论文部分内容阅读
将模拟退火思想和加速遗传特性相结合,改进选择策略和遗传算子,建立加速遗传模拟退火算法(AGSA);基于支持向量机(SVM)的非线性回归和改进混合遗传算法的因子筛选,构造了支持向量机模型参数的自适应优化算法,提出连续微滤系统(CMF)产水预测模型;通过实测中试规模连续微滤系统产水量变化对模型进行验证,结果表明:该模型较好地揭示了CMF系统产水变化规律,模拟与实测结果间的误差小、相关性强(R2=0.91、MAE=0.0132、SSE=0.0055、RMSE=0.0155),说明模型具有较强的预测能力;采用留一法对训练样本进行交叉验证(R2=0.89、MAE=0.0164、SSE=0.0073、RMSE=0.0178),表明该模型同时具有良好的稳健性。此外,将基于AGSA-SVM的模型与神经网络BP算法进行了比较,结果显示:应用AGSA-SVM建立的模型在稳健性和预测能力都优于BP算法,因此该算法更适合于对CMF系统进行产水预测研究。
Combining the idea of simulated annealing with the acceleration of genetic characteristics, we improve the selection strategy and genetic operators to establish an accelerated Genetic Simulated Annealing Algorithm (AGSA), nonlinear regression based on Support Vector Machine (SVM) and genetic algorithm with improved genetic algorithm, The adaptive optimization algorithm based on support vector machine model parameters was proposed and the prediction model of continuous microfiltration system (CMF) was put forward. The model was validated by the measured water production of pilot scale continuous microfiltration system. The results showed that the model was better (R2 = 0.91, MAE = 0.0132, SSE = 0.0055, RMSE = 0.0155), which shows that the model has a strong ability to predict; the use of retention One method was used to cross-validate the training samples (R2 = 0.89, MAE = 0.0164, SSE = 0.0073, RMSE = 0.0178), indicating that the model has good robustness. In addition, comparing the AGSA-SVM-based model with the BP neural network algorithm, the results show that: the model established by AGSA-SVM is superior to BP algorithm in robustness and predictive ability, so the algorithm is more suitable for CMF system Research on Water Production Prediction.