本文關(guān)鍵詞： 決策 創(chuàng)新技術(shù)擴(kuò)散 動態(tài)影響因素 能量模型 反向傳播算法 多智能體 蟻群算法　出處：《華中科技大學(xué)》2008年博士論文　論文類型：學(xué)位論文

【摘要】： 當(dāng)今時代,科學(xué)技術(shù)飛速發(fā)展,誰擁有先進(jìn)技術(shù)誰就擁有主動權(quán),誰就有可能在經(jīng)濟(jì)上領(lǐng)先于其它國家。作為推動科學(xué)技術(shù)發(fā)展的主要源泉——科技創(chuàng)新是實(shí)現(xiàn)經(jīng)濟(jì)可持續(xù)增長的根本動力,這一點(diǎn)已經(jīng)成為了共識。要想提高經(jīng)濟(jì)增長的質(zhì)量和效益,改變現(xiàn)有的高耗能、低產(chǎn)出的生產(chǎn)水平,需要大力進(jìn)行技術(shù)創(chuàng)新。作為廣義技術(shù)創(chuàng)新的后續(xù)子過程——創(chuàng)新技術(shù)擴(kuò)散是技術(shù)創(chuàng)新研究的一個重要領(lǐng)域。一項(xiàng)創(chuàng)新技術(shù)只有借助擴(kuò)散,它的潛在經(jīng)濟(jì)效益才能最大限度地發(fā)揮出來。企業(yè)作為智能體采納創(chuàng)新技術(shù)是創(chuàng)新技術(shù)擴(kuò)散的主要方式。是否采納、何時采納、優(yōu)化選擇采納創(chuàng)新技術(shù),是提高企業(yè)采納創(chuàng)新技術(shù)成功率并進(jìn)一步形成核心競爭力的關(guān)鍵。要較好的達(dá)到這一目標(biāo),重點(diǎn)分析影響創(chuàng)新技術(shù)擴(kuò)散的動態(tài)因素,建立企業(yè)采納創(chuàng)新技術(shù)多因素能量模型,綜合運(yùn)用反向傳播學(xué)習(xí)算法、多智能體、蟻群算法等工具,研究競爭性創(chuàng)新技術(shù)擴(kuò)散演化系統(tǒng)及互補(bǔ)性創(chuàng)新技術(shù)擴(kuò)散演化系統(tǒng)的規(guī)律,對于更好的推廣創(chuàng)新技術(shù)、快速提高企業(yè)競爭實(shí)力無疑具有非常重要的理論意義和現(xiàn)實(shí)意義。創(chuàng)新技術(shù)采納決策與擴(kuò)散問題主要從以下幾個方面進(jìn)行研究: 借鑒國內(nèi)外相關(guān)研究成果,針對創(chuàng)新技術(shù)擴(kuò)散規(guī)律的復(fù)雜非線性特征和多樣性要求,采用從理論到模型、仿真再到案例分析的研究思路,力求在理論和實(shí)際應(yīng)用方面有所創(chuàng)新。首先提出了影響企業(yè)采納創(chuàng)新技術(shù)的動態(tài)影響因素——稱為企業(yè)動能的概念;其次在此基礎(chǔ)上根據(jù)資源約束條件下效用函數(shù)優(yōu)選算法建立企業(yè)優(yōu)化采納創(chuàng)新技術(shù)多因素能量模型;然后運(yùn)用反向傳播學(xué)習(xí)算法結(jié)合能量模型進(jìn)行定量分析,克服了多個影響因素權(quán)重配置的問題,完善了企業(yè)采納創(chuàng)新技術(shù)評價(jià)體系,有效解決了企業(yè)采納創(chuàng)新技術(shù)中是否采納、何時采納的問題;最后通過虛擬企業(yè)生產(chǎn)技術(shù)選擇采納例證,為企業(yè)決策優(yōu)化采納提供了高效率的預(yù)測手段和分析方法。 在給定多項(xiàng)競爭性創(chuàng)新技術(shù)擴(kuò)散系統(tǒng)定義及假設(shè)的基礎(chǔ)上,首先利用多智能體技術(shù)提出了一種新的研究方法,將多智能體的建模理論與方法應(yīng)用到競爭性創(chuàng)新技術(shù)擴(kuò)散的分析當(dāng)中,建立了多項(xiàng)競爭性創(chuàng)新技術(shù)擴(kuò)散系統(tǒng)的動態(tài)演化模型;其次詳細(xì)分析了在競爭性創(chuàng)新技術(shù)擴(kuò)散系統(tǒng)中,當(dāng)擴(kuò)散速度等于或大于零時的系統(tǒng)平衡問題;然后通過大量的仿真實(shí)驗(yàn),剖析了不同的企業(yè)多因素能量強(qiáng)度和競爭作用強(qiáng)度對擴(kuò)散過程的影響;最后以通信技術(shù)的擴(kuò)散為對象進(jìn)行了案例分析,將動態(tài)演化模型、多智能體模型和實(shí)際數(shù)據(jù)進(jìn)行了對比,對未來的手機(jī)用戶擴(kuò)散數(shù)量進(jìn)行了預(yù)測。分析結(jié)果驗(yàn)證了該動態(tài)演化模型的可行性、有效性和實(shí)用性,有利于企業(yè)根據(jù)競爭性創(chuàng)新技術(shù)擴(kuò)散規(guī)律優(yōu)選采納決策、實(shí)現(xiàn)經(jīng)濟(jì)持續(xù)增長。 在互補(bǔ)性創(chuàng)新技術(shù)擴(kuò)散研究領(lǐng)域,首先分析群集智能研究中的蟻群算法(ACA)。發(fā)現(xiàn)兩對相似因素:蟻群找到從巢穴到食物源的最短路徑和企業(yè)成功采納創(chuàng)新技術(shù)達(dá)到效用最大化,信息激素遺留濃度和企業(yè)成功采納次數(shù)的累積。以此為基礎(chǔ)提出了一種新的研究思路,進(jìn)行了基于蟻群作用原理的創(chuàng)新技術(shù)采納方案的詳細(xì)設(shè)計(jì)。其次引入創(chuàng)新技術(shù)互補(bǔ)作用系數(shù),建立了多項(xiàng)互補(bǔ)性創(chuàng)新技術(shù)擴(kuò)散系統(tǒng)的動態(tài)演化模型。然后以上述原理與技術(shù)為基礎(chǔ),開發(fā)相應(yīng)的軟件支撐工具,進(jìn)行仿真分析。最后以無線關(guān)聯(lián)投影機(jī)為實(shí)例,利用相關(guān)的互補(bǔ)傳輸技術(shù)進(jìn)行了案例分析,找出互補(bǔ)性創(chuàng)新技術(shù)中對擴(kuò)散系統(tǒng)演化行為影響較大的關(guān)鍵技術(shù)。分析結(jié)果驗(yàn)證了該動態(tài)演化模型的實(shí)用性和有效性,便于企業(yè)根據(jù)互補(bǔ)性創(chuàng)新技術(shù)擴(kuò)散規(guī)律優(yōu)化采納決策。 基于上述研究成果,對生物芯片生產(chǎn)技術(shù)采納及檢測技術(shù)有效性選擇應(yīng)用實(shí)例進(jìn)行總體優(yōu)化設(shè)計(jì),并開發(fā)相應(yīng)的軟件支撐工具,進(jìn)行案例及仿真分析,進(jìn)一步驗(yàn)證了本文理論、方法、模型的正確性、實(shí)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.

【學(xué)位授予單位】：華中科技大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2008
【分類號】：F224;F062.4

【引證文獻(xiàn)】

相關(guān)期刊論文 前1條

1 趙劍冬;黃戰(zhàn);;基于Agent的經(jīng)濟(jì)社會系統(tǒng)建模與仿真研究[J];復(fù)雜系統(tǒng)與復(fù)雜性科學(xué);2011年04期

相關(guān)博士學(xué)位論文 前2條

1 馮纓;中小企業(yè)電子商務(wù)采納—實(shí)施—評價(jià)影響因素及方法研究[D];江蘇大學(xué);2010年

2 趙劍冬;基于Agent的產(chǎn)業(yè)集群企業(yè)競爭模型與仿真研究[D];華南理工大學(xué);2010年

相關(guān)碩士學(xué)位論文 前1條

1 孫永軍;基于多智能體的多元技術(shù)創(chuàng)新擴(kuò)散仿真研究[D];杭州電子科技大學(xué);2012年

，

本文編號：1536825