Refinement Identification Method Of Digital Drilling For Tunnel Geological

There are plenty of tunnel engineering in our country’s infrastructure construction, the water inrush disasters seriously limit the tunnel construction process, due to the geological body in front of the tunnel face is unknown, reconnaissance data cannot fully reflect construction ahead, brings great blindness to the site operation, so often appear unpredictable geological disasters, including water inrush, collapse and harmful gas, such as engineering disaster caused by tunnel excavation is unable to selectivity, diversity, refractory and sudden, in case of disasters occuring, in general case,machinery and equipment were washed out tunnel, cause interruption of normal construction; Or cause heavy casualties,and huge economic loss, even part of me underground engineering will be forced to relocate or suspend the project, this has become a serious impact on the construction process. Conventional drilling core geotechnical engineering investigation method time-consuming, laborious, the cost is very high, but geophysical exploration methods can’t overcome the shortcomings of information interpretation uncertainty and multiple solutions, as far as we know, a kind of geophysical exploration methods can’t solve those problems at the same time,such as interface recognition, adverse geological body and surrounding rock classification.To solve above problems, and puts forward the tunnel geological figures into fine recognition method, so regard exploring the correlation between drilling parameters and the stratigraphic information index as the main line, regard drillability index, drilling energy and drilling parameter data as the core,with the help of the bionic k-medoids algorithm and quantitative inheritance - RBF neural network approach, on the basis of the design of parameter acquisition system and the data transformation, to realize the recognition of stratigraphic information, and gets the test and application in the engineering practice, a series of innovative research results as the following:(1) Optimizing drilling energy conversion formula, add the neglectful drill water power energy to the formula of drilling energy, at the same time, considering the specific energy of penetrating and specific energy of machinery had a certain error when reflecting the rock drillability, in order to more accurately reflect the objective reality of rock mass, proposed the concept of weighting specific work, and using the efficacy coefficient method and rough set theory, determine the weighted value of weighting specific work formula, defines the weighting specific work formula.These two parts provide a basis for future researchof TEM detection. The modified TEM configuration acquires data in three dimensions and provides a basis for future three dimensional inversions together with the three dimensional modeling.(2) Designed a new digital drilling information acquisition system, improved the parameter acquisition devices in the experimental carrier drill, redesigned the acquisition method of parameters, such as drilling depth, drilling speed, torque, pressure and revolving speed. Improved the data conversion system, used DSP microprocessor to build drilling information mathematical model and convert modulus, and based on the concept of drilling energy and weight ratio work to generate the available numerical drilling energy information, at the same time established a digital drilling information acquisition and data processing system and realized the storage and analysis of digital drilling information.(3) Use the interface recognition method based on the drilling speed-bit speed-propulsion coupling drilling drillability index, to process and analyze the collecting information of the stratum depth, drilling speed, torque, revolving speed, propulsion and so on. Then can get the interface information within the drilling stratum, and in combination with geological analysis to determine the levels of degree of weathering surrounding rock and its properties between the strata interface.(4) As to the large amount of data getting by the digital drilling information acquisition system, use the theory of digital drilling energy transfer them into drilling energy and weight ratio work data sets, then using the bionic k-medoids clustering algorithm to analyze a large number of drilling energy and weight ratio work dates, to identify the unfavorable geological body efficiently and accurately.(5) For surrounding rock classification problem, based on quantum genetic RBF (radial basis function) neural network algorithm, digital drilling surrounding rock classification method is proposed, this method is based on the digital drilling technology, extract useful information from drilling parameters, construct the surrounding rock classification index system, and improve the genetic algorithm by using the theory of quantum computing, regenerate populations by quantum qubit encoding and quantum revolving door, to determine the parameters of RBF neural network, construct a new surrounding rock classification system based on quantum genetic algorithm-RBF neural network.(6) Study on the poor geological digital drilling fine recognition method, as to the F4-4 fault of Qingdao Jiaozhou bay tunnel, use parameters which collected by the digital drilling system, such as drilling speed, torque, rotational speed, propulsion, drilling energy and weight ratio work, to recognize the internal interface and poor geological body and classify the surrounding rock in F4-4 fault.