Real-time nonlinear MPC and MHE for a large-scale mechatronic application

Progress in optimization algorithms and in computational hardware made deployment of Nonlinear Model Predictive Control (NMPC) and Moving Horizon Estimation (MHE) possible to mechatronic applications. This paper aims to assess the computational performance of NMPC and MHE for rotational start-up of Airborne Wind Energy systems. The capabilities offered by an automatic code generation tool are experimentally verified on a real physical system, using a model comprising 27 states and 4 inputs at a sampling frequency of 25 Hz. The results show the feedback times less than 5 ms for the NMPC with more than 1500 variables.     

Nonlinear model predictive control of fed-batch fermentations using dynamic flux balance models

Fed-batch fermentation is an important production technology in the biochemical industry. Using fed-batch Saccharomyces cerevisiae fermentation as a prototypical example, we developed a general methodology for nonlinear model predictive control of fed-batch bioreactors described by dynamic flux balance models. The control objective was to maximize ethanol production at a fixed final batch time by adjusting the glucose feeding rate and the aerobic–anaerobic switching time. Effectiveness of the closed-loop implementation was evaluated by comparing the relative performance of NMPC and the open-loop optimal controller. NMPC was able to compensate for structural errors in the intracellular model and parametric errors in the substrate uptake kinetics and cellular energetics by increasing ethanol production between 8.0% and 14.7% compared with the open-loop operating policy. Minimal degradation in NMPC performance was observed when the biomass, glucose, and ethanol concentration and liquid volume measurements were corrupted with Gaussian white noise. NMPC based on the dynamic flux balance model was shown to improve ethanol production compared to the same NMPC formulation based on a simpler unstructured model. To our knowledge, this study represents the first attempt to utilize a dynamic flux balance model within a nonlinear model-based control scheme.     

Advanced control with parameter estimation of batch transesterification reactor

The objective of this work is to enhance the economic performance of a batch transesterification reactor producing biodiesel by implementing advanced, model based control strategies. To achieve this goal, a dynamic model of the batch reactor system is first developed by considering reaction kinetics, mass balances and heat balances. The possible plant-model mismatch due to inaccurate or uncertain model parameter values can adversely affect model based control strategies. Therefore, an evolutionary algorithm to estimate the uncertain parameters is proposed. It is shown that the system is not observable with the available measurements, and hence a closed loop model predictive control cannot be implemented on a real system. Therefore, the productivity of the reactor is increased by first solving an open-loop optimal control problem. The objective function for this purpose optimizes the concentration of biodiesel, the batch time and the heating and cooling rates to the reactor. Subsequently, a closed-loop nonlinear model predictive control strategy is presented in order to take disturbances and model uncertainties into account. The controller, designed with a reduced model, tracks an offline determined set-point reactor temperature trajectory by manipulating the heating and cooling mass flows to the reactor. Several operational scenarios are simulated and the results are discussed in view of a real application. With the proposed optimization and control strategy and no parameter mismatch, a revenue of 2.76 $ min⁻¹ can be achieved from the batch reactor. Even with a minor parameter mismatch, the revenue is still 2.01 $ min⁻¹. While these values are comparable to those reported in the literature, this work also accounts for the cost of energy. Moreover, this approach results in a control strategy that can be implemented on a real system with limited online measurements.     

Simultaneous dissolved oxygen and glucose regulation in fed-batch methionine production using decoupled input–output linearizing control

Methionine is one of the essential amino acids produced by fed-batch fermentation. The synthesis of methionine at the cellular level is strictly regulated and its process dynamics shows a nonlinear interaction between dissolved oxygen and glucose concentration. For controlling this process, a decoupled input–output linearizing controller (DIOLC) is derived. The model used for this purpose contains an exponential kinetic structure for describing the nonlinearities and metabolic switching function for describing oxygen dependency. The control system is square having two inputs and two outputs. The zero dynamics stability for internal variables and error convergence is proved. The performance of DIOLC is examined under high and low oxygen demand conditions. Four case studies are used to demonstrate that the DIOLC action is decoupled. The DIOLC also exhibited robust performance even for random variations up to ±20% in some parameters. In simulated experiments, using DIOLC produces 16 g l^?1 of methionine. The performance of PI controllers under identical conditions is given for comparison.     

Adaptive local tracking of a temperature profile in tubular reactor with partial measurements

In this work, a local constrained adaptive output feedback is presented for a class of exothermic tubular reactors models described by a nonlinear partial differential equations. The considered output is the measured temperature in a fixed zone of the reactor to regulate the temperature throughout the reactor to a ball with radius λ (arbitrarily small) centered at the fixed temperature profile. For a given measurement zone with length given in terms of the desired profile and λ and for initial temperature in a fixed domain, it is shown that the tracking error through the reactor tends asymptotically to a ball of arbitrary prescribed radius λ > 0, centered at the given temperature profile. Numerical simulations have been performed to illustrate the performance of the proposed approach.     

Characterization of DNA hybridization kinetics in a microfluidic flow channel

In this investigation we report on the influence of volumetric flow rate, flow velocity, complementary DNA concentration, height of a microfluidic flow channel and time on DNA hybridization kinetics. A syringe pump was used to drive Cy3-labeled target DNA through a polydimethylsiloxane (PDMS) microfluidic flow channel to hybridize with immobilized DNA from the West Nile Virus. We demonstrate that a reduction of channel height, while keeping a fixed volumetric flow rate or a fixed flow velocity, enhances mass transport of target DNA to the capture probes. Compared to a passive hybridization, the DNA hybridization in the microfluidic flow channel generates higher fluorescence intensities for lower concentration of target DNA during the same fixed period of time. Within a fixed 2 min time period the fastest DNA hybridization at a 50 pM concentration of target DNA is achieved with a continuous flow of target DNA at the highest flow rate and the lowest channel height.     

Free-living inferential modeling of blood glucose level using only noninvasive inputs

The goal of this work is to present a causation modeling methodology with the ability to accurately infer blood glucose levels using a large set of highly correlated noninvasive input variables over an extended period of time. These models can provide insight to improve glucose monitoring, and glucose regulation through advanced model-based control technologies. The efficacy of this approach is demonstrated using real data from a type 2 diabetic (T2D) subject collected under free-living conditions over a period of 25 consecutive days. The model was identified and tested using eleven variables that included three food variables as well as several activity and stress variables. The model was trained using 20 days of data and validated using 5 days of data. This gave a fitted correlation coefficient of 0.70 and an average absolute error (AAE) (i.e., the average of the absolute values for the measured glucose concentration minus modeled glucose concentration) of 13.3 mg/dL for the validation data. This AAE result was significantly better than the subject’s personal glucose meter AAE of 15.3 mg/dL for replicated measurements.     

Estimation and adaptive nonlinear model predictive control of selective catalytic reduction systems in automotive applications

The demands of high NO_x conversion efficiency and low tailpipe ammonia slip for urea-based selective catalytic reduction (SCR) systems have been substantially increased in the past decade, as NO_x emission legislations for Diesel engines are becoming more stringent than ever before. Since catalyst aging has a significant impact on SCR performance, robust and adaptive SCR control has been preferred for degraded SCR systems to realize emission control objectives. The purpose of this paper is twofold. Firstly, a robust ammonia coverage ratio observer was designed for estimating the ammonia coverage ratio reference for catalysts with different aging levels. An ammonia storage capacity observer was developed for estimating the actual ammonia storage capacity which can be reduced due to catalyst aging. An adaptive ammonia coverage ratio reference design was then developed to estimate the desired ammonia coverage ratio ranges at each instantaneous engine operating point for both single-cell and two-cell SCR systems at different aging levels based on a singular perturbation method. Secondly, to ensure the estimated ammonia coverage ratio falls in the desired ranges for most of engine operating conditions, robust nonlinear model predictive control (NMPC) algorithms were designed for both single-cell and two-cell SCR systems. Experimental data over US06 cycle were collected from a Diesel engine and aftertreatment system platform for controller verification. Simulation results under US06 test cycle demonstrate that the proposed NMPC algorithms were capable of consistently achieving high NO_x conversion efficiency (>95.6%) and constrained tailpipe ammonia slip (<10 ppm on average and <12 ppm on the peak) for both fresh catalyst and aged catalyst with 30% loss of ammonia storage capacity.     

Perchlorate-selective polymeric membrane electrode based on a cobaloxime as a suitable carrier

A cobaloxime ([chlorobis(dimethylglyoximeato)(triphenylphosphine)] cobalt (III), [Co(dmgH)₂pph₃Cl]) incorporated in a plasticized poly(vinyl chloride) membrane was used to develop a perchlorate-selective electrode. The influence of membrane composition on the electrode response was studied. The electrode exhibits a Nernstian response over the perchlorate concentration range 1.0 × 10⁻⁶ to 1 × 10⁻¹ mol l⁻¹ with a slope of −56.8 ± 0.7 mV per decade of concentration, a detection limit of 8.3 × 10⁻⁷, a wide working pH range (3–10) and a fast response time (<15 s). The electrode shows excellent selectivity towards perchlorate with respect to many common anions. The electrode was used to determine perchlorate in water and human urine.     

Software sensors are a real alternative to true sensors

At the Ejby Mølle WWTP in Odense Denmark a software sensor predicts the ammonium and nitrite + nitrate concentration in real-time based on ammonium and redox potential measurements. The predicted ammonium concentration is used to control the length of the nitrification phase in a Biodenipho^® activated sludge unit because the software sensor has a shorter response time and a better up-time than the ammonium meter. The software sensor simplifies meter service and can reduce maintenance costs. The computed nitrite + nitrate concentration is an added benefit of the software sensor. On 4 different days, series of grab samples of the mixed liquor were collected in the aeration tanks. The average difference between the ammonium concentrations in the grab samples and the predicted ammonium concentration was 0.2 mgN L^?1 and the average difference between the predicted and the measured nitrite + nitrate concentration was 0.3 mgN L^?1. The agreement between the predicted and the measured ammonium concentration in the grab samples was better than the agreement between the ammonium meter and the grab samples. This was due to the shorter response time of the software sensor compared with the ammonium meter.     

Piezoelectric unimorph valve assembled on an LTCC substrate

In this paper a piezoelectric initially open valve was designed in low temperature co-fired ceramic (LTCC), manufactured using standard processes, and tested with integrated gas channels inside the LTCC module. Actuation of the valve was based on a piezoelectric unimorph with a diameter of 15 mm and thickness of 0.35 mm glued onto the fired LTCC substrate. Subsequently, a series of tests, including flow, displacement and switching time measurements, was carried out. Measurements of the valve revealed a flow of 143 ml/min under 1 bar pressure, leakage levels of 4%, valve displacement of 1.3 μm, and closing times less than 30 ms. Additional miniaturization and integration of an embedded valve in the LTCC will be pursued, enabling improved manufacturing as a batch process and micro- and nano-litre fluid management for various applications.     

Experimental blood glucose interval identification of patients with type 1 diabetes

Many problems are confronted when characterizing a type 1 diabetic patient such as model mismatches, noisy inputs, measurement errors and huge variability in the glucose profiles. In this work we introduce a new identification method based on interval analysis where variability and model imprecisions are represented by an interval model as parametric uncertainty.The minimization of a composite cost index comprising: (1) the glucose envelope width predicted by the interval model, and (2) a Hausdorff-distance-based prediction error with respect to the envelope, is proposed. The method is evaluated with clinical data consisting in insulin and blood glucose reference measurements from 12 patients for four different lunchtime postprandial periods each.Following a “leave-one-day-out” cross-validation study, model prediction capabilities for validation days were encouraging (medians of: relative error = 5.45%, samples predicted = 57%, prediction width = 79.1 mg/dL). The consideration of the days with maximum patient variability represented as identification days, resulted in improved prediction capabilities for the identified model (medians of: relative error = 0.03%, samples predicted = 96.8%, prediction width = 101.3 mg/dL). Feasibility of interval models identification in the context of type 1 diabetes was demonstrated.     

Development of wireless compressive/relaxation-stress measurement system integrated with pressure-sensitive carbon black-filled silicone rubber-based sensors

In order to detect the installation compressive stress and monitor the stress relaxation between two bending surfaces on a defensive furnishment, a wireless compressive-stress/relaxation-stress measurement system based on pressure-sensitive sensors is developed. The flexible pressure-sensitive stress sensor array is fabricated by using carbon black-filled silicone rubber-based composite. The wireless stress measurement system integrated with this sensor array is tested with compressive stress in the range from 0 MPa to 3 MPa for performance evaluation. Experimental results indicate that the fractional change in electrical resistance of the pressure-sensitive stress sensor changes linearly and reversibly with the compressive stress, and its fractional change goes up to 355% under uniaxial compression; the change rate of the electrical resistance can track the relaxation stress and give out a credible measurement in the process of stress relaxation. The relationship between input (compressive stress) and output (the fractional change in electrical resistance) of the pressure-sensitive sensor is ΔR/R₀ = σ × 1.2 MPa^?1. The wireless compressive stress measurement system can be used to achieve sensitivity of 1.33 V/MPa to the stress at stress resolution of 920.3 Pa. The newly developed wireless stress measurement system integrated with pressure-sensitive carbon black-filled silicone rubber-based sensors has advantages such as high sensitivity to stress, high stress resolution, simple circuit and low energy consumption.     

A MEMS-based resonant-scanning lamellar grating Fourier transform micro-spectrometer with laser reference system

A lamellar grating Fourier transform infra-red (FTIR) micro-spectrometer is presented in which the device is electromagnetically actuated in resonant mode so as to achieve larger displacements with a lower driving voltage. By actuating at resonance, we can also have a design with a higher spring stiffness design such that the micro-spectrometer will have little influence from external perturbation. A data acquisition electronic system is designed such that the interferogram of the IR source can still be acquired at a fixed optical path distance (OPD) intervals. This is achieved by using a reference laser source. Working at a resonant frequency of 330 Hz, a 100 μm (bi-directional) displacement is achieved by the device with an input voltage of 2.2 V. A tunable laser source is used to demonstrate the system performance. The peak of the recorded spectra is very close to the actual wavelength of the IR, with a maximum difference of less than 5 nm.     

Electrochemical determination of ascorbic acid in fruits on a vanadium oxide polypropylene carbonate modified electrode

A novel vanadium oxide polypropylene carbonate modified glassy carbon electrode was developed and used for the measurement of ascorbic acid (AA). The electrode was prepared by casting a mixture of vanadium tri(isopropoxide) oxide (VO(OC₃H₇)₃) and poly(propylene carbonate) (PPC) onto the surface of a glassy carbon electrode. The electrochemical behavior of the VO(OC₃H₇)₃–PPC film modified glassy carbon electrode was investigated by cyclic voltammetry and amperometry. This modified electrode exhibited electrocatalytic response to the oxidation of ascorbic acid. Compared with a bare glassy carbon electrode, the modified electrode exhibits a 220 mV shift of the oxidation potential of ascorbic acid in the cathodic direction and a marked enhancement of the current response. The response current revealed a good linear relationship with the concentration of ascorbic acid in the range of 4 × 10⁻⁸ and 1 × 10⁻⁴ mol L⁻¹ and the detection limit of 1.5 × 10⁻⁸ mol L⁻¹ (S/N = 3) in the pH 8.06 Britton–Robinson solution. Quantitative recovery of the ascorbic acid in synthetic samples has been obtained and the interferences from different species have been studied. The method has been successfully applied to the determination of ascorbic acid in fruits. The concentrations of ascorbic acid measured by this method are in good agreement with the literature value. It is much promising for the modified films to be used as an electrochemical sensor for the detection of ascorbic acid.     

Moisture sensor at glass/polymer interface for monitoring of photovoltaic module encapsulants

A sensor developed for measurement of water concentration inside glass/polymer encapsulation structures with a particular application area in accelerated aging of photovoltaic module encapsulants is described. An approximately 5 μm thick porous TiO₂ film applied to a glass substrate with a conductive coating acts as the moisture-sensitive component. The response is calibrated with weather chamber experiments for sensors open to the environment and with diffusion experiments for sensors laminated under an encapsulant. For the interpretation of diffusion experiment results, a transport model describing the diffusion of water across the polymer/TiO₂ interface is developed. The logarithm of AC resistance shows a linear dependence on water concentration in both open and encapsulated calibration. The first measurable response from an encapsulated 3.5 mm × 8 mm size sensor is obtained when approximately 10 μg of water has entered the film. Implications of the calibration results for sensor usage in accelerated aging tests are discussed.     

Design and performance of a microcantilever-based hydrogen sensor

This paper describes the design of, and the effects of basic environmental parameters on, a microelectromechanical (MEMS) hydrogen sensor. The sensor contains an array of 10 micromachined cantilever beams. Each cantilever is 500 μm wide×267 μm long×2 μm thick and has a capacitance readout capable of measuring cantilever deflection to within 1 nm. A 20-nm-thick coating of 90% palladium–10% nickel bends some of the cantilevers in the presence of hydrogen. The palladium–nickel coatings are deposited in ultra-high-vacuum (UHV) to ensure freedom from a “relaxation” artifact apparently caused by oxidation of the coatings. The sensor consumes 84 mW of power in continuous operation, and can detect hydrogen concentrations between 0.1 and 100% with a roughly linear response between 10 and 90% hydrogen. The response magnitude decreases with increasing temperature, humidity, and oxygen concentration, and the response time decreases with increasing temperature and hydrogen concentration. The 0–90% response time of an unheated cantilever to 1% hydrogen in air is about 90 s at 25 °C and 0% humidity.     

Investigation of Ni-based thermal bimaterial structure for sensor and actuator application

Thermal bimaterial structures made of Ni and Ni-diamond nanocomposite for sensor and actuator application are proposed, fabricated, and tested. Two deflection types of thermal bimaterial structures, including upward and downward bending types, can be easily fabricated by controlling electroplating sequence of Ni and Ni-diamond nanocomposite. According to thermal performance measurement, the tip deflection of upward and downward types can reach about 82.5 μm and ?22.5 μm for a temperature change of 200 °C, respectively. In the condition, the thermomechanical sensitivity and output force are 412.5 nm/K and 97.0 μN for upward type thermal bimaterial structure; and ?112.5 nm/K and ?26.5 μN for downward type one. Due to the low electroplating process temperature (～50 °C) for both Ni-based layers, diminutive pre-deformation of as-fabricated structure and strong interlaminar bonding strength are verified by SEM and vibrational test. The resonant frequency of the structure remains unchanged after 10⁹ cycles.     

Fabrication and frequency response of dual-element ultrasonic transducer using PZT-5A thick film

Ultrasonic transducers based on PZT-5A thick films deposited onto polycrystalline Al₂O₃ substrates using screen-printing were successfully fabricated. Considering the relatively high sintering temperature of PZT-5A thick films and better impedance matching characteristics with PZT-5A, polished polycrystalline Al₂O₃ were used as substrates. For electrodes, high quality platinum (Pt) was deposited by a thin film process, because the surface state of electrodes greatly affects the quality of piezoelectric films. Applying Pt/PZT-5A/Pt/Al₂O₃ structures, dual-element ultrasonic transducers were assembled. The assembled transducers included a wear plate (normally alumina with 40.21 × 10⁶ kg/m² s of impedance), backing (tungsten carbide-epoxy), electrical matching, an epoxy glue layer, and a housing. The optimum measurement ranges of 5 and 10 MHz ultrasonic transducers were 2.51–300.2 and 2.50–250.1 mm, respectively. From the time and frequency response measurements of the assembled 10 MHz DEUTs, the value of −20 dB level waveform duration and the −6 dB bandwidth was 481.8 ns and 34.4%, respectively. Also, the measurement accuracies of both 5 and 10 MHz DEUTs assembled in this study were below 0.1 and 0.4%, respectively.