustering method, optimal cluster number will be guaranteed by between-within proportion index; 2) contribution rate method based on an improved partial least squares (PLS) model, which better performances are provided; 3) classification of online measurements hybrid PLS...

Projection to latent structures (PLS) model has been widely used in quality-related process monitoring, as it can establish a mapping relationship between variables and quality index variables. To enhance the adaptivity of PLS, kernel PLS (KPLS) an advanced version proposed for nonlinear processes. In this paper, we discuss new total (T-KPLS) monitoring. The divides input spaces into four parts instead two KPLS, where individual subspace is responsible predicting output, are utilized...

This article develops a robust fault tolerant (FT) control scheme for an n-link uncertain robotic system with actuator failures. In order to eliminate the influence of both uncertainties and failures on performance, Gaussian radial basis function neural networks are used compensate dynamics. An adaptive observer is designed external disturbance. addition, in accelerate recovery stability after failure, nonsingular fast terminal sliding mode given. Finally, simulation results two-link...

This paper is concerned with the recursive filtering problem for a class of uncertain systems amplify-and-forward (AF) relays. The parameter uncertainties are described by set norm-bounded matrices. An AF relay located between sensor and remote filter to forward signal received from filter. A random variables certain probability distribution introduced characterise transmission power transmitting measurement. By utilising average power, robust first constructed stochastic system. Then, an...

In modern industrial processes, production quality, system performance, process reliability, and safety issues have received considerable attention. This paper proposes a plug-and-play (PnP) monitoring control architecture that results in simple reliable design procedure. The proposed PnP is integrated with systems, by which performance can be enhanced without modifying or replacing the existing system. Based on architecture, disturbance compensation for rolling mills proposed. effectiveness...

This paper proposes a common and individual (CnI) feature extraction-based process monitoring (PM) method for tracking the operating performance product quality of processes with multiple modes. Different from traditional methods that separately develop PM models concerning only each mode data, new seeks to build model simultaneously all including acquire subspace captures behind different modes, reflects unique mode. The newly proposed framework is achieved using conventional principal...

In actual production processes, the occurrence probability of multiple faults is much higher than that a single fault, which will affect process industry operating performance and final products quality. This paper concerned with industrial practices theoretical approaches for detection location key indicator (KPI) related in industries. First, new KPI-related fault monitoring scheme addressed from subprocess level based on developed correlation-based canonical variable analysis model. Then,...

Due to the complex static, dynamic, and large-scale characteristics for modern industrial processes, in this article, we propose a double-layer distributed monitoring approach based on multiblock slow feature analysis independent component analysis. To end, processed dataset is divided into static dynamic blocks basis of sequential information each variable first layer. Considering correlations between variables correlation matrices two are calculated, which serves as second-layer block...

Accurate estimation of the remaining useful life (RUL) and health state for rollers is great significance to hot rolling production. It can provide decision support roller management so as improve productivity process. In addition, RUL prediction helpful in transitioning from current regular maintenance strategy conditional-based maintenance. Therefore, a new method that extract coarse-grained fine-grained features batch data predict proposed this paper. Firstly, deep learning network...

The industrial process monitoring and operating performance assessment techniques are of great significance to ensure the safety efficiency production improve comprehensive economic benefits for modern enterprises. In this paper, a new key indicator (KPI) oriented nonlinear method is proposed based on improved Hessian locally linear embedding (HLLE), in view problems strong nonlinearity, high dimension information redundancy actual data. Firstly, order characterise similarities samples both...

This article takes into account the problem of adaptive fixed-time control for nonlinear systems in a strict form via finite-time command-filtered backstepping. Our presented scheme gives consideration to rapidity by using and command filter, its prime objective is ensure that system output can be guided from any initial conditions go after an ideal variable. Meanwhile, this strategy makes sure all states other signals are bounded at finite time, convergence time does not have connection...

Quality prediction is important for precise control of industrial processes and improvements product quality. The nonlinear dynamic features time series data can be effectively extracted by the appropriate soft sensor models quality prediction. However, temporal may often considered conventional methods, ignoring spatial features, which makes not have stronger generalization abilities because insufficient feature information Moreover, due to unstable transmission signals, equipment failures,...