Agent-based Modeling Approach Research Articles

Knowledge sharing in a dynamic, multi-level organization: an agent-based modeling approach

Knowledge sharing in a dynamic, multi-level organization: an agent-based modeling approach

Evaluating Peer-to-Peer Electricity Markets across the U.S. Using an Agent-Based Modeling Approach

The diffusion of distributed energy resources can overcome some challenges associated with the historical centralized model of electric power distribution. Decentralized generation by residential solar photovoltaic cells creates the potential for peer-to-peer (P2P) electricity trading, where households can act as consumers and prosumers to buy and sell renewable electricity. P2P energy markets are emerging at locations across the globe, and market performance is affected by various social, economic, and environmental factors. This research applies an agent-based modeling (ABM) framework to simulate electricity trades between heterogeneous households in a decentralized market. The P2P system is tested for 15 locations in the United States that vary in climate parameters and local economic factors. The results from these simulations are compared to assess how differences in climate, demand pattern, retail rate, and irradiance affect market performance. Simulations demonstrate that market outcomes rely on the ratio of prosumers to consumers, environmental factors, and geographic conditions. Battery energy storage overcomes limitations associated with faulty forecasting and improves the flexibility of household-generated solar resources to increase the proportion of production that is sold in the P2P market. The application of the agent-based modeling framework demonstrates how P2P markets can be expected to perform for various locations and can be applied to assess alternative locations for market performance.

Examining construction group’s safety attitude resilience under major disruptions: An agent-based modelling approach

The construction industry is undergoing a series of disruptions and industrial changes. In disturbances of major events, construction workers’ safety attitude is at risk of being sabotaged due to factors such as resource reallocations, job insecurity, and mental pressure. The extend of such impact on the safety attitude by the disruptive event is complicated and complex as it depends on individual, group and external factors. This research extended and applied successfully the resilience engineering theory to model the response-recovery process of individual workers’ safety attitude under disruptions. With data collected from questionnaire surveys, an agent-based model was built to describe the dynamic changes of safety attitude under disruptions. The model examined the influence of factors at individual, group and external environmental aspects on both response and recovery stages. The results showed that construction workers’ safety attitude was resilient under external disruptions, and group and external level factors determined the length of recovery process, while individual factors determined if the safety attitude could be eventually fully recovered. When the safety parameters decreased by 28%, the construction group’s average safety attitude would not fully recover, unless compensated by the learning adaptability (LA) increasing by 10%. Furthermore, the results revealed that a high level of group interactions (doubled active-ratio) absorbed the external disruptions and the influence on construction workers with high degrees (doubled centre-value) had greater impact on the overall safety attitude resilience. This research shed lights on developing managerial suggestions to cope external disruptions and maintain the positive and stable safety attitude.

Modeling social interaction and metapopulation mobility of the COVID-19 pandemic in main cities of highly populated Java Island, Indonesia: An agent-based modeling approach

IntroductionCoronavirus transmission is strongly influenced by human mobilities and interactions within and between different geographical regions. Human mobility within and between cities is motivated by several factors, including employment, cultural-driven, holidays, and daily routines.MethodWe developed a sustained metapopulation (SAMPAN) model, an agent-based model (ABM) for simulating the effect of individual mobility and interaction behavior on the spreading of COVID-19 viruses across main cities on Java Island, Indonesia. The model considers social classes and social mixing affecting the mobility and interaction behavior within a sub-population of a city in the early pandemic. Travelers’ behavior represents the mobility among cities from central cities to other cities and commuting behavior from the surrounding area of each city.ResultsLocal sensitivity analysis using one factor at a time was performed to test the SAMPAN model, and we have identified critical parameters for the model. While validation was carried out for the Jakarta area, we are confident in implementing the model for a larger area with the concept of metapopulation dynamics. We included the area of Bogor, Depok, Bekasi, Bandung, Semarang, Surakarta, Yogyakarta, Surabaya, and Malang cities which have important roles in the COVID-19 pandemic spreading on this island.DiscussionOur SAMPAN model can simulate various waves during the first year of the pandemic caused by various phenomena of large social mobilities and interactions, particularly during religious occasions and long holidays.

Simulating rumor spreading and rebuttal strategy with rebuttal forgetting: An agent-based modeling approach

Rumor is a typical social psychological phenomenon, and forgetting is a universal psychological trait. Exploring the role of forgetting will improve the understanding of rumor spreading and its control. Based on the classical SIR epidemic framework, the conceptual model of rumor spreading and rebuttal strategy is proposed with consideration of forgetting, and the mathematical model at the individual level is constructed from a competitive innovation diffusion perspective. A Java-based Agent-based Modeling approach is adopted to conduct comprehensive simulation experiments. Results indicate that both rebuttal strategy and rebuttal forgetting would affect the process and the result of rumor spreading and control, and complex interactions exist among different factors. Although the influence of control strategy is dominant in most cases, more active control might be not optimal in extreme cases because of the moderation of forgetting, the nonlinearity of influence increases the control challenge.

A coupled agent-based model to analyse human-drought feedbacks for agropastoralists in dryland regions

Drought is a persistent hazard that impacts the environment, people's livelihoods, access to education and food security. Adaptation choices made by people can influence the propagation of this drought hazard. However, few drought models incorporate adaptive behavior and feedbacks between adaptations and drought. In this research, we present a dynamic drought adaptation modeling framework, ADOPT-AP, which combines socio-hydrological and agent-based modeling approaches. This approach is applied to agropastoral communities in dryland regions in Kenya. We couple the spatially explicit hydrological Dryland Water Partitioning (DRYP) model with a behavioral model capable of simulating different bounded rational behavioral theories (ADOPT). The results demonstrate that agropastoralists respond differently to drought due to differences in (perceptions of) their hydrological environment. Downstream communities are impacted more heavily and implement more short-term adaptation measures than upstream communities in the same catchment. Additional drivers of drought adaptation concern socio-economic factors such as wealth and distance to wells. We show that the uptake of drought adaptation influences soil moisture (positively through irrigation) and groundwater (negatively through abstraction) and, thus, the drought propagation through the hydrological cycle.

Simulating the Diffusion of Residential Rooftop Photovoltaic, Battery Storage Systems and Electric Cars in Italy. An Exploratory Study Combining a Discrete Choice and Agent-Based Modelling Approach

Rooftop solar photovoltaic (PV) systems could significantly contribute to renewable energy production and reduce domestic energy costs. In Italy, as in other countries, the current incentives generate a modest annual increase after the generous fiscal incentives that kick-started the PV market in the 2008–2013 period. Several factors are, however, at play that can speed up the installation process, such as the improvements in PV technology at declining prices, the increased availability of battery-storage (BS) systems, the growing use of electric appliances, the uptake of electric cars, and the increased environmental awareness. We integrate two research methodologies, discrete choice modeling and agent-based modeling, to understand how these factors will influence households’ decisions regarding PV and BS installations and how agents interact in their socioeconomic environment. We predict that in Italy, given the preference structure of homeowners, the continuing decline in costs, and the social interaction, 40–45% of homeowners will have PV or PV and BS installed by 2030, thanks to the existing investment tax credit policy.

Unraveling the Most Influential Determinants of Residential Segregation in Jakarta: A Spatial Agent-Based Modeling and Simulation Approach

This study involves the analysis of the residential segregation patterns in Jakarta, Indonesia, one of the largest global metropolitan cities. Our objective is to determine whether similarities in religion or socioeconomic status are more dominant in shaping residential segregation patterns in Jakarta. To do so, we extended Schelling’s segregation agent-based model incorporating the random discrete utility choice approach to simulate the relocation decisions of the inhabitants. Utilizing actual census data from the 2010–2013 time period and the Jakarta GIS map, we simulated the relocation movements of the inhabitants at the subdistrict level. We set the inhabitants’ socioeconomic and religious similarities as the independent variables and the housing constraints as the moderating variable. The segregation parameters of the inhabitants (i.e., dissimilarity and Simpson indexes) and the spatial patterns of residential segregation (i.e., Moran index and segregation maps) were set as the dependent variables. Additionally, we further validated the simulation outcomes for various scenarios and contrasted them with their actual empirical values. This study concludes that religious similarity is more dominant than socioeconomic status similarity in shaping residential segregation patterns in Jakarta.

Economic and Labour Market Impacts of Migration in Austria: An Agent-Based Modelling Approach

Economic and Labour Market Impacts of Migration in Austria: An Agent-Based Modelling Approach

Simulating the Effects of Law on Patent Licensing Process: Agent-Based Modeling Approach

Simulating the Effects of Law on Patent Licensing Process: Agent-Based Modeling Approach

Social Mixing and Network Characteristics of COVID-19 Patients Before and After Widespread Interventions: A Population-based Study.

Abstract SARS-CoV-2 rapidly spreads among humans via social networks, with social mixing and network characteristics potentially facilitating transmission. However, limited data on topological structural features has hindered in-depth studies. Existing research is based on snapshot analyses, preventing temporal investigations of network changes. Comparing network characteristics over time offers additional insights into transmission dynamics. We examined confirmed COVID-19 patients from an eastern Chinese province, analyzing social mixing and network characteristics using transmission network topology before and after widespread interventions. Between the two time periods, the percentage of singleton networks increased from 38.9 $ \% $ to 62.8 $ \% $ $ (p<0.001) $ ; the average shortest path length decreased from 1.53 to 1.14 $ (p<0.001) $ ; the average betweenness reduced from 0.65 to 0.11 $ (p<0.001) $ ; the average cluster size dropped from 4.05 to 2.72 $ (p=0.004) $ ; and the out-degree had a slight but nonsignificant decline from 0.75 to 0.63 $ (p=0.099). $ Results show that nonpharmaceutical interventions effectively disrupted transmission networks, preventing further disease spread. Additionally, we found that the networks’ dynamic structure provided more information than solely examining infection curves after applying descriptive and agent-based modeling approaches. In summary, we investigated social mixing and network characteristics of COVID-19 patients during different pandemic stages, revealing transmission network heterogeneities.

Имитационное моделирование движения поезда по перегону с применением тяговых расчётов

Within the framework of the conducted research, the necessity for developing a highly detailed simulation model of interval regulation of train traffic is substantiated. The advantages and disadvantages of current universal and specialized programs of simulation modeling of processes in railway transport are analyzed. By means of the universal simulation tool AnyLogic, the model of train movement along the running line is constructed using traction calculations and implementation of interval regulation. In the model, the agent-based modeling approach is used to simulate train movement modes, the process approach is used to implement the logic of the operation of interval control systems, and the elements of system dynamics are used to calculate the speed of the train. The adequacy of the model is justified by analyzing the train travel times in traction mode along the lines belonging to the four types of track profile with varying the mass of rolling stock and the power of train locomotives.

Evaluating synchromodal transport with agent-based simulation

Synchromodal transport implies using dynamically several modes of transport. Two relevant aspects of synchromodality are collaboration between actors and flexibility, aiming to make as many decisions as possible in real-time, imposing several changes on the logistic system's performance. This study aims to model the interactions between actors and their reactions to disruptions within the transport network. Using an agent-based modelling approach, the system is modelled as a collection of heterogeneous agents with complex interactions. Different collaboration scenarios and reaction strategies are modeled. The result shows improvement in the average cost, as the collaboration between logistics service providers and the flexibility of their response to disruptions increases. This model provides a testbed for stakeholders to assess different scenarios and ideas inexpensively and risk-free.

Optimisation of Electric Vehicle Battery Size

Energy storage and battery technologies have taken centre stage in the race to meet the UK government target to ban new petrol and diesel cars by 2030. However, underlying key issues such as resource demand and negative public opinion must be solved before the high uptake of electric vehicles. The research conducted in this paper proposed viable solutions to these challenges through modelling of real driver data utilising an agent based modelling approach. Per month state of charge analysis confirmed that the current charging infrastructure in circulation will not accommodate the miniaturisation of electric vehicle battery size. Thus, an improved alternative charging infrastructure was proposed, which enabled the optimal battery size to be reduced by up to 40%. The users stop times were analysed to assign an optimal battery size based upon monthly driving behaviour concluding daily inner city drivers require a 30kWh battery and daily long distance drivers require a 40kWh battery. When decreasing the battery size by the proposed 40% there is a £2650.60 saving and a 6.4kg lithium demand decrease per battery when compared to the current average battery size.

Agent-Based Modeling of Malaria Transmission

Despite global efforts to eradicate malaria, it remains a major health threat to nearly half of the world’s population. Recent statistics show that globally, there are over 200 million malaria cases, and estimated deaths are close to half a million. Africa alone accounts for almost 90% of these cases. Numerous studies have been conducted to understand the transmission dynamics of malaria. Mathematical methods have been widely used to model and understand disease dynamics and outbreak patterns. Although mathematical methods have provided good results for homogeneous populations, they impose significant limitations for investigating certain heterogeneities in a population, such as age and vaccine status. This paper proposes an agent-based modeling approach that permits the capture of heterogeneous mixing and agent interactions, thus enabling a better understanding of malaria dynamics and outbreak patterns. Our approach is illustrated in a case study using climate and demographic data of the cities of Tripura, Limpopo, and Benin. Our agent-based simulation was validated against the reported cases of malaria collected in these cities. Furthermore, the efficiency of the proposed model was compared with the mathematical models used as a benchmark. A statistical test confirmed that the agent-based model is robust and has the potential to accurately predict the peak seasons of malaria. These results can help healthcare providers and policymakers have intervention mechanisms in place in advance, which can potentially help reduce the malaria transmission rate.

An agent-based model approach for simulating e-scooter routing

Modeling of Electric scooters (e-scooters) and prediction of their impacts in the transport system is challenging due to their dual drive behavior i.e., from vehicles’ roadway to pedestrians’ sidewalk. This study aims to develop a model to integrate e-scooter routing process in an agent-based model, namely: MATSim. The developed model is built on the hypothesis that perceived safety affects travel behavior of e-scooter users and is related to road environment. By performing a network analysis, links are classified in seven safety levels; these classifications are later introduced in the simulation process where agents balance safety with time and distance to make choices. Four network scenarios are simulated. The simulation outputs show that there are some distance thresholds for which e-scooter users are not willing to travel longer even if safety levels for such routes are higher.

Effect of global warming on Western Mediterranean seagrasses: a preliminary agent-based modelling approach

The Mediterranean Sea is warming more rapidly than the global average, leading to dire consequences for its inhabiting organisms. Seagrasses are key structural elements in coastal ecosystems, and studying how temperature affects these species is crucial to anticipate the implications of global warming. In this work, we use an empirically based numerical model to study the combined dynamics of Posidonia oceanica and Cymodocea nodosa and their resilience to sea warming. The model is parametrised using seagrass growth rates measured in the Western Mediterranean Sea. Under favourable growth conditions, our simulations predict the emergence of a coexistence region at the front between monospecific meadows. This region can be characterised by its width and local shoot densities, which depend on the coupling parameter between P. oceanica and C. nodosa. Such regions have been empirically observed in Ses Olles de Son Saura (Balearic Islands). A comparison between the field measurements at the study site with the model predictions was used to fit the value of the coupling parameter. Field data also relate the width of the coexistence region to the average length of P. oceanica leaves at the front. Remarkably, a linear relationship was found between the coupling parameter and leaf length. In the presence of sea warming, the model predicts an exponential decay in the population of P. oceanica, which is highly sensitive to temperature. This behaviour is a direct consequence of the clonal nature of the plant and can be characterised by the model parameters. Considering a scenario of high greenhouse emissions, our model forecasts that P. oceanica meadows will experience a 70% population decline by the year 2050. C. nodosa, with higher thermal resilience, acts as an opportunistic species and will colonise the space left by the degraded P. oceanica.

A framework proposal for machine learning-driven agent-based models through a case study analysis

Agent-based modeling (ABM) has been widely employed by researchers in various domains. Developing valid and useful agent-based models (ABMs) imposes challenges on the modelers. Using machine learning (ML) techniques in ABMs may facilitate the development of these models and improve their performance. This paper provides a detailed overview of the relationship between ML and ABM approaches. The benefits and drawbacks of data-driven ABMs are evaluated. A main scheme for utilizing ML techniques in ABMs is provided and explored through references to the relevant studies. As part of the primary scheme, a framework for modeling agent behaviors in ABMs utilizing ML approaches is proposed. In the framework, theoretical support is also combined with ML approaches in order to increase the accuracy of agent behavior generated by ML approaches. Using the suggested framework, a real-world case study is performed to investigate the application of ML techniques to improve the accuracy of ABMs and facilitate their creation. The findings indicate that ML approaches may facilitate the construction of ABMs.

A multiscale mathematical model describing the growth and development of bambara groundnut

A principal objective in agriculture is to maximise food production; this is particularly relevant with the added demands of an ever increasing population, coupled with the unpredictability that climate change brings. Further improvements in productivity can only be achieved with an increased understanding of plant and crop processes. In this respect, mathematical modelling of plants and crops plays an important role. In this paper we present a two-scale mathematical model of crop yield that accounts for plant growth and canopy interactions. A system of nonlinear ordinary differential equations (ODEs) is formulated to describe the growth of each individual plant, where equations are coupled via a term that describes plant competition via canopy–canopy interactions. A crop of greenhouse plants is then modelled via an agent based modelling approach in which the growth of each plant is described via our system of ODEs. The model is formulated for the African drought tolerant legume bambara groundnut (Vigna subterranea), which is currently being investigated as a food source in light of climate change and food insecurity challenges. Our model allows us to account for plant diversity and also investigate the effect of individual plant traits (e.g. plant canopy size and planting distance) on the yield of the overall crop. Informed with greenhouse data, model results show that plant positioning relative to other plants has a large impact on individual plant yield. Variation in physiological plant traits from genetic diversity and the environmental effects lead to experimentally observed variations in crop yield. These traits include plant height, plant carrying capacity, leaf accumulation rate and canopy spread. Of these traits plant height and ground cover growth rates are found to have the greatest impact on crop yield. We also consider a range of different planting arrangements (uniform grid, staggered grid, circular rings and random allocation) and find that the staggered grid leads to the greatest crop yield (6% more compared to uniform grid). Whilst formulated specifically for bambara groundnut, the generic formulation of our model means that with changes to certain parameter’s, it may be extended to other crop species that form a canopy.

Detecting Multiple Perturbations on Swine using Data from Simulation of Precision Feeding Systems

Industry 4.0 brings transformation to all industries, including agriculture. Smart livestock has been replacing traditional livestock as a trend of the agricultural industry in the world. Precision feeding is one of the areas of smart husbandry that combines many modern multidisciplinary technologies which are prominent such as AI, IoT, Big Data, etc. To obtain that for pigs, a precision feeding system needs to be implemented. Components of the system include automatic feeders connected to a computer system to collect and process data on daily feed intake of fishes and animals, and/or from ambient sensors. Perturbations such as heat stress or sanitation issues have a significant impact on the nutritional profile of group housed pigs. However, perturbation is often detected only after it has occurred and is recognized late by the consequences left on the animal. Although the cause of perturbations might be unknown, the effect on the animal can be observed early throughout the data of voluntary feed intake. By the precision feeding system, the data are processed and analysed based on mathematical models following a two-step approach: (1) estimation of target trajectory of cumulative feed intake using linear and quadratic functions, and (2) detection of perturbations based on deviations from the target cumulative feed intake. However, implementing such a system requires huge costs and is often beyond the capabilities of farms, production households and small/medium laboratories. In this paper, we introduce an agent-based modeling approach to simulate precision feeding systems for swine, whose data can be used to early detect multiple perturbations which may have appeared. Experiments were carried out on GAMA simulation platform to demonstrate the efficiency in detecting multiple perturbations of group housed pigs, and also prove the usefulness of simulation of precision feeding systems.