Workplace stress has a significant impact on productivity, since keeping workers’ stress on an adequate level results a key factor for companies to increase their performance. While a high stress level may conduct to anxiety or absenteeism, a low level may also have undesirable consequences, such as lack of motivation. To identify and understand all the elements which interfere on workers’ stress results a key factor in order to improve workers’ performance. However, the complexity of human behavior increases the difficulty of recognizing the influence of these stressors and finding a way to regulate workers’ stress. This paper proposes the use of agent-based simulation techniques for addressing the challenge of analyzing workers’ behavior and stress regulation policies. The main contributions of the paper are: (i) the definition of a stress model that takes into account work and ambient conditions to calculate the stress and the productivity of workers; (ii) the implementation of this model in an agent-based simulation system, enabling the analysis of workplace stress and productivity for different stress regulation policies; (iii) the analysis of four different stress regulation policies; and (iv) the validation of the model with a sensitivity analysis and with its application to a living lab.

工作场所的压力对生产率产生重大影响，因为将工人的压力保持在适当水平是公司提高绩效的关键因素。高压力水平可能导致焦虑或旷工，而低压力水平也可能产生不良后果，例如缺乏动力。识别和理解影响工人压力的所有因素是提高工人绩效的关键因素。但是，人类行为的复杂性增加了识别这些压力源的影响并找到调节工人压力的方法的难度。本文提出使用基于代理的模拟技术来解决分析工人行为和压力调节策略的挑战。该论文的主要贡献是：（i）定义压力模型，该模型考虑到工作和环境条件来计算工人的压力和生产率；（ii）在基于代理的模拟系统中实施此模型，从而可以分析不同压力调节政策的工作场所压力和生产率；（iii）分析四种不同的压力调节政策；（iv）通过敏感性分析对模型进行验证，并将其应用于实际实验室。