High-speed railway roadbed filler parameter optimization method and system

Abstract

The invention relates to the technical field of railway subgrade engineering, in particular to a high-speed railway subgrade filler parameter optimization method and system. The method comprises the following steps: acquiring to-be-detected high-speed railway data; carrying out roadbed surface multi-frequency electromagnetic wave scanning on the to-be-detected high-speed railway data to obtain original filler electromagnetic response signal data; response signal space-time registration is carried out on the original filler electromagnetic response signal data, roadbed filler physical feature extraction is carried out, and roadbed filler physical feature data are generated; performing abnormal region spatial feature statistics according to the roadbed filler physical feature data, setting a partition filler construction scheme, and generating intelligent partition filler construction scheme data; and according to the intelligent partition filling construction scheme data, subgrade filling parameter ratio adjustment is carried out, so that high-speed railway subgrade filling parameter optimization operation is achieved. According to the method, accurate optimization of filler parameters is realized through railway roadbed filler partition defect evaluation, and the roadbed construction efficiency and quality are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of railway roadbed engineering, and in particular to a method and system for optimizing parameters of high-speed railway roadbed filler. Background Art

[0002] The construction of high-speed railways involves a large amount of earthwork, among which roadbed filling is one of the key links. The stability and durability of the roadbed directly affect the safety and comfort of high-speed railway operations, as well as the maintenance cost throughout the life cycle. Therefore, how to effectively control the quality of the roadbed filling material to ensure the long-term stability of the roadbed. The existing high-speed railway roadbed construction and filling material parameter adjustment are adjusted through manual experience adjustment, manual compaction and ring knife sampling. For example, manual compaction depends on the experience and operation level of the construction personnel, and it is difficult to ensure that the compaction of the filling material meets the design requirements, and it is difficult to detect the compaction inside the filling material. Although ring knife sampling can obtain the compaction data of some filling materials, the number of samples is limited, and the selection of sampling positions is somewhat arbitrary, which cannot fully and accurately reflect the overall compaction quality of the filling material. This local sampling method is prone to ignore local uneven or insufficiently compacted areas, resulting in problems such as subsidence, deformation, and even cracking of the roadbed in the later stage, causing huge economic losses and safety hazards. However, traditional high-speed railway subgrade construction mainly relies on manual experience and simple detection methods to adjust the filler parameter ratio and compaction process, which makes it difficult to achieve precise control of the compaction quality of the subgrade filler, resulting in the filler compaction effect being difficult to achieve the optimal state, increasing the risk of later subsidence, deformation and even cracking of the subgrade. Summary of the invention

[0003] Based on this, the present invention provides a method and system for optimizing parameters of high-speed railway roadbed filling materials to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above object, a method for optimizing parameters of high-speed railway roadbed filling material comprises the following steps:

[0005] Step S1: acquiring the high-speed railway data to be detected; performing electromagnetic wave scanning frequency processing on the high-speed railway data to be detected, and generating an electromagnetic wave emission control strategy; performing multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy, and obtaining the original filler electromagnetic response signal data and the roadbed sampling position data respectively;

[0006] Step S2: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; performing electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data;

[0007] Step S3: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; performing roadbed filler physical feature extraction based on the multi-frequency electromagnetic response signal grouping data to generate roadbed filler physical feature data; performing abnormal area spatial feature statistics based on the roadbed filler physical feature data to generate abnormal area spatial feature data; performing partition filler construction plan setting based on the abnormal area spatial feature data to generate intelligent partition filler construction plan data;

[0008] Step S4: Perform roadbed filling construction processing according to the intelligent partition filling construction plan data, and adjust the roadbed filling parameter ratio to achieve high-speed railway roadbed filling parameter optimization operation.

[0009] The present invention utilizes the sensitivity of electromagnetic waves to the internal structure of the filler. Through multi-frequency electromagnetic wave scanning and fine signal analysis, it can deeply detect the compaction degree, uniformity and potential defects of the roadbed filler, and completely get rid of the subjectivity and uncertainty of traditional manual experience judgment. Through the spatiotemporal registration of electromagnetic response signals and effective signal screening, it can accurately identify the differences in physical characteristics inside the filler, and then accurately judge whether the compaction quality meets the standard, effectively avoiding the problem of insufficient or uneven local compaction, and laying a solid foundation for the long-term stability and safety of the roadbed. The automated electromagnetic wave scanning technology adopted can quickly and widely acquire filler data, and through intelligent data processing and analysis, quickly generate filler physical characteristic data and abnormal area spatial characteristic data, greatly shortening the detection time and improving construction efficiency. At the same time, the intelligent partition filler construction plan generated according to the analysis results can guide the construction personnel to carry out targeted filler parameter ratio adjustment and compaction operations, ensuring that each layer of filler can achieve the best compaction effect, avoiding blind construction and repeated rework, and significantly improving the overall construction quality of the roadbed. By accurately controlling the compaction quality of the roadbed filler, it is possible to fundamentally prevent the occurrence of problems such as roadbed settlement, deformation and cracking. The long-term stability of the roadbed is guaranteed, reducing the frequency and workload of later maintenance and repair, thereby reducing the maintenance cost of the entire life cycle. At the same time, since the safety hazards caused by roadbed quality problems are avoided, the safety and reliability of high-speed railway operations are also improved, ensuring the travel safety of passengers and the smooth operation of railway transportation. Therefore, a high-speed railway roadbed filler parameter optimization method of the present invention analyzes the filler characteristics of different regions through multi-frequency electromagnetic response signal grouping and physical feature extraction, based on the spatial feature statistics of abnormal areas, intelligently identifies and targetedly processes insufficient compaction or other abnormal areas, and realizes the refined adjustment of roadbed filler parameters, ultimately improving the roadbed construction quality and stability, and effectively reducing the risk of later settlement and deformation.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: acquiring the high-speed railway data to be detected; extracting the roadbed filling material and the path width from the high-speed railway data to be detected, and obtaining the railway roadbed width data and the roadbed filling material parameters respectively;

[0012] Step S12: using a preset mobile inspection vehicle to carry a multi-band electromagnetic wave transmitting and receiving device to construct an electromagnetic integrated scanning device;

[0013] Step S13: activating the detection sensor array of the electromagnetic integrated scanning device based on the railway roadbed width data to generate sensor array distribution data;

[0014] Step S14: determining the initial position of the roadbed for the preset mobile inspection vehicle, setting the travel speed according to the high-speed railway data to be inspected, and generating the travel speed data of the inspection vehicle;

[0015] Step S15: performing electromagnetic wave scanning frequency processing according to the roadbed filling material parameters and the detection vehicle driving speed data, and performing electromagnetic wave emission control processing on the electromagnetic integrated scanning device through the sensor array distribution data to generate an electromagnetic wave emission control strategy;

[0016] Step S16: Based on the electromagnetic wave emission control strategy, the electromagnetic integrated scanning device is used to perform multi-frequency electromagnetic wave scanning on the roadbed surface, and the electromagnetic wave reflection signal is synchronously received to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively.

[0017] The present invention provides a key basis for the formulation of subsequent electromagnetic wave scanning strategies by extracting the roadbed width and filling material parameters, ensuring the pertinence and accuracy of the detection. Secondly, by using a preset mobile inspection vehicle equipped with an electromagnetic integrated scanning device, automated, large-scale roadbed detection is achieved, which greatly improves the detection efficiency and reduces the errors caused by manual intervention. At the same time, the sensor array activated based on the roadbed width data ensures the comprehensive coverage of the electromagnetic wave signal and the integrity of the data collection. By accurately controlling the driving speed of the inspection vehicle and combining the roadbed filling material parameters, the electromagnetic wave scanning frequency is intelligently adjusted, the penetration and resolution of the electromagnetic wave signal are optimized, and the detection results are more precise and reliable. The generated electromagnetic wave emission control strategy further ensures the orderliness and efficiency of the electromagnetic wave scanning process. Finally, by synchronously receiving the electromagnetic wave reflection signal, the original electromagnetic response signal data containing rich filler physical characteristic information and accurate roadbed sampling position data can be obtained, laying a solid data foundation for subsequent signal processing and compaction quality evaluation.

[0018] Preferably, step S15 comprises the following steps:

[0019] Step S151: Analyze the material magnetic permeability characteristics according to the roadbed filling material parameters to generate roadbed material characteristic data;

[0020] Step S152: Processing the transmission frequency of the transmitting device according to the driving speed data of the detection vehicle to generate electromagnetic transmission frequency data;

[0021] Step S153: matching the electromagnetic wave scanning frequency range with the electromagnetic emission frequency data to generate electromagnetic wave scanning frequency range data;

[0022] Step S154: dividing the electromagnetic wave scanning frequency range data into high and low frequency scanning frequency segments according to the roadbed material characteristic data to generate electromagnetic wave scanning frequency sequence data;

[0023] Step S155: Control the phased array beam direction of the electromagnetic integrated scanning device through the sensor array distribution data, and input the electromagnetic wave scanning frequency according to the electromagnetic wave scanning frequency sequence data, so as to obtain the electromagnetic wave emission control strategy.

[0024] The present invention analyzes the magnetic permeability characteristics of the roadbed filling material, and the obtained roadbed material characteristic data provides a scientific basis for the precise setting of the electromagnetic wave scanning frequency, ensuring the optimal coupling of the electromagnetic wave and the material, thereby improving the signal penetration and data validity. Secondly, the transmission frequency of the transmitting device is adjusted according to the driving speed of the detection vehicle, and combined with the electromagnetic wave scanning frequency range matching, the continuity and integrity of data acquisition are ensured, and signal distortion or omission caused by speed changes is avoided. The scanning frequency is segmented into high and low frequencies according to the characteristics of the roadbed material, and electromagnetic wave scanning frequency sequence data is generated, so that the electromagnetic wave can be targeted for scanning different material characteristics, thereby improving the detection accuracy of the roadbed composed of complex materials. Finally, the sensor array distribution data is used to control the phased array beam direction, and combined with the precise electromagnetic wave scanning frequency sequence input, the directional scanning of the electromagnetic beam and the fine adjustment of the frequency are realized, and the detection capability of specific depths and areas is enhanced.

[0025] Preferably, step S2 comprises the following steps:

[0026] Step S21: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data;

[0027] Step S22: performing adaptive wavelet denoising on the time series data of the electromagnetic response signal of the filler, and performing response signal error correction to obtain corrected electromagnetic response signal time series data;

[0028] Step S23: performing short-time Fourier transform on the time series data of the corrected electromagnetic response signal to generate frequency domain data of the filler electromagnetic response;

[0029] Step S24: performing signal amplitude and phase value processing on the filler electromagnetic response frequency domain data, and performing signal energy response threshold calculation to generate signal energy response threshold data;

[0030] Step S25: Screening the effective electromagnetic signals of the filler electromagnetic response signal time series data according to the preset electromagnetic frequency determination rules and signal energy response threshold data to generate effective electromagnetic response signal data.

[0031] The present invention generates the time series data of the electromagnetic response signal of the filler by time-space registration of the response signal, ensures the uniformity of the signal data in time and space, and lays the foundation for subsequent analysis. Secondly, by using adaptive wavelet denoising and error correction, the noise and interference in the signal are effectively removed, the purity and authenticity of the signal are improved, and the subsequent frequency domain analysis is more accurate. The time series data is converted into frequency domain data by short-time Fourier transform, which can reveal the distribution characteristics of the signal at different frequencies, and provide key information for in-depth analysis of the physical properties of the filler. Subsequently, the intensity and energy distribution of the signal are quantified by signal amplitude and phase value processing and energy response threshold calculation, providing a clear standard for the screening of electromagnetic effective signals. Finally, based on the preset electromagnetic frequency judgment rules and signal energy response threshold data, the electromagnetic response signal of the filler is effectively screened to ensure that the extracted effective electromagnetic response signal can truly reflect the compaction state and physical properties of the filler, and eliminate the interference of irrelevant signals.

[0032] Preferably, step S3 comprises the following steps:

[0033] Step S31: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data according to the roadbed sampling position data to generate multi-frequency electromagnetic response signal group data;

[0034] Step S32: performing frequency band penetration depth analysis on the multi-frequency electromagnetic response signal group data to generate frequency band penetration depth data;

[0035] Step S33: extracting the signal propagation time of the multi-frequency electromagnetic response signal group data, and calculating the electromagnetic wave propagation speed according to the frequency band penetration depth data to generate multi-frequency electromagnetic wave propagation speed data;

[0036] Step S34: extracting the physical characteristics of the roadbed filler through the multi-frequency electromagnetic wave propagation velocity data to generate the physical characteristic data of the roadbed filler;

[0037] Step S35: Perform spatial feature statistics of abnormal areas based on the physical feature data of the roadbed filler to generate spatial feature data of the abnormal areas; set up a partitioned filler construction plan based on the spatial feature data of the abnormal areas to generate intelligent partitioned filler construction plan data.

[0038] The present invention groups multi-frequency electromagnetic response signals, and the scheme can analyze the signal characteristics of different frequencies, and more accurately capture the physical state of the filler at different depths and regions. Secondly, through the analysis of frequency band penetration depth and the calculation of electromagnetic wave propagation velocity, the scheme can accurately obtain the propagation characteristics of electromagnetic waves in the filler, providing key data support for the subsequent extraction of physical characteristics. By using multi-frequency electromagnetic wave propagation velocity data to extract the physical characteristics of roadbed fillers, the scheme can quantify key parameters such as the compaction degree and uniformity of the filler, providing a scientific basis for evaluating the quality of the filler. Subsequently, by performing spatial feature statistics of abnormal areas on these physical feature data, the scheme can accurately identify areas where compaction does not meet standards or where potential defects exist, avoiding the limitations of traditional methods that rely on empirical judgment. Finally, the intelligent partitioned filler construction plan generated based on the spatial feature data of abnormal areas provides clear guidance for construction personnel, and realizes targeted construction and refined management.

[0039] Preferably, step S34 includes the following steps:

[0040] Step S341: performing dielectric constant inversion according to the multi-frequency electromagnetic wave propagation velocity data to generate dielectric constant distribution data;

[0041] Step S342: performing electromagnetic frequency weighting on the dielectric constant distribution data, and performing weighted fusion of multi-frequency data to generate weighted fusion dielectric constant data;

[0042] Step S343: Calculate the electromagnetic wave attenuation coefficient according to the multi-frequency electromagnetic wave propagation velocity data and the weighted fusion dielectric constant data, and perform water content contribution processing to generate roadbed water content distribution data;

[0043] Step S344: using a preset filler density-dielectric constant linear model to estimate the filler density value of the weighted fused dielectric constant data, and generating filler density distribution characteristic data;

[0044] Step S345: performing filler particle distribution index processing on the roadbed moisture content distribution data through filler density distribution characteristic data to generate roadbed particle distribution index parameters;

[0045] Step S346: Integrate the filler density distribution characteristic data, the roadbed moisture content distribution data and the roadbed particle distribution index parameters into the roadbed filler physical characteristic data to generate the roadbed filler physical characteristic data.

[0046] The present invention effectively improves the accuracy and reliability of dielectric constant data through the inversion of dielectric constant and weighted fusion of multi-frequency data, laying the foundation for subsequent physical characteristic analysis. As an important electromagnetic parameter of the material, the accurate measurement of dielectric constant is crucial for evaluating the compaction and uniformity of the filler. The scheme realizes the accurate measurement of the water content distribution of the roadbed by combining the electromagnetic wave propagation velocity and the weighted fusion dielectric constant data to calculate the electromagnetic wave attenuation coefficient and perform water content contribution processing. Water content is an important factor affecting the mechanical properties of the filler, and its accurate measurement helps prevent the settlement and deformation of the roadbed. In addition, the filler density is estimated using the filler density-dielectric constant linear model, which further enriches the dimension of the filler's physical characteristics. Through the filler density data, the water content data is processed by the filler particle distribution index, and the scheme successfully extracts the roadbed particle distribution index parameters, and together with the density and water content data, it constitutes a complete roadbed filler physical characteristic data. The implementation of this series of steps not only realizes the comprehensive quantification of key physical characteristics such as filler density, water content and particle distribution, but also improves the accuracy and reliability of these characteristics through data fusion and comprehensive analysis.

[0047] Preferably, step S35 includes the following steps:

[0048] Step S351: Acquire historical railway roadbed filling engineering data; construct a mapping relationship between filler physical characteristics and compaction degree based on the historical railway roadbed filling engineering data to obtain filler physical characteristics-compactness mapping relationship data;

[0049] Step S352: using a preset convolutional neural network model to perform model transfer learning on filler physical characteristics-compactness mapping relationship data, and using historical railway roadbed filler engineering data to perform model training to generate a roadbed filler compaction assessment model;

[0050] Step S353: transmitting the roadbed filler physical characteristic data to the roadbed filler compaction assessment model for compaction assessment to generate filler compaction data;

[0051] Step S354: matching roadbed sampling points according to the roadbed filler physical characteristic data, and performing compaction spatial distribution difference according to the filler compaction data to generate roadbed compaction distribution data;

[0052] Step S355: Process the roadbed defect construction plan based on the roadbed compaction distribution data and the roadbed filler physical characteristic data to generate intelligent roadbed filler construction plan data.

[0053] The present invention constructs a mapping relationship between filler physical characteristics and compaction by using historical engineering data, and applies this relationship to the transfer learning and training of the convolutional neural network model. The scheme successfully constructs a prediction model specifically for the compaction evaluation of roadbed fillers. This process makes full use of existing engineering experience and data resources, and improves the prediction accuracy and generalization ability of the model. The extracted roadbed filler physical characteristic data is input into the trained compaction evaluation model, which can quickly and accurately generate filler compaction data and realize the quantitative evaluation of compaction. This machine learning-based evaluation method avoids the interference of human factors in traditional methods and improves the objectivity and reliability of the evaluation. At the same time, through the matching of roadbed sampling points and the difference processing of compaction spatial distribution, intuitive roadbed compaction distribution data is generated, so that construction personnel can clearly understand the spatial distribution characteristics of filler compaction. Based on the compaction distribution data and filler physical characteristic data, the roadbed defect construction plan is processed, and the generated intelligent roadbed filler construction plan can solve the areas with insufficient compaction or defects in a targeted manner, realizing the intelligent and refined management of the construction plan. This data-driven construction plan optimization method not only improves construction efficiency, but also reduces construction costs, and ultimately ensures the long-term stability and safety of the high-speed railway subgrade.

[0054] Preferably, step S355 includes the following steps:

[0055] Step S3551: using the preset roadbed filler early warning index library to perform cluster analysis on the measured roadbed condition on the roadbed filler physical characteristic data and roadbed compaction distribution data, and marking abnormal areas to generate preliminary filler abnormal area data;

[0056] Step S3552: Perform spatial connectivity analysis on the preliminary filler abnormal area data, and merge adjacent abnormal areas to obtain roadbed filler abnormal area data;

[0057] Step S3553: Perform regional spatial feature statistics on the abnormal area data of the roadbed filler to generate abnormal area spatial feature data;

[0058] Step S3554: Set up the roadbed filling construction plan according to the spatial characteristic data of the abnormal area, and generate intelligent roadbed filling construction plan data.

[0059] The present invention realizes accurate identification of abnormal areas of subgrade fillers and formulation of targeted construction plans by deeply analyzing and intelligently processing the physical characteristics and compaction data of subgrade fillers, thereby greatly improving the quality and efficiency of high-speed railway subgrade construction. First, by using the preset subgrade filler early warning index library to cluster the measured subgrade conditions, the scheme can automatically identify abnormal areas that do not meet the early warning indicators, mark them, and generate preliminary filler abnormal area data. This process avoids the limitations of relying on manual experience judgment in traditional methods and improves the accuracy and efficiency of abnormal area identification. By performing spatial connectivity analysis and merging adjacent areas on the preliminary filler abnormal area, the scheme further optimizes the division of abnormal areas, making the final determined subgrade filler abnormal area more in line with the actual situation and avoiding misjudgment caused by isolated abnormal points. Subsequently, the spatial feature statistics of the abnormal area are performed, and the generated abnormal area spatial feature data describes in detail the key information such as the location, size, and shape of the abnormal area. The construction plan of the roadbed filling material is set based on the spatial characteristic data of the abnormal area. The generated intelligent roadbed filling material construction plan can provide customized construction strategies according to the characteristics of different abnormal areas, such as adjusting the filling material ratio, increasing the number of compaction times, or using special construction technology. This intelligent construction plan can not only effectively solve the abnormal problem of roadbed filling material, but also improve construction efficiency, reduce construction costs, and ultimately ensure the long-term stability and safety of the high-speed railway roadbed.

[0060] Preferably, step S3554 includes the following steps:

[0061] The abnormal area data of the roadbed filling abnormal area is located through the abnormal area spatial feature data, and the roadbed filling risk area is divided to obtain the roadbed filling high risk area data and the roadbed filling low risk area data respectively;

[0062] Perform fine grid division on the high-risk area data of the roadbed filler to generate fine grid data of the high-risk area;

[0063] The physical characteristic data of roadbed filler is used to calculate the filler physical anomaly value of the refined grid data in the high-risk area, and the filler physical anomaly data in the high-risk area is generated;

[0064] The compaction loss of high-risk areas is calculated based on the refined grid data of high-risk areas through the compaction distribution data of the roadbed and the physical anomaly data of the filler in the high-risk areas, and the compaction loss data of the high-risk areas is generated;

[0065] Set up multi-dimensional reinforcement plans based on the compaction loss data of high-risk areas and generate reinforcement plan data for high-risk areas;

[0066] Use the physical characteristic data of roadbed fill to conduct routine quality assessment of low-risk areas of roadbed fill, and generate assessment benchmark values ​​for low-risk areas;

[0067] Conduct construction process adaptability analysis based on the low-hazard area assessment benchmark, optimize standard construction parameters, and generate construction plan data for low-hazard areas;

[0068] The construction plan data of low-hazard areas and the reinforcement plan data of high-hazard areas are integrated into partitioned construction plans to obtain intelligent partitioned filling material construction plan data.

[0069] The present invention locates and divides the risks of abnormal areas through the spatial characteristic data of abnormal areas, clarifies the high-risk and low-risk areas, and lays the foundation for subsequent differentiated processing. The high-risk areas are finely gridded and the physical abnormal values ​​of the fillers and the compaction loss are calculated, so that the identification of the problem is more specific and quantitative, and provides an accurate basis for the formulation of targeted reinforcement plans. The multi-dimensional reinforcement plan based on compaction loss data ensures that the high-risk areas are effectively handled and reduces potential safety risks. For low-risk areas, the plan implements standardized management of low-risk areas by using the physical characteristic data of the fillers for conventional quality assessment and generating assessment benchmark values. On this basis, the construction process adaptability analysis and standard construction parameter optimization are carried out, and the generated low-risk area construction plan not only ensures the construction quality, but also avoids resource waste. This differentiated processing strategy based on risk level enables the construction resources to be reasonably allocated and improves construction efficiency. The high-risk area reinforcement plan and the low-risk area construction plan are integrated to generate an intelligent partition filler construction plan, which realizes the comprehensive optimization of the entire roadbed filler construction. The solution not only solves the hidden dangers in high-risk areas, but also improves the construction quality in low-risk areas, ensuring the overall stability and durability of the roadbed. Through this refined and differentiated construction management, the solution effectively reduces the subsequent maintenance costs and improves the operational safety and reliability of high-speed railways.

[0070] Preferably, the present invention further provides a high-speed railway roadbed filling parameter optimization system, which executes the high-speed railway roadbed filling parameter optimization method as described above, and the high-speed railway roadbed filling parameter optimization system comprises:

[0071] The roadbed electromagnetic scanning module is used to obtain the high-speed railway data to be detected; perform electromagnetic wave scanning frequency processing on the high-speed railway data to be detected to generate an electromagnetic wave emission control strategy; perform multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively;

[0072] The electromagnetic response signal analysis module is used to perform time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; perform electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data;

[0073] The roadbed construction processing module is used to perform multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; extract the physical characteristics of the roadbed filler according to the multi-frequency electromagnetic response signal grouping data to generate the physical characteristic data of the roadbed filler; perform spatial characteristic statistics of the abnormal area according to the physical characteristic data of the roadbed filler to generate the spatial characteristic data of the abnormal area; set the partition filler construction plan according to the spatial characteristic data of the abnormal area to generate the intelligent partition filler construction plan data;

[0074] The subgrade filling parameter optimization module is used to carry out subgrade filling construction processing according to the intelligent partition filling construction plan data, and adjust the subgrade filling parameter ratio to achieve high-speed railway subgrade filling parameter optimization operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A schematic diagram of the steps of a method for optimizing parameters of high-speed railway roadbed fillers according to the present invention;

[0076] Figure 2 for Figure 1 Detailed implementation steps of step S1 in FIG.

[0077] Figure 3 for Figure 1 Detailed implementation steps of step S2 in FIG.

[0078] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0079] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0080] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0081] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0082] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for optimizing parameters of high-speed railway roadbed filling materials, comprising the following steps:

[0083] Step S1: acquiring the high-speed railway data to be detected; performing electromagnetic wave scanning frequency processing on the high-speed railway data to be detected, and generating an electromagnetic wave emission control strategy; performing multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy, and obtaining the original filler electromagnetic response signal data and the roadbed sampling position data respectively;

[0084] Step S2: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; performing electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data;

[0085] Step S3: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; performing roadbed filler physical feature extraction based on the multi-frequency electromagnetic response signal grouping data to generate roadbed filler physical feature data; performing abnormal area spatial feature statistics based on the roadbed filler physical feature data to generate abnormal area spatial feature data; performing partition filler construction plan setting based on the abnormal area spatial feature data to generate intelligent partition filler construction plan data;

[0086] Step S4: Perform roadbed filling construction processing according to the intelligent partition filling construction plan data, and adjust the roadbed filling parameter ratio to achieve high-speed railway roadbed filling parameter optimization operation.

[0087] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a schematic diagram of a step flow chart of a method for optimizing parameters of a high-speed railway roadbed filler according to the present invention. In this embodiment, the method for optimizing parameters of a high-speed railway roadbed filler comprises the following steps:

[0088] Step S1: acquiring the high-speed railway data to be detected; performing electromagnetic wave scanning frequency processing on the high-speed railway data to be detected, and generating an electromagnetic wave emission control strategy; performing multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy, and obtaining the original filler electromagnetic response signal data and the roadbed sampling position data respectively;

[0089] In an embodiment of the present invention, the high-speed railway subgrade data to be detected is obtained. These data include the design parameters, construction records, historical detection data, etc. of the subgrade. For example, the design parameters such as the length, width, height, filler type, compaction degree, and past settlement monitoring data, deformation monitoring data, etc. of a certain section of the high-speed railway subgrade can be obtained. Then, the acquired subgrade data is subjected to electromagnetic wave scanning frequency processing. According to factors such as the type, thickness, and water content of the subgrade filler, a suitable electromagnetic wave frequency range is selected, and the number of scans and intervals of each frequency are determined. For example, for clay fillers, a lower frequency range (e.g., 1MHz-10MHz) can be selected, while for gravel fillers, a higher frequency range (e.g., 10MHz-100MHz) can be selected. By analyzing the propagation characteristics of electromagnetic waves of different frequencies in the subgrade filler, an electromagnetic wave emission control strategy is generated, which includes parameters such as the emission power, emission time, and scanning path of electromagnetic waves of different frequencies. For example, the emission angle and scanning path of the electromagnetic wave can be set according to the geometry of the subgrade and the distribution of the filler to ensure that the electromagnetic wave can effectively penetrate the subgrade filler and obtain sufficient reflected signals. Finally, based on the generated electromagnetic wave emission control strategy, a multi-frequency electromagnetic wave emission device is used to scan the roadbed surface. The emission device is installed on a mobile platform, and the roadbed surface is scanned along a predetermined path, and the three-dimensional coordinate information of each sampling point is recorded as the roadbed sampling position data. At the same time, the receiving device receives the electromagnetic wave signal reflected back by the roadbed filler, and records the electromagnetic wave amplitude, phase and other information at each frequency as the original filler electromagnetic response signal data.

[0090] Step S2: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; performing electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data;

[0091] In an embodiment of the present invention, the acquired original filler electromagnetic response signal data is time-space aligned. Due to the time delay and position deviation in the process of electromagnetic wave emission and reception, the original data needs to be corrected to ensure that each data point corresponds to the correct sampling time and position. For example, the sampling position can be accurately located using GPS technology, and the sampling time can be synchronized using timestamp technology. Through time-space alignment, the electromagnetic response signal data collected at different frequencies and times can be integrated to generate filler electromagnetic response signal time series data. The time series data contains the electromagnetic response signal information of each sampling point at different times and frequencies. Then, the filler electromagnetic response signal time series data is screened for effective signals. Due to the influence of factors such as environmental noise and instrument errors, the original data contains some invalid signals or interference signals. It is necessary to remove these invalid signals according to a pre-set threshold or filtering algorithm to retain effective electromagnetic response signals. For example, the validity of the signal can be judged based on the signal-to-noise ratio, amplitude, phase and other characteristics of the signal. Signal processing methods such as wavelet transform and Fourier transform can also be used to denoise and filter the signal to improve the quality of the signal, and after effective signal screening, effective electromagnetic response signal data is generated.

[0092] Step S3: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; performing roadbed filler physical feature extraction based on the multi-frequency electromagnetic response signal grouping data to generate roadbed filler physical feature data; performing abnormal area spatial feature statistics based on the roadbed filler physical feature data to generate abnormal area spatial feature data; performing partition filler construction plan setting based on the abnormal area spatial feature data to generate intelligent partition filler construction plan data;

[0093] In an embodiment of the present invention, effective electromagnetic response signal data is grouped using roadbed sampling position data. According to the spatial position of the sampling point, the electromagnetic response signal data of adjacent areas are divided into a group to generate multi-frequency electromagnetic response signal group data. Each group represents a specific roadbed area, and its electromagnetic response signal reflects the characteristics of the filler in the area. Then, the physical characteristics of the roadbed filler are extracted according to the multi-frequency electromagnetic response signal group data. For example, the physical parameters such as the dielectric constant, conductivity, and magnetic permeability of the filler can be calculated according to the attenuation degree and phase change of electromagnetic waves of different frequencies. These physical parameters reflect the characteristics of the filler such as density, water content, and particle size. The extracted physical characteristic data is compared with the pre-established filler characteristic database to determine the type, quality, and whether there is an abnormality in the filler. Then, according to the physical characteristic data of the roadbed filler, the spatial characteristics of the abnormal area are statistically analyzed. The physical characteristic parameters of each group, such as the average value, standard deviation, maximum value, minimum value, etc., are statistically analyzed. The physical characteristic parameters are compared with the preset threshold value to identify the area with abnormal parameters, and the spatial distribution, area, shape, and other characteristics of these abnormal areas are statistically analyzed to generate the spatial characteristic data of the abnormal area. Finally, the construction plan of the zoning filler is set according to the spatial characteristic data of the abnormal area. According to the spatial distribution and filler characteristics of the abnormal area, a targeted filler construction plan is formulated. For example, for areas with insufficient density, compaction treatment can be used; for areas with too high water content, drainage measures can be used; for areas where the filler type does not meet the requirements, it can be replaced or improved. By setting the construction plan by zoning, it is possible to achieve refined control of the roadbed filler, improve construction efficiency and quality, and generate intelligent zoning filler construction plan data.

[0094] Step S4: Perform roadbed filling construction processing according to the intelligent partition filling construction plan data, and adjust the roadbed filling parameter ratio to achieve high-speed railway roadbed filling parameter optimization operation.

[0095] In an embodiment of the present invention, the intelligent partition filler construction plan data includes specific construction plans for different areas of the roadbed, such as areas where filler replacement, compaction treatment, grouting reinforcement and other operations are required, as well as the required filler type, quantity, parameter ratio and other information. Based on these data, construction personnel can be guided to perform precise construction operations. For example, the plan data indicates that the filler needs to be replaced in area A, and stipulates that the required filler type is graded crushed stone, the quantity is 100 cubic meters, and the maximum particle size is 50 mm. The construction personnel need to remove the original filler in area A, and purchase and transport 100 cubic meters of graded crushed stone in accordance with regulations to ensure that its maximum particle size does not exceed 50 mm, and then fill it into area A. For areas that need to be compacted, the plan data will specify parameters such as compaction target value, compaction number, and roller type. Construction personnel need to operate the roller according to these parameters and monitor the compaction in real time until the target value is reached. For example, area B needs to be compacted, and the plan data stipulates that the compaction target value is 95%, the compaction number is 8 times, and a vibratory roller is used. The construction personnel need to use a vibratory roller to compact area B eight times and measure the compaction degree after each compaction to ensure that the final compaction degree reaches 95%. For example, for a certain area, the data of the intelligent partition filling construction plan indicates that there is a local loose problem in the area and reinforcement treatment is required, and the specific reinforcement material ratio and construction process are given. During the roadbed filling construction process, it is necessary to continuously monitor the changes in the filling parameters, and dynamically adjust the filling parameter ratio to ensure that the various parameters of the roadbed filling are in the best state, and finally realize the optimization of high-speed railway roadbed filling parameters and improve the stability and durability of the roadbed.

[0096] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S1 in the flowchart, in this example, step S1 includes:

[0097] Step S11: acquiring the high-speed railway data to be detected; extracting the roadbed filling material and the path width from the high-speed railway data to be detected, and obtaining the railway roadbed width data and the roadbed filling material parameters respectively;

[0098] In an embodiment of the present invention, the digital design information, construction records and historical inspection data of the high-speed railway section to be inspected are obtained. These data contain information such as the geometry of the roadbed, design parameters, type of filling material used, compaction degree, etc. For example, the cross-sectional view and longitudinal section of the roadbed of a certain section, as well as the type of filler used in the section (such as graded crushed stone, clay, etc.), design compaction degree and other parameters can be obtained. Then, the roadbed filling material parameters and path width data are extracted from these data. The roadbed filling material parameters include material type, particle size distribution, density, water content, etc. The path width data refers to the width of the top surface of the roadbed. For example, by analyzing the roadbed cross-sectional view, the width of the roadbed and the type and parameters of fillers used at different depths can be determined.

[0099] Step S12: using a preset mobile inspection vehicle to carry a multi-band electromagnetic wave transmitting and receiving device to construct an electromagnetic integrated scanning device;

[0100] In the embodiment of the present invention, a mobile inspection vehicle suitable for high-speed railway roadbed inspection is selected as a platform. The mobile inspection vehicle should have stable driving performance and sufficient carrying capacity to carry electromagnetic wave transmitting and receiving devices and related auxiliary equipment. A multi-band electromagnetic wave transmitting device and a receiving device are installed on the mobile inspection vehicle. The transmitting device can generate electromagnetic wave signals of multiple different frequencies, and the receiving device is responsible for receiving the electromagnetic wave signals reflected from the roadbed filler. The selection of these devices needs to be determined according to the characteristics of the roadbed filler to be detected and the detection accuracy requirements. For example, for the situation where deep fillers need to be detected, it is necessary to select a device that can emit low-frequency electromagnetic waves; and for the situation where high-precision imaging is required, it is necessary to select a device that can emit high-frequency electromagnetic waves. The transmitting device and the receiving device are integrated to form a complete electromagnetic integrated scanning device. The device should have the ability to synchronously transmit and receive signals, and can accurately record the time information of transmission and reception. At the same time, in order to improve the detection efficiency and accuracy, multiple transmitting and receiving units can be combined into an array form to form a multi-channel electromagnetic integrated scanning device. For example, a linear array or a planar array layout can be used to achieve rapid scanning and high-resolution imaging of the roadbed. The resulting electromagnetic integrated scanning device is a key device that can efficiently and accurately collect electromagnetic response data of roadbed fillers.

[0101] Step S13: activating the detection sensor array of the electromagnetic integrated scanning device based on the railway roadbed width data to generate sensor array distribution data;

[0102] In an embodiment of the present invention, the detection sensor array range that needs to be activated on the electromagnetic integrated scanning device is determined according to the railway roadbed width data extracted in step S11. Since the roadbed width determines the scanning range, it is necessary to adjust the activation range of the sensor array according to the roadbed width to ensure that the electromagnetic wave can cover the entire roadbed width. For example, if the roadbed width is 12 meters and the total width of the sensor array is 15 meters, it is necessary to activate the part of the sensor array covering a width of 12 meters according to the position of the roadbed centerline. After activating the sensor array, the position information of each activated sensor is recorded to generate sensor array distribution data. These data describe the relative position of each sensor in the array, which is crucial for subsequent data processing and analysis. For example, the sensor array distribution data may include information such as the number, coordinates, and direction of each sensor. This information can be used to determine the spatial position corresponding to the data collected by each sensor, thereby realizing the spatial distribution analysis of the roadbed filler characteristics. In addition, the sensor array distribution data can also be used to calibrate the position deviation between sensors to improve detection accuracy.

[0103] Step S14: determining the initial position of the roadbed for the preset mobile inspection vehicle, setting the travel speed according to the high-speed railway data to be inspected, and generating the travel speed data of the inspection vehicle;

[0104] In an embodiment of the present invention, before starting the roadbed detection, it is necessary to determine the initial driving position of the mobile detection vehicle. Select a suitable starting point to ensure that the detection vehicle can perform continuous scanning along the roadbed direction to cover the entire area to be detected. For example, the initial position of the detection vehicle can be set at the starting point of the roadbed or a known reference point. After determining the initial position, record its coordinate information to facilitate spatial positioning during subsequent data processing. At the same time, the driving speed of the mobile detection vehicle is set according to the relevant information in the high-speed railway data to be detected. The setting of the driving speed needs to comprehensively consider multiple factors, including detection accuracy requirements, data acquisition rate, roadbed length, etc. For example, if high-precision detection data is required, the driving speed should be reduced to increase the data sampling density; if the detection task needs to be completed quickly, the driving speed can be appropriately increased. The set driving speed should ensure that it remains stable throughout the detection process to avoid uneven detection data caused by speed changes.

[0105] Step S15: performing electromagnetic wave scanning frequency processing according to the roadbed filling material parameters and the detection vehicle driving speed data, and performing electromagnetic wave emission control processing on the electromagnetic integrated scanning device through the sensor array distribution data to generate an electromagnetic wave emission control strategy;

[0106] In the embodiment of the present invention, since different materials have different response characteristics to electromagnetic waves of different frequencies, it is necessary to select a suitable scanning frequency according to the type and characteristics of the filling material to obtain the best detection effect. For example, for soil with high water content, a lower frequency electromagnetic wave should be selected to increase the penetration depth; for materials with high density, the frequency needs to be adjusted to enhance the intensity of the reflected signal. At the same time, the driving speed of the detection vehicle will also affect the selection of frequency, because the speed determines the data sampling density, and the frequency needs to be adjusted according to the speed to ensure the continuity and integrity of the data. Combined with the sensor array distribution data, an electromagnetic wave emission control strategy is formulated. The sensor array distribution data provides the position information of each sensor in the array. Based on this information, the timing and frequency of each sensor emitting electromagnetic waves can be determined. For example, each sensor can be activated in sequence in a certain order for emission, or multiple sensors can be activated at the same time for multi-channel scanning. At the same time, according to the requirements of different filling materials and driving speeds, a corresponding emission frequency is assigned to each sensor. The electromagnetic wave emission control strategy finally generated should include parameters such as the activation order, emission frequency, and emission power of each sensor to ensure that the electromagnetic wave can effectively cover the entire roadbed width and obtain high-quality detection data.

[0107] Step S16: Based on the electromagnetic wave emission control strategy, the electromagnetic integrated scanning device is used to perform multi-frequency electromagnetic wave scanning on the roadbed surface, and the electromagnetic wave reflection signal is synchronously received to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively.

[0108] In an embodiment of the present invention, an electromagnetic integrated scanning device is started to perform multi-frequency electromagnetic wave scanning on the roadbed surface. The mobile inspection vehicle moves at a constant speed along the roadbed at a set driving speed, and the electromagnetic integrated scanning device activates each sensor in the sensor array in turn according to the control strategy to emit electromagnetic waves of different frequencies. The emitted electromagnetic waves propagate in the roadbed filler, and reflection and refraction occur when encountering different medium interfaces. The receiving device synchronously receives the electromagnetic wave signals reflected back from the roadbed filler. These reflected signals contain information about the internal structure and characteristics of the filler. During the scanning process, the electromagnetic wave reflection signal intensity, phase and other parameters of each sampling point are synchronously recorded to generate the original filler electromagnetic response signal data. At the same time, a high-precision positioning system (such as GPS, inertial navigation, etc.) is used to record the spatial position of each sampling point to generate roadbed sampling position data. The sampling position data corresponds to the electromagnetic response signal data one by one. For example, the electromagnetic response signal can be drawn into a two-dimensional or three-dimensional image of the roadbed filler according to the sampling position data, and the spatial distribution of the filler characteristics can be intuitively displayed.

[0109] Preferably, step S15 comprises the following steps:

[0110] Step S151: Analyze the material magnetic permeability characteristics according to the roadbed filling material parameters to generate roadbed material characteristic data;

[0111] Step S152: Processing the transmission frequency of the transmitting device according to the driving speed data of the detection vehicle to generate electromagnetic transmission frequency data;

[0112] Step S153: matching the electromagnetic wave scanning frequency range with the electromagnetic emission frequency data to generate electromagnetic wave scanning frequency range data;

[0113] Step S154: dividing the electromagnetic wave scanning frequency range data into high and low frequency scanning frequency segments according to the roadbed material characteristic data to generate electromagnetic wave scanning frequency sequence data;

[0114] Step S155: Control the phased array beam direction of the electromagnetic integrated scanning device through the sensor array distribution data, and input the electromagnetic wave scanning frequency according to the electromagnetic wave scanning frequency sequence data, so as to obtain the electromagnetic wave emission control strategy.

[0115] In an embodiment of the present invention, the magnetic permeability characteristics of the roadbed filling material are analyzed. Different types of fillers, such as sand, clay, fly ash, etc., have different magnetic permeabilities. The material database can be consulted or laboratory tests can be performed to obtain the magnetic permeability values ​​of different fillers. The filler type and its corresponding magnetic permeability value are stored as roadbed material characteristic data. For example, if the magnetic permeability of a certain graded crushed stone is 1.2, then "graded crushed stone, 1.2" is stored in the roadbed material characteristic data. According to the driving speed of the inspection vehicle, the range of the electromagnetic wave emission frequency is determined. The faster the driving speed, the higher the sampling frequency is required to ensure sufficient spatial resolution. For example, if the speed of the inspection vehicle is 10m / s, in order to achieve a spatial resolution of 0.1m, the electromagnetic wave emission frequency needs to be at least 100Hz. The calculated suitable frequency range is stored as electromagnetic emission frequency data. The electromagnetic emission frequency data is matched with the preset electromagnetic wave scanning frequency range. The preset scanning frequency range is usually relatively wide, such as 1MHz to 100MHz. According to the electromagnetic emission frequency data, a suitable sub-range is selected, such as 10MHz to 20MHz, to improve the scanning efficiency and accuracy. According to the response characteristics of different materials to electromagnetic waves of different frequencies, the scanning frequency range is divided into a high frequency band and a low frequency band. For example, for materials with high magnetic permeability or high electrical conductivity, selecting a low frequency band can improve the penetration depth; for situations where surface details need to be detected or fine structures need to be distinguished, selecting a high frequency band can improve the resolution. The division process can be performed according to empirical values ​​or theoretical calculation results. After the division is completed, a specific scanning frequency point is determined in each frequency band to generate electromagnetic wave scanning frequency sequence data. The phased array of the electromagnetic integrated scanning device is controlled in the beam direction. The phased array technology realizes electronic scanning and directional emission of the beam by adjusting the phase of each antenna unit in the array. According to the geometric structure of the sensor array and the roadbed detection requirements, the phase delay required for each antenna unit is calculated to realize directional scanning of a specific area of ​​the roadbed. At the same time, according to the electromagnetic wave scanning frequency sequence data generated in step S154, each scanning frequency point is input into the electromagnetic integrated scanning device in sequence. In this way, the device will transmit in a predetermined beam direction at each scanning frequency point. Finally, the beam direction control parameters and scanning frequency sequence data are integrated to generate a complete electromagnetic wave emission control strategy.

[0116] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S3 in the flowchart, in this example, step S3 includes:

[0117] Step S21: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data;

[0118] In an embodiment of the present invention, the original filler electromagnetic response signal data is obtained, and these data are usually collected by an electromagnetic integrated scanning device when it moves and scans the roadbed surface. Due to the influence of factors such as the driving speed of the mobile detection vehicle and the electromagnetic wave emission frequency, these data are not aligned on the time axis and the space axis. In order to perform accurate analysis, it is necessary to perform time-space registration of the response signal on the original data. The specific operation of time-space registration includes: according to the roadbed sampling position data, each electromagnetic response signal data is associated with the corresponding spatial position; according to the electromagnetic wave emission timestamp and the driving speed of the detection vehicle, the exact time of each signal data acquisition is determined. Then, the signal data is resampled using an interpolation algorithm (such as linear interpolation, spline interpolation, etc.) to make it evenly distributed in time and space. For example, the data can be resampled at fixed time intervals or spatial intervals to generate a continuous, equally spaced signal sequence to obtain the filler electromagnetic response signal time series data.

[0119] Step S22: performing adaptive wavelet denoising on the time series data of the electromagnetic response signal of the filler, and performing response signal error correction to obtain corrected electromagnetic response signal time series data;

[0120] In an embodiment of the present invention, due to the influence of environmental noise and system errors, the filler electromagnetic response signal time series data contains noise and errors. An adaptive wavelet denoising method is used to denoise the signal. Adaptive wavelet denoising can automatically adjust the wavelet basis function and the number of decomposition layers according to the local characteristics of the signal, thereby effectively removing noise while retaining the detailed information of the signal. For example, for areas with large signal fluctuations, a smaller number of decomposition layers can be selected to retain more details; for areas with small signal fluctuations, a larger number of decomposition layers can be selected to better remove noise. In addition, according to a pre-established error model or calibration data, the signal is error corrected, such as removing systematic deviations or compensating for temperature effects. The denoised and corrected data is stored as corrected electromagnetic response signal time series data.

[0121] Step S23: performing short-time Fourier transform on the time series data of the corrected electromagnetic response signal to generate frequency domain data of the filler electromagnetic response;

[0122] In an embodiment of the present invention, a short-time Fourier transform (STFT) is performed on the time series data of the corrected electromagnetic response signal. STFT can convert a time domain signal into a frequency domain signal and retain the time information of the signal. Through STFT, the energy distribution of different frequency components in different time periods can be analyzed. For example, it can be observed that the energy of certain frequency components is enhanced or attenuated in a specific time period, which is related to the characteristic changes of the roadbed filler. The transformed data is stored as filler electromagnetic response frequency domain data, which contains the response amplitude and phase information of each position at different frequencies.

[0123] Step S24: performing signal amplitude and phase value processing on the filler electromagnetic response frequency domain data, and performing signal energy response threshold calculation to generate signal energy response threshold data;

[0124] In an embodiment of the present invention, the frequency domain data of the electromagnetic response of the filler is processed with signal amplitude and phase values, for example, the signal amplitude and phase at each frequency are calculated. Then, based on the amplitude and phase values ​​of the signal, the energy response of the signal is calculated. For example, the square of the signal amplitude can be used as the signal energy. The signal energy response threshold is calculated by statistically analyzing the signal energy distribution at different positions and frequencies. The signal energy response threshold is calculated using the amplitude spectrum data. The signal energy response threshold is used to distinguish between valid signals and noise signals. For example, the mean and standard deviation of the amplitude spectrum can be calculated, and the mean plus several times the standard deviation can be used as the threshold; or, a suitable threshold can be selected based on the background noise level of the signal so that the amplitude of the noise signal is lower than the threshold. The calculated signal energy response threshold will be recorded as signal energy response threshold data.

[0125] Step S25: Screening the effective electromagnetic signals of the filler electromagnetic response signal time series data according to the preset electromagnetic frequency determination rules and signal energy response threshold data to generate effective electromagnetic response signal data.

[0126] In an embodiment of the present invention, based on the preset electromagnetic frequency determination rules, it is determined which frequency ranges of signals need to be paid attention to. These rules are formulated based on the physical properties of the roadbed filler and the detection target. For example, if the focus is on the change in moisture content in the deep layer of the roadbed, a lower frequency range will be selected; if the focus is on the defects in the surface layer of the roadbed, a higher frequency range will be selected. Then, the signal energy response threshold data calculated in step S24 is used to filter the time series data. The specific operations include: performing a short-time Fourier transform on the time series data to obtain the spectrum at each time point; extracting the spectrum amplitude value within the frequency range of interest; comparing these amplitude values ​​with the signal energy response threshold value. If the amplitude value is greater than the threshold, the signal is considered to be a valid signal, otherwise it is considered to be a noise signal. Finally, the time points and amplitude values ​​corresponding to all valid signals are retained to generate valid electromagnetic response signal data.

[0127] Preferably, step S3 comprises the following steps:

[0128] Step S31: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data according to the roadbed sampling position data to generate multi-frequency electromagnetic response signal group data;

[0129] Step S32: performing frequency band penetration depth analysis on the multi-frequency electromagnetic response signal group data to generate frequency band penetration depth data;

[0130] Step S33: extracting the signal propagation time of the multi-frequency electromagnetic response signal group data, and calculating the electromagnetic wave propagation speed according to the frequency band penetration depth data to generate multi-frequency electromagnetic wave propagation speed data;

[0131] Step S34: extracting the physical characteristics of the roadbed filler through the multi-frequency electromagnetic wave propagation velocity data to generate the physical characteristic data of the roadbed filler;

[0132] Step S35: Perform spatial feature statistics of abnormal areas based on the physical feature data of the roadbed filler to generate spatial feature data of the abnormal areas; set up a partitioned filler construction plan based on the spatial feature data of the abnormal areas to generate intelligent partitioned filler construction plan data.

[0133] In an embodiment of the present invention, effective electromagnetic response signal data is grouped according to the roadbed sampling position data. Each group corresponds to a specific roadbed area. For example, sampling points with similar distances can be divided into a group. Each group contains electromagnetic response signal data of multiple frequencies to form multi-frequency electromagnetic response signal group data. The multi-frequency electromagnetic response signal group data generated in step S31 is subjected to frequency band penetration depth analysis. Electromagnetic waves of different frequencies have different penetration depths in roadbed fillers. High-frequency electromagnetic waves decay faster and have shallower penetration depths; low-frequency electromagnetic waves decay slower and have deeper penetration depths. The analysis method can be based on a theoretical model of electromagnetic wave propagation. For example, the penetration depth of electromagnetic waves of different frequencies can be calculated based on their attenuation coefficients. The penetration depth of electromagnetic waves of different frequencies in specific fillers can also be obtained by experimental measurement. The analysis results will be stored in the form of frequency band penetration depth data, and the penetration depth value corresponding to each frequency band will be recorded. The multi-frequency electromagnetic response signal group data is analyzed to extract the propagation time of the signal from transmission to reception. The propagation time can be obtained by analyzing the time domain waveform or frequency domain characteristics of the signal. For example, the propagation time can be determined by identifying the peak value or zero crossing point of the signal. Then, in combination with the frequency band penetration depth data obtained in step S32, the propagation speed of electromagnetic waves in different frequency bands is calculated. The propagation speed is equal to the penetration depth divided by the propagation time. The calculation results will be stored in the form of multi-frequency electromagnetic wave propagation speed data, recording the electromagnetic wave propagation speed corresponding to each frequency band. According to the multi-frequency electromagnetic wave propagation speed data, the physical characteristics of the roadbed filler are extracted. The propagation speed of electromagnetic waves in different media is related to its physical properties such as dielectric constant and magnetic permeability. For example, the faster the propagation speed, the smaller the dielectric constant is usually. By establishing a relationship model between the electromagnetic wave propagation speed and the physical properties of the filler, the physical characteristics of the filler such as dielectric constant, magnetic permeability, density, water content, etc. can be inverted and calculated based on the multi-frequency electromagnetic wave propagation speed data to generate the physical characteristic data of the roadbed filler. According to the physical characteristic data of the roadbed filler, such as density, water content, etc., it is judged whether there is an abnormal area in the roadbed. For example, if the density of a certain area is significantly lower than that of other areas, there is a problem of insufficient compaction in the area. The spatial distribution, area, shape and other characteristics of the abnormal area are statistically analyzed to generate the spatial characteristic data of the abnormal area. According to the spatial characteristics and filler characteristics of the abnormal area, a targeted filler construction plan is formulated. For example, for areas with insufficient compaction, the number of compaction passes can be increased; for areas with too high water content, drainage measures can be taken. The construction plans of different areas are integrated to generate intelligent partition filler construction plan data.

[0134] Preferably, step S34 includes the following steps:

[0135] Step S341: performing dielectric constant inversion according to the multi-frequency electromagnetic wave propagation velocity data to generate dielectric constant distribution data;

[0136] Step S342: performing electromagnetic frequency weighting on the dielectric constant distribution data, and performing weighted fusion of multi-frequency data to generate weighted fusion dielectric constant data;

[0137] Step S343: Calculate the electromagnetic wave attenuation coefficient according to the multi-frequency electromagnetic wave propagation velocity data and the weighted fusion dielectric constant data, and perform water content contribution processing to generate roadbed water content distribution data;

[0138] Step S344: using a preset filler density-dielectric constant linear model to estimate the filler density value of the weighted fused dielectric constant data, and generating filler density distribution characteristic data;

[0139] Step S345: performing filler particle distribution index processing on the roadbed moisture content distribution data through filler density distribution characteristic data to generate roadbed particle distribution index parameters;

[0140] Step S346: Integrate the filler density distribution characteristic data, the roadbed moisture content distribution data and the roadbed particle distribution index parameters into the roadbed filler physical characteristic data to generate the roadbed filler physical characteristic data.

[0141] In an embodiment of the present invention, the relationship between the electromagnetic wave propagation velocity and the dielectric constant is used to invert and calculate the multi-frequency electromagnetic wave propagation velocity data to obtain the dielectric constant at different frequencies. Since electromagnetic waves of different frequencies have different sensitivities to the dielectric constant, dielectric constant values ​​corresponding to multiple frequencies can be obtained. These data are indexed by spatial position and stored as dielectric constant distribution data. According to the sensitivity of electromagnetic waves of different frequencies to the characteristics of roadbed fillers, electromagnetic frequency weights are assigned to the dielectric constant distribution data. For example, for a specific type of filler, electromagnetic waves of certain frequencies can better reflect its dielectric properties, and these frequencies are given higher weights. Then, the dielectric constant data of different frequencies are weighted averaged to obtain a fused dielectric constant value, and weighted fused dielectric constant data are generated. Combined with the multi-frequency electromagnetic wave propagation velocity data and the weighted fused dielectric constant data, the attenuation coefficient of the electromagnetic wave in the roadbed filler is calculated. When electromagnetic waves propagate in the medium, energy attenuation occurs. The attenuation coefficient reflects the degree of attenuation and is related to the characteristics of the medium such as the conductivity and dielectric constant. After calculating the attenuation coefficient, the contribution of water content to the attenuation coefficient is further analyzed. The water content will significantly affect the conductivity and dielectric constant of the filler, and then affect the attenuation coefficient of the electromagnetic wave. By establishing a relationship model between the attenuation coefficient and the water content, such as an empirical formula or a theoretical model, the water content distribution data of the roadbed filler can be calculated based on the attenuation coefficient inversion. Using the pre-established filler density-dielectric constant linear model, the weighted fusion dielectric constant data generated in step S342 is used to estimate the filler density value. The linear model describes the linear relationship between the filler density and the dielectric constant, which can be obtained by experimental measurement or theoretical derivation. Substitute the weighted fusion dielectric constant data into the linear model to calculate the filler density value at the corresponding position. Combine the calculated density value with the spatial position information to generate filler density distribution characteristic data. Combine the filler density distribution characteristic data generated in step S344 and the roadbed moisture content distribution data generated in step S343, and perform filler particle distribution index processing. The particle distribution of the filler will affect its density and moisture content. For example, fillers with uniform particle grading usually have higher density and lower moisture content. By analyzing the spatial distribution characteristics of the filler density and moisture content, the filler particle distribution can be inferred. For example, the non-uniformity coefficient of particle distribution or other particle distribution indexes can be calculated based on the correlation between density and water content. The filler density distribution characteristic data, the roadbed water content distribution data and the roadbed particle distribution index parameters are integrated together to generate the roadbed filler physical characteristic data. This data contains a variety of physical property information of the roadbed filler.

[0142] Preferably, step S35 includes the following steps:

[0143] Step S351: Acquire historical railway roadbed filling engineering data; construct a mapping relationship between filler physical characteristics and compaction degree based on the historical railway roadbed filling engineering data to obtain filler physical characteristics-compactness mapping relationship data;

[0144] Step S352: using a preset convolutional neural network model to perform model transfer learning on filler physical characteristics-compactness mapping relationship data, and using historical railway roadbed filler engineering data to perform model training to generate a roadbed filler compaction assessment model;

[0145] Step S353: transmitting the roadbed filler physical characteristic data to the roadbed filler compaction assessment model for compaction assessment to generate filler compaction data;

[0146] Step S354: matching roadbed sampling points according to the roadbed filler physical characteristic data, and performing compaction spatial distribution difference according to the filler compaction data to generate roadbed compaction distribution data;

[0147] Step S355: Process the roadbed defect construction plan based on the roadbed compaction distribution data and the roadbed filler physical characteristic data to generate intelligent roadbed filler construction plan data.

[0148] In an embodiment of the present invention, historical railway roadbed filling engineering data are collected, and these data should include physical characteristic parameters of the filler (such as density, water content, particle grading, etc.) and corresponding compaction data. The compaction data can be obtained through field testing, such as by using methods such as light dynamic probing or heavy dynamic probing. Then, the historical data is analyzed to construct a mapping relationship between the physical characteristics of the filler and the compaction. For example, regression analysis, neural network and other methods can be used to establish a model, using the physical characteristics of the filler as input and the compaction as output. For example, ResNet, VGG, etc., as basic models. The filler physical characteristics-compactness mapping relationship data obtained in step S351 is used for model transfer learning. Transfer learning refers to applying a pre-trained model to a new related task, which can effectively improve the training efficiency and generalization ability of the model. The specific operation includes initializing the weights of the pre-trained model to the new model, and then fine-tuning the new model using historical railway roadbed filling engineering data. During the training process, the physical characteristics of the filler are used as input and the compaction is used as output, and the model parameters are continuously adjusted so that it can accurately predict the compaction of the roadbed filling. The model finally trained will be used as the roadbed fill compaction evaluation model. The model will predict the compaction of the roadbed fill based on the input physical characteristic parameters. For example, the physical characteristic data such as density and water content of each sampling point are input into the model, and the model outputs the compaction value of the corresponding sampling point. The compaction prediction values ​​of all sampling points are integrated to generate filler compaction data. According to the physical characteristic data of the roadbed filler, the filler compaction data is matched with the corresponding roadbed sampling point. Ensure that each compaction value corresponds to the correct spatial position. Since there is a spatial interval between the sampling points, it is necessary to perform a difference in the compaction spatial distribution to generate continuous roadbed compaction distribution data. The difference method can use linear interpolation, Kriging interpolation and other methods. For example, the compaction value of any position between adjacent sampling points can be calculated based on the compaction values ​​of adjacent sampling points. According to the roadbed compaction distribution data and the physical characteristic data of the roadbed filler, the defects in the roadbed are identified, such as insufficient compaction, excessive water content, unreasonable particle grading, etc. According to different defect types, corresponding construction plans are formulated. For example, for areas with insufficient compaction, the number of compaction passes can be increased or the compaction equipment can be replaced; for areas with too high water content, drainage measures can be taken or dry materials can be added. The construction plans for different areas can be integrated to generate intelligent roadbed filler construction plan data.

[0149] Preferably, step S355 includes the following steps:

[0150] Step S3551: using the preset roadbed filler early warning index library to perform cluster analysis on the measured roadbed condition on the roadbed filler physical characteristic data and roadbed compaction distribution data, and marking abnormal areas to generate preliminary filler abnormal area data;

[0151] Step S3552: Perform spatial connectivity analysis on the preliminary filler abnormal area data, and merge adjacent abnormal areas to obtain roadbed filler abnormal area data;

[0152] Step S3553: Perform regional spatial feature statistics on the abnormal area data of the roadbed filler to generate abnormal area spatial feature data;

[0153] Step S3554: Set up the roadbed filling construction plan according to the spatial characteristic data of the abnormal area, and generate intelligent roadbed filling construction plan data.

[0154] In the embodiment of the present invention, a roadbed filler early warning index library is pre-established, which contains characteristic indicators of various filler defects, such as normal density ranges, water content ranges, compaction ranges, etc. of different types of fillers. Using this index library, cluster analysis is performed on the physical characteristic data of the roadbed filler and the roadbed compaction distribution data. The measured data is compared with the early warning index, and the area that does not meet the early warning index is marked as an abnormal area. Spatial connectivity analysis is performed on the preliminary filler abnormal area data. If two adjacent areas are marked as abnormal areas, and their defect types are the same or similar, they are merged into a larger abnormal area. For example, if two adjacent areas both have insufficient compaction problems, they are merged into a larger insufficient compaction area. The statistical content includes geometric features such as the area, perimeter, shape, depth, and center coordinates of each abnormal area, as well as statistical values ​​of the physical characteristic parameters of the filler (such as density, water content, compaction, etc.) in the abnormal area, such as the average value, standard deviation, maximum value, minimum value, etc. These statistical information are integrated to generate abnormal area spatial characteristic data. According to the abnormal area spatial characteristic data and the roadbed filler early warning index library, a targeted construction plan is formulated. For example, large compaction equipment can be used to handle large areas with insufficient compaction; for areas with complex shapes and difficult to handle, manual or small equipment can be used for refined construction; for different types of defects, different treatment methods can be used, such as grouting, replacement filling, etc. The construction plans for different areas are integrated to generate intelligent roadbed filling construction plan data.

[0155] Preferably, step S3554 includes the following steps:

[0156] The abnormal area data of the roadbed filling abnormal area is located through the abnormal area spatial feature data, and the roadbed filling risk area is divided to obtain the roadbed filling high risk area data and the roadbed filling low risk area data respectively;

[0157] Perform fine grid division on the high-risk area data of the roadbed filler to generate fine grid data of the high-risk area;

[0158] The physical characteristic data of roadbed filler is used to calculate the filler physical anomaly value of the refined grid data in the high-risk area, and the filler physical anomaly data in the high-risk area is generated;

[0159] The compaction loss of high-risk areas is calculated based on the refined grid data of high-risk areas through the compaction distribution data of the roadbed and the physical anomaly data of the filler in the high-risk areas, and the compaction loss data of the high-risk areas is generated;

[0160] Set up multi-dimensional reinforcement plans based on the compaction loss data of high-risk areas and generate reinforcement plan data for high-risk areas;

[0161] Use the physical characteristic data of roadbed fill to conduct routine quality assessment of low-risk areas of roadbed fill, and generate assessment benchmark values ​​for low-risk areas;

[0162] Conduct construction process adaptability analysis based on the low-hazard area assessment benchmark, optimize standard construction parameters, and generate construction plan data for low-hazard areas;

[0163] The construction plan data of low-hazard areas and the reinforcement plan data of high-hazard areas are integrated into partitioned construction plans to obtain intelligent partitioned filling material construction plan data.

[0164] In an embodiment of the present invention, based on the spatial characteristic data of the abnormal area (such as area, shape, depth, etc.), the abnormal area of ​​the roadbed filler is located, and according to the preset threshold or rule, the abnormal area is divided into a high-risk area and a low-risk area. For example, an area with an area greater than 10 square meters and a compaction degree less than 90% can be divided into a high-risk area. The high-risk area is finely gridded, for example, the area is divided into 1m×1m grids. This can more finely analyze and process the differences within the high-risk area. Using the physical characteristic data of the roadbed filler (such as density, water content, etc.), the abnormal value of the physical characteristics of the filler in each grid is calculated. For example, the deviation of the density of each grid from the average density can be calculated. Combined with the roadbed compaction distribution data and the filler physical abnormality data, the compaction loss value of each grid is calculated. For example, the theoretical maximum compaction degree of the filler can be calculated based on its density and water content, and then compared with the measured compaction degree. According to the compaction loss value of each grid and the filler physical abnormality data, a multi-dimensional reinforcement scheme is set. For example, for areas with severe compaction loss and low density, strong tamping reinforcement can be used; for areas with high water content, drainage measures can be used. Generate reinforcement plan data for high-risk areas, including specific reinforcement measures for each grid. Use the physical characteristics data of roadbed fillers to conduct routine quality assessments on low-risk areas, such as calculating average density, average water content, average compaction, etc., to generate assessment benchmark values ​​for low-risk areas. According to the assessment benchmark values ​​for low-risk areas, analyze the adaptability of different construction processes and optimize standard construction parameters. For example, if the filler quality in the low-risk area is good, standard compaction processes and parameters can be used; if the filler quality is poor, it is necessary to adjust parameters such as the number of compaction passes and rolling speed. Generate construction plan data for low-risk areas, including specific construction processes and parameters for each area. Integrate the reinforcement plan data for high-risk areas and the construction plan data for low-risk areas to generate the final intelligent partition filler construction plan data. This data contains detailed construction plans for all areas, guiding construction personnel to perform targeted operations to ensure that the quality of roadbed fillers meets the requirements.

[0165] Preferably, the present invention further provides a high-speed railway roadbed filling parameter optimization system, which executes the high-speed railway roadbed filling parameter optimization method as described above, and the high-speed railway roadbed filling parameter optimization system comprises:

[0166] The roadbed electromagnetic scanning module is used to obtain the high-speed railway data to be detected; perform electromagnetic wave scanning frequency processing on the high-speed railway data to be detected to generate an electromagnetic wave emission control strategy; perform multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively;

[0167] The electromagnetic response signal analysis module is used to perform time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; perform electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data;

[0168] The roadbed construction processing module is used to perform multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; extract the physical characteristics of the roadbed filler according to the multi-frequency electromagnetic response signal grouping data to generate the physical characteristic data of the roadbed filler; perform spatial characteristic statistics of the abnormal area according to the physical characteristic data of the roadbed filler to generate the spatial characteristic data of the abnormal area; set the partition filler construction plan according to the spatial characteristic data of the abnormal area to generate the intelligent partition filler construction plan data;

[0169] The subgrade filling parameter optimization module is used to carry out subgrade filling construction processing according to the intelligent partition filling construction plan data, and adjust the subgrade filling parameter ratio to achieve high-speed railway subgrade filling parameter optimization operation.

[0170] The present application is to utilize the sensitivity of electromagnetic waves to the internal structure of the filler, through multi-frequency electromagnetic wave scanning and fine signal analysis, to deeply detect the compaction degree, uniformity and potential defects of the roadbed filler, and completely get rid of the subjectivity and uncertainty of traditional manual experience judgment. Through the spatiotemporal registration of electromagnetic response signals and effective signal screening, the physical characteristic differences inside the filler can be accurately identified, and then the compaction quality can be accurately judged whether it meets the standard, effectively avoiding the problem of insufficient or uneven local compaction. Through intelligent data processing and analysis, the physical characteristic data of the filler and the spatial characteristic data of the abnormal area are quickly generated, which greatly shortens the detection time and improves the construction efficiency. At the same time, the intelligent partition filler construction plan generated according to the analysis results can guide the construction personnel to carry out targeted filler parameter ratio adjustment and compaction operations, ensuring that each layer of filler can achieve the best compaction effect, avoiding blind construction and repeated rework, and significantly improving the overall construction quality of the roadbed. Through the precise control of the compaction quality of the roadbed filler, the occurrence of problems such as roadbed settlement, deformation and cracking can be fundamentally prevented. The long-term stability of the roadbed is guaranteed, reducing the frequency and workload of later maintenance and repair, thereby reducing the maintenance cost of the entire life cycle. At the same time, by avoiding safety hazards caused by roadbed quality problems, the safety and reliability of high-speed railway operations are improved, ensuring the travel safety of passengers and the smooth operation of railway transportation.

[0171] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0172] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for optimizing parameters of high-speed railway roadbed filling material, characterized in that: The following steps are involved: Step S1: Acquire high-speed railway data to be detected; Perform electromagnetic wave scanning frequency processing on the high-speed railway data to be inspected and generate an electromagnetic wave emission control strategy; Based on the electromagnetic wave emission control strategy, multi-frequency electromagnetic wave scanning is performed on the roadbed surface to obtain the electromagnetic response signal data of the original filler and the roadbed sampling position data respectively; Step S2: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; performing electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data; Step S3: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; performing roadbed filler physical feature extraction based on the multi-frequency electromagnetic response signal grouping data to generate roadbed filler physical feature data; performing abnormal area spatial feature statistics based on the roadbed filler physical feature data to generate abnormal area spatial feature data; performing partition filler construction plan setting based on the abnormal area spatial feature data to generate intelligent partition filler construction plan data; Step S4: Perform roadbed filling construction processing according to the intelligent partition filling construction plan data, and adjust the roadbed filling parameter ratio to achieve high-speed railway roadbed filling parameter optimization operation.

2. The high-speed railway roadbed filling parameter optimization method according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: acquiring the high-speed railway data to be detected; extracting the roadbed filling material and the path width from the high-speed railway data to be detected, and obtaining the railway roadbed width data and the roadbed filling material parameters respectively; Step S12: using a preset mobile inspection vehicle to carry a multi-band electromagnetic wave transmitting and receiving device to construct an electromagnetic integrated scanning device; Step S13: activating the detection sensor array of the electromagnetic integrated scanning device based on the railway roadbed width data to generate sensor array distribution data; Step S14: determining the initial position of the roadbed for the preset mobile inspection vehicle, setting the travel speed according to the high-speed railway data to be inspected, and generating the travel speed data of the inspection vehicle; Step S15: performing electromagnetic wave scanning frequency processing according to the roadbed filling material parameters and the detection vehicle driving speed data, and performing electromagnetic wave emission control processing on the electromagnetic integrated scanning device through the sensor array distribution data to generate an electromagnetic wave emission control strategy; Step S16: Based on the electromagnetic wave emission control strategy, the electromagnetic integrated scanning device is used to perform multi-frequency electromagnetic wave scanning on the roadbed surface, and the electromagnetic wave reflection signal is synchronously received to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively.

3. The high-speed railway roadbed filling parameter optimization method according to claim 2 is characterized in that: Step S15 includes the following steps: Step S151: Analyze the material magnetic permeability characteristics according to the roadbed filling material parameters to generate roadbed material characteristic data; Step S152: Processing the transmission frequency of the transmitting device according to the driving speed data of the detection vehicle to generate electromagnetic transmission frequency data; Step S153: matching the electromagnetic wave scanning frequency range with the electromagnetic emission frequency data to generate electromagnetic wave scanning frequency range data; Step S154: dividing the electromagnetic wave scanning frequency range data into high and low frequency scanning frequency segments according to the roadbed material characteristic data to generate electromagnetic wave scanning frequency sequence data; Step S155: Control the phased array beam direction of the electromagnetic integrated scanning device through the sensor array distribution data, and input the electromagnetic wave scanning frequency according to the electromagnetic wave scanning frequency sequence data, so as to obtain the electromagnetic wave emission control strategy.

4. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; Step S22: performing adaptive wavelet denoising on the time series data of the electromagnetic response signal of the filler, and performing response signal error correction to obtain corrected electromagnetic response signal time series data; Step S23: performing short-time Fourier transform on the time series data of the corrected electromagnetic response signal to generate frequency domain data of the filler electromagnetic response; Step S24: performing signal amplitude and phase value processing on the filler electromagnetic response frequency domain data, and performing signal energy response threshold calculation to generate signal energy response threshold data; Step S25: Screening the effective electromagnetic signals of the filler electromagnetic response signal time series data according to the preset electromagnetic frequency determination rules and signal energy response threshold data to generate effective electromagnetic response signal data.

5. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data according to the roadbed sampling position data to generate multi-frequency electromagnetic response signal group data; Step S32: performing frequency band penetration depth analysis on the multi-frequency electromagnetic response signal group data to generate frequency band penetration depth data; Step S33: extracting the signal propagation time of the multi-frequency electromagnetic response signal group data, and calculating the electromagnetic wave propagation speed according to the frequency band penetration depth data to generate multi-frequency electromagnetic wave propagation speed data; Step S34: extracting the physical characteristics of the roadbed filler through the multi-frequency electromagnetic wave propagation velocity data to generate the physical characteristic data of the roadbed filler; Step S35: Perform spatial feature statistics of abnormal areas based on the physical feature data of the roadbed filler to generate spatial feature data of the abnormal areas; set up a partitioned filler construction plan based on the spatial feature data of the abnormal areas to generate intelligent partitioned filler construction plan data.

6. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 5, characterized in that: Step S34 includes the following steps: Step S341: performing dielectric constant inversion according to the multi-frequency electromagnetic wave propagation velocity data to generate dielectric constant distribution data; Step S342: performing electromagnetic frequency weighting on the dielectric constant distribution data, and performing weighted fusion of multi-frequency data to generate weighted fusion dielectric constant data; Step S343: Calculate the electromagnetic wave attenuation coefficient according to the multi-frequency electromagnetic wave propagation velocity data and the weighted fusion dielectric constant data, and perform water content contribution processing to generate roadbed water content distribution data; Step S344: using a preset filler density-dielectric constant linear model to estimate the filler density value of the weighted fused dielectric constant data, and generating filler density distribution characteristic data; Step S345: performing filler particle distribution index processing on the roadbed moisture content distribution data through filler density distribution characteristic data to generate roadbed particle distribution index parameters; Step S346: Integrate the filler density distribution characteristic data, the roadbed moisture content distribution data and the roadbed particle distribution index parameters into the roadbed filler physical characteristic data to generate the roadbed filler physical characteristic data.

7. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 5, characterized in that: Step S35 includes the following steps: Step S351: Acquire historical railway roadbed filling engineering data; construct a mapping relationship between filler physical characteristics and compaction degree based on the historical railway roadbed filling engineering data to obtain filler physical characteristics-compactness mapping relationship data; Step S352: using a preset convolutional neural network model to perform model transfer learning on filler physical characteristics-compactness mapping relationship data, and using historical railway roadbed filler engineering data to perform model training to generate a roadbed filler compaction assessment model; Step S353: transmitting the roadbed filler physical characteristic data to the roadbed filler compaction assessment model for compaction assessment to generate filler compaction data; Step S354: matching roadbed sampling points according to the roadbed filler physical characteristic data, and performing compaction spatial distribution difference according to the filler compaction data to generate roadbed compaction distribution data; Step S355: Process the roadbed defect construction plan based on the roadbed compaction distribution data and the roadbed filler physical characteristic data to generate intelligent roadbed filler construction plan data.

8. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 7, characterized in that: Step S355 includes the following steps: Step S3551: using the preset roadbed filler early warning index library to perform cluster analysis on the measured roadbed condition on the roadbed filler physical characteristic data and roadbed compaction distribution data, and marking abnormal areas to generate preliminary filler abnormal area data; Step S3552: Perform spatial connectivity analysis on the preliminary filler abnormal area data, and merge adjacent abnormal areas to obtain roadbed filler abnormal area data; Step S3553: Perform regional spatial feature statistics on the abnormal area data of the roadbed filler to generate abnormal area spatial feature data; Step S3554: Set up the roadbed filling construction plan according to the spatial characteristic data of the abnormal area, and generate intelligent roadbed filling construction plan data.

9. The method for optimizing parameters of high-speed railway roadbed filling materials according to claim 8, characterized in that: Step S3554 includes the following steps: The abnormal area data of the roadbed filling abnormal area is located through the abnormal area spatial feature data, and the roadbed filling risk area is divided to obtain the roadbed filling high risk area data and the roadbed filling low risk area data respectively; Perform fine grid division on the high-risk area data of the roadbed filler to generate fine grid data of the high-risk area; The physical characteristic data of roadbed filler is used to calculate the filler physical anomaly value of the refined grid data in the high-risk area, and the filler physical anomaly data in the high-risk area is generated; The compaction loss of high-risk areas is calculated based on the refined grid data of high-risk areas through the compaction distribution data of the roadbed and the physical anomaly data of the filler in the high-risk areas, and the compaction loss data of the high-risk areas is generated; Set up multi-dimensional reinforcement plans based on the compaction loss data of high-risk areas and generate reinforcement plan data for high-risk areas; Use the physical characteristic data of roadbed fill to conduct routine quality assessment of low-hazard area data of roadbed fill to generate assessment benchmark values ​​of low-hazard area; Conduct construction process adaptability analysis based on the low-hazard area assessment benchmark, optimize standard construction parameters, and generate construction plan data for low-hazard areas; The construction plan data of low-hazard areas and the reinforcement plan data of high-hazard areas are integrated into partitioned construction plans to obtain intelligent partitioned filling material construction plan data.

10. A high-speed railway roadbed filling parameter optimization system, characterized in that: Used to execute the high-speed railway roadbed filling parameter optimization method as claimed in claim 1, the high-speed railway roadbed filling parameter optimization system comprises: The roadbed electromagnetic scanning module is used to obtain the high-speed railway data to be detected; perform electromagnetic wave scanning frequency processing on the high-speed railway data to be detected to generate an electromagnetic wave emission control strategy; perform multi-frequency electromagnetic wave scanning on the roadbed surface based on the electromagnetic wave emission control strategy to obtain the original filler electromagnetic response signal data and the roadbed sampling position data respectively; The electromagnetic response signal analysis module is used to perform time-space registration of the original filler electromagnetic response signal data to generate filler electromagnetic response signal time series data; perform electromagnetic effective signal screening according to the filler electromagnetic response signal time series data to generate effective electromagnetic response signal data; The roadbed construction processing module is used to perform multi-frequency electromagnetic response signal grouping on the effective electromagnetic response signal data through the roadbed sampling position data to generate multi-frequency electromagnetic response signal grouping data; extract the physical characteristics of the roadbed filler according to the multi-frequency electromagnetic response signal grouping data to generate the physical characteristic data of the roadbed filler; perform spatial characteristic statistics of the abnormal area according to the physical characteristic data of the roadbed filler to generate the spatial characteristic data of the abnormal area; set the partition filler construction plan according to the spatial characteristic data of the abnormal area to generate the intelligent partition filler construction plan data; The subgrade filling parameter optimization module is used to carry out subgrade filling construction processing according to the intelligent partition filling construction plan data, and adjust the subgrade filling parameter ratio to achieve high-speed railway subgrade filling parameter optimization operation.

Images