An ant colony optimisation algorithm for scheduling in agile manufacturing

Producing customised products in a short time at low cost is one of the goals of agile manufacturing. To achieve this goal an assembly-driven differentiation strategy has been proposed in the agile manufacturing literature. In this paper, we address a manufacturing system that applies the assembly-driven differentiation strategy. The system consists of machining and assembly stages, where there is a single machine at the machining stage and multiple identical assembly stations at the assembly stage. An ant colony optimisation (ACO) algorithm is developed for solving the scheduling problem of determining the sequence of parts to be produced in the system so as to minimise the maximum completion time (or makespan). The ACO algorithm uses a new dispatching rule as the heuristic desirability and variable neighbourhood search as the local search to make it more efficient and effective. To evaluate the performance of heuristic algorithms, a branch-and-bound procedure is proposed for deriving the optimal solution to the problem. Computational results show that the proposed ACO algorithm is superior to the existing algorithm, not only improving the performance but also decreasing the computation time.     

Ant colony optimisation of decision tree and contingency table models for the discovery of gene–gene interactions

In this study, ant colony optimisation (ACO) algorithm is used to derive near‐optimal interactions between a number of single nucleotide polymorphisms (SNPs). This approach is used to discover small numbers of SNPs that are combined into a decision tree or contingency table model. The ACO algorithm is shown to be very robust as it is proven to be able to find results that are discriminatory from a statistical perspective with logical interactions, decision tree and contingency table models for various numbers of SNPs considered in the interaction. A large number of the SNPs discovered here have been already identified in large genome‐wide association studies to be related to type II diabetes in the literature, lending additional confidence to the results.Inspec keywords: genetics, genomics, DNA, polymorphism, molecular biophysics, molecular configurations, ant colony optimisation, decision trees, bioinformatics, diseasesOther keywords: ant colony optimisation, decision tree, contingency table models, gene‐gene interactions, ACO algorithm, near‐optimal interactions, single nucleotide polymorphisms, SNP, genome‐wide association studies, type II diabetes     

Ant Colony Optimisation-based radiation pattern manipulation algorithm for Electronically Steerable Array Radiator Antennas

A new algorithm for manipulating the radiation pattern of Electronically Steerable Array Radiator Antennas is proposed. A continuous implementation of the Ant Colony Optimisation (ACO) technique calculates the optimal impedance values of reactances loading different parasitic radiators placed in a circle around a centre antenna. By proposing a method to obtain a suitable sampling frequency of the radiation pattern for use in the optimisation algorithm and by transforming the reactance search space into the search space of associated phases, special care was taken to create a fast and reliable implementation, resulting in an approach that is suitable for real-time implementation. The authors compare their approach to analytical techniques and optimisation algorithms for calculating these reactances. Results show that the method is able to calculate near-optimal solutions for gain optimisation and side lobe reduction.     

Efficient modelling and optimisation of hybrid multilayered plates subject to ballistic impact

Among the key challenges present in the modelling and optimisation of composite structures against impact is the computational expense involved in setting up accurate simulations of the impact event and then performing the iterations required to optimise the designs. It is of more interest to find good designs given the limitations of the resources and time available rather than the best possible design. In this paper, low cost but sufficiently accurate finite element (FE) models were generated in LS Dyna for several experimentally characterised materials by semi-automating the modelling process and using existing material models. These models were then used by an optimisation algorithm to generate new hybrid offspring, leading to minimum weight and/or cost designs from a selection of isotropic metals, polymers and orthotropic fibre-reinforced laminates that countered a specified impact threat. Experimental validation of the optimal designs thus identified was then successfully carried out using a single stage gas gun. With sufficient computational hardware, the techniques developed in this pilot study can further utilise fine meshes, equations of state and sophisticated material models, so that optimal hybrid systems can be identified from a wide range of materials, designs and threats.     

An enhanced ant colony optimiser for multi-attribute partner selection in virtual enterprises

Increasing global competition drives enterprises, especially small and medium-sized enterprises, to collaborate in order to respond faster to customers’ needs, reduce operating costs, increase capacity, and produce customised products to reach the market quicker. A virtual enterprise (VE) is an important manufacturing paradigm to address this trend in the dynamic global economy. Partner selection is a key issue tightly coupled to the success of a VE coalition, and because of its complexity, it is considered a multi-attribute optimisation problem. In this paper, an enhanced ant colony optimiser (ACO) is proposed to address the partner selection problem. Five attributes (namely, cost, time, quality, reputation, and risk) considering both qualitative and quantitative aspects have been investigated to evaluate the candidate partners. Experiments have been conducted to validate the performance of the enhanced ACO algorithm, and the results show that the enhanced ACO algorithm can produce better results in terms of search accuracy and computing time.     

Order-oriented cooperative sequencing optimisation in multi-mix-model assembly lines

Order-oriented products assembly sequence among different assembly lines becomes a critical problem for mass customisation manufacturing systems. It significantly affects system productivity, delivery time, and manufacturing cost. In this paper, we propose a new approach to extend the traditional products sequencing from mixed model assembly line (MMAL) to multi-mixed model assembly lines (MMMALs) to obtain the optimal assembly sequence with the objectives of minimising consumption waviness of each material in the lines, assembly line setup cost, and lead-time. A multi-objective optimisation algorithm based on variable neighbourhood search methods (VNS) is developed. We perform an industrial case study in order to demonstrate the practicality and effectiveness of the proposed approach.     

Multi-objective optimal design of sandwich panels using a genetic algorithm

In this study, an optimization problem concerning sandwich panels is investigated by simultaneously considering the two objectives of minimizing the panel mass and maximizing the sound insulation performance. First of all, the acoustic model of sandwich panels is discussed, which provides a foundation to model the acoustic objective function. Then the optimization problem is formulated as a bi-objective programming model, and a solution algorithm based on the non-dominated sorting genetic algorithm II (NSGA-II) is provided to solve the proposed model. Finally, taking an example of a sandwich panel that is expected to be used as an automotive roof panel, numerical experiments are carried out to verify the effectiveness of the proposed model and solution algorithm. Numerical results demonstrate in detail how the core material, geometric constraints and mechanical constraints impact the optimal designs of sandwich panels.     

Spatial scheduling strategy for irregular curved blocks based on the modified genetic ant colony algorithm (MGACA) in shipbuilding

This paper proposes a scheduling strategy for irregular curved blocks to address the complex spatiotemporal coupling scheduling problem related to the entered time, the entered sequence, the setting positions and the rotated angles for the curved blocks in a shipbuilding yard. The strategy presents a makespan-based curved blocks – classification and selection rule to fulfil the programming time for the entry of the curved blocks into the workplace and realises the suppression on the delay. Useless stepping search of curved blocks in occupied workplace is avoided by combining the lowest centre-of-gravity rule with the calculation method of the remained workplace proposed in this paper. A modified genetic ant colony algorithm was proposed, which apply the ease to premature characteristics of GA and the excellent local optimisation ability of ACO, to let and promote the algorithm falls into local optimum. Then the large-scale and full-range mutation will be implemented to make the algorithm jump out of the original local optimisation to search more local optimal solutions so that the global optimal solution can be achieved. Finally, a software system for algorithm verification was developed which conducts the comparative analysis of the algorithms and verifies the validity of the algorithm proposed.     

Multilayered composite structure design optimisation using distributed/parallel multi-objective evolutionary algorithms

This paper presents a research work on stacking sequence design optimisation for multilayered composite plate using a parallel/distributed evolutionary algorithm. The stacking sequence of fibres has a dramatic influence on the strength of multilayered composite plates. Multiple layers of fibre-reinforced material systems offer versatility in engineering material design due to the fact that the stacking sequence of each orthotropic layer can offer full advantage of superior mechanical properties. Numerical results show that the optimal composite structures have lower weight, higher stiffness and also affordable cost when compared to the extreme and intermediate composite structures. In addition, the benefits of using a parallel optimisation system are also presented.     

基于智能混合策略的测试集优化

利用基于粒子群和蚁群算法的智能混合优化策略,删除冗余测试向量以解决测试集的优化问题. 利用蚁群算法的并行搜索能力构造初始解集,通过粒子群优化算法将解集维数降低,确定每次迭代的个体最优解和全局最优解,并利用新粒子信息更新信息素,最终通过多次迭代找到一个或多个最优测试集. 通过多组数据实例分析可知: 该智能混合优化策略与蚁群算法等其他测试集优化算法相比,可得到多个可行性最优测试集;与蚁群算法相比可提高收敛速度,并降低蚁群算法参数选取对收敛结果的影响,从而避免次优解的出现.