Assembly type underground structure construction assembly quality control system and method

Abstract

The invention discloses an assembly type underground structure construction assembly quality control system and method. The method comprises the following steps: S1, determining a mathematical relationship model between piezoelectric signals and duct piece joint opening amount; s2, dividing a plurality of early warning levels of the shield tunnel joint opening amount, and determining piezoelectric signal ranges corresponding to the early warning levels; s3, before shield tunnel construction, flexible piezoelectric sensing devices are arranged between shield tunnel segments; s4, constructing a shield tunnel segment joint opening prediction model; and S5, monitoring the splicing state of the shield tunnel segments in real time by using a flexible piezoelectric sensing device, predicting the splicing state of the shield tunnel segments by using the shield tunnel segment joint opening amount prediction model, and adjusting construction parameters in time according to a prediction result. According to the method, on the basis of sensing the mechanical response in the segment splicing process in real time, the safety state of the shield tunnel segment joint opening amount is determined, and a reasonable basis is provided for adjustment of follow-up shield tunnel construction parameters and real-time monitoring and control of the segment splicing state.

Description

Technical Field

[0001] The present invention relates to the technical field of shield tunnel construction, and in particular to a quality control system and method for assembling a prefabricated underground structure. Background Art

[0002] In recent years, shield tunnels have developed rapidly in the fields of transportation and municipal administration, and the number of curved shield tunnels has increased day by day. Affected by factors such as the tunnel line shape and the tunneling posture of the shield machine, the assembly quality of the segments is difficult to control. Especially when the curvature radius is small, there is often excessive pressure on the inner arc side of the curve and the capping block during segment assembly, resulting in damage to the corners of the segments, which brings hidden dangers to tunnel operation and may shorten the service life of the tunnel accordingly.

[0003] At present, the main engineering difficulties in the assembly construction of curved shield tunnel segments include: 1) Since it is difficult to control the axis of the curved shield tunnel, one end of the segment is prone to severe squeezing after deviating from the designed axis; 2) During the assembly process of the curved shield tunnel, the jack thrust generates a horizontal force, and the segment is prone to large displacement under the action of the horizontal component force.

[0004] During the segment assembly process of shield tunnels, we often manually measure the opening of the segment annular joints and control the segment assembly process based on experience to prevent local excessive stress and local damage. However, it is easy to lose control during the actual construction process. Therefore, it is urgent to develop an intelligent control system for the assembly quality of shield tunnel joints to achieve intelligent perception and control of the shield tunnel assembly process, so as to ensure the construction quality of shield tunnels and safety during the operation stage. Summary of the invention

[0005] The purpose of the present invention is to provide a prefabricated underground structure construction assembly quality control system that solves the problems of pipe segment damage and the like encountered during the current curved shield tunnel construction assembly process.

[0006] Another object of the present invention is to provide an assembly quality control method implemented by using the above-mentioned prefabricated underground structure construction assembly quality control system.

[0007] To this end, the technical solution of the present invention is as follows:

[0008] A prefabricated underground structure construction assembly quality control system, comprising:

[0009] The piezoelectric sensing experimental device for shield tunnel joints is used to determine the mathematical relationship model between the piezoelectric signal and the opening amount of the segment joints;

[0010] A plurality of flexible piezoelectric sensing devices are arranged one by one between each two adjacent shield tunnel segments to monitor in real time the pressure point sensing signals between the shield tunnel segments to obtain the joint opening amount between the shield tunnel segments at the current moment;

[0011] A shield tunnel joint prediction device, which is a shield tunnel segment joint opening prediction model constructed and trained based on a BP neural network deep learning model, so as to obtain a predicted value of the joint opening between each shield tunnel segment at the next moment according to the joint opening between each shield tunnel segment and construction parameters at the current moment; and

[0012] The control system monitors in real time whether the piezoelectric signals transmitted by multiple flexible piezoelectric sensing devices meet the normal construction requirements, and optimizes the construction parameters in real time according to the prediction results of the shield tunnel joint prediction device.

[0013] Furthermore, the piezoelectric sensing experimental device for shield tunnel joints is composed of an electronic universal testing machine and a steel mold; the steel mold is composed of two steel pressing plates, and strip grooves with built-in rubber water stop strips are symmetrically opened on the contacting surfaces of the two; wherein the rubber water stop strip is the same as the rubber water stop strip arranged on the cross-section of the shield tunnel pipe segment, and its arrangement method in the strip groove is the same as its arrangement method on the cross-section of the shield tunnel pipe segment; the two steel pressing plates are symmetrically fixed on the bottom surface of the upper pressure head and the top surface of the lower pressure head of the electronic universal testing machine.

[0014] Furthermore, the flexible piezoelectric sensing device is composed of two flexible piezoelectric cable sensors and two signal processors; the signal processor is composed of a charge amplifier, a data collector and a wireless data transmitter which are arranged in a box and connected in sequence; each flexible piezoelectric cable sensor is connected to the charge amplifier in the corresponding signal processor through a wire.

[0015] Furthermore, the flexible piezoelectric cable sensor adopts a flexible piezoelectric cable with a length of 10 cm; the wireless data transmitter adopts Bluetooth or GPS to connect with the control end.

[0016] A quality control method implemented by using the above-mentioned prefabricated underground structure construction assembly quality control system comprises the following steps:

[0017] S1. Construct the same environmental conditions as the shield tunnel segment assembly process in the laboratory, and conduct a shield tunnel joint piezoelectric sensing experiment to determine the mathematical relationship model I between the piezoelectric signal on the rubber water stop and the segment joint opening, and 2) the mathematical relationship model II between the piezoelectric signal on the segment section and the segment joint opening;

[0018] S2. According to the maximum and minimum values ​​of the allowable opening of the segment joints, the segment joint opening is linearly normalized, and based on the normalized results of the segment joint opening, a plurality of warning levels of the shield tunnel joint opening are divided; based on the mathematical relationship model I and the mathematical relationship model II determined in step S1, the piezoelectric signal range corresponding to each warning level is determined;

[0019] S3. Before the construction of the shield tunnel, two pairs of flexible piezoelectric sensing devices are arranged between the longitudinal seam sections of each two adjacent shield tunnel segments; two groups of flexible piezoelectric sensing devices are symmetrically arranged in the horizontal direction and the vertical direction between the ring seam sections of the segments between each two adjacent shield tunnel rings, and each group of flexible piezoelectric sensing devices is composed of two pairs of flexible piezoelectric sensing devices; among which, in the two pairs of flexible piezoelectric sensing devices between each adjacent section, the flexible piezoelectric cable sensors of one pair of flexible piezoelectric sensing devices are fixed on the rubber water stop strip of the section, and the flexible piezoelectric cable sensors of the other pair of flexible piezoelectric sensing devices are fixed on the section through the rubber force transfer pad; the signal processor of each flexible piezoelectric sensing device is arranged on the outside of the shield tunnel segment;

[0020] S4. Based on the deep learning model of BP neural network, a prediction model for the opening amount of shield tunnel segments joints is constructed and trained, so as to obtain the predicted value of the opening amount of the joints between the shield tunnel segments at the next moment by inputting the opening amount of the joints between the shield tunnel segments at the current moment and the construction parameters;

[0021] S5. During the construction of the shield tunnel, the piezoelectric signals between the shield tunnel segments are obtained in real time by the flexible piezoelectric sensing devices arranged in step S3, and the joint opening between the shield tunnel segments is obtained according to the pressure point signal, so as to determine the warning level according to the joint opening between the shield tunnel segments at the current moment, and confirm whether the current assembly state is normal; at the same time, the shield tunnel segment joint opening prediction model of step S4 is used to obtain the predicted value of the joint opening between the shield tunnel segments at the next moment by inputting the joint opening between the shield tunnel segments at the current moment and the current construction parameters, so as to determine the warning level according to the joint opening between the shield tunnel segments at the next moment, and confirm whether the assembly state is normal at the next moment; if the current assembly state or the confirmation result of the assembly state at the next moment is an abnormal state, the construction parameters are adjusted according to different warning levels until the assembly state corresponding to the predicted value of the joint opening between the shield tunnel segments at the next moment is normal.

[0022] Furthermore, in step S1, the specific implementation steps of the shield tunnel joint piezoelectric sensing experiment are:

[0023] S101. Build a piezoelectric sensing experimental device for shield tunnel joints, which is composed of an electronic universal testing machine and a steel mold; the steel mold is composed of two steel pressing plates, and the surfaces in contact with each other are symmetrically provided with strip grooves with built-in rubber water stop strips; wherein the rubber water stop strip is the same as the rubber water stop strip arranged on the cross section of the shield tunnel pipe segment, and the arrangement method of the rubber water stop strip in the strip groove is the same as the arrangement method of the rubber water stop strip on the cross section of the shield tunnel pipe segment; the two steel pressing plates are symmetrically fixed on the bottom surface of the upper pressure head and the top surface of the lower pressure head of the electronic universal testing machine;

[0024] S102, fixing two flexible piezoelectric cable sensors of a flexible piezoelectric sensing device between rubber water stop strips on two steel pressing sheets in the same setting as between shield tunnel segments; controlling a uniaxial pressurizing device connected to an upper pressure head to gradually move downward in the axial direction, so as to use two steel pressing sheets to simulate the process from opening to closing between shield tunnel segments, and at the same time monitoring in real time the change in the spacing between the two steel pressing sheets during the process, and the change in the piezoelectric signal collected by the flexible piezoelectric sensing device;

[0025] S103, repeating step S102, taking the joint between the two steel pressing plates from fully opened to fully closed as the test benchmark, conducting 3 to 5 indoor compression tests with different joint opening amounts, so as to determine the mathematical relationship between the piezoelectric signal and the joint opening amount.

[0026] Furthermore, in step S2, the linear normalization processing formula of the segment joint opening amount is: Where x' is the normalized segment joint opening, x is the original segment joint opening, and x max is the maximum allowable opening of the segment joint, x min It is the minimum allowable opening of the segment joint.

[0027] Further, in step S2, the four warning levels of shield tunnel joint opening are divided into:

[0028] When x'<0 or x'>1, the shield tunnel joint opening level is particularly severe;

[0029] When 0<x'<0.15 or 0.85<x'<1, the shield tunnel joint opening level is severe;

[0030] When 0.15<x'<0.3 or 0.7<x'<0.85, the shield tunnel joint opening level is heavier;

[0031] When 0.3<x'<0.7, the shield tunnel joint opening level is normal.

[0032] Further, in step S3, the two sets of flexible piezoelectric sensing devices between adjacent sections are arranged as follows:

[0033] 1) For two groups of flexible piezoelectric sensing devices arranged between the longitudinal seam sections of two adjacent shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the longitudinal seam section of a shield tunnel segment in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the length direction of the shield tunnel segment, and the two flexible piezoelectric sensing devices of the second group are centrally fixed on the longitudinal seam section of the same shield tunnel segment through a rubber force transfer pad in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the width direction of the shield tunnel segment;

[0034] 2) For the two groups of flexible piezoelectric sensing devices arranged on the annular seam sections of two shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the annular seam section of a shield tunnel segment in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the circumferential direction of the shield tunnel pipe; the two flexible piezoelectric sensing devices of the second group are symmetrically fixed on both sides of the annular seam section of the same shield tunnel segment through rubber force transmission pads in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along a direction perpendicular to the two flexible piezoelectric cable sensors of the first group.

[0035] Furthermore, in step S4, the shield tunnel segment joint opening prediction model consists of an input layer, a hidden layer and an output layer; the input layer consists of ten neurons, the hidden layer consists of ten neurons, and the output layer consists of one neuron; the calculation expression of the relationship between the neurons in each layer is:

[0036]

[0037]

[0038] In the formula, is the (l-1)th th Layer i th The output value of a neuron (l can be 2, 3, i = 1, ..., 10), where The first neuron in the input layer corresponds to the maximum thrust of the jack. is the jack loading rate corresponding to the second neuron in the input layer, The third neuron in the input layer corresponds to the over-digging amount of the cutter head. is the grouting parameter corresponding to the fourth neuron in the input layer, The 5th neuron in the input layer corresponds to the segment assembly point. is the tunnel depth corresponding to the sixth neuron in the input layer, is the stratum density corresponding to the 7th neuron in the input layer, is the Poisson's ratio of the 8th neuron in the input layer, is the compression modulus of the stratum corresponding to the 9th neuron in the input layer, The seam type (vertical / circular) corresponding to the 10th neuron in the input layer; is the hidden layer (i=1,…,10), which has the same number of neurons as the input layer, a total of 10 neurons; is the output layer, which contains 1 neuron, corresponding to the opening of the pipe joint. m is the lth th The number of neurons in the layer, It is the first th Layer j th The neurons and the (l-1)th th Layer i th The connection weights between neurons, Is with the first th Layer j th The bias value associated with each neuron, and Can be continuously optimized through on-site construction data,γ l j It is the first th Layer j th The initial linear value of each neuron, f() is the activation function, specifically the hyperbolic tangent function.

[0039] Furthermore, in step S4, the learning samples used to train the shield tunnel segment joint opening prediction model are taken from the construction data of similar projects and / or the construction data of the previous construction section in the current project.

[0040] Compared with the prior art, the beneficial effects of the assembly quality control system for prefabricated underground structure construction are:

[0041] (1) The system and method successively conduct piezoelectric sensing experiments on shield tunnel joints to determine the mathematical relationship model between the piezoelectric signal and the opening of the segment joints, determine the warning level of the segment joint opening according to the maximum and minimum values ​​of the allowable opening of the segment joints, determine the relationship between the warning level and the piezoelectric signal variable, and achieve real-time sensing of the mechanical response of the segments during the shield tunnel assembly process to determine the safe state of the shield tunnel segment joint opening, thereby providing a reasonable basis for the subsequent adjustment of shield tunnel construction parameters and the real-time monitoring and control of the segment assembly status.

[0042] (2) The system and method can combine the construction parameters of similar projects and previous construction sections to effectively predict the opening of the subsequent segment assembly joints, optimize and adjust the construction parameters in a timely manner in response to unfavorable conditions until the deduced results meet the construction standards, and provide recommended construction parameters to effectively assist construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1It is a flow chart of the method for controlling the quality of construction and assembly of a prefabricated underground structure of the present invention;

[0044] Figure 2 It is a schematic diagram of the arrangement of two flexible piezoelectric sensing devices on the segment annular seam sections between two adjacent shield tunnel rings in step S3 of the method for controlling the quality of construction and assembly of an assembled underground structure of the present invention;

[0045] Figure 3 It is a schematic diagram of the arrangement of the flexible piezoelectric sensing device on the longitudinal seam sections of each of two adjacent shield tunnel segments in step S3 of the method for controlling the quality of construction and assembly of an assembled underground structure of the present invention;

[0046] Figure 4 It is a schematic diagram of the arrangement of the flexible piezoelectric sensing device between the pipe segments on a single shield tunnel pipe in step S3 of the method for controlling the quality of construction and assembly of an assembled underground structure of the present invention;

[0047] Figure 5 It is a flow chart of step S4 of the method for controlling the quality of construction and assembly of a prefabricated underground structure of the present invention;

[0048] Figure 6 This is a flow chart of step S5 of the method for controlling the quality of construction and assembly of a prefabricated underground structure of the present invention. DETAILED DESCRIPTION

[0049] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but the following embodiments are by no means intended to limit the present invention in any way.

[0050] Example 1

[0051] The prefabricated underground structure construction assembly quality control system includes a shield tunnel joint piezoelectric sensing experimental device, multiple flexible piezoelectric sensing devices, a shield tunnel joint prediction device and a control system.

[0052] The piezoelectric sensing experimental device for shield tunnel joints is used to determine the mathematical relationship model between the piezoelectric signal and the opening amount of the pipe segment joints; specifically, the piezoelectric sensing experimental device for shield tunnel joints is composed of an electronic universal testing machine and a steel mold; the steel mold is composed of two steel pressing plates, and the contacting surfaces of the two are symmetrically provided with strip grooves with built-in rubber water stop strips; wherein, the rubber water stop strip is the same as the rubber water stop strip arranged on the cross section of the shield tunnel pipe segment, and the arrangement method of the strip groove is the same as the arrangement method of the strip groove on the cross section of the shield tunnel pipe segment; the two steel pressing plates are arranged on the cross section of the shield tunnel pipe segment. It is fixed on the bottom surface of the upper pressure head and the top surface of the lower pressure head of the electronic universal testing machine; when in use, two flexible piezoelectric cable sensors of a flexible piezoelectric sensing device are fixed between the rubber water stop strips on the two steel pressing plates in the same setting as between the shield tunnel segments; the uniaxial pressurizing device connected to the upper pressure head is controlled to gradually move downward along the axial direction, so as to use the two steel pressing plates to simulate the process from opening to closing between the shield tunnel segments, and at the same time, the change in the spacing between the two steel pressing plates during the process and the change in the piezoelectric signal collected by the flexible piezoelectric sensing device are monitored in real time.

[0053] Multiple flexible piezoelectric sensing devices are arranged one by one between each two adjacent shield tunnel segments to monitor the pressure point sensing signals between the shield tunnel segments in real time, so as to obtain the joint opening amount between the shield tunnel segments at the current moment; specifically, see Figure 2 to Figure 4 The flexible piezoelectric sensing device is used to sense the annular seam opening or longitudinal seam opening of the shield tunnel segment 1 in real time; specifically, the flexible piezoelectric sensing device is composed of two flexible piezoelectric cable sensors 2 and two signal processors 6; wherein, the flexible piezoelectric cable sensor 2 adopts a flexible piezoelectric cable with a length of 10 cm; the signal processor 6 is composed of a charge amplifier, a data collector and a wireless data transmitter which are arranged in a box and connected in sequence; each flexible piezoelectric cable sensor 2 is connected to the charge amplifier in the corresponding signal processor 6 through a wire 5; the wireless data transmitter specifically adopts Bluetooth, which realizes data transmission with the control system located at the construction site through Bluetooth; in actual applications, the number of flexible piezoelectric sensing devices is adapted to the number of shield tunnel segments 1.

[0054] The shield tunnel joint prediction device is a shield tunnel segment joint opening prediction model built and trained based on the BP neural network deep learning model, which can obtain the predicted value of the joint opening between each shield tunnel segment at the next moment according to the joint opening between each shield tunnel segment and the construction parameters at the current moment.

[0055] The control system is used to monitor in real time whether the piezoelectric signals transmitted by multiple flexible piezoelectric sensing devices meet the normal construction requirements, and optimize the construction in real time according to the prediction results of the shield tunnel joint prediction device.

[0056] Example 2

[0057] See also Figure 1 The specific implementation steps of the construction assembly quality control method implemented by the prefabricated underground structure construction assembly quality control system of Example 1 are as follows:

[0058] S1. Construct the same environmental conditions as those in the shield tunnel segment assembly process in the laboratory. Through the piezoelectric sensing experiment of the shield tunnel joints, determine the following mathematical relationship model I between the piezoelectric signal on the rubber water stop and the segment joint opening amount, and 2) the mathematical relationship model II between the piezoelectric signal on the segment cross section and the segment joint opening amount.

[0059] In step S1, the specific implementation steps of the shield tunnel joint piezoelectric sensing experiment are:

[0060] S101. Construct a piezoelectric sensing experimental device for shield tunnel joints; specifically, the joint opening evaluation device consists of an electronic universal testing machine and a steel mold; the electronic universal testing machine specifically adopts a 100kN microcomputer-controlled electronic universal testing machine-UTM5105; the steel mold is used to simulate the shield tunnel joint, which is composed of two steel pressing plates, and the surfaces of the two contacting each other are symmetrically provided with strip grooves with built-in rubber water stop strips; specifically, the rubber water stop strip is the same as the rubber water stop strip arranged on the cross section of the shield tunnel segment, and the arrangement method of the rubber water stop strip in the strip groove is the same as the arrangement method of the rubber water stop strip on the cross section of the shield tunnel segment; the electronic universal testing machine is used to control the opening of the shield tunnel joint, and the two steel pressing plates are symmetrically fixed on the bottom surface of the upper pressure head and the top surface of the lower pressure head of the electronic universal testing machine, and the top of the upper pressure head is connected to a uniaxial pressurizing device, and the uniaxial pressurizing device is controlled to gradually move downward along the axial direction to simulate the assembly process between the shield tunnel segments;

[0061] S102, fixing two flexible piezoelectric cable sensors of a flexible piezoelectric sensing device between rubber water stop strips on two steel pressing sheets in the same setting as between shield tunnel segments; controlling a uniaxial pressurizing device connected to an upper pressure head to gradually move downward in the axial direction, so as to use two steel pressing sheets to simulate the process from opening to closing between shield tunnel segments, and at the same time monitoring in real time the change in the spacing between the two steel pressing sheets during the process, and the change in the piezoelectric signal collected by the flexible piezoelectric sensing device;

[0062] S103, repeating step S102, taking the joint between the two steel pressing plates from fully opened to fully closed as the test benchmark, conducting 3 to 5 indoor compression tests with different joint opening amounts, so as to determine the mathematical relationship between the piezoelectric signal and the joint opening amount.

[0063] S2. According to the maximum and minimum values ​​of the allowable opening of the segment joints, the segment joint opening is linearly normalized, and based on the normalized results of the segment joint opening, multiple warning levels of the shield tunnel joint opening are divided; based on the mathematical relationship model I and the mathematical relationship model II determined by step S1, the piezoelectric signal range corresponding to each warning level is determined.

[0064] The specific implementation steps of step S2 are:

[0065] S201. Determine the maximum and minimum values ​​of the allowable opening of the segment joints according to the construction requirements;

[0066] S202. The linear normalization formula for the segment joint opening is: Where x' is the normalized segment joint opening, x is the original segment joint opening, and x max is the maximum allowable opening of the segment joint, x min The minimum allowable opening of the segment joint;

[0067] S203. Based on the normalized result of the segment joint opening amount, four warning levels of the shield tunnel joint opening amount are divided, including extremely serious, serious, relatively serious and normal.

[0068] S3. Before the construction of the shield tunnel, two pairs of flexible piezoelectric sensing devices are arranged between the longitudinal seam sections of each two adjacent shield tunnel segments; two groups of flexible piezoelectric sensing devices are symmetrically arranged in the horizontal direction and the vertical direction respectively between the annular seam sections of the segments between each two adjacent shield tunnel rings, and each group of flexible piezoelectric sensing devices is composed of two pairs of flexible piezoelectric sensing devices; among which, in the two pairs of flexible piezoelectric sensing devices between each adjacent section, the flexible piezoelectric cable sensors of one pair of flexible piezoelectric sensing devices are fixed on the rubber water stop strip of the section, and the flexible piezoelectric cable sensors of the other pair of flexible piezoelectric sensing devices are fixed on the section through rubber force transfer pads; the signal processor of each flexible piezoelectric sensing device is arranged on the outside of the shield tunnel segment.

[0069] In step S3, the two sets of flexible piezoelectric sensing devices between two adjacent sections are arranged as follows:

[0070] 1) See Figure 3, for two groups of flexible piezoelectric sensing devices arranged between the longitudinal seam sections of two adjacent shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the longitudinal seam section of a shield tunnel segment in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the length direction of the shield tunnel segment, and the two flexible piezoelectric sensing devices of the second group are centrally fixed on the longitudinal seam section of the same shield tunnel segment through a rubber force transfer pad in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the width direction of the shield tunnel segment. Specifically, the two flexible piezoelectric sensing devices of the second group are located between the two openings 4 on the longitudinal seam section of the shield tunnel segment;

[0071] 2) See Figure 2 For the two groups of flexible piezoelectric sensing devices arranged on the annular seam sections of two shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the annular seam section of a shield tunnel segment in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the circumferential direction of the shield tunnel pipe; the two flexible piezoelectric sensing devices of the second group are symmetrically fixed on both sides of the annular seam section of the same shield tunnel segment through rubber force transmission pads in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along a direction perpendicular to the two flexible piezoelectric cable sensors of the first group. Specifically, the two flexible piezoelectric sensing devices of the second group are located on both sides of the opening 4 on the annular seam section of the shield tunnel segment.

[0072] Specifically, Figure 4 For example, based on the segmentation of a single shield tunnel segment ring into six segments, there are four groups of flexible piezoelectric sensing devices 7 for sensing the annular seam opening, which are symmetrically arranged in the horizontal and vertical directions; and there are six flexible piezoelectric sensing devices 8 for sensing the longitudinal seam opening, which are respectively arranged on the longitudinal seam sections of each two adjacent segments. In this embodiment, the rubber water stop strip is 15mm thick and arranged in the groove with a groove depth of about 7.5mm. The other two flexible piezoelectric cable sensors are symmetrically fixed on the shield rubber force transfer pads on both sides of the bolt hole, and the force transfer pads are about 3mm thick.

[0073] The flexible piezoelectric cable sensor is used to collect force change signals between segments and output piezoelectric signals. The piezoelectric signal changes are used to determine the annular gap or longitudinal gap between segments, and then determine the assembly quality of shield tunnel segments. In practical applications, repeaters and gateways can be deployed inside the shield tunnel as needed to ensure that the data collected by the wireless data collector is transmitted to the control system server in real time.

[0074] S4. Based on the deep learning model of BP neural network, a prediction model for the opening amount of shield tunnel segments joints is constructed and trained, so as to obtain the predicted value of the opening amount of the joints between the shield tunnel segments at the next moment by inputting the opening amount of the joints between the shield tunnel segments at the current moment and the construction parameters;

[0075] In step S4, the shield tunnel segment joint opening prediction model is constructed to predict the joint opening of subsequent segments according to the current real-time splicing status of the segments during the construction phase, so as to adjust the corresponding construction parameters in time to ensure that the joint opening meets the construction safety requirements;

[0076] See also Figure 5 The shield tunnel segment joint opening prediction model consists of an input layer, a hidden layer and an output layer; the input layer consists of ten neurons, each of which serves as a data input end, corresponding to the maximum jack thrust, jack loading rate, cutter head over-excavation, grouting parameters, segment assembly points, tunnel burial depth, stratum density, stratum Poisson's ratio, stratum compression modulus, and joint type (i.e., longitudinal joint or annular joint); the hidden layer also consists of ten neurons, which are used to improve the model learning and computing capabilities. Each neuron has no actual physical meaning and is an intermediate quantity in the calculation process; the output layer consists of one neuron, which is used to output the segment joint opening.

[0077] The computational expression of the relationship between neurons in each layer is:

[0078]

[0079]

[0080] In the formula, is the (l-1)th th Layer i th The output value of a neuron (l can be 2, 3, i = 1, ..., 10), where The first neuron in the input layer corresponds to the maximum thrust of the jack. is the jack loading rate corresponding to the second neuron in the input layer, The third neuron in the input layer corresponds to the over-digging amount of the cutter head. is the grouting parameter corresponding to the fourth neuron in the input layer, The 5th neuron in the input layer corresponds to the segment assembly point. is the tunnel depth corresponding to the sixth neuron in the input layer, is the stratum density corresponding to the 7th neuron in the input layer, is the Poisson's ratio of the 8th neuron in the input layer, is the compression modulus of the stratum corresponding to the 9th neuron in the input layer, The seam type (vertical / circular) corresponding to the 10th neuron in the input layer; is the hidden layer (i=1,…,10), which has the same number of neurons as the input layer, a total of 10 neurons; is the output layer, which contains 1 neuron, corresponding to the opening of the pipe joint. m is the lth th The number of neurons in the layer, It is the first th Layer j th The neurons and the (l-1)th th Layer i th The connection weights between neurons, Is with the first th Layer j th The bias value associated with each neuron, and It can be continuously optimized through on-site construction data. It is the first th Layer j th The initial linear value of each neuron, f() is the activation function, and the hyperbolic tangent function (tanh) is selected.

[0081] In practical applications, the learning samples of the intelligent prediction model for joint opening are taken from similar projects and the previous construction sections of the current project, that is, the joint opening of subsequent segments is predicted based on the real-time construction parameters of the current project.

[0082] S5. During the construction of the shield tunnel, the piezoelectric signals between the shield tunnel segments are obtained in real time by the flexible piezoelectric sensing devices arranged in step S3, and the joint opening between the shield tunnel segments is obtained according to the pressure point signal, so as to determine the warning level according to the joint opening between the shield tunnel segments at the current moment, and confirm whether the current assembly state is normal; at the same time, the shield tunnel segment joint opening prediction model of step S4 is used to obtain the predicted value of the joint opening between the shield tunnel segments at the next moment by inputting the joint opening between the shield tunnel segments at the current moment and the current construction parameters, so as to determine the warning level according to the joint opening between the shield tunnel segments at the next moment, and confirm whether the assembly state is normal at the next moment; if the current assembly state or the confirmation result of the assembly state at the next moment is an abnormal state, the construction parameters are adjusted according to different warning levels until the assembly state corresponding to the predicted value of the joint opening between the shield tunnel segments at the next moment is normal.

[0083] Specifically, see Figure 6Based on the real-time construction parameters, the shield tunnel segment joint opening amount prediction model of step S4 is used to predict the segment joint opening amount at the next moment. When the assembly status prediction result is not normal, the construction parameters of the maximum jack thrust, jack loading rate, cutter head over-excavation and grouting parameters are optimized according to the severity level to make the assembly status prediction result normal, and the control end adjusts the actual construction parameters according to the optimized construction parameters to ensure that the shield tunnel segment assembly state at the next moment is normal.

Claims

1. A quality control system for the construction and assembly of prefabricated underground structures, characterized in that: include: The piezoelectric sensing experimental device for shield tunnel joints is used to determine the mathematical relationship model between the piezoelectric signal and the opening amount of the segment joints; A plurality of flexible piezoelectric sensing devices are arranged one by one between each two adjacent shield tunnel segments to monitor in real time the pressure point sensing signals between the shield tunnel segments to obtain the joint opening amount between the shield tunnel segments at the current moment; A shield tunnel joint prediction device, which is a shield tunnel segment joint opening prediction model built and trained based on a BP neural network deep learning model, so as to obtain a predicted value of the joint opening between each shield tunnel segment at the next moment according to the joint opening between each shield tunnel segment and the construction parameters at the current moment; and The control system monitors in real time whether the piezoelectric signals transmitted by multiple flexible piezoelectric sensing devices meet the normal construction requirements, and optimizes the construction parameters in real time according to the prediction results of the shield tunnel joint prediction device.

2. The assembly quality control system for prefabricated underground structure construction according to claim 1 is characterized in that: The flexible piezoelectric sensing device consists of two flexible piezoelectric cable sensors and two signal processors; the signal processor consists of a charge amplifier, a data collector and a wireless data transmitter which are arranged in a box and connected in sequence; each flexible piezoelectric cable sensor is connected to the charge amplifier in the corresponding signal processor through a wire.

3. The assembly quality control system for prefabricated underground structure construction according to claim 1 is characterized in that: The flexible piezoelectric cable sensor adopts a flexible piezoelectric cable with a length of 10 cm; the wireless data transmitter adopts Bluetooth or GPS to connect with the control end.

4. A quality control method implemented by using the assembly quality control system for prefabricated underground structure construction as described in any one of claims 1 to 3, characterized in that: Here are the steps: S1. The same environmental conditions as the shield tunnel segment assembly process were constructed in the laboratory. Through the shield tunnel joint piezoelectric sensing experiment, the following mathematical relationship models were determined between adjacent segments: 1) the piezoelectric signal on the rubber water stop strip and the segment joint opening amount, and 2) the mathematical relationship model II between the piezoelectric signal on the segment section and the segment joint opening amount. S2. According to the maximum and minimum values ​​of the allowable opening of the segment joints, the segment joint opening is linearly normalized, and multiple warning levels of the shield tunnel joint opening are divided based on the normalized results of the segment joint opening; Based on the mathematical relationship model I and the mathematical relationship model II determined in step S1, the piezoelectric signal range corresponding to each warning level is determined; S3. Before the construction of the shield tunnel, two pairs of flexible piezoelectric sensing devices are arranged between the longitudinal seam sections of each two adjacent shield tunnel segments; two groups of flexible piezoelectric sensing devices are symmetrically arranged in the horizontal direction and the vertical direction between the ring seam sections of the segments between each two adjacent shield tunnel rings, and each group of flexible piezoelectric sensing devices is composed of two pairs of flexible piezoelectric sensing devices; among which, in the two pairs of flexible piezoelectric sensing devices between each adjacent section, the flexible piezoelectric cable sensors of one pair of flexible piezoelectric sensing devices are fixed on the rubber water stop strip of the section, and the flexible piezoelectric cable sensors of the other pair of flexible piezoelectric sensing devices are fixed on the section through the rubber force transfer pad; the signal processor of each flexible piezoelectric sensing device is arranged on the outside of the shield tunnel segment; S4. Based on the deep learning model of BP neural network, a prediction model for the opening amount of shield tunnel segments joints is constructed and trained, so as to obtain the predicted value of the opening amount of the joints between the shield tunnel segments at the next moment by inputting the opening amount of the joints between the shield tunnel segments at the current moment and the construction parameters; S5. During the construction of the shield tunnel, the piezoelectric signals between the shield tunnel segments are obtained in real time by the flexible piezoelectric sensing devices arranged in step S3, and the joint opening between the shield tunnel segments is obtained according to the pressure point signal, so as to determine the warning level according to the joint opening between the shield tunnel segments at the current moment, and confirm whether the current assembly state is normal; at the same time, the shield tunnel segment joint opening prediction model of step S4 is used to obtain the predicted value of the joint opening between the shield tunnel segments at the next moment by inputting the joint opening between the shield tunnel segments at the current moment and the current construction parameters, so as to determine the warning level according to the joint opening between the shield tunnel segments at the next moment, and confirm whether the assembly state is normal at the next moment; if the current assembly state or the confirmation result of the assembly state at the next moment is an abnormal state, the construction parameters are adjusted according to different warning levels until the assembly state corresponding to the predicted value of the joint opening between the shield tunnel segments at the next moment is normal.

5. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 4, characterized in that: In step S1, the specific implementation steps of the shield tunnel joint piezoelectric sensing experiment are as follows: S101. Build a piezoelectric sensing experimental device for shield tunnel joints, which is composed of an electronic universal testing machine and a steel mold; the steel mold is composed of two steel pressing plates, and the surfaces in contact with each other are symmetrically provided with strip grooves with built-in rubber water stop strips; wherein the rubber water stop strip is the same as the rubber water stop strip arranged on the cross section of the shield tunnel pipe segment, and the arrangement method of the rubber water stop strip in the strip groove is the same as the arrangement method of the rubber water stop strip on the cross section of the shield tunnel pipe segment; the two steel pressing plates are symmetrically fixed on the bottom surface of the upper pressure head and the top surface of the lower pressure head of the electronic universal testing machine; S102, fixing two flexible piezoelectric cable sensors of a flexible piezoelectric sensing device between rubber water stop strips on two steel pressing sheets in the same setting as between shield tunnel segments; controlling a uniaxial pressurizing device connected to an upper pressure head to gradually move downward in the axial direction, so as to use two steel pressing sheets to simulate the process from opening to closing between shield tunnel segments, and at the same time monitoring in real time the change in the spacing between the two steel pressing sheets during the process, and the change in the piezoelectric signal collected by the flexible piezoelectric sensing device; S103, repeating step S102, taking the joint between the two steel pressing plates from fully opened to fully closed as the test benchmark, conducting 3 to 5 indoor compression tests with different joint opening amounts, so as to determine the mathematical relationship between the piezoelectric signal and the joint opening amount.

6. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 4, characterized in that: In step S2, the linear normalization processing formula of the segment joint opening is: Where x' is the normalized segment joint opening, x is the original segment joint opening, and x max is the maximum allowable opening of the segment joint, x min It is the minimum allowable opening of the segment joint.

7. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 4, characterized in that: In step S2, the four warning levels of shield tunnel joint opening are divided into: When x'<0 or x'>1, the shield tunnel joint opening level is particularly severe; When 0<x'<0.15 or 0.85<x'<1, the shield tunnel joint opening level is severe; When 0.15<x'<0.3 or 0.7<x'<0.85, the shield tunnel joint opening level is heavier; When 0.3<x'<0.7, the shield tunnel joint opening level is normal.

8. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 4, characterized in that: In step S3, the two sets of flexible piezoelectric sensing devices between two adjacent sections are arranged as follows: 1) For two groups of flexible piezoelectric sensing devices arranged between the longitudinal seam sections of two adjacent shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the longitudinal seam section of a shield tunnel segment in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the length direction of the shield tunnel segment, and the two flexible piezoelectric sensing devices of the second group are centrally fixed on the longitudinal seam section of the same shield tunnel segment through a rubber force transfer pad in a manner that the two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the width direction of the shield tunnel segment; 2) For the two groups of flexible piezoelectric sensing devices arranged on the annular seam sections of two shield tunnel segments, the two flexible piezoelectric sensing devices of the first group are centrally fixed on the rubber water stop strip of the annular seam section of a shield tunnel segment in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along the circumferential direction of the shield tunnel pipe; the two flexible piezoelectric sensing devices of the second group are symmetrically fixed on both sides of the annular seam section of the same shield tunnel segment through rubber force transmission pads in a manner that their two flexible piezoelectric cable sensors are arranged in parallel and at intervals along a direction perpendicular to the two flexible piezoelectric cable sensors of the first group.

9. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 4, characterized in that: In step S4, the shield tunnel segment joint opening prediction model consists of an input layer, a hidden layer and an output layer; the input layer consists of ten neurons, the hidden layer consists of ten neurons, and the output layer consists of one neuron; the calculation expression of the relationship between neurons in each layer is: In the formula, is the (l-1)th th Layer i th The output value of a neuron (l can be 2, 3, i = 1, ..., 10), where The first neuron in the input layer corresponds to the maximum thrust of the jack. is the jack loading rate corresponding to the second neuron in the input layer, The third neuron in the input layer corresponds to the over-digging amount of the cutter head. is the grouting parameter corresponding to the fourth neuron in the input layer, The 5th neuron in the input layer corresponds to the segment assembly point. is the tunnel depth corresponding to the sixth neuron in the input layer, is the stratum density corresponding to the 7th neuron in the input layer, is the Poisson's ratio of the 8th neuron in the input layer, is the compression modulus of the stratum corresponding to the 9th neuron in the input layer, The seam type (vertical / circular) corresponding to the 10th neuron in the input layer; is the hidden layer (i=1,…,10), which has the same number of neurons as the input layer, a total of 10 neurons; is the output layer, which contains 1 neuron, corresponding to the opening of the pipe joint. m is the lth th The number of neurons in the layer, It is the first th Layer j th The neurons and the (l-1)th th Layer i th The connection weights between neurons are Is with the first th Layer j th The bias value associated with each neuron, and Can be continuously optimized through on-site construction data, Y l j It is the first th Layer j th The initial linear value of each neuron, f() is the activation function, specifically the hyperbolic tangent function.

10. The method for controlling the quality of construction and assembly of an assembled underground structure according to claim 9, characterized in that: In step S4, the learning samples used to train the shield tunnel segment joint opening prediction model are taken from the construction data of similar projects and / or the construction data of the previous construction section in the current project.

Images