CERES :: Browsing by Author "Cao, Yi"

Browsing by Author "Cao, Yi"

Now showing 1 - 20 of 65

Open Access
Active control of hydrodynamic slug flow
(Cranfield University, 2013-04) Inyiama, Fidelis Chidozie; Cao, Yi
Multiphase flow is associated with concurrent flow of more than one phase (gas-liquid, liquid-solid, or gas-liquid-solid) in a conduit. The simultaneous flow of these phases in a flow line, may initiate a slug flow in the pipeline. Hydrodynamic slug flow is an alternate or irregular flow with surges of liquid slug and gas pocket. This occurs when the velocity difference between the gas flow rate and liquid flow rate is high enough resulting in an unstable hydrodynamic behaviour usually caused by the Kelvin-Helmholtz instability. Active feedback control technology, though found effective for the control of severe slugs, has not been studied for hydrodynamic slug mitigation in the literature. This work extends active feedback control application for mitigating hydrodynamic slug problem to enhance oil production and recovery. Active feedback Proportional-Integral (PI) control strategy based on measurement of pressure at the riser base as controlled variable with topside choking as manipulated variable was investigated through Olga simulation in this project. A control system that uses the topside choke valve to keep the pressure at the riser base at or below the average pressure in the riser slug cycle has been implemented. This has been found to prevent liquid accumulation or blockage of the flow line. OLGA (olga is a commercial software widely tested and used in oil and gas industries) has been used to assess the capability of active feedback control strategy for hydrodynamic slug control and has been found to give useful results and most interestingly the increase in oil production and recovery. The riser slugging was suppressed and the choke valve opening was improved from 5% to 12.65% using riser base pressure as controlled variable and topside choke valve as the manipulated variable for the manual choking when compared to the automatic choking in a stabilised operation, representing an improvement of 7.65% in the valve opening. Secondly, implementing active control at open-loop condition reduced the riser base pressure from 15.3881bara to 13.4016bara.
Open Access
Bidirectional branch and bound for controlled variable selection. Part I: principles and minimum singular value criterion.
(Elsevier Science B.V., Amsterdam., 2008-10-17T00:00:00Z) Cao, Yi; Kariwala, Vinay
The minimum singular value (MSV) rule is a useful tool for selecting controlled variables (CVs) from the available measurements. However, the application of the MSV rule to large-scale problems is difficult, as all feasible measurement subsets need to be evaluated to find the optimal solution. In this paper, a new and efficient branch and bound (BAB) method for selection of CVs using the MSV rule is proposed by posing the problem as a subset selection problem. In traditional BAB algorithms for subset selection problems, pruning is performed downwards (gradually decreasing subset size). In this work, the branch pruning is considered in both upward (gradually increasing subset size) and downward directions simultaneously so that the total number of subsets evaluated is reduced dramatically. Furthermore, a novel bidirectional branching strategy to dynamically branch solution trees for subset selection problems is also proposed, which maximizes the number of nodes associated with the branches to be pruned. Finally, by replacing time-consuming MSV calculations with novel determinant based conditions, the efficiency of the bidirectional BAB algorithm is increased further. Numerical examples show that with these new approaches, the CV selection problem can be solved incredibly fast.
Open Access
Bidirectional branch and bound for controlled variable selection. Part II: exact local method for self-optimizing control
(Elsevier Science B.V., Amsterdam., 2009-08-12T00:00:00Z) Kariwala, Vinay; Cao, Yi
The selection of controlled variables (CVs) from available measurements through enumeration of all possible alternatives is computationally forbidding for large-dimensional problems. In Part I of this work [Cao, Y., & Kariwala, V. (2008). Bidirectional branch and bound for controlled variable selection: Part I. Principles and minimum singular value criterion. Comput. Chem. Eng., 32 (10),2306-2319], we proposed a bidirectional branch and bound (BAB) approach for subset selection problems and demonstrated its efficiency using the minimum singular value criterion. In this paper, the BAB approach is extended for CV selection using the exact local method for self-optimizing control. By redefining the loss expression, we show that the CV selection criterion for exact local method is bidirectionally monotonic. A number of novel determinant based criteria are proposed for fast pruning and branching purposes resulting in a computationally inexpensive BAB approach. We also establish a link between the problems of selecting a subset and combinations of measurements as CVs and present a partially bidirectional BAB method for selection of measurements, whose combinations can be used as CVs. Numerical tests using randomly generated matrices and binary distillation column case study demonstrate the computational efficiency of the proposed methods. (C) 2009 Elsevier Ltd. All rights reserved.
Open Access
Bidirectional branch and bound for controlled variable selection. Part III: local average loss minimization
(IEEE, 2010-02-05T00:00:00Z) Kariwala, Vinay; Cao, Yi
The selection of controlled variables (CVs) from available measurements through exhaustive search is computationally forbidding for large-scale processes. We have recently proposed novel bidirectional branch and bound (B-3) approaches for CV selection using the minimum singular value (MSV) rule and the local worst- case loss criterion in the framework of self-optimizing control. However, the MSV rule is approximate and worst-case scenario may not occur frequently in practice. Thus, CV selection by minimizing local average loss can be deemed as most reliable. In this work, the B-3 approach is extended to CV selection based on local average loss metric. Lower bounds on local average loss and, fast pruning and branching algorithms are derived for the efficient B-3 algorithm. Random matrices and binary distillation column case study are used to demonstrate the computational efficiency of the proposed method.
Open Access
A branch and bound method for isolation of faulty variables through missing variable analysis
(Elsevier Science B.V., Amsterdam., 2010-12-31T00:00:00Z) Kariwala, Vinay; Odiowei, P. E.; Cao, Yi; Chen, T.
Fault detection and diagnosis is a critical approach to ensure safe and efficient operation of manufacturing and chemical processing plants. Although multivariate statistical process monitoring has received considerable attention, investigation into the diagnosis of the source or cause of the detected process fault has been relatively limited. This is partially due to the difficulty in isolating multiple variables, which jointly contribute to the occurrence of fault, through conventional contribution analysis. In this work, a method based on probabilistic principal component analysis is proposed for fault isolation. Furthermore, a branch and bound method is developed to handle the combinatorial nature of problem involving finding the contributing variables, which are most likely to be responsible for the occurrence of fault. The efficiency of the method proposed is shown through benchmark examples, such as Tennessee Eastman process, and randomly generated cases.
Open Access
Branch and bound method for multiobjective pairing selection
(Elsevier Science B.V., Amsterdam., 2010-05-31T00:00:00Z) Kariwala, Vinay; Cao, Yi
Most of the available methods for selection of input-output pairings for decentralized control require evaluation of all alternatives to find the optimal pairings. As the number of alternatives grows rapidly with process dimensions, pairing selection through an exhaustive search can be computationally forbidding for large-scale processes. Furthermore, the different criteria can be conflicting necessitating pairing selection in a multiobjective optimization framework. In this paper, an efficient branch and bound (BAB) method for multiobjective pairing selection is proposed. The proposed BAB method is illustrated through a biobjective pairing problem using selection criteria involving the relative gain array and the mu-interaction measure. The computational efficiency of the proposed method is demonstrated by using randomly generated matrices and the large-scale case study of cross-direction control. (C) 2010 Elsevier Ltd. All rights reserved.
Open Access
Branch and bound method for regression-based controlled variable selection
(Elsevier, 2013-03-27) Kariwala, Vinay; Ye, Lingjian; Cao, Yi
Self-optimizing control is a promising method for selection of controlled variables (CVs) from available measurements. Recently, Ye, Cao, Li, and Song (2012) have proposed a globally optimal method for selection of self-optimizing CVs by converting the CV selection problem into a regression problem. In this approach, the necessary conditions of optimality (NCO) are approximated by linear combinations of available measurements over the entire operation region. In practice, it is desired that a subset of available measurements be combined as CVs to obtain a good trade-off between the economic performance and the complexity of control system. The subset selection problem, however, is combinatorial in nature, which makes the application of the globally optimal CV selection method to large-scale processes difficult. In this work, an efficient branch and bound (BAB) algorithm is developed to handle the computational complexity associated with the selection of globally optimal CVs. The proposed BAB algorithm identifies the best measurement subset such that the regression error in approximating NCO is minimized and is also applicable to the general regression problem. Numerical tests using randomly generated matrices and a binary distillation column case study demonstrate the computational efficiency of the proposed BAB algorithm.
Open Access
Can a wind turbine learn to operate itself? Evaluation of the potential of a heuristic, data-driven self-optimizing control system for a 5MW offshore wind turbine
(Elsevier, 2017-12-15) Gueorguiev Iordanov, Stefan; Collu, Maurizio; Cao, Yi
Larger and more expensive offshore wind turbines, subject to more complex loads, operating in larger wind farms, could substantially benefit from more advanced control strategies. Nonetheless, the wind industry is reluctant to adopt such advanced, more efficient solutions, since this is perceived linked to a lower reliability. Here, a relatively simple self-optimizing control strategy, capable to “learn” (data-driven) which is the optimum control strategy depending on the objective defined, is presented. It is proved that it “re-discovers”, model-free, the optimum strategy adopted by commercial wind turbine in region 2. This methodology has the potential to achieve advanced control performance without compromising its simplicity and reliability.
Open Access
Canonical variate analysis for performance degradation under faulty conditions
(Elsevier, 2016-06-01) Ruiz Cárcel, Cristóbal; Lao, Liyun; Cao, Yi; Mba, David
Condition monitoring of industrial processes can minimize maintenance and operating costs while increasing the process safety and enhancing the quality of the product. In order to achieve these goals it is necessary not only to detect and diagnose process faults, but also to react to them by scheduling the maintenance and production according to the condition of the process. The objective of this investigation is to test the capabilities of canonical variate analysis (CVA) to estimate performance degradation and predict the behavior of a system affected by faults. Process data was acquired from a large-scale experimental multiphase flow facility operated under changing operational conditions where process faults were seeded. The results suggest that CVA can be used effectively to evaluate how faults affect the process variables in comparison to normal operation. The method also predicted future process behavior after the appearance of faults, modeling the system using data collected during the early stages of degradation.
Open Access
Canonical variate dissimilarity analysis for process incipient fault detection
(IEEE, 2018-02-28) Salgado Pilario, Karl Ezra; Cao, Yi
Early detection of incipient faults in industrial processes is increasingly becoming important, as these faults can slowly develop into serious abnormal events, an emergency situation, or even failure of critical equipment. Multivariate statistical process monitoring methods are currently established for abrupt fault detection. Among these, canonical variate analysis (CVA) was proven to be effective for dynamic process monitoring. However, the traditional CVA indices may not be sensitive enough for incipient faults. In this work, an extension of CVA, called the canonical variate dissimilarity analysis (CVDA), is proposed for process incipient fault detection in nonlinear dynamic processes under varying operating conditions. To handle non-Gaussian distributed data, kernel density estimation was used for computing detection limits. A CVA dissimilarity-based index has been demonstrated to outperform traditional CVA indices and other dissimilarity-based indices, namely DISSIM, RDTCSA, and GCCA, in terms of sensitivity when tested on slowly developing multiplicative and additive faults in a CSTR under closed-loop control and varying operating conditions.
Open Access
Combination of process and vibration data for improved condition monitoring of industrial systems working under variable operating conditions
(Elsevier, 2015-06-19) Ruiz Cárcel, Cristóbal; Jaramillo, Víctor H.; Mba, David; Ottewill, James R.; Cao, Yi
The detection and diagnosis of faults in industrial processes is a very active field of research due to the reduction in maintenance costs achieved by the implementation of process monitoring algorithms such as Principal Component Analysis, Partial Least Squares or more recently Canonical Variate Analysis (CVA). Typically the condition of rotating machinery is monitored separately using vibration analysis or other specific techniques. Conventional vibration-based condition monitoring techniques are based on the tracking of key features observed in the measured signal. Typically steady-state loading conditions are required to ensure consistency between measurements. In this paper, a technique based on merging process and vibration data is proposed with the objective of improving the detection of mechanical faults in industrial systems working under variable operating conditions. The capabilities of CVA for detection and diagnosis of faults were tested using experimental data acquired from a compressor test rig where different process faults were introduced. Results suggest that the combination of process and vibration data can effectively improve the detectability of mechanical faults in systems working under variable operating conditions.
Open Access
Design and analysis of robust controllers for directional drilling tools
(Cranfield University, 2016-08) Ghole, Amaan; Whidborne, James F.; Cao, Yi
Directional drilling is a very important tool for the development of oil and gas deposits. Attitude control which enables directional drilling for the efficient placement of the directional drilling tools in petroleum producing zones is reviewed along with the various engineering requirements or constraints. This thesis explores a multivariable attitude governing plant model as formulated in Panchal et al. (2010) which is used for developing robust control techniques. An inherent input and measurement delay which accounts for the plant's dead-time is included in the design of the controllers. A Smith Predictor controller is developed for reducing the effect of this dead-time. The developed controllers are compared for performance and robustness using structured singular value analysis and also for their performance indicated by the transient response of the closed loop models. Results for the transient non-linear simulation of the proposed controllers are also presented. The results obtained indicate that the objectives are satisfactorily achieved.
Open Access
Differential recurrent neural network based predictive control.
(Elsevier Science B.V., Amsterdam., 2008-07-24T00:00:00Z) Al Seyab, Rihab Khalid Shakir; Cao, Yi
In this paper an efficient algorithm to train general differential recurrent neural network (DRNN) is developed. The trained network can be directly used in the nonlinear model predictive control (NMPC) context. The neural network is represented in a general nonlinear state-space form and used to predict the future dynamic behavior of the nonlinear process in real time. In the new training algorithms, the ODEs of the model and the dynamic sensitivity are solved simultaneously using Taylor series expansion and automatic differentiation (AD) techniques. The same approach is also used to solve the online optimization problem in the predictive controller. The efficiency and effectiveness of the DRNN training algorithm and the NMPC approach are demonstrated through a two-CSTR case study. A good model fitting for the nonlinear plant at different sampling rates is obtained using the new method. A comparison with other approaches shows that the new algorithm can considerably reduce network training time and improve solution accuracy. The DRNN based NMPC approach results in good control performance under different operating conditions.
Open Access
Direct and indirect gradient control for static optimisation
(Springer Science Business Media, 2005-01-01T00:00:00Z) Cao, Yi
Static “self-optimising” control is an important concept, which provides a link between static optimisation and control (Skogestad, 2000). According to the concept, a dynamic control system could be configured in such a way that when a set of certain variables are maintained at their setpoints, the overall process operation is automatically optimal or near optimal at steady-state in the presence of disturbances. A novel approach using constrained gradient control to achieve “self-optimisation” has been proposed by Cao (2004). However, for most process plants, the information required to get the gradient measure may not be available in real-time. In such cases, controlled variable selection has to be carried out based on measurable candidates. In this work, the idea of direct gradient control has been extended to controlled variable selection based on gradient sensitivity analysis (indirect gradient control). New criteria, which indicate the sensitivity of the gradient function to disturbances and implementation errors, have been derived for selection. The particular case study shows that the controlled variables selected by gradient sensitivity measures are able to achieve near optimal perf
Open Access
Dynamic latent variable modelling and fault detection of Tennessee Eastman challenge process
(IEEE, 2016-05-19) Cao, Yi; Samuel, Raphael
Dynamic principal component analysis (DPCA) is commonly used for monitoring multivariate processes that evolve in time. However, it is has been argued in the literature that, in a linear dynamic system, DPCA does not extract cross correlation explicitly. It does not also give the minimum dimension of dynamic factors with non zero singular values. These limitations reduces its process monitoring effectiveness. A new approach based on the concept of dynamic latent variables is therefore proposed in this paper for extracting latent variables that exhibit dynamic correlations. In this approach, canonical variate analysis (CVA) is used to capture process dynamics instead of the DPCA. Tests on the Tennessee Eastman challenge process confirms the workability of the proposed approach.
Open Access
Experimental and simulation studies on performance of a compact gas/liquid separation system
(Cranfield University, 2013-01) Zhou, Ying Hui; Cao, Yi
The need of exploiting the offshore oil reserves and reducing the equipment costs becomes the motivation for developing new compact separation techniques. In the past years, the development of compact separators has almost solely focused on the cyclonic type separators made of pipes, because of their simple construction, relatively low cost of manufacturing and being able to withstand high pressures. Considerable effort has been put into the separator test program and qualification, and consequently notable advances in the compact separation technique have been made. However the application has been held back due to lacking of reliable predicting and design tools. The objectives of this study were threefold. Firstly, an experimental study was carried out aiming at understanding the separation process and flow behaviours in a compact separator, named Pipe-SEP, operating at high inlet gas volume fraction (GVF). Secondly it is to gain insight of the gas and liquid droplet flow in the compact separator by means of Computational Fluid Dynamics (CFD) simulations. Last but not least, the understanding and insight gained above were used to develop a comprehensive performance predictive model, based on which, a reliable optimizing design procedure is suggested. An experimental study was carried out to test a 150-mm Pipe-SEP prototype with a water-air mixture. Three distinct flow regimes inside the Pipe-SEP were identified, namely swirled, agitated, and gas blow-by. The transition of the flow regimes was found to be affected by inlet flow characteristics, mixture properties, geometry of the separator, and downstream conditions. A predictive model capable of predicting the transition of flow regimes and the separation efficiency was developed. A comparison between the predicted result and experiment data demonstrated that the model could serve as a design tool to support decision-making in early design stages ... [cont.].
Open Access
Experimental study on severe slugging mitigation by applying wavy pipes
(Elsevier Science B.V., Amsterdam, 2013-01-31T00:00:00Z) Xing, Lanchang; Yeung, Hoi; Shen, Joseph; Cao, Yi
Wavy pipes were installed in the pipeline for mitigating severe slugging in pipeline/riser systems. Experimental results have revealed that: with a wavy pipe applied, the operating region of severe slugging is reduced; the severity of severe slugging and oscillation flow is mitigated; the wavy pipe performs better with its outlet located upstream of the riser base. The wavy pipe is essentially reducing the slug length. For severe slugging the wavy pipe works by accelerating the movement of the gas in the pipeline to the riser base; for the oscillation flow it works by mixing the gas/liquid two phases.
Open Access
A Formulation of nonlinear model predictive control using automatic differentiation
(Elsevier Science B.V., Amsterdam., 2005-12-01T00:00:00Z) Cao, Yi
An efficient algorithm is developed to alleviate the computational burden associated with nonlinear model predictive control (NMPC). The new algorithm extends an existing algorithm for solutions of dynamic sensitivity from autonomous to non-autonomous differential equations using the Taylor series and automatic differentiation (AD). A formulation is then presented to recast the NMPC problem as a standard nonlinear programming problem by using the Taylor series and AD. The efficiency of the new algorithm is compared with other approaches via an evaporation case study. The comparison shows that the new algorithm can reduce computational time by two orders of magnitude.
Open Access
Global approximation of self-optimizing controlled variables with average loss minimization
(American Chemical Society, 2015-11-23) Ye, Lingjian; Cao, Yi; Yuan, Xiaofeng
Self-optimizing control (SOC) constitutes an important class of control strategies for real-time optimization (RTO) of chemical plants, by means of selecting appropriate controlled variables (CVs). Within the scope of SOC, this paper develops a CV selection methodology for a global solution which aims to minimise the average economic loss across the entire operation space. A major characteristic making the new scheme different from existing ones is that each uncertain scenario is independently considered in the new solution without relying on a linearised model, which was necessary in existing local SOC methods. Although global CV selection has been formulated as a nonlinear programming (NLP) problem, a tractable numerical algorithm for a rigorous solution is not available. In this work, a number of measures are introduced to ease the challenge. Firstly, we suggest to represent the economic loss as a quadratic function against the controlled variables through Taylor expansion, such that the average loss becomes an explicit function of the CV combination matrix, a direct optimizing algorithm is proposed to approximately minimize the global average loss. Furthermore, an analytic solution is derived for a suboptimal but much more simplified problem by treating the Hessian of the cost function over the entire operating space as a constant. This approach is found very similar to one of existing local methods, except that a matrix involved in the new solution is constructed from global operating data instead of using a local linear model. The proposed methodologies are applied to three simulated examples, where the effectiveness of proposed algorithms are demonstrated.
Open Access
Global self-optimizing control for uncertain constrained process systems
(Elsevier, 2017-10-18) Ye, Lingjian; Cao, Yi; Skogestad, Sigurd
Self-optimizing control is a promising control strategy to achieve real-time optimization (RTO) for uncertain process systems. Recently, a global self-optimizing control (gSOC) approach has been developed to extend the economic performance to be globally acceptable in the entire uncertain space spanned by disturbances and measurement noise. Nevertheless, the gSOC approach was derived based on the assumption of no change in active constraints, which limits the applicability of the approach. To address this deficiency, this paper proposes a new CV selection approach to handle active constraint changes. It ensures that all constraints are within their feasible regions when the selected CVs are maintained at constant setpoints for all expected uncertainties. In particular, constraints of interest are linearized at multiple operating conditions to get better estimates of their values and then incorporated into the optimization formulation when solving the globally self-optimizing CVs. The new CV selection approach is able to ensure an improved operational economic performance without potential constraint violations, as illustrated in an evaporator case study.