Research On Reliability Optimal Control Of Underground Uranium Mine Production Safety Multi-Agent System

Underground uranium mine production safety system is a flexible system which is related to multi-factor, multi-level, dynamic,involving many subsystems of mine production safety such as mining drilling-blasting, mine ventilation, radon and radon daughters precipitation, mine transportation and hoist; although the mechanism of each one's occurrence varies, but the factors that causes the accident are interrelated and interactive. Therefore, for the complexity of the uranium mining operating environment, network character of production safety systems and the diversity character of accident causes, carrying out research on the reliability and optimal control of uranium mine production safety systems has become an urgent task to solve the problem of optimal control of underground uranium mine production safety systems. Therefore, multi-agent technology and reliability theory are applied to build structure and consultation mechanisms of multi-agent systems reliability model of underground uranium mine production safety,applying immune genetic optimization algorithm to achieve optimal control of underground uranium mine safety production of multi-agentsystems. The main research work is as follows:(1) Building the reliability structural system of underground uranium mine production safety systems based on Multi-Agent technology.Through in-depth understanding of production technology and production characteristics of underground uranium mine production safety system,combined with multi-agent theory, the underground uranium mine production safety system is divided into Human-computer interaction interface, Master Agent, Blasting Agent, Exploitation and Loading(ore drawing) Agent, Transportation and Hoisting Agent, Ventilation Agent,Radioactive contamination control Agent, and each agent has its own decision-making, computing and independent communication ability, but their own perceptions are limited, requiring judgment based on information from other Agents. Internal structure is designed for functional attributes of each agent model.(2) Research on multi-agent systems negotiation mechanism of underground uranium mine production safety. By the application of internal mechanism of the system(series, parallel and hybrid) and consultation mechanism of multi-agents, we can eventually solve the problem of internal consultations of mine production safety system.During MAS's running, it requires Agents with different capabilities to collaborate to achieve goal that the reliability of the whole system to the highest as to complete complex system tasks.(3) Research on optimal control of immune genetic optimization of multi-agent systems of production safety. Integrating intelligent optimal algorithm into immune genetic optimal algorithm and optimal control theory, building high performance intelligent optimal algorithm, we can use immune genetic algorithm respectively to solve multi-agent system structure integrated system by multiple iterative optimization to meet the needs of overall situation optimization of multi-agent structure multi-objective system, which can address the problem that overall situation optimal control of multi-agent production safety systems of underground uranium mine.(4) Analysis on the case of Uranium mine safety production optimization and multi-agent system. Take a large underground uranium mine as an example to model and integrate multi-agent production safety systems; set constraints and operational processes between the respective Agents based on mine's actual production situation. Multi-agent technology, reliability theory and immune genetic algorithm are applied to solve the problem of coordination and optimal control of the underground uranium mine production safety system, and at the same time, computer programming and debugging are also accomplished.In summary, in this paper, underground uranium mine production safety multi-agent systems reliability model and immune genetic optimization control research based on multi-agent technology andreliability theory is completed and solve the problems that it's difficult to model of reliability of the study of complex systems, to obtain realistic reliability index and to solve the problem about the system optimal of control overall situation. It can play an important guiding role to achieve optimization of maintenance strategy of underground uranium mines and reasonable arrangements for production planning. Moreover, empirical analysis and test of the effectiveness and scope of validation of the research results is also completed.