Simulation has been used to model combat for a long time. Recently, it has been accepted that combat is a complex adaptive system (CAS). Multi-agent systems (MAS) are also considered as a powerful modelling and development environment to simulate combat. Agent-based distillations (ABD) - proposed by the US Marine Corp - are a type of MAS used mainly by the military for exploring large scenario spaces. ABDs that facilitated the analysis and understanding of combat include: ISAAC, EINSTein, MANA, CROCADILE and BactoWars. With new concepts such as networked forces, previous ABDs can implicitly simulate a networked force. However, the architectures of these systems limit the potential advantages gained from the use of networks. In this thesis, a novel network centric multi-agent architecture (NCMAA) is pro-posed, based purely on network theory and CAS. In NCMAA, each relationship and interaction is modelled as a network, with the entities or agents as the nodes. NCMAA offers the following advantages: 1. An explicit model of interactions/relationships: it facilitates the analysis of the role of interactions/relationships in simulations; 2. A mechanism to capture the interaction or influence between networks; 3. A formal real-time reasoning framework at the network level in ABDs: it interprets the emergent behaviours online. For a long time, it has been believed that it is hard in CAS to reason about emerging phenomena. In this thesis, I show that despite being almost impossible to reason about the behaviour of the system by looking at the components alone because of high nonlinearity, it is possible to reason about emerging phenomena by looking at the network level. This is undertaken through analysing network dynamics, where I provide an English-like reasoning log to explain the simulation. Two implementations of a new land-combat system called the Warfare Intelligent System for Dynamic Optimization of Missions (WISDOM) are presented. WISDOM-I is built based on the same principles as those in existing ABDs while WISDOM-II is built based on NCMAA. The unique features of WISDOM-II include: 1. A real-time network analysis toolbox: it captures patterns while interaction is evolving during the simulation; 2. Flexible C3 (command, control and communication) models; I 3. Integration of tactics with strategies: the tactical decisions are guided by the strategic planning; 4. A model of recovery: it allows users to study the role of recovery capability and resources; 5. Real-time visualization of all possible information: it allows users to intervene during the simulation to steer it differently in human-in-the-loop simulations. A comparison between the fitness landscapes of WISDOM-I and II reveals similarities and differences, which emphasise the importance and role of the networked architecture and the addition of strategic planning. Lastly but not least, WISDOM-II is used in an experiment with two setups, with and without strategic planning in different urban terrains. When the strategic planning was removed, conclusions were similar to traditional ABDs but were very different when the system ran with strategic planning. As such, I show that results obtained from traditional ABDs - where rational group planning is not considered - can be misleading. Finally, the thesis tests and demonstrates the role of communication in urban ter-rains. As future warfighting concepts tend to focus on asymmetric warfare in urban environments, it was vital to test the role of networked forces in these environments. I demonstrate that there is a phase transition in a number of situations where highly dense urban terrains may lead to similar outcomes as open terrains, while medium to light dense urban terrains have different dynamics

我們把經濟體視為一個複雜適應系統(complex adaptive system), 強調系統中異質性(heterogeneous)agent的學習適應行為與agent之間的互動性交互作用, 此時主流經濟學裡的分析架構, 如:代表性個人模型(represesentive agent model)、理性預期(rational expectation)、固定點均衡分析(fixed-point equilibrium analysis)等將不再適用, 取而代之的是演化經濟學(evolutionary economics)的研究典範, 這樣的研究架構下, 並沒有適當的數學分析工具可資運用, 因此我們改以agent-based建模(agent-based modelng)的社會模擬(social simulation)來建構一個人工的經濟體(artificial economy), 以此為主要研究方法, 這就是agent-based計算經濟學(agent-based computational economics)或稱人工經濟生命(artificial economic life)。 本文中以股票市場為主要的研究課題, 我們以遺傳規劃(genetic programming)的人工智慧(artificial intelligence)方法來模擬股市中有限理性(bounded rational)異質交易者的交易策略學習行為, 建構出一個人工股票市場(artificial stock market), 在這樣的架構下, 我們成功地產生出類似真實股票市場的股價時間序列特性, 我們同時也檢定了人工股票市場中價量的因果關係, 說明了在沒有外生因素之下, 人工股票市場的複雜系統可自發地產生出雙向的價量因果關係, 進一步地, 我們研究下層agent(交易者)行為與上層股價時間序列行為的關聯性, 我們也發現個體的行為並不能直接加總或推論出複雜適應系統的總體行為, 這就是突現性質(emergent property)的發生, 最後, 本文描述了agent-based計算經濟學研究架構的優勢與缺點, 再附帶介紹一個用以進行agent-based建模相關研究的軟體程式庫-SWARM。

在複雜動態競爭市場中，生產者的價格競爭行為一直是一個研究的重點，相較於生產者動態價格競爭，零售商的價格競爭行為鮮少被探討，因此本研究針對零售商價格競爭行為進行研究。針對零售商之間的價格競爭行為，除了考量零售商與對手零售商的價格互動，不可忽略的是上游供應商的競爭互動與下游消費者的學習行為在市場中與零售商端互動下錯綜複雜的動態影響，緣此，本研究以零售商端的角度，想了解供應商競爭與消費者學習行為對零售商競爭的影響，再以單一零售商角度，分析各情況下所應對的價格調整策略。 本研究將零售商、供應商及消費者互動形成之競爭市場視為一個複雜適應性系統（Complex Adaptive System ，簡稱CAS），應用代理人基塑模與模擬（Agent-based Modeling and Simulation，簡稱ABMS）方式建構考量供應商競爭與消費者學習行為之零售商價格競爭模式，將演化賽局理論應用於價格競爭中，探討不同的消費者學習及供應商價格競爭行為如何動態影響零售商價格競爭型態，以及不同價格調整策略之績效表現。 研究結果發現一，市場中消費者呈現不同的學習行為，對零售商競爭將造成不同的衝擊。「貨比三家無學習」型消費者將造成零售商端低價競爭，使其平均價格最低及獲利最低。「自我式學習」型消費者將造成零售商高價合作，使其平均價格最高及獲利最高。「群體式學習」型消費者同樣使零售商端偏向高價合作，且其平均價格及獲利相當接近自我式學習市場，雖然兩種學習行為具有近似的平均價格與獲利，「群體式學習」卻會導致零售商價格競爭之型態轉為劇烈，包括獲利領先轉換方式由漸進轉為瀑布，領先方式從勢均力敵轉為大幅領先，領先互換的頻率由低轉為高。另外，消費者購買決策之理性程度對零售商端競爭形態有影響，不論在何種供應商行為下，高理性購買決策在群體式學習下將導致零售商端價格競爭較激烈，在自我式學習下卻導致零售商端競爭行為較緩和。 研究發現二，市場中供應商的價格競爭行為會對零售商端的價格、獲利與競爭型態造成衝擊。供應商呈現價格競爭行為下，在「貨比三家無學習」之消費者行為市場中，將減緩零售商價格競爭，使零售商端之平均價格及獲利提高。在「自我」與「群體式」學習消費者市場中，將增強零售商價格競爭強度，使零售商端之平均價格及獲利降低。 研究發現三，不同的競爭市場中，零售商之最佳價格調整策略也將不同。基本上在供應商無競爭行為下，無論消費者呈現何種行為，零售商採取開放式價格調整策略具有明顯優勢。在供應商呈現競爭行為下，開放式價格調整策略在「無學習」及「群體式學習高理性程度」行為市場仍為優勝策略，在「自我式學習」及「群體式學習低理性程度」下，保守型價格調整策略則表現較佳。 在實務意涵上，若零售商可使消費者行為偏向自我或群體式學習，並穩定供應商價格競爭下，整體而言零售商端競爭可獲得最高的獲利，若當此刻競爭零售商採取保守型價格策略，而本身採取開放式價格調整策略，則獲利最大。然而面臨群體式學習消費者，由於競爭強度的增加，需留意市場動態，須隨時靈活調整本身價格策略，避免因價格策略的僵化，而成為虧損之零售商。 / The pricing competitive model traditionally assumes that consumers will buy from the firm selling the homogeneous product at the lowest price, thus discarding any possibility of learning behavior on the demand side. But if, as in real competition, consumers learn adaptively and competition is a dynamic process, then some attention should be paid to consumers' behavior. In a multiple supplier – multiple retailer supply chain, multiple price competitive forces interact to influence firm price decisions. These forces include: (1) the supplier level competition each supplier faces from others producing the same product, (2) the retailer level competition among the retailers selling the same set of goods, and (3) the vertical interaction competition between the retailer and supplier. We are interest in these questions: How does the consumer learning behavior affect the retailer pricing competitive model? How does the competition of supplier affect the retailer pricing competitive model? What is the optimal adaptive pricing strategy for retailer performance in such competitive market including retailers, suppliers and consumers. Therefore, this research study a version of the pricing competitive (Bertrand) model in which consumer exhibit dynamic adaptive learning behavior when deciding from what retailers they will buy. And we consider to join the supplier competitive pricing behavior into the retailer pricing competitive model and formulate their interaction as evolutional game and to analyze the competition of supplier effect and its impact on the pricing competition of retailers. This research uses a complex adaptive system perspective to construct a retailer pricing competitive model which considers both competitive supplier and learning consumer behavior. Using agent-based modeling and simulation (ABMS) to construct the competitive market include retailers, suppliers and consumers, and use the fuzzy logic, genetic algorithms to model the pricing decision and learning behavior of retailers and suppliers, and use reinforcement learning and swarm algorithms to model consumers’ learning behavior. The simulation results demonstrate that: The retailer level obtains the highest profit when the consumer behavior following reinforcement learning. When the consumer behavior displays swarm learning, the retailer level also obtains high profit near the highest profit. However swarm learning increases the competitive intensity on the retailer level. The competitive supplier increases the competitive intensity and decrease profit on the retailer level when the consumer behavior displays reinforcement learning and swarm learning. The performance of retailer following a closed adaptive pricing strategy (high exploitation low exploration) exceeds that of retailer following an open adaptive pricing strategy (low exploitation high exploration) when the consumer behavior displays reinforcement learning and supplier display competitive behavior. However when the consumer behavior displays swarm learning and supplier display competitive behavior, the performance of retailer following an open adaptive pricing strategy exceeds that of retailer following a closed adaptive pricing strategy. The proposed pricing competitive model with adaptive learning of consumer behavior and competition of supplier can help retailers to analyze pricing strategy and further discovery and design the more optimal pricing strategy.

Infrastructure is presenting significant national and global challenges. Whilst often seen as performing well, infrastructure tends to do so against only limited terms of reference and short-term objectives. Given that the world is facing a new infrastructure bill of ~£40T, improving the benefits delivered by existing infrastructure is vital (Dobbs et al., 2013). This thesis investigates strategic intent and the management of infrastructure systems; how factors such as organisational structure and business practice affect outcomes and the ways in which those systems — not projects — are managed. To date, performance has largely been approached from the perspective of project investment and/or delivery, or the assessment of latent failures arising from specific shocks or disruptive events (e.g. natural disaster, infrastructure failures, climate change). By contrast, the delivery of system-level services and outcomes across the infrastructure system has been rarely examined. This is where infrastructure forms an enduring system of services, assets, projects, and networks each at different stages of their lifecycle, and affecting one another as they develop, then age. Yet system performance, which also includes societal, organisational, administrative and technical factors, is arguably the level relevant to, and the reality of, day-to-day public infrastructure management. This research firstly investigated industry perceptions in order to test and confirm the problem: the nub of which was the inability to fully deliver appropriate and relevant infrastructure outcomes over the long term. Three detailed studies then explored the reasons for this problem through different lenses; thereby providing an evidence-base for a range of issues that are shared by the wider infrastructure industry. In confirming its hypothesis that “the strategic intent and the day-to-day management of infrastructure systems are often misaligned, with negative consequences for achieving the desired long-term infrastructure system outcomes”, this research has increased our understanding of the ways in which that misalignment occurs, and the consequences that result. It found those consequences were material, and frequently not visible within the sub-system accountable for the delivery of those outcomes. That public infrastructure exists, not in its own right, but to be of benefit to society, is a central theme drawn from the definition of infrastructure itself. This research shows that it is not enough to be focused on technical outcomes. Infrastructure needs to move beyond how society interacts with an asset, to the outcomes that reflect the needs, beliefs, and choices of society as well as its ability to respond to change (aptitude). Although the research has confirmed its hypothesis and three supporting propositions, the research does not purport to offer ‘the solution’. Single solutions do not exist to address the challenges facing a complex adaptive system such as infrastructure. But the research does offer several system-oriented sense-making models at both the detailed and system-level. This includes the probing methodology by way of a diagnostic roadmap. These models aim to assist practitioners in managing the transition of projects, assets, and services into a wider infrastructure system, their potential, and in (re)orienting the organisation to the dynamic nature of the system and its societal imperative.