Environment regulation and control method for tunnel excavation of three-arm drill jumbo

Abstract

The invention provides an environment regulation and control method for tunnel excavation of a three-arm drill jumbo, which is suitable for the field of tunnel excavation and comprises the steps of arrangement and installation of monitoring sensors, collection and transmission of data, calculation of a representative value of dust concentration and analysis of the data. According to the environment regulation and control method for tunnel excavation of the three-arm drill jumbo, real-time monitoring, accurate prediction and intelligent regulation and control of the dust concentration in the tunnel excavation process are achieved, the construction safety and efficiency are greatly improved, and the environment regulation and control method can be widely applied to the field of tunnel excavation.

Description

Technical Field

[0001] The invention relates to an environment control method for tunnel excavation using a three-arm rock drilling trolley, and is applicable to the field of tunnel excavation. Background Art

[0002] In tunnel excavation projects, the monitoring and control of dust concentration is an important part of ensuring the health and safety of construction workers. Especially when using a three-arm drilling rig for tunnel excavation, the dust concentration at the construction site is very likely to exceed the standard, leading to a deterioration of the construction environment and potentially having an adverse effect on the health of operators. Therefore, how to effectively monitor and control the dust concentration at the construction site has become a technical problem that needs to be solved in the field of tunnel construction.

[0003] Traditional dust monitoring methods usually rely on simple concentration detection and static early warning mechanisms, which make it difficult to achieve dynamic prediction and real-time control of dust concentration changes. These methods often have shortcomings such as delayed data processing, untimely early warnings, and inflexible control methods, and cannot meet the efficiency and safety requirements of modern tunnel excavation operations. To this end, it is particularly necessary to develop a system that can monitor, predict and control dust concentration in real time during tunnel excavation by combining advanced sensor technology, data analysis algorithms and artificial intelligence technology. Summary of the invention

[0004] This paper proposes a tunnel excavation environment control method based on a three-arm drilling rig. This method uses laser dust monitoring sensors, M estimation method, recursive neural network and other technical means to achieve dynamic monitoring and intelligent control of dust concentration during tunnel excavation.

[0005] The purpose of the present invention can be achieved by adopting the following technical solutions:

[0006] A method for controlling the environment of a tunnel excavation using a three-arm drilling trolley, characterized in that the method comprises the following steps:

[0007] S101, arrangement and installation of monitoring sensors;

[0008] The arrangement and installation of the monitoring sensors include selecting appropriate locations near the tunnel excavation face and on the three-arm drilling rig to arrange a total of n dust monitoring sensors, marked as A1 to A n The dust monitoring sensor can monitor the dust concentration in the nearby air to obtain a dust concentration monitoring value. The dust monitoring sensor measures the dust concentration in the surrounding air based on the laser principle.

[0009] S102, data collection and transmission;

[0010] The data collection and transmission includes starting the dust monitoring sensors A1 to A1 when using a three-arm drilling rig to excavate a tunnel. n , obtain the dust concentration monitoring value N at a certain time interval T ij , N ij It represents the dust concentration monitoring value measured by the i-th sensor at the j-th time interval, i = 1~n;

[0011] S103, calculation of representative value of dust concentration;

[0012] The calculation of the dust concentration representative value includes the dust concentration monitoring value N obtained for the jth time interval 1j ~N nj , use the M estimation method to calculate the representative value M at this time j ;

[0013] S104, analysis of data;

[0014] The data analysis includes using a recurrent neural network based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 ;

[0015] S105, achievement of regulatory objectives;

[0016] The achievement of the control purpose includes the following steps: j+1 , and formulate the dust concentration threshold of the face in combination with the actual situation of the face and the regulations. If the predicted value M j+1 If the dust concentration threshold of the palm surface is exceeded, an early warning will be issued to achieve the purpose of early warning and regulation.

[0017] Furthermore, in S103, the representative value M is calculated using the M estimation method. j The steps are:

[0018] a) The dust concentration monitoring value N obtained at the jth time interval 1j ~N nj ,

[0019] b) Assume that each dust concentration monitoring value N 1j ~N nj The weight values ​​are equal, that is, the weights are all 1.

[0020] c) Calculate the dust concentration monitoring value N 1j ~N nj The initial mean value of is μ,

[0021] d) Calculate the residual σ of each monitoring value using the initial mean μ j ,

[0022] e) Calculate the weight ω(M) of each monitoring value based on the residual result j ),

[0023] f) Recalculate the weighted representative value M' j , and check convergence. If the difference is small, it is considered converged. Otherwise, adjust the weight again and continue to iterate until the representative value converges. Finally, the representative value M is obtained. j .

[0024] Furthermore, the residual σ of each monitoring value is calculated j The method is to use formula (1) to calculate and obtain:

[0025] σ j =M j -μ (1)

[0026] Furthermore, the weight ω(M) of each monitoring value is calculated according to the residual result. j ) is calculated by using formula (2).

[0027]

[0028] Furthermore, the recalculated weighted representative value M' j The calculation formula of is obtained by using formula (3).

[0029]

[0030] Furthermore, in the above S104, the recursive neural network is used based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 The steps are:

[0031] a) Select representative value data M j , according to the time acquisition order, a representative value sequence data set C = {M1, M2...Mj} is formed,

[0032] b) Use the maximum and minimum method to standardize the data,

[0033] c) Divide the data set C into training set and test set according to a certain ratio,

[0034] d) Use a sliding window method to convert the data into a format suitable for RNN input,

[0035] e) constructing a recursive neural network model using a deep learning framework, setting the number of layers, number of units, activation function and dropout rate of the recursive neural network,

[0036] f) using the training set data to carry out model training on the recursive neural network model, and using the test set to evaluate and optimize the recursive neural network model, wherein the evaluation refers to using the mean square error as the loss function to evaluate the prediction performance of the recursive neural network model, and the optimization refers to adjusting the parameters of the recursive neural network model according to the evaluation results.

[0037] g) Finally, the optimized recursive neural network model is obtained to obtain the predicted value M j+1 .

[0038] The beneficial effects of the present invention are:

[0039] The beneficial effects of the present invention are: (1) Real-time monitoring and precise control: by arranging multi-point dust monitoring sensors, the dust concentration data of the tunnel excavation site is obtained in real time to ensure the dynamic monitoring of the dust concentration on site. Combined with the M estimation method and the recursive neural network model, the change of dust concentration is accurately predicted, and timely warning and control can be carried out to effectively prevent the dust concentration from exceeding the standard. (2) Improving construction safety: predicting the trend of dust concentration changes in advance can enable construction personnel to take corresponding protective measures before the dust concentration may exceed the standard, reduce the harm of harmful dust to the health of operators, and improve the safety of tunnel construction. (3) Robustness of data processing: using the M estimation method to process dust concentration data can effectively reduce the errors caused by abnormal data or noise, improve the accuracy and robustness of the calculation of the representative value of dust concentration, and thus improve the reliability of the overall monitoring and prediction results. (4) Intelligent control means: using recursive neural networks to predict dust concentration, breaking through the limitations of traditional static monitoring means, realizing the intelligence and automation of dust concentration control, reducing the need for manual intervention, and improving control efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 :Flow chart of a method for controlling the environment of tunnel excavation using a three-arm drilling trolley of the present invention,

[0041] Figure 2 : A dust concentration monitoring data diagram of an embodiment of an environmental control method for tunnel excavation using a three-arm drilling rig of the present invention,

[0042] Figure 3 : Representative value data diagram of an embodiment of an environmental control method for tunnel excavation using a three-arm drilling rig of the present invention. DETAILED DESCRIPTION

[0043] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings; it should be understood that the specific embodiments given here are only used to illustrate and explain the present invention and cannot be used to limit the present invention.

[0044] The following is a specific embodiment of an environmental control method for tunnel excavation using a three-arm drilling rig.

[0045] like Figure 1 Shown is a flow chart of an environmental control method for tunnel excavation using a three-arm drilling rig according to the present invention.

[0046] The project name of this embodiment is a tunnel construction project of the East Malaysia Rail Project. The overview of the tunnel project is as follows: the total length is 9882m, of which the import mileage is C2CH4+642 and the export mileage is C2CH14+524. The line descends from the import end to the export end C2CH14+236 at a slope of 6.7‰, with a maximum burial depth of 460m and a minimum burial depth of 3m. The tunnel is designed as a composite lining structure with a standard line spacing of 4.2m. The design of the Shuangwendan 2# tunnel has a total of 4 inclined shaft auxiliary construction and rescue channels with a total length of 5029m. The clear section size of the inclined shaft is 5m*8m, the maximum comprehensive slope is -10%, and a trackless double-lane design is adopted, with an anchor sprayed lining + molded lining structure.

[0047] The tunnel is located in Ulu Selangor, Selangor State. Selangor State is located in the middle of the west coast of Peninsular Malaysia, bordering Perak State to the north, Pahang State to the east, Negeri Sembilan State to the south, and the Strait of Malacca to the west. The Serendah 2# Tunnel is located in a forest reserve, and the vegetation in the tunnel site area is well developed.

[0048] The surrounding rock of the tunnel is mainly composed of silty clay and granite. Silty clay is mainly distributed in the tunnel entrance section, and the tunnel body is mainly granite with local fragmentation and cracks. The main distribution along the East Sea Railway in Malaysia is Quaternary alluvial and diluvial layers, slope deposits, Jurassic-Cretaceous granite, quartzite, Ordovician-Silurian phyllite, schist, shale, limestone, marble, etc. The Quaternary cover layer is relatively thin in the central mountain area.

[0049] A method for controlling the environment of a tunnel excavation using a three-arm drilling trolley, characterized in that the method comprises the following steps:

[0050] S101, arrangement and installation of monitoring sensors;

[0051] The arrangement and installation of the monitoring sensors include selecting appropriate locations near the tunnel excavation face and on the three-arm drilling rig to arrange a total of n dust monitoring sensors, marked as A1 to A n The dust monitoring sensor can monitor the dust concentration in the nearby air to obtain a dust concentration monitoring value. The dust monitoring sensor measures the dust concentration in the surrounding air based on the laser principle.

[0052] In this embodiment, taking a certain section of the tunnel as an example, a total of 4 dust sensors are arranged, marked as A1 to A4, and the index parameters of the dust sensors are shown in Table 1.

[0053] Table 1 Product technical parameters of dust sensor

[0054] project Specifications principle Laser Technology Range 0~200mg / m3 Resolution 0.001mg / m3 Indication error ±10％ Detection particle size PM1.0, PM2.5, PM5, PM10, TSP Sampling method Pump Suction Air pump flow 2.0L±5％ Interface display Touch LCD display Alarm output Sound alarm

[0055] S102, data collection and transmission;

[0056] The data collection and transmission includes starting the dust monitoring sensors A1 to A1 when using a three-arm drilling rig to excavate a tunnel. n , obtain the dust concentration monitoring value N at a certain time interval T ij , N ij It represents the dust concentration monitoring value measured by the i-th sensor at the j-th time interval, i = 1~n;

[0057] In this embodiment, dust sensors A1-A4 obtain dust concentration monitoring values ​​at a certain time interval T=5min, and a total of 5 monitoring values ​​are obtained, which are recorded as N 11 ~N 15 , N 21 ~N 25 , N 31 ~N 35 , N 41 ~N 45 , N 51 ~N 55 , the distribution of the data is as follows Figure 2 Dust concentration monitoring value is shown.

[0058] S103, calculation of representative value of dust concentration;

[0059] The calculation of the dust concentration representative value includes the dust concentration monitoring value N obtained for the jth time interval 1j ~N nj , use the M estimation method to calculate the representative value M at this time j ;

[0060] Furthermore, in S103, the representative value M is calculated using the M estimation method. j The steps are:

[0061] a) The dust concentration monitoring values ​​N1j~Nnj obtained at the jth time interval,

[0062] b) Assume that the weight of each dust concentration monitoring value N1j~Nnj is equal, that is, the weight is 1.

[0063] c) Calculate the initial mean value of the dust concentration monitoring values ​​N1j~Nnj as μ,

[0064] d) Calculate the residual σ of each monitoring value using the initial mean μ j ,

[0065] e) Calculate the weight ω(M) of each monitoring value based on the residual result j ),

[0066] f) Recalculate the weighted representative value M' j , and check convergence. If the difference is small, it is considered converged. Otherwise, adjust the weight again and continue to iterate until the representative value converges. Finally, the representative value M is obtained. j .

[0067] Furthermore, the residual σ of each monitoring value is calculated j The method is to use formula (1) to calculate and obtain:

[0068] σ j =M j -μ (1)

[0069] Furthermore, the weight ω(M) of each monitoring value is calculated according to the residual result. j ) is calculated by using formula (2).

[0070]

[0071] Furthermore, the recalculated weighted representative value M' j The calculation formula of is obtained by using formula (3).

[0072]

[0073] In this embodiment, the representative values ​​M1-M5 obtained by using the M estimation method are 0.100, 0.133, 0.137, 0.143 and 0.173 respectively. Figure 3 Representative values ​​are described in the data figure.

[0074] S104, analysis of data;

[0075] The data analysis includes using a recurrent neural network based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 ;

[0076] Furthermore, in the above S104, the recursive neural network is used based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 The steps are:

[0077] a) Select representative value data M j , according to the time acquisition order, a representative value sequence data set C = {M1, M2...Mj} is formed,

[0078] b) Use the maximum and minimum method to standardize the data,

[0079] c) Divide the data set C into training set and test set according to a certain ratio,

[0080] d) Use a sliding window method to convert the data into a format suitable for RNN input,

[0081] e) constructing a recursive neural network model using a deep learning framework, setting the number of layers, number of units, activation function and dropout rate of the recursive neural network,

[0082] f) using the training set data to carry out model training on the recursive neural network model, and using the test set to evaluate and optimize the recursive neural network model, wherein the evaluation refers to using the mean square error as the loss function to evaluate the prediction performance of the recursive neural network model, and the optimization refers to adjusting the parameters of the recursive neural network model according to the evaluation results.

[0083] g) Finally, the optimized recursive neural network model is obtained to obtain the predicted value M j+1 .

[0084] In this embodiment, a recursive neural network is used to construct a prediction model based on the representative values ​​M1 to M5 to obtain a prediction value M6=0.172;

[0085] S105, achievement of regulatory objectives;

[0086] The achievement of the control purpose includes the following steps: j+1 , and formulate the dust concentration threshold of the face in combination with the actual situation of the face and the regulations. If the predicted value M j+1 If the dust concentration threshold of the palm surface is exceeded, an early warning will be issued to achieve the purpose of early warning and regulation.

[0087] In this embodiment, according to the obtained predicted value M6=0.172, combined with the tunnel face conditions, close attention is required at this time.

[0088] In the above embodiments, the present invention discloses an environmental control method for tunnel excavation with a three-arm drilling rig; including the arrangement and installation of monitoring sensors, data collection and transmission, calculation of representative values ​​of dust concentration, and data analysis; the present invention proposes an environmental control method for tunnel excavation with a three-arm drilling rig, which realizes real-time monitoring, accurate prediction and intelligent control of dust concentration during tunnel excavation, greatly improves construction safety and efficiency, and can be widely used in the field of tunnel excavation.

[0089] The above description is a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the environment of a three-arm drilling rig tunnel excavation, characterized in that: The method for controlling the environment of tunnel excavation by a three-arm drilling rig comprises the following steps: 1) Arrangement and installation of monitoring sensors; The arrangement and installation of the monitoring sensors include selecting appropriate locations near the tunnel excavation face and on the three-arm drilling rig to arrange a total of n dust monitoring sensors, marked as A1 to A n The dust monitoring sensor can monitor the dust concentration in the nearby air to obtain a dust concentration monitoring value. The dust monitoring sensor measures the dust concentration in the surrounding air based on the laser principle. 2) Data collection and transmission; The data collection and transmission includes starting the dust monitoring sensors A1 to A1 when using a three-arm drilling rig to excavate a tunnel. n , obtain the dust concentration monitoring value N at a certain time interval T ij , N ij It represents the dust concentration monitoring value measured by the i-th sensor at the j-th time interval, i = 1~n; 3) Calculation of representative value of dust concentration; The calculation of the dust concentration representative value includes the dust concentration monitoring value N obtained for the jth time interval 1j ~N nj , use the M estimation method to calculate the representative value M at this time j ; 4) Data analysis; The data analysis includes using a recurrent neural network based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 ; 5) Achievement of regulatory objectives; The achievement of the control purpose includes the following steps: j+1 , and formulate the dust concentration threshold of the face in combination with the actual situation of the face and the regulations. If the predicted value M j+1 If the dust concentration threshold of the palm surface is exceeded, an early warning will be issued to achieve the purpose of early warning and regulation.

2. The method for controlling the environment of tunnel excavation by a three-arm drilling rig according to claim 1, characterized in that: In the above 3), the representative value M is calculated using the M estimation method. j The steps are: a) The dust concentration monitoring value N obtained at the jth time interval 1j ~N nj , b) Assume that each dust concentration monitoring value N 1j ~N nj The weight values ​​are equal, that is, the weights are all 1. c) Calculate the dust concentration monitoring value N 1j ~N nj The initial mean value of is μ, d) Calculate the residual σ of each monitoring value using the initial mean μ j , e) Calculate the weight ω(M) of each monitoring value based on the residual result j ), f) Recalculate the weighted representative value M' j , and check convergence. If the difference is small, it is considered converged. Otherwise, adjust the weight again and continue to iterate until the representative value converges. Finally, the representative value M is obtained. j .

3. A method for controlling the environment of tunnel excavation using a three-arm drilling rig according to claim 2, characterized in that: The residual σ of each monitoring value is calculated j The method is to use formula (1) to calculate it. s j =M j -m (1).

4. A method for controlling the environment of tunnel excavation using a three-arm drilling rig according to claim 2, characterized in that: The weight ω(M) of each monitoring value is calculated according to the residual result. j ) is calculated using formula (2).

5. A method for controlling the environment of tunnel excavation using a three-arm drilling rig according to claim 2, characterized in that: The recalculated weighted representative value M' j The calculation formula of is obtained by using formula (3).

6. The method for controlling the environment of tunnel excavation by a three-arm drilling rig according to claim 1, characterized in that: In the above 4, the recursive neural network is used based on the representative value M j Construct a prediction model to obtain the predicted value M j+1 The steps are: a) Select representative value data M j , according to the time acquisition order, a representative value sequence data set C = {M1, M2...M j }, b) Use the maximum and minimum method to standardize the data, c) Divide the data set C into training set and test set according to a certain ratio, d) Use a sliding window method to convert the data into a format suitable for RNN input, e) constructing a recursive neural network model using a deep learning framework, setting the number of layers, number of units, activation function and dropout rate of the recursive neural network, f) using the training set data to carry out model training on the recursive neural network model, and using the test set to evaluate and optimize the recursive neural network model, wherein the evaluation refers to using the mean square error as the loss function to evaluate the prediction performance of the recursive neural network model, and the optimization refers to adjusting the parameters of the recursive neural network model according to the evaluation results. g) Finally, the optimized recursive neural network model is obtained to obtain the predicted value M j+1 .

Images