A Hybrid Genetic Algorithm-Neural Network Approach for Pricing Cores and Remanufactured Cores

Abstract

Sustainability has become a major issue in most economies, causing many leading companies to focus on product recovery and reverse logistics. Remanufacturing is an industrial process that makes used products reusable. One of the important aspects in both reverse logistics and remanufacturing is the pricing of returned and remanufactured products (called cores). In this paper, we focus on pricing the cores and remanufactured cores. First we present a mathematical model for this purpose. Since this model does not satisfy our requirements, we propose a simulation optimisation approach. This approach consists of a hybrid genetic algorithm based on a neural network employed as the fitness function. We use automata learning theory to obtain the learning rate required for training the neural network. Numerical results demonstrate that the optimal value of the acquisition price of cores and price of remanufactured cores is obtained by this approach. OPSOMMINGVolhoubaarheid het 'n belangrike saak geword in die meeste ekonomieë, wat verskeie maatskappye genoop het om produkherwinning en omgekeerde logistiek te onder oë te neem. Hervervaardiging is 'n industriële proses wat gebruikte produkte weer bruikbaar maak. Een van die belangrike aspekte in beide omgekeerde logistiek en hervervaardiging is die prysbepaling van herwinne en hervervaardigde produkte. Hierdie artikel fokus op die prysbepalingsaspekte by wyse van 'n wiskundige model.

Decision makings in key remanufacturing activities to optimise remanufacturing outcomes: A review

“…Some of the most commonly used modelling approaches are game theory (Bulmus et al, 2013), optimal control (Zhou and Yu, 2011), Markov chain (Vercraene et al, 2014) and mixed integer programming (Nenes and Nikolaidis, 2012). Our findings also uncover other modelling approaches such as non-linear programming (Seidi and Kimiagari, 2010), real option valuation (Wei and Tang, 2014), Bayesian estimation of distributed lag model (Clottey, 2012, Clottey, 2016 and multi-period stochastic dynamic programming (Xie et al, 2015). Table 6 illustrates that most papers about core acquisition (14 of 20) use optimal control as the modelling approach since optimal control is a mature mathematical discipline in science and engineering.…”

“…The majority of research papers (8 of 12 papers) assumed that the acquisition price is a linear function as follows. Acquisition price (A) = uN , where u is a constant unit acquisition cost and N refers to the number of acquired products/components (Galbreth and Blackburn, 2006, Yang et al, 2014, Bulmus et al, 2014, Teunter and Flapper, 2011, Seidi and Kimiagari, 2010. However, some researchers have suggested that acquisition can also be a non-linear function because the uncertain return rate of usedproduct returns and the fluctuating demand for remanufactured products will influence the acquisition price dynamically.…”

“…If discrete distribution is applied, cores of the same grade have the same quality level and remanufacturing costs. The number of cores at each grade is assumed to be deterministic (Guide et al, 2003, Galbreth and Blackburn, 2006, Seidi and Kimiagari, 2010, Teunter and Flapper, 2011, Pokharel and Liang, 2012, Yang et al, 2014, Yang et al, 2016a. With regard to continuous distribution, Ferguson (2011) has proposed that the returned cores have quality q ∈ [0, 1] as shown in Figure 7, where 0 is the minimum quality of returned cores, 1 is the maximum quality of returned cores and the quality probability density function changes over time (Wei et al, 2015) .…”

Returned Product Acquisition Pricing by Adaptive Neuro Fuzzy Inference System

“…Liang ve ark tarafından yapılan çalışmada, birden fazla alıcı ve satıcının bulunduğu serbest Pazar ortamında sadece yeniden üretim süreci için değil aynı zamanda çekirdek ürünlerin gelecek satış fiyatları için çekirdek fiyatlaması yapılmıştır [7]. Dönen ürünlerdeki belirsizliğin yüksek olması nedeniyle fiyatlandırmada genellikle simülasyon ve stokastik yaklaşımlar kullanılmıştır (bknz [7][8][9][10]). Telekomünikasyon ürünlerinin farklı tersine lojistik işlemlerine uygun olması ve son kullanıcılarla perakendeci arasındaki ilişkiler (yenileme programları, servis hizmetleri vs.) sebebiyle fiyatlandırma ve diğer kararlar sıklıkla çalışılmıştır (bknz [11][12][13]).…”