Browse
Record Types
Records with Type: Postprint
A 2-stage Approach to Parameter Estimation of Differential Equations using Neural ODEs
William Bradley, Fani Boukouvala
November 7, 2021 (v1)
Keywords: Neural ODEs, Neural-Networks, Nonlinear programming, parameter estimation
Modeling physio-chemical relationships using dynamic data is a common task in fields throughout science and engineering. A common step in developing generalizable, mechanistic models is to fit unmeasured parameters to measured data. However, fitting differential equation-based models can be computation intensive and uncertain due to the presence of nonlinearity, noise, and sparsity in the data, which in turn causes convergence to local minima and divergence issues. This work proposes a merger of Machine Learning (ML) and mechanistic approaches by employing ML models as a means to fit nonlinear mechanistic ODEs. Using a two-stage indirect approach, Neural ODEs are used to estimate state derivatives, which are then used to estimate the parameters of a more interpretable mechanistic ODE model. In addition to its computational efficiency, the proposed method demonstrates the ability of Neural ODEs to better estimate derivative information than interpolating methods based on algebraic... [more]
High frequency injection maximum power point tracking for thermoelectric generators
Romina Rodriguez, Jing Guo, Matthias Preindl, James S. Cotton, Ali Emadi
October 9, 2019 (v1)
Keywords: DC-DC power converter, Energy harvesting, High frequency injection (HFI), Maximum power point tracking (MPPT), Perturb & observe (P&O), Thermoelectric generator (TEG)
Thermoelectric generators (TEGs) can harvest thermal energy from waste heat sources to supply various power levels due to the Seebeck effect. The power generated by a TEG is dependent not only on the temperature difference across them but also on the electrical load applied. Typically, waste heat sources have variable operating conditions which means maximum power point tracking (MPPT) must be employed through the use of power converters to produce the desired operating point of the system and thus increase the system efficiency. This paper presents a new MPPT scheme which has not been previously applied to thermoelectric generators, the high frequency injection (HFI) scheme to achieve a fast and accurate tracking of the maximum power operation point for TEGs. The proposed MPPT scheme is implemented with a power converter, and the tracking scheme performance is experimentally evaluated on a commercial TEG module through three different experiments. The proposed scheme is also compared... [more]
Finding better limit cycles of semicontinuous distillation. Part 1: Back-stepping design Methodology
Pranav Bhaswanth Madabhushi, Thomas Adams II
July 9, 2019 (v1)
Keywords: Hybrid dynamic system, Limit Cycle, Particle Swarm Optimization, Process Design, Semicontinuous Distillation
Semicontinuous ternary zeotropic distillation is a periodic process that is carried out
in a single distillation column and a tightly integrated external middle vessel. In the
state-of-the-art design procedure of this process, a continuous distillation process that
separates the top and bottoms products to the desired purity is used to generate an arbitrary
initial state for simulating the dynamics of the semicontinuous distillation process.
Although this method is useful in estimating the limit cycle, it was later found that the
operation of the process in this limit cycle was economically sub-optimal. In this study,
a new algorithmic design procedure, called the back-stepping design methodology, is
proposed to find better limit cycles for zeotropic ternary semicontinuous distillation
using the aspenONE Engineering suite. The proposed methodology was applied to two
different case studies using feed mixtures with different chemical components. A comparison
with the current d... [more]
Supply Chain Optimization of Flare-Gas-To-Butanol Processes in Alberta
Leila Hoseinzade, Thomas A Adams II
August 15, 2018 (v1)
Subject: Optimization
Keywords: Portable technology, Supply chain optimization, Sustainability
In this work, the economic feasibility of combining a novel portable gas-to-methanol process with a novel methanol-to-butanol process is examined. The gas-to-methanol process converts waste flare gas into methanol using a series of truck-mounted devices deployed at oil production wellheads. The methanol-to-butanol process uses a new proprietary catalyst which produces butanol via a diketene intermediate at a large centralized facility. The goal of this work is to identify the best ways of commercializing this technology in Alberta. To do this, a supply chain optimization model is formulated which considers specifically how many gas-to-methanol trucks should be used and where specifically in Alberta they should be deployed, the specific suppliers of CO2 to use, where the location of the central methanol-to-butanol facility should be chosen, and the costs of transportation of materials between locations. The model framework also considers the possibility of getting methanol in full or in... [more]
Application of a Two-Level Rolling Horizon Optimization Scheme to a Solid-Oxide Fuel Cell and Compressed Air Energy Storage Plant for the Optimal Supply of Zero-Emissions Peaking Power
Jake Nease, Nina Monteiro, Thomas A. Adams II
June 19, 2018 (v1)
We present a new two-level rolling horizon optimization framework applied to a zero-emissions coal-fueled solid-oxide fuel cell power plant with compressed air energy storage for peaking applications. Simulations are performed where the scaled hourly demand for the year 2014 from the Ontario, Canada market is met as closely as possible. It was found that the proposed two-level strategy, by slowly adjusting the SOFC stack power upstream of the storage section, can improve load-following performance by 86% compared to the single-level optimization method proposed previously. A performance analysis indicates that the proposed approach uses the available storage volume to almost its maximum potential, with little improvement possible without changing the system itself. Further improvement to load-following is possible by increasing storage volumes, but with diminishing returns. Using an economically-focused objective function can improve annual revenue generation by as much as 6.5%, but no... [more]
Techno-economic comparison of Acetone-Butanol-Ethanol fermentation using various extractants
Giancarlo Dalle Ave, Thomas A. Adams II
June 12, 2018 (v1)
Keywords: Acetone, Butanol, Cost of CO2 Emissions Avoided, Ethanol, Extraction, Fermentation, Technoeconomic Analysis
This work compares various chemicals for use as extractants in second-generation Acetone-Butanol-Ethanol fermentation on economic and environmental bases. Both non-toxic and toxic extractants are considered in this study. The combinative extractive-distillation separation process was modelled using a combination of Microsoft Excel 2013, MATLAB 2015 and Aspen Plus v8.8. Separation trains were designed and optimized for each extractant to best take advantage of extractant properties. Upstream units considered in this analysis include: biomass (switchgrass) solids processing, biomass pre-treatment and saccharification, and fermentation. Downstream processes considered include utility generation and wastewater treatment. The cost of CO2 equivalent emissions avoided (CCA) was used as the metric to compare the environmental impact of each process as compared to conventional petroleum-based gasoline. The economic and environmental best extractant is shown to be 2-ethyl-hexanol with a minimum... [more]
Direct Steam Generation Concentrated Solar Power Plant with a Decalin/Naphthalene Thermochemical Storage System
Haoxiang Lai, Thomas A. Adams II
November 20, 2018 (v2)
This study presents the design and analysis of a new integrated direct steam generation (DSG) concentrated solar power (CSP) plant with a decalin/naphthalene thermochemical storage system. Model simulations were performed in accordance to historical hourly solar radiation data over a year, using a combination of Aspen Plus v10, MATLAB 2016b, and Microsoft Excel VBA. It was found that the proposed plant feasibly stored and discharged energy, based on the solar radiation and chemical storage availability, to maintain base-load power productions (250 MW or 120 MW) with an overall efficiency of 14.6%. The effectiveness of the designed storage system was found to be comparable to a molten salt storage system which is currently used in existing CSP plants. The proposed integrated DSG CSP plant with a decalin/naphthalene thermochemical storage system shows promise for being an alternative to existing CSP plants.
Dynamic modeling of integrated mixed reforming and carbonless heat systems
Leila Hoseinzade, Thomas A. Adams II
June 12, 2018 (v1)
Keywords: Carbonless Heat, Dry Reforming, Dynamic Modelling, Integrated Systems, Steam Reforming, Syngas
In the previous study, a dynamic and two-dimensional model for a steam methane reforming process integrated with nuclear heat production was developed. It was shown that the integrated high temperature gas-cooled reactor (HTGR)/steam methane reforming (SMR) is an efficient process for applications such as hydrogen production. In this study, it is demonstrated that combining nuclear heat with the mix of steam and dry reforming process can be a promising option to achieve certain desired H2/CO ratios for Fischer-Tropsch or other downstream energy conversion processes. The model developed in the previous study is extended to the combined steam and dry reforming process. The resulting model was validated using reported experimental data at non-equilibrium and equilibrium conditions. The dynamic and steady state performance of the integrated mixed reforming of methane and nuclear heat system was studied and it was found that in addition to desired H2/CO ratios, higher methane conversion and... [more]
[Show List of Record Types]