关键词: Epsilon策略, Pareto算法, 伯努利预测, 多属性决策, 供应商参与, 二分类Logistic回归, 困难度, 伸及界面, 伸及能力, 生物力学负荷, 坐姿手伸及

Abstract: Aiming at the limitations of the traditional product scheme planning methods, in which the suppliers participate, existing in two aspects of components optimization and suppliers evaluation, an integrated two-stage planning method consisting of components multi-objective optimization and suppliers fuzzy dynamic multi-attribute decision evaluation is proposed. A math optimization model, whose objective is for quality, cost and delivery time of components combination scheme, is built. In the meanwhile, Epsilon strategy and predominant constraint trick are adopted to improve strength Pareto evolutionary algorithm. Afterwards, the improved algorithm is employed to solve abovementioned model and to get finite Pareto solutions. Fourier series is introduced to amend errors of Bernoulli forecasting model, and the amended model and fuzzy dynamic multi-attribute decision are integrated to evaluate suppliers pertaining to Pareto solutions, and then the optimal supplier for each component is selected and the best product planning scheme is acquired. Taking the supplier involved development process of air-separating equipment as a case, the result of which illustrates the feasibility and validation of the proposed method through simulation and computation.

Key words: Bernoulli forecasting, Epsilon strategy, Multi-attribute decision, Pareto algorithm, Supplier-involved, Binary Logistic Regression, Biomechanics loads, Difficulty, Reach capability, Reach envelope, Seated reach