Font Size: a A A

Establishment Of Dynamic Prediction Model And Implementation Of The Algorithm For Mining Subsidence

Posted on:2018-12-02Degree:DoctorType:Dissertation
Country:ChinaCandidate:B ZhangFull Text:PDF
GTID:1311330518468962Subject:Geodesy and Survey Engineering
Abstract/Summary:PDF Full Text Request
At present,Chinese main energy is still dominated by coal,and its consumption accounts for more than 60%of the total energy consumption.In the long run,with the increasingly serious form of air pollution in China,the trend that clean energy will gradually replace coal energy has been inevitable,but the coal-based consumption structure will continue to maintain in the short term.Huge amount of coal consumption has led to a great deal of coal mining,and coal mining leads o land subsidence on the farmland and the environment hazards in return,which is now very serious,at the same time,due to the rapid development of China's urbanization,a direct result of that is that the amount of coal under the buildings,rivers and railways?referred to"three under"?is increasing year by year,so is the"three under"coal mining.And the hazards of surface subsidence and economic losses are also growing,which bring about a series of serious social problems.The study of the collapse of ground surface and overlying strata caused by mining can be traced back to the end of the 19th century.However,according to the literature analysis,it is found that the study proportion of movement deformation of surface and overlying strata after the stability of mining subsidence is relatively large,and the dynamic study of surface and overlying strata during mining is rather few.The dynamic prediction of surface movement and deformation caused by mining is one of the important contents of mining subsidence.To master the mining process with the time of the dynamic changing,not only can grasp the regulations of the surface and overlying strata movement and deformation,but also obtain the maximum deformation value of the time and space position in real-time.It is of great significance to evaluate the deformation process and the damage degree of the surface buildings in the process of mining,to formulate the maintenance and reinforcement scheme of the buildings and to determine the implementation time.On the basis of other scholars'research,this paper has carried on the thorough research on some problems which exist in the prediction of mining subsidence,mainly including the following aspects.Firstly,several time function models which can be commonly used to predict the mining subsidence are studied,and their advantages,disadvantages and applicability are pointed out.The segmented Knothe time function is studied,and it is pointed out that there are problems in practical application,and a correction method for the problem is put forward.Secondly,the prediction method for dynamic movement deformation is studied and the corresponding algorithm is designed for the strike main section and the incline main section and any point on the surface in the mining of rectangular face.Thirdly,the triangular division method of the irregular work area is studied and the corresponding algorithm is designed.Fourthly,the methods and algorithms of dynamic forecasting of irregular face are studied.Fifthly,the algorithm proposed by the dissertation proves to be practical through Matlab software,and the prediction system for mining subsidence is developed.Last but not least,the orthogonal experiment method of numerical simulation is studied to determine the rock mechanics parameters,and the dynamic problem of mining subsidence is studied by numerical simulation method,and a new method is proposed for numerical modeling of complex terrain.The main results of this paper are as follows:?1?The segmented Knothe time function is studied and optimizedThe two main problems of the segmented Knothe time function model are pointed out through studying.First,at the time pointt,the value of the time function does not match the theoretical value,if the time function is used for dynamic prediction,the prediction error will occur at the time of segmentation.This error will increase as the product of the two parameters of the time function decreases,and the error can be effectively reduced only when the product of the two parameters is greater than a certain value,and this problem greatly limits the applicability of the dynamic prediction when using the time function.Second,the segmented Knothe time function value does not converge to 1 when the product of its two parameters is smaller,but converges to a value less than 1,and the smaller the product is,the greater the deviation of the convergence value is 1,which decreases the prediction accuracy of the subsidence?maximum?value of the ground surface.Aiming at the problems existing in the segmented Knothe time function,this paper proposes a method,which is used to optimize original function.The function value of the optimized model,no matter how its value of the parameterstand c are given,is equal to the theoretical value at the point of segmentation,and the time function value finally converges to 1,which solves some theoretical problems existing in the practical application of the original segmented Knothe time function.?2?In this paper,a new time function--normal distribution time function is introduced,and the variation range and the variation rule of the time function are compared and analyzed.The conditions that can be used to predict the time function of mining subsidence surface are discussed.The significance of the normal distribution time function parameter is studied,and the dynamic predictive adaptability of the time function is studied.The results are as follows:first,the subsidence velocity and acceleration of the normal distribution time function are consistent with the subsidence law of the surface;second,when the value of the parameterdof the normal distribution time function is different,the time function value does not always converge to 1.When the value of the parameters change from large to small,the error of the time function value converges to 1 will be small to large,which shows that the smaller the value ofd,the less suitable as a dynamic time function,in the case of dynamic prediction with low accuracy requirements,the value ofdcan be given a value of 2 or greater than 2,and when the prediction accuracy is high,the value ofdneed to be given a value of 3 or more than 3,and whendis less than 1,using it as time function for dynamic prediction,the prediction error will be big.?3?Based on the probability integral method and the Knothe time function model optimized in this paper,the division method of dynamic mining unit on rectangular working face,the method for determining parameters of time function and the method for calculating the time function value of each dynamic mining unit are studied.In the dynamic prediction of the incline main section,the corresponding parameters are different due to the different mining depths of each dynamic mining unit,so the formula for calculating the probability integral parameters of the dynamic elements of the incline main section is established.Finally,the dynamic prediction model of the strike main section,the incline main section and the arbitrary point of the rectangular working face on the surface is established,and the corresponding algorithm is given.?4?In order to solve the problem of determining the parameters of the segment Knothe time function of being optimized,this paper presents a method of“Comparison of methods of determining parameters based on measured data”,and it is simple and reliable.In addition,based on the probability integral parameters of the mining area and the Knothe's"graphic method"proposed by Knothe,Poland scholar,the direct calculation method of determining time parameters is studied.According to this method,we can accurately obtain the c value of the time function parameter in the studying mine if we have obtained the probability integral prediction parameters of this mining area,and the calculated c value is no longer an interval,but a fixed value,which is very convenient for computer programming.?5?In this paper,the triangle division method for irregular working face is studied,and the principle of division of triangle integral region with different shapes is given,and the method of determining the upper and lower bounds of integral interval is derived,and finally the concept of"triangulation"of irregular working area is proposed,and the corresponding algorithm is designed.The algorithm can be used to static subsidence prediction for level coal seam and inclined coal seam,and can also be used to multi mining face with different depths and different shapes.In addition,in order to make the prediction algorithm brief and efficient?before the dynamic prediction,the National coordinate system is usually converted to the coordinate system of working face?,as well as realizing the mutual transformation between the national coordinate system and the coordinate system or other coordinate system,the method of mutual transformation between coordinate systems is studied,and the method for determining the parameters of the conversion is given.?6?Based on the method of triangulation,the dynamic prediction method for the mining of convex polygon is studied,and a method for determining the real-time integral field of irregular working face and a method to determine the time interval of each dynamic unit at any given time are proposed.In addition,two kinds of dynamic prediction methods are studied when the vertices of the working face are sorted in order or in any way,and the corresponding algorithm is designed.The dynamic prediction method for the mining of special concave polygon working face is studied,and a computing method is proposed,which is based on the change of the advancing direction of the working face,the coordinate system of the working face is changed accordingly.And then,the calculation results of different coordinate systems are summed.Therefore,the problem of determining the probability integral interval of concave polygon surface is solved.?7?In this paper,matlab software is used to implement the various algorithms proposed in this paper,and the prediction system of mining subsidence is developed.Under the condition of mining in horizontal or gently inclined coal seam,the system can be used for the dynamic prediction or static prediction when the mining face are with arbitrary shape.It is possible to predict the surface movement and deformation of multi-working face mining under different levels and different geological mining conditions.In addition,the system has beautiful interface and complete functions and has powerful function of data processing and 3D visualization.It can deal with the prediction results directly and it needs no other softwares,and thus the efficiency of the application is improved.Though the verifying by the real examples,it shows that the relative error of the prediction results can be controlled within 7%,and the reliability of the system is proved.?8?Aiming at the problem of numerical modeling for complex terrain,based on Auto CAD and Matlab software,this paper presents a new method to build numerical model for complex terrain.Using this method to build the model,the data processing workload is small and the modeling speed is fast,moreover,it can be used to establish the three-dimensional numerical model of any complex terrain,and the deviation between the surface of the numerical model and the actual terrain surface is very small.?9?The numerical simulation of mining in 6206 working face with complex shape in Wang Zhuang coal mine is carried out.The method of establishing the numerical model of the working face is introduced in detail.Aiming at the difficult problem of obtaining physical and mechanical parameters of rock mass in numerical simulation,according to the surface subsidence data of the strike observation line and incline observation line,the method of calculating the rock mechanics parameters based on orthogonal experiment and numerical simulation is discussed in detail.?10?Through the numerical simulation,it points out that when the mining speed is very large,because of the large area of one-time mining,in a short period of time,the movement range of the strata and surface caused by mining has been basically formed,but the values are relatively small.When there are faults in the model,the vertical displacement of strata increases obviously on both sides of the fault.At a certain range of the both sides of fault line,the rock has large vertical displacement,pyramid development.With the increase of mining time,pyramid range and vertical displacement increases.Then,with the increase of time,the scope of the tower gradually disappears,and the difference between the displacements of the surrounding rocks is gradually reduced,however,the displacement on both sides of the fault is obviously smaller than that of the surrounding rocks,and the displacement of the strata will still be symmetrical along the fault.?11?In the numerical simulation of irregular working face,no matter how long the mining length is,in the initial stage,when the mining area is small,with the increase of the length of the mining,the calculation time of the stability of the model increases rapidly.However,when the mining length increases to a certain value,although the increased length of the mining is the same,the increase of the calculation step is obviously reduced.Based on the analysis,the relationship between the number of calculation steps and the mining area is obtained,that is:n=8.83A0.7032-0.2481A?...
Keywords/Search Tags:Mining subsidence, Dynamic prediction model, Irregular Working face, Surface movement and deformation, Prediction algorithm
PDF Full Text Request
Related items