A bim-based cable construction method

Abstract

The application discloses a kind of cable construction methods based on BIM, comprising: obtaining BIM model data including cable design path, design bending radius and design stress;In the process of cable laying, the optical fiber sensor is fixedly attached to the outer sheath surface of the cable to be laid, and the optical fiber sensor is continuously distributed along the length direction of the cable;Real-time acquisition of the actual strain distribution data and the actual bending radius distribution data measured by the optical fiber sensor, and comparison with the design stress data, the design bending radius data in the BIM model;If the actual strain value of any position point exceeds the strain threshold range corresponding to the design stress, or the actual bending radius value is less than the design bending radius value, it is determined that the position point has construction deviation;Based on the position point information with construction deviation, generate adjustment instructions for subsequent cable laying operation.The application can effectively improve the cable laying accuracy and construction safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cable construction. More particularly, the present application relates to a cable construction method based on BIM. BACKGROUND

[0002] In the process of cable construction, the bending radius and stress state of the cable need to be strictly controlled to avoid damage to the cable insulation layer, shorten the service life, or even safety accidents caused by excessive bending or excessive stress. In the existing construction, the monitoring of the actual bending radius and strain of the cable mainly relies on manual inspection or discrete point measurement. Manual inspection relies on experience and is highly subjective, and it is difficult to capture continuous distribution of subtle deviations. Although discrete point measurement can obtain local data, it cannot cover the entire length of the cable and is prone to miss abnormalities at key positions. Moreover, the actual measurement data in the construction and the cable path, bending radius and stress parameters in the design stage lack effective spatial position correlation. The design parameters usually exist in the form of drawings or models, and the construction personnel need to manually compare the design data to determine whether the measured values are compliant, which is time-consuming and prone to misjudgment due to spatial correspondence deviation. When the actual bending radius is smaller than the design value or the actual strain exceeds the allowable range, it is often difficult to find out in time, and when the problem is found in the subsequent detection, a certain construction deviation has been caused, which requires rework adjustment, increasing the construction cost and the time pressure.

[0003] Therefore, it is necessary to design a technical solution that can overcome the above-mentioned defects to some extent. SUMMARY

[0004] An object of the present application is to provide a cable construction method based on BIM, which can effectively improve the cable laying accuracy and construction safety.

[0005] In order to achieve these objects and other advantages of the present application, according to one aspect of the present application, a cable construction method based on BIM is provided, comprising: S1: obtaining BIM model data containing cable design path, design bending radius and design stress; S2: during the process of cable laying, fixing an optical fiber sensor tightly on the surface of the outer sheath of the cable to be laid, the optical fiber sensor being continuously distributed along the length direction of the cable; S3: real-time collecting the actual strain distribution data and the actual bending radius distribution data measured by the optical fiber sensor, and comparing them with the design stress data and the design bending radius data in the BIM model in terms of spatial position correlation; S4: if the actual strain value of any position point exceeds the strain threshold range corresponding to the design stress, or the actual bending radius value is smaller than the design bending radius value, it is determined that there is a construction deviation at the position point; S5: generating adjustment instructions for subsequent cable laying operations based on the position point information of the construction deviation.

[0006] Further, in S2, before or during the cable laying operation, the distributed optical fiber sensing element is arranged along the length direction of the cable to be laid; the arranged distributed optical fiber sensing element is attached and fixed to the outer sheath surface of the cable to be laid using at least one of a heat shrinkable sleeve, an adhesive, or a mechanical binding belt.

[0007] Further, positioning units are arranged at intervals on the cable laying path, and the positioning units include UWB positioning tags and time stamp generators synchronized with the optical fiber sensor; the real-time data stream collected by the optical fiber sensor is divided into a plurality of data segments, and each data segment is bound to a cable design path segment corresponding to a coordinate in the BIM model through a mapping relationship; temperature compensation is performed on the actual strain data of each data segment: the temperature distribution is calculated through the Brillouin scattered light power value of the distributed optical fiber sensor, and the calibrated strain value is calculated according to the formula ε cal =ε m -α·ΔT, where ε m is the measured strain value, ε cal is the calibrated strain value, α is the thermal expansion coefficient of the cable material, and ΔT is the difference between the current temperature and the reference temperature; the calibrated actual strain distribution data and the actual bending radius distribution data are compared in real time with the design stress data and the design bending radius data of the same spatial position in the BIM model.

[0008] Further, the real-time sensing data stream is processed by a pre-trained construction deviation judgment model, and the sensing data stream includes the calibrated strain value ε cal distribution, the actual bending radius distribution, and their space-time derivative quantities; the space-time derivative quantities include the strain change rate ▽ε and the bending radius deviation amount ΔR: the calculation formula of the strain change rate ▽ε is ▽ε(x i )=(ε cal (x i+1 )-ε cal (x i-1 )) / (x i+1 -x i-1 ), x i is the coordinate in the length direction of the cable; the calculation formula of the bending radius deviation amount ΔR is ΔR(x i )=|R design (x i )-R actual (x i )|, R design (x i ) and R actual (x i ) are the actual bending radius and the design bending radius, respectively; the construction deviation judgment model is constructed by: collecting sensing data stream samples of historical laying projects, and the samples include ε caldistribution, actual bending radius distribution, and calculated by the above formula, Δε, ΔR; mark the space-time coordinates of the true construction deviation events in the sample; use convolution long short-term memory network to extract the space-time joint features of the sensor data stream; the output layer generates a binary decision vector, the first component represents the strain overrun probability P ε , and the second component represents the bending radius deficiency probability P R ; when any position point satisfies P ε greater than or equal to the probability threshold θ ε or P R greater than or equal to the probability threshold θ R , it is determined that the position point has construction deviation.

[0009] Further, the input length of the real-time sensor data stream is dynamically controlled according to the current construction state parameters, including the cable laying speed, the real-time bending radius distribution data of the laying path, and the minimum design bending radius; when the cable laying speed increases, the effective time length of the sensor data stream is shortened; when there is a value less than 1.5 times the minimum design bending radius in the real-time bending radius distribution data, the spatial length covered by the sensor data stream is extended; the minimum spatial length of the sensor data stream needs to cover the path interval corresponding to the minimum design bending radius, and the minimum time length is not less than the interval of three adjacent effective data collection of the distributed optical fiber sensing element.

[0010] Further, the construction deviation judgment model includes a basic sub-model, a first sub-model and a second sub-model arranged in parallel, each sub-model is trained and generated for different lengths of data stream; when the cable laying speed exceeds a preset threshold, the first sub-model with shortened time length after input length adaptation is enabled, and the convolution kernel step of the first sub-model is greater than that of the basic sub-model; when it is detected that there is a section in the laying path with an actual bending radius value less than 1.5 times the minimum design bending radius, the second sub-model with extended spatial length after input length adaptation is enabled, and the bending radius risk weighting module is added to the basic sub-model structure; the bending radius risk weighting module extracts the bending radius deviation ΔR(x i ) in the real-time sensor data stream, divides the risk level according to the ratio of ΔR(x i ) to the minimum design bending radius, and gives higher weight to the position points with high risk level; the weighted bending radius deviation is input into the feature fusion layer of the convolution long short-term memory network; when the above conditions are not met, the basic sub-model is enabled.

[0011] Further, all the position points with construction deviation in the BIM model are marked, the corresponding construction risk grade is calculated based on the actual bending radius deviation amount AR and the actual strain deviation amount DE of each position point, the influence domain range is dynamically determined according to the minimum design bending radius and the upper limit of the strain threshold value with each deviation position point as the center, the influence domain covers the cable path not laid around the point; the risk propagation analysis is carried out on the not laid path segment in the influence domain; the risk contribution value of the deviation position point to the not laid path point is calculated based on the spatial distance attenuation principle; the risk contribution values of all the deviation position points are superimposed to generate the comprehensive risk weight of the not laid path point.

[0012] Further, according to the comprehensive risk weight, three-level adjustment instructions are divided: for the section with the comprehensive risk weight greater than or equal to the high risk threshold value, a first-level adjustment instruction is generated: the subsequent cable bending radius is increased to a compensation value greater than the design bending radius; for the section with the comprehensive risk weight less than the high risk threshold value but greater than or equal to the medium risk threshold value, a second-level adjustment instruction is generated: the cable laying speed is reduced, and the data acquisition frequency of the optical fiber sensor is increased; for the section with the comprehensive risk weight less than the medium risk threshold value, a third-level adjustment instruction is generated: the current cable laying parameters are maintained, but the data acquisition frequency of the optical fiber sensor is increased.

[0013] The present application at least includes the following beneficial effects:

[0014] The present application realizes real-time continuous monitoring of the cable construction state by combining the BIM model with the optical fiber sensing technology, and overcomes the defects of strong subjectivity of traditional manual inspection and incomplete coverage of discrete point measurement. By accurately associating and comparing the actual strain and bending radius data with the design parameters of the corresponding spatial position in the BIM model, construction deviation can be identified in time, and the misjudgment problem caused by the disconnection between design and construction data is solved. The rapid determination of construction deviation can avoid deviation accumulation, reduce the need for subsequent rework adjustment, reduce construction cost and construction period pressure, and effectively improve the cable laying accuracy and construction safety.

[0015] Other advantages, objects, and features of the present application will be apparent to those skilled in the art from the following description, and will be understood by those skilled in the art upon reading and understanding the specification. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 Flowchart of an embodiment of the present application. DETAILED DESCRIPTION

[0017] The present application will be further described in detail below, so that those skilled in the art can implement it according to the specification.

[0018] It should be understood that the terms such as "have", "contain", and "include" used in the embodiments of the present application are not exclusive to the presence or addition of one or more other elements or combinations thereof. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative position relationship, movement condition, etc. between components in a certain posture, and if the certain posture changes, the directional indications also change accordingly. When an element is referred to as "fixed to" or "disposed on" another element, it can be directly on another element or can have a middle element. When an element is referred to as "connected to" another element, it can be directly connected to another element or indirectly connected to another element through a middle element. The descriptions of "first", "second", etc. in the embodiments of the present application are only for descriptive purposes, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features.

[0019] It should be noted that the technical solutions among the various embodiments of the present application can be combined with each other, but it must be based on the implementation of a person of ordinary skill in the art, and when the combination of technical solutions is contradictory or cannot be implemented, it should be considered that the combination of technical solutions does not exist, and is not within the protection scope required by the present application.

[0020] As shown in Figure 1 The embodiments of the present application provide a cable construction method based on BIM, which comprises: S1: obtaining BIM model data containing cable design path, design bending radius and design stress; S2: in the process of cable laying, a fiber optic sensor is fixed tightly on the surface of the outer sheath of the cable to be laid, and the fiber optic sensor is continuously distributed along the length direction of the cable; S3: real-time collection of actual strain distribution data and actual bending radius distribution data measured by the fiber optic sensor, and comparison with the design stress data and the design bending radius data in the BIM model; S4: if the actual strain value of any position point exceeds the strain threshold range corresponding to the design stress, or the actual bending radius value is less than the design bending radius value, it is determined that there is a construction deviation at the position point; S5: based on the position point information of the existing construction deviation, an adjustment instruction for subsequent cable laying operation is generated.

[0021] Exemplarily, in S1, the BIM model data refers to a digital model data containing key parameters of cable construction, which is constructed by building information modeling technology. The cable design path is a pre-planned cable laying space route, the design bending radius can be 1.5 meters or 2 meters (determined according to the cable model), and the design stress can be 50 MPa or 80 MPa (determined by the mechanical properties of the cable material). The data can be obtained by exporting Autodesk Revit or other BIM modeling software. In S2, the cable laying process refers to the construction process of pulling the cable from the storage location to the installation path. The optical fiber sensor can be a distributed optical fiber sensor, which can continuously sense the change of physical quantity along the length direction. The outer sheath surface is the outer surface of the outermost protective structure of the cable. The optical fiber sensor is tightly attached to the outer sheath surface by using epoxy resin adhesive or stainless steel mechanical binding belt, and one sensing unit is arranged every meter along the length direction of the cable to realize continuous distribution. In S3, the real-time acquisition is completed by the optical fiber sensing demodulator. The actual strain distribution data is the data reflecting the deformation degree of the cable at different positions under stress, and the actual bending radius distribution data is the data representing the bending degree of each section of the cable. The spatial position correlation comparison refers to one-to-one correspondence of the cable position corresponding to the measured data and the coordinates of the design path in the BIM model through the GPS positioning module, and then point-by-point comparison of the difference between the actual data and the design data. In S4, the strain threshold range corresponding to the design stress can be calculated according to Hooke's law, i.e. ε = σ / E, where σ is the design stress and E is the elastic modulus of the cable material. If the cable is made of cross-linked polyethylene material, E can be selected as 1.2 GPa or 1.5 GPa. When the design stress is 50 MPa, the strain threshold range is 50 MPa / 1.2 GPa≈0.00042 to 50 MPa / 1.5 GPa≈0.00033; when the design stress is 80 MPa, the corresponding range is about 0.00053 to 0.00067. If the actual strain value exceeds the range, it is considered abnormal. If the design bending radius is 1.5 meters and the actual measured value is 1.2 meters, it is less than the design value, and it is determined that there is a construction deviation at this position. In S5, the position point information includes the BIM coordinates of the deviation point and the deviation degree. The generated adjustment instruction can be to slow down the laying speed to 0.5 meters per second or to increase the radius of the subsequent bending section to 1.8 meters, etc.

[0022] In the prior art, the local bending radius is measured by manual tape measurement, and the strain of a small number of discrete points is measured by strain gauge. Then, whether it meets the design requirements is determined by manual comparison with paper drawings. In the embodiment, the BIM model is combined with the optical fiber sensing to realize full-length continuous monitoring and automatic correlation comparison, which can timely find the deviations easily missed by manual inspection, reduce the rework caused by late discovery of problems, and improve the construction accuracy.

[0023] In another embodiment, in S2, the distributed optical fiber sensing element is arranged along the length direction of the cable to be laid before or during the cable laying operation; the arranged distributed optical fiber sensing element is attached and fixed to the outer sheath surface of the cable to be laid using at least one of the fixing methods of heat shrink sleeve, adhesive or mechanical binding belt.

[0024] Exemplarily, in S2, before the cable laying operation starts refers to the preparation stage of the cable before being installed by traction, and during the operation refers to the construction stage of the guide wire being moved by traction. The distributed optical fiber sensing element is an element capable of sensing full-length physical quantity changes through optical fiber transmission, and its length can match the length of the cable, such as 100 meters or 200 meters. Arranging along the length direction of the cable to be laid means that the optical fiber element is unfolded parallel to the cable axis to ensure that the sensing range covers the full length of the cable. The heat shrink sleeve can be a polyethylene adhesive heat shrink tube with a diameter of 8 mm or 10 mm; the adhesive can be an acrylic ester structural adhesive with temperature resistance; and the mechanical binding belt can be a nylon binding belt with a width of 3 mm or 5 mm. When attaching and fixing, first clean the dust and oil stains on the outer sheath surface of the cable, if using a heat shrink sleeve, wrap the sleeve around the optical fiber and the cable at the contact position, and heat it to 120-150°C with a hot air gun to shrink and fix it; if using an adhesive, evenly apply a 0.5-1 mm thick adhesive layer at the contact position of the optical fiber and the cable, and let it stand for 24 hours to cure; if using a binding belt, bind it every 30-50 centimeters, and the tightness should be appropriate so as not to damage the optical fiber and not to be loose.

[0025] In the prior art, optical fiber sensors are often fixed by simple adhesive tape winding, which is easy to fall off due to laying friction or environmental temperature changes, resulting in data collection interruption. The present embodiment uses a combination of multiple fixing methods to adapt to different construction environments, ensure that the sensing element is closely attached to the cable, reduce monitoring errors caused by fixing failure, and improve the stability of construction monitoring.

[0026] In another embodiment, positioning units are arranged at intervals on the cable laying path, the positioning units include UWB positioning tags and a time tag generator synchronized with the optical fiber sensor; the real-time data stream collected by the optical fiber sensor is divided into several data segments, each data segment is bound to the cable design path segment corresponding to the coordinate in the BIM model through a mapping relationship; the actual strain data of each data segment is temperature compensated: the temperature distribution is calculated through the Brillouin scattering light power value of the distributed optical fiber sensor, and the calibrated strain value is calculated according to the formula ε cal =ε m -α·ΔT, where ε m is the measured strain value, and ε calFor the calibrated strain value, a is the thermal expansion coefficient of the cable material, and ΔT is the difference between the current temperature and the reference temperature; the calibrated actual strain distribution data and the actual bending radius distribution data are compared in real time with the design stress data and the design bending radius data at the same spatial position in the BIM model.

[0027] Exemplarily, positioning units are arranged at intervals on the cable laying path, and the interval distance can be 10 meters or 15 meters. The positioning units can be installed on supports or walls along the path. The UWB positioning tag is a positioning device based on ultra-wideband technology, and the positioning accuracy can reach 0.3 meters or 0.5 meters. The time tag generator can generate a synchronous time signal to ensure that the optical fiber data and the positioning information are time consistent. Before the mapping relationship is established, the three-dimensional coordinate interval of the cable design path segment in the BIM model needs to be preset, and each segment includes a starting coordinate and an ending coordinate. After the real-time data stream is segmented, the actual on-site position coordinates corresponding to the data segment are obtained through the UWB positioning tag, and the on-site coordinates are mapped to the BIM model coordinate system through a coordinate conversion algorithm (such as a seven-parameter conversion), so that each data segment is accurately matched to the design path segment in the same spatial range in the BIM. The real-time data stream is segmented into 10-meter or 15-meter data segments according to the interval distance of the positioning units, and the mapping relationship is established through the preset path coordinate table, so that the data segment corresponds to the design path segment of the same length in the BIM model. When temperature compensation is performed, the Brillouin scattering optical power value is calculated by the optical fiber demodulator, the reference temperature can be set to 25°C, the current temperature is measured synchronously by the sensing element, and ΔT is the difference between the measured temperature and 25°C; a is the thermal expansion coefficient of the cable material (such as cross-linked polyethylene), and the value can be selected as 20×10 -6 / ℃ or 25×10 -6 / ℃, and ε cal is obtained by substituting the formula to eliminate the temperature effect. Real-time comparison is realized through a data processing terminal, and the calibrated measured data is compared with the design value (such as a design stress of 50 MPa corresponding to a strain of 0.0002) at the corresponding coordinate point in the BIM model point by point.

[0028] In the prior art, strain measurement often ignores the influence of temperature, and the measured data and the design path lack accurate coordinate correspondence, resulting in large deviation of the comparison result. In this embodiment, the measured data and the design data are accurately matched in the time and space dimensions through positioning synchronization and temperature compensation, the accuracy of the deviation determination is improved, and the environmental factor interference is reduced.

[0029] In another embodiment, the real-time sensing data stream is processed by a pre-trained construction deviation determination model, and the sensing data stream includes the calibrated strain value ε cal distribution, the actual bending radius distribution, and their time and space derivatives; the time and space derivatives include the strain change rate ▽ε and the bending radius deviation ΔR: the calculation formula of the strain change rate ▽ε is ▽ε(x i )=((εcal (x i+1 )-ε cal (x i-1 )) / (x i+1 -x i-1 ), x i The coordinates are along the length of the cable; the formula for calculating the bending radius deviation ΔR is ΔR(x...). i )=|R design (x i )-R actual (x i )|,R design (x i ) and R actual (x i The actual bending radius and the designed bending radius are respectively; the construction method of the construction deviation judgment model includes: collecting sensor data stream samples of historical laying projects, the samples containing ε cal Distribution, actual bending radius distribution, and ▽ε and ΔR calculated according to the above formulas; spatiotemporal coordinates of actual construction deviation events in the labeled samples; spatial-temporal joint features of the sensor data stream are extracted using a convolutional long short-term memory network; the output layer generates a binary decision vector, where the first component represents the strain exceedance probability P. ε The second component represents the probability P of insufficient bending radius. R When any point satisfies P ε Greater than or equal to the probability threshold θ ε or P R Greater than or equal to the probability threshold θ R At that time, it was determined that there was a construction deviation at that location.

[0030] For example, the pre-trained construction deviation judgment model is deployed in an edge computing server, and the real-time sensor data stream is transmitted to the server via a fiber optic demodulator. cal The distribution consists of calibrated strain data at various points along the entire cable length; the actual bending radius distribution is calculated from fiber optic sensing data. In the calculation of spatiotemporal derivatives, x... i x represents the coordinates (in meters) along the length of the cable. i+1 With x i-1 For x iThe coordinates of the front and rear adjacent points with a distance of 0.5 meters or 1 meter are substituted into the formula to obtain △ε, which reflects the rate of change of strain with length; ΔR is calculated by the absolute value difference between the designed bending radius (such as 1.5 meters) and the actual measured value (such as 1.2 meters). When the model is constructed, the historical samples are collected from 10 or 15 cable laying projects, and the deviation event coordinates are labeled in combination with the construction records and detection reports. When the model is constructed, in order to improve the generalization ability, the historical samples need to cover different cable models (such as 10kV cross-linked polyethylene cables, low-voltage plastic force cables), laying environments (indoor bridge, underground direct burial, pipe laying) and climate conditions (temperature range of-10℃ to 35℃), and the sample size is not less than 500 groups. Through data enhancement technology, ±5% random noise and ±10% scale stretching are added to the samples to simulate different construction disturbances, and the deviation event coordinates are labeled in combination with the construction records and detection reports. The convolutional long short-term memory network includes 2 convolutional layers and 1 LSTM layer, which is trained through the TensorFlow framework; the probability threshold θ ε may be 0.8, and θ R may be 0.7. When a point P ε =0.85 or P R =0.72, it is determined that there is a construction deviation.

[0031] In the prior art, deviation determination relies on manual setting of fixed thresholds, which is difficult to adapt to dynamic changes in complex construction environments. The present embodiment learns the characteristics of historical data through the model, and realizes probabilistic determination combined with the time and space derived quantities, thereby improving the recognition ability of subtle deviations and complex scenes and reducing the misjudgment rate.

[0032] In another embodiment, the input length of the real-time sensing data stream is dynamically controlled according to the current construction state parameters, including the cable laying speed, the real-time bending radius distribution data of the laying path, and the minimum designed bending radius; when the cable laying speed increases, the effective time length of the sensing data stream is shortened; when there is a value less than 1.5 times the minimum designed bending radius in the real-time bending radius distribution data, the spatial length covered by the sensing data stream is lengthened; the minimum spatial length of the sensing data stream needs to cover the path interval corresponding to the minimum designed bending radius, and the minimum time length is not less than the interval of three times effective data collection of the distributed optical fiber sensing element.

[0033] Exemplarily, the current construction state parameters are collected in real time by sensors and construction equipment, the cable laying speed is measured by a speed sensor of the traction machine, in units of meters / minute; the real-time bending radius distribution data are generated by the optical fiber sensing system; the minimum design bending radius is determined according to the cable model, such as 1 meter or 1.2 meters. When the laying speed increases from 3 meters / minute to 6 meters / minute, the effective time length can be shortened from 10 seconds to 5 seconds, reducing data delay. 1.5 times the minimum design bending radius, i.e. 1.5 x 1 meter = 1.5 meters, if a bending radius of 1.3 meters appears in the real-time data, the space length is extended from 2 meters to 3 meters. The minimum space length needs to cover the path arc length corresponding to the minimum design bending radius, such as 2 meters of arc length corresponding to 1 meter radius; if the interval of adjacent three data collection is 1 second, the minimum time length is not less than 1 second, ensuring data continuity.

[0034] In the prior art, the data stream length is fixed, and when the construction speed is high, the data lag is prone to miss detection deviation, and the complex bending section is prone to inaccurate judgment due to insufficient data. The embodiment dynamically adjusts the data stream length, while ensuring real-time, meeting the data needs of key sections, and improving the monitoring adaptability under different construction states.

[0035] In another embodiment, the construction deviation judgment model includes a basic sub-model, a first sub-model and a second sub-model arranged in parallel, each sub-model being trained and generated for different scales of data stream length; when the cable laying speed exceeds a preset threshold, the first sub-model with an input length adapted to shorten the time length is enabled, and the convolution kernel step of the first sub-model is greater than that of the basic sub-model; when it is detected that there is a section in the laying path with an actual bending radius value less than 1.5 times the minimum design bending radius, the second sub-model with an input length adapted to extend the space length is enabled, and the second sub-model adds a bending radius risk weighting module on the structure of the basic sub-model; the bending radius risk weighting module extracts the bending radius deviation ΔR(x i ) in the real-time sensing data stream, divides the risk level according to the ratio of ΔR(x i ) to the minimum design bending radius, and gives higher weight to the high-risk level position; the weighted bending radius deviation is input into the feature fusion layer of the convolutional long short-term memory network; when the above conditions are not met, the basic sub-model is enabled.

[0036] Exemplarily, the three sub-models are all deployed on the same AI inference platform, and the basic sub-model is adapted to a regular data flow length (e.g., a time length of 5 seconds or a space length of 2 meters). A preset threshold can be set to 5 meters / minute, when the laying speed reaches 6 meters / minute, the first sub-model is enabled, and the convolution kernel step length is 2 (the basic sub-model step length is 1), which accelerates the data processing speed. The minimum design bending radius is 1.5 times, such as 1.5 meters, when a section of 1.2 meters is detected, the second sub-model is enabled, and the space length is adapted to 3 meters. For example, the basic sub-model contains 2 convolution layers (convolution kernel size 3x3, step length 1) and 1 LSTM layer (hidden unit number 64), which is adapted to regular data flow. The first sub-model has the same layer structure as the basic sub-model, and the convolution kernel step length is increased to 2 to speed up processing. The second sub-model is provided with 3 convolution layers (3x3 for the first two layers and 5x5 for the third layer, step length 1) and 1 LSTM layer, and a bending radius risk weighting module is added, which is adapted to long space data flow. In the risk level division, ΔR(x i ) and the ratio of the minimum design radius >0.3 is high risk (e.g., ΔR=0.4 meters, ratio 0.4 / 1=0.4), and the weight is 1.5 (regular weight 1.0), and the weighted data is fused with the strain data through the feature fusion layer.

[0037] A single model is difficult to adapt to different construction scenes, resulting in a decline in determination performance under complex conditions. The embodiment dynamically switches multiple sub-models, optimizes the model structure for speed and bending risk, and improves the real-time performance and accuracy of deviation determination.

[0038] In another embodiment, all position points with construction deviations are marked in the BIM model, and based on the actual bending radius deviation ΔR and the actual strain deviation Δε of each position point, the corresponding construction risk level is calculated; based on the minimum design bending radius and the upper limit of the strain threshold, the influence domain range is dynamically determined with each deviation position point as the center, and the influence domain covers the un-laid cable path around the point; the risk propagation analysis is performed on the un-laid path segment in the influence domain: based on the spatial distance attenuation principle, the risk contribution value of the deviation position point to the un-laid path point is calculated; the risk contribution values of all deviation position points are superimposed to generate the comprehensive risk weight of the un-laid path point.

[0039] Exemplarily, the deviation point in the BIM model is marked by coordinates, ΔR is the difference between the actual and designed bending radius (e.g. 0.3 meters), Δε is the difference between the actual strain and the threshold value (e.g. 0.0001), and the risk level is divided into low, medium and high according to the product of the two (e.g. the product < 0.00003 is low risk). The influence domain range is set as an area with a radius of 2 meters based on the minimum designed bending radius of 1 meter, or an area of 3 meters based on the upper limit of the strain threshold of 0.0005, covering the surrounding un-laid path. In the risk propagation analysis, the un-laid point 1 meter away from the deviation point contributes 0.8 of the initial risk, and 2 meters away contributes 0.5 (distance attenuation coefficient 0.3), and the comprehensive risk weight is obtained after superimposing the contribution values of multiple deviation points (e.g. 0.6 or 0.8).

[0040] In the prior art, only the existing deviation points are concerned, and the potential impact on subsequent construction is ignored, which is easy to lead to a chain of deviations. The present embodiment predicts the risk of the un-laid section through risk propagation analysis, provides a basis for subsequent adjustment, and reduces the cumulative effect of deviation.

[0041] In another embodiment, three-level adjustment instructions are divided according to the comprehensive risk weight: for the section with a comprehensive risk weight greater than or equal to the high risk threshold, a first-level adjustment instruction is generated: increase the subsequent cable bending radius to a compensation value that exceeds the designed bending radius; for the section with a comprehensive risk weight less than the high risk threshold but greater than or equal to the medium risk threshold, a second-level adjustment instruction is generated: reduce the cable laying speed and increase the data acquisition frequency of the optical fiber sensor; for the section with a comprehensive risk weight less than the medium risk threshold, a third-level adjustment instruction is generated: maintain the current cable laying parameters but increase the data acquisition frequency of the optical fiber sensor.

[0042] Exemplarily, the comprehensive risk weight ranges from 0 to 1, the high risk threshold is set to 0.7, and the medium risk threshold is set to 0.4. In the first-level adjustment instruction, when the designed bending radius is 1.5 meters, the compensation value is set to 0.3 meters, and the subsequent bending radius is increased to 1.8 meters. In the second-level adjustment instruction, the laying speed is reduced from 4 meters / minute to 2 meters / minute, and the data acquisition frequency is increased from 1 Hz to 5 Hz. The third-level adjustment instruction maintains the current laying speed of 3 meters / minute and the acquisition frequency of 2 Hz, and only increases the frequency to 3 Hz. The instructions are sent to the traction machine and the sensing equipment through the construction control system for real-time adjustment of the construction parameters.

[0043] In the prior art, construction adjustment lacks grading basis, often over-adjusting affects efficiency or under-adjusting leads to risk. The present embodiment classifies according to the risk weight, controls the risk while considering the construction efficiency, and realizes precise construction control.

[0044] While embodiments of the application have been disclosed in connection with the above specification, it will be apparent to those skilled in the art that numerous modifications can be made thereto without departing from the overall concept of the application. Accordingly, it is intended that all such modifications be included within the scope of the claims and their equivalents.

Claims

1. A BIM-based cable construction method, characterized by, The method comprises the following steps: S1: obtaining BIM model data containing cable design path, design bending radius and design stress; S2: fixing the optical fiber sensor tightly on the surface of the outer sheath of the cable to be laid during the cable laying process, and the optical fiber sensor is continuously distributed along the length direction of the cable; S3: real-time acquisition of the actual strain distribution data and the actual bending radius distribution data measured by the optical fiber sensor, and comparison with the design stress data and the design bending radius data in the BIM model; S4: if the actual strain value of any position point exceeds the strain threshold value corresponding to the design stress, or the actual bending radius value is less than the design bending radius value, it is determined that there is a construction deviation at the position point; S5: generating adjustment instructions for subsequent cable laying operation based on the position point information of the construction deviation; The real-time sensing data stream containing calibrated strain values ε is processed by the pre-trained construction deviation determination model cal Distribution, actual bending radius distribution and its spatiotemporal derivative; the spatiotemporal derivative includes strain change rate ∂ε and bending radius deviation ΔR: the calculation formula of strain change rate ∂ε is ∂ε(x i )=(ε cal (x i+1 )-ε cal (x i-1 )) / (x i+1 -x i-1 ), x i is the cable length direction coordinate; the calculation formula of bending radius deviation ΔR is ΔR(x i )=|R design (x i )-R actual (x i )|, R design (x i ) and R actual (x i ) are actual bending radius and design bending radius respectively; The construction deviation determination model construction method comprises the following steps: collecting historical laying project sensing data stream samples, the samples containing ε cal distribution, actual bending radius distribution, and calculated by the above formula, ΔR; labeling the space-time coordinates of the true construction deviation event in the sample; using a convolutional long short-term memory network to extract the space-time joint features of the sensing data stream; The output layer generates a binary decision vector, the first component of which represents the strain overrun probability P ε , and the second component represents the insufficient bending radius probability P R ; when any position point satisfies P ε greater than or equal to the probability threshold θ ε or P R greater than or equal to the probability threshold θ R , it is determined that the position point has construction deviation. According to the current construction state parameters, the input length of the real-time sensing data stream is dynamically controlled, and the construction state parameters include the cable laying speed, the real-time bending radius distribution data of the laying path and the minimum design bending radius; when the cable laying speed increases, the effective time length of the sensing data stream is shortened; when there is a value less than 1.5 times the minimum design bending radius in the real-time bending radius distribution data, the spatial length covered by the sensing data stream is lengthened; the minimum spatial length of the sensing data stream needs to cover the path interval corresponding to the minimum design bending radius, and the minimum time length is not less than the interval of three times effective data acquisition of the distributed optical fiber sensing element.

2. A BIM-based cabling method according to claim 1, characterized in that, In S2, the distributed optical fiber sensing element is arranged along the length direction of the cable to be laid before or during the cable laying operation; The arranged distributed optical fiber sensing element is attached and fixed to the outer sheath surface of the cable to be laid by using at least one fixing method of heat shrink sleeve, adhesive or mechanical binding belt.

3. The BIM-based cabling method of claim 1, wherein, Positioning units are arranged at intervals on the cable laying path, and the positioning units include UWB positioning tags and time tag generators synchronized with the optical fiber sensor; The real-time data stream collected by the optical fiber sensor is divided into several data segments, and each data segment is bound to the cable design path segment of the corresponding coordinate in the BIM model through a mapping relationship; Temperature compensation is performed on the actual strain data of each data segment: the temperature distribution is calculated by the Brillouin scattered light power value of the distributed optical fiber sensor, and the calibrated strain value is calculated according to the formula ε cal =ε m -α·ΔT, wherein ε m is the measured strain value, ε cal is the calibrated strain value, α is the thermal expansion coefficient of the cable material, and ΔT is the difference between the current temperature and the reference temperature; The calibrated actual strain distribution data and actual bending radius distribution data are compared with the design stress data and design bending radius data of the same spatial position in the BIM model in real time.

4. The BIM-based cable construction method of claim 1, wherein, The construction deviation judgment model includes a basic sub-model, a first sub-model and a second sub-model arranged in parallel, and each sub-model is trained and generated for different lengths of data stream; When the cable laying speed exceeds the preset threshold value, the first sub-model with shortened time length is enabled for input length adaptation, and the convolution kernel step of the first sub-model is greater than that of the basic sub-model; When it is detected that there is a section in the laying path with an actual bending radius value less than 1.5 times the minimum design bending radius, a second sub-model of input length adaptation to elongated space length is enabled, and the second sub-model adds a bending radius risk weighting module on the basis of the sub-model structure; the bending radius risk weighting module extracts the bending radius deviation ΔR(x i ) in the real-time sensing data stream, divides the risk level according to the ratio of ΔR(x i ) to the minimum design bending radius, and gives higher weights to the high-risk level position points. The weighted bending radius deviation amount is input into the feature fusion layer of the convolution long short-term memory network; When the above conditions are not met, the basic sub-model is enabled.

5. The BIM-based cable construction method of claim 1, wherein, All position points with construction deviation are marked in the BIM model, and the construction risk level corresponding to each position point is calculated based on the actual bending radius deviation amount ΔR and the actual strain deviation amount Δε of the position point. The influence domain range is dynamically determined according to the minimum design bending radius and the upper limit of the strain threshold, and the influence domain covers the cable path that is not laid around the deviation position point; Risk propagation analysis is performed on the un-laid path segment in the influence domain: based on the spatial distance attenuation principle, the risk contribution value of the deviation position point to the un-laid path point is calculated; The risk contribution values of all deviation position points are superimposed to generate the comprehensive risk weight of the un-laid path point.

6. The BIM-based cable construction method of claim 5, wherein, According to the comprehensive risk weight, three-level adjustment instructions are divided: For the section with a comprehensive risk weight greater than or equal to the high-risk threshold, a first-level adjustment instruction is generated: increase the subsequent cable bending radius; For the section with a comprehensive risk weight less than the high-risk threshold but greater than or equal to the medium-risk threshold, a second-level adjustment instruction is generated: reduce the cable laying speed and increase the data acquisition frequency of the optical fiber sensor; For the section with a comprehensive risk weight less than the medium-risk threshold, a third-level adjustment instruction is generated: maintain the current cable laying parameters and increase the data acquisition frequency of the optical fiber sensor.

Images