Fed-batch Fermentation Process Research Articles

Animal-free production of hen egg ovalbumin in engineered Saccharomyces cerevisiae via precision fermentation

To enable the sustainable production of ovalbumin (OVA) without relying on animal sources, the generally recognized as safe (GRAS) host Saccharomyces cerevisiae was used for the precision fermentation-based production of recombinant OVA. For this purpose, a signal peptide derived from EPX1, the most abundant extracellular protein produced by Pichia pastoris, was identified as a novel signal peptide for the efficient secretion of OVA in S. cerevisiae. To improve OVA secretion and cell growth, three helper proteins (PDI1, KAR2, and HAC1) present in the endoplasmic reticulum were expressed individually or in combination. Notably, the +P1/K2 strain coexpressing PDI1 and KAR2 with OVA produced 2 mg/L of OVA in the medium fraction; this value was 2.6-fold higher than the corresponding value for the control strain without helper proteins. Finally, a glucose-limited fed-batch fermentation process using the +P1/K2 strain yielded 132 mg/L of total OVA with 8 mg/L of extracellular OVA.

Combinatorial Metabolic Engineering for Improving Betulinic Acid Biosynthesis in Saccharomyces cerevisiae.

Betulinic acid (BA) is a lupane-type triterpenoid with potent anticancer and anti-HIV activities. Its great potential in clinical applications necessitates the development of an efficient strategy for BA synthesis. This study attempted to achieve efficient BA biosynthesis in Saccharomyces cerevisiae using systematic metabolic engineering strategies. First, a de novo BA biosynthesis pathway in S. cerevisiae was constructed, which yielded a titer of 14.01 ± 0.21 mg/L. Then, by enhancing the BA synthesis pathway and dynamic inhibition of the competitive pathway, a greater proportion of the metabolic flow was directed toward BA synthesis, achieving a titer of 88.07 ± 5.83 mg/L. Next, acetyl-CoA and NADPH supply was enhanced, which increased the BA titer to 166.43 ± 1.83 mg/L. Finally, another BA synthesis pathway in the peroxisome was constructed. Dual regulation of the peroxisome and cytoplasmic metabolism increased the BA titer to 210.88 ± 4.76 mg/L. Following fed-batch fermentation process modification, the BA titer reached 682.29 ± 8.16 mg/L. Overall, this work offers a guide for building microbial cell factories that are capable of producing terpenoids with efficiency.

Optimization of the fermentative production of L-methionine by engineered Escherichia coli

As the only essential amino acid containing elemental sulphur, L-methionine has important physiological and biochemical functions in living organisms. However, the fermentative production of L-methionine has not met the requirements of industrial production because of its low production level. In this paper, the fermentation process of an efficient L-methionine producing strain E. coli W3110ΔIJAHFEBC trc-fliY trc-malY/PAM glyA-22 metF constructed previously was systematically optimized. Based on the optimal initial glucose concentration, the effects of different fed-batch fermentation processes, including DO-Stat, pH-Stat, controlling residual sugar control at different level and feeding glucose with constant rate, on L-methionine fermentation were studied. It was found that the control of glucose concentration greatly affected the fermentation process. Subsequently, an optimal fed-batch fermentation process was developed, where the L-methionine titer was increased to 31.71 g/L, the highest yield reported to date, while the fermentation time was shortened to 68 h. Meanwhile, a fermentation kinetics model under the optimal fed-batch fermentation conditions was established, which fitted well with the biosynthesis process of L-methionine. This study may facilitate further development of the fermentative production of L-methionine.

Construction and Regulation of the Abscisic Acid Biosynthesis Pathway in Yarrowia lipolytica.

Abscisic acid (ABA) is an important plant hormone with a variety of physiological functions such as regulating plant growth and helping plants to resist an adverse growth environment. However, at present, the ABA yield of heterologous biosynthesis by metabolic engineering is still low for industrial production. Therefore, five Botrytis cinerea genes (bcaba1, bcaba2, bcaba3, bcaba4, and bccpr1) related to ABA biosynthesis were expressed in Yarrowia lipolytica PO1h; its ABA production was 24.33 mg/L. By increasing the copy number of IDI and ERG12S, ERG20YMT, and bcaba3, bcaba1 genes, the yield of ABA was increased to 54.51 mg/L. By locating HMG-CoA reductase and HMG-CoA synthase in mitochondria, acetyl-CoA in mitochondria was converted into mevalonate; this increased the ABA yield to 102.12 mg/L. Finally, in the fed-batch fermentation process with the addition of dodecane, the ABA yield was up to 1212.57 mg/L, which is the highest yield of heterologous production of ABA by metabolic engineering.

An improved transfer learning approach based on geodesic flow kernel for multiphase batch process soft sensor modeling

For multiphase batch process, the characteristics of process data under various batches differ. Consequently, the soft sensor model built for a particular working condition is inapplicable to other working conditions. Besides, each batch can be divided into several phases whose characteristics are probably different. To address the above problems, a soft sensor model based on phase division and transfer learning strategy is proposed. First, transfer learning strategy is adopted to construct a soft sensor model applicable to various working conditions. Specifically, geodesic flow kernel based on linear local tangent space alignment (LLTSA-GFK) algorithm is designed. By projecting process data to the common manifold subspace, the distribution difference of data between various batches is reduced and the performance of the soft sensor model is enhanced. In addition, sequence-based fuzzy clustering and just-in-time learning (JITL) are adopted to solve the multistage characteristic for batch process. The root-mean-square error ( RMSE), coefficient of determination [Formula: see text], and mean absolute error ( MAE) are adopted to compare the conventional soft sensing approach (i.e., partial least-square regression based on JITL, support vector regression, and back propagation neural network) with the proposed approach. The superiority of the proposed model is verified by a fed-batch penicillin fermentation process.

Oxidative stress resistance prompts pyrroloquinoline quinone biosynthesis in Hyphomicrobium denitrificans H4-45

Pyrroloquinoline quinone (PQQ) is a natural antioxidant with diverse applications in food and pharmaceutical industries. A lot of effort has been devoted toward the discovery of PQQ high-producing microbial species and characterization of biosynthesis, but it is still challenging to achieve a high PQQ yield. In this study, a combined strategy of random mutagenesis and adaptive laboratory evolution (ALE) with fermentation optimization was applied to improve PQQ production in Hyphomicrobium denitrificans H4-45. A mutant strain AE-9 was obtained after nearly 400 generations of UV-LiCl mutagenesis, followed by an ALE process, which was conducted with a consecutive increase of oxidative stress generated by kanamycin, sodium sulfide, and potassium tellurite. In the flask culture condition, the PQQ production in mutant strain AE-9 had an 80.4% increase, and the cell density increased by 14.9% when compared with that of the initial strain H4-45. Moreover, batch and fed-batch fermentation processes were optimized to further improve PQQ production by pH control strategy, methanol and H2O2 feed flow, and segmented fermentation process. Finally, the highest PQQ production and productivity of the mutant strain AE-9 reached 307 mg/L and 4.26 mg/L/h in a 3.7-L bioreactor, respectively. Whole genome sequencing analysis showed that genetic mutations in the ftfL gene and thiC gene might contribute to improving PQQ production by enhancing methanol consumption and cell growth in the AE-9 strain. Our study provided a systematic strategy to obtain a PQQ high-producing mutant strain and achieve high production of PQQ in fermentation. These practical methods could be applicable to improve the production of other antioxidant compounds with uncleared regulation mechanisms.Key points• Improvement of PQQ production by UV-LiCl mutagenesis combined with adaptive laboratory evolution (ALE) and fermentation optimization.• A consecutive increase of oxidative stress could be used as the antagonistic factor for ALE to enhance PQQ production.• Mutations in the ftfL gene and thiC gene indicated that PQQ production might be increased by enhancing methanol consumption and cell growth.

Enzyme selection, optimization, and production toward biodegradation of post-consumer poly(ethylene terephthalate) at scale.

Poly(ethylene terephthalate) (PET) is one of the world's most widely used polyester plastics. Due to its chemical stability, PET is extremely difficult to hydrolyze in a natural environment. Recent discoveries in new polyester hydrolases and breakthroughs in enzyme engineering strategies have inspired enormous research on biorecycling of PET. This study summarizes our research efforts toward large-scale, efficient, and economical biodegradation of post-consumer waste PET, including PET hydrolase selection and optimization, high-yield enzyme production, and high-capacity enzymatic degradation of post-consumer waste PET. First, genes encoding PETase and MHETase from Ideonella sakaiensisand the ICCG variant of leaf-branch compost cutinase (LCCICCG ) were codon-optimized and expressed in Escherichia coliBL21(DE3) for high-yield production. To further lower the enzyme production cost, a pelB leader sequence was fused to LCCICCG so that the enzyme can be secreted into the medium to facilitate recovery. To help bind the enzyme on the hydrophobic surface of PET, a substrate-binding module in a polyhydroxyalkanoate depolymerase from Alcaligenes faecalis (PBM) was fused to the C-terminus of LCCICCG . The resulting four different LCCICCG variants (LCC, PelB-LCC, LCC-PBM, and PelB-LCC-PBM), together with PETase and MHETase, were compared for PET degradation efficiency. A fed-batch fermentation process was developed to produce the target enzymes up to 1.2g/L. Finally, the best enzyme, PelB-LCC, was selected and used for the efficient degradation of 200g/L recycled PET in a well-controlled, stirred-tank reactor. The results will help develop an economical and scalable biorecycling process toward a circular PET economy. This article is protected by copyright. All rights reserved.

Single and multi-objective dynamic optimization study of an industrial scale fed batch reactor

Abstract The present work focuses on obtaining optimal operational policies of an industrial scale penicillin fed-batch fermentation process using dynamic optimization (DO). The three process objectives considered in this work include, maximization of total penicillin formed (J 1), minimization of total biomass formed (J 2), and minimization of fed-batch operation time (J 3). The control variables for the optimization study include, the feed flowrates of sugar, soyabean oil, and phenylacetic. We perform a single-objective optimization study with J 1 as the objective function and showcased the importance of DO. Consequently, we solve three multi-objective optimization (MOO) problems corresponding to all the pairwise combinations of the aforementioned three objectives. Outcomes of MOO problems are presented in terms of several non-dominated Pareto-optimal solutions. Furthermore, descriptive analysis of three representative Pareto points for all the MOO problems is presented, which highlights the importance of MOO study of the penicillin bio-reactor.

Data-Driven Soft Sensing for Batch Processes Using Neural Network-Based Deep Quality-Relevant Representation Learning

Soft sensors provide a means to reliably estimate unmeasurable variables, thereby playing a prevalent role in formulating closed-loop control in batch processes. In soft sensor development, enhancing quality-relevant information and eliminating quality-irrelevant information are important. This study proposes a neural network-based deep quality-relevant representation learning approach to improve the soft sensing performance in dynamic batch processes. The structure of a deep neural network is optimized in a layer-by-layer manner. First, given the generally abundant predictor variables, the maximal relevanceminimal redundancy criterion is used to optimize the input layer, select the most beneficial variables, and eliminate modeling redundancy. Second, mutual information-based quality-relevant representation selection is performed in the middle layers to enhance the quality-relevant information and eliminate the influence of irrelevant representations. Third, deep quality-relevant representations are extracted, and a soft sensor model is developed. The proposed method is tested on a fed-batch penicillin fermentation process and an industrial injection molding process. Lastly, it shows that the proposed method outperforms several state-of-the-art approaches, thereby confirming its effectiveness.

Enhanced production of nicotinamide mononucleotide by high cell density culture of engineered Escherichia coli

Nicotinamide mononucleotide (NMN), an intermediate product of NAD+ biosynthesis, is commonly found intracellularly in living species. Recently, NMN has attracted great interest for the treatment of different diseases. Here, we developed a high cell density fed-batch culture system to produce NMN at higher concentrations using an engineered E. coil strain. Fed-batch culture using linear feeding strategy exhibited 9.38 ± 0.8, 14.8 ± 0.56, and 5.04 ± 0.21 g/L NMN concentrations, and 59.5 ± 0.44, 68.8 ± 0.49, and 36.2 ± 0.32 g/L dry cell weights, using different concentrations of nicotinamide (NAM) of 3, 5, and 10 g/L, respectively. A maximum NMN concentration of 17.2 ± 1.3 g/L with a productivity of 0.956 g/L/h was achieved using two intermittent additions of 5 g/L NAM after 6 h and 17 h. NMN obtained from the fermentation broth was purified using size exclusion chromatography, generating a yield of 53.9 % NMN. Liquid chromatography-mass spectrometry and 1H nuclear magnetic resonance analysis was additionally used to validate the purity of NMN after purification. The fed-batch fermentation process applied in this study, using cheaper feed-stocks, demonstrated a remarkable enhancement in the production of therapeutically important NMN, making this a promising strategy for industrial-scale production of NMN.

Pseudo label estimation based on label distribution optimization for industrial semi-supervised soft sensor

In process industry, the lack of sufficient labeled data often leads to poor performance of traditional supervised soft sensors. Thus, a pseudo label estimation method based on label distribution optimization (PLELDO) is proposed. PLELDO first converts the pseudo label estimation into an explicit label distribution optimization problem, and then obtains high-confidence pseudo labels. Further, a semi-supervised ensemble soft sensor modeling framework namely EnPLELDO is developed. EnPLELDO first obtains sufficient good pseudo labeled data by repeating small-scale pseudo label estimation. Then, a set of diverse base models are constructed from the extended training sets. Finally, these enhanced base models are combined through stacking strategy. The application results from an industrial fed-batch fermentation process show that, compared with several state-of-the-art methods, PLELDO can obtain better pseudo label estimations and EnPLELDO can deliver more accurate predictions of key quality variables by leveraging the scarce labeled data and large-scale unlabeled data.

Efficient production of α-pinene through identifying the rate-limiting enzymes and tailoring inactive terminal of pinene synthase in Escherichia coli

α-Pinene has been indicated as an ideal biofuel or fuel precursors. The higher energy value and broad application of pinene dimers has recently been evaluated as a promising high-density jet fuel. In this work, the microbial de novo synthesis of α-pinene inEscherichia coliwas optimized by applying a systematic approach. Firstly, a complete biosynthetic process to α-pinene was constructed by bioprospecting the key enzymes pinene synthase (PtPS1Q457L) and geranyl diphosphate synthase (AgGPPS). Subsequently,the appropriate N-terminal truncated sites of pinene synthase were determined by analyzing the hydrophilicity/hydrophobicity of PtPS1Q457L. The accumulation of extracellular α-pinene was further increased 2.5 times by tailoring an appropriate length of inactive terminal of pinene synthase. Fine-tuning of IPTG concentration and induction point led to the synthesis of α-pinene at the yield of 541.8 mg/L at the shake flask level.Finally, further optimization of the fed-batch fermentation process enabled a α-pinene titer of 1035.1 mg/L in a 1.3L bioreactor, which is the highest according to currently available reports.

Model Predictive Control—A Stand Out among Competitors for Fed-Batch Fermentation Improvement

The fed-batch cultivation is in many ways a benchmark for fermentation processes, and it has been an attractive choice for the biotechnological production of various products in the past decades. The majority of biopharmaceuticals that are presently undergoing clinical trials or are available on the market are manufactured through fed-batch fermentations. A crucial process parameter in fed-batch cultivations is the substrate feed rate, which directly influences the overall process productivity, product quality and process repeatability; henceforth, effective control of this parameter is imperative for a successful fed-batch fermentation process. Two distinct control strategies can be distinguished—open-loop and closed-loop (feedback) control. Each of these methods has its own set of benefits, limitations and suitability for specific bioprocesses. This article surveys and compares the most popular open- and closed-loop methods for substrate feed rate control in fed-batch fermentations. Emphasis is placed on model-predictive feed rate control (MPC)—a stand out among other methods that offers a promising application perspective. The authors also demonstrate a practical example of the implementation of a robust, flexible MPC solution that is suitable for various cultures and runs on standard computer hardware, thus overcoming one of the main reported MPC drawbacks—high computational requirements.

Improved terephthalic acid production from p-xylene using metabolically engineered Pseudomonas putida

Terephthalic acid (TPA) is an important commodity chemical used as a monomer of polyethylene terephthalate (PET). Since a large quantity of PET is routinely manufactured and consumed worldwide, the development of sustainable biomanufacturing processes for its monomers (i.e. TPA and ethylene glycol) has recently gained much attention. In a previous study, we reported the development of a metabolically engineered Escherichia coli strain producing 6.7 g/L of TPA from p-xylene (pX) with a productivity and molar conversion yield of 0.278 g/L/h and 96.7 mol%, respectively. Here, we report metabolic engineering of Pseudomonas putida KT2440, a microbial chassis particularly suitable for the synthesis of aromatic compounds, for improved biocatalytic conversion of pX to TPA. To develop a plasmid-free, antibiotic-free, and inducer-free biocatalytic process for cost-competitive TPA production, all heterologous genes required for the synthetic pX-to-TPA bioconversion pathway were integrated into the chromosome of P. putida KT2440 by RecET-based markerless recombineering and overexpressed under the control of constitutive promoters. Next, TPA production was enhanced by integrating multiple copies of the heterologous genes to the ribosomal RNA genes through iteration of recombineering-based random integration and subsequent screening of high-performance strains. Finally, fed-batch fermentation process was optimized to further improve the performance of the engineered P. putida strain. As a result, 38.25 ± 0.11 g/L of TPA was produced from pX with a molar conversion yield of 99.6 ± 0.6%, which is equivalent to conversion of 99.3 ± 0.8 g pX to 154.6 ± 0.5 g TPA. This superior pX-to-TPA biotransformation process based on the engineered P. putida strain will pave the way to the commercial biomanufacturing of TPA in an industrial scale.

Development of a novel defined minimal medium for Gluconobacter oxydans 621H by systematic investigation of metabolic demands

BackgroundHistorically, complex media are used for the cultivation of Gluconobacter oxydans in industry and research. Using complex media has different drawbacks like higher costs for downstream processing and significant variations in fermentation performances. Synthetic media can overcome those drawbacks, lead to reproducible fermentation performances. However, the development of a synthetic medium is time and labour consuming. Detailed knowledge about auxotrophies and metabolic requirements of G. oxydans is necessary. In this work, we use a systematic approach applying the in-house developed μRAMOS technology to identify auxotrophies and develop a defined minimal medium for cultivation of G. oxydans fdh, improving the production process of the natural sweetener 5-ketofructose.ResultsA rich, defined synthetic medium, consisting of 48 components, including vitamins, amino acids and trace elements, was used as a basis for medium development. In a comprehensive series of experiments, component groups and single media components were individually omitted from or supplemented to the medium and analysed regarding their performance. Main components like salts and trace elements were necessary for the growth of G. oxydans fdh, whereas nucleotides were shown to be non-essential. Moreover, results indicated that the amino acids isoleucine, glutamate and glycine and the vitamins nicotinic acid, pantothenic acid and p-aminobenzoic acid are necessary for the growth of G. oxydans fdh. The glutamate concentration was increased three-fold, functioning as a precursor for amino acid synthesis. Finally, a defined minimal medium called ‘Gluconobacter minimal medium’ was developed. The performance of this medium was tested in comparison with commonly used media for Gluconobacter. Similar/competitive results regarding cultivation time, yield and productivity were obtained. Moreover, the application of the medium in a fed-batch fermentation process was successfully demonstrated.ConclusionThe systematic investigation of a wide range of media components allowed the successful development of the Gluconobacter minimal medium. This chemically defined medium contains only 14 ingredients, customised for the cultivation of G. oxydans fdh and 5-ketofructose production. This enables a more straightforward process development regarding upstream and downstream processing. Moreover, metabolic demands of G. oxydans were identified, which further can be used in media or strain development for different processes.

Pretreatment Strategies to Enhance Enzymatic Hydrolysis and Cellulosic Ethanol Production for Biorefinery of Corn Stover.

There is a rising interest in bioethanol production from lignocellulose such as corn stover to decrease the need for fossil fuels, but most research mainly focuses on how to improve ethanol yield and pays less attention to the biorefinery of corn stover. To realize the utilization of different components of corn stover in this study, different pretreatment strategies were used to fractionate corn stover while enhancing enzymatic digestibility and cellulosic ethanol production. It was found that the pretreatment process combining dilute acid (DA) and alkaline sodium sulfite (ASS) could effectively fractionate the three main components of corn stover, i.e., cellulose, hemicellulose, and lignin, that xylose recovery reached 93.0%, and that removal rate of lignin was 85.0%. After the joint pretreatment of DA and ASS, the conversion of cellulose at 72 h of enzymatic hydrolysis reached 85.4%, and ethanol concentration reached 48.5 g/L through fed-batch semi-simultaneous saccharification and fermentation (S-SSF) process when the final concentration of substrate was 18% (w/v). Pretreatment with ammonium sulfite resulted in 83.8% of lignin removal, and the conversion of cellulose and ethanol concentration reached 86.6% and 50 g/L after enzymatic hydrolysis of 72 h and fed-batch S-SSF, respectively. The results provided a reference for effectively separating hemicellulose and lignin from corn stover and producing cellulosic ethanol for the biorefinery of corn stover.

Thermal analysis of extractive fed-batch ethanol fermentation with CO2 stripping: Modeling and simulation

Ethanol production from sugarcane is an exothermic process. The heat generated during the fermentation is removed by heat exchangers, using cooling water. However, it is difficult to keep the fermentation process close to the optimum yeast temperature, especially in hot months. In this work, extractive fermentation with CO2 stripping was evaluated as a way to overcome this operational problem besides reducing the water demand in the process. A mathematical model was developed to describe conventional and extractive fed-batch ethanol fermentation processes. Simulations were performed to determine the effect of different gas flow rates on heat removal from fermentations at different must substrate concentrations (Csm). With a Csm of 285.7 g/L and a CO2 flow rate of 0.7 vvm, the consumption of cooling water in the extractive fermentation decreased by up to 62.1% compared to the conventional process, keeping the broth temperature constant. Thus, the potential of this technique to increase the total ethanol content and reduce the cooling water consumption was demonstrated. Furthermore, there were decreases in the amount of vinasse and steam consumption in the distillation step.

Metabolic recycling of storage lipids promotes squalene biosynthesis in yeast

BackgroundMetabolic rewiring in microbes is an economical and sustainable strategy for synthesizing valuable natural terpenes. Terpenes are the largest class of nature-derived specialized metabolites, and many have valuable pharmaceutical or biological activity. Squalene, a medicinal terpene, is used as a vaccine adjuvant to improve the efficacy of vaccines, including pandemic coronavirus disease 2019 (COVID-19) vaccines, and plays diverse biological roles as an antioxidant and anticancer agent. However, metabolic rewiring interferes with inherent metabolic pathways, often in a way that impairs the cellular growth and fitness of the microbial host. In particular, as the key starting molecule for producing various compounds including squalene, acetyl-CoA is involved in numerous biological processes with tight regulation to maintain metabolic homeostasis, which limits redirection of metabolic fluxes toward desired products.ResultsIn this study, focusing on the recycling of surplus metabolic energy stored in lipid droplets, we show that the metabolic recycling of the surplus energy to acetyl-CoA can increase squalene production in yeast, concomitant with minimizing the metabolic interferences in inherent pathways. Moreover, by integrating multiple copies of the rate-limiting enzyme and implementing N-degron-dependent protein degradation to downregulate the competing pathway, we systematically rewired the metabolic flux toward squalene, enabling remarkable squalene production (1024.88 mg/L in a shake flask). Ultimately, further optimization of the fed-batch fermentation process enabled remarkable squalene production of 6.53 g/L.ConclusionsOur demonstration of squalene production via engineered yeast suggests that plant- or animal-based supplies of medicinal squalene can potentially be complemented or replaced by industrial fermentation. This approach will also provide a universal strategy for the more stable and sustainable production of high-value terpenes.

Batch-to-Batch Iterative Learning Control of a Fed-Batch Fermentation Process

In this work, Iterative Learning Control on a fed-batch fermentation process using linearised models has been studied. The repetitive nature of batch processes enables ILC to obtain information from a previous batch in order to improve the performance of the current batch such that the product quality converges asymptotically to the desired trajectory The basic batch to batch ILC law presents the control action of a current batch as a summation of the control action from the previous batch and the deviation of the output trajectory from the desired reference trajectory incorporation with a learning rate. In a bid to address the issue of the process non-linearity, the control policy and the output trajectory were linearised around their respective nominal trajectories. The linearised models were then identified using Multi Linear Regression (MLR), Principal Component Analysis (PCR) and Partial Least Squares (PLS). In order to curb the effects of plant-model mismatches and process variations, the linearised models were reidentified after each batch operation. This was done by selecting the immediate previous batch as the nominal batch and then adding the recently obtained process data into the historical data batch on completion of the current batch run. The weighting matrices in the objective function were carefully selected taking into consideration that they have a major influence on the robust performance of the process. In using PLS and PCR models the issue of process collinearity was effectively addressed. The proposed batch to batch ILC strategy was applied to a simulated fed-batch fermentation process for the production of secreted protein. The results of the optimal control policy were comparable to that obtained in using full mechanistic model. ILC, a simple but yet an effective optimal control strategy has demonstrated to be a viable option in complex processes such as batch processes where mechanistic models are difficult to develop. Keywords: Iterative Learning Control, batch process, fed-batch fermentation, batch to batch ILC, control policy. DOI: 10.7176/CMR/14-3-02 Publication date: August 31 st 2022

Two-dimensional multiphase batch process monitoring based on sparse canonical variate analysis

Most industrial batch processes involve inherent dynamic characteristics in both within-batch time direction and batch-wise direction. In order to ensure process safety and improve process performance, the two-dimensional dynamics should be analyzed during batch process monitoring. In this work, two-dimensional region of support (2D-ROS) is first constructed to select and preserve the relevant samples for the current measured sample by calculating autoregressive orders with Akaike information criterion (AIC) in time direction and measuring the similarity with the weighted Euclidean distance in batch-wise direction. Afterwards, sparse canonical variate analysis (SCVA) algorithm is performed to yield sparse canonical vectors, which is especially advantageous for eliminating the irrelevant variables and facilitating the interpretation of underlying relationships of process variables. Meanwhile, given most measurements are subject to the non-Gaussian distribution, the upper control limits (UCLs) in 2D-SCVA can be estimated using kernel density estimation (KDE). The achieved results obtained from a numerical dynamic example and the benchmark fed-batch penicillin fermentation process clearly verify that the proposed method performs well for detecting abnormal operation for the batch processes.