海南省波浪能开发智能化适宜性分析模型研究

Research on Intelligent Model for Suitability Analysis of Wave Energy Development in Hainan Province

针对目前波浪能开发适宜性分析方法存在的问题及工作量大、成本高的特点,基于地理信息系统(Geographic information system, GIS)、改进的多准则决策方法(Multi-criteria decision making, MCDM)和人工神经网络(Artificial neural network, ANN)提出海南省波浪能开发智能化适宜性分析模型。对于该模型:首先,使用模糊层次分析(Fuzzy analytic hierarchy process, FAHP)法计算评价指标权重,降低专家主观偏差,更好地描述信息的不确定性;其次,提出灰色关联分析-折衷优化(Grey relation analysis-vlse kriterijumska optimizacijaⅠkompromisno resenje, GRA-VIKOR)法计算开发适宜性指数,解决评价过程中部分信息丢失、评价结果不准确的问题;最后依托反向传播(Back propagation, BP)神经网络进行模型训练并进行验证,实现海南省波浪能开发适宜性分析的智能化,不仅提高工作效率,而且降低计算成本。通过模型得到海南省波浪能可开发区域和开发适宜性等级,为海南省波浪能选址决策奠定基础,同时可填补海南省波浪能开发适宜性分析在智能化领域的空白。

In view of the problems existing in the current wave energy development suitability analysis method and the characteristics of large workload and high cost,this paper proposes an intelligent suitability analysis model for wave energy development in Hainan Province based on geographic information system(GIS),improved multi-criteria decision making(MCDM)and artificial neural network(ANN).In this model,fuzzy analytic hierarchy process(FAHP)is used to calculate the weight of evaluation criteria,which can reduce the subjective bias of experts and better describe the uncertainty of information.Grey relation analysis-vlse kriterijumska optimizacijaⅠkompromisno resenje(GRA-VIKOR)is proposed to calculate the suitability index,which solves the problems of partial information loss and inaccurate evaluation results in the evaluation process.The model is trained and verified based on back propagation(BP)neural network to realize the intelligence.The feasible areas and suitability levels of wave energy in Hainan province are obtained by the model,which lays a foundation for the site selection decision making of wave energy in Hainan Province and fills the gap in the field of intelligent wave energy development suitability analysis.