Modeling and optimization of developed cocoa beans extractor parameters using box behnken design and artificial neural network
Introduction
Cocoa is an excellent primary material for the manufacture of chocolates, cosmetics, health drinks, pharmaceuticals, etc. It contains about 50 percent fat, which is useful for producing a candle, ointments, pharmaceutical products, cosmetics, etc. (Porter, 2006, Prosapio and Norton, 2019). The cocoa pod is the best source for the creation of potash manure, soap, biogas, and speck boards (Babayemi et al., 2010, Adjin-Tetteh et al., 2018). The global needs for cocoa beans have been steadily increased over recent decades due to the increased consumption of chocolate and chocolate-flavored products (Sunoj et al., 2016).
At present, the cocoa pods have been broken manually using wood and machete. Thus, it is a strenuous and time-consuming process. The manual method of cocoa pod breaking leads to damages to the beans, increase the percentage of bean losses, and reduces the profit. The strenuous task causes the persistent weakness and sickness of the labor and farmer, resulting in a low standard of health (Huang et al., 2010, Patrício and Rieder, 2018).
The extensive requirement of labour during the manual method results in the high cost of production. Adewumi and Fatusin, 2006, Chamsing et al., 2006 have been developed a hand operated mechanical cocoa pod breaker. The automated method of extracting the cocoa beans will reduce fatigue for the farmers and the labours. Also, the losses usually occur during the manual breaking of the pods and separating beans from the placenta with wood, and cutlass will be reduced, which enhances income for the farmers. The time required for extracting the beans using the manual method will be reduced by the mechanical way (Kate et al., 2018).
The optimization of machine design conditions becomes a problem which has been vigorously analyzed as concern agriculture, biosystems, and food processing industries pertained different devices and abilities were applied for this view in order to accomplish reasonable quality solutions (Sudha et al., 2016, Le Chau et al., 2019). BBD from Response Surface Methodology (RSM) can be depicted as an empirical modeling system and could be utilized for developing, improving, and optimizing complex processes (Madadlou et al., 2009, Chau et al., 2018, Chau et al., 2019). RSM has the advantage of decreasing the count of experimental runs, which is enough to furnish statistically acceptable results (Nourbakhsh et al., 2014). RSM used to optimize agricultural oriented machine design (Saldaña-Robles et al., 2020), post-harvest processing machine design (Sofu et al., 2016), food processing machine designs (Umani et al., 2019), dairy processing machine design (Madadlou et al., 2009), food product development optimization (Kothakota et al., 2013a, Kothakota et al., 2013b, Kothakota et al., 2016, Kothakota et al., 2017, Nagpal et al., 2018, Pandiselvam et al., 2019a), food process optimization (Santana et al., 2018, Sagarika et al., 2018, Shameena Beegum et al., 2019) and food preservation optimization (Nourbakhsh et al., 2014, Gaurh et al., 2017). Artificial neural network (ANN) have great learning capacity and potential of determining and modeling the complicated non-linear relationship among the input and the output of a system in comparison to RSM. ANN has appeared as an extra robust and excellent modeling tool, as a result of its ability to determine from observations and to draw conclusions using rationalization and predictive modeling from the behaviour of complex nonlinear method. Numerous investigations had stated on ANN applications for design and performance of the machine in agricultural processing (Saldaña-Robles et al., 2020), post-harvest processing of crops (Nourbakhsh et al., 2014, Santana et al., 2018), food storage (Sanaeifar et al., 2018), food preservation (Santana et al., 2018), dairy processing (Madadlou et al., 2009) and process optimization for product development (Nagpal et al., 2018, Pandiselvam et al., 2019a).
Considering the aforementioned factors, the post-harvest processing of cocoa has to be mechanized.. Hence, this present study aimed for 1) design and development of continuous cocoa beans extractor. 2) modeling and optimizing the machine performance parameters by applying BBD and RSM 3) cost analysis and field testing of a developed prototype in comparision with the manual process.
Section snippets
Raw material
Matured cocoa fruits (Theobroma cacao L.) of two verities (Criollo and Forastero) were procured from M/s Cadbury unit of Kerala Agricultural University, Thrissur, India, and the progressive farmer from Karuvarakundu, Malappuram, India. The cocoa pods were cleaned and sorted from cracked, skin injuries, and disease affected pods. Moisture content of cocoa pod husk was determined by AOAC (2000) method. The experiment was carried out at fixed moisture 81.21 ± 1% wb (for Criollo veriety) and
Experimental design
The performance evaluation and experimental optimization from a designed prototype machine (Fig. 1) have been executed through the Box-behnken design (BBD) and artificial neural network (ANN).
Engineering properties of cocoa pod
The engineering properties of cocoa pods could be useful for development of pod breaking machine. The physical properties of cocoa pod such as lenth (124–188 mm), diameter (73–97 mm), sphericity (0.62–0.70), volume (293–785 cm3), bulk density (384–417 kg/m3) and porosity (47–55%), were used to design the hopper. The frictional properties (SS-0.29 to 0.33, AS-0.32 to 0.34, GI-0.34 to 0.40 and PW- 0.33 to 0.39) on different materials would be useful for fabrication material selection and
Conclusion
The present study concentrated on the design and development of continuous cocoa bean extractor. During design, the outcome from the operation valuation disclosed that the shape of the pod, roller speed, and inclination of the strainer of the machine had more impact on machine performance pursed by roller clearance. BBD and ANN model have been applied for modeling and optimizing machine performance conditions. BBD and ANN regression coefficients have been found to maximum 0.994 and 0.999,
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
The financial support received from the Department of Food and Agricultural Process Engineering, Kerala Agricultural University, is gratefully acknowledged.
References (47)
- et al.
Thermochemical conversion and characterization of cocoa pod husk a potential agricultural waste from Ghana
Ind. Crops Prod.
(2018) - et al.
Development and testing of an automated grain drinks processing machine
Food Bioprod Process
(2017) - et al.
Development of soft computing and applications in agricultural and biological engineering
Comput. Electron. Agric.
(2010) - et al.
Cracking characteristics of walnut
Biosyst. Eng.
(2004) - et al.
Response surface optimization of an artificial neural network for predicting the size of re-assembled casein micelles
Comput. Electron. Agric.
(2009) - et al.
Ultrasound, microwave and Box-Behnken Design amalgamation offered superior yield of gum from Abelmoschus esculentus: Electrical, chemical and functional peculiarity
Comput. Electron. Agric.
(2018) - et al.
Prediction of red plum juice permeate flux during membrane processing with ANN optimized using RSM
Comput. Electron. Agric.
(2014) - et al.
Computer vision and artificial intelligence in precision agriculture for grain crops: A systematic review
Comput. Electron. Agric.
(2018) - et al.
Design, development and testing of hand-operated decorticator for Jatropha fruit
Appl. Energy
(2010) - et al.
Development of fat-reduced chocolate by using water-in-cocoa butter emulsions
J. Food Eng.
(2019)
Design and development of secondary controlled industrial palm kernel nut vegetable oil expeller plant for energy saving and recuperation
J. Food Eng.
Structural design of an agricultural backhoe using TA, FEA, RSM and ANN
Comput. Electron. Agric.
Fusion of dielectric spectroscopy and computer vision for quality characterization of olive oil during storage
Comput. Electron. Agric.
Optimization of vacuum cooling treatment of postharvest broccoli using response surface methodology combined with genetic algorithm technique
Comput. Electron. Agric.
Evaluation of a centrifugal impaction-type decorticator for shelling tung fruits
Ind. Crops Prod.
Design of an automatic apple sorting system using machine vision
Comput. Electron. Agric.
Optimization of process parameters in feed manufacturing using artificial neural network
Comput. Electron. Agric.
Nondestructive determination of cocoa bean quality using FT-NIR spectroscopy
Comput. Electron. Agric.
Development and testing of an automated contact plate unit for a cassava grater
Comput. Electron. Agric.
Novel method and machine for dehulling of guar seeds and optimisation of dehulling process
Food Bioprod. Process.
Design, fabrication and testing of an impact-type hand operated cocoa pod breaker
Int. J. Agric. Eng.CIGR J.
Development and performance evaluation of a cocoa pod breaking machine
J. Adv. Mater. Res.
Official Methods of Analysis
Cited by (32)
Developing digital twin design for enhanced productivity of an automated anodizing industry and process prediction using hybrid deep neural network
2023, Engineering Applications of Artificial IntelligenceMulti-objective optimization of seeding performance of a pneumatic precision seed metering device using integrated ANN-MOPSO approach
2023, Engineering Applications of Artificial IntelligenceCitation Excerpt :Its better universal function approximation capability, more remarkable generalization ability, and less sensitive nature to noisy or missing data (Luk et al., 2000; Haykin, 2009) make it suitable for complex non-linear modelling without prior knowledge about the input–output relationships, which is difficult to handle with the statistical approaches. In the past, this modelling technique was successfully implemented in various process modelling studies of agricultural machines (Anantachar et al., 2010; Kumar et al., 2009; Gundoshmian et al., 2010; Anantachar et al., 2011; Hiregoudar et al., 2011; Chandel et al., 2019; Zhang et al., 2019; Gundoshmian et al., 2019; Nadai et al., 2020; Srikanth et al., 2020; Pareek et al., 2021). In the majority of these studies, the most commonly used ANN configuration, i.e., the multilayer perception neural network (MLPNN), was applied for modelling purposes.
Design consideration and modelling studies of ultrasound and ultraviolet combined approach for shelf-life enhancement of pine apple juice
2022, Ultrasonics SonochemistryCitation Excerpt :For thermal pasteurisation, pineapple juice was heated at 80 °C for 15 min and cooled to ambient temperature [31]. The total cost of production of combined treated pineapple juice was calculated using the procedure reported by Srikanth et al. [52] according to our specifications. Colour is an essential visual indicator to determine juice quality and consumer acceptance of fruit juices [36].
A novel approach for improved in-situ biodiesel production process from gamma-irradiated castor seeds using synergistic ultrasound and microwave irradiation: Process optimization and kinetic study
2022, Industrial Crops and ProductsCitation Excerpt :Since the range of each parameter is properly understood and defined, the application of BBD is considered due to its advantage of excellence in predicting the responses with comparatively fewer experimental runs (Ideris et al., 2021). As reported in the literature, both the methods RSM based BBD and artificial intelligence based artificial neural network (ANN) method have proven their competency in predictive capabilities (de Luna et al., 2020; Srikanth et al., 2020). However, the BBD method has the advantage of using fewer experimental runs than ANN, which requires more experiments to train the neural network (de Luna et al., 2020).