An Optimal Production Plan for Cashew Nuts Community Enterprise Using Metaheuristic Algorithms

Introduction

In recent years, the optimal production plan problem became one of the most important factors in the manufacturing process. The good plan can help the manufacturers in reducing the cost and the waste of the product. Thailand is a country full of the agricultural products. In order to increase the value, the products are always fed into various kinds of manufacturing system, for example, transformation, extending life cycle and packaging. Cashew nuts are a well-known product of Thailand exported to world-wide market. Thailand ranks as the third most important cashew nuts producing in Asia. In 2016, Thailand has only 14,704.64 hectare with major area in Uttaradit, Chonburi and Ubonratchathani, respectively (Department of Agricultural Extention, 2017). However, cashew nuts product trends to greatly decrease due to poor fruit set, cut down and substitute with other trees and low maintenance. Therefore, the proper production plan for cashew nuts during the manufacturing process is needed.

Optimization algorithms play an important role in various fields of study such as economic (Abdi et al., 2018), business (Wang et al., 2019), environment (Longo et al., 2019), biology (Remeseiro & Canedo, 2019), engineering (Houssein et al., 2020), computer science (Devikanniga et al., 2019), electronic (Janprom et al., 2020) and especially in industry. The main concept of optimization is designed to find the optimal solutions in the aspect of maximum or minimum value. It can be classified into deterministic and heuristic approaches (Lin et al., 2012). For industrial application, optimization is widely contributed to solve the optimal production plan or lot sizing problem. Based on related studies, a deterministic model is the most efficient method to solve production scheduling problem in various industrial sectors. However, as the problem becomes larger and more complex, the time taking to solve the problem will also increase. Silver meal algorithm (SM) (Silver & Miltenburg, 1984) is specifically designed to determine simply and effectively a replenishment strategy for the case of a time-varying, but deterministic demand pattern. The solution is to minimize average cost in each period. In addition, Rezaei and Davoodi (2008) use the deterministic model to solve the problem of supply chain with multiple suppliers and multiple products. Based on classical optimization methods, social and cultural data could not be analyzed in the model. The genetic algorithm (GA) is therefore applied to solve the problem. Khakdaman et al. (2015) develop a new optimization model through the development of linear programming to incofse production planning efficiency for hybrid make-to-stock–make-to-order business. The results show that the presented model can be applied in real life problem. However, complex mathematical processing requires high resources, so the integration of artificial intelligence can increase processing efficiency. However, deterministic model is unable to consider any uncertainties. The heuristic algorithm is developed to solve large-scale production scheduling. For example, Ho et al. (2007) propose production planning methods based on the effects of inventory deterioration. Three heuristic methods are improved as follows; net least period cost (nLPC), part-period algorithm (PPA), least total cost (LTC). It is the improvement of nLPC is the best performance under 100 conditions. Beck et al. (2015) propose a dynamic lot-sizing approach to inventory management using leinz–bossert–habenicht (LBH) method. Groff’s rule (GR) and least unit cost (LUC) are applied in LBH and called LBH-LUC and LBH-GR, respectively. The results show that LBH-LUC can be to reduce the cost variability of LUC compared to the WW method.