创新技术采纳决策与扩散问题研究及应用

发布时间：2018-02-26 06:12

本文关键词： 决策创新技术扩散动态影响因素能量模型反向传播算法多智能体蚁群算法　出处：《华中科技大学》2008年博士论文　论文类型：学位论文

【摘要】： 当今时代,科学技术飞速发展,谁拥有先进技术谁就拥有主动权,谁就有可能在经济上领先于其它国家。作为推动科学技术发展的主要源泉——科技创新是实现经济可持续增长的根本动力,这一点已经成为了共识。要想提高经济增长的质量和效益,改变现有的高耗能、低产出的生产水平,需要大力进行技术创新。作为广义技术创新的后续子过程——创新技术扩散是技术创新研究的一个重要领域。一项创新技术只有借助扩散,它的潜在经济效益才能最大限度地发挥出来。企业作为智能体采纳创新技术是创新技术扩散的主要方式。是否采纳、何时采纳、优化选择采纳创新技术,是提高企业采纳创新技术成功率并进一步形成核心竞争力的关键。要较好的达到这一目标,重点分析影响创新技术扩散的动态因素,建立企业采纳创新技术多因素能量模型,综合运用反向传播学习算法、多智能体、蚁群算法等工具,研究竞争性创新技术扩散演化系统及互补性创新技术扩散演化系统的规律,对于更好的推广创新技术、快速提高企业竞争实力无疑具有非常重要的理论意义和现实意义。创新技术采纳决策与扩散问题主要从以下几个方面进行研究: 借鉴国内外相关研究成果,针对创新技术扩散规律的复杂非线性特征和多样性要求,采用从理论到模型、仿真再到案例分析的研究思路,力求在理论和实际应用方面有所创新。首先提出了影响企业采纳创新技术的动态影响因素——称为企业动能的概念;其次在此基础上根据资源约束条件下效用函数优选算法建立企业优化采纳创新技术多因素能量模型;然后运用反向传播学习算法结合能量模型进行定量分析,克服了多个影响因素权重配置的问题,完善了企业采纳创新技术评价体系,有效解决了企业采纳创新技术中是否采纳、何时采纳的问题;最后通过虚拟企业生产技术选择采纳例证,为企业决策优化采纳提供了高效率的预测手段和分析方法。在给定多项竞争性创新技术扩散系统定义及假设的基础上,首先利用多智能体技术提出了一种新的研究方法,将多智能体的建模理论与方法应用到竞争性创新技术扩散的分析当中,建立了多项竞争性创新技术扩散系统的动态演化模型;其次详细分析了在竞争性创新技术扩散系统中,当扩散速度等于或大于零时的系统平衡问题;然后通过大量的仿真实验,剖析了不同的企业多因素能量强度和竞争作用强度对扩散过程的影响;最后以通信技术的扩散为对象进行了案例分析,将动态演化模型、多智能体模型和实际数据进行了对比,对未来的手机用户扩散数量进行了预测。分析结果验证了该动态演化模型的可行性、有效性和实用性,有利于企业根据竞争性创新技术扩散规律优选采纳决策、实现经济持续增长。在互补性创新技术扩散研究领域,首先分析群集智能研究中的蚁群算法(ACA)。发现两对相似因素:蚁群找到从巢穴到食物源的最短路径和企业成功采纳创新技术达到效用最大化,信息激素遗留浓度和企业成功采纳次数的累积。以此为基础提出了一种新的研究思路,进行了基于蚁群作用原理的创新技术采纳方案的详细设计。其次引入创新技术互补作用系数,建立了多项互补性创新技术扩散系统的动态演化模型。然后以上述原理与技术为基础,开发相应的软件支撑工具,进行仿真分析。最后以无线关联投影机为实例,利用相关的互补传输技术进行了案例分析,找出互补性创新技术中对扩散系统演化行为影响较大的关键技术。分析结果验证了该动态演化模型的实用性和有效性,便于企业根据互补性创新技术扩散规律优化采纳决策。基于上述研究成果,对生物芯片生产技术采纳及检测技术有效性选择应用实例进行总体优化设计,并开发相应的软件支撑工具,进行案例及仿真分析,进一步验证了本文理论、方法、模型的正确性、实用性和有效性。
[Abstract]:Nowadays, the rapid development of science and technology, who have advanced technology who will have the initiative, who will probably be in the economy ahead of other countries. As the fundamental driving force to promote the development of science and technology, science and technology innovation is the main source to achieve sustainable economic growth, it has become a consensus in order to improve the quality and efficiency. Economic growth, change the existing high energy consumption, low production level, the need to vigorously carry out technological innovation. As the subsequent general technology innovation process, innovation technology diffusion is an important field in the research of technological innovation. An innovative technology only with diffusion, maximize the potential economic benefits to its play enterprise as the agent. The adoption of innovative technology is the main way of innovation technology diffusion. If adopted, when adopted, optimizing the adoption of innovative technologies, is to improve the enterprise. The key that technology innovation success rate and further form the core competitiveness. To achieve this goal, investigating dynamic factors affecting innovation technology diffusion, the establishment of enterprise innovation technology adoption multi factors energy model, using the backpropagation learning algorithm, multi agent, ant colony algorithm and other tools of competitive innovation diffusion the evolution of technology innovation diffusion system and complementary technology evolution system, for the promotion of innovation and technology better, it has very important theoretical significance and practical significance to rapidly improve the competitive power of enterprises. The adoption of decision making and innovation diffusion problems mainly from the following aspects:
From the relevant research results at home and abroad, aiming at the complex nonlinear characteristics of diffusion of technology innovation and diversity requirements, using the theoretical model, simulation and research of case analysis, and strive to be innovative in theory and practical application. Firstly puts forward the dynamic factors influencing enterprise innovation technology adoption -- known as the concept of enterprise on the basis of the kinetic energy; secondly according to the resource constraints effect the establishment of enterprise with the function optimization algorithm to optimize the multi factors energy innovation technology adoption model; then using the back-propagation learning algorithm can combine quantitative analysis model, to overcome the many factors affecting the weight allocation, improve the enterprise innovation technology adoption evaluation system, effective solution the adoption of innovative technology in the enterprise is adopted, when adopted; finally through the selection of virtual enterprise production technology adoption cases It provides an efficient method of forecasting and analysis for the optimization and adoption of enterprise decision.
In a given number of competitive technology diffusion system definitions and assumptions, first proposed a new method based on multi-agent technology, the multi-agent modeling theory and method applied to the analysis of the competition of innovation technology diffusion, established a dynamic evolution model of multiple competing technology innovation diffusion system the second; analyzes in detail the competitive technology diffusion system, when the diffusion speed is equal to or greater than zero balance system; then through lots of experiments, analyzes the influence factors of different enterprise energy intensity and the effect of competition strength on diffusion process; finally, case analysis to spread communication technology as the object a dynamic evolution model, compares the multi-agent model and actual data, diffusion number of future mobile phone users were predicted. The results verify The feasibility, effectiveness and practicability of the dynamic evolution model are favorable for enterprises to adopt decisions according to the diffusion rules of competitive innovation technology and achieve sustained economic growth.
In the complementary innovation technology diffusion field, first analysis of ant colony algorithm (ACA). The study found two pairs of similar factors: the ant colony to find the shortest path from the nest to the food source and the enterprise's successful adoption of innovative technologies to achieve maximum utility, the cumulative information hormone concentration and the number of successful adoption as. This paper proposes a new research method, the detailed design principle of ant colony innovation technology adoption scheme based on the introduction of technology innovation. Secondly complementary effect coefficient, the establishment of a number of complementary dynamic evolution model of innovation technology diffusion system. Then based on the principle and technology of the supporting software tool is developed. Simulation analysis is carried out. Finally the wireless Association projector as an example, a case is analyzed using complementary transmission technology, to find out the complementary innovation technology diffusion The key technologies that influence the evolution behavior of dispersed system are analyzed. The analysis results verify the practicability and effectiveness of the dynamic evolution model, and facilitate enterprises to optimize adoption decisions based on complementary innovation technology diffusion rule.
Based on the above research results, the overall optimization design of bio chip production technology adoption and detection technology of the effective selection of application examples, and supporting software tool is developed, by case analysis and simulation, this paper further verifies the theory, method, model accuracy, practicality and effectiveness.

【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2008
【分类号】：F224;F062.4

【引证文献】