Two Stage Mini-Max Algorithm for Grid-Based Wind Farm Layout Optimization

The power maximizing grid-based wind farm layout optimization problem seeks to determine the layout of a given number of turbines from a grid of possible locations such that wind farm power output is maximized. The problem in general is a nonlinear discrete optimization problem which cannot be solved to optimality, so heuristics must be used. This article proposes a new two stage heuristic that first finds a layout that minimizes the maximum pairwise power loss between any pair of turbines. The initial layout is then changed one turbine at a time to decrease sum of pairwise power losses. The proposed heuristic is compared to the greedy algorithm using real world data collected from a site in Iowa. The results suggest that the proposed heuristic produces layouts with slightly higher power output, but are less robust to changes in the dominant wind direction.

Design of a Reconfigurable Dynamic Testbed for Co-Design Method Validation

Physical testing as a technique for validation of engineering design methods can be a valuable source of insights not available through simulation alone. Physical testing also helps to ensure that design methods are suitable for design problems with a practical level of detail, and can reveal issues related to interactions not captured by physics-based computer models. Construction of physical and testing of physical prototypes, however, is costly and time consuming so it is not often used when investigating new design methods for complex systems. This gap is addressed through an innovative testbed presented here that can be reconfigured to achieve a range of different prototype design properties, including kinematic behavior and different control system architectures. Thus, a single testbed can be used for validation of numerous design geometries and control system architectures. The testbed presented here is a mechanically and electronically reconfigurable quarter-car suspension testbed with nonlinear elements that is capable of testing a wide range of both optimal and sub-optimal design prototypes using a single piece of equipment. Kinematic suspension properties can be changed in an automated way to reflect different suspension linkage designs, spring and damper properties can be adjusted in real time, and control system design can be changed easily through streamlined software modifications. While the specific case study is focused on development of a reconfigurable system for validation of co-design methods, the concept extends to physical validation using reconfigurable systems for other classes of design methods.

CAD-Integrated Cost Estimation and Build Orientation Optimization to Support Design for Metal Additive Manufacturing

Cost estimation techniques for Additive Manufacturing (AM) have limited synchronization with the metadata of 3D CAD models. This paper proposes a method for estimating AM build costs through a commercial 3D solid modeling program. Using an application programming interface (API), part volume and surface data is queried from the CAD model and used to generate internal and external support structures as solid-body features. The queried data along with manipulation of the part’s build orientation allows users to estimate build time, feedstock requirements, and optimize parts for AM production while they are being designed in a CAD program. A case study is presented with a macro programmed using the SolidWorks API with costing for a metal 3D-printed automotive component. Results reveal that an imprecise support angle can under-predict support volume by 34% and build time by 20%. Orientation and insufficient build volume packing can increase powder depreciation costs by nearly twice the material costs.

Design of an Agent-Based Technique for Controlling Interconnected Distributed Energy Resource Transactions

Electricity has traditionally been a commodity that is bought and sold through a rigid marketplace between an electric utility and a ratepayer. Today, however, the electricity market is rapidly evolving to be comprised of distributed energy resources and microgrids that change the structure of the technical and financial relationship between utilities and ratepayers. Regulation, a reduction in cost of renewable energy technologies, interoperability and improved communications, and public interest in green power are facilitating this transition. Microgrids require an additional layer of control, often use preprogrammed rule sets, and lack bi-directional self-awareness, self-management, and self-diagnostics necessary to dynamically adapt to changes on-site and in the grid. Research is needed in optimization and controls. This study explores the viability of self-organizing control algorithms to manage multiple distributed energy resources within a distribution network and reduce electricity cost to one or more ratepayers having such resources installed on-site. Such research provides insight into the transition from a traditional power distribution architecture into a flexible smart network that is better prepared for future technological advances, renewables integration, and customer-side control. Agent-based techniques are employed for least-cost optimization and implements these to manage transactions between three decentralized distributed energy resource systems within an electrical network.

Ensemble Prognostics With Degradation-Dependent Weights: Prediction of Remaining Useful Life for Aircraft Engines

The objective of this research is to introduce a new ensemble prognostics method with degradation-dependent weights. Specifically, this method assigns an optimized, degradation-dependent weight to each learner (i.e., learning algorithm) such that the weighted sum of the prediction results from all the learners predicts the RUL of mechanical components with better accuracy. The ensemble prognostic algorithm is demonstrated using a data set collected from an engine simulator. Analysis results show that the predictive model trained by the ensemble learning algorithm outperform the existing methods.

Linking Creativity Measurements to Product Market Favorability: A Data-Mining Approach

While creative ideas can lead to market success and payoff, they are also associated with high risks and uncertainties. One way to reduce these uncertainties is to provide decision makers with valuable information about the innovative potential and future success of an idea. Even though several metrics have been proposed in the literature to evaluate the creativity of early design-stage ideas, these metrics do not provide information about the future product success or market favorability of new product ideas. Hence, existing metrics fail to link the creativity of early-stage ideas to their future market favorability. In order to bridge this gap, the current work proposes a new metric to estimate early design-stage ideas’ favorability and analyzes its relationship with current creativity metrics. A data-mining driven method to assess the future favorability of new product ideas using customers’ reviews of current market products that shared similar features with the new ideas of interest is presented. The results suggest that the new product idea favorability is positively correlated with relative creativity metrics and existing product market favorability ratings. This method can be used to help designers gain a better insight into the creativity and market favorability potential of new product ideas in early design-stages via a systematic approach; hence, helping reduce the risks and uncertainties associated with early-phase ideas during the screening and selecting process.

Employing Multidimensional Data Visualization Tools to Assess the Impact of Constraint Uncertainties on Complex Design Problems

A potential source of uncertainty within multi-objective design problems can be the exact value of the underlying design constraints. This uncertainty will affect the resulting performance of the selected system commensurate with the level of risk that decision-makers are willing to accept. This research focuses on developing visualization tools that allow decision-makers to specify uncertainty distributions on design constraints and to visualize their effects in the performance space using multidimensional data visualization methods to solve problems with high orders of computational complexity. These visual tools will be demonstrated using an example portfolio design scenario in which the goal of the design problem is to maximize the performance of a portfolio with an uncertain budget constraint.

A Guided Particle Swarm Optimizer for Distributed Operation of Electric Vehicle to Building Integration

Relevant research has demonstrated that more potential benefits can be achieved when energy and information are transacted and exchanged locally among different energy consumers. With increasing number of electric vehicles (EVs), various models and solution strategies have been developed for collaboration between building and EV charging station to achieve greater energy efficiency. However, most of the existing research employs centralized decision model which is time consuming for large scale problems and cannot protect private information for each participator. To bridge these research gaps, a guided particle swarm optimizer based distributed decision approach is proposed to study the energy transaction between building and EV charging station. In the proposed decision approach, the marginal price signal of transactive energy is collected to guide iterative direction of particle’s velocity and position which can maximally protect private information of building and EV charging station. A study case based on a commercial building and a nearby charging station in Chicago area is designed for illustration. The experimental results demonstrate that our proposed marginal price guided particle swarm optimizer is more stable and efficient comparing with canonical particle swarm optimizer and two state-of-the-art distributed decision algorithms.

Analogy Retrieval Through Textual Inference

Analogy-making has been deemed one of the core cognitive mechanisms which play a role in human creative thinking activities such as design and art. Designers can make use of analogies in various stages of design including ideation, planning and evaluation. However, human analogy-making is limited by experience and reliance of human memory on superficial attributes rather than relational or causal structure during analogy retrieval. In this regard, different design-by-analogy tools have been developed to assist designers in analogical reasoning. Analogical reasoning tools can be viewed as either based on hand-coded structured knowledge or natural-language-based design-by-analogy tools. The former are naturally limited in extent and scope to that which was hand coded [1]. Alternatively, natural language analogical reasoning can leverage the abundantly available textual resources. Current text-based analogy research for design have relied on analogies between individual word meanings. This leaves open consideration of the relational structure of the language where the relational similarity of texts can indicate a significant analogy. In this article, we develop four computational models of analogy that capture relational structure of the text. This includes spatial representation of semantics, multi-level deep neural reasoning, graph matching based model and transformation-based model. The models are then combined together into an ensemble model to achieve acceptable level of analogical accuracy for the end-user. The underlying design-related knowledge upon which analogies were drawn includes engineering ontologies, function hierarchy and raw patent texts. Instantiating this analogical reasoning model in design concept analogy retrieval system, we show this approach can help retrieve meaningful analogies from the World Intellectual Property Organization (WIPO) patent repository. We demonstrate this for a particular design problem.

Hybrid Procedure-Based Design Strategies Augmented With Optimization

Procedure-based design is well-established, supporting engineers via expert knowledge codified in resources such as handbooks, tables, and heuristic if-then rules of thumb. These procedures enable even inexperienced designers to benefit from the knowledge obtained by more experienced counterparts through years of practice and discovery. While procedural approaches have many advantages, they do have limitations. They tend to produce only satisficing, rather than optimal, solutions. In addition, they are based on historical designs, so offer little assistance for new system types, and are often descriptive rather than normative in nature. In contrast, normative methods — such as constrained optimization — can resolve many of these issues, but at the cost of significant development effort. Here we present a synergistic hybrid strategy with the objective of capitalizing on established procedure-based design methods for a subset of design problem elements, while incorporating normative strategies for the remaining elements. A design procedure is analyzed to identify steps that involve specification of design variables, and a subset of rule-based steps that could be replaced with optimization algorithms. A single-stage spur gear train design example is used to illustrate this process, and for comparing alternative hybrid solution strategies. Initial results indicate that solution quality can be improved significantly over purely procedure-based design when incorporating limited optimization elements, while maintaining a reasonable level of additional modeling effort.