Two-step Optimization Research Articles

A two-step treatment planning strategy incorporating knowledge-based planning for head-and-neck radiotherapy.

There has been much research interest in automated head-and-neck (HN) planning with the goal of reducing planning time and inter-planner variability while improving plan quality. However, clinical uses are still limited and institution-dependent due to the plan complexity. This work aims to investigate whether the use of a novel semi-automated two-step optimization method (TSP) can improve the quality and efficiency of planning while providing a simple framework that other institutions can follow. Forty patients (two and three prescription isodose levels) were retrospectively studied. Plans were generated by TSP which incorporates a knowledge-based planning solution. Comparisons were performed for plan conformity and selected dose-volume indices between clinical plan (CP) and TSP. Blind reviews were carried out by 15 clinicians to determine preference between the CP and TSP, as well as clinical suitability. For majority of patients studied, TSP had similar or slightly better conformity for the high-dose PTV, and better conformity for the low-dose PTV and 45Gy isodose lines compared to CP. The only statistically significant difference observed for the serial organs was a reduction of the spinal cord maximum dose with TSP. Except for left parotid gland (Dmean and V30 for both 2R× and 3R× groups) and oral cavity (Dmean for 3R× group), TSP had significant dose reductions for all parallel organs compared to CP. Blind reviewers either showed preference/no preference for 57.2%/21.7% (2R×) and 57.5%/27.8% (3R×) of TSP compared with CP. Excluding no preference votes, 60% of TSP were preferred. TSP was selected majority of the time when looking at the vote distribution for each patient individually. Our TSP allows plans to be created within 90-min time frame while offering improvements in plan quality and less inter-planner variability as compared to traditional planning techniques.

Statistics-Based Optimization of Cellulase and Xylanase Production by the Endophytic Fungus Talaromyces Funiculosus using Agricultural Waste Materials

Lignocellulosic biomass can be utilized as a low-cost, renewable, and sustainable feedstock for obtaining non-fossil energy sources with low CO2 emission. One of the most promising technologies for producing 2G biofuels is the saccharification of agricultural waste materials with the help of cellulolytic enzymes, followed by yeast fermentation of sugars into cellulosic ethanol. Cellulases are multi-component enzymes involved in the degradation of cellulose, which can synergistically degrade cellulose and includes three major categories: endoglucanase (EC 3.2.1.4), exoglucanase or cellobiohydrolase (EC 3.2.1.91), and β-glucosidase (EC 3.2.1.21). The core enzyme used for the degradation of the xylan skeleton of hemicellulose is endo-β-1,4-xylanase (EC 3.2.1.8). The high cost of enzymes synthesized by fungi is a bottleneck for the production of cellulosic ethanol. Optimization of the nutrient medium composition is an important factor in increasing the production of enzymes and the efficiency of lignocellulosic biomass hydrolysis. The aim of the current study was to optimize the production of cellulolytic and xylanolytic enzymes through cultivation of filamentous fungus Talaromyces funiculosus on low-cost nutrient media with non-pretreated agricultural waste materials. Methods. Filamentous fungus Talaromyces funiculosus was grown on potato-dextrose agar for 10—14 days at 26±2 °С. To obtain the culture filtrate, the fungus was cultivated under submerged conditions in an Erlenmeyer flask for 4 days. The nutrient medium composition was varied according to the factor experiment design. A two-step optimization of the nutrient medium composition was used. A screening experiment with the Plackett-Burman fractional factorial design and response surface methodology with the Box-Behnken design were used to optimize cellulase production. The enzymatic activity was determined by measuring the reduced sugar production after the enzymes hydrolysis with specific substrates: exoglucanase with filter paper, endoglucanase with carboxymethylcellulose, and xylanase with beech wood xylan, using the colorimetric DNS method with glucose or xylose as a standard. The activity of β-glucosidase was determined by the hydrolysis reaction of p-nitrophenyl-β-D-glucopyranoside, which results in the formation of p-nitrophenol, quantified at 410 nm. Results. As a result of experiments with using agricultural waste, including wheat straw, corn stalk, and corn cob as carbon sources of the culture medium, it was shown that T. funiculosus is able to grow and produce cellulase and xylanase on all non-pretreated substrates studied. The two-step sequential optimization of the nutrient medium composition for T. funiculosus cultivation according to the Plackett-Berman and Box-Behnken designs made it possible to increase the activity of cellulolytic and xylanolytic enzymes by 2.4—2.6 times. The optimized cultivation medium does not contain such expensive components as Avicel, peptone, and yeast extract and has the following composition, g/L: corn stalks — 50.0; urea — 0.86; NaNO3 — 1.0; KH2PO4 — 6.0; KCl — 0.25; MgSO4 — 0.25; FeSO4 — 0.01. Conclusions. The studied strain of T. funiculosus produces a lignocellulosic enzyme complex with a high level of β-glucosidase activity when cultivated on an optimized nutrient medium with untreated agricultural waste and is promising for the conversion of lignocellulosic biomass into fermentable sugars.

Synergistic band modulation and precipitates: Achieving high quality factor in SnTe

Breaking the thermoelectric figure of merit zT barrier of SnTe enables it to become a promising alternative to PbTe; however, the inferior and strongly coupled physicochemical properties of pristine SnTe severely restrict the efficient optimization. Herein, we doped trivalent Sb in SnTe and incorporated SnS particles to achieve high quality factor B through a two-step optimization strategy of tuning the valence band structure and intercalating heterostructural precipitates, and well predicted the potential prospects. The high solubility limit of Sb not only reduced the carrier concentration nH but also significantly optimized the valence band structure and improved the Seebeck coefficient, thereby enhancing the weight mobility μw in the all-temperature region. Furthermore, the additional SnS, which tends to exist as precipitates with different micrometer-scale sizes, enhanced low-medium-frequency phonon scattering in a wider frequency range except for point defects scattering, suppressing the lattice thermal conductivity to 0.55 W m−1 K−1. As a result of this synergistic effect, a high B-factor of ∼0.82 greater than triple pure SnTe was obtained in Sn0.91Sb0.09Te-10%SnS, with an enhanced zT of ∼1.15 at 850 K. More importantly, the high B-factor accurately predicted an excellent zT value of ∼1.65 at the optimal Fermi level, which highlights the great potential of Sn1-xSbxTe-y%SnS-based materials. This work provides an effective route for stepwise optimization of electrical and thermal performance from the B-factor perspective and has guiding significance for other thermoelectric materials.

Two-step optimization design of a periodic beam with acoustic black holes

This paper presents a two-step optimization method to design the acoustic black holes (ABH) profiles according to local property in the target frequency region. In the first step, the concept of 'local index' is used to identify the position of local vibration. Then, as the second step, the ABH parameters are multi-objective optimized. The numerical examples indicate that the two-step optimization provide a set of Pareto solutions for lightweight ABH design with board bandgaps. To reveal the mechanism of the attenuation band, the vibration modes of the structure are analyzed and the local property of the ABH is shown. Using dynamic and static characteristics as objective functions, the proposed method provides a way to merge physical meaning in the process of data-based optimization.

TWO-STEP OPTIMIZATION OF TUNED INERTIAL MASS DAMPERS FOR ELASTIC-PLASTIC MDOF STRUCTURES USING TIME-FREQUENCY DUAL DOMAIN PERFORMANCE EVALUATION

The purpose of this paper is to propose a two-step optimization method to design tuned inertial mass dampers for elastic-plastic MDOF structures using the time-frequency dual domain performance evaluation. Although the frequency-domain-based approach allows us to obtain the optimal design independent of the characteristics of disturbances, it cannot consider the elastic-plastic response of structures. On the other hand, the time-domain design is highly dependent on the characteristics of design earthquake ground motions. The proposed two-step method resolves these issues by use of dual-domain information. Finally, the response analysis for recorded ground motions demonstrates the effectiveness of the proposed design method.

Precise Position Estimation Using Smartphone Raw GNSS Data Based on Two-Step Optimization.

This paper presents a high-precision positioning method using raw global navigation satellite system (GNSS) observations from smartphones in the Google smartphone decimeter challenge (GSDC). Compared to commercial GNSS receivers, smartphone GNSS observations are noisy owing to antenna limitations, making it difficult to apply conventional high-precision positioning methods. In addition, it is important to exclude outliers in GSDC because GSDC includes data in environments where GNSS is shielded, such as tunnels and elevated structures. Therefore, this study proposes a smartphone positioning method based on a two-step optimization method, using factor graph optimization (FGO). Here, the velocity and position optimization process are separated and the velocity is first estimated from Doppler observations. Then, the outliers of the velocity estimated by FGO are excluded, while the missing velocity is interpolated. In the next position-optimization step, the velocity estimated in the previous step is adopted as a loose state-to-state constraint and the position is estimated using the time-differenced carrier phase (TDCP), which is more accurate than Doppler, but less available. The final horizontal positioning accuracy was 1.229 m, which was the first place, thus demonstrating the effectiveness of the proposed method.

A strategy to train machine learning material models for finite element simulations on data acquirable from physical experiments

A high quality finite element analysis of structural components requires an adequate, often complex, description of the material behaviour. Data-driven models based on artificial neural networks have been shown to possess the potential to substitute current material models, as they promise fast computation, high adaptability to new materials and uncoupling from closed analytical formulations of the model, among others. However, the training of these Machine Learning Material Models (MLMM) is currently limited to data acquired from simulations using classical material models, since this gives the stress-output as a response to a given strain input directly, required to train the MLMM using supervised learning. Although, from experiments only surficial strain field and global force are measurable, local stress values in an inhomogeneous field are not obtainable.We propose a methodology to uncouple MLMM from classical modelling, making them no longer surrogate models, with the promise to unleash their full potential of data-driven materials modelling. In a two-step optimization, starting with a pre-trained MLMM on a similar material and using only data practically acquirable from experiments, we re-train or re-calibrate the MLMM onto a new material. The first and major step is denoted pseudogradient-ES (evolutionary strategy), an external optimization of the neural network’s parameters, such as weights and biases, using mechanically-motivated loss-functions as the optimization objective. It uses a unique combination of evolutionary strategy and gradient based optimization, with a special formulation of a vector serving as an alternative to a gradient in the optimization process. The final step is required to reduce the amount of physical tests by taking advantage of the material symmetry, using our proposed rotation-backpropagation strategy, where new training data is constructed and used in a final training stage of the MLMM using backpropagation. The methodology is shown as a proof of concept on elastic virtual experiments, which however only uses data acquirable in real physical experiments. The final MLMM describing the new material is shown running in finite element analyses.

Photovoltaic single diode model parameter extraction by dI/dV-assisted deterministic method

In this work, a state-of-the-art deterministic method is proposed for photovoltaic single diode model parameter extraction from experimental current–voltage (I-V) curves. This new method takes advantage of the numerical derivative (dI/dV) information, such that a high-quality solution can be obtained by linear least squares technique. The solution can be further improved through a two-step nonlinear optimization procedure using the trust-region-reflective or Levenberg-Marquardt algorithm. Compared with previous methods, the proposed method has achieved the best accuracy with the lowest computation cost up till now in the benchmark test of two commonly-used case studies. When comparing the computation costs of different works, it is pointed out for the first time that not only the number of iteration steps but also the single-step computation complexity should be taken into account if different deterministic optimization algorithms are implemented. Furthermore, the robustness of the proposed method is verified through a large-scale dataset containing more than one million I-V curves, showing that the method can be used for realtime monitoring of photovoltaic modules. Finally, a MATLAB app is developed and freely provided to the public.

Verification of improved YOLOX model in detection of greenhouse crop organs: Considering tomato as example

Efficient and accurate monitoring crop growth is an important prerequisite for modern greenhouse crop growth control and management. However, it is still challenging to achieve efficient in coordinated detection of tomato organs. Therefore, in this study, a multiobjective detection model was established to detect the blossom, fruit, and fruit color of crop population under dense planting, leaf shading, fruit overlap, and interference from light change in greenhouses. Specifically, sample equalization, random transformation, and mosaic enhancement were applied to identify multiple targets and small-sample images. A new detection network, YOLOX, was used, which introduces a decoupled head structure and redesigns a detection head according to the anchor-free concept to effectively improve the recognition of small-target occlusion. An efficient tomato flower and fruit recognition model was constructed based on a two-step fine-tuned multi-round optimization strategy training. The results showed that the YOLOX target detection model improved the average recognition accuracy (mAP) by 52.49–10.75 %, was lighter, was only 14 % original YOLOv4, and improved the detection speed by 25.53 FPS compared to six main target detection algorithms. It can be used as a method for the collaborative recognition of tomato flowers and fruits. Accordingly, two sets of lightweight detection networks with different performances were designed according to different agricultural scenarios. YOLOXMOB modeled with a small size (only 34.2 MB), high mAP of 81.39 %, and detection speed of 61.85 FPS. It is recommended for use as a deployment model for mobile devices. YOLOXPC, with a deep network layer and strong performance, achieved 86.16 % mAP, and is recommended for monitoring and dispatching computers. This study puts forward a set of multiobjective detection model scheme adapted to the real environment of greenhouse, which provides ideas for the construction of a highly intelligent monitoring system platform for the agricultural Internet of Things.

Hybrid uncertainty analysis and optimisation based on probability box for bus powertrain mounting system

In engineering practice, the bus powertrain mounting system (BPMS) may have both epistemic and aleatory uncertainty under the influence of manufacturing, measurement, and assembly errors. The hybrid uncertainty in BPMS may result in over-design or insufficient design. Therefore, the probability box (p-box) model, which can handle both aleatory and epistemic variables, is introduced into the uncertainty analysis of BPMS. Considering the elastic connection between the compressor and powertrain, a 12-degree-of-freedom dynamic model is constructed to calculate the inherent characteristic of BPMS. A rejection sampling method based on the fast envelope function (RSMBFEF) is proposed to propagate the hybrid uncertainties. Then double-loop Monte Carlo method is used to be compared with RSMBFEF. To reduce the number of uncertainty analyses, a two-step uncertainty optimisation method is proposed. Finally, the proposed method's efficacy and accuracy are verified through a numerical case. The applicability of the p-box model is illustrated by comparing it with the BPMS model with only pure aleatory or pure interval variables.

Visible imaging system optical design by continuous optimization of glasses

Choice of lenses materials in optical design is crucial to reduce aberrations down to an acceptable level. Commercial glasses do not cover a continuous range of refractive indices and must be selected in a discrete library making them discrete variables in any optimization design process to achieve the final optical design to be manufactured. This paper proposes an alternative method to avoid the complicated discrete variables optimization process thanks to a two-steps continuous optimization methodology starting with fictitious glasses models before jumping to the real glasses optimization design. The illustration of this process and achieved results are presented on an example of optical system which validates our proposed method.

Vehicles for the Service Economy: Early-stage Vehicle Concept Designs for Vehicle-based Service

Future vehicles are connected, autonomous and shared platforms that render a multitude of services in our everyday lives. The vehicle's driving function will ultimately converge to a commodity sourced by superstructure manufactures that develop service-oriented functionality. For the early stage vehicle concept design, this new functionality imposes a new complexity for superstructure engineers. Novel service offerings translate into new functional requirements and technical components that need to be optimally sized and positioned inside the vehicle package to increase the amount of service units that can be supplied to the customer. Today, vehicle concept design optimization only focuses on powertrain components. In this paper, we introduce a novel two-step package design optimization for vehicle superstructures. First, functional requirements are derived from service units, the smallest unit of value to the vehicle-based service customer. We then develop component weight and size regression models from data retrieved from online warehouses. Components and their installation constrains are represented by scalable containers. A combined packing heuristic and parameter optimization arranges the containers within the vehicle's available installation space and created layout alternatives. We apply this procedure to a real case study based on the aCar, a light electric utility vehicle for rural sub-Saharan Africa. This is a first step to enable superstructure manufactures to design and optimize vehicle concept designs based on service capacity.

A convex optimisation approach to parallel magnetic resonance imaging reconstruction

In parallel magnetic resonance imaging (pMRI), the image reconstruction with unknown coil sensitivity functions is known as a non-convex problem in the existing literatures. The analysis of this paper shows that there exists a convex solution region in the space of the magnitude image and sensitivity encoded image functions, which contains the true magnitude image solution. The derivation of the convex solution region resolves the non-convex difficulty and leads to a convex optimisation formulation of the pMRI reconstruction problem. The formulated problem consists of two steps. Each of the steps solves a regularised convex optimisation problem and provides a globally optimal solution, in the sense that the solution optimises the performance index and is independent of the initial conditions. The applications of the proposed two-step optimisation to in vivo and phantom data sets result in superior pMRI reconstruction performance compared with state-of-the-art algorithms.

A synergistic two-step optimization design enables high capacitive energy storage in lead-free Sr0.7Bi0.2TiO3-based relaxor ferroelectric ceramics

A two-step design is developed to realize multi-objective synergistic optimizations including high activity/ultrafine PNRs and ultrasmall grain size with compact grain boundaries, showing huge application potential in advanced pulsed power devices.

A two-step optimization process for efficient scheduling of heterogeneous fleets of electric buses with synchronized transfers

Public transport systems are undergoing a significant change in most cities thanks to the advances in technology. The focus is not only on sustainability but also on efficiency and the experience of the passenger. In this work, we propose a complete redesign of the public transportation that considers a heterogeneous fleet of electric buses that dynamically changes in terms of the needs of each bus line at different hours, as well as a new bus schedule that minimizes passenger waiting time at bus stops by synchronizing different bus lines. In order to get optimal solutions, two optimization steps are needed. Preliminary results suggest that it is possible to implement an electrical bus fleet with different passenger and battery capacities that addresses the passenger demand, including the transfers on extensive trips where the users reduce travel times due to the synchronized schedule.

Multistatic Localization by Differential Time Delays and Time Differences of Arrival in the Absence of Transmitter Position

We can only extract the differential time delay (DTD) measurements between the direct and reflected paths in multistatic localization when there is no synchronization in time between the transmitter and receiver and among the receivers. This paper first addresses the problem of multistatic localization of a fixed object when the transmitter position is not available by using the DTD measurements. We propose a two-step optimization method for jointly estimating the object and transmitter positions. In the first step, we formulate a non-convex constrained weighted least squares (CWLS) problem by transforming the DTD measurement model and introducing nuisance variables. Such a non-convex CWLS problem is then relaxed to a tractable convex semidefinite programming (SDP) problem by applying semidefinite relaxation. In the second step, the error coming from relaxation and approximation in the SDP solution is reduced iteratively through solving a generalized trust region subproblem (GTRS) in each iteration. If the receivers are synchronized such that the time difference of arrival (TDOA) measurements can be acquired in addition to DTD, we formulate a different CWLS problem by utilizing both DTD and TDOA measurements, which is solved by convex relaxation as well. The relaxed SDP problem can achieve the optimal solution of the CWLS problem, and further refinement is no longer needed. We conduct the mean square error (MSE) analysis to validate that both proposed methods are able to achieve the Cramer-Rao Lower Bound (CRLB) performance under small Gaussian noise, which is also validated by simulations.

Controlling Bias Between Categorical Attributes in Datasets: A Two-Step Optimization Algorithm Leveraging Structural Equation Modeling

Controlling Bias Between Categorical Attributes in Datasets: A Two-Step Optimization Algorithm Leveraging Structural Equation Modeling

A New Method to Compute Minimal Detectable and Isolable Faults of Active Fault Diagnosis*

This paper proposes a new method for the computation of guaranteed minimal detectable and isolable faults of set-based active fault diagnosis for multiplicative actuator faults. Compared with our previous work, the proposed method in this paper has the important advantage of low computational complexity. In particular, by handling the coupling of faults and inputs and using an objective function describing the minimal detectable and isolable faults, a new two-step optimization method is proposed to analyze and compute the minimal detectable and isolable faults. At the end of this paper, two examples are used to show the effectiveness of the proposed method and the advantage of low computational complexity.

Design of screen-printed potentiometric platform for sensitive determination of mirabegron in spiked human plasma; molecular docking and transducer optimization.

The integration of molecular modelling simulation and electrochemical sensors is of high interest. Herein, for the first time, a portable solid-contact potentiometric electrode was designed for the sensitive determination of mirabegron (MIR) in human plasma and pharmaceutical formulation. A two-step optimization protocol was investigated for the fabrication of an ion on sensing polymeric membrane. First, molecular docking was used for optimum ionophore selection. Calix[6]arene showed the highest affinity towards MIR with a better docking score (-4.35) and potential energy (-65.23) compared to other calixarene derivatives. Second, carbon nanotubes and gold nanoparticles were investigated as ion-electron transducers using a drop-casting procedure. Gold nanoparticle-based sensors showed better slope, potential stability, and rapid response compared to carbon nanotubes. The proposed solid contact sensors (V-VII) showed comparable sensitivity and ease of handling compared to liquid contact sensors (I-IV). The optimized gold nanoparticles sensor VII produced a Nernstian response over the range of 9.77 × 10-7 to 1 × 10-3 M with LOD of 2.4 × 10-7 M. It has also been used to determine MIR in its pharmaceutical formulation in the presence of a co-formulated antioxidant butylated hydroxytoluene and spiked human plasma. This would offer a feasible and economic platform for monitoring MIR in pharmaceutical preparation and biological fluids.

Joint sensor registration and multi-object tracking with PHD filter in distributed multi-sensor networks

This paper addresses the joint sensor registration and multi-object tracking in distributed sensor networks wherein the accurate positions of sensors are unknown. Each local sensor runs the probability hypothesis density filter (PHD) in its local coordinate system, having no knowledge about the relative positions of other sensors in the network. The generalized covariance intersection (GCI) fusion rule is employed to fuse the local densities. To tackle the sensor registration problem, a cost function expressing the information discrepancy between local densities is constructed based on the GCI divergence, then the relative positions of sensors are estimated by minimizing such a cost function. As the constructed cost function leads to a high-dimension optimization problem, for computational efficiency, a pairwise sensor registration strategy is proposed which is proved equally sound with the high-dimension registration method but more computationally efficient. Lastly, we demonstrate that the formulated registration method involves a non-convex problem. To solve such a non-convex problem, a two-step optimization method is proposed which can transform the non-convex problem into a convex one. The efficacy of the proposed method is demonstrated in a challenging tracking scenario via numerical experiments.