Search results

The purpose of this paper is to demonstrate the applicability of swarm and evolutionary techniques for regularized machine learning. Generally, by defining a proper penalty function, regularization laws are embedded into the structure of common least square solutions to increase the numerical stability, sparsity, accuracy and robustness of regression weights. Several regularization techniques have been proposed so far which have their own advantages and disadvantages. Several efforts have been made to find fast and accurate deterministic solvers to handle those regularization techniques. However, the proposed numerical and deterministic approaches need certain knowledge of mathematical programming, and also do not guarantee the global optimality of the obtained solution. In this research, the authors propose the use of constraint swarm and evolutionary techniques to cope with demanding requirements of regularized extreme learning machine (ELM).

To implement the required tools for comparative numerical study, three steps are taken. The considered algorithms contain both classical and swarm and evolutionary approaches. For the classical regularization techniques, Lasso regularization, Tikhonov regularization, cascade Lasso-Tikhonov regularization, and elastic net are considered. For swarm and evolutionary-based regularization, an efficient constraint handling technique known as self-adaptive penalty function constraint handling is considered, and its algorithmic structure is modified so that it can efficiently perform the regularized learning. Several well-known metaheuristics are considered to check the generalization capability of the proposed scheme. To test the efficacy of the proposed constraint evolutionary-based regularization technique, a wide range of regression problems are used. Besides, the proposed framework is applied to a real-life identification problem, i.e. identifying the dominant factors affecting the hydrocarbon emissions of an automotive engine, for further assurance on the performance of the proposed scheme.

Through extensive numerical study, it is observed that the proposed scheme can be easily used for regularized machine learning. It is indicated that by defining a proper objective function and considering an appropriate penalty function, near global optimum values of regressors can be easily obtained. The results attest the high potentials of swarm and evolutionary techniques for fast, accurate and robust regularized machine learning.

The originality of the research paper lies behind the use of a novel constraint metaheuristic computing scheme which can be used for effective regularized optimally pruned extreme learning machine (OP-ELM). The self-adaption of the proposed method alleviates the user from the knowledge of the underlying system, and also increases the degree of the automation of OP-ELM. Besides, by using different types of metaheuristics, it is demonstrated that the proposed methodology is a general flexible scheme, and can be combined with different types of swarm and evolutionary-based optimization techniques to form a regularized machine learning approach.

Additive manufacturing (AM) can reduce the process supply chain and encourage manufacturing innovation in remote or austere environments by producing an array of replacement/spare parts from a single raw material source. The wide variety of AM technologies, materials, and potential use cases necessitates decision support that addresses the diverse considerations of deployable manufacturing. The paper aims to discuss these issues.

Semi-structured interviews with potential users are conducted in order to establish a general deployable AM framework. This framework then forms the basis for a decision support tool to help users determine appropriate machines and materials for their desired deployable context.

User constraints are separated into process, machine, part, material, environmental, and logistical categories to form a deployable AM framework. These inform a “tiered funnel” selection tool, where each stage requires increased user knowledge of AM and the deployable context. The tool can help users narrow a database of candidate machines and materials to those appropriate for their deployable context.

Future work will focus on expanding the environments covered by the decision support tool and expanding the user needs pool to incorporate private sector users and users less familiar with AM processes.

The framework in this paper can influence the growth of existing deployable manufacturing endeavors (e.g. Rapid Equipping Force Expeditionary Lab – Mobile, Army’s Mobile Parts Hospital, etc.) and considerations for future deployable AM systems.

This work represents novel research to develop both a framework for deployable AM and a user-driven decision support tool to select a process and material for the deployable context.

The purpose of this paper is to realize the integrated optimization of process planning and scheduling in printed circuit board assembly (PCBA).

Logical and numerical contour matrix is used to describe the constituent of component and machine for different PCBA processes on the basis of polychromatic sets (PS) theory, and a PS model of PCBA is built. A hybrid genetic algorithm (GA) is developed to optimize the component allocation, PCB assignment and assembly sequence simultaneously.

Integration of PCBA process planning and scheduling (PCBAPPS) can bridge the gap between design and manufacturing to guarantee the assembly quality and improve the production efficiency. However, PCBAPPS have to search for the optimal result in their own vast solution space. They are complex combinatorial optimization problems. The optimization of PCBAPPS constructs a unified solution space which includes two sub‐solution space stated above. In this paper, dynamic optimization of PCBAPPS is implemented and the solution efficiency is improved.

PS model holds unified standard form on the basis of logical contour and numerical matrix. It is adopted to describe the static structure and dynamic characteristic of PCBA system and combine with GA to solve the integrated optimization problem of PCBAPPS effectively and dynamically.

The main contribution of this manuscript is to suggest new approaches in order to deal with dynamic lot-sizing and maintenance problem under aspect energetic and risk analysis. The authors introduce a new maintenance strategy based on the centroid approach to determine a common preventive maintenance plan for all machines to minimize the total maintenance cost. Thereafter, the authors suggest a risk analysis study further to unforeseen disruption of availability machines with the aim of helping the production stakeholders to achieve the obtained forecasting lot-size plan.

The authors tackle the dynamic lot-sizing problem using an efficient hybrid approach based on random exploration and branch and bound method to generate possible solutions. Indeed, the feasible solutions of random exploration method are used as input for branch and bound to determine the near-optimal solution of lot-size plan. In addition, our contribution to the maintenance part is to determine the optimal common maintenance plan for M machines based on a new algorithm called preventive maintenance (PM) periods means.

First, the authors have funded the optimal lot-size plan that should satisfy the random demand under service level requirement and energy constraint while minimizing the costs of production and inventory. Indeed, establishing a best lot-size plan is to determine the appropriate number of available machines and manufactured units per period. Second, for risk analysis study, the solution of subcontracting is proposed by specifying a maximum cost of subcontractor in the context of a calling of tenders.

For maintenance problem, the originality consists in regrouping the maintenance plans of M machines into only one plan. This approach lets us to minimize the total maintenance cost and reduces the frequent breaks of production. As a second part, this paper contributed to the development of a new risk analysis study further to unforeseen disruption of availability machines. This risk analysis developed a decision-making system, for production stakeholders, in order to achieve the forecasting lot-size plan and keeps its profitability, by specifying the unit cost threshold of subcontractor in the context of a calling of tender.

The purpose of this paper is to provide a comparison of synchronous permanent magnet machine types for wide constant power speed range operation.

A combination of analytical models and finite element analysis is used to conduct this study.

The paper has presented a detailed comparison between various types of synchronous PM machines for applications requiring a wide speed range of constant‐power operation. Key observations include: surface permanent magnet (SPM) and interior permanent magnet (IPM) machines can both be designed to achieve wide speed ranges of constant‐power operation. SPM machines with fractional‐slot concentrated windings offer opportunities to minimize machine volume and mass because of their short winding end turns and techniques for achieving high‐slot fill factors via stator pole segmentation. High back‐emf voltage at elevated speeds is a particular issue for SPM machines, but also poses problems for IPM machine designs when tight maximum limits are applied. Magnet eddy‐current losses pose a bigger design issue for SPM machines, but design techniques can be applied to significantly reduce the magnitude of these losses. Additional calculations not included here suggest that the performance characteristics of the inverters accompanying each of the four PM machines are quite similar, despite the differences in machine pole number and electrical frequency.

The paper is targeting traction applications where a very wide speed range of constant‐power operation is required.

Results presented are intended to provide useful guidelines for engineers faced with choosing the most appropriate PM machine for high‐constant power speed ratio applications. As in most real‐world drive design exercises, the choice of PM machine type involves several trade‐offs that must be carefully evaluated for each specific application.

The paper provides a comprehensive comparison between different types of synchronous PM machines, which is very useful in determining the most suitable type for various applications.

Inventory control models deal with production planning in order to minimize inventory and shortage cost, while cellular manufacturing analysis mainly addresses how machines should be grouped and parts be produced. A mathematical programming model is developed using an integrated approach for production and inventory planning in a cellular manufacturing environment. The mathematical programming model minimizes inter‐cell material handling cost, finished‐good inventory cost and system set‐up cost. The non‐linear mixed integer programming model cannot be directly solved for real size practical problems due to its NP‐complexity. A decomposition‐based heuristic algorithm was then developed to efficiently solve the integrated planning and control problem. Numerical examples are provided to test and illustrate the model and the solution method presented in this paper.

In many manufacturing systems, machines are subject to preventive maintenance. This paper aims to schedule the operations of jobs and preventive maintenance tasks in such a way that the completion time of jobs and preventive maintenance tasks is minimized.

An heuristic approach based on artificial immune algorithm is proposed for solving the multiple‐route job shop‐scheduling problem subject to fixed periodic and age‐dependent preventive maintenance tasks. Under fixed periodic assumption, the time between two consecutive preventive maintenance tasks is assumed constant. Under age‐dependent assumption, a preventive maintenance task is triggered if the machine operates for a certain amount of time. The goal is to schedule the jobs and preventive maintenance task subject to makespan minimization.

In addition to presenting mathematical formulation for the multiple‐route job shop‐scheduling problem, this paper proposes a novel approach by which one can tackle the complexity that is raised in scheduling and sequencing the jobs and the preventive maintenance simultaneously and obtain the required schedule in reasonable time.

Integrating preventive maintenance tasks into the scheduling procedure is vital in many manufacturing systems. Using the proposed approach, one can obtain a schedule that defines the production route through which each part is processed, the time each operation must be started, and when preventive maintenance must be carried out on each machine. This, in turn, results in overall manufacturing cost reduction.

Using the approach proposed in this paper, good solutions, if not optimal, can be obtained for scheduling jobs and preventive maintenance task in one of the most complicated job shop configurations, namely, multiple‐route job shop. Thus, the approach can dominate all other simpler configurations.

Presents a mixed‐integer programme to simultaneously select part routeings and form machine cells in the presence of alternate process plans so that the total cost of operating and intercell material handling is minimized. Demand for parts, machine capacities, number of cells to be formed, and number of machines in a cell are included in the model. Includes an example to illustrate the solution technique of the problem of practical instance.

The purpose of this paper is to propose an algorithm based on genetic algorithm (GA) to solve the block erection scheduling problem in shipbuilding.

The block erection scheduling problem is defined as the identical parallel machine‐scheduling problem with precedence constraints and machine eligibility (PCME) restrictions. A GA is proposed to find near optimal solution, and a few lower bounds and the percentage of the reduced makespan are defined to evaluate the performance of the proposed algorithm. Finally, the GA for block erection scheduling problem in a shipyard is illustrated by using the real data from a local shipyard.

The proposed GA produces lesser values of makespan against the random heuristic algorithm for the collected real instances.

The proposed GA can solve other similar parallel machine‐scheduling problems with PCME to minimize makespan.

Based on the proposed GA, the developed scheduling system for block erection in a shipyard can reduce the makespan of block erection, and contribute to the productivity improvement.

The allocation of block erection to the crane is modeled as a parallel machine‐scheduling problem with PCME, and the GA is developed to solve this problem to minimize makespan.

It is common knowledge that during the last decade markets have become extremely competitive with product variety increasing continuously and product life cycles shortening. Many manufacturing companies, which hitherto satisfied their customers while operating specific production systems, were recently obliged to reconsider because of the potential superiority of other “manufacturing philosophies”. In the literature, we meet a great variety of production systems and manufacturing philosophies, while, on the other side, in industry we usually find different combinations of “primary” productions systems. In this paper, we present the existing “state‐of‐the‐art” theoretical and experiential knowledge about productions systems, as well as describe their basic characteristics in a useful, exact and comprehensive way for practitioners and software houses who want to have a knowledge base for further research and practical implementation in the wider field of production management, planning and scheduling.