48 results about "Dry density" patented technology

The invention discloses a compaction degree measuring and calculating method based on multi-physics field joint inversion. The method comprises the following steps: acquiring multi-physics field observation data and rock-soil compaction characteristic parameters of a to-be-detected area; establishing a mapping relationship between the moisture content and the observation data and between the dry density and the observation data by using a pre-established forward model; executing joint inversion iteration, and updating the state variable based on the observation residual error; in the iteration process, a physical feasible region is constructed by utilizing rock-soil compaction characteristics, and the physical feasible region is embedded into the inversion process as a constraint condition to constrain the value ranges of the dry density and the water content; and in response to the condition that the joint inversion iteration process meets a preset convergence condition, outputting the dry density after inversion convergence and calculating the compaction degree. According to the method, the problems that a traditional inversion solution violates the soil engineering compaction rule and cannot adapt to compaction under variable work conditions are solved, and the physical authenticity and precision of compaction quality detection are improved.

The invention discloses a vibration-free foundation construction method for collapsible loess. The construction method provided by the invention comprises the following steps: dividing a construction area into grids, taking the intersection points of the grids as compaction points, detecting the initial dry density of the compaction points, drawing a density distribution thermodynamic diagram according to the initial dry density, and setting the initial pressure of compaction treatment on each compaction point according to the density distribution thermodynamic diagram; according to the initial pressure, compaction treatment is carried out at the compaction point, and gradient pressurization is carried out in the compaction treatment process until the preset depth is reached; regulating the water content to 18-22% according to the water content of the loess in the construction area; and the compaction point and the soil body compactness around the compaction point and the vibration acceleration of buildings around the construction area are monitored in real time, and when the soil body compactness change rate is smaller than 1% and the vibration acceleration is smaller than 0.1 g in at least three continuous detection periods, foundation treatment reaches the standard.

The invention relates to an autoclaved aerated concrete block based on an all-solid waste material and a preparation method thereof, and relates to the technical field of building material processing. Comprising the following components in parts by weight: 25-30 parts of an all-solid waste cementing material, 60-70 parts of construction waste micro powder, 0.05-0.1 part of aluminum powder, 0.02-0.06 part of a foam stabilizer and 60-70 parts of water, the all-solid waste cementing material is prepared from slag, fly ash, steel slag, desulfurized gypsum and red mud according to a mass ratio of (40-60): (20-35): (10-20): (5-15): (5-10); the construction waste micro powder is prepared from recycled concrete micro powder and brick powder according to a mass ratio of (60-70): (30-40). The dry density and compressive strength of the product prepared by the method can stably reach B07 and A5.0 grade standards, and the cost is lower.

The utility model relates to engineering detection technical field, discloses a kind of quick detection device of dry density specifically, to solve the problems such as traditional detection method operation cumbersome, low efficiency, existing radiation risk. Device core includes standard container, weighing sensor, signal conditioning circuit, single-chip microcontroller, hardware proportioning calculator, stable judging module, display screen and ventilation / heating module. Through RFID, container volume V is automatically read, weighing sensor measures soil sample mass m, proportioning calculator calculates dry density ρ=m / V in real time, and the result is output through display screen;Ventilation / heating module accelerates moisture evaporation, shortens detection time. The device has the advantages of real-time display, simple operation, avoids artificial error, no radiation risk, and is suitable for fine-grained soil, coarse-grained soil and frozen soil detection dry density requirements, etc., and is suitable for water conservancy engineering, such as earth dam filling and embankment engineering, building engineering and other fields of soil compaction quality rapid detection.

The invention discloses an internal tamping carrier pile backfill soil quality monitoring system and method, and belongs to the technical field of soil mass quality monitoring, the system comprises a monitoring assembly which comprises a plurality of sensor assemblies and is used for collecting pressure data and actually measured moisture content of backfill soil at different depths of an internal tamping carrier pile in the tamping process; the data transmission module is used for amplifying the data acquired by the sensor assembly and transmitting the data to the data processing unit; the data processing unit is used for converting the pressure data into equivalent additional stress and converting the equivalent additional stress into a void ratio; calculating the dry density of the backfill based on the void ratio and the actually measured water content; and finally, calculating the compaction coefficient of the backfill soil by adopting the dry density and the maximum dry density of the tamping foundation experiment. And the evaluation and decision-making unit is used for receiving the compaction coefficient calculated by the data processing unit, determining whether the compacted backfill soil on the layer where the sensor assembly is located is qualified or not according to the compaction coefficient, and giving a construction suggestion when the backfill soil is unqualified.

The invention discloses a calibration-free soil dry density in-situ measurement method, and belongs to the technical field of soil physical property detection in geotechnical engineering and agricultural engineering. The method comprises the following steps: firstly, acquiring a volume dielectric constant epsilon and an environment temperature of soil in situ, and acquiring a mass water content omega of the soil in situ; secondly, determining physical parameters of the soil, wherein the physical parameters comprise a soil particle dielectric constant, a soil particle relative density and a water dielectric constant determined according to the temperature T; and finally, substituting the volume dielectric constant, the mass water content omega and the physical parameters of the soil into a calculation model constructed on the basis of a soil solid-liquid-gas three-phase mixed medium theory, and solving to obtain the dry density of the soil. According to the method, a calculation model is constructed based on physical characteristics of a soil solid-liquid-gas three-phase mixed medium, parameters which can be directly measured are obtained in situ, and the dry density is directly solved after the parameters are substituted into the model, so that tedious steps of field calibration in a traditional method are avoided, and in-situ, rapid and calibration-free measurement of the dry density of the soil is realized.

The invention relates to the technical field of civil engineering detection, and discloses a dry density-based fine-grained soil compression modulus rapid determination method, which comprises the following steps: carrying out sample preparation and standardization on target fine-grained soil according to fixed dry soil mass and a unified initial height value; executing a limit compression test on the sample, and collecting a load-displacement data sequence until the curve slope is sharply increased and is close to be vertical; processing the load-displacement data sequence, and constructing a dry density-compression modulus reference relation model of the target fine-grained soil; measuring the on-site dry density of the on-site to-be-measured compacted soil body which is the same as the target fine-grained soil; and substituting the field dry density into the reference relation model, and calculating and determining the field compression modulus. According to the method, the dry density-compression modulus reference model of the soil is provided through a one-way extreme compaction experiment and digital model construction, and rapid and accurate measurement of the compression modulus is realized.

The invention provides an MICP-based sandy soil reinforcement stability evaluation method, device, equipment and medium, and relates to the technical field of soil reinforcement, and the method comprises the following steps: respectively adding an ammonium standard solution into sodium modified zeolite and alkalization modified fly ash to determine the ammonia nitrogen removal rate, and screening candidate mass ratios; on the basis of the candidate mass ratio, the sodium modified zeolite and the alkalized modified fly ash are doped into the standard sand columns correspondingly, the bacterial liquid and the cementing liquid are injected for an MICP reaction, and a solidified sand column is obtained; the dry density and the calcium carbonate generation amount of the multiple solidified sand columns are measured, an unconfined compressive strength experiment is conducted, and an unconfined compressive strength value is obtained; the distribution form of calcium carbonate crystals is observed through a scanning electron microscope for the multiple solidified sand columns, and gelling substances are analyzed through X-ray diffraction; and judging an unconfined compressive strength value based on a microscopic analysis result, and when the unconfined compressive strength value meets a sandy soil solidification threshold value, outputting a solidification grade and a sandy soil strength value to finish stability evaluation. The problem that sandy soil reinforcement is not stable is solved.

This invention relates to the field of machine learning technology, specifically to a method for calculating soil water distribution based on function fitting and mathematical integration. The method includes the following steps: discrete data is generated by collecting voltage and three-dimensional coordinates; a soil state function is fitted via regression; the radial dry density is analyzed and the dominant direction is selected; a directional integral sequence is generated based on normal distance attenuation; the density difference between adjacent coordinates is calculated and connected point sets are extracted to generate structural weights; the soil state, water demand, and weight integration are coupled to obtain the water distribution calculation result; and optimized scheduling sequences are generated by smoothing out boundary regions. In this invention, discrete states are formed by splicing sensor electrical signals and three-dimensional coordinates; the soil state function is fitted via spatial regression; directional attenuation and density connectivity extraction characterize the extension trend and structural weights; and Gaussian integration couples soil state and water demand information. The water distribution result is more suitable for non-uniform distributions, smooths out boundary regions, and improves the targeting of spraying scheduling.

The invention discloses a dam rockfill material compaction quality prediction method based on THO-XGBoost, and the method comprises the steps: obtaining input parameters and output parameters of a dam rockfill material compaction process, the input parameters at least comprise a material source characteristic parameter and a rolling characteristic parameter, and the output parameter is rolling dry density; converting the material source characteristic parameters, the rolling characteristic parameters and the rolling dry density into a unified space-time reference coordinate system by utilizing multi-source heterogeneous information integration and space-time association, realizing space-time registration of input and output data, and generating a compaction quality data set; and constructing an XGBoost integrated learning model based on the compaction quality data set, performing global optimization on hyper-parameters of the XGBoost model by adopting a Heman optimization algorithm introducing Tent chaotic mapping, establishing a compaction quality prediction model based on THO-XGBoost, and realizing high-precision prediction of the compaction quality. According to the method, the precision and stability of rockfill material compaction quality prediction are effectively improved, the dependence on detection of a large number of test pits is reduced, and a basis is provided for construction quality control and process optimization.

The invention discloses a river basin sediment production estimation method based on a dynamic evolution process of a river-blocking barrier lake, and the method comprises the steps: carrying out the comparative analysis of key physicochemical characteristic parameters through an identification event layer in a residual barrier lake deposition sequence, and constructing a disaster chain extreme surface process river basin sediment production estimation method. Water level changes, hydrodynamic conditions and material source changes of the barrier lake are comprehensively analyzed, different sand production stages from the stable period, the decline period and the extinction period of the barrier lake are divided, the sediment change history from lacustrine facies suspended load to flood plain bed load is analyzed, and a complete barrier lake formation-evolution full life cycle evolution process mode is constructed. A layered triangular irregular network model is constructed to obtain layered accumulated sediment volume data, the layered volume and mass are calculated based on a dry density field of sedimentary stratification and spatial interpolation so as to calculate the total sediment yield, a physical information neural network model inverts a sedimentary thickness field, sediment yield estimation is achieved, and the non-stationary dynamic evolution characteristics of water and sediment of the barrier lake are quantified.

The application discloses a compaction degree measuring and calculating method based on multi-physical field joint inversion, comprising the following steps: obtaining multi-physical field observation data and geotechnical compaction characteristic parameters of a region to be detected; using a pre-constructed forward model to establish a mapping relationship between water content, dry density and observation data; performing joint inversion iteration, updating state variables based on observation residuals; in the iteration process, using geotechnological compaction characteristics to construct a physically feasible region, and embedding it as a constraint condition into the inversion process to constrain the value range of dry density and water content; in response to the joint inversion iteration process meeting a preset convergence condition, outputting the dry density after inversion convergence and calculating the compaction degree. The application solves the problems that the traditional inversion solution violates the soil compaction rule and cannot adapt to variable power compaction, and improves the physical authenticity and accuracy of compaction quality detection.

This invention relates to the field of foundation construction technology, specifically to a construction method for treating thick, collapsible loess foundations, comprising the following steps: Step 1, construction preparation, including determining the natural moisture content, maximum dry density, and optimum moisture content of the foundation soil and backfill soil, and simultaneously watering and mixing the backfill soil for material preparation; Step 2, drilling humidification holes, injecting filter material into each humidification hole after drilling, and then tamping and backfilling; Step 4, after tamping and backfilling, leveling the construction site, watering and moistening it, and after the piles are formed in the pile holes, leveling and fully tamping; Step 5, inspection. The treatment method provided by this invention has a better effect in eliminating collapsibility, resulting in higher foundation bearing capacity, more uniform overall stiffness, smaller settlement deformation, and an effectively shortened construction period, forming a high-bearing-capacity composite foundation.

The invention discloses low-thermal-conductivity lightweight concrete and a preparation method thereof. The low-thermal-conductivity lightweight concrete is prepared from the following components in parts by weight and does not contain auxiliaries: 95-125 parts of water; 100 to 534 parts of ceramsite; 0-514 parts of fine aggregate; 315 to 500 parts of cement; 35 to 42 parts of silica fume; 3.15 to 4.2 parts of a water reducing agent; the EPS particles are added according to the volume mixing amount of 0.08 m < 3 > / m < 3 >; the fine aggregate is perlite. According to the low-heat-conductivity lightweight concrete formula provided by the invention, collaborative optimization of three properties of light weight, high strength and low heat conductivity is realized, the problem of strength loss when aggregate is replaced by perlite can be solved, and the 28d strength is not greatly reduced while the dry density is reduced by 10-15% by optimizing a replacement object (preferentially replacing sand instead of coarse aggregate) of the perlite and a doping proportion.

The present application relates to geotechnical test technical field, especially to a kind of filling type fissure rock mass geological strength index quantification method under the influence of different buried depth freeze-thaw environment, including obtaining different buried depth temperature data, the fissure filling material taken from mine is prepared to be consistent with the actual dry density, moisture content of sample in situ, carry out freeze-thaw cycle test, calculate the cohesion and internal friction angle of fissure filling material;Filling type fissure rock mass numerical model is constructed and numerical simulation under triaxial compression condition is carried out, the volume of characterization unit is determined, and the degradation process of the mechanical parameters of filling type fissure rock mass under the influence of different buried depth freeze-thaw environment is quantified;Freeze-thaw correction coefficient of fissure filling material is introduced in structural plane grade parameter SCR, and the correction of filling type fissure rock mass geological strength index under the influence of different buried depth freeze-thaw environment is realized;The present application significantly improves the accuracy and application range of strength prediction, and overcomes the limitations of traditional method ignoring freeze-thaw damage.

The application discloses a kind of detection methods of subgrade compactness, comprising: extracting soil sample from the construction site to be measured, it is divided into multiple indoor soil sample samples according to moisture content size;Detect the moisture content and dry density of each indoor soil sample sample, draw relationship curve I;According to the method that steel ball free fall impacts the top surface of soil sample sample, obtain the depth value of each indoor soil sample sample hemispherical pit, establish fitting relationship;Randomly select a point in the subgrade to be detected area as the test point of outdoor subgrade soil moisture content and dry density, obtain the moisture content and dry density of measured subgrade soil sample;According to the steel ball impact method, obtain the depth value of each detection point hemispherical pit in the subgrade to be detected area;Finally, the subgrade compactness of each detection point is fitted to obtain.The application can quickly and directly obtain the subgrade compactness, the detection period is short, the detection efficiency is high, the detection result has high reliability and accuracy, can meet the large-scale large-area continuous filling construction requirement of subgrade.

The invention relates to the technical field of reservoir structure safety control in water conservancy projects, in particular to an anti-sliding stability control method under the working condition of two-way sudden water level drop of a plain reservoir partition dam, which comprises the following steps: S1, working condition initialization and boundary condition definition; s2, carrying out seepage-stress coupling analysis; s3, solving an anti-sliding stability coefficient; s4, performing multi-objective optimization on the structural parameters; s5, verifying the efficiency of the waterproof and drainage system; s6, construction parameter simulation: fitting a soil material dry density-water content relationship through compaction test data, and determining an optimal construction water content and compaction degree parameter; and S7, performing dynamic early warning in an operation period. The anti-slip stability coefficient is accurately solved by constructing the seepage-stress bidirectional coupling analysis model and combining the strip division method, the internal mechanical response and potential slip crack surface characteristics of the dam body under the working condition of bidirectional sudden water level drop can be accurately captured, the anti-slip stability evaluation error is reduced, and a reliable quantitative basis is provided for safety management and control of the partition dam.

This invention discloses a performance testing device for autoclaved aerated concrete (AAC) blocks, relating to the field of filtration equipment. Addressing the shortcomings of prior art, such as the inability to comprehensively test the dry density and moisture content of AAC blocks, and the low testing efficiency due to the limitation of a single electrically heated drying oven processing only one AAC block sample, the present invention proposes the following solution: A base is included, with an insulated drying chamber mounted on top of the base. The insulated drying chamber comprises a support frame bolted to the outer wall of the base top and a box body internally coated with a ceramic fiber insulation layer. Three heating mechanisms, each including several fans, are located within the support frame. This invention enables simultaneous drying of two sets of AAC blocks. After drying, one AAC block can be immediately subjected to a permeability test without waiting for the previous block to dry again, thus improving testing efficiency.

The present application belongs to the technical field of hydraulic engineering design, and particularly relates to a method for evaluating dam building conditions of earth-rock dams. The present application adopts evaluation indexes of dam shell material deformation stability u1, dam shell material dry density u2, dam shell material reserve abundance u3, anti-filter material particle size u4, dam foundation covering layer natural density u5, dam foundation covering layer depth u6, and dam foundation covering layer compression modulus u7, and uses Vikor method to establish a model to obtain evaluation results of dam building conditions of the proposed earth-rock dam. The present application can scientifically and accurately evaluate common earth-rock dam building conditions, solve the problem of lack of quantitative evaluation and demonstration of existing earth-rock dam selection, and provide scientific and effective quantitative basis for earth-rock dam type demonstration and selection.

The application discloses a kind of earth-rock filling body compaction degree nondestructive testing method, collect the multi-source original feature data of earth-rock filling body, obtain dynamic deformation modulus, and establish dry density benchmark prediction model based on calibration test;Collect the original feature vector including dynamic deformation modulus, material characteristic parameter, construction state parameter and equipment impact parameter, and corresponding measured dry density;Establish benchmark relationship function, calculate target correction term;The original feature is standardized, and the standardized feature vector is obtained;Depth neural network model is constructed, second stage random forest model training, based on the parameter set and standardization feature, respectively calculate physical correction term and compensation term, sum to obtain comprehensive correction term;Final compaction degree prediction value.The application effectively solves the problems of backwardness of conventional earth-rock filling body density value detection means, damage to filling surface and low efficiency, and the obtained earth-rock filling body density value test meets the requirements of detection accuracy and construction progress.

The invention discloses a method for accurately determining the rolling dry density of a gravel material through numerical simulation, and provides a method for accurately determining the rolling dry density by performing model simulation by utilizing truncation grading and then compensating a simulation result, so that not only is the operation efficiency of model simulation ensured, but also an accurate simulation result is obtained, and the calculation efficiency is improved. The calculation efficiency and accuracy of determining the dry rolling density of the gravel material through numerical simulation are effectively improved; after the simulation result of the truncation grading is obtained, the pores are compensated through theoretical calculation, the maximum dry density value after compensation is calculated, then the maximum dry density value after compensation is compared with the maximum dry density value of a field rolling test, and the accuracy of the compensation method is verified. According to the method, the gravel material rolling dry density is accurately determined by utilizing numerical simulation, and the calculation precision and the calculation efficiency are both considered.

The invention relates to a method for preparing a roadbed material from silted soil, which comprises the following steps: uniformly mixing silted soil and magnesium oxychloride cement with or without water to obtain a mixture; curing the mixture to obtain consolidated sludge, adding microorganism induced calcium carbonate precipitation reaction liquid, and curing for 7-14 days to obtain a treated body; and adding lime into the treated body, and maintaining for 2-5 days. According to the method disclosed by the invention, the compressive strength, the dry density and the impermeability of the modified sludge can be remarkably improved, and the unconfined compressive strength after 10-14 days of maintenance can reach 200-350kPa; the process period is short, the construction is convenient, and the manpower and material cost can be greatly reduced; and the method is green, environment-friendly, high in adaptability and suitable for rapid treatment and resource utilization of soft soil foundations such as urban sludge disposal sites, river channel desilting and mud flat reclamation.

This invention discloses a method and apparatus for preparing precast piles from engineering waste soil. The waste soil is sieved according to particle size. Based on the particle size composition of the waste soil, particle flow numerical simulation is used to simulate the compaction characteristics of waste soil with different particle sizes under ideal conditions, obtaining the maximum compaction density and gradation. Based on the maximum compaction density gradation under ideal conditions and the characteristic particle size of the cementitious material, different amounts of cementitious material are added to prepare the waste soil mixture. According to the evolution model of compaction density with pressure, the close packing gradation of the waste soil and cementitious material mixture is obtained. Experiments are conducted on the variation of compaction dry density with moisture content, clay mineral content, and cementitious material water demand under different pressure conditions. Based on a multi-parameter artificial neural network machine learning model, a model of dry density with moisture content, clay mineral content, and cementitious material water demand is constructed to obtain the optimal moisture content and maximum dry density under different pressure conditions. A variable frequency micro-amplitude vibration bidirectional compaction device is used for compaction and molding using a four-stage process: static pressure-strong vibration-weak vibration-static pressure.

A kind of soil detection distinguishing device, first detection instrument, second detection instrument and third detection instrument arranged along the conveying direction of first conveying belt can respectively obtain the moisture content in the soil to be measured, cross-sectional area and mass parameter, the dry density in the soil to be measured is obtained by calculation, and then compared with the moisture content threshold value, cross-sectional area threshold value and dry density threshold value in control system, according to the comparison result, control system controls the screening body to be communicated with first discharge port or second discharge port at inlet end, so as to distribute the soil to be measured to clay or sandy soil collection area, realize the automatic detection and distinction of soil, eliminate manual sorting, improve work efficiency.

The invention discloses an intelligent rolling monitoring method and system for a high-fill roadbed, and belongs to the field of roadbed monitoring, and the method comprises the steps: dividing a working plane into a plurality of monitoring subareas, building a plurality of compression wave paths in a manner of burying a sensor network in a filling layer and carrying out the excitation vibration at a planned point location through rolling equipment, and carrying out the monitoring of the monitoring subareas; the method comprises the following steps: calculating a path wave velocity corresponding to a compressed wave path, and screening out an effective path subset capable of representing a soil state in each monitoring partition to obtain a wave velocity representative value; and for each monitoring subarea, inputting a plurality of characteristic parameters containing a wave velocity representative value into the dry density evaluation model to directly obtain a dry density predicted value of the monitoring subarea, comparing the dry density predicted value with a dry density design value, and accurately identifying an under-compaction area. And the limitation of a traditional surface monitoring method is overcome.

The application discloses a loess collapsibility evaluation method and system based on geophysical prospecting test technology, relates to the field of land exploration, and comprises the following steps: arranging a regular detection grid in a loess site to be evaluated, and constructing a site geographic coordinate system; collecting Rayleigh wave signals in real time, inverting a shear wave velocity profile, and collecting apparent resistivity data; drilling shallow soil samples, measuring soil parameters, constructing a model according to the mapping relationship between site geophysical parameters and soil parameters; inputting the collected data into a wave velocity-moisture content correlation model and a resistivity-dry density correlation model, generating a moisture content distribution field and a dry density distribution field; calculating the collapsibility of loess in each detection grid based on a geophysical response equation of the collapsibility, outputting a collapsibility contour map, and determining the collapsibility grade of each grid and the distribution range of collapsible loess. The application has the advantages that: through accurate non-invasive data collection and model analysis, a collapsibility distribution field and a contour map are generated, and a scientific basis is provided for engineering construction.

The invention discloses a preparation method of a remolded sample for reducing in-situ mechanical properties of sandy soil based on a static sounding test. The method comprises the following steps: carrying out the static sounding test to obtain conical tip resistance; calculating the in-situ relative density of the sand through an empirical formula; after drilling sampling, obtaining the maximum dry density, the minimum dry density and the natural water content through an indoor soil test; inversely calculating the in-situ dry density through the in-situ relative density, the maximum dry density and the minimum dry density; calculating the in-situ wet density through the in-situ dry density and the natural water content; and preparing the remolded sample by taking the wet density as a target control index. Compared with the prior art, the static sounding test data and the remolding sample preparation process system are combined for the first time, the target density directly derives from in-situ testing, the in-situ state is scientifically restored, the representativeness is high, operation is easy and convenient, cost is low, and the principle is clear; the process standardization is suitable for various engineering investigation and scientific research projects, is easy to popularize and apply in engineering practice, and is especially suitable for sandy soil environments such as oceans, riverbeds and loose settled layers in which undisturbed samples are difficult to obtain.

The invention belongs to the technical field of road paving, and particularly discloses a construction method for large-thickness paving of a cement-stabilized gravel base layer, which comprises the following steps: screening a gravel material, controlling the ratio of cement to the gravel material during mixing, controlling the water content to be 4.77%, adding water according to the environment temperature, controlling the maximum dry density to be 2.365 g / m < 3 >, and constructing an earth shoulder during paving. Controlling the thickness and the speed of virtual paving by paving, rolling by using road rollers with different tonnages, finishing the surface until no wheel mark exists after initial pressing, re-pressing and final pressing, then performing depth detection, and finally performing film covering and watering maintenance; by strictly controlling the content of raw materials and cement, the water content, the maximum dry density, the paving thickness and grinding, shrinkage cracks caused by sudden change of water gradient are reduced, so that the problems that the water evaporation speed of the raw materials of the cement-stabilized base layer in an arid climate region is too high in the subsequent construction stage, the shrinkage cracks appear on the cement-stabilized base layer, and the construction quality is influenced are solved. Therefore, the strength is reduced.

The invention discloses an anti-erosion improvement method for red-layer purple soil, and belongs to the technical field of soil improvement. The improvement method comprises the following steps: crushing and screening sandstone collapse deposits in the hilly area in Sichuan to obtain aggregates with the particle size of 1-2mm, pretreating, compounding the aggregates with red-layer purple soil in a layered manner according to the mixing amount of 30% by mass, spreading the aggregates by adopting a grid method, turning over and stirring in multiple directions, compacting, controlling the dry density, and standing to be stable. According to the method, the effects that the soil cohesion is increased by more than or equal to 319.7%, the internal friction angle is increased by more than or equal to 23.8%, the stable permeability coefficient is increased by more than or equal to 35.7%, the total sand production of the slope surface is less than or equal to 4082 g, and the average erosion depth is less than or equal to 5.04 cm are achieved, the problem that the red-layer purple soil is easy to wash is effectively solved, local idle deposits are recycled, the method has the advantages of low cost and environmental friendliness, and the method is suitable for popularization and application. The method is suitable for improvement of slope cropland with the gradient of 11-12 degrees in hilly areas in Sichuan.

A method for intelligent prediction of seepage characteristic parameters of soil and rock materials is disclosed. The method involves: initially selecting and processing a dataset of indoor seepage test results for soil and rock materials; constructing an interpretable machine learning model to quantify the influence of different feature parameters on the seepage characteristic parameters of soil and rock materials, thereby achieving dimensionality reduction and optimization of the dataset; establishing an intelligent combined prediction model that integrates bidirectional long short-term memory neural networks, sparrow search algorithm, and attention mechanism optimization; training the prediction model based on the optimized dataset; and inputting relevant feature parameters to achieve intelligent prediction of seepage characteristic parameters of soil and rock materials. This invention can assist engineering technicians in rapidly and intelligently predicting the seepage characteristic parameters (permeability coefficient and critical permeability gradient) of soil and rock materials based on their gradation characteristics, porosity, and dry density, thereby reducing testing costs.