China
Africa
Explore chapters and articles related to this topic
Preventive Maintenance Modeling
Published in Mangey Ram, Reliability Engineering, 2019
Sylwia Werbińska-Wojciechowska
Moreover, some research studies are based on the implementation of linear programming (see [194]), genetic algorithms (see [195,196]), dynamic programming (see [197]), theory of optimal stopping (see [198]), fuzzy modeling approach (see [199]), and simulations (see [200]). A generalized modeling method for maintenance optimization of single- and multi-unit systems is given in [182]. Moreover, a Bayesian perspective in opportunistic maintenance is investigated in [201], where the authors propose a PM policy for multi-component systems based on dynamic Bayesian networks (DBN)—Hazard and Operability Study (HAZOP) model. The use of expert judgment to parameterize a model for degradation, maintenance, and repair is provided in [202].
Virtual metrology in long batch processes using machine learning
Published in Materials and Manufacturing Processes, 2023
Ritam Guha, Anirudh Suresh, Jared DeFrain, Kalyanmoy Deb
In Fig. 1, we illustrate an example of the sensor readings present in the dataset.2 Two different batches are stopped at two different timesteps. But we do not know if stopping the processes at these timesteps was optimal. Manually analyzing the processes and stopping them is tedious and not optimal. VM can help us decide on predicting an optimal stopping time. In deployment, the idea is to collect the process parameters for timesteps and simulate[14] the process for additional timesteps and predict the process outcome at the final timestep . By varying ta, we can analyze the response for stopping at different timesteps and then we can select a suitable ta that leads to a reasonable output. This is how the stopping time can be decided by taking help from VM. Moreover, we can notice that the two batches in Fig. 1 have different sensor readings along the timeline. This indicates that there is a data drift observed in the dataset, which can happen due to multiple reasons like recipe change, sensor modifications, data collection process updates, etc. Due to the data drift, the mapping problem becomes non-trivial and a generalized VM framework should be able to capture this drift to make efficient predictions.
How to design a dynamic feed-in tariffs mechanism for renewables – a real options approach
Published in International Journal of Production Research, 2020
Li Li, Junqi Liu, Lei Zhu, Xiao-Bing Zhang
The bounded is integral and . Therefore, an optimal investment trigger exists. Similarly, with the pricing of a European call option on payoff function , the solution of an optimal stopping-time problem, namely, , is involved, with the threshold ; if , the discounted factor is where and are confluent hypergeometric functions (Dixit and Pindyck 1994; Takashima et al. 2012). Accordingly,
Supporting small suppliers through buyer-backed purchase order financing
Published in International Journal of Production Research, 2018
Boray Huang, Andy Wu, David Chiang
Early researchers in inventory control begin their examination of financial issues by putting a budget limitation on purchase or production costs, see Hadley and Whitin (1963).In 2000s, payment terms are considered into the decision process for the economic order quantity, see Mahata and Goswami (2007) and Huang and Huang (2008). However, the financial requirements in these studies are presented as the exogenous factors rather than endogenous decisions. Some scholars began to integrate financial decisions with operational strategies. For instance, Xu and Birge (2004) consider the joint strategy of debt financing and production capacity. Babich and Sobel (2004) use an optimal stopping time model to link the IPO decision with inventory operations. Gaur and Seshadri (2005), Caldentey and Haugh (2006), and Ding, Dong, and Kouvelis (2007) coordinate hedging strategies with inventory operations. Kouvelis and Zhao (2011, 2015) examine how the bankruptcy costs affect the retailers’ optimal order quantities. Li, Shubik, and Sobel (2013) analyse joint decision models about production, borrowing and dividend policies.