Punishment Mechanism Research Articles

Multi-Agent System Anti-Interference Algorithm Based on Deep Reinforcement Learning

As multi-agent systems are increasingly applied in various environments, achieving stable control under communication interference has become a key challenge. This paper proposes a stability control algorithm based on a Markov model and reinforcement learning, which enables the multi-agent system to achieve its goals through a reward and punishment mechanism under interference. The results demonstrate that the algorithm effectively handles communication interference, converges well, and ultimately completes control tasks, showing promising application prospects for the proposed solution.

Application of Reinforcement Learning in Controlling Quadrotor UAV Flight Actions

Most literature has extensively discussed reinforcement learning (RL) for controlling rotorcraft drones during flight for traversal tasks. However, most studies lack adequate details regarding the design of reward and punishment mechanisms, and there is a limited exploration of the feasibility of applying reinforcement learning in actual flight control following simulation experiments. Consequently, this study focuses on the exploration of reward and punishment design and state input for RL. The simulation environment is constructed using AirSim and Unreal Engine, with onboard camera footage serving as the state input for reinforcement learning. The research investigates three RL algorithms suitable for discrete action training. The Deep Q Network (DQN), Advantage Actor–Critic (A2C), and Proximal Policy Optimization (PPO) were combined with three different reward and punishment design mechanisms for training and testing. The results indicate that employing the PPO algorithm along with a continuous return method as the reward mechanism allows for effective convergence during the training process, achieving a target traversal rate of 71% in the testing environment. Furthermore, this study proposes integrating the YOLOv7-tiny object detection (OD) system to assess the applicability of reinforcement learning in real-world settings. Unifying the state inputs of simulated and OD environments and replacing the original simulated image inputs with a maximum dual-target approach, the experimental simulation achieved a target traversal rate of 52% ultimately. In summary, this research formulates a set of logical frameworks for an RL reward and punishment design deployed with real-time Yolo’s OD implementation synergized as a useful aid for related RL studies.

Revolutionizing automated pear picking using Mamba architecture

With the emergence of the new generation vision architecture Vmamba and the further demand for agricultural yield and efficiency, we propose an efficient and high-accuracy target detection network for automated pear picking tasks based on Vmamba, aiming to address the issue of low efficiency in current Transformer architectures. The proposed network, named SRSMamba, employs a Reward and Punishment Mechanism (RPM) to focus on important information while minimizing redundancy interference. It utilizes 3D Selective Scan (SS3D) to extend scanning dimensions and integrates global information across channel dimensions, thereby enhancing the model's robustness in complex agricultural environments and effectively adapting to the extraction of complex features in pear orchards and farmlands. Additionally, a Stacked Feature Pyramid Network (SFPN) is introduced to enhance semantic information during the feature fusion stage, particularly improving the detection capability for small targets. Experimental results show that SRSMamba has a low parameter count of 21.1 M, GFLOPs of 50.4, mAP of 72.0%, mAP50 reaching 94.8%, mAP75 at 68.1%, and FPS at 26.9. Compared with other state-of-the-art (SOTA) object detection methods, it achieves a good trade-off between model efficiency and detection accuracy.

How to Promote Data Sharing of Financial Service Providers Based on Federated Learning Technology

Data sharing between financial service providers (FSPs) can promote the reduction of financial risks, but the reality is that the FSPs may be unwilling to share data with each other due to data security problems. The emergence of federated data technology provides a convenient way to address this issue. In this article, the author proposes using a federated data sharing model for FSPs and two dynamic evolutionary games for evaluating influencing factors. Cooperation results when both FSPs reach a certain level of mutual payoff. The reward and punishment mechanisms are important means by which FSPs actively implement real data sharing. The interaction between the FSP and the federated technology service provider (FTSP) demonstrates the practicality of this new technology. The results show that the data sharing improved the data trust and supported regulatory mechanisms.

Evolution of corporate carbon information disclosure considering the reward and punishment mechanism and new media environment

This study examines the internal logic and dynamic evolution of corporate carbon information disclosure decision making. Drawing on the game-theoretical relationship among enterprises, third-party carbon verification agencies, and local governments, we analyze the intrinsic mechanisms by which rewards, punishment, and media supervision influence corporate carbon information disclosure behavior. We further examine each participant's strategy selection process in the corporate carbon information disclosure decision making under different conditions and consider the impact of key factors on the process. The findings show that the best evolutionary stabilization strategy consists of true disclosure, compliance, and negative support. A stronger reward and punishment mechanism of carbon information disclosure helps achieve greater self-regulation and reduce false disclosure. Enterprises are increasingly choosing true disclosure as the probability of media exposure increases. Decision makers involved in the carbon information disclosure process are influenced by image loss, disclosure costs, and government regulatory costs.

Taxation Can Significantly Enhance The Effectiveness of Reward and Punishment Mechanisms in The N-Person Snowdrift Game

We extend the reward and punishment mechanisms in the N-person evolutionary snowdrift game (NSG) model by constructing a tax-based pure punishment NSG model and a tax-based pure reward NSG model. Using replicator dynamics, we established two sets of dynamic equations that describe the evolution of the frequencies of the three strategies. Over longer time scales, both systems exhibit stable states involving two strategies, with the exclusion of the third. Introducing tax factors into the reward and punishment mechanisms significantly increases the proportion of cooperators in the steady state and eliminates defectors, with the time to reach the steady state depending on the tax rate. Moreover, we found that tax-based pure reward strategies are more conducive to fostering cooperative behavior within the system compared to tax-based pure punishment strategies. We employed numerical algorithms to simulate the replicator dynamics, and the results of the dynamic equations were in perfect agreement with the numerical simulations.

Threshold incentive mechanisms for the sustainable management of public resources.

Maintaining global ecological balance is inextricably linked to the effective governance of public resources. Previous studies have proposed the introduction of incentive mechanisms to mitigate the defection behavior within the scope of public resource management to facilitate sustainable utilization. However, these incentive strategies often present a considerable financial burden. Here, we first propose a two-threshold incentive scheme based on group state objectives, where the reward mechanism is triggered when the number of cooperators is below the threshold M, and the punishment mechanism is triggered when the number of cooperators is above the threshold Q. We further introduce an incentive mechanism by considering both resource and group thresholds, where incentives are implemented separately when the number of cooperators and the current quantity of resources meet thresholds. Our theoretical results reveal that introducing these two new mechanisms can effectively maintain cooperative behavior and sustainable resources while reducing management costs. However, if the initial resource level is low, the tragedy of the commons will still occur. It is recommended that the number of developers be reduced, the allocation be lowered, or the development simply suspended and the natural regeneration of resources be waited for.

Evolutionary game and simulation study of public transport under government incentive and punishment mechanism.

Public transport plays an indispensable role in the whole public transport system. This paper makes an in-depth study on how public transport can provide passengers with higher service quality while meeting the needs of passengers. In order to achieve this research goal, this paper organically incorporates the three key subjects of government supervision, public transport and passengers into the research framework. Evolutionary game theory is used to construct the corresponding research model. It has been found that the decision-making behaviours of government regulators, public transport and passengers are intricately intertwined to influence each other in the evolutionary process. It is particularly noteworthy that the incentive or punishment measures adopted by the government have a great impact on the quality of public transport services. In addition, timely supervision and inspection of government regulatory authorities by higher authorities proved to be crucial for buses to provide stable and high-quality services. This study reveals the mechanisms of interaction between different subjects in the public transport system, particularly the government-guided incentive measures and supervision mechanism to promote the overall service level. To further support the research conclusions, this paper carries on the simulation analysis, and puts forward the countermeasures and suggestions for the bus to provide high-quality service according to the simulation results. These recommendations will help guide the government, public transport and passengers to make better decisions in the synergistic development process, thereby improving the overall level of service.

Governance structure, organizational form, and business performance: A study of the Shanxi piaohao (banks), 1820s–1930s

This paper studies Shanxi piaohao, China’s first and largest private remittance banking group, established by a group of merchants from Shanxi province. Drawing on extensive primary sources, our study offers the first empirical analysis of the piaohao’s business organisation and performance. We show that piaohao thrived but were also constrained by their own private rules. In their earlier development, they successfully overcame the commitment problem by implementing a townsmen-guarantor-based recruitment system, a centralised M-form structure, and a multilateral punishment mechanism. To align the interests of shareholders and employees, they created employment shareholders and offered a competitive compensation scheme. However, piaohao’s governance structure became obsolete as shareholders increasingly failed to monitor head managers in the later period. A drastic increase in business volume and a changing business climate also exposed the defects of a centralised structure. As modern Chinese banks rose to prominence, the piaohao were gradually outcompeted and declined.

An evolutionary game analysis of the prefabricated building supply side under dynamic government reward and punishment mechanisms

From the standpoint of the supply side of prefabricated buildings, an evolutionary game model between the government and developers is constructed under the static and dynamic reward and punishment mechanisms in order to promote the development of prefabricated buildings and realize the transformation and upgrading of the construction industry. The simulation study of the example is conducted, and the impact of policy cost, subsidy upper limit, and tax upper limit on the system’s evolutionary stability strategy is examined. The findings demonstrate that when the government chooses static subsidy and static tax policies, there is no evolutionary stable strategy in the game system. The flaws of the static reward and punishment system can be adequately compensated for by the dynamic subsidy and dynamic tax policy, leading to an evolutionary stable state. The likelihood of developers producing prefabricated structures is negatively connected with policy cost and subsidy upper limits and favorably correlated with tax upper limits, according to the dynamic reward and punishment mechanism. The findings of the study have significant implications for government policy that will encourage the widespread use of prefabricated structures in China.

Tripartite evolutionary game analysis of government reward and punishment mechanisms and agricultural resilience under a dual carbon context

PurposeUnder the “dual carbon” framework, the article explores the equilibrium points among the government, agricultural enterprises and village committees, and uses sensitivity analysis to reveal the dynamic factors affecting these stakeholders, thereby proposing methods to enhance agricultural disaster resilience.Design/methodology/approachThe article uses MATLAB to construct a game model for the three parties with interests: agribusiness, government and village council. It examines the stability of strategies among these entities. Through graphical simulation, the paper analyzes the sensitivity of agricultural enterprises carbon emissions and village committees’ rent-seeking behaviors in the decision-making process, focusing on significant factors such as government carbon tax and regulatory policies.FindingsA single government reward and punishment mechanism is insufficient to influence the strategic choices of enterprises and village committees. The cost of rent-seeking does not affect the strategic choices of enterprises and village committees. A key factor influencing whether the village committee engages in rent-seeking is the level of labor income of the village committee as an “intermediary”.Originality/valueThis paper focuses on the dynamic game between three stakeholders (the government, agricultural enterprises and village committees), seeking dynamic equilibrium and conducting sensitivity analysis through visualization to provide the government with optimal policy recommendations.

Render help or stand by? The effect of group size on third-party punishment and its neural mechanisms

Third-party punishment (TPP) is the punishment that an individual executes on a violator as a third party or observer to maintain social norms. Many studies have provided insights into the neural mechanisms of third-party punishment in group environments. Still, only some studies have focused on the neural mechanisms of third-party punishment in different group sizes. This study used EEG analysis to explore the effects of group size on third-party punishment and its neural activity characteristics from the context of gain and loss. The results show that the punishment rate and amount of the third party in the small group size and loss context were significantly higher than that in the large group size and gain context. EEG results showed that third-party punishment in small groups induced greater P2 than in large groups. In the loss context, the third-party punishment in the large group size induced more negative LNP and activated more theta band activation than in the small group. The results showed that the motivation of the third party to seek a positive reputation in the small size exceeds the balance of its economic interests and tends to punish the violator for maintaining fair norms. The loss context plays a promoting role in this process. However, in the large size, the third-party consideration of its interests was stronger than the willingness to maintain social norms. This study provided neuroscientific evidence for third-party punishment to maintain fair norms in a group environment and further explanations from neuroscience for understanding Indirect Reciprocity Theory.

Evolutionary Game and Simulation Analysis of New-Energy Vehicle Promotion in China Based on Reward and Punishment Mechanisms

In China, new-energy vehicles are viewed as the ultimate goal for the automobile industry, given the current focus on the “dual-carbon” target. Therefore, it is important to promote the sustainable development of this new-energy market and ensure a smooth transition from fuel-driven vehicles to new-energy vehicles. This study constructs a tripartite evolutionary game model involving vehicle enterprises, consumers, and the government. It improves the tripartite evolutionary game through the mechanisms of dynamic and static rewards and punishments, respectively, using real-world data. The results show the following. (1) A fluctuation is present in the sales of new-energy vehicles by enterprises and the active promotional behavior of the government. This fluctuation leads to instability, and the behavior is difficult to accurately predict, which is not conducive new-energy vehicles’ promotion and sales. (2) A static reward and punishment mechanism can change the fluctuation threshold or peak value. Nevertheless, the stability of the system’s strategy is not the main reason that the government has been actively promoting it for a long time. However, enterprises are still wavering between new-energy and fuel vehicles. (3) The linear dynamic reward and punishment mechanism also has its defects. Although they are considered the stability control strategy of the system, they are still not conducive to stability. (4) The nonlinear dynamic reward and punishment mechanism can help the system to achieve the ideal stabilization strategy.

Research on the Green Transition Path of Airport Development under the Mechanism of Tripartite Evolutionary Game Model

Since existing studies primarily explore green development measures from the static perspective of a single airport stakeholder, this paper constructs an evolutionary game model to analyze the strategic choices of three key stakeholders: airport authorities, third-party organizations, and government departments, based on evolutionary game theory. By solving the stable strategy of the tripartite evolution using the Jacobian matrix, the green transition of airport development can be divided into three stages: “initiation”, “development”, and “maturity”, allowing for the exploration of key factors influencing the green transition of airport development. A simulation analysis is conducted based on real Guangzhou Baiyun International Airport data. The results indicate that the tripartite evolutionary game strategy is stable at E4(0,0,1) and the green transition of Baiyun Airport remains in the development stage. By improving the reward and punishment mechanisms of government departments, the evolutionary game strategy can be stabilized at E8(1,1,1), promoting the green transition of airport development toward the mature stage. By adjusting the game parameters, the dynamic process of green transition in airports at different levels of development and under varying regulatory environments can be effectively captured, supporting the precise formulation of corresponding policies.

Digital transformation of construction industry under government reward and punishment mechanism based on evolutionary game modeling

ABSTRACT Current research on digital transformation in the construction industry requires an in-depth analysis of the behavioral strategy choices of relevant key players. This paper bridges this gap by combining the characteristics of the construction industry and focusing on the key behavioral players in the digital transformation of the construction industry. It constructs an evolutionary game model involving the government, construction units, and construction enterprises, and examines the impact of government incentives and penalties on the strategic choices of each game participant. MATLAB is used to simulate, validate and analyze the effects of various factors on strategic choices to verify the validity of the model. The results suggest that (1) increasing incentives and subsidies for construction units and enterprises will facilitate the advancement of digital transformation in the construction industry. However, they must be kept within a specific range or they may lead to speculative behavior. (2) By setting a loss limit and penalizing local governments, higher-level governments can stimulate the digital transformation of construction enterprises. Through the analysis of the evolutionary results, several countermeasure suggestions are put forward, aiming to promote the digital transformation of the construction industry.

Research on the Game Strategy of Mutual Safety Risk Prevention and Control of Industrial Park Enterprises under Blockchain Technology

Systematic management of corporate safety risks in industrial parks has become a hot topic. And risk prevention and control mutual aid is a brand-new model in the risk and emergency management of the park. In the context of blockchain, how to incentivize enterprises to actively invest in safety risk prevention and control mutual aid has become a series of key issues facing government regulators. This paper innovatively combines Prospect Theory, Mental Accounting, and Evolutionary Game Theory to create a hypothetical model of limited rationality for the behavior of key stakeholders (core enterprises, supporting enterprises, and government regulatory departments) in mutual aid for safety risk prevention and control. Under the static prize punishment mechanism and dynamic punishment mechanism, the evolutionary stabilization strategy of stakeholders was analyzed, and numerical simulation analysis was performed through examples. The results show: (1) Mutual aid for risk prevention and control among park enterprises is influenced by various factors, including external and subjective elements, and evolves through complex evolutionary paths (e.g., reference points, value perception). (2) Government departments are increasingly implementing dynamic reward and punishment measures to address the shortcomings of static mechanisms. Government departments should dynamically adjust reward and punishment strategies, determine clearly the highest standards for rewards and punishments, and the combination of various incentives and penalties can significantly improve the effectiveness of investment decisions in mutual aid for safety risk prevention and control. (3) Continuously optimizing the design of reward and punishment mechanisms, integrating blockchain technology with management strategies to motivate enterprise participation, and leveraging participant feedback are strategies and recommendations that provide new insights for promoting active enterprise investment in mutual aid for safety risk prevention and control. The marginal contribution of this paper is to reveal the evolutionary pattern of mutual safety risk prevention and control behaviors of enterprises in chemical parks in the context of blockchain.

Study on the Problems and Countermeasures of Specially Invited Mediation’s System in China Under the Diversified Dispute Resolution Mechanisms

The Provisions of the Supreme People’s Court on Specially Invited Mediation in the People’s courts, issued in 2016, formally established the specially invited mediation’s system, but the lack of uniform standards for the rules applicable to the specially invited mediation’s system, the unsoundness of safeguard mechanism, and the inefficiency of its operation have not been effectively resolved. Therefore, the scope of application of the specially invited mediation’s system should be clarified through legislation, the entry threshold of specially invited mediators should be raised, a perfect financial guarantee system for specially invited mediation should be established, and a reward and punishment mechanism should be set up to strengthen the punishment for false mediation, preventing false mediation, and increasing the supervision of the specially invited mediation procedure, so as to promote the orderly conduct of specially invited mediation and efficiently resolve disputes.

Periodic update rule with Q-learning promotes evolution of cooperation in game transition with punishment mechanism

Cooperative behavior assumes a critical role in resolving conflicts arising between collective and individual interests, while punishment measures serve as a robust deterrent against opportunistic free-riding. Within this context, evolutionary game theory (EGT) emerges as an indispensable paradigm for addressing this multifaceted issue. When it comes to introspection behaviors, reinforcement learning (RL) methods exhibit remarkable capabilities to capture agents’ cognitive processes. Nonetheless, previous research has often focused on a static and time-invariant update rule, neglecting the dynamic nature of real-world scenarios where individuals can flexibly transit between strategies in periodic time-dependent patterns. Here, we propose periodic update rules with Q-learning algorithm and game transition model with a punishment mechanism that grants cooperative agents the autonomy to exercise discretion in deciding whether to initiate punishment actions. The agents display dynamic rules periodically through game model transitions, thus ensuring EGT’s inherent adaptability. By employing Monte Carlo (MC) simulations, we analyze the emergence of cooperation that underscores the substantial enhancement of cooperative behavior through the proposed periodic update rules with Q-learning algorithm and game transitions in the presence of punishment. Our study highlights the indispensable significance of appropriate periodic intervals for updating rules and determining optimal punishment costs in the game transition model as critical elements for fostering the evolution of cooperation in real-world scenarios.

The motive matters: Experimental evidence on the expressive function of punishment

The literature on punishment and prosocial behavior has presented conflicting findings. In some settings, punishment crowds out prosocial behavior and backfires; in others, however, it promotes prosociality. We examine whether the punisher's motives can help reconcile these results through a novel experiment in which the agent's outcomes are identical in two environments, but in one the pre-emptive punishment scheme is self-serving (i.e., potentially benefits the punisher), while in the other it is other-regarding (i.e., potentially benefits a third party). We find that self-serving punishment reduces the social stigma of selfish behavior, while other-regarding punishment does not. Self-serving punishment is thus less effective at encouraging compliance and is more likely to backfire. We further show that the normative message is somewhat weaker when punishment is less costly for the punisher. Our findings have implications for the design of punishment mechanisms and highlight the importance of the punisher's motives in expressing norms.

Tripartite evolutionary game and simulation analysis of electric bus charging facility sharing under the governmental reward and punishment mechanism

Electric vehicles (EVs) are environmentally friendly with the features of low noise levels and zero tailpipe emissions. However, limited charging facilities have restricted the large-scale development of EVs, and electric bus charging facility sharing is a promising solution to this issue. While reducing charging time and improving facility utilization, this measure involves the interests of the government, bus operators, and electric car owners, requiring a systemic analysis of their interactions. Therefore, this study constructs a tripartite evolutionary game model under the governmental reward and punishment mechanism. The model examines the strategic decision-making process, influential factors, and evolutionary stability of three players, followed by simulation analysis. The results show that: (1) The evolutionary system may exhibit three stable states: (0,0,0), (1,0,0) and (1,1,1), and enhance the initial ratio of the three stakeholders will help evolve towards the ideal state. (2) Increasing rewards and penalties can both be meaningful, when penalties exceed rewards, the measures can be more effective. (3) Bus operators are more sensitive to incentive and penalty measures than electric car owners, and governments should pay more attention to them in the implementation process. The findings can provide references for the policy formulation on electric bus charging facility sharing.