优质期刊论文发表网提供专业的论文发表、论文写作以及期刊推广服务!QQ:820771224

电话:18796337551
当前位置:优质期刊论文发表网工商论文 → 文章正文

我国石油石化企业国际竞争力评价研究

作者:优质期刊论文发表网  来源:www.yzqkw.com  发布时间:2019/9/16 9:36:50  

摘要:随着全球贸易自由化和经济一体化的不断推进和加深,我国的石油石化企业也越来越广泛地参与到国际市场的竞争之中。随之而来,如何更大程度和更高效率地提高我国石油石化企业国际竞争能力,是我国石油石化产业界和理论界将要面对的一个重要问题。

从改革开放到市场经济的发展,再到我国加入WTO步入21新世纪,对比评估我国石油石化企业的国际竞争能力的强弱以及分析如何使我国石油石化企业的潜在比较优势转化为实际竞争优势,会有利于指导我国石油石化企业更好地面向国际市场进行生产运营,从而将有助于解决我国的能源问题,为我国政府采取合理的宏观经济政策提供客观依据。基于这样一种认识,本文借鉴传统国际竞争力评价模式,从企业资源、生产规模、资金管理、技术创新和社会效益等五大维度,构建石油石化企业国际竞争力评价指标体系,运用神经网络原理建立石油石化企业国际竞争力比较分析模型。通过对样本企业国际竞争力评价体系进行模型训练和仿真验证,对其结果进行分析,并提出以下对策建议。第一,提高石油石化企业管理效率。第二,加强石油石化企业创新能力。第三,推动石油石化企业国际化进程。第四,打造中国石油石化企业国际品牌。第五,推进石油石化企业的全球化能源互联网建设。

本文与以往的研究不同之处主要有三点。第一,在算法的设计上,我们通过无规律的选取一组权值,统一按照误差逆向传播的方向进行最速下降法的执行,解得待求权。另外通过多层信号的交互式传播,自动提取运算的求解规则,以达到准确无误地识别复杂数据相互之间隐性的非线性关系,进一步使网络训练函数无限地接近动态变化的函数模型。第二,在研究算法上,我们尝试采用自适应学习速率,并应用改进后的附加动量法,来避免传统神经网络方法中极易陷入局部极小值状态及收敛速度慢的问题。第三,在BP神经网络模型的基础上,应用Elman模型算法在其隐含层中增加承接层,进一步精化训练函数的筛选过程,使训练函数的选择更具有处理动态信息的能力,进一步增加对样本数据的敏感程度。

With the deepening of global tradeliberalization and economic integration, Chinesepetroleum and petrochemical enterprises are increasingly widely participatingin the international market competition. Along with this, how to improve theinternational competitiveness of China's petroleum and petrochemicalenterprises to a greater extent and more efficiently is an important issue thatChina's petroleum and petrochemical industry and the theoretical community willface.

From reform and opening up to thedevelopment of market economy, to China's accession to the WTO and entering the21st century, comparing and assessing the strength of China's petroleum andpetrochemical enterprises' international competitiveness and analyzing how toturn the potential comparative advantages of China's petroleum andpetrochemical enterprises into actual competitive advantages are useful. Itwill help guide China's petroleum and petrochemical enterprises to better operatein the international market, which will help solve China's energy problems andprovide an objective basis for the Chinese government to adopt reasonablemacroeconomic policies. Based on such an understanding, this paper draws on thetraditional international competitiveness evaluation model, and builds theinternational competitiveness evaluation index system of petroleum andpetrochemical enterprises from the five dimensions of enterprise resources,production scale, capital and management, technology and innovation and socialbenefits. Through the model training and simulation verification of the samplecompany's international competitiveness evaluation system, the results areanalyzed and corresponding countermeasures are proposed. First, improve the managementefficiency of petroleum and petrochemical enterprises. Second, strengthen theinnovation capability of petroleum and petrochemical enterprises. Third,promote the internationalization of petroleum and petrochemical enterprises.Fourth, build an international brand of China's petroleum and petrochemicalenterprises. Fifth, promote the construction of a global energy Internet forpetroleum and petrochemical enterprises.

There are three main points in this paperthat differ from previous studies. First, in the design of the algorithm, werandomly select a set of weights, and uniformly perform the method of thesteepest descent method according to the direction of the error reversepropagation, and obtain the right to be solved. In addition, through theinteractive propagation of multi-layer signals, the algorithm for solving theoperation is automatically extracted to achieve accurate and unambiguousidentification of the implicit nonlinear relationship between complex data, andfurther to make the network training function infinitely close to thedynamically changing function model. Second, in the research algorithm, we tryto adopt the adaptive learning rate and apply the improved additional momentummethod to avoid the problem that the traditional neural network method is easyto fall into the local minimum state and the convergence speed is slow.Thirdly, based on the BP neural network model, the Elman model algorithm isused to increase the acceptance layer in its hidden layer to refine the screeningprocess of the training function, so that the selection of the trainingfunction has the ability to process dynamic information, and further increasesensitivity to sample data.

关键词:石油石化;国际竞争力;神经网络

Petroleum and Petrochemical;International Competitiveness;Neural NetworkModel

联系方式

客服QQ 820771224
客服热线18796337551
网站地址 www.yzqkw.com
郑重承诺 高效,快速,包发表!
优质期刊论文发表网真诚欢迎新老客户的光临与惠顾!