Social-spider optimization algorithm for improving ANFIS to predict biochar yield

Abstract

The production of renewable and sustainable energy has more attention because the traditional energy sources such as fossil fuel are decreasing dramatically. The prediction of biochar yield from manure pyrolysis is considered as one type of renewable energy that used to produce energy. However, the experimental methods that used to produce energy from biochar yield are time-consuming and expensive, therefore, computational methods are applied to solve this problem. There are many methods applied to predict the biochar like least square-support vector machine (LS-SVM) and neural network. However, these methods can get stuck in local point and time complexity. To avoid these drawbacks, this paper works to improve the Adaptive Neuro-Fuzzy Inference System (ANFIS) using Social-Spider Optimization algorithm to predict biochar yield. The results of the proposed method are compared to classic ANFIS, artificial bee colony, particle swarm optimization, and LS-SVM. The results of ANFIS-SSO approach outperformed the standard ANFIS and they are better than other approaches.