在这项研究中,我们正在报告一种新颖的预测模型,用于预测微藻的二氧化碳(CO2)固定,该模型基于自适应神经模糊推理系统(ANFIS)和遗传算法(GA)的混合方法。收集各种藻类菌株的CO2固定率和微藻的培养条件,例如温度,pH值,CO2%,氮和磷的量(mg/L)作为输入变量,而CO2固定率作为输出变量。利用MATLAB通过遗传算法(GA)对ANFIS参数进行优化,形成优化后的模糊模型结构,以实现最优的预测能力和工业适用性。利用均方根误差(RMSE)等统计分析参数,计算出最佳拟合模型,回归系数(R2),和平均绝对相对偏差(AARD)。根据分析,GA-ANFIS模型描述了比ANFIS模型更高的预测能力。RMSE,在训练阶段,GA-ANFIS的R2和AARD分别为0.000431、0.97865和0.044354,在测试阶段分别为0.00056、0.98457和0.032156。分别,GA-ANFIS模型。因此,可以得出结论,提出的GA-ANFIS模型是一种有效的技术,具有很高的潜力来准确预测CO2固定速率。
In this study, we are reporting a novel prediction model for forecasting the carbon dioxide (CO2) fixation of microalgae which is based on the hybrid approach of adaptive neuro-fuzzy inference system (ANFIS) and genetic algorithm (GA). The CO2 fixation rate of various algal strains was collected and the cultivation conditions of the microalgae such as temperature, pH, CO2 %, and amount of nitrogen and phosphorous (mg/L) were taken as the input variables, while the CO2 fixation rate was taken as the output variable. The optimization of ANFIS parameters and the formation of the optimized fuzzy model structure were performed by genetic algorithm (GA) using MATLAB in order to achieve optimum prediction capability and industrial applicability. The best-fitting model was figured out using statistical analysis parameters such as root mean square error (RMSE), coefficient of regression (R2), and average absolute relative deviation (AARD). According to the analysis, GA-ANFIS model depicted a greater prediction capability over ANFIS model. The RMSE, R2, and AARD for GA-ANFIS were observed to be 0.000431, 0.97865, and 0.044354 in the training phase and 0.00056, 0.98457, and 0.032156 in the testing phase, respectively, for the GA-ANFIS Model. As a result, it can be concluded that the proposed GA-ANFIS model is an efficient technique having a very high potential to accurately predict the CO2 fixation rate.