A framework for design optimization across multiple concepts.

In engineering design, there often exist multiple conceptual solutions to a given problem. Concept design and selection is the first phase of the design process that is estimated to affect up to 70% of the life cycle cost of a product. Currently, optimization methods are rarely used in this phase, since standard optimization methods inherently assume a fixed (given) concept; and undertaking a full-fledged optimization for each possible concept is untenable. In this paper, we aim to address this gap by developing a framework that searches for optimum solutions efficiently across multiple concepts, where each concept may be defined using a different number, or type, of variables (continuous, binary, discrete, categorical etc.). The proposed approach makes progressive data-driven decisions regarding which concept(s) and corresponding solution(s) should be evaluated over the course of search, so as to minimize the computational budget spent on less promising concepts, as well as ensuring that the search does not prematurely converge to a non-optimal concept. This is achieved through the use of a tree-structured Parzen estimator (TPE) based sampler in addition to Gaussian process (GP), and random forest (RF) regressors. Aside from extending the use of GP and RF to search across multiple concepts, this study highlights the previously unexplored benefits of TPE for design optimization. The performance of the approach is demonstrated using diverse case studies, including design of a cantilever beam, coronary stents, and lattice structures using a limited computational budget. We believe this contribution fills an important gap and capitalizes on the developments in the machine learning domain to support designers involved in concept-based design.

A linearly decreasing deterministic annealing algorithm for the multi-vehicle dial-a-ride problem.

Dial a ride problem (DARP) is a complex version of the pick-up and delivery problem with many practical applications in the field of transportation. This study proposes an enhanced deterministic annealing algorithm for the solution of large-scale multi-vehicle DARPs. The proposed method always explores the feasible search space; therefore, a feasible solution is guaranteed at any point of termination. This method utilises advanced local search operators to accelerate the search for optimal solutions and it relies on a linearly decreasing deterministic annealing schedule to limit poor jumps during the course of search. This study puts forward a systematic series of experiments to compare the performance of solution methods from various angles. The proposed method is compared with the most efficient methods reported in the literature i.e., the Adaptive Large Neighbourhood Search (ALNS), Evolutionary Local Search (ELS), and Deterministic Annealing (DA) using standard benchmarks. The results suggest that the proposed algorithm is on average faster than the state-of-the-art algorithms in reaching competitive objective values across the range of benchmarks.

EV Hosting Capacity Enhancement in a Community Microgrid Through Dynamic Price Optimization-Based Demand Response.

Community microgrids, as an emerging technology, offer resiliency in operation for smart grids. Microgrids are seeing an increased penetration of eco-friendly electric vehicles (EVs) in recent years. However, the uncontrolled charging of EVs can easily overwhelm such electric networks. In this work, we propose an efficient demand response (DR) scheme based on dynamic pricing to enhance the capacity of the microgrid to securely host a large number of EVs. A hierarchical two-level optimization framework is introduced to realize the DR scheme. At the upper level, the dynamic prices for the participating users in DR are optimized while at the lower level, each user optimizes its energy consumption based on the price signal from the upper level. An evolutionary algorithm and a mixed-integer linear programming model is employed to solve the upper and lower level problems, respectively. Energy scheduling problems of the users are solved in a distributed manner which adds to the scalability of the approach. The proposed DR scheme is tested on a microgrid system adopted from the IEEE European low-voltage distribution network. Numerical experiments confirm the effectiveness of the proposed DR scheme compared to the benchmark pricing policies from the literature.

A multi-objective optimization of stent geometries.

Stents are scaffolding cardiovascular implants used to restore blood flow in narrowed arteries. However, the presence of the stent alters local blood flow and shear stresses on the surrounding arterial wall, which can cause adverse tissue responses and increase the risk of adverse outcomes. There is a need for optimization of stent designs for hemodynamic performance. We used multi-objective optimization to identify ideal combinations of design variables by assessing potential trade-offs based on common hemodynamic indices associated with clinical risk and mechanical performance of the stents. We studied seven design variables including strut cross-section, strut dimension, strut angle, cell alignment, cell height, connector type and connector arrangement. Optimization objectives were the percentage of vessel area exposed to adversely low time averaged WSS (TAWSS) and adversely high Wall Shear Stress (WSS) assessed using computational fluid dynamics modeling, as well as radial stiffness of the stent using FEA simulation. Two multi-objective optimization algorithms were used and compared to iteratively predict ideal designs. Out of 50 designs, three best designs with respect to each of the three objectives, and two designs in regard to overall performance were identified.

An Enhanced Decomposition-Based Evolutionary Algorithm With Adaptive Reference Vectors.

Multiobjective optimization problems with more than three objectives are commonly referred to as many-objective optimization problems (MaOPs). Development of algorithms to solve MaOPs has garnered significant research attention in recent years. "Decomposition" is a commonly adopted approach toward this aim, wherein the problem is divided into a set of simpler subproblems guided by a set of reference vectors. The reference vectors are often predefined and distributed uniformly in the objective space. Use of such uniform distribution of reference vectors has shown commendable performance on problems with "regular" Pareto optimal front (POF), i.e., those that are nondegenerate, smooth, continuous, and easily mapped by a unit simplex of reference vectors. However, the performance deteriorates for problems with "irregular" POF (i.e., which deviate from above properties), since a number of reference vectors may not have a solution on the POF along them. While adaptive approaches have been suggested in the literature that attempt to delete/insert reference directions conforming to the geometry of the evolving front, their performance may in turn be compromised for problems with regular POFs. This paper presents a generalized version of previously proposed decomposition-based evolutionary algorithm with adaptive reference vectors, intended toward achieving competitive performance for both types of problems. The proposed approach starts off with a set of uniform reference vectors and collects information about feasibility and nondominance of solutions that associate with the reference vectors over a learning period. Subsequently, new reference directions are inserted/deleted, while the original directions may assume an active or inactive role during the course of evolution. Numerical experiments are conducted over a wide range of problems with regular and irregular POFs with up to 15 objectives to demonstrate the competence of the proposed approach with the state-of-the-art methods.

An Enhanced Memetic Algorithm for Single-Objective Bilevel Optimization Problems.

Bilevel optimization, as the name reflects, deals with optimization at two interconnected hierarchical levels. The aim is to identify the optimum of an upper-level  leader problem, subject to the optimality of a lower-level follower problem. Several problems from the domain of engineering, logistics, economics, and transportation have an inherent nested structure which requires them to be modeled as bilevel optimization problems. Increasing size and complexity of such problems has prompted active theoretical and practical interest in the design of efficient algorithms for bilevel optimization. Given the nested nature of bilevel problems, the computational effort (number of function evaluations) required to solve them is often quite high. In this article, we explore the use of a Memetic Algorithm (MA) to solve bilevel optimization problems. While MAs have been quite successful in solving single-level optimization problems, there have been relatively few studies exploring their potential for solving bilevel optimization problems. MAs essentially attempt to combine advantages of global and local search strategies to identify optimum solutions with low computational cost (function evaluations). The approach introduced in this article is a nested Bilevel Memetic Algorithm (BLMA). At both upper and lower levels, either a global or a local search method is used during different phases of the search. The performance of BLMA is presented on twenty-five standard test problems and two real-life applications. The results are compared with other established algorithms to demonstrate the efficacy of the proposed approach.

Optimum oil production planning using infeasibility driven evolutionary algorithm.

In this paper, we discuss a practical oil production planning optimization problem. For oil wells with insufficient reservoir pressure, gas is usually injected to artificially lift oil, a practice commonly referred to as enhanced oil recovery (EOR). The total gas that can be used for oil extraction is constrained by daily availability limits. The oil extracted from each well is known to be a nonlinear function of the gas injected into the well and varies between wells. The problem is to identify the optimal amount of gas that needs to be injected into each well to maximize the amount of oil extracted subject to the constraint on the total daily gas availability. The problem has long been of practical interest to all major oil exploration companies as it has the potential to derive large financial benefit. In this paper, an infeasibility driven evolutionary algorithm is used to solve a 56 well reservoir problem which demonstrates its efficiency in solving constrained optimization problems. Furthermore, a multi-objective formulation of the problem is posed and solved using a number of algorithms, which eliminates the need for solving the (single objective) problem on a regular basis. Lastly, a modified single objective formulation of the problem is also proposed, which aims to maximize the profit instead of the quantity of oil. It is shown that even with a lesser amount of oil extracted, more economic benefits can be achieved through the modified formulation.