مدل رفتار ناایمن در معادن زغال‌سنگ بر اساس نظریه بازی ها

Behavior choice, coal mine monitoring, and control intensity are combined in a general mathematical model established from the perspective of a behavioral game. A case study is provided with effective conditions of monitoring provided. This paper defines the expected value difference of control return and behavior cost difference and discusses the measurement and optimization of variable indexes, including the monitoring intensity and costs of control. The results imply that the control of unsafe behavior can be more effective when monitoring and control of coal mines are both improved. Monitoring will be useful when the rewards for displaying safe behavior, and the monitoring of unsafe behavior, are improved to a high level.

Compared to developed countries the safety situation in Chinese coal mines is still unsatisfactory. This is despite increasingly safe production and fewer accidents and casualties. In fact statistics from coal mine accident investigations show that in China over 80% of the accidents result from direct or indirect behavior of the staff. Coal mine accidents can be reduced through prevention and control of unsafe behaviors. Foreign scholars have studied unsafe human behavior earlier than those in China. In 1919, M Greenwood and H H Woods first started the qualitative study of unsafe behavior through statistical tests on accident occurrences. Their study examined different statistical distributions. The results showed that accidents happen more often with some of the workers. After the end of the 20th century, foreign studies on factors influencing individual safe behavior advanced. Jane Mullen conducted a qualitative study on these influencing factors and presented organizational and social factors related to an individual’s engagement in dangerous work [1]. Hofmann and Stitzer thought that working pressure, resulting from overwork, could influence safe behavior and that insufficient working hour, training, and resources could influence job performance [2]. Laboratory equipment and experimental methods for the study of human factors in China is relatively backward compared to foreign studies. However, since the 1980’s Chinese scholars have been doing more studies on human factors and have made some remarkable achievements. Lin Zeyan discovered that accidents and death tollscaused by human adventure both exceeded those caused by unexpected accidents [3]. This was true in state coal mines, in state owned local mines, or in township mines [3]. Cheng Weimin and Zhou Gang thought human unsafe behavior should be controlled from three aspects: establishing and maintaining an operator’s interest in safe work, job standardization, and safety management [4]. Mi Chuming determined the coal mine, personnel accident cause analysis frame leading to a three level prevention scheme consisting of organization, work team, and the individual. They constructed the correction model of human errors in coal mine accidents [5]. Despite the great value of the above mentioned studies they have only laid importance to the study of specific unsafe behavior monitoring strategies. This results in insufficient quantitative study from the perspective of microscopic mechanisms. A general mathematical model of coal miners’ unsafe behavior has been established from the perspective of behavioral game theory to find new ways of quantitatively studying coal mine safety management practice. This model is based on the above discussion and uses a relative theoretical model for reference [6] and [7]. It refers to staff choice in behavior, coal mine monitoring, and control intensity.

(1) Coal miner behavior can be monitored. Considering this and control intensity a general mathematical model of unsafe behavior monitoring has been established from the perspective of behavioral game theory. (2) Utility analysis is used to show the monitoring of unsafe behavior can be better in the case of more control intensity and can cause the return of safe behavior. The larger the expected value of natural return for unsafe behavior compared to safe behavior, or the larger the cost of safe behavior compared to unsafe behavior, the more difficult monitoring is. (3) Only when the rewards for safe behavior and the monitoring of unsafe behavior reach a high level will the monitoring intensity be positive. Only then will the monitoring be effective. It should be pointed out that this paper presents managers and coal miners as economic men pursuing utility maximization. In reality, decision makers are not always completely rational and are, rather, irrational to some extent. The model can be brought closer to reality and made more explanatory by integrating these irrational characteristics into the unsafe behavior monitoring model. In addition, the measurement of monitoring intensity and the innovation of safe behavior monitoring methods requires further study.