Quantum computing has emerged as a new field that may have the potential to revolutionize the landscape of information processing and computational power, although physically constructing quantum hardware has proven difficult,and quantum computers in the current Noisy Intermediate Scale Quantum (NISQ) era are error prone and limited in the number of qubits they contain.A sub-field within quantum algorithms research which holds potential for the NISQ era, and which has seen increasing activity in recent years, is quantum machine learning, where researchers apply approaches from classical machine learning to quantum computing algorithms and explore the interplay between the two. This master thesis investigates feature selection and autoencoding algorithms for quantum computers. Our review of the prior art led us to focus on contributing to three sub-problems: A) Embedded feature selection on quantum annealers, B) short depth quantum autoencoder circuits, and C)embedded compressed feature representation for quantum classifier circuits.For problem A, we demonstrate a working example by converting ridge regression to the Quadratic Unconstrained Binary Optimization (QUBO) problem formalism native to quantum annealers, and solving it on a simulated backend. For problem B we develop a novel quantum convolutional autoencoder architecture and successfully run simulation experiments to study its performance.For problem C, we choose a classifier quantum circuit ansatz based on theoretical considerations from the prior art, and experimentally study it in parallel with a classical benchmark method for the same classification task,then show a method from embedding compressed feature representation onto that quantum circuit. / Kvantberäkning är ett framväxande område som potentiellt kan revolutionera informationsbehandling och beräkningskraft. Dock är praktisk konstruktion av kvantdatorer svårt, och nuvarande kvantdatorer i den s.k. NISQ-eran lider av fel och begränsningar i antal kvantbitar de kan hantera. Ett lovande delområde inom kvantalgoritmer är kvantmaskininlärning, där forskare tillämpar klassiska maskininlärningsmetoder på kvantalgoritmer och utforskar samspelet mellande två områdena.. Denna avhandling fokuserar på kvantalgoritmer för funktionsval,och datakompression (i form av s.k. “autoencoders”). Vi undersöker tre delproblem: A) Inbäddat funktionsval på en kvantannealer, B) autoencoder-kvantkretsar för datakompression, och C) inbyggt funktionsval för kvantkretsar för klassificering. För problem A demonstrerar vi ett fungerande exempel genom att omvandla ridge regression till problemformuleringen "Quadratic Unconstrained Binary Optimization"(QUBO) som är nativ för kvantannealers,och löser det på en simulerad backend. För problem B utvecklar vi en ny konvolutionerande autoencoder-kvantkrets-arkitektur och utför simuleringsexperimentför att studera dess prestanda. För problem C väljer vi en kvantkrets-ansats för klassificering baserad på teoretiska överväganden från tidigare forskning och studerar den experimentellt parallellt med en klassisk benchmark-metod församma klassificeringsuppgift, samt visar en metod för inbyggt funktionsval (i form av datakompression) i denna kvantkrets.

Medium Mn (med-Mn) third generation advanced high strength steels (3G AHSSs) are promising candidates for meeting automotive weight reduction requirements without compromising passenger safety. However, the thermal processing of these steels should be compatible with continuous galvanizing line (CGL) processing capabilities as it provides cost-effective, robust corrosion protection for autobody parts. Hence, the main objective of this Ph.D. research is to develop a CGL-compatible thermal processing route for a prototype 0.2C-6Mn-1.5Si-0.5Al-0.5Cr-xSn (wt%) (x = 0 and 0.05 wt%) med-Mn steel that will result in the 3G AHSS target mechanical properties (24,000 MPa%  UTS × TE  40,000 MPa%) and high-quality galvanized coatings via enhanced reactive wetting. It was found that the starting microstructure, intercritical annealing (IA) time/temperature, and Sn micro-alloying had a significant effect on the retained austenite volume fraction and stability and, thereby, the mechanical properties of the prototype med-Mn steel. For the as-received cold-rolled (CR) starting microstructure, the intercritical austenite nucleated and grew on dissolving carbide particles and resulted in blocky retained austenite. However, Sn micro-alloying significantly effected the intercritical austenite chemical stability by segregating to the carbide/matrix interface and retarding C partitioning to the intercritical austenite. This resulted in lower volume fractions of low stability retained austenite which transformed to martensite (via the TRIP effect) at low strains, thereby quickly exhausting the TRIP effect and resulting in a failure to sustain high work hardening rates and delay the onset of necking. Consequently, the Sn micro-alloyed CR starting microstructure was unsuccessful in achieving 3G AHSS target mechanical properties regardless of the IA parameters employed. Contrastingly, the CR starting microstructure without Sn micro-alloying was able to meet target 3G mechanical properties via intercritical annealing at 675 °C × 60 s and 120 s, and at 690 °C × 60 s owing to sufficiently rapid carbide dissolution and C/Mn partitioning into the intercritical austenite such that it had sufficient mechanical and chemical stability to sustain a gradual deformation-induced transformation to martensite and maintain high work hardening rates. On the other hand, the martensitic (M) starting microstructure produced higher volume fractions of chemically and mechanically stable lamellar retained austenite regardless of Sn micro-alloying. Intercritical annealing at 650 °C × 60 s and 675 °C × 60 s and 120 s produced 3G AHSS target mechanical properties. It was shown that the stable lamellar retained austenite transformed gradually during deformation. Furthermore, deformation-induced nano-twin formation in the retained austenite was observed, suggesting the TWIP effect being operational alongside the TRIP effect. As a result, a continuous supply of obstacles to dislocation motion was maintained during deformation, which aided in sustaining a high work hardening rate and resulted in a high strength/ductility balance, meeting 3G AHSS target properties. Based on these results, the martensitic starting microstructure without Sn micro-alloying and the M-675 °C × 120 s IA condition were chosen for the selective oxidation and reactive wetting studies. The selective oxidation study determined the effect of a N2-5H2-xH2O (vol%) process atmosphere pO2 (–30, –10, and +5 °C dew point (Tdp)) on the composition, morphology, and spatial distribution of the external and internal oxides formed during the austenitizing and subsequent intercritical annealing cycles. The objective of this study was to identify the process atmosphere for the promising M-675 °C × 120 s heat treatment that would result in a pre-immersion surface that could be successfully galvanized in a conventional galvanizing (GI) bath. The austenitizing heat treatment (775 °C × 600 s) used to produce the martensitic starting microstructure resulted in thick (~ 200 nm) external oxides comprising MnO, MnAl2O4, MnSiO3/Mn2SiO4, and MnCr2O4, regardless of the process atmosphere pO2. However, intermediate flash pickling was successful in dissolving the external oxides to a thickness of approximately 30 nm along with exposing metallic Fe in areas which contained relatively thin external oxides. Furthermore, extruded Fe nodules that were trapped under the external oxides were revealed during the flash pickling process. Overall, flash pickling resulted in a surface consisting of dispersed external oxide particles with exposed metallic substrate and extruded Fe nodules. This external surface remained unchanged during IA owing to the multi-micron (~ 2–8 µm) solute-depleted layer that formed during the austenitizing heat treatment. Subsequent galvanizing in a 0.2 wt% (dissolved) Al GI bath with an immersion time of 4 s at 460 °C was successful in achieving high-quality, adherent galvanized coatings through multiple reactive wetting mechanisms. The dispersed nodule-type external oxides along with exposed substrate and extruded Fe nodules on the pre-immersion surface facilitated direct wetting of the steel substrate and promoted the formation of a robust and continuous Fe2Al5Znx interfacial layer at the steel/coating interface. Additionally, oxide lift-off, oxide wetting, bath metal ingress, and aluminothermic reduction were operational during galvanizing. The galvanized med-Mn steels met 3G AHSS target mechanical properties. Overall, this Ph.D. research showed that it is possible to employ a CGL-compatible thermal processing route for med-Mn steels to successfully produce 3G AHSS target mechanical properties as well as robust galvanized coatings. / Thesis / Doctor of Philosophy (PhD) / One of the largest challenges associated with incorporating the next generation of advanced high strength steels into the automotive industry lies in processing these steels in existing industrial production lines. In that regard, a two-stage heat treatment with an intermediate flash pickling stage and process atmosphere compatible with existing industrial continuous galvanizing line technology was developed for a prototype medium-Mn steel. The heat-treated prototype steel met the target mechanical properties outlined for the next generation of advanced high strength steels. Furthermore, the heat treatment and process atmosphere utilised in this research produced a surface that facilitated the successful galvanizing of the prototype medium-Mn steel. This adherent and high-quality galvanized coating will provide robust corrosion protection if the candidate medium-Mn steel is used in future automotive structural applications.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.