Multi-frequency data fusion intelligent construction site personnel high-precision real-time positioning method based on construction progress dynamic adaptation

Abstract

The invention discloses a multi-frequency data fusion intelligent construction site personnel high-precision real-time positioning method based on construction progress dynamic adaptation. The method comprises the steps that construction project information is collected, and construction stages and regions are divided; setting a positioning precision grade in combination with a regional operation type and a space parameter, and matching a safety helmet signal adaptive frequency range; defining a frequency-region matching degree and constructing an evaluation table, and setting a signal frequency and calculating a positioning weight according to the evaluation table; sensing node signals are fused and are subjected to time domain segmentation and DFT processing to obtain same-frequency narrow-band signals; signal and noise subspaces are separated through covariance matrix decomposition, a cost function is constructed in combination with positioning weights, and personnel positions are positioned after areas are gridded. According to the method, the construction progress is associated with the regional characteristic dynamic adaptive parameters, the problems of poor adaptability and information loss of a traditional positioning method are solved, the positioning precision and stability in a construction site time-varying scene are improved, and the practicability is high.

Description

Technical Field

[0001] This invention relates to the field of wireless positioning technology, specifically to a method for high-precision real-time positioning of personnel in smart construction sites based on multi-frequency data fusion dynamically adapted to construction progress. Background Technology

[0002] With the deepening of smart city construction, safety management and efficiency improvement in the construction industry increasingly rely on information and intelligent technologies. Among them, real-time, high-precision location sensing of personnel on construction sites is a key technology for achieving safety early warning (such as entering dangerous areas), work scheduling, emergency rescue, and digital management.

[0003] Currently, the main solutions in this field are as follows: Global Positioning Systems (GPS, BeiDou) perform well in open outdoor areas, but their signals attenuate severely indoors and in areas with dense steel structures, failing to meet the full coverage requirements of complex construction site environments. Furthermore, their conventional civilian accuracy is only at the meter level, making it difficult to support sophisticated personnel management. While Wi-Fi or Bluetooth-based wireless positioning technologies are convenient to deploy, they are susceptible to environmental interference, with positioning accuracy typically limited to five to ten meters. They are mostly used for area sensing and cannot achieve precise trajectory tracking. Ultra-wideband (UWB) technology can theoretically achieve decimeter-level accuracy and has become the mainstream choice for high-precision positioning. However, in actual construction site scenarios, its signal is easily blocked by walls, equipment, and scaffolding, leading to positioning drift or even interruption under non-line-of-sight conditions. Relying solely on UWB makes it difficult to guarantee the continuous robustness of the system. Inertial Navigation Systems (IMUs) do not rely on external signals and can temporarily compensate for wireless positioning blind spots, but their errors accumulate over time, lacking independent long-term positioning capabilities.

[0004] The invention disclosed in CN115407266A presents a direct positioning method based on the orthogonality of cross-spectral subspaces. Compared with other direct positioning methods, it has higher accuracy and lower complexity, and maintains high estimation performance even in multipath environments. However, this type of method is still not well applied to actual construction sites. This is because the construction process on actual construction sites is characterized by dynamic iteration. From the foundation pit operation in the foundation stage, to the floor construction in the main structure stage, to the interior finishing work in the decoration stage, the layout of key operation areas, high-risk areas, and large equipment changes significantly as the process progresses. This dynamism requires that the positioning system cannot adopt a one-size-fits-all fixed configuration. For example, in the high-altitude operation sub-stage of the main structure construction, the accuracy requirement for personnel positioning in dangerous areas such as floor edges and adjacent openings reaches the centimeter level, requiring a switch to high-precision tracking mode; while in the foundation pit excavation area of ​​the foundation stage, only area-level positioning is needed to meet the collision avoidance requirements of machinery and personnel.

[0005] Meanwhile, interference sources on construction sites exhibit strong scene-related characteristics. For example, densely reinforced areas during the main construction phase significantly attenuate UWB signals, while metal ceilings during the decoration phase are prone to signal multipath effects. These interferences directly impact the reception quality of safety helmet signals. Furthermore, the positioning requirements differ significantly between different trades and even within the same trade at different construction stages. For instance, electricians need sub-meter level static positioning to avoid electric shock risks when working on embedded pipelines, while tower crane operators require continuous dynamic trajectory tracking to coordinate with hoisting operations. However, existing positioning technologies generally suffer from limitations due to their reliance on single data sources (e.g., using only UWB or Bluetooth). They cannot adapt to the time-varying interference environment of construction sites, nor can they simultaneously meet the accuracy requirements of different scenarios, the continuity of positioning during personnel movement, and the system reliability under complex working conditions. Therefore, a new high-precision positioning method that can integrate multi-source information and adapt to dynamic scenarios is urgently needed. Summary of the Invention

[0006] The purpose of this invention is to provide a high-precision real-time positioning method for personnel in smart construction sites based on multi-frequency data fusion with dynamic adaptation to construction progress. It introduces a regional positioning strategy that is aware of the construction phase. By associating the process characteristics of the construction progress, the risk level of the area, and the interference type of the scene, it dynamically adjusts the positioning accuracy threshold of the corresponding area and the parameter configuration of the multi-frequency fusion algorithm to achieve adaptive matching for time-varying scenes on the construction site.

[0007] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0008] A method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion, the method comprising the following steps:

[0009] S1, collect relevant information about the construction project, including project type, total duration, work site drawings, list of large equipment, and list of construction personnel; based on this information, divide the construction schedule into... There are several construction phases, each corresponding to... One construction area;

[0010] S2, for each construction progress, select the construction area one by one, and set the initial positioning accuracy level according to the operation type and spatial parameters of the construction area. , Representation phase ,area The initial positioning accuracy level is determined by the value, with a higher value indicating a higher positioning accuracy requirement. This is then combined with the interference type and regional characteristics of the construction area to match the appropriate frequency range for the safety helmet signal.

[0011] S3, within the adapted frequency range, defines the frequency-area matching degree for each construction area at different construction stages based on the physical characteristics of the construction area. , Indicates the stage ,area Frequency of use Applicability degree; frequency-region matching degree based on definition Construct a frequency-region matching evaluation table;

[0012] S4. Define the current construction progress and construction area. Set the frequency of the radiation signal emitted by the safety helmets worn by construction workers according to the frequency-area matching degree evaluation table, and find the corresponding frequency-area matching degree. Combined with frequency-region matching degree The positioning weights used for multi-frequency data fusion are calculated based on the initial positioning accuracy level.

[0013] S5, at each sensing node, receives the radiation signal emitted by the safety helmets worn by construction workers and uploads it to the data processing center. The signals received by all sensing nodes are fused. The fused received signal is divided into J segments in the time domain, each segment containing N data points. DFT processing is performed on each segment. Data with the same frequency are taken from the DFT results of the J segments and combined into N narrowband received signals of the same frequency.

[0014] S6. Calculate the covariance matrix of N narrowband received signals at the same frequency and perform eigenvalue decomposition to separate the signal subspace and noise subspace. Construct a cost function using the orthogonality relationship of the noise subspace at each frequency point and the positioning weight. Then, grid the smart construction site area and calculate the function value of the cost function at each grid point. Find the largest number of peaks of the cost function and use the corresponding grid positions as the real-time positions of the corresponding personnel.

[0015] Step S1 further includes:

[0016] Collect relevant information about the construction project, including project type, total construction period, work area drawings, list of large equipment, and list of construction personnel. The work area drawings include the dimensions of the construction area, the distribution of obstacles, and the deployment locations of sensing nodes. The list of large equipment includes equipment models, operating ranges, and electromagnetic interference parameters. The list of construction personnel includes the number of personnel and their job types.

[0017] The construction schedule is broken down according to the total project duration. The construction process is divided into several phases, and within each phase, the site is further divided according to the work area drawings and the distribution of large equipment. Each construction area is clearly defined, specifying its operational boundaries and types.

[0018] Step S2 further includes:

[0019] For each construction progress, a construction area is selected one by one. Based on the spatial parameters of the construction area related to the openness of the space, the type of obstruction, and the electromagnetic environment, the spatial complexity of the construction area is analyzed. Then, the corresponding operation risk level is analyzed based on the operation type of the construction area.

[0020] The initial positioning accuracy level is set according to the operational risk level and spatial complexity of the construction area. ;

[0021] Based on the openness of the construction area, the type of obstructions, and the initial positioning accuracy level The usable frequency bands were initially screened out; then, based on the regional spatial parameters related to the electromagnetic environment, the frequency bands that overlapped with the interference frequency bands of the construction area were eliminated, and finally the suitable frequency range of the construction area was determined.

[0022] In step S4, the positioning weights used for multi-frequency data fusion are calculated using the following formula. :

[0023]

[0024] In the formula, N is the total number of frequency points in the signal frequency domain, where n is the nth frequency point. This represents the signal frequency at the nth frequency point.

[0025] Step S5 further includes:

[0026] L sensing nodes are placed at the edge of the construction area, with the following locations: Within the construction area There are [number] staff members, located at [location]. Assuming that each worker's smart safety helmet emits uncorrelated broadband stationary signals, the first... Each sensing node Received signal at any time Represented as:

[0027]

[0028] in The number of sampling points. Additive noise, The signal emitted by the smart safety helmet worn by the k-th worker is represented; the received signals from all sensing nodes are integrated to obtain the fused received signal. for:

[0029] ;

[0030] For the fused received signal Perform time-domain segmentation, dividing it into J segments. After segmentation, each segment has... The data in the column is expressed in the following specific formula:

[0031]

[0032] The expression for each segment after segmentation is as follows:

[0033]

[0034] in

[0035] ;

[0036] in This represents the j-th segment of the signal emitted by the smart safety helmet worn by the k-th worker. It is noise signal during transmission;

[0037] Perform DFT processing on the signal of each segment to obtain :

[0038]

[0039] in yes The frequency domain representation, It is the frequency domain representation of the noise signal. The expression is as follows

[0040]

[0041] in ,in , These are the x and y coordinates of the l-th sensing node in the coordinate system. and is the x and y coordinates of the kth worker, and c is the speed of electromagnetic signal propagation in space, i.e., the speed of light;

[0042] Extracting the same column of data from each segment to form a new signal data, resulting in N narrowband received signals at the same frequency:

[0043] .

[0044] Step S6 further includes:

[0045] Calculate the covariance matrix of N narrowband received signals at the same frequency and perform eigenvalue decomposition on it to obtain:

[0046]

[0047]

[0048] in, and It is a diagonal matrix composed of eigenvalues. and It is a matrix composed of eigenvectors, with the eigenvalues ​​sorted from largest to smallest. It is the signal subspace composed of the eigenvectors corresponding to the first K eigenvalues. It is a noise subspace composed of the eigenvectors corresponding to the remaining LK eigenvalues;

[0049] Based on the orthogonality between the noise subspace and the direction vector, we obtain

[0050]

[0051] in

[0052]

[0053] It is the expression for the guide vector, where It is the coordinate vector in the smart construction site area, in It is a location The propagation delay with the l-th sensing node is used to construct the cost function as follows:

[0054] ;

[0055] The smart construction site area is gridded, and the function value of the cost function is calculated at each grid point;

[0056] Find the largest peaks in the cost function and use their corresponding grid positions as the real-time positions of the corresponding personnel.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] First, the high-precision real-time positioning method for personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion of the present invention dynamically adjusts the positioning accuracy threshold and multi-frequency fusion parameters by associating the process characteristics of construction progress, regional risk level and scene interference type. It can adapt to time-varying scenarios such as open-air excavation, underground operation and fine decoration construction on construction sites, and has stronger scene adaptability, solving the problem that the parameters of traditional methods are fixed and cannot match complex working conditions.

[0059] Secondly, the high-precision real-time positioning method for personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion of the present invention, on the one hand, dynamically sets the accuracy threshold based on the risk level of the construction area, and automatically matches higher accuracy standards for high-risk areas (such as high-altitude operation areas); on the other hand, it integrates the information of all sensing nodes through multi-frequency data fusion, avoiding the data loss defects of traditional methods, and significantly reducing the positioning error in scenarios with occlusion and multipath interference, thus improving both positioning accuracy and stability.

[0060] Third, the high-precision real-time positioning method for personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion of the present invention matches the frequency range based on the characteristics of the construction area, and selects the optimal frequency band through frequency-area matching degree. Combined with multi-frequency fusion algorithm, it makes full use of the signal advantages of different frequency bands (such as the penetration of low frequency and the high precision of high frequency), resulting in higher signal utilization efficiency. Compared with the single-frequency positioning of the AOA method, the signal anti-interference capability is effectively improved.

[0061] Fourth, the high-precision real-time positioning method for personnel in smart construction sites based on multi-frequency data fusion dynamically adapted to construction progress of the present invention sets parameters (such as spatial complexity and operational risk level) based on the actual collectable information of the construction project, dynamically adjusts the logic to fit the on-site management habits of the construction site, does not require additional hardware costs, and is easier to implement and promote on construction sites than traditional methods, with better engineering practicality. Attached Figure Description

[0062] Figure 1 This is a flowchart of the method for high-precision real-time positioning of personnel in a smart construction site based on dynamic adaptation of construction progress and multi-frequency data fusion.

[0063] Figure 2 A scene diagram illustrating high-precision real-time positioning of personnel at a smart construction site;

[0064] Figure 3 This is a schematic diagram of the two-dimensional spatial spectrum of the present invention;

[0065] Figure 4 A scatter plot showing the results of locating multiple radiation sources using the method of this invention;

[0066] Figure 5 A schematic diagram illustrating the root mean square error performance of the positioning results of the present invention and the traditional positioning method under different signal-to-noise ratios;

[0067] Figure 6 This diagram illustrates the root mean square error performance of the positioning results of the present invention and the traditional positioning method under different snapshot numbers. Detailed Implementation

[0068] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0069] Symbol representation: In this invention, bold uppercase letters, bold lowercase letters, and italic letters, such as... , and , representing matrices, vectors, and scalars, respectively. , and These represent the transpose, conjugate transpose, and inverse operations of a matrix, respectively. , , , , These letters, each with a triangular superscript, represent the expectation, F-norm, trace, real part, and transformation of a vector into a diagonal matrix, respectively. express The estimated value.

[0070] This invention provides a method for high-precision real-time positioning of personnel in smart construction sites based on multi-frequency data fusion with dynamic adaptation to construction progress, such as... Figure 1 As shown, during the project implementation process, this invention dynamically allocates optimal positioning weights to the monitoring frequency points (or frequency bands) corresponding to each stage according to a preset construction schedule. The core of this process lies in mapping the project's time nodes to dynamic adjustments for the positioning accuracy requirements of different areas. For example, in the initial foundation construction stage, the positioning weights for secondary areas can be reduced, and computational resources can be concentrated on key structural points, using a high-resolution but computationally demanding algorithm configuration. As construction enters the later precision installation stage, the positioning weights for the core equipment area are dynamically increased, and a robust algorithm that balances real-time performance and accuracy is switched to. Through adaptive allocation of positioning weights, a precise match between computational resources, positioning accuracy, and actual project requirements is achieved, ensuring that the differentiated monitoring requirements of each stage are met while optimizing the overall algorithm's operating efficiency and energy consumption. Specifically, the following steps are included:

[0071] S1, collect relevant information about the construction project, including project type, total duration, work site drawings, list of large equipment, and list of construction personnel; based on this information, divide the construction schedule into... There are several construction phases, each corresponding to... One construction area.

[0072] Step S1 is used to complete the basic data collection and spatial-temporal dimension segmentation of the entire construction scenario, providing data support for subsequent positioning strategy adaptation. In practical applications, a dedicated information collection team can be formed to collect project type (building construction / municipal construction / bridge construction, etc.), overall construction period planning documents, construction site plan, and detailed drawings of each area, based on the construction specifications of the corresponding project, clarifying the dimensions of the work surface, the distribution of obstacles, and the preset locations of sensing nodes. Simultaneously, a list of large equipment (including tower cranes, welding machines, etc., models, working range, and electromagnetic interference frequency band parameters) and a list of construction personnel (including the number of personnel, job types, working hours, and unique safety helmet numbers) are collected.

[0073] Based on this, the construction schedule is broken down into W construction phases according to the total construction period. The division of construction phases must be consistent with key process nodes (such as foundation construction, main structure, and decoration). For each construction phase, combined with the work area drawings, equipment distribution, and process requirements, a two-dimensional division method based on physical boundaries and work functions is adopted to divide the construction site into R construction areas. The geographical boundaries, work content, and personnel flow range of each area are clearly defined, forming a construction phase-construction area division comparison table, marking the core work characteristics of each construction area.

[0074] S2, for each construction progress, select the construction area one by one, and set the initial positioning accuracy level according to the operation type and spatial parameters of the construction area. , Representation phase ,area The initial positioning accuracy level, for example The larger the value, the higher the positioning accuracy requirement. Then, the appropriate frequency range of the safety helmet signal is matched by the interference type and regional characteristics of the construction area.

[0075] The purpose of step S2 is to determine the positioning accuracy standard based on regional characteristics and to select suitable signal frequency bands to ensure the effectiveness of the positioning foundation.

[0076] Preferably, for each construction progress stage, a construction area is selected one by one. Based on the spatial parameters of the construction area related to its openness, type of obstructions, and electromagnetic environment, the spatial complexity of the construction area is analyzed. Then, the corresponding operational risk level is analyzed based on the operation type of the construction area. Next, the initial positioning accuracy level is set according to the operational risk level and spatial complexity of the construction area. Regarding the initial positioning accuracy level The setting method can employ methods such as the risk matrix method, combining the operational risk level and spatial complexity of the construction area r under the construction stage w. The operational risk level can be divided into four levels: major, significant, general, and low. Spatial complexity is divided into three levels: high, medium, and low based on the density of obstructions and the degree of spatial enclosure. Based on this, high-risk, high-complexity areas such as deep foundation pits and tall formwork operation areas are assigned the highest level; medium-risk, medium-complexity areas such as steel structure construction layers and open-air excavation areas are assigned the medium level; and low-risk, low-complexity areas such as material storage areas and office areas are assigned the lowest level, forming a rule table corresponding to the construction area and the initial positioning accuracy level.

[0077] Then, based on the openness of the construction area, the type of obstructions, and the initial positioning accuracy level... The process begins with initial screening to identify usable frequency bands. For example, in enclosed spaces such as underground parking garages and pipeline corridors, low-frequency bands with strong penetration are selected; in open areas requiring medium to high precision positioning, mid-frequency bands with stable transmission are chosen; and in areas with high precision requirements, high-frequency bands with high positioning accuracy are selected. Finally, based on the regional spatial parameters related to the electromagnetic environment, frequency bands overlapping with the interference frequency bands of the construction area are eliminated, ultimately determining the suitable frequency range for the construction area. For example, if there is electromagnetic interference from equipment such as tower cranes and welding machines in the construction area, the interfering frequency bands are avoided first.

[0078] S3, due to different frequencies Different regions have different applicability; for example, high-frequency signals work well in open areas (value close to 1), but are difficult to use behind concrete walls (value close to 0). Therefore, within the suitable frequency range, the frequency-region matching degree for each construction area at different construction stages is defined according to the physical characteristics of the construction area. , Indicates the stage ,area Frequency of use The applicability of the signal strength is determined by the measured signal reception power at the construction area, the anti-interference capability is assessed based on the bit error rate of the signal under interference conditions, and the transmission stability is assessed by the number of signal transmission interruptions in one hour. This is based on the defined frequency-area matching degree. A frequency-region matching evaluation table is constructed. Preferably, an adaptation threshold can be set simultaneously to directly eliminate unqualified candidate frequencies. Table 1 shows the frequency-region matching evaluation table for some construction areas.

[0079] Table 1 Frequency-Region Matching Assessment Table, Percentage (0%-100%)

[0080]

[0081] S4. Define the current construction progress and construction area. Set the frequency of the radiation signal emitted by the safety helmets worn by construction workers according to the frequency-area matching degree evaluation table, and find the corresponding frequency-area matching degree. The candidate frequency with the highest matching degree is selected as the radiation signal frequency of the safety helmet in the construction area. Preferably, the frequency parameters are written through the built-in wireless communication module of the safety helmet to ensure that the signal frequency of the safety helmets of all workers in the same construction area is consistent. Different construction areas can independently set frequencies based on the evaluation results to avoid cross-regional signal interference; combined with the frequency-region matching degree... The positioning weights used for multi-frequency data fusion are calculated based on the initial positioning accuracy level. :

[0082] .

[0083] S5, at each sensing node, receives the radiation signal emitted by the safety helmets worn by construction workers and uploads it to the data processing center. The signals received by all sensing nodes are fused. The fused received signal is divided into J segments in the time domain, each segment containing N data points. DFT processing is performed on each segment. Data with the same frequency are taken from the DFT results of the J segments and combined into N narrowband received signals of the same frequency.

[0084] Considering the location area is as follows Figure 2 The localization scenario involves placing L sensing nodes at the edge of a construction area, with their positions as follows: Assuming there are within the location of the construction area There are [number] staff members, and their positions are as follows: Assume that each worker's smart safety helmet emits uncorrelated broadband stationary signals, with the signal's start frequency and cutoff frequency being respectively... and . No. Received signals from each sensing node It can be represented as:

[0085]

[0086] in The number of sampling points. As additive noise, it is generally assumed to obey high-order noise. Additive white noise with a Sigmund distribution. This represents the signal emitted by the smart safety helmet worn by the kth worker.

[0087] To improve processing efficiency, it is necessary to integrate the received signals from all sensing nodes. The expression for the fused received signal is as follows:

[0088]

[0089] The fused received signal is segmented in the time domain, and then... Divided into J segments, each segment then has The data in the column is expressed in the following specific formula.

[0090]

[0091] We can get The expression for each segment after segmentation is as follows:

[0092]

[0093] in

[0094]

[0095] By performing DFT processing on each segment of the signal, we can obtain

[0096]

[0097] in yes The frequency domain representation, It is the frequency domain representation of the noise signal. The expression is as follows

[0098]

[0099] in

[0100] .

[0101] Perform DFT processing on the signal of each segment to obtain Each of these J segments has N columns of data. Taking the common columns from each segment to form a new signal data, this data is the narrowband received signal at the same frequency, defined as...

[0102] .

[0103] S6. Calculate the covariance matrix of N narrowband received signals at the same frequency and perform eigenvalue decomposition to separate the signal subspace and noise subspace. Construct a cost function using the orthogonality relationship of the noise subspace at each frequency point and the positioning weight. Then, grid the smart construction site area and calculate the function value of the cost function at each grid point. Find the largest number of peaks of the cost function and use the corresponding grid positions as the real-time positions of the corresponding personnel.

[0104] Calculating the covariance matrix of these N sets of data and performing eigenvalue decomposition yields the following results:

[0105]

[0106]

[0107] in, and It is a diagonal matrix composed of eigenvalues. and It is a matrix composed of eigenvectors. It is a signal subspace composed of eigenvectors corresponding to K large eigenvalues. It is a noise subspace composed of eigenvectors corresponding to LK small eigenvalues.

[0108] Based on the orthogonality between the noise subspace and the direction vector, we can obtain

[0109]

[0110] in

[0111]

[0112] It is the expression for the guide vector. yes The propagation delay between the current location and the l-th sensing node can be used to obtain the cost function:

[0113]

[0114] In smart construction sites, the area is typically a two-dimensional space, divided into several grids. Then, the corresponding data for each grid point is... Substitute the values ​​into the equation above, calculate the corresponding function values, and find the K peaks with the largest cost function. The corresponding grid points are the location results.

[0115] like Figure 3 As shown, the two-dimensional spatial spectrum calculated within the working area using the method of this invention clearly characterizes the location information of the staff. It can be observed that the spatial spectrum exhibits two highly significant spectral peaks at coordinates (24, 52) m and (68, 78) m, which perfectly match the preset actual location of the staff. This phenomenon reveals in principle that the actual spatial location of the staff is directly reflected as a local maximum point in the spatial spectrum function. Therefore, the positioning mechanism proposed in this invention is theoretically sound. Based on this principle, the positioning process can be transformed into an efficient peak search problem; that is, by identifying and extracting the coordinates corresponding to the K largest peaks in the two-dimensional spatial spectrum, the positioning results of multiple radiation sources can be directly output.

[0116] Figure 4 This is a scatter plot of the localization results of the method described in the invention, where two radiation sources are considered to be located at... and The four observation stations are located at... , , and Take the number of snapshots. , the number of array elements The signal-to-noise ratio is 10dB. From Figure 4 As can be seen, the method of the present invention can locate radiation sources with high precision.

[0117] The performance estimation standard of this invention is the root mean square error (RMSE). The RMSE for radiation source localization is defined as follows:

[0118]

[0119] in, Indicates the first The estimation results of the location of the k-th radiation source in the second experiment. This indicates the number of Monte Carlo simulations. In this section, the number of Monte Carlo simulations is set to 1000.

[0120] Figure 5 The positioning performance of the proposed method and the comparative method is demonstrated, presented as a curve of root mean square error versus signal-to-noise ratio. The simulation environment and parameter settings are consistent with... Figure 4 Maintain consistency. Analysis Figure 5 As can be seen, under different signal-to-noise ratio conditions, the method of the present invention consistently exhibits a lower root mean square error, demonstrating its significant advantage in positioning accuracy.

[0121] Figure 6 The positioning performance of the proposed method and the comparative method is further demonstrated, presented as a curve of root mean square error versus the number of signal snapshots. Experimental setup and... Figure 4 Same. From Figure 6 The results show that the method of the present invention achieves better positioning accuracy in all test intervals, further verifying the effectiveness of the algorithm.

[0122] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0123] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for high-precision real-time positioning of personnel in smart construction sites based on multi-frequency data fusion with dynamic adaptation to construction progress, characterized in that... The method Includes the following steps: S1, collect relevant information about the construction project, including project type, total duration, work site drawings, list of large equipment, and list of construction personnel; based on this information, divide the construction schedule into... There are several construction phases, each corresponding to... One construction area; S2, for each construction progress, select the construction area one by one, and set the initial positioning accuracy level according to the operation type and spatial parameters of the construction area. , Representation phase ,area The initial positioning accuracy level is determined by the value, with a higher value indicating a higher positioning accuracy requirement. This is then combined with the interference type and regional characteristics of the construction area to match the appropriate frequency range for the safety helmet signal. S3, within the adapted frequency range, defines the frequency-area matching degree for each construction area at different construction stages based on the physical characteristics of the construction area. , Indicates the stage ,area Frequency of use Applicability degree; frequency-region matching degree based on definition Construct a frequency-region matching evaluation table; S4. Define the current construction progress and construction area. Set the frequency of the radiation signal emitted by the safety helmets worn by construction workers according to the frequency-area matching degree evaluation table, and find the corresponding frequency-area matching degree. Combined with frequency-region matching degree The positioning weights used for multi-frequency data fusion are calculated based on the initial positioning accuracy level. S5, at each sensing node, receives the radiation signal emitted by the safety helmets worn by construction workers and uploads it to the data processing center. The signals received by all sensing nodes are fused. The fused received signal is then segmented in the time domain into J segments, each segment containing N data points. DFT processing is performed on each segment. Take the same frequency data from the DFT results of the J-band signal and combine them into N narrowband received signals of the same frequency; S6. Calculate the covariance matrix of N narrowband received signals at the same frequency and perform eigenvalue decomposition to separate the signal subspace and noise subspace; construct a cost function using the orthogonality relationship of the noise subspace at each frequency point and the positioning weight; then grid the smart construction site area and calculate the function value of the cost function at each grid point. Find the largest peaks in the cost function and use their corresponding grid positions as the real-time positions of the corresponding personnel.

2. The method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion as described in claim 1, is characterized in that... Step S1 further includes: Collect relevant information about the construction project, including project type, total construction period, work area drawings, list of large equipment, and list of construction personnel. The work area drawings include the dimensions of the construction area, the distribution of obstacles, and the deployment locations of sensing nodes. The list of large equipment includes equipment models, operating ranges, and electromagnetic interference parameters. The list of construction personnel includes the number of personnel and their job types. The construction schedule is broken down according to the total project duration. The construction process is divided into several phases, and within each phase, the site is further divided according to the work area drawings and the distribution of large equipment. Each construction area is clearly defined, specifying its operational boundaries and types.

3. The method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion as described in claim 1, is characterized in that... Step S2 further includes: For each construction progress, a construction area is selected one by one. Based on the spatial parameters of the construction area related to the openness of the space, the type of obstruction, and the electromagnetic environment, the spatial complexity of the construction area is analyzed. Then, the corresponding operation risk level is analyzed based on the operation type of the construction area. The initial positioning accuracy level is set according to the operational risk level and spatial complexity of the construction area. ; Based on the openness of the construction area, the type of obstructions, and the initial positioning accuracy level The usable frequency bands were initially screened out; then, based on the regional spatial parameters related to the electromagnetic environment, the frequency bands that overlapped with the interference frequency bands of the construction area were eliminated, and finally the suitable frequency range of the construction area was determined.

4. The method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion as described in claim 1, is characterized in that... In step S4, the positioning weights used for multi-frequency data fusion are calculated using the following formula. : ； In the formula, N is the total number of frequency points in the signal frequency domain, where n is the nth frequency point. This represents the signal frequency at the nth frequency point.

5. The method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion as described in claim 1, characterized in that, Step S5 further includes: L sensing nodes are placed at the edge of the construction area, with the following locations: , l represents the l-th sensing node; within the construction area there are There are [number] staff members, located at [location]. Let k represent the k-th worker; assuming that each worker's smart safety helmet emits uncorrelated broadband stationary signals, the k-th worker... Each sensing node The received signal at time t is represented as: ； in The number of sampling points. Additive noise, The signal emitted by the smart safety helmet worn by the k-th worker is represented by: The received signals from all sensing nodes are integrated to obtain the fused received signal: ； For the fused received signal Perform time-domain segmentation, dividing it into J segments. After segmentation, each segment has... The data in the column is expressed in the following specific formula: ； The expression for each segment after segmentation is as follows: ； in ; in This represents the j-th segment of the signal emitted by the smart safety helmet worn by the k-th worker. It is noise signal during transmission; Perform DFT processing on the signal of each segment to obtain : ； in yes The frequency domain representation, It is the frequency domain representation of the noise signal. The expression is as follows ； in ,in , These are the x and y coordinates of the l-th sensing node in the coordinate system. and is the x and y coordinates of the kth worker, and c is the propagation speed of electromagnetic signals in space; Extracting the same column of data from each segment to form a new signal data, resulting in N narrowband received signals at the same frequency: 。 6. The method for high-precision real-time positioning of personnel in smart construction sites based on dynamic adaptation of construction progress and multi-frequency data fusion as described in claim 1, is characterized in that... Step S6 further includes: Calculate the covariance matrix of N narrowband received signals at the same frequency and perform eigenvalue decomposition on it to obtain: ； ； in, and It is a diagonal matrix composed of eigenvalues. and It is a matrix composed of eigenvectors, with the eigenvalues ​​sorted from largest to smallest. It is the signal subspace composed of the eigenvectors corresponding to the first K eigenvalues. It is a noise subspace composed of the eigenvectors corresponding to the remaining LK eigenvalues; Based on the orthogonality between the noise subspace and the direction vector, we obtain ; in It is the expression for the guide vector, where It is the coordinate vector in the smart construction site area, in It is a location The propagation delay with the l-th sensing node is used to construct the cost function as follows: ； The smart construction site area is gridded, and the function value of the cost function is calculated at each grid point; Find the largest peaks in the cost function and use their corresponding grid positions as the real-time positions of the corresponding personnel.

Images