Data-driven soft sensor modeling based on twin support vector regression for cane sugar crystallization

Meng, Yanmei; Lan, Qiliang; Qin, Johnny; Yu, Shuangshuang; Pang, Haifeng; Zheng, Kangyuan

doi:10.1016/j.jfoodeng.2018.07.035

Cited by 36 publications

(18 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In Equations (5) and (6), S r represents the relative supersaturation, and is defined by Equation 8:…”

Section: Of 12mentioning

confidence: 99%

“…In Equations (5) and (6), the proposed mathematical model includes seven kinetic parameters (K b , b, j, p, K g , g, and q) specific for each batch crystallization operating condition [10]. These values need to be adjusted so that the model provides representative results of this process.…”

Section: Parameter Value Unitmentioning

confidence: 99%

“…The instrumentation of massecuite vacuum pans is a challenge for automation companies due to the high values of the sensors and large amount of equipment to be monitored [5,6]. In order to reduce these costs, two models were developed to relate the mean diameter of the crystals D (4,3) and the total formed crystal mass (FCM)-the data of which are difficult to measure-with the equipment concentration, volume, and temperature; one for A massecuite vacuum pan and the other for B massecuite vacuum pan [7].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fed-Batch Sucrose Crystallization Model for the B Massecuite Vacuum Pan, Solution by Deterministic and Heuristic Methods

et al. 2020

View full text Add to dashboard Cite

Fed-batch crystallization is a crucial step for sugar production. In order to relate parameters that are difficult to measure (average diameter of the crystals and total mass formed) to other easier to measure parameters (volume, temperature, and concentration), a model was developed for a B massecuite vacuum pan composed of mass and energy balances together with empirical relations that describe the crystal development inside equipment. The generated system of ordinary differential equations (ODE) had eight parameters which were adjusted through minimization of relative differences between the model results and experimental data. It was solved through the function fmincon, available in MATLABTM, which is a deterministic and gradient-based optimization method. The objective of this paper is to improve the model obtained and, for this purpose, two metaheuristic functions were used: genetic algorithm and particle swarm. To compare the results, the convergence time of each algorithm was used as well as the resulting quadratic deviation. The particle swarm method was the best option among the three used, presenting a shorter execution time and lower quadratic relative deviation.

show abstract

“…In Equations (5) and (6), S r represents the relative supersaturation, and is defined by Equation 8:…”

Section: Of 12mentioning

confidence: 99%

Section: Parameter Value Unitmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fed-Batch Sucrose Crystallization Model for the B Massecuite Vacuum Pan, Solution by Deterministic and Heuristic Methods

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Due to the nonlinearity and increasing complexity in the modern process industry, there are demands for data‐based approaches 6, 20, 38–42. A data‐based approach is primarily disadvantaged by its requirement of large amounts of raw historical data.…”

Section: Introductionmentioning

confidence: 99%

A Review on Data‐Driven Learning Approaches for Fault Detection and Diagnosis in Chemical Processes

et al. 2021

View full text Add to dashboard Cite

Fault detection and diagnosis for process plants has been an active area of research for many years. This review presents a concise overview on supervised and unsupervised data‐driven approaches for fault detection and diagnosis in chemical processes. Methods based on supervised and unsupervised data‐driven techniques are reviewed, and the challenges in the field of fault detection and diagnosis have also been highlighted. It is observed that most of the data‐driven approaches are application specific, in that no single method can be used to obtain a generalized solution for nearly all purposes. The methods reviewed differ significantly from one to the other, and hence it is difficult to generalize any key similarity. The majority of the works proposed in the literature focused mainly on single fault detection, and do not cover the root‐cause diagnosis of the detected faults. In cases where both detection and diagnosis are performed, the focus is mainly for a single fault. In addition, majority of the articles do not extend to the diagnosis of the root cause for multiple and simultaneous faults.

show abstract

“…Damour, Benne, Grondin‐Perez, and Chabriat (2010) designed a model‐based soft‐sensor to improve the process monitoring and control in industrial sugar crystallization. Meng, Lan, et al (2019) and Meng, Yu, et al (2019) put forward hybrid model for cane sugar crystallization process based on mechanism and data‐driven, which can be used in actual industrial production process. These are the applied method of data‐driven modeling in sugar crystallization, and show effect of data‐driven modeling method when it was difficult to build mechanism model or optimize variables directly through mechanism modeling, however, there are less model for CSD in cane sugar crystallization process to improve the quality of sugar crystal and increase production efficiency.…”

Section: Introductionmentioning

confidence: 99%

Data‐driven modeling for crystal size distribution parameters in cane sugar crystallization process

Meng

Tao

et al. 2021

J Food Process Engineering

Self Cite

View full text Add to dashboard Cite

Crystal size distribution (CSD) is important in evaluating crystal quality in cane sugar crystallization process. Due to the complex non-linearity, time-delay and strong coupling in cane sugar crystallization process, it is difficult to directly modeling in the mechanism of cane sugar crystallization process to obtain CSD parameters. In order to obtain two main CSD parameters so that to achieve better control and production of cane sugar, this article constructs a data-driven model based on least squares support vector regression (LSSVR) and particle swarm optimization (PSO). Based on LSSVR, the model takes the easy to measureable variables (massecuite brix, massecuite level, massecuite temperature, steam pressure, feeding volume, and vacuum degree) as input variables, and outputs sugar CSD parameters (crystal average size, coefficient of variation of crystal size). PSO algorithm is used to optimize the key parameters of primary model to get better modeling performance. Compared with other modeling methods such as back propagation, extreme learning machine, radial basis function, and SVR, the constructed PSO-LSSVR model has obvious advantages over other models in learning speed and predictive effect, generalization ability. This model has potential to be applied to the control system of cane sugar crystallization process and get better product of sugar.

show abstract

Data-driven soft sensor modeling based on twin support vector regression for cane sugar crystallization

Cited by 36 publications

References 19 publications

Fed-Batch Sucrose Crystallization Model for the B Massecuite Vacuum Pan, Solution by Deterministic and Heuristic Methods

Fed-Batch Sucrose Crystallization Model for the B Massecuite Vacuum Pan, Solution by Deterministic and Heuristic Methods

A Review on Data‐Driven Learning Approaches for Fault Detection and Diagnosis in Chemical Processes

Data‐driven modeling for crystal size distribution parameters in cane sugar crystallization process

Contact Info

Product

Resources

About