As a classic multivariate statistical method, principal component analysis(PCA) has been widely used in monitoring industrial processes, but it is still necessary to make improvements in having a timely and effective ...
详细信息
As a classic multivariate statistical method, principal component analysis(PCA) has been widely used in monitoring industrial processes, but it is still necessary to make improvements in having a timely and effective access to variation information. It is known that the transformation matrix generated from real-time PCA model indicates inner relations between original variables and new produced components, so this matrix would be different when some variables deviate from the original values area due to the change of the operating condition. Based on this theory, this paper proposes a novel real-time monitoring approach which utilizes polygon area method to measure the variation degree of the transformation matrices and then constructs a statistic for monitoring purpose. The on-line data are collected through a combined moving window(CMW), which contain normal and monitored data at the same time in order to distinguish the faulty data. To evaluate the performance of the proposed method, a simple numerical simulation and the classic Tennessee Eastman process are employed for illustration, with some PCA-based methods used for comparison.
Moving window local outlier probability(MWLo OP)is an outlier detecting method which was proposed for monitoring time-varying industrial processes;however,for the practical industrial processes,besides the time-varyin...
详细信息
Moving window local outlier probability(MWLo OP)is an outlier detecting method which was proposed for monitoring time-varying industrial processes;however,for the practical industrial processes,besides the time-varying characters caused by deactivation of catalyst,measuring instrument drifting and so on,the operation mode is often switched as the adjusting of the feedstock,changes in market demands and so *** the WMLo OP algorithm can deal with the time-varying process data,the multi-mode process data will lead to a mass of fault *** solve this problem,an external analysis moving window local outlier probability(EA-MWLo OP)algorithm is proposed in this *** external analysis is employed to eliminate the influence of operation mode change on the process data,then the MWLo OP method can deal with complex distribution time-varying data,and give an outlier ***,the corresponding statistic and control limit are constructed to detect the process *** addition,while the monitoring model updated,the control limit is not necessary to *** performance of this method is evaluated through a case study of a non-isothermal continuous stirred tank reactor(CSTR).
Measurement data usually do not reflect actual chemicalprocesses correctly because of inevitable errors in measurement, which is known as unbalance of measurement data. Data reconciliation and gross error detection a...
详细信息
Color night-vision technology increases the representation ability of monochrome night-vision imagery by adding color to it, making observers’ understanding easier. Usually the color night-vision methods require the ...
详细信息
The flow shop scheduling problem with limited buffers is widely existing in manufacturing systems. This article proposes a hybrid discrete harmony search algorithm for the problem to minimize total flow time. The algo...
详细信息
The flow shop scheduling problem with limited buffers is widely existing in manufacturing systems. This article proposes a hybrid discrete harmony search algorithm for the problem to minimize total flow time. The algorithm presents a novel discrete improvisation and a differential evolution scheme with the job-permutation-based representation. Moreover, the discrete harmony search is hybridized with the problem-dependent local search based on insert neighborhood to balance the global exploration and local exploitation. In addition, an orthogonal experiment design is employed to provide a receipt for turning the adjustable parameters of the algorithm. Comparisons based on the Taillard benchmarks indicate the superiority of the proposed algorithm in terms of effectiveness and efficiency.
To understand the behavior of potential network invaders, this paper considers a system attack problem from the perspective of an invader. The invader intends to attack a system, where a group of sensors measure a pro...
详细信息
ISBN:
(纸本)9781467374439
To understand the behavior of potential network invaders, this paper considers a system attack problem from the perspective of an invader. The invader intends to attack a system, where a group of sensors measure a process state and send the measurements to a remote estimator for state estimation, by launching Denial-of-Service(Do S) attacks to block the communication channels. As the invader has a power budget and cannot block all the channels, he needs to decide which sensors to attack so that the estimation performance can be mostly affected, which is studied in this paper. In the scenario where the sensing abilities of the sensors have a full order, an explicit solution is provided. When the order does not exist, the problem is transformed into a convex optimization problem and is solved using efficient numerical algorithms.
Many systems composed by several interacting subsystems are usually controlled by a distributed control framework. Distributed Model Predictive control(DMPC) strategy, in which each subsystem is controlled by a local ...
详细信息
ISBN:
(纸本)9781479947249
Many systems composed by several interacting subsystems are usually controlled by a distributed control framework. Distributed Model Predictive control(DMPC) strategy, in which each subsystem is controlled by a local MPC controller, has advantages of accommodating constraints, less computational cost and high flexibility. In order to improve the global performance and guarantee the system stability, a stabilized DMPC strategy is proposed in this paper, in which subsystems interact through inputs. At first, local initial feasible solutions are achieved based on a Minkowski functional to guarantee the local closed-loop system stabilization. And then the global optimal solutions are obtained through coordination strategy for the sake of reducing iteration time and accelerating the convergence speed efficiently. Finally, the accuracy and efficiency of the proposed scheme is put to test through simulation.
In the present study, twelve volunteers were participated in a 2 h continuous mental arithmetic task without any break, which was designed to induce mental fatigue. The negative influence was investigated through EEG ...
详细信息
Many solder joints usually have to be traversed for spot welding robots, and reasonable welding sequence will improve welding efficiency. Intelligent optimization algorithms have been used to study path optimization p...
详细信息
This paper studies the dynamic output feedback consensus problem of multi-agent systems over analog fading *** the case of undirected communication topology,both sufficient and necessary conditions are presented for m...
详细信息
ISBN:
(纸本)9781479947249
This paper studies the dynamic output feedback consensus problem of multi-agent systems over analog fading *** the case of undirected communication topology,both sufficient and necessary conditions are presented for mean square consensus of discrete-time LTI multi-agent systems over analog fading *** is further shown that in the case of single output,the sufficient condition is also necessary,while for other cases,the gap between the sufficient condition and the necessary condition may be ***,sufficient and necessary conditions are also provided for the mean square consensus over a balanced directed communication topology by using Lyapunov *** the derived criteria demonstrate intricately how system dynamics,communication quality and network topological structure interplay with each other to allow the existence of a linear distributed consensus controller.
暂无评论