Search

Business process evaluation is a common norm in small-medium-large industries globally and information obtained during such evaluation have been used in simulating future performance of most industries using mathematical models such as Autoregressive Integrated Moving Average (ARIMA) and artificial neural network (ANN). This study explored the possibility of predicting port productivity and berth effectiveness of seaport using ANN and ARIMA. A comparative analysis of multi-layer perceptron (MLP) back propagation algorithm and ARIMA performance was carried out based on ships days at port, days at berth and tonnage which serves as model inputs, while port productivity and berth effectiveness were the model outputs. The MLP-ANN and ARIMA (1, 0, 4- port productivity) and (1, 0, 4-berth effectiveness) results were compared based on their coefficient of correlation and mean square error. The coefficient of correlation for port productivity prediction using MLP-ANN was 0.998. This value outperformed that of ARIMA (0.9862) for port productivity. The coefficient of correlation of 0.9956 and 0.9928 were obtained for berth effectiveness using MLR and ARIMA, respectively.

Prediction is necessary in industrial practices according to the norms of modern scientific development. Recognising maintenance problems before they occur potentially helps to improve workforce performance as effective preventive maintenance actions are developed. A large number of industrial systems worldwide owe their gains to predictive analysis. Furthermore, precise predictions can aid maintenance policy makers in making decisions as well as promoting the quality of decision making in maintenance, particularly on maintenance workforce issues. This study reports the use of grey relational analysis cum artificial neural networks (ANN) for maintenance workforce size prediction through the selection of an optimal neural network architecture. Workforce cost, workload, productivity and effectiveness represent the input parameters utilised in the ANN framework. The method competitively determines the most suitable ANN architecture for maintenance workforce size. Comparison of the ANN model results reveal that its performance is better than a fuzzy inference system. Conclusively, the application of the framework advanced in this work was found useful with practical data obtained from a process industry that operates as a brewery. The efficiency of the proposed approach was documented.

In manufacturing systems, stiff competition in the international markets coupled with the global economic crisis has exerted immense pressure to search for new maintenance approaches. This study focuses on determining the optimal maintenance time and workforce size, as well as a low overall life-cycle cost. Current literature has completely ignored the influence of these measures and the significance of prediction has been downplayed. Consequently, the necessity to develop an approach based on scientific foundations is affirmed. Keeping in view the status of the literature concerning this issue, a surrogate model (an evolutionary algorithm technique and fuzzy inference system (FIS)) for maintenance workforce cost prediction is presented. The key influencing parameters include production volume, routine maintenance time, workforce size and spare parts costs. The surrogate model was tested using process industry data. The result (root mean square error) obtained from the surrogate model was better than those of an artificial neural network (ANN) and FIS. The implication of these results is that the surrogate model could serve as a basis for determining maintenance workforce costs during decision-making processes. The novelty of this study is that it uses a differential evolution algorithm to improve the performance of FIS when dealing with maintenance workforce cost prediction.

Previous studies have shown that soft computing models are excellent predictive models for demand management problems. However, their applications in solving water demand forecasting problems have been scantily reported. In this study, feedforward artificial neural networks (ANNs) and a support vector machine (SVM) were used to forecast water consumption. Two ANN models were trained using different algorithms: differential evolution (DE) and conjugate gradient (CG). The performance of these soft computing models was investigated with real-world data sets from the City of Ekurhuleni, South Africa, and compared with conventionally used exponential smoothing (ES) and multiple linear regression (MLR). The results obtained showed that the ANN model that was trained with DE performed better than the CG-trained ANN and other predictive models (SVM, ES and MLR). This observation further demonstrates the robustness of evolutionary computation techniques amongst soft computing techniques.