Establishing an accurate equivalent model is a critical foundation to describe the energy conversion characteristics of photovoltaic system, which can support research fault analysis, output power prediction, and performance analysis system. However, widely used models are highly nonlinear have many unknown parameters, making it difficult identify these parameters accurately. Our previous work found that gaining-sharing knowledge-based algorithm (GSK) shows promising in solving this problem....

Although several data-driven soft sensors are available, online reliable prediction of the Mooney viscosity in industrial rubber mixing processes is still a challenging task. A robust semi-supervised sensor, called ensemble deep correntropy kernel regression (EDCKR), proposed. It integrates strategy, brief network (DBN), and (CKR) into unified sensing framework. The multilevel DBN-based unsupervised learning stage extracts useful information from all secondary variables. Sequentially,...

The polishing of product parts is a difficult problem in the field industrial robots. In order to adapt robots needs polishing, it necessary effectively control contact stress generated during actual grinding process. Based on principle and method force control, this paper puts forward constant end actuator its for robot polishing. Aiming at process, relevant device are analyzed. design scheme cushion cylinder as power output element put forward, working A feedback system based sensor...

In fault diagnosis studies two main approaches are mostly used. The first one consists in designing the full physical or empirical model of system healthy and faulty conditions. major drawback this approach is difficulty to obtain an accurate reflecting all operating conditions phenomena. second approach, used work, using signal processing techniques for characterization behaviors. This paper deals with study a detection isolation procedure on three phase inverter feeding induction machine...

Optimal power flow is one of the fundamental optimal operation problems for systems. With increasing scale solar energy integrated into systems, uncertainty brings intractable challenges to system operation. The multi-objective (MOOPF) considering becomes a hotspot issue. In this study, MOOPF model proposed. Both scenarios overestimation and underestimation are modeled penalized in form operating cost. order solve optimization effectively, study proposes clustering-based differential...

In this paper, distributed model predictive control (DMPC) for island DC micro-grids (MG) with wind/photovoltaic (PV)/battery power is proposed, which coordinates all generations (DG) to stabilize the bus voltage together insurance of having computational efficiency under a real-time requirement. Based on feedback voltage, deviation current dispatched each DG according cost over prediction horizon. Moreover, avoid excessive fluctuation battery power, both discharge-charge switching times and...

Reliable uncertainty estimation plays a crucial role in various safety-critical applications such as medical diagnosis and autonomous driving. In recent years, Bayesian neural networks (BayesNNs) have gained substantial research industrial interests due to their capability make accurate predictions with reliable estimation. However, the algorithmic complexity resulting hardware performance of BayesNNs hinder adoption real-life applications. To bridge this gap, paper proposes an algorithm...

With the determination of goal "carbon peak and carbon neutrality", it is urgent for coal-fired power plants to save energy reduce consumption emissions. This paper discusses architecture design smart plants, especially taking Panji Power Plant in Anhui, China, as an example. The fivelayer architecture, including device layer, network control information management data visualization presented discussed detail. To achieve a plant, fault diagnosis health are most crucial. Thus, system...

Existing safety control methods for non-stochastic systems become undefined when the system operates outside maximal robust controlled invariant set (RCIS), making those vulnerable to unexpected initial states or unmodeled disturbances. In this work, we propose a novel framework that can work both inside and RCIS, by identifying worst-case disturbance be handled at each state constructing inputs model. We show such models jointly computed considering an invariance problem auxiliary system....

Intelligent diagnosis is the development direction of mechahnical condition monitoring and fault diagnosis.Conbined improved EEMD with SVM in intelligent researched this paper.To bearing normal as an example,impove decomposed 9D normalized energy for characteristic vector applied to binary classification identification.Compared accuracy using different kernel functions that linear,polynomial,RBF Sigmoid function.In same parameters,SVM based on linear polynomial function a hundred...

Abstract Rolling bearings play an important role in the aerospace industry, manufacturing, and nuclear engineering. To ensure reliable stable operation of various mechanical equipment, research on bearing fault diagnosis is very practical critical. With rapid development smart manufacturing industrial big data, deep learning has become effective solution for emerging identification. Due to different distributions training samples test faults, researchers have introduced many transfer methods...

The control of traffic signals is fundamental and critical to alleviate congestion in urban areas. However, it challenging since dynamics are complicated real-world scenarios. Because the high complexity optimisation problem for modelling traffic, experimental settings existing works often inconsistent. Moreover, not trivial multiple intersections properly real complex scenarios due its vast state action space. Failing take intersection topology relations into account also results inferior...