基于进化稳定策略的施工项目安全管理研究

On the safety management of the construction project based on the evolutionary stable strategy

在大规模建设施工过程中,施工安全是工程项目得以有效进行的前提,安全管理应引起高度重视。采集施工损失的数据,分析了安全事故的变化趋势,进而通过收益矩阵建立了施工安全的复制动态方程模型,探讨局中人进行施工安全管理的必要性,并根据局中人的进化稳定策略做盈亏平衡分析。结果表明,在追求利益最大化的前提下,局中人会以最小损失的方式选择博弈策略。以复制动态方程为例,当x*i≤0时,函数F(xi)在[0,1]上的进化稳定策略为x*i=1,得到安全管理成本Ci≤Ri+P1(1-P2)Yi,表明较低的成本会使所有局中人都愿意进行安全管理。最后,从算例可知,当x*i分别在区间[-∞,0]、[0,1]、[1,+∞]上取值时,局中人的策略也会随之变化。

In the process of large-scale construction, safety is a pre- requisite for the project to carry out effectively. Therefore, for the re- search purpose, we have collected the data of construction safety ac- cidents that have taken place in China from 1998 to 2012, and have made an analysis of the trend of these accidents. In proceeding with our research, we have further explored the strategy of the construction safety management in the payoff matrix, and have set up the replica- tor dynamics of the evolutionary game theory for the sake of discussing the necessity of the safety management. Besides, we have also made a break-through analysis by using the evolutionary stable strategy （ESS）. The results of our study have shown that, in case when we were trying to seek measures to maximize the benefits, the players in the matrix game would like to select the optimal strategy with the min- imal loss. To clarify the problems of construction safety, the replica- tor dynamics function can be expressed by the equation of f（x） = f（x, C,R, Y, PI,P2） （x ε [0,1]） . Thus, it can easily be seen that the evolutionary stable strategy （ESS） of f（ x ） is stationary if x ≤ 0 , which can then be denoted by x = 1 . Thus we have gained C 〈 R ＋ Pl（1 - P2）Y , which means that all the players should be able to promote the safety management and control the influential fac- tors of the safety accidents on the engineering cost. What＇ s more, the evolutionary stable strategy （ESS） should be equal to f（x） , which can then be represented by x = 0 if x ≥ 1. Hence, we gain C R ＋ P1 Y. Nevertheless, all the players in this case wouldn＇ t be able to promote the safety management for the cost is out of the con- trol. And, then, if 0〈 x 〈 1, the evolutionary stable strategy （ESS） off（x） should be able to be rewritten both by x = 0 and x = 1. However, in the context of seeking to minimal loss, the evolution- ary stable strategy （ESS） of all the players should be increasingly de- noted by x = 0. Thus, through the aforementioned examples illustrat- ed in the paper, it is possible for us to arrive at a conclusion that all the players should be able to change their strategies if the element x is in the different intervals of the replicator dynamics.