基于极限学习机的绞吸挖泥船产量预测研究

Research on output prediction of cutter suction dredger based on extreme learning machine

绞吸挖泥船在施工过程中,实际施工产量受土壤、水文、气象和水下杂物等许多因素的影响,难以获得准确的预测结果,且绞吸式挖泥船产量直接关系到工程效益。针对目前行业对绞吸挖泥船产量的预测过于依赖施工经验且实时性差的缺点,提出了一种基于极限学习机的预测方法,根据实际施工数据,构建产量预测模型。实验结果表明,该方法预测精度高且稳定性好,可以应用于挖泥船疏浚作业过程中,具有一定的工程实用价值。

During the construction of the cutter suction dredger,the actual construction output is affected by many factors such as soil,hydrology,meteorology and underwater debris.It is difficult to obtain accurate prediction results,and the output of the cutter suction dredger is directly related to the project benefit.Aiming at the shortcomings of the current industry that the output prediction of cutter suction dredgers is too dependent on construction experience and the real-time performance is poor,a prediction method based on extreme learning machine(ELM)was proposed,and a production prediction model was constructed based on actual construction data.The experimental results show that the method has high prediction accuracy and good stability,can be applied to the dredging process of dredgers,and has certain engineering practical value.