Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Mathematical optimization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 361 records · Page 20

Overview of Upcoming Process Improvement Efforts at the Saltstone Facilities at the Savannah River Site - 20474

The Saltstone Facilities at the Savannah River Site (SRS) process and dispose of low activity salt solution waste generated by other liquid waste facilities. Today in the Liquid Waste System (LWS), approximately 125 million liters of salt waste remain to be processed. The Saltstone Production Facility (SPF) receives the decontaminated salt waste, which is mixed with dry feeds consisting of cement, fly ash and slag to produce the saltstone. The mixed slurry is pumped to the Saltstone Disposal Units (SDUs) for safe disposal in the Saltstone Disposal Facility (SDF). To support continuous Salt Waste Processing Facility (SWPF) operations at annual processing volumes many times higher than present, significant modifications are required to ensure that the SPF has the capacity to support the throughput of salt waste. These changes include several infrastructure upgrades that are being worked through the Enhanced Low Activity Waste Disposal (ELAWD) project. In addition to infrastructure upgrades, Savannah River Remediation (SRR), the Savannah River Site's Liquid Waste Contractor for the U.S. Department of Energy, has identified a flowsheet improvement opportunity of implementing a cement-free saltstone formulation to eliminate cement from the current three component formulation. This two-component formulation would simplify the procurement, transport, off-loading, and storage of the premix materials, and increase the storage capacity for the individual components. This flowsheet, when implemented, will reduce operational risk and improve throughput necessary to achieve disposal rates necessary for SWPF operations. Research has been completed to confirm the feasibility of this formulation, with the new cement-free flowsheet expected to be implemented in the next few years. Improvements have also been identified downstream of the SPF to increase SDU utilization. Specifically, a concentrated effort has taken place to raise the fill height in SDU 6 to the full disposal height, 13 meters, to maximize the usage of each SDU in support of the liquid waste disposal mission. This improvement has been made through additional thermal modeling and improved process understanding to create a SDU flammability model and control strategy more representative of facility operations. This work will allow for the completion of construction and operational readiness for SDU 7 while filling SDU 6 to ensure there is no lapse in available space once SWPF starts up. A summary of these improvements as they relate to increased throughput will be provided along with associated status. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

The impacts of convex piecewise linear cost formulations on AC optimal power flow

Despite strong connections through shared application areas, research efforts on power market optimization (e.g., unit commitment) and power network optimization (e.g., optimal power flow) remain largely independent. A notable illustration of this is the treatment of power generation cost functions, where nonlinear network optimization has largely used polynomial representations and market optimization has adopted piecewise linear encodings. This work combines state-of-the-art results from both lines of research to understand the best mathematical formulations of the nonlinear AC optimal power flow problem with piecewise linear generation cost functions. An extensive numerical analysis of non-convex models, linear approximations, and convex relaxations across fifty-four realistic test cases illustrates that nonlinear optimization methods are surprisingly sensitive to the mathematical formulation of piecewise linear functions. The results indicate that a poor formulation choice can slow down algorithm performance by a factor of ten, increasing the runtime from seconds to minutes. Furthermore, these results provide valuable insights into the best formulations of nonlinear optimal power flow problems with piecewise linear cost functions, an important step towards building a new generation of energy markets that incorporate the nonlinear AC power flow model.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

An optimization approach for fitting canonical tensor decompositions.

Tensor decompositions are higher-order analogues of matrix decompositions and have proven to be powerful tools for data analysis. In particular, we are interested in the canonical tensor decomposition, otherwise known as the CANDECOMP/PARAFAC decomposition (CPD), which expresses a tensor as the sum of component rank-one tensors and is used in a multitude of applications such as chemometrics, signal processing, neuroscience, and web analysis. The task of computing the CPD, however, can be difficult. The typical approach is based on alternating least squares (ALS) optimization, which can be remarkably fast but is not very accurate. Previously, nonlinear least squares (NLS) methods have also been recommended; existing NLS methods are accurate but slow. In this paper, we propose the use of gradient-based optimization methods. We discuss the mathematical calculation of the derivatives and further show that they can be computed efficiently, at the same cost as one iteration of ALS. Computational experiments demonstrate that the gradient-based optimization methods are much more accurate than ALS and orders of magnitude faster than NLS.

97 MATHEMATICS AND COMPUTING↗

Practical and Optimal Sequential Bayesian Experimental Design for Complex Systems Incorporating Human Experimenter Preferences (Final Scientific/Technical Report)

Experiments are indispensable for developing models of complex systems. Carefully designed experiments can provide substantial savings for these expensive data-acquisition opportunities. However, designs based on heuristics are often suboptimal for systems with multiphysics, nonlinear dynamics, and uncertain and noisy environments. Optimal experimental design, while leveraging predictive models, seeks to systematically quantify and maximize the value of experiments. In this project, we focused on the design of multiple experiments, where current approaches are largely suboptimal: batch-design does not adapt to new data acquired during the experiment campaign (no feedback), and greedy/myopic design ignores future dynamics and consequences (no lookahead). We developed the mathematical framework and computational methods for sequential optimal experimental design (sOED) for complex systems. We enabled tractable model-based sOED in a rigorous manner through novel algorithms based on reinforcement learning, and investigated the effects of human experimenters on the design process. Our methods are fully Bayesian, able to quantify and update uncertainty in a principled manner. The traits aimed by our approach—mathematical rigor and optimality, human effects and uncertainty quantification, computational practicality—are crucial for elevating the standards of artificial intelligence (AI) to support decision-making in scientific domains, and contribute toward trust and realistic adoption of AI in experimental design practice.

97 MATHEMATICS AND COMPUTING↗

Designing networks: A mixed-integer linear optimization approach

Designing networks with specified collective properties is useful in a variety of application areas, enabling the study of how given properties affect the behavior of network models, the downscaling of empirical networks to workable sizes, and the analysis of network temporal evolution. Despite the importance of the task, there currently exists a gap in our ability to systematically generate networks that adhere to theoretical guarantees for the given property specifications. In this work, we propose the use of Mixed-Integer Linear Optimization modeling and solution methodologies to address this Network Generation Problem. We present useful modeling techniques and apply them to mathematically express and constrain a broad class of network properties in the context of an optimization formulation. We derive complete formulations for the generation of networks that attain specified levels of connectivity, spread, assortativity and robustness, and we illustrate these via a number of computational case studies.

97 MATHEMATICS AND COMPUTING↗

Stochastic Strategic Participation of Active Distribution Networks With High-Penetration DERs in Wholesale Electricity Markets

With the increasing penetration of distributed energy resources (DERs), traditional distribution networks as load-serving entities in wholesale electricity markets, now evolve towards active distribution networks (ADNs) which can proactively participate in wholesale markets by optimally controlling the DERs in their networks. A stochastic bilevel optimization model is proposed in this paper for the strategic participation of ADNs and DERs to provide energy and grid services in wholesale electricity markets. The bilevel optimization model can capture the interactions between the ADN and the wholesale energy and ancillary service markets, considering the uncertainties of DERs in the ADN. In the upper-level model, the ADN makes optimal decisions on energy and reserve bidding considering the availability, uncertainties, and flexibility of DERs. The joint energy and reserve market-clearing of the independent system operator (ISO) is modeled as the lower-level problem. Using strong duality theory and Karush-Kuhn Tucker (KKT) conditions, the proposed bilevel optimization problem is reformulated as mathematical programming with equilibrium constraints (MPEC) problem and further converted into a computationally-solvable mixed-integer second-order-cone programming (MISOCP) model. The simulation results demonstrate the effectiveness of the model and the interactions between an ADN and wholesale electricity markets.

active distribution network↗

Modeling design and control problems involving neural network surrogates

Here, we consider nonlinear optimization problems that involve surrogate models represented by neural networks. We demonstrate first how to directly embed neural network evaluation into optimization models, highlight a difficulty with this approach that can prevent convergence, and then characterize stationarity of such models. We then present two alternative formulations of these problems in the specific case of feedforward neural networks with ReLU activation: as a mixed-integer optimization problem and as a mathematical program with complementarity constraints. For the latter formulation we prove that stationarity at a point for this problem corresponds to stationarity of the embedded formulation. Each of these formulations may be solved with state-of-the-art optimization methods, and we show how to obtain good initial feasible solutions for these methods. We compare our formulations on three practical applications arising in the design and control of combustion engines, in the generation of adversarial attacks on classifier networks, and in the determination of optimal flows in an oil well network.

97 MATHEMATICS AND COMPUTING↗

Shape Optimization for Navier-Stokes Equations with Algebraic Turbulence Model: Numerical Analysis and Computation

We study the shape optimization problem for the paper machine headbox which distributes a mixture of water and wood fibers in the paper making process. The aim is to find a shape which a priori ensures the given velocity profile on the outlet part. The mathematical formulation leads to the optimal control problem in which the control variable is the shape of the domain representing the header, the state problem is represented by the generalized Navier-Stokes system with nontrivial boundary conditions. This paper deals with numerical aspects of the problem.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Applied Mathematics at the U.S. Department of Energy: Past, Present and a View to the Future

Over the past half-century, the Applied Mathematics program in the U.S. Department of Energy's Office of Advanced Scientific Computing Research has made significant, enduring advances in applied mathematics that have been essential enablers of modern computational science. Motivated by the scientific needs of the Department of Energy and its predecessors, advances have been made in mathematical modeling, numerical analysis of differential equations, optimization theory, mesh generation for complex geometries, adaptive algorithms and other important mathematical areas. High-performance mathematical software libraries developed through this program have contributed as much or more to the performance of modern scientific computer codes as the high-performance computers on which these codes run. The combination of these mathematical advances and the resulting software has enabled high-performance computers to be used for scientific discovery in ways that could only be imagined at the program's inception. Our nation, and indeed our world, face great challenges that must be addressed in coming years, and many of these will be addressed through the development of scientific understanding and engineering advances yet to be discovered. The U.S. Department of Energy (DOE) will play an essential role in providing science-based solutions to many of these problems, particularly those that involve the energy, environmental and national security needs of the country. As the capability of high-performance computers continues to increase, the types of questions that can be answered by applying this huge computational power become more varied and more complex. It will be essential that we find new ways to develop and apply the mathematics necessary to enable the new scientific and engineering discoveries that are needed. In August 2007, a panel of experts in applied, computational and statistical mathematics met for a day and a half in Berkeley, California to understand the mathematical developments required to meet the future science and engineering needs of the DOE. It is important to emphasize that the panelists were not asked to speculate only on advances that might be made in their own research specialties. Instead, the guidance this panel was given was to consider the broad science and engineering challenges that the DOE faces and identify the corresponding advances that must occur across the field of mathematics for these challenges to be successfully addressed. As preparation for the meeting, each panelist was asked to review strategic planning and other informational documents available for one or more of the DOE Program Offices, including the Offices of Science, Nuclear Energy, Fossil Energy, Environmental Management, Legacy Management, Energy Efficiency & Renewable Energy, Electricity Delivery & Energy Reliability and Civilian Radioactive Waste Management as well as the National Nuclear Security Administration. The panelists reported on science and engineering needs for each of these offices, and then discussed and identified mathematical advances that will be required if these challenges are to be met. A review of DOE challenges in energy, the environment and national security brings to light a broad and varied array of questions that the DOE must answer in the coming years. A representative subset of such questions includes: (1) Can we predict the operating characteristics of a clean coal power plant? (2) How stable is the plasma containment in a tokamak? (3) How quickly is climate change occurring and what are the uncertainties in the predicted time scales? (4) How quickly can an introduced bio-weapon contaminate the agricultural environment in the US? (5) How do we modify models of the atmosphere and clouds to incorporate newly collected data of possibly of new types? (6) How quickly can the United States recover if part of the power grid became inoperable? (7) What are optimal locations and communication protocols for sensing devices in a remote-sensing network? (8) How can new materials be designed with a specified desirable set of properties? In comparing and contrasting these and other questions of importance to DOE, the panel found that while the scientific breadth of the requirements is enormous, a central theme emerges: Scientists are being asked to identify or provide technology, or to give expert analysis to inform policy-makers that requires the scientific understanding of increasingly complex physical and engineered systems. In addition, as the complexity of the systems of interest increases, neither experimental observation nor mathematical and computational modeling alone can access all components of the system over the entire range of scales or conditions needed to provide the required scientific understanding.

99 GENERAL AND MISCELLANEOUS↗

graphenv: a Python library for reinforcement learning on graph search spaces

Many important and challenging problems in combinatorial optimization (CO) can be expressed as graph search problems, in which graph vertices represent full or partial solutions and edges represent decisions that connect them. Graph structure not only introduces strong relational inductive biases for learning (Battaglia et al., 2018) - in this context, by providing a way to explicitly model the value of transitioning (along edges) between one search state (vertex) and the next - but lends itself to problems both with and without clearly defined algebraic structure. For example, classic CO problems on graphs such as the Traveling Salesman Problem (TSP) can be expressed as either pure graph search or integer programs. Other problems, however, such as molecular optimization, do no have concise algebraic formulations and yet are readily implemented as a graph search (V. et al., 2022; Zhou et al., 2019). Such "model-free" problems constitute a large fraction of modern reinforcement learning (RL) research owing to the fact that it is often much easier to write a forward simulation that expresses all of the state transitions and rewards, than to write down the precise mathematical expression of the full optimization problem. In the case of molecular optimization, for example, one can use domain knowledge alongside existing software libraries to model the effect of adding a single bond or atom to an existing but incomplete molecule, and let the RL algorithm build a model of how good a given decision is by "experiencing" the simulated environment many times through. In contrast, a model-based mathematical formulation that fully expresses all the chemical and physical constraints is intractable. In recent years, RL has emerged as an effective paradigm for optimizing searches over graphs and led to state-of-the-art heuristics for games like Go and chess, as well as for classical CO problems such as the TSP. This combination of graph search and RL, while powerful, requires non-trivial software to execute, especially when combining advanced state representations such as Graph Neural Networks (GNN) with scalable RL algorithms.

97 MATHEMATICS AND COMPUTING↗

Time complexity and gate complexity

We formulate and investigate the simplest version of time-optimal quantum computation theory (TO-QCT), where the computation time is defined by the physical one and the Hamiltonian contains only one- and two-qubit interactions. This version of TO-QCT is also considered as optimality by sub-Riemannian geodesic length. The work has two aims: One is to develop a TO-QCT itself based on a physically natural concept of time, and the other is to pursue the possibility of using TO-QCT as a tool to estimate the complexity in conventional gate-optimal quantum computation theory (GO-QCT). In particular, we investigate to what extent is true the following statement: Time complexity is polynomial in the number of qubits if and only if gate complexity is also. In the analysis, we relate TO-QCT and optimal control theory (OCT) through fidelity-optimal computation theory (FO-QCT); FO-QCT is equivalent to TO-QCT in the limit of unit optimal fidelity, while it is formally similar to OCT. We then develop an efficient numerical scheme for FO-QCT by modifying Krotov's method in OCT, which has a monotonic convergence property. We implemented the scheme and obtained solutions of FO-QCT and of TO-QCT for the quantum Fourier transform and a unitary operator that does not have an apparent symmetry. The former has a polynomial gate complexity and the latter is expected to have an exponential one which is based on the fact that a series of generic unitary operators has an exponential gate complexity. The time complexity for the former is found to be linear in the number of qubits, which is understood naturally by the existence of an upper bound. The time complexity for the latter is exponential in the number of qubits. Thus, both the targets seem to be examples satisfyng the preceding statement. The typical characteristics of the optimal Hamiltonians are symmetry under time reversal and constancy of one-qubit operation, which are mathematically shown to hold in fairly general situations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Comprehensive Energy Assessment: EE and RE Project Optimization Modeling for United States Pacific Command (USPACOM) American Recovery and Reinvestment Act (ARRA) FEMP Technical Assistance

This report summarizes a study and corresponding model development conducted in support of the United States Pacific Command (USPACOM) as part of the Federal Energy Management Program (FEMP) American Reinvestment and Recovery Act (ARRA). This research was aimed at developing a mathematical programming framework and accompanying optimization methodology in order to simultaneously evaluate energy efficiency (EE) and renewable energy (RE) opportunities. Once developed, this research then demonstrated this methodology at a USPACOM installation - Camp H.M. Smith, Hawaii. We believe this is the first time such an integrated, joint EE and RE optimization methodology has been constructed and demonstrated.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Network Optimization of the Electrosynthesis of Chemicals from CO2

Carbon dioxide electroreduction (ECO2R) is gaining attention due to its capacity to mitigate CO2 emissions while using electricity that would otherwise be curtailed. Its foreseeable industrial implementation requires of holistic methods to assess the technological and economic performance of ECO2R processes and integrate them in current chemical supply chains and power systems. Here, we combine techno-economic assessment and mathematical programming to find the optimal paths to electroreduce CO2 into valuable chemicals under variable electricity prices. The proposed approach is tested with a case study addressing the CO2 capture from flue gas or direct air and its electricity-powered reduction into carbon monoxide, formic acid or multi-carbon compounds. The results obtained demonstrate the ability of the framework to build ECO2R networks and provide operation profiles that respond to fluctuating electricity prices.

carbon dioxide↗

Aerodynamic Improvement of a Compact Centrifugal Compressor

For the example of the JetCat P-200RX turbojet engine, optimization of the diffuser and straightening unit of a compact centrifugal compressor is considered, on the basis of a mathematical model of viscous turbulent air flow. The goal of optimization is to increase the isentropic (adiabatic) efficiency of the compressor.

42 ENGINEERING↗

STUDY OF THE DESIGN REPRESENTATION METHODS FOR THE OPTIMIZATION OF MULTI-LAYER COMPOSITE STRUCTURES

The purpose of multi-layer composite structure optimization is to find the optimal composite layout, such that superior structure performances and lightweight can be achieved. However, the existing optimization methods have a low efficiency when applied to the multi-component, multi-layer composite structure. Such low efficiency is caused by the high dimensionality and the inherent shortcomings of the existing design representation methods. In this work, two composite layout representation methods are compared to better understand their impacts on optimization. The root cause of the low efficiency is identified, and its influences are also quantified using three metrics. Furthermore, a new Representation-Switch Optimization (RSO) strategy is proposed. This strategy improves the search efficiency by switching the design representation methods during the optimization process. Three benchmark studies, two mathematical examples and one real engineering example, are conducted to demonstrate the impact of design representation methods on the optimization results, as well as the effectiveness of the proposed optimization strategy.

Xu, Hongyi↗

Outcomes and Insights From Simplified Analytic Trajectory Optimization for a Tethered Underwater Kite

This letter formulates and solves a periodic trajectory optimization problem for a tethered underwater kite. The goal is to maximize the average mechanical power harvested by the kite. The type of kite considered in this letter extracts electricity from ocean currents by moving cross-current as its reel away from its base station, and consumes electricity to reel back. The problem of optimizing this kite’s trajectory is challenging due to the high dimensionality and nonlinearity of its dynamics. To tackle this challenge, the literature often separates the problem into two subproblems focusing on optimizing the cross-current and the reel-in/reel-out components of the trajectory, respectively, which may be sub-optimal. In contrast, this letter solves for the combined cross-current and reel-in/reel-out trajectory by linearizing the dynamics of the kite around a zero-power reference equilibrium trajectory in spherical coordinates. This allows the trajectory optimization problem to be solved analytically for simple sinusoidal input perturbations from equilibrium. Here we use linear quadratic regulation to enable the nonlinear kite model to track the optimized trajectory. The result is a computationally efficient approach that achieves an attractive Loyd factor of 19.9%, while providing important insights into the nature of the optimal trajectory.

42 ENGINEERING↗

Identification of parameters of discrete-continuous models

In the paper, the parameters of a discrete-continuous model have been identified on the basis of experimental investigations and formulation of optimization problem. The discrete-continuous model represents a cantilever stepped Timoshenko beam. The mathematical model has been formulated and solved according to the Lagrange multiplier formalism. Optimization has been based on the genetic algorithm. The presented proceeding’s stages make the identification of any parameters of discrete-continuous systems possible.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗