32 results about "Diffusion field" patented technology

The invention provides a semiconductor chip defect detection positioning method and system, and relates to the technical field of semiconductor chip detection, and the method comprises the steps: obtaining an original image and electrical test data; forming a defect candidate area based on local density analysis; variational mode decomposition is carried out on the electrical data, and electrical characteristic abnormal points are marked; mapping the image features to an electrical characteristic space and calculating mutual information values to determine defect points; and calculating the distribution and path of a diffusion field and determining the severity of the defect. According to the invention, accurate positioning and severity evaluation of chip defects are realized, and the detection efficiency and accuracy are improved.

The invention relates to the technical field of optical elements, and discloses a fog display point diffusion optical system and a preparation method thereof, and the preparation method comprises the steps: obtaining user visual parameters, and building a visual behavior probability density model through kernel density estimation; designing the front surface of the lens based on the model, and generating an asymmetric defocus system coupled with a gazing habit; the rear surface of the lens is designed, and the distribution density of a fog display point diffusion unit is cooperatively modulated by the front surface defocusing amount, the fixation probability and the pupil diameter; and finally integrating front and rear surface design to form an optical lens body. According to the method, the asymmetric defocus field on the front surface and the fog display point diffusion field on the rear surface are subjected to collaborative design, and the defocus intensity and the contrast modulation intensity are accurately applied to the effective retina area of the user through the visual behavior probability density model; the technical problem that a traditional out-of-focus lens is fixed in signal and cannot adapt to individual staring habits and physiological parameter changes is solved.

The invention provides a real-time monitoring method and system for the construction quality of a jet grouting pile and a storage medium, and relates to the technical field of the construction quality of the jet grouting pile. Pressure sensors, flow sensors and micro-seismic sensors are arranged at an outlet of a jet grouting pile pipeline and a drill bit, multi-dimensional time sequence data with the high signal-to-noise ratio are collected, self-adaptive filtering and drifting correction are conducted, and the real-time monitoring of the construction quality of the jet grouting pile is achieved. Generating a dynamic trajectory based on Takens phase space reconstruction, calculating a rheological entropy and constructing a four-dimensional space-time diffusion field; after the field is projected to a construction plane, an underground karst cave network model is continuously corrected by utilizing Bayesian updating, multi-stage early warning is triggered in a quantitative and graded manner according to an entropy value and a filling rate, and a filling quality thermodynamic diagram is generated, so that high-precision, dynamic and visual monitoring and guidance of a slurry diffusion process are realized.

The invention provides a nuclear radiation diffusion analysis method based on a Gaussian mixture atmospheric diffusion model. The method comprises the steps that nuclear facility source item parameters are acquired; integrating meteorological, geography, remote sensing and radiation monitoring data by using a heterogeneous data fusion engine, extracting atmospheric turbulence, terrain and building interference characteristics and source item parameters, and forming a dynamic input parameter set; constructing an unsteady state diffusion model containing a multi-modal distribution function, processing features, obtaining a mixed distribution model, and correcting the mixed distribution model; predicting spatial and temporal distribution of radiation pollutants, and generating multi-scale diffusion field simulation; and optimizing a diffusion path, and generating a three-dimensional risk field. According to the method, multi-source data are integrated, an unsteady state diffusion model is constructed, Gaussian component weights are dynamically adjusted, the model is corrected by using real-time data, and a geographic information system is coupled to carry out spatial constraint optimization. And through multi-scale feature coupling, dynamic parameter adjustment and geographic space constraint correction, the depicting capability of the atmospheric diffusion model is remarkably improved, and the nuclear radiation diffusion prediction precision is greatly improved.

The invention provides an urban traffic ubiquitous perception method based on a federal aided large model, and belongs to the technical field of data processing, and the method specifically comprises the steps: 1, building a space-time diffusion field based on multi-source heterogeneous traffic data of a target city, and dynamically dividing traffic perception subtasks according to the space-time diffusion field, and obtaining a subtask set; 2, scene training is carried out according to each traffic perception sub-task in the sub-task set based on federated learning, and a traffic perception sub-model corresponding to each traffic perception sub-task is obtained; and step 3, performing hierarchical aggregation and splicing on all the traffic perception sub-models to obtain a hierarchical perception large model, and realizing multi-task parallel perception and dynamic management of target urban traffic. Through the scheme of the invention, the generalization ability and accuracy are improved.

The invention relates to the technical field of civil engineering and computational mechanics, in particular to a load and environmental corrosion coupled reinforced concrete structure analysis method. The method comprises the following steps: carrying out structural finite element modeling; carrying out mechanical field solving; the diffusion characteristics of the crack-driven corrosive medium are updated; solving a corrosive medium diffusion field; the material performance is corrected; carrying out double-time-domain self-adaptive solution; by means of the mode, the coupling effect of complex loads and multiple environmental factors in the reinforced concrete structure can be comprehensively and accurately simulated within the engineering acceptable calculation cost, and reliable prediction of the full life cycle performance is achieved.

The invention discloses an intelligent ecological water resource management system, and the system comprises a sensing layer which is responsible for the collection and preprocessing of multi-source heterogeneous data, and is connected with satellite remote sensing, an unmanned plane LiDAR, a ground sensor, and social media public opinion data; the transmission layer is responsible for data transmission and exchange; the data layer adopts a neural radiation field, loads basin terrain / land utilization GIS data, injects real-time meteorological data to drive hydrological process simulation, constructs a dynamic digital twinborn body, embeds a human activity agent to simulate the influence of different governance policies, selects a variable order fractional derivative kernel function according to a medium fractal dimension, and obtains a variable order fractional derivative kernel function; calculating a migration path of the pollutant in the multi-layer heterogeneous aquifer, simulating an abnormal diffusion process of the pollutant in a heterogeneous medium, and generating diffusion field data; and the application layer is used for generating pollution propagation synthetic data covering long-tail distribution and water quality recovery paths under different treatment policies, and calculating profit and loss distribution of different schemes to upstream and downstream stakeholders.

The invention belongs to the technical field of intelligent grouting simulation, and particularly discloses an intelligent grouting simulation method and system based on a large model.The method comprises the steps that grouting pressure and grouting flow data monitored on a grouting site are obtained, and grouting pressure, flow and grout diffusion form data are obtained through a grouting experiment; obtaining multiphase flow field data through numerical simulation; training a teacher model by utilizing the preprocessed numerical simulation data, grouting experiment data and part of field monitoring data; in the computational domain, for a set prediction working condition, a differential knowledge distillation strategy is adopted, non-field data and field data are processed respectively, and generalization knowledge of the teacher model is migrated to the student model; and acquiring parameters under a to-be-measured working condition, and inputting the parameters into the trained student model to obtain prediction results of the slurry diffusion field, the pressure field and the velocity field. According to the method, the requirement of quickly adapting to new working conditions is met, and the efficiency requirement of real-time decision making in grouting engineering is met.

A city traffic ubiquitous perception method of a federal auxiliary large model is provided in the application, and belongs to the technical field of data processing, and specifically comprises the following steps: step 1, constructing a space-time diffusion field based on multi-source heterogeneous traffic data of a target city, and dynamically dividing traffic perception subtasks to obtain a subtask set; step 2, performing situational training according to each traffic perception subtask in the subtask set based on federated learning to obtain a traffic perception submodel corresponding to each traffic perception subtask; and step 3, layering and aggregating and splicing all traffic perception submodels to obtain a layered perception large model, and realizing multi-task parallel perception and dynamic management of the target city traffic. Through the scheme of the present application, the generalization ability and accuracy are improved.

The present application relates to a kind of low odor multilayer composite adhesive tape VOC release prediction method, comprising the following steps: S1: the original data of multilayer composite adhesive tape is obtained and preprocessed, and multimodal input data set is obtained;S2: atomic-bond diagram representation is constructed, and input graph neural network is carried out feature extraction, obtains molecular level characterization vector, molecular level feature is fused with material macroscopic attribute, and unified material characteristic representation vector X1 is obtained;S3: based on material characteristic representation vector and environmental parameter as input, establish the neural network model satisfying diffusion equation constraint, obtain VOC diffusion field data X2 for preliminary prediction;S4: the diffusion field data X2 obtained is converted into time series input, and bidirectional recurrent neural network model is constructed to carry out time series feature extraction, and the feature vector X3 containing time dependence is obtained;S5 based on integrated prediction model, output target prediction value.The present application realizes the high-precision, generalizable prediction of the VOC release behavior of multilayer composite adhesive tape.

The invention provides an ecological environment monitoring system and method, and the method comprises the steps: building a three-dimensional digital twinborn model of a designated city in a target region, and enabling the three-dimensional digital twinborn model to comprise thermal environment dynamic data of different environment nodes; determining spatial structure characteristics of different environment nodes, and constructing a spatial communication relationship through the spatial structure characteristics; determining diffusion intensity indexes of each environment node in different thermal diffusion ranges, and further determining the heat island effect intensity of each environment node by combining all the diffusion intensity indexes with the spatial communication relationship; based on the heat island effect intensity, evaluating the influence of the heat island effect intensity on a pollution diffusion field, and obtaining the heat influence pollution diffusion risk of each environment node in different heat diffusion ranges; and generating a multi-scale thermal pollution report of the target area through all the thermal influence pollution diffusion risks. By adopting the scheme of the invention, multi-scale monitoring can be carried out on urban thermal pollution based on a space-time cooperation mechanism of digital twinning.

This invention relates to the field of water quality monitoring technology, specifically to an AI-driven intelligent detection method for water pollution events. The method includes: acquiring pipeline geometric parameters, numerical water quality indicators, and visual images of the water body using intelligent sensors, and extracting multi-dimensional water quality monitoring indicators to generate a water pollution feature tensor; generating a physical constraint deviation term in the loss function using the hydrodynamic convection-diffusion equation, and performing drift compensation on the measured values ​​of the intelligent sensors based on this physical constraint deviation term to construct a water quality evolution prediction model; inputting the water pollution feature tensor into the water quality evolution prediction model to output a three-dimensional dynamic diffusion field of water pollutants; and performing source tracing using a spatiotemporal field inversion operator to calculate the coordinate probability distribution of potential pollution sources and generate early warning commands. This invention improves the robustness and source tracing confidence of complex pipeline monitoring systems through physical consistency constraints and drift compensation.

The invention discloses a swarm robot deployment method and device in a dynamic diffusion field, a server and a medium, and belongs to the technical field of robot deployment. Comprising the following steps: dividing a dynamic diffusion field into sub-regions, and arranging sensing robots in the sub-regions; constructing a parabolic pollution diffusion field model, and constructing a global field density observer; constructing a projection operator of the diffusion coefficient, and constructing an adaptive updating law of the diffusion coefficient; obtaining an estimated value of a diffusion coefficient, and obtaining an estimated value of global field density; dividing the dynamic diffusion field into a plurality of sub-regions to obtain a Voronoi cell cavity; calculating an estimated centroid; constructing a sliding mode variable and an auxiliary intermediate variable, and constructing a distributed guidance law; and driving each execution robot to move towards the target deployment position to realize deployment. By constructing a global field density observer and a speed-free distributed guidance law, deployment of swarm robots in a dynamic diffusion field under the conditions that a diffusion coefficient is unknown, the speed is unmeasurable and resources are limited is realized.

The application discloses a kind of intelligent material sound insulation quantity testing device and method, the intelligent material sound insulation quantity testing device and method include using sound source in reverberation chamber to the sound intensity probe in sound absorption chamber occurs sound source, wherein sound source in reverberation chamber forms the diffusion field that energy density is uniform everywhere, in each propagation direction makes random distribution;Using sound intensity probe receives the sound of sound source through the sound material sample to be tested;According to the sound insulation volume of reverberation chamber sound pressure level and sound intensity probe received sound pressure level, the sound material sample to be tested is calculated.This application provides a kind of intelligent material sound insulation quantity testing device and method has the automatic test of using intelligent testing device, can efficiently carry out test and test according to requirement.

The invention provides a pollution diffusion prediction method and device based on gamma radioactivity and a wind field, and relates to the technical field of pollutant monitoring, and the method comprises the steps: carrying out the feature peak extraction, nuclide recognition and concentration calculation of a gamma radioactivity energy spectrum data set, and obtaining a nuclide concentration set; constructing a three-dimensional wind field diffusion field through the topographic data and the wind field data set; performing coupling calculation and correlation matching on the nuclide concentration set and the three-dimensional wind field diffusion field through a Gaussian diffusion model, and constructing a correlation matrix of the nuclide concentration and the wind field; obtaining a nuclide half-life period, a gamma-ray emissivity and a terrain, and performing weighted fusion on elements in the incidence matrix through the nuclide half-life period and the gamma-ray emissivity to obtain a fusion coefficient; correcting the fusion coefficient through the terrain to obtain a diffusion trend coefficient set; and constructing a radioactive pollution prediction model, and according to the diffusion trend coefficient set, performing diffusion prediction in a future time period through the radioactive pollution prediction model to obtain a pollution diffusion prediction result.

The invention discloses a vehicle-mounted loudspeaker capable of adjusting a pointing angle and a control method thereof, and relates to the technical field of automotive electronics and electroacoustics, and the vehicle-mounted loudspeaker comprises an in-vehicle microphone array, a main control unit, an electric lifting and rotating loudspeaking unit, an auxiliary loudspeaking unit and a self-adaptive digital signal processing power amplifier. The main control unit collects a voice signal, identifies an instruction intention, and calculates the head center coordinate of a sound production passenger; and planning a physical path according to the instruction and the coordinates, and generating an angle control instruction. Executing acoustic parameter coupling calculation: calculating a delay amount in a high-fidelity mode; a diffusion field equalization parameter is configured in a surrounding mode; and calculating the gain and the phase delay of the auxiliary loudspeaker in the privacy mode. And the main control unit synchronously drives the loudspeaker to perform mechanical action and DSP signal processing. According to the method, deterministic mapping of the mechanical degree of freedom and the acoustic parameters is established, dynamic reconstruction of the acoustic path in the vehicle is achieved, the problem of acoustic image positioning drift is effectively solved, and auditory experience is improved.

The present invention is applicable to the field of geophysical exploration technology and provides an inversion method for HVSR spectral ratio curves based on the diffusion field theory, including the following steps: construction of a spectral ratio calculation equation based on the diffusion field theory; artificial synthesis of microtremor records with a random seismic source distribution; model setting and microtremor data acquisition; calculation of HVSR spectral ratio curves; construction and optimization of an inversion objective function; and output of inversion results. Based on the diffusion field theory, the present invention significantly reduces the limitations of traditional single-wavefield assumptions by comprehensively considering various wavefield components such as body waves and surface waves, and improves the inversion accuracy. This method shows excellent adaptability under shallow structure and complex medium conditions, and at the same time has strong noise robustness, and is applicable to complex environments such as urban underground space exploration. In addition, this method reduces the dependence on the initial model, avoids the problem of local optimum, and can be extended to fields such as engineering seismic safety assessment and underground water resource exploration, with broad application prospects.

The present invention provides a global pollutant diffusion field simulation method and system based on sparse data, which relates to the technical field of pollutant diffusion simulation, including: S1, collecting meteorological data, pollution source data and site monitoring concentration data at each natural time; S2, generating an initial particle set at natural time T based on meteorological data and pollution source data at natural time T and before natural time T; S3, combining the particle set at natural time T and particle time t with the meteorological data and pollution source data at natural time T, estimating the state of the particle set at natural time T and particle time t+1, and performing error evaluation. When the error meets the acceptable condition, the RF estimation field at natural time T is estimated based on the state of the particle set at natural time T and particle time t+1, and going to step S4; S5, predicting the predicted pollutant concentration field at natural time T+1 through a generator; S6, evaluating the predicted pollutant concentration field at natural time T+1 through a discriminator.

The application discloses a kind of dynamic diffusion field under group robot deployment method, device, server and medium, belong to robot deployment technical field.It includes: dynamic diffusion field is divided into sub-region, sensor robot is arranged in sub-region;Parabolic pollution diffusion field model is constructed, global field density observer is constructed;The projection operator of diffusion coefficient is constructed, the adaptive update law of diffusion coefficient is constructed;The estimated value of diffusion coefficient is obtained, the estimated value of global field density is obtained;Dynamic diffusion field is divided into several sub-regions, and Voronoi cell is obtained;Estimated centroid is calculated;Sliding mode variable and auxiliary intermediate variable are constructed, and distributed guidance law is constructed;Drive each execution robot to move to the target deployment position, to realize deployment.By constructing global field density observer and speedless distributed guidance law, the deployment of group robot in dynamic diffusion field is realized under the condition that diffusion coefficient is unknown, speed is not measurable and resource is limited.

The invention discloses a water body micro-plastic magnetic trapping system and method for sewage treatment, and relates to the technical field of sewage treatment.The method comprises the following steps that a density gradient field, a shear-diffusion field and a floating tendency probability field of a water body are obtained; performing interactive mapping on the density gradient field, the shear-diffusion field and the floating tendency probability field to generate an adding amount matrix; and performing short-term prediction on the density gradient field, the shear-diffusion field and the floating tendency probability field in adjacent time windows by taking the dosage matrix as a response variable. According to the method, the time window is divided, the data sequence is constructed, the hybrid predictor is combined to capture the time sequence relevance, and meanwhile, a dynamic fusion mechanism of the vulnerability coefficient and the feeding risk coefficient is introduced, so that the problem that the feeding opportunity is not matched with the micro-plastic concentration peak value due to neglecting of the dynamic change of the water body in the prior art is solved.

The application discloses a kind of escape path planning methods based on flow field global modal characteristics, comprising: the establishment of computational model is carried out to leakage scene, grid division, boundary condition setting, calculation method is determined, non-constant value calculation.The eigenvalue orthogonal decomposition is carried out to the characteristic section of chlorine diffusion field, and the mode is sorted according to the energy proportion, and the first-order mode with the maximum energy proportion is determined as the dominant mode.The escape path node grid file is established, and the dominant mode amplitude is projected onto the grid node.According to the maximum value of mode amplitude and corresponding time coefficient, the risk coefficient partition is determined, and the equivalent path length is calculated using the risk coefficient.According to the equivalent path length, the escape path is planned based on Dijkstra algorithm.The escape path obtained by the application is more universal in the whole leakage process.

The invention discloses a snow crystal morphology prediction method based on structure-dependent dynamics. The method comprises the following steps: step 1, acquiring micro-physical environment parameters of snow crystal growth; step 2, performing simulation initialization of snow crystal growth based on environmental parameters, specifically, configuring a data acquisition and edge calculation module for receiving and processing an original data stream from a sensor in real time, performing data synchronization and screening, and simulating a snow crystal growth initial process; 3, establishing a core attachment dynamics basic model; step 4, constructing a structure dependence correction module; step 5, constructing a growth iteration engine coupled with the diffusion field and dynamics; step 6, coupling in the step 4 and the step 5 automatically triggers and simulates edge sharpening instability; and step 7, form output and database generation. The method is high in prediction precision and physical fidelity, and breaks through the limitation of an empirical map.

The application discloses a water body micro-plastic magnetic trapping system and method for sewage treatment, and relates to the technical field of sewage treatment, and comprises the following steps: obtaining a density gradient field, a shear-diffusion field and a floating tendency probability field of a water body; performing interactive mapping on the density gradient field, the shear-diffusion field and the floating tendency probability field to generate a dosage matrix; and performing short-term prediction on the density gradient field, the shear-diffusion field and the floating tendency probability field in adjacent time windows with the dosage matrix as a response variable. The application solves the problem that the existing technology fails to match the dosage time and the peak value of the micro-plastic concentration due to the neglect of the dynamic changes of the water body by dividing the time window, constructing a data sequence, combining a hybrid predictor to capture the time sequence correlation, and introducing a dynamic fusion mechanism of the vulnerability coefficient and the dosage risk coefficient.

The application belongs to the technical field of intelligent grouting simulation, and specifically discloses a grouting intelligent simulation method and system based on a large model, which comprises the following steps: obtaining grouting pressure and grouting flow data monitored on site, obtaining grouting pressure, flow, and grout diffusion form data through grouting experiments, and obtaining multiphase flow field data through numerical simulation; training a teacher model by using preprocessed numerical simulation data, grouting experiment data, and part of the on-site monitoring data; in the calculation domain, for a set of predicted working conditions, adopting a differentiated knowledge distillation strategy, processing non-field data and field data respectively, and migrating the generalization knowledge of the teacher model to a student model; obtaining parameter input under a to-be-tested working condition and inputting the parameter into the trained student model to obtain the prediction results of the grout diffusion field, the pressure field, and the velocity field. The application realizes the requirement of quickly adapting to new working conditions and meets the efficiency requirement of real-time decision-making in grouting engineering.

The invention provides an oil and gas ground information physical system DoS attack intrusion detection method and system based on cross-domain diffusion field self-adaption, and belongs to the field of oil and gas ground information network security protection. The problems that an oil and gas ground information physical system is prone to being affected by DoS attacks under variable working conditions and cross-scene conditions, and an existing detection method is insufficient in generalization and robustness when domain migration is processed are solved. According to the method, the diffusion model trained on the target domain data is utilized, the source domain data is gradually translated into the target domain data through the reverse denoising process, and the pseudo target domain data is generated; a discriminator-guided gradient regularization mechanism is designed, and a back diffusion trajectory is constrained, so that the authenticity and stability of pseudo target data are improved; a semantic consistency loss function based on a support sample is constructed, semantic information of a source domain is gradually reserved, and the recognition precision of a detector on a pseudo target domain Dos attack is enhanced.

The invention discloses an intelligent material sound reduction index testing device and method.The intelligent material sound reduction index testing method comprises the steps that a sound source in a reverberation chamber is used for generating a sound source to a sound intensity probe arranged in an anechoic chamber, and a diffusion field which is uniform in energy density and randomly distributed in all propagation directions is formed in the reverberation chamber; receiving the sound of the sound source passing through the to-be-tested acoustic material sample piece by using a sound intensity probe; and according to the sound pressure level of the reverberation chamber and the sound pressure level received by the sound intensity probe, calculating the sound insulation amount of the to-be-tested acoustic material sample. According to the intelligent material sound insulation index testing device and method provided by the invention, the intelligent testing device is used for automatic testing, and test detection and testing can be efficiently carried out according to requirements.

The application provides a cross-domain multi-modal information source perception distribution method and system, and belongs to the technical field of multi-source information management; firstly, the method quantitatively calculates the information demand similarity between different user groups; secondly, the score weight of the similar group is migrated to the target group based on the similarity to generate high-quality initial weight; then, the user group is innovatively modeled as a time-varying probability distribution in a feature space, and an evolutionary dynamics equation fused with the influence of information diffusion field is used to model and predict the dynamic evolution process thereof; finally, a model predictive control framework is adopted, and based on the prediction of the future group state, the score weight is prospectively optimized and adjusted; the application realizes intelligent migration of scoring knowledge among groups, dynamic and accurate prediction of user demand, and rapid response to information environment mutation, and significantly improves the adaptability, accuracy and user experience of the information distribution system.

The invention discloses an escape path planning method based on global modal characteristics of a flow field. The escape path planning method comprises the steps of calculation model establishment, grid division, boundary condition setting, calculation method determination and unsteady numerical calculation for a leakage scene. And carrying out eigenorthogonal decomposition on the chlorine diffusion field of the characteristic cross section, sorting modes according to the energy ratio, and determining the first-order mode with the maximum energy ratio as a dominant mode. And establishing an escape path node grid file, and projecting the dominant modal amplitude to grid nodes. And determining a danger coefficient partition according to the modal amplitude and the maximum value of the corresponding time coefficient, and calculating the equivalent path length by using the danger coefficient. And planning an escape path based on a Dijkstra algorithm according to the equivalent path length. The escape path obtained by the method is more universal in the whole leakage process.

The present invention relates to the field of intelligent inspection technology, and specifically to an intelligent identification system for abnormal hidden dangers in pipeline inspections based on drones. In the present invention, by constructing a three-dimensional grid for the data collected by the drone during the inspection process, the spatial division of the inspection area is achieved, so that the detection of gas pipelines has higher accuracy. Combined with hyperspectral data analysis, the determination of methane concentration can be accurately mapped to spatial grid points, reducing the interference of environmental factors and improving the accuracy of preliminary identification of leakage areas. Based on the calculation of local concentration gradients, through data smoothing and diffusion optimization methods, the boundaries of the leakage area are made clearer, the interference of environmental noise is eliminated, and the concentration value is dynamically adjusted to enhance the adaptability to the changing trend of the leakage point. With the construction of a virtual diffusion field, combined with the spatial occupancy of terrain obstacles, the analysis of the gas flow path is made more accurate, solving the problem of misjudgment caused by the complexity of the terrain.

The present invention provides a real-time monitoring method, system and storage medium for the construction quality of jet grouting piles, and relates to the technical field of jet grouting pile construction quality. The present invention deploys pressure, flow and microseismic sensors at the jet grouting pile pipeline outlet and drill bit, collects multi-dimensional time series data with a high signal-to-noise ratio, and adaptively filters and drift corrects it. Based on Takens phase space reconstruction, dynamic trajectories are generated, rheological entropy is calculated, and a four-dimensional space-time diffusion field is constructed. After projecting the field onto the construction plane, the underground cave network model is continuously corrected using Bayesian updating, multi-level warnings are triggered according to quantitative classification of entropy values ​​and filling rates, and a filling quality heat map is generated, thereby achieving high-precision, dynamic and visual monitoring and guidance of the slurry diffusion process.