| The Hard Rock Tunnel Boring Machine(TBM)is a large-scale underground engineering equipment for forming tunnel sections of hard rock excavation,slag transport,step-by-step propulsion and support lining.TBM construction is facing the issue of safe and efficient driving and intelligent control,which has become one of the major technical challenges and frontier hot issues in the field of tunnel engineering.The research results on the intelligent tunnel boring machine of the domestic and overseas mainly focued on the shield machine used in the excavation of soft soil strata.There was relatively little research results on the tunnel boring machine for the hard rock tunnel.Moreover,R&D is relatively late,the relationship between rock and machine is not yet clear,the theory of optimal decision-making is not mature,geological information is difficult to obtain in large quantities,the construction workers can not operate the TBM accurately according to the current geological conditions,and the TBM can not work in the optimal excavation status.Therefore,it is necessary to find out the relationship between rock and machine,and then study TBM operating parameters to predict tunneling load,rock grade intelligent identification system and TBM optimization decision theory.This issue was supported by the National Basic Research Program of China(973).Based on abundant data collected from the construction site-Water Supply Project from Songhua River in Jilin province,relevant studies about the relationship between cutterhead torque and tunneling parameters were carried out.The load forecasting model based on the nonlinear support vector regression and the sequence minimum optimization algorithm was established.The model was trained,tested and corrected by field data.Combining the load forecasting model and the improved Particle Swarm Optimization algorithm(PSO),an optimization model of operating parameters considering the energy consumption and the duration was established.According to the fitting relationship between load and penetration at the beginning of TBM tunneling cycle,an intelligent recognition system of rock type was established.According to the load forecasting model of tunneling,optimization decision theory and rock type identification system,a TBM intelligent tunneling optimization decision-making closed-loop system was proposed.The main research work of this paper is as follows.In the first chapter,the development history and research status of TBM technology of the domestic and overseas were introduced,and the development characteristics and development plan were summarized.Then based on the research of this paper,the research status of TBM tunneling load,optimization decision theory and rock type identification methods were expounded,the existing problems in TBM engineering were introduced,the research needs of TBM and the existing research deficiencies were analyzed.Finally,the main contents and significance of this research subject were pointed out.In the second chapter,the Water Supply Project from Songhua River in Jilin province and its database system was Introduced.The correlation between tunneling load and tunneling parameters is analyzed.The reasons for choice of tunneling parameters are expounded.Tunneling data and operating parameters were preprocessed.The set of feature points used to study the relationship of parameters were extracted.Finally,the relationship between operating parameters and tunneling load was studied by using chart analysis method.It was concluded that under the hard rock geological conditions,the tunneling load is mainly affected by the advance speed,and under the soft rock geological conditions,the tunneling load is mainly affected by the cutterhead rotation rate.That is,in the hard rock conditions to get a larger tunneling load need to adjust the advance speed,in the soft rock conditions need to adjust the cutterhead rotation rate.Qualitative description of the role of tunneling load and operating parameters of the relationship between the third chapter to further study the quantitative relationship between the specified direction.This section described the qualitative relationship between tunneling load and operating parameters and pointed out the direction for further study of the quantitative relationships in Chapter Three.In the third chapter,the nonlinear support vector regression(NSVR)prediction model of TBM tunneling load was proposed,and the sequence minimum optimization algorithm was used to speed up the solution.This NSVR model was then applied in the field to calculate tunneling load of 19 854 training samples and 19 854 test samples,and the established NSVR model was trained,tested and adjusted.The prediction results showed that,under the condition of given advance speed,cutterhead rotation rate and surrounding rock type,the average relative prediction error of training sample sets and test sample sets were less than 13.1%.All the data this paper attained demonstrates that provided with specified cutterhead rotation rate,advance speed and rock types,cutterhead torque NSVR model can afford a relatively high accuracy,and can be applied to the fourth chapter tunneling optimization decision theory model.In the fourth chapter,the optimization methods of decision objective were summarized,the function relations of power consumption and operation parameters,tunneling construction period and operating parameters were derived.Based on NSVR prediction model of tunneling load and improved PSO algorithm,the decision of minimum power consumption and shortest construction period was proposed to guide the optimization of operating parameters.The calculation flow chart of two kinds of optimization decision-making theory and algorithm pseudo-code were given.Finally,the results of the optimal operating parameters for the decision of the minimum tunneling energy consumption and the fastest advance speed under different geological conditions were given by using the established optimization decision theory.In the fifth chapter,based on the binary state discriminant function,the excavation cycle data was extracted.The relationship between the total propulsion force and the penetration at the beginning of TBM tunneling cycle was obtained by using nonlinear regression fitting method.The relationship between the cutterhead torque and the penetration at the beginning of TBM tunneling cycle was obtained by using linear regression fitting method.The fitting curves and fitting parameters of tunneling load and penetration were obtained.The relationship between the fitting results and surrounding rocks types was analyzed.It was found that the distribution of the curve is regional and the fitting parameter values appear clustering phenomenon.Based on the clustering boundary of fitting parameters,the dividing line of surrounding rock type was established,and the discriminant equation of surrounding rock boundary based on fitting parameters was obtained.Then,an intelligent identification system of surrounding rock was established.By fitting the unsteady state data,the fitting parameter values were obtained,and then rock types of the current geology were obtained according to the fitting parameters and the discriminant equations of the surrounding rock boundaries.According to the load forecasting model of tunneling,optimization decision theory and rock type identification system,a TBM intelligent tunneling optimization decision-making closed-loop system was proposed.In the sixth chapter,the full text of the work was summarized,and the following research work was prospected. |
PDF Full Text Request