Research On Load Reconstruction Method Of Coal And Rock Cutting By Conical Pick

It is difficult to collect and measure the cutting load directly due to the influence of the actual working state of the shearer in the mine and the complicated severe environment.In general,the inverse problem theory and method are used to reconstruct the cutting load indirectly.Reconstruction is a deconvolution process and belongs to the category of inverse problems,which are often ill-posed.The ill-conditioned system matrix and model uncertainty may seriously deviate the reconstructed cutting load from the actual cutting load.Therefore,it is of great value and significance to research the reconstruction method for effectively dealing with ill-conditioned problems and its application in cutting load reconstruction of coal and rock.In this paper,a systematic study is carried out on how to effectively solve the ill-posed and inverse methods of load reconstruction,and the efficient and practical methods of load reconstruction for coal and rock cutting are explored,and some effective improvements are made on the stability and accuracy of load reconstruction.Therefore,the paper carries out and completes the following work:The cutting load reconstruction model of coal and rock with arbitrary installation angle is established based on the kernel function theory and the limited measured response data.The reconstructed cutting load of coal and rock is expressed as a series of superposition of kernels,and the measurement response is defined as a convolution expression of kernels,reconstructed load and angle coefficient function in the time domain.The reconstructed model of coal-rock cutting load is discretized into a set of linear equations under the assumption of linear time-invariant,the uncertainty of load reconstructed is discussed by using singular value decomposition and Picard criterion,and the methods of selecting regular parameter are expounded.At the same time,the direct regularization method and its iteration method for solving the reconfiguration model are investigated,and scope of application and constraint condition are analyzed.Several kinds of high efficiency load reconstruction methods which are focused on in this paper are given.The load reconstruction method of a novel filter function regularization is proposed based on the basic theory of regularization.The expression of filter function is redefined and a new piecewise filter function is constructed by combining TSVD with Tikhonov regularized filter function,whose purpose is to avoid all singular values being modified and altered in order to eliminate the smoothing phenomenon in the process of load reconstruction.The method proposed in this paper is used to analyze the morbidity of the kernel function matrix,and the convergence of the proposed method is obtained by choosing the prior information of the regular parameters.The feasibility and validity of the method proposed in this chapter are verified by numerical examples and concrete examples of coal and rock cutting.Meanwhile,the regularization parameter selection method and different load reconstruction methods are studied.The load reconstruction method of an estimated extended entropy term regularization is proposed based on norm and entropy theory.The traditional regularization term is replaced by a combination of 1-norm and entropy function,which is defined as the estimated extended entropy term,the purpose of this method is to eliminate the false Peak Phenomena in load reconstruction.The process of dealing with the minimum objective function is transformed into a class of unconstrained optimization problems.A new conjugate gradient method is used to solve the objective function,which avoids the cumulative error caused by repeated solving the regular equation,reduces the number of iterations and improves the accuracy of the load reconstruction solution.The stability of the method is proved from the mathematical point of view,the applicability and validity of the method are verified by numerical examples.The method is applied to reconstruct the cutting load of coal and rock,the influence of different estimated parameters on the cutting load of coal and rock is obtained.At the same time,the optimum combination of parameters is determined.A fractional order method for load reconstruction of coal and rock is proposed based on the mathematical idea of fractional calculus.Integer order regularization is deduced to fractional order.Regularization matrix is defined as the product of diagonal matrix,adjustment coefficient matrix and right singular vector transposition.The purpose of introducing fractional order is to eliminate the useless components of small singular values,which can alleviate the rate of truncation and select useful information to be retained,and the purpose of introducing adjustment coefficient matrix is to reduce the sudden change of load reconfiguration.The expressions of integer order,fractional order and improved fractional order filter functions and their asymptotic properties are obtained,the stability and accuracy of the proposed method are verified and analyzed by numerical examples and coal-rock cutting examples,and the influence of the combination of different regularization parameters,kernel function coefficients and fractional order on load reconfiguration results is investigated.An interval analysis method for reconstructing of coal and rock cutting load is presented due to the non-uniformity,anisotropy and uncertainty of structural parameters of coal.An interval model for reconstructing cutting load of coal and rock is established based on interval set theory,which is suitable for arbitrary uncertain horizontal parameters.The load boundary reconstruction for lower uncertainties,the uncertain parameters in the first kind of model are defined in interval form,only the midpoint value and gradient of the uncertain parameters are needed based on the first order Taylor approximation,and the first order partial derivatives of the uncertain parameters are calculated by finite difference method to avoid the cumulative error caused by nested solution.Load boundary reconstruction for higher uncertainties,the larger uncertain level interval is divided into several sub-interval combinations,the load boundary reconstruction method with lower uncertainty level is used to solve the problem.Because of the morbidity of the interval model,the accuracy of the model solution is analyzed by using multi-parameter regularization soft threshold technology.The method proposed in this chapter is verified and analyzed by the practical application in coal and rock cutting load reconstruction.