Multivariable Strategy Research Articles

Optimization and Validation of a Liquid Formulation for a New Recombinant Veterinary Product using QbD Approach

Protein formulation and drug characterization are one of the most difficult and time-consuming tasks because of the complexity of biotherapeutic proteins. Hence, maintaining a protein drug in its active state typically requires preventing changes in its physical and chemical properties. Quality by Design (QbD) is a systematic approach emphasizing product and process understanding. Design of Experiments (DoE) is one of the most important QbD tools, allowing the possibility to modify the formulation attributes within a defined design space. Here, we report the validation of a RP-HPLC assay for recombinant equine chorionic gonadotropin (reCG) that demonstrated a high correlation with the in vivo potency biological assay. QbD concepts were then applied to obtain an optimized liquid formulation of reCG with a predefined quality product profile. The developed strategy demonstrates the importance of applying multivariable strategies as DoE to simplify formulation stages, improving the quality of the obtained results. Moreover, it is important to highlight that this is the first time that a liquid formulation is reported for an eCG molecule, since, up to now, the only eCG products available in the market for veterinary use consisted in partially purified preparations of pregnant mare serum gonadotropin (PMSG) presented as a lyophilized product.

A Fault Level-Based System to Control Voltage and Enhance Power Factor Through an On-Load Tap Changer and Distributed Generators

The on-load tap changer is critical for voltage regulation, however, its voltage regulation philosophy is suitable for one-directional power flow power systems. Since distributed generators introduce bidirectional power flow, the conventional operation of the on-load tap changer will be highly impacted, resulting in high voltage magnitudes that exceed acceptable limits, frequent voltage magnitude fluctuations, and the increase in overall reactive power supply when distributed generators start absorbing/injecting reactive power. Therefore, an adaptive control system is proposed on this paper that alters the existing voltage regulation philosophy of the on-load tap changer and also coordinate the voltage regulation capabilities of distributed generators with those of the on-load tap changer for optimal overall voltage regulation and the reduction of reactive power that flows through the power system, enhancing the power factor thereof. The proposed control system will change the centralized, and single-variable busbar voltage monitoring technique of the conventional philosophy with a decentralized, and multi-variable strategy that controls voltage based on the average voltage of the entire distribution network. The control system will calculate the on-load tap changer and distributed generators setpoints based on the overall average network voltage, average voltage deviation from nominal voltage, generators points of connection voltage and the total reactive power that flows through the power transformer. The control system was tested on a 22kV network modelled in MATLAB SIMULINK and the results attained revealed that the proposed multi-stage control system can successfully control voltage and simultaneously improve power factor.

Experimental Assessment of a Multi-Variable Control Strategy of a Micro-Cogeneration Solar-ORC Plant for Domestic Application

Suitability to off-design operation, applicability to combined thermal and electrical generation in a wide range of low temperatures and pressures and compliance with safety and environmental limitations qualify small-scale Organic Rankine Cycle plants as a viable option for combined heat and power generation in the residential sector. As the plants scale down, the electric and thermal output maximization has to account for issues, spanning from high pump power absorption, compared to the electric output of the plant, to intrinsically low plant permeability induced by the expander, to the intermittent availability of thermal power, affected by the heat demand for domestic hot water (DHW) production. The present paper accounts for a flat-plate solar thermal collector array, bottomed by an ORC unit featuring a sliding vane expander and pump and flat-plate heat exchangers. A high-temperature buffer vessel stores artificially heated water – electric heaters, simulating the solar collector - and feeds either the hot water line for domestic use or the ORC evaporator, depending on the instantaneous demand (i.e., domestic hot water or electric power), the temperature conditions inside the tank and the stored mass availability. A low-temperature receiver acts like the heat sink of the ORC unit and harvests the residual thermal power, downstream the expander: a dedicated control, modelled to properly modulate the mass addition/subtraction to this storage unit allows to restore the operating points of the cycle and to limit the incidence of off-design operation, via real-time adjustment of the cycle operating parameters. Indeed, the possibility of continuous ORC generation depends on (i) the nature of the demand and (ii) the amount of hot water withdrawn from the high-temperature buffer vessel. The time-to-temperature for the mass stored inside the buffer affects the amount of ORC unit activations and eventually the maximum attainable generation of electric energy. The plant energy performance is experimentally assessed, and various characteristic operating points are mapped, based on test runs carried out on a real-scale ORC pilot unit.

Predicting tumor mutational burden in head and neck squamous cell carcinoma based on CT imaging features: A TCGA/TCIA study.

e15254 Background: Immunotherapy has substantially changed the therapeutic strategies for cancers. Unfortunately, only 20–50% of patients with advanced solid tumours respond to treatment. There is therefore a need for the development of methods to identify patients who are most likely to respond to immunotherapy. Tumor mutation burden (TMB) have been served as the most prevalent biomarkers to predict immunotherapy response. This study was designed to investigate the ability of radiomics to predict TMB status in patients with head and neck squamous cell carcinoma (HNSCC). Methods: TMB values were calculated using genomic data obtained from the HNSCC dataset in The Cancer Genome Atlas (TCGA).We identified matching patients (n = 100) who underwent contrast-enhanced CT scan prior to treatment from The Cancer Imaging Archive (TCIA),and patients were grouped based on the cutoff value; high group(>4.2 mutations/Mb) and low group(≤4.2 mutations/Mb). A total of 249 radiomics features(9 non-texture features and 240 scan-texture-parameter features) were extracted from CT images of the tumor. The incorporation of features into multivariable models was performed using logistic regression. The multivariable modeling strategy involved imbalance-adjusted bootstrap resampling in the following four steps leading to final prediction model construction: (1) feature set reduction; (2) feature selection; (3) prediction performance estimation; and (4) computation of model coefficients. The performance was evaluated in terms of area under the curve (AUC), sensitivity, and specificity. Results: Among all the features, twenty features were found to have the most impact on the predictive value; the two top texture parameters were GLCM-Variance and GLCM-Sum Average. In multivariable analysis, the best performance was obtained using a combination of seven texture features that can discriminate between high mutation burden versus low mutation burden. The AUC, sensitivity, and specificity of this model were 0.97 ± 0.01, 0.92 ± 0.04, and 0.92 ±0.01, respectively. Conclusions: The proposed CT-derived predictive model can accurately predict TMB status in patients with HNSCC. It may be helpful in guiding immunotherapy in clinical practice and deserves further analysis.

Improvement of Small Wind Turbine Control in the Transition Region

Wind energy conversion systems are very challenging from the control system viewpoint. The control difficulties are even more challenging when wind turbines are able to operate at variable speed and variable pitch. The contribution of this work is focused on designing a combined controller that significantly alleviates the wind transient loads in the power tracking and power regulation modes as well as in the transition zone. In a previous work, the authors studied the applicability of different multivariable decoupling methodologies. The methodologies were tested in simulation and verified experimentally in a lab-scale wind turbine. It was demonstrated that multivariable control strategies achieve a good closed-loop response within the transition region, where the interaction level is greater. Nevertheless, although such controllers showed an acceptable performance in the power tracking (region II) and power regulation (region IV) zones, appreciable improvement was possible. To this end, the new proposed methodology employs a multivariable gain-scheduling controller with a static decoupling network for the transition region and monovariable controllers for the power tracking and power regulation regions. To make the transition between regions smoother, a gain scheduling block is incorporated into the multivariable controller. The proposed controller is experimentally compared with a standard switched controller in the lab-scale wind turbine. The experiments carried out suggest that the combination of the proposed multivariable strategy for the transition region to mitigate wind transient loads combined with two monovariable controllers, one dedicated to region II and other to region IV, provide better results than traditional switched control strategies.

Abstract P4-05-12: Genetic determinants of the molecular portraits of epithelial cancers

Abstract Background: Tumorigenesis is often driven by multiple types of aberrations in DNA leading to diseases of enormous complexity and heterogeneity. With the effort of many consortiums including The Caner Genome Atlas, large-scale multi-platform genomic data are now available, providing an opportunity to more accurately characterize cancer phenotypic heterogeneity on a molecular level and through using multiple technologies. In particular, many gene expression signatures have been developed to define specific cancer phenotypes varying from proliferation rates, to disease subtypes, to features of the tumor microenvironment. These mRNA expression features, along with protein expression, somatic mutations, and clinical features provide a comprehensive molecular portrait of tumors. Integrating multi-platform genomic data together to elucidate the relationship between genotype and phenotypes is critical to understanding the “driver” features of tumor behavior. Methods: In this study, we utilized tumor DNA information and an extensive archive of gene expression signatures as a framework to characterize multiple aspects of tumor biology. We present an integrative computational approach using a genome-wide association analysis, and an Elastic Net prediction method, to analyze the relationship between DNA copy number alterations (CNAs) and gene expression signatures. A multivariable Elastic Net modelling strategy was used to build objective predictive models for multiple molecular features including protein and gene expression patterns, somatic mutations and clinical phenotypes. Results: Across breast cancers we were able to identify known and novel associations between DNA CNAs and many gene expression signatures. Moreover, we were able to accurately predict many gene expression signatures levels within individual tumors using Elastic Net models based upon DNA copy number features alone; these successful models could predict proliferation status and Estrogen-signaling pathway activity. We were also able to build predictive models for many other key phenotypes including intrinsic molecular subtypes, some protein expression features including clinical estrogen receptor status, and somatic mutation status of TP53 and CDH1. This approach was also successfully applied to 24 other tumor types (Pan-Cancer), which identified a number of repeatedly predictable signatures across multiple tumor types including immune cell features in squamous/basal-like cancers. These Elastic Net DNA predictors could also be called from commonly used DNA-based gene panels containing only hundreds of genes, thus facilitating their use for “non-genetic” tumor information that could assist in guiding therapeutic decision making. Conclusions: Our results demonstrate an ability to build DNA CNA-based predictors for multiple complex cancer phenotypes for breast tumors. Once appropriately validated, a whole new set of prognostic and predictive biomarkers could be read out from existing DNA-based gene panels, thus providing more guidance for precision medicine at no additional cost. Citation Format: Youli Xia, Cheng Fan, Katherine A Hoadley, Joel S Parker, Charles M Perou. Genetic determinants of the molecular portraits of epithelial cancers [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-05-12.

Mode decomposition approach in robust control design for horizontal axis wind turbines

AbstractA robust multivariable strategy for pitch and torque control design of variable‐speed variable‐pitch wind turbines in the full load region is introduced in this paper. The pitch and torque control loops that share the tracking and active damping of drivetrain torsional mode objectives are designed simultaneously using a novel decomposition scheme. This permits the systematic design of robust multivariable controllers for wind turbines in a manner that facilitates industrial application. We achieve this by making the process of weighting function design fast, intuitive, and simple and by giving the designer a clear insight on the compromising aspects of the various control system objectives. FAST simulation is used to demonstrate application of the method.

Multivariable Control Strategy for a refrigeration plant

Multivariable Control Strategy for a refrigeration plant

MULTIVARIABLE ROBUST GRID CURRENTS CONTROL OF A SQUIRREL CAGE INDUCTION GENERATOR CONNECTED TO A WIND TURBINE

In this paper, a multivariable robust control strategy of a variable speed wind energy conversion system based on a squirrel cage induction generator (SCIG) is presented. Optimal speed control of the SCIG is achieved by a conventional PI controller combined with a Maximum Power Point Tracking (MPPT) strategy. The Indirect Field Oriented Control (IFOC) technique based on a simple Park (d-q) reference model is used to implement the proposed controller. Active and reactive power flow between the grid and the inverter is controlled via simultaneous H-infinity control of active and reactive currents of the grid and the DC link voltage. The switching technique of the inverter which is connected directly to the grid is controlled by a programmed PWM look-up table wave shaping. The robustness and performance of the proposed control scheme is evaluated on a comprehensive model of the wind energy conversion system using Matlab/Simulink.

Multimodel Modeling and Predictive Control for Direct-Drive Wind Turbine with Permanent Magnet Synchronous Generator

The safety and reliability of the wind turbines wholly depend on the completeness and reliability of the control system which is an important problem for the validity of the wind energy conversion systems (WECSs). A method based on multimodel modeling and predictive control is proposed for the optimal operation of direct-drive wind turbine with permanent magnet synchronous generator in this paper. In this strategy, wind turbine with direct-drive permanent magnet synchronous generator is modeled and a backpropagation artificial neural network is designed to estimate the wind speed loaded into the turbine model in real time through the estimated turbine shaft speed and mechanical power. The nonlinear wind turbine system is presented by multiple linear models. The desired trajectory of the nonlinear system is decomposed to be suitable for the reference trajectory of multiple models that are presented by the linear models of the nonlinear system, which simplifies the nonlinear optimization problems and decreases the calculation difficulty. Then a multivariable control strategy based on model predictive control techniques for the control of variable-speed variable-pitch wind turbines is proposed. Finally, simulation results are given to illustrate the effectiveness of the proposed strategy, and the conclusion that multiple model predictive controller (MMPC) has better control performance than the PI control method is obtained.

A Design Method of Multi-variable Economic Optimization Objective Function Based on Interval Predictive Control

For that there is still a lack of effective methods improving economic performance of the complex process systems in Model Predictive Control (MPC), a design method of multi-variable economic optimization objective function based on interval predictive control is proposed in this paper. The objective function consists of constraint item, control item and economic item, which are separately used to handle the constraint targets, coordinate the operational control and improve the economic indicators. In the process of solving the objective function, the multi-variable economic control optimization strategy is used. Fist, a multi-variable coordination optimization method is used preferentially to satisfy the target set point and constraint interval control of the Controlled Variables (CV s), then the weights of the economic item are adjusted according to the variable priority to achieve optimization of the Manipulated Variables (MV s). The simulation results show that the application of this objective function in MPC is not only able to achieve requirements of the product quality target, but also can significantly increase the economic benefits.

Hybrid DSO-GA-based sensorless optimal control strategy for wind turbine generators

The operation of a wind turbine generator involves natural uncertainty due to aerodynamic characteristics, resulting in a system that performs inefficiently. In general, the conventional controller now in wide use is not suitable for every operating point, because its tuning parameters and set-points do not meet the varying system characteristics. A study into an optimal control technique is conducted to reduce the negative effects of inherent uncertainty in system operation. In order to resolve the uncertainty problem, an optimal control method for an effective wind turbine generator is designed on the basis of a sensorless frame by utilizing a hybrid of the direct search optimization method (DSO) and the genetic algorithm (GA). This method is easy to implement and computation of functional derivatives is not necessary. The conventional GA is well known for its high performance in global optimization and its effectiveness in making ideal choices for control variables. The proposed DSO-GA hybrid differs from the conventional GA in terms of the sampling survey and the crossover operation. Moreover, the proposed multivariable optimal control strategy is a sensorless optimization technique that determines the pitch angle of the blades and the yaw angle of the nacelles to produce stable maximum power from a wind turbine system under steady-state operation. The proposed DSO-GA controller is implemented for a lab-scale wind turbine generator exposed to artificial wind, and the experimental results constitute a 3-D performance surface model of output voltage, which is used as an objective function for simulation. The optimization procedure with the objective function is carried out by means of the conventional and proposed methods, whose results reveal that the proposed DSO-GA optimizer yields far better performance in terms of generation number, convergence rate, and robustness. Both techniques are applied to a wind power generator through simulation and experiments. The performances are compared, and conclusions are drawn for each case.

A Mutivariable Methodology for Fast Visual Servoing of Flexible Manipulators Moving in a Restricted Workspace

High-speed visual servoing of manipulators including flexibilities is considered. In the proposed methodology, a model taking into account all the dynamics of the system is identified from the measurements usually available in visual servoing. This model is valid around the working position of the arm and can be easily recomputed after camera displacement. Two multivariable strategies are proposed for controlling the position of the end-effector: generalized predictive control and control. Comparative simulation and experimental results on a 2-degrees of freedom planar manipulator are given. A robustness evaluation step shows that the available working space is quite large.

Can matching improve the performance of boosting for identifying important genes in observational studies?

When two groups of individuals are to be compared with respect to gene expression there will often be some potentially confounding variables that differ between the groups. Matching is an established approach for obtaining comparable groups and enabling subsequent univariate tests for each gene. Alternatively, the confounders might be incorporated directly into a multivariable regression model for adjustment. In contrast to univariate tests, such models can consider all genes simultaneously. Aiming to combine the advantages of both approaches, matching and multivariable modeling, we consider a matching-based boosting procedure for fitting risk prediction models in two-group settings. This possibly allows to identify and automatically remove problematic observations that might negatively affect the regression model. Therefore, we compare the ability to identify important covariates for this combination of matching and boosting with only boosting for different covariate correlation structures in a simulation study. Furthermore, we analyze the prediction performance of these approaches on two gene expression microarray studies. The first study comprises patients with B-cell and T-cell type acute lymphoblastic leukemia and the second patients with acute megakaryoblastic leukemia. While the matching component can in principle guard against problematic observations, the combined approach is seen to neither improve identification of important covariates nor to improve prediction performance. Therefore, a combination of the two approaches cannot be recommended. Adjustment for potential confounders is seen to provide the best performance, i.e. a pure multivariable regression modeling strategy seems to be promising even in presence of considerable heterogeneity.

Multi-Variable Model Predictive Control of Automatic Variable-Load in Air Separation Unit Production Process

The process of air separation unit has the characteristics of many types of equipment, long flow, and strong relativity and complicates structures. A multi-variable model predictive control strategy to achieve the goal of automatic variable-load for the control limits in the variable-load process of air separation unit was proposed. Regard model predictive controller as the core, the problems of coupling in the various technological parameters were solved. Moreover, the non-linear control request for global variable-loading process was satisfied, and the perfect operation character and rate were ensured. Production characteristics, structure, scheme and application processes for the control strategy were described in detail. The system succeeds in applying to the air separation unit in an iron and steel enterprise, the multi-variable control system shows its effectiveness and efficiency in the automatic variable-loading.

Risk factors for cessation of breastfeeding prior to six months postpartum among a community sample of women in Calgary, Alberta.

To describe the rates of breastfeeding initiation and breastfeeding for at least six months and identify risk factors for failing to breastfeed for six months among a community sample of mothers in Calgary, Alberta. A cohort of women (n=1737) who participated in a longitudinal study of prenatal support and who could be contacted when their child was three-years-old (n=1147) were invited to participate in a follow-up telephone questionnaire. Of these 1147 women, 780 (69% participating rate) participated and provided breastfeeding data. Risk factors for early cessation of breastfeeding prior to six months were identified using bivariate and multivariable strategies. Of the 780 women, 95.6% initiated breastfeeding and 71.6% continued to breastfeed for at least six months. Risk factors identified for early cessation included younger maternal age, obesity prior to pregnancy, lower maternal education, working full-time or intending to within the first year, history of depression, depression or anxiety during pregnancy, poor social support, and smoking during pregnancy (all p<0.05). Multivariable analysis revealed that working full-time or intending to within the first year, lower maternal education, obesity prior to pregnancy and anxiety during pregnancy most increased a woman's risk of early cessation (all p<0.05). Nearly all mothers initiated breastfeeding and 70% continued to breastfeed for six months, although subgroups of women remained at an elevated risk of early cessation. Research to better understand breastfeeding decisions among women with the risk factors identified is needed to facilitate the development of more effective breastfeeding promotion strategies.

Risk Factors for Cessation of Breastfeeding Prior to Six Months Postpartum among a Community Sample of Women in Calgary, Alberta

Objectives: To describe the rates of breastfeeding initiation and breastfeeding for at least six months and identify risk factors for failing to breastfeed for six months among a community sample of mothers in Calgary, Alberta. Methods: A cohort of women (n=1737) who participated in a longitudinal study of prenatal support and who could be contacted when their child was three years old (n=1147) were invited to participate in a follow-up telephone questionnaire. Of these 1147 women, 780 (69% participating rate) participated and provided breastfeeding data. Risk factors for early cessation of breastfeeding prior to six months were identified using bivariate and multivariable strategies. Results: Of the 780 women, 95.6% initiated breastfeeding and 71.6% continued to breastfeed for at least six months. Risk factors identified for early cessation included younger maternal age, obesity prior to pregnancy, lower maternal education, working full-time or intending to within the first year, history of depression, depression or anxiety during pregnancy, poor social support, and smoking during pregnancy (all p Conclusion: Nearly all mothers initiated breastfeeding and 70% continued to breastfeed for six months, although subgroups of women remained at an elevated risk of early cessation. Research to better understand breastfeeding decisions among women with the risk factors identified is needed to facilitate the development of more effective breastfeeding promotion strategies. Key words: Breastfeeding; cessation; risk factors; epidemiology

Control-Oriented Modeling and Analysis of Air Management System for Fuel Reforming Fuel Cell Vehicle

With the purpose of meeting the specifically restrictive requirements of fuel reforming fuel cell vehicle, this paper brings into focus the issues of the transient operation of fuel cell systems and presents a control-oriented dynamic model of fuel cell air management system, suited for multivariable controller design, system optimization, and supervisory control strategy. In a first step, the dual analytical approach based on lumped and distributed parameter models is detailed: The partial differential equations deduced from mass/energy conservation laws and inertial dynamics are reduced to ordinary differential equations using spatial discretization and then combined with semiempirical actuator models to form the overall air system model. In a second step, a classical approach is followed to obtain a local linearization of the model. A validation of both nonlinear and linearized versions is performed by computational fluid dynamics simulations and experiments on a dedicated air system test bench. Thanks to dynamic analysis (pole/zero map), operating point impact and model order reduction are investigated. Finally, the multiinput multioutput state-space model—which balances model fidelity with model simplicity—can be coupled with reformer, stack, and thermal models to understand the system complexity and to develop model-based control methodologies.

Recurrent neural networks-based multivariable system PID predictive control

A nonlinear proportion integration differentiation (PID) controller is proposed on the basis of recurrent neural networks, due to the difficulty of tuning the parameters of conventional PID controller. In the control process of nonlinear multivariable system, a decoupling controller was constructed, which took advantage of multi-nonlinear PID controllers in parallel. With the idea of predictive control, two multivariable predictive control strategies were established. One strategy involved the use of the general minimum variance control function on the basis of recursive multi-step predictive method. The other involved the adoption of multistep predictive cost energy to train the weights of the decoupling controller. Simulation studies have shown the efficiency of these strategies.

A Novel Control Methodology for a Pilot Plant Azeotropic Distillation Column

A fundamental dynamic model was successfully used in the implementation of multiple control methodologies via a novel inferential control strategy using HYSYS to treat missing process measurements. Results from steady state and dynamic testing of an azeotropic distillation system of methanol, normal pentane, and cyclohexane were obtained. Steady-state equilibrium and nonequilibrium models were developed and validated with experimental data from a packed distillation unit operated at finite reflux. Dynamic multicomponent distillation experiments were also carried out, and experimental process data were collected using the pilot scale experimental unit. Two different variable pairings were studied, and the results from individual control loop configurations were compared with a multivariable control strategy using the model predictive control (MPC) software Predict Pro.