| The mining subsidence prediction model and its parameter system are the core basic theories of "under three" coal mining design.Inverting probability integral parameters accurately and reliably is prerequisite for implementing accurate surface subsidence prediction.The parameter inversion model of probability integration method is a typical multi-parameters complex nonlinear model,and there is a correlation among the parameters.Therefore,it is difficult to obtain satisfactory results by estimating the probability integration parameters by using the conventional method.In view of the above deficiencies,this paper proposes new methods based on "quantum+intelligent algorithm",and the simulated experiment and engineering experiment have a good effect.In addition,in view of the problem in accurately detecting the characteristics of the goaf which restricts the development and utilization of closed/abandoned mine resources,this paper considers that surface subsidence is the essence of the external manifestation of coal mining space propagating to the surface according to a specific mechanism,and constructs a mining time-space characteristics inversion model based on surface movement deformation observations and a"quantum+intelligent algorithm" parameter inversion method.Through the above research,this paper has achieved the following results:(1)Construction of parameters inversion model of probability integration method based on "quantum+" intelligent optimization algorithm.Aiming at the shortcomings of genetic algorithm(GA)and simulated annealing(SA)which are more commonly used in the parameters inversion of probability integration method,combined with quantum bit coding,quantum revolving gate and quantum fluctuation mechanism,the parameters inversion models of probability integration method based on quantum genetic algorithm(QGA)and quantum annealing(QA)are constructed,respectively.The simulation experiment data proves that QGA and QA have obvious improvement in reliability,anti-random error ability,anti-gross error ability and anti-loss points ability compared with traditional GA and SA.The experimental results of actual engineering application prove that the parameter inversion model based on"quantum+" probability integral method has higher fitting degree and more reliability.(2)Establishing the identification model for characteristic parameters of the goaf based on "quantum+" intelligent optimization algorithm.The characteristic parameters of the rectangular working face are taken as unknown parameters,and the parameters of probability integration method is used as the known parameter,the identification model for characteristic parameters of the goaf based on probability integration,method is established.The simulation experiments are carried out to study the reliability,anti-random error ability,anti-gross error and anti-loss point ability of the two algorithms used in identification method for characteristic parameters of the goaf based on probability integration method.And combined with engineering application experiments,the accuracy of the identification method for characteristic parameters of the goaf based on "quantum+" intelligent optimization algorithm is verified.It also proves that compared with the traditional GA and SA,the superiority of QGA and QA in identification method for characteristic parameters of the goaf based on "quantum+"intelligent optimization algorithm.(3)Based on the single-phase absolute deformation data and the two-phase relative deformation data of the surface observation station,the corresponding dynamic inversion model of the parameters of the probability integral method is constructed,respectively.The absolute deformation data and the two-phase relative deformation data are taken as the fitting criterion respectively;the Knothe time function and the probability integration method dynamic prediction model are combined to establish a dynamic inversion model of the probability integration method.Taking 1414(1)working face in Huainan mining area as the research object,the variation law between the probability integration parameters obtaind by a dynamic inversion model of the probability integration method based on the inversion of single-phase observation data and the ratio of the width to the depth of the working face is studied.Meanwhile,the feasibility of the dynamic inversion method for predicting parameters of probability integral method based on relative deformation data is studied.It is also proved that the larger the time span of the two phases of data,the higher the accuracy of dynamic inversion of parameters.Figure[61]table[19]reference[105]... |
PDF Full Text Request