Papers by behrouz Kiani Talaei
IEEE SENSORS JOURNAL, 2024
This study aims to design a soft sensor (SS) using the state-dependent parameter (SDP) method wit... more This study aims to design a soft sensor (SS) using the state-dependent parameter (SDP) method within the control loop. SSs estimate variables based on measured values of other variables, but measurement noise affects their performance. The advantage of the SDP method is its reliance on the system state for model parameter dependency, leading to a lower model order and a reduction in number of input variables. However, this introduces measurement noise into the SS outputs. The research focuses on developing a reliable SS that effectively handles measurement noise, improving accuracy and reliability within the control loop. The innovation of this article lies in enhancing the SDP soft sensor (SDPSS) using the dynamic data reconciliation (DDR) filter. The proposed approach's superiority stems from utilizing linear regression (LR) models instead of complex dynamic process models as redundant models in the DDR filter's development. The performance of the SDP method is evaluated by employing an actual industrial sulfur recovery unit (SRU) as a reference point for assessing the effectiveness of SSs. Quantitative results conclusively demonstrate that SDP-based SSs significantly enhance estimation accuracy compared with alternative soft sensing techniques. A simulated continuous stirred-tank reactor (CSTR) is also used to evaluate the effectiveness of the proposed approach in two scenarios: external disturbance elimination and setpoint tracking. The DDR-SDPSS's performance is compared with that of the statistical filter-based SDPSS, demonstrating improved variable monitoring, reduced susceptibility to fluctuations, and satisfactory performance across a wide range of measurement noise variances without the need for controller readjustment.
Enhancing Soft Sensor Performance in Control Loops: Integrating State-Dependent Parameter Method with Regression-based Dynamic Data Reconciliation Filter
IEEE Sensors Journal, Dec 31, 2022
This study aims to design a soft sensor (SS) using the state-dependent parameter (SDP) method wit... more This study aims to design a soft sensor (SS) using the state-dependent parameter (SDP) method within the control loop. SSs estimate variables based on measured values of other variables, but measurement noise affects their performance. The advantage of the SDP method is its reliance on the system state for model parameter dependency, leading to a lower model order and a reduction in number of input variables. However, this introduces measurement noise into the SS outputs. The research focuses on developing a reliable SS that effectively handles measurement noise, improving accuracy and reliability within the control loop. The innovation of this article lies in enhancing the SDP soft sensor (SDPSS) using the dynamic data reconciliation (DDR) filter. The proposed approach's superiority stems from utilizing linear regression (LR) models instead of complex dynamic process models as redundant models in the DDR filter's development. The performance of the SDP method is evaluated by employing an actual industrial sulfur recovery unit (SRU) as a reference point for assessing the effectiveness of SSs. Quantitative results conclusively demonstrate that SDP-based SSs significantly enhance estimation accuracy compared with alternative soft sensing techniques. A simulated continuous stirred-tank reactor (CSTR) is also used to evaluate the effectiveness of the proposed approach in two scenarios: external disturbance elimination and setpoint tracking. The DDR-SDPSS's performance is compared with that of the statistical filter-based SDPSS, demonstrating improved variable monitoring, reduced susceptibility to fluctuations, and satisfactory performance across a wide range of measurement noise variances without the need for controller readjustment.
Systems and Soft Computing, 2023
The continuous stirred-tank reactors have complex dynamic behavior. In the reactor, the residual ... more The continuous stirred-tank reactors have complex dynamic behavior. In the reactor, the residual concentration decreases during an exothermic chemical reaction. In this transition proper cooling is essential to stabilize the reaction, and to prevent reactor overheating. This paper presents the data-driven metaheuristic state-dependent parameter proportional-integration-plus (M-SDP-PIP) control as an alternative of gain-scheduled control for adjusting the coolant temperature of the reactor. In fact, the parameters of the discrete transfer function were identified in the non-minimal state space using state-dependent parameter which is a nonlinear method; and weighting matrixes of linear quadratic servomechanism cost function of the optimal controller were achieved by metaheuristic methods, genetic algorithm, and particle swarm optimization, according to the minimization of the integral absolute error index and energy consumption. The proposed controller was compared to gain-scheduled and non-gain-scheduled controllers in the closed-loop condition in terms of tracking performance. The results show particle swarm optimization algorithm converge to the same optimal solution as genetic algorithm using fewer number of function evaluation, so that proposed controller has more effective offset-free servo-regulatory performance than other control schemes. In addition, it is simpler and more comprehensible because it requires no auxiliary equations and models to determine and manage the control gain, as opposed to the gain-scheduled structure.
There are different methods to separate Natural Gas Liquid (NGL) from natural gas. One of these m... more There are different methods to separate Natural Gas Liquid (NGL) from natural gas. One of these methods is refrigeration. Temperature reduction occurs in the dew point adjustment part to condense the NGL. The aim of this paper is to present a methodology for optimizing the NGL production process by calculating the best quantity for some set-points (the temperature, pressure, and etc for some vessels or other equipment) and at the same time try to reduce the energy consumption. To do this, we use a hybrid algorithm including a Genetic Algorithm (NSGA II) and Artificial Neural Network system (ANN). Indeed, in this research, we define a multi-objective problem and try to investigate more solutions for finding the best pareto-front. Therefore, it needs more time to evaluate solutions, thus, using an ANN to evaluate the amount of objective functions according to different solutions, is faster than the simulation methods. We solve the defined multi-objective problem by using of NSGAII. In...
Conference Presentations by behrouz Kiani Talaei
The 4th National Conference on Application of Experimental and Numerical Methods in Chemical and Mineral Industries , 2025
This study presents a Machine Learning-based (ML) framework for predicting and optimizing CO₂ cap... more This study presents a Machine Learning-based (ML) framework for predicting and optimizing CO₂ capture efficiency in post-combustion chemical absorption systems. A Random Forest model was trained on a synthetic dataset spanning a wide range of operational conditions, including variations in gas temperature, pressure, MEA concentration, gas flow rate, and liquid flow rate. The predictive model demonstrated strong performance with a Mean Absolute Error (MAE) of 0.0196 and a Coefficient of Determination (R²) of 0.878. Initial optimization efforts using a Grid Search approach focused solely on maximizing CO₂ capture efficiency, achieving a predicted efficiency of 98.2%. To enhance practical applicability, an integrated optimization framework was later developed, incorporating economic and energy consumption considerations and capping the maximum achievable efficiency at 97%. The refined optimization strategy resulted in a more economically viable operational point, achieving approximately 71% capture efficiency while minimizing solvent and energy costs. This work highlights the critical role of Machine Learning models in accelerating process optimization and underscores the importance of integrating practical constraints to design efficient and feasible carbon capture systems. Future work will focus on experimental validation and the application of advanced optimization algorithms to further improve computational and practical outcomes.
Uploads
Papers by behrouz Kiani Talaei
Conference Presentations by behrouz Kiani Talaei