124 results about "Simulation error" patented technology

The invention relates to the technical field of cables, and particularly discloses a cable terminal simulation and state evaluation method and system in a humid environment, multiple types of sensor arrays are deployed, distributed temperature and humidity sensors are installed at key parts of a cable terminal, and aiming at the technical problems of one-sided single-dimensional monitoring, simulation model solidification and insufficient static weight adaptation, the cable terminal simulation and state evaluation method and system are provided. Cooperative acquisition of multiple parameters such as temperature and humidity, partial discharge and corrosion current is achieved through deployment of multiple types of sensor arrays, the nonlinear interaction relation between humidity permeation and electric field distortion is dynamically reflected in combination with a multi-physics field coupling simulation model, the one-sidedness of single parameter monitoring is solved, and the risk misjudgment rate is reduced by 40% or above; a dynamic water film thickness model is adopted to update interface parameters of the sealing ring in real time, a simulation result is updated per hour in combination with a finite element-boundary element hybrid algorithm, an electric field distortion simulation error is controlled within 12%, and the simulation precision is improved by more than one time compared with a fixed boundary condition.

The invention relates to the technical field of wind field simulation, and provides a wind field numerical simulation method based on multiple meteorological data sources. The objective of the invention is to solve the problems of large simulation error, rough terrain boundary processing and poor turbulence model parameter adaptability caused by non-uniform coverage of a single data source, insufficient precision and unscientific multi-source fusion. The method is characterized by comprising the following steps: step 1, constructing a CFD three-dimensional grid based on a geometric model of a target area; 2, collecting original data of a ground station, satellite remote sensing, numerical forecasting and the like; 3, performing standardized preprocessing (abnormal value elimination, missing interpolation, radiation / geometric correction and resampling), determining a multi-source fusion weight by combining historical data analysis, and generating comprehensive meteorological data by adopting a weighted average method; and 4, inputting the comprehensive data into a CFD-RANS model, dynamically adjusting turbulence parameters, accurately setting terrain boundary conditions, and obtaining a wind field space-time distribution result through numerical solution. According to the invention, through combination of multi-source data fusion and CFD simulation, the wind field simulation precision and stability are improved.

The invention relates to the technical field of data processing, and provides a resistance line-associated underwater bench blasting numerical simulation error correction method, which is characterized in that land and underwater measured data are matched through the same resistance line and similar material parameters, and the resistance line and the water depth are used as basic input of an error prediction model; a reference basis is provided for error correction of subsequent numerical simulation; calculating a target result in combination with a prediction error after initial simulation, comparing a simulation result with the target result during iterative simulation to generate an iterative error, and analyzing and determining sensitivity coefficients of different error classifications in a resistance line and the iterative error so as to adjust material parameters to reduce an invalid iterative process; simulation results are processed according to stress areas in a zoning mode through an equidistant slicing method, adjustment precision is verified in combination with the root rate and the damage area, and the reliability of the simulation results is improved; and meanwhile, parameter data passing verification are input into a back propagation neural network, a correction parameter mapping library is constructed, and the design efficiency of an underwater blasting scheme is improved.

The invention belongs to the technical field of hydrology and water resources, and discloses a cold and dry mountainous area drainage basin runoff simulation method based on an MIKE SHE / MIKE 11 coupling model in combination with GIS / RS data. According to the multi-source data fusion technology, a cold and dry mountainous area distributed input data set is constructed through TRMM satellite rainfall data, an MODIS LAI product and an ASTER DEM, and the rainfall space representation precision is improved by 40%. According to the model coupling mechanism, an MIKE SHE / MIKE 11 collaborative simulation framework is provided, earth surface-groundwater flow data interaction is achieved through river channel linkage, and the complex river network flood routing simulation error is reduced to 10% or below. According to the parameter partition calibration method, a drainage basin is divided into five hydrogeological regions, sensitive parameters such as a Manning coefficient and a hydraulic conduction coefficient are optimized in combination with a genetic algorithm, and the basic flow simulation precision is improved by 25%.

The invention provides an aero-engine complete machine full-three-dimensional performance simulation error transmission and control method, and relates to the technical field of aero-engine performance simulation, and the method comprises the following steps: decomposing an aero-engine into a plurality of core parts along an airflow direction, and building a pneumatic and thermal coupling simulation model between the parts; identifying and quantifying a simulation error source of each part, and defining a simulation error value of each part; constructing a whole machine error transfer network, wherein the whole machine error transfer network comprises a sensitivity weight and a coupling transfer coefficient; based on the sensitivity weight and the coupling transfer coefficient, calculating a whole machine accumulated error; in the whole machine full-three-dimensional simulation process, outlet airflow parameter simulation data of all the components are collected, the simulation data are compared with test data, and iteration correction is conducted on simulation models of all the components and whole machine accumulated errors according to comparison results. According to the method, the whole machine simulation precision is improved, the whole machine simulation error is finally reduced to be within 2.5%, and the engineering reliability of a simulation result is improved.

According to the large-span steel structure refined construction method based on the simulation model, through whole-process refined simulation and optimization, the construction level of a large-span steel structure is remarkably improved, the construction scheme is determined based on a multi-objective optimization algorithm, accurate design of parameters such as hoisting, supporting and monitoring is achieved, a temperature field model is combined with actual measurement correction, and the construction efficiency is improved. The integration of optimization closure temperature, node semi-rigidity and initial defects enables simulation errors, support stability evaluation and optimization guarantee of safety coefficients, most construction risks can be predicted in advance on the whole, the field adjustment time is shortened, the formal construction period is shortened, the construction efficiency and safety are greatly improved, the obvious engineering value and economic benefits are achieved, and the method is suitable for popularization and application. The problem that an existing large-span steel structure construction method lacks a calculation model for predicting the risk in advance and shortening the construction period by combining multi-aspect data is solved.

The invention provides an ecological system service flow simulation and collaborative tradeoff relation quantification method based on a graph neural network, and relates to the technical field of ecological system service simulation and quantification. According to the method, the Huaihe River basin serves as a research area, firstly, a landscape is abstracted into a graph structure, ecological patches serve as nodes and are endowed with multi-dimensional features, ecological process flow paths serve as edges, and flow attributes are quantified; then, constructing a graph neural network model, and learning inherent rules of generation, flow and consumption of an ecological system service through training; and finally, realizing ecological system service, particularly dynamic simulation of supply service and adjustment service, and quantifying the collaborative tradeoff relation by adopting a multi-index fusion method. According to the method, the nonlinear correlation of the ecological process can be accurately captured, the flow simulation and relation quantification precision is improved, scientific support is provided for ecological planning and management of the Huaihe River basin, and compared with an InVEST model, the simulation error is reduced by 20%-30%, and the collaborative tradeoff recognition accuracy is improved by 15% or above.

The invention discloses an industrial steam pipe network control method and system based on multi-source data fusion and adaptive learning, and the method comprises the steps: obtaining multi-source real-time operation data of an industrial steam pipe network system, and carrying out the calculation based on a digital twin model to obtain a state variable calculation value of an industrial steam pipe network; comparing the state variable calculated value with a state variable measured value in real-time operation data to generate a residual error, taking minimization of the residual error as an optimization target, and automatically inverting and updating one or more time-varying physical parameters in the digital twinborn model on line through an optimization algorithm, so as to obtain a digital twinborn model. And performing system simulation, state monitoring or optimization control by using the updated digital twin model. The simulation error is controlled within an extremely low range allowed by engineering for a long time; a self-adaptive learning mechanism enables the model to be quickly self-adjusted and keep effective when facing working conditions such as pipe network topology change and equipment performance degradation, so that the adaptability of the digital twin system to a complex dynamic environment is greatly enhanced.

The invention relates to a lithium ion battery internal temperature estimation method based on machine learning and data generation, and belongs to the technical field of batteries. The method solves the problems that traditional internal temperature estimation data acquisition is high in cost, the error between a simulation model and an actual system is large, and the generalization ability of the model is insufficient. According to the technical scheme, the method comprises the following steps: establishing an electrothermal coupling model for a target battery type such as a cylindrical or square battery to generate simulation data covering various parameters and working conditions; pre-training a gated circulation unit neural network or a long-short-term memory neural network based on the simulation data set, and capturing a mapping relation of input features; the model is finely tuned through an unsupervised domain adaptive transfer learning algorithm by using actual battery unmarked data, including pseudo tag generation and feature alignment to compensate simulation errors; and finally verifying the effectiveness of the method by combining a real battery experiment. Accurate, robust and adaptive internal temperature estimation is realized, and the generalization ability of the model is ensured.

The invention discloses a method for forecasting flood in a drainage basin influenced by strong human activities based on a hybrid model. The method comprises the following steps: firstly, constructing a distributed hydrological model combining sub-drainage basin division and a Muskingum method, and performing calibration; using a hydrological model simulation error and multi-step watershed average rainfall based on river channel propagation time as characteristics, and using a Bayesian optimized LSTM (Long Short Term Memory) to construct an error correction model; and finally coupling the two models to realize real-time forecasting. Verification of the Lanxi river basin shows that compared with a traditional model, the Nash efficiency coefficient of the mixed model is remarkably improved, the flood peak relative error is smaller than or equal to 8%, the peak current time difference is smaller than or equal to 2 hours, precision is stable within the 12-hour forecast period, a long series of continuous historical data is not needed to serve as support, and only data in the session flood period are needed; the method is suitable for hour-scale flood forecasting of the watershed lack of data and strong human activity, an efficient and reliable technical means is provided for flood control and disaster reduction, and the method has wide engineering application prospects.

The invention discloses a high-voltage submarine cable multi-parameter multi-physics field simulation calculation method, system, device, medium and program, and belongs to the technical field of cable detection. The method comprises the following steps: according to a submarine cable structure, establishing initial finite element models of soil direct burial, a seabed, a cable trench and a transition section; according to the initial finite element model, mesh generation is carried out on the submarine cable cross section, and a finite element simulation model after mesh generation is established; performing multi-physical field coupling calculation through the finite element simulation model after mesh generation to obtain the temperature of the submarine cable, and performing simulation calculation of the submarine cable core according to the temperature of the submarine cable; and comparing the cable core temperature of the submarine cable in the simulation calculation result with the temperature measured in an optical fiber temperature measuring device arranged in the submarine cable to obtain a simulation error and an actual error. According to the method, the distribution conditions of the temperature field, the magnetic field and the fluid velocity field around the cable and the temperature change condition of the laying transition section can be observed simultaneously, and the temperature of the submarine cable during change is obtained.

The invention discloses a boiler heating surface three-dimensional digital twin construction method and system based on multi-source data fusion, and relates to the field of boiler monitoring. Comprising the following steps: carrying out multi-source data space-time alignment and structured processing, and unifying CAD design coordinates, installation welding seam GPS positioning, DCS operation measuring points and ultrasonic detection point positions to a boiler absolute coordinate system by adopting a space-time reference conversion technology; analyzing the unstructured data through a data cleaning engine, generating a welded junction level equipment ledger, and forming a time-space aligned standardized database; carrying out multi-scale point cloud fusion and three-dimensional model reconstruction, fusing laser scanning point cloud, BIM parametric modeling and welding seam ray imaging data, and constructing a multi-scale model by adopting an octree spatial index; cT scanning point cloud is introduced into the key area to enhance the precision. Compared with traditional BIM modeling, the geometric accuracy is improved, the creep damage simulation error of the high-temperature pipeline is reduced, and a real geometric base is provided for life evaluation.

The invention relates to the technical field of data processing, in particular to a user and entity behavior modeling anomaly detection system, which comprises a timestamp generation module for receiving user power consumption data, equipment monitoring data and power grid monitoring data to generate data vectors; the dynamic association module interpolates the low-frequency data in the same time window to generate a virtual sampling point, and outputs a multi-dimensional association tensor; the behavior model generation module generates a joint behavior model and a confidence probability; the detection decision-making module adopts four parallel optimization channels to output three-dimensional decision-making indexes including abnormal probabilities; the simulation optimization module drives multi-algorithm simulation in the digital twin environment and writes back optimal parameters; and the feedback correction module generates a time reversal data flow reconstruction behavior model based on the simulation error, and outputs an equipment binding early warning signal to a power grid safety system. According to the method, the problem of dynamic association missing caused by multi-source data time sequence asynchronization is solved, and efficient collaborative diagnosis of composite anomalies such as electricity stealing is realized.

The invention provides an intelligent identification method and equipment for a simulation error dominant area of a power system, relates to the technical field of power systems, and solves the problems of unstable aggregation result, limited partitioning hierarchy, insufficient dynamic feature expression, lack of credibility feedback and the like in simulation error identification of the power system in the prior art. According to the scheme, a power grid weighted adjacency matrix is constructed; performing spectrum serialization and preliminary node rearrangement; community priority partitioning and intra-block optimization are carried out; performing hybrid dynamic simulation and credibility evaluation; and multi-level block recursion and error locking are carried out. According to the method, the weighted adjacency matrix with line impedance or admittance as the weight is constructed, so that mathematical layering and node coupling feature mapping of the power grid are realized; the method comprises the following steps of: extracting a sub-matrix from an error salient region, repeatedly executing spectrum aggregation and community division, realizing automatic locking of an error dominant node, forming a closed loop structure of spectrum aggregation-decoupling-simulation-feedback in combination with hybrid simulation feedback, and realizing multi-level error identification from global to local.

The invention relates to the technical field of dwelling house performance evaluation, in particular to a dwelling house all-climate performance dynamic evaluation system and method based on a full-scale scene. A five-step closed-loop evaluation system of domestic climate data accurate docking-full-scale scene climate dynamic replication-comfortable temperature field double-model control-overall energy consumption real-time recording-component performance comparison evaluation is constructed, and the pain points of traditional component one-way detection, large simulation error and no overall evaluation are mainly solved. A real living scene is restored through a 1: 1 full-scale model, and in combination with domestic high-precision climate data and a dynamic climate duplicating technology, precise quantification of overall energy consumption and component performance of a house is realized, and a core technical support is provided for overall low energy consumption, component optimal matching and regional adaptation of a good house.

The invention discloses a permeability coefficient field inversion method based on physical fusion large language model multi-agent, and belongs to the technical field of hydrogeological modeling and artificial intelligence crossing. The method aims at solving the problems that in the prior art, efficiency is low due to dependence on artificial experience, a pure data driving method lacks physical consistency, and closed-loop error correction cannot be achieved in large language model application. The method comprises the following steps: constructing a collaboration framework comprising a reasoning agent and an execution agent; generating an inversion code for calling a groundwater numerical simulator by the reasoning agent according to a natural language task; and the execution agent runs the code and obtains a simulation result, a code execution error and a simulation error are fed back to the reasoning agent through a double feedback mechanism of an inner-layer logic loop and an outer-layer physical loop, and the reasoning agent is driven to dynamically correct and iteratively optimize the code or an inversion strategy until the result is converged. According to the method, full-process automation of permeability coefficient field inversion is realized, and the physical rationality and reliability of an inversion result are ensured while manual dependence is reduced.

The invention discloses an electronic product reliability evaluation method fusing a failure mechanism and data analysis, and relates to the technical field of electronic device evaluation, the method comprises the following steps: obtaining a preset sensitive parameter data set of a to-be-evaluated device, and processing the preset sensitive parameter data set according to a preset degradation model to obtain a degradation data set; performing micromechanism and device structure simulation based on the degradation data set to obtain a life simulation error; fitting the degradation data set according to a preset life distribution model to obtain a reliable confidence coefficient; and obtaining an evaluation result based on the reliability confidence and the life simulation error. The invention is used for providing a degradation mode recognition system combined with a microscopic failure mechanism to improve the reliability index.

The invention relates to the crossing field of water conservancy projects and computer simulation technologies, in particular to a river and canal protection engineering failure dynamic judgment and boundary updating method, which comprises the following steps of: marking a protection engineering area and setting the protection engineering area as a non-scour boundary during simulation initialization, and calculating hydrodynamic parameters such as water depth and flow velocity in each time step; comparing with the water depth-flow velocity relationship of a gabion, the design flood control standard of a stilling pool, the bottom elevation of an anti-scour wall and other anti-scour thresholds, if the thresholds are exceeded, judging that the boundary is invalid and dynamically updating the boundary to be a scoured boundary, and continuing simulation by combining dynamic dry and wet grid judgment and GPU parallel calculation; according to the method, water flow-sediment-riverbed-protection engineering collaborative simulation is achieved, the progressive damage process is precisely restored, the erosion and deposition simulation error is lower than 9%, universality is high, and a reliable quantification tool is provided for flood control safety evaluation and optimization design of wading projects such as river and canal crossing engineering.

The invention relates to a plate whole-process thermal-mechanical coupling simulation method and device, a storage medium and equipment, in particular to the technical field of finite element simulation, and the method comprises the steps that a plate finite element model corresponding to a blank is constructed based on structure parameters of the blank; according to the process processing data of the blank, a multi-process thermal coupling analysis step corresponding to the plate finite element model is configured, and the multi-process thermal coupling analysis step comprises a rough rolling analysis step, a temperature waiting analysis step, a finish rolling analysis step, a rapid cooling analysis step and a straightening analysis step; and according to the multi-process thermal coupling analysis step, full-process thermal coupling simulation processing of the plate finite element model is carried out, a plate simulation result of the blank after process processing is obtained, and the plate simulation result comprises the thicknesses of different positions of the plate formed after the blank is subjected to the full-process thermal coupling simulation processing. The obtained plate simulation result is closer to the actual production working condition, analog simulation of the complete process of the blank is achieved, and the simulation error is effectively reduced.

The application provides a kind of steam turbine governing system electric power step response simulation error analysis method and device, it is related to electric power equipment performance detection field, including: according to the parameter sample of steam turbine governing system electric power step response, preset sample polynomial is fitted, corresponding peak power arrival time is obtained;According to the sample polynomial and historical sample of fitting, the starting step time and steady state entering time of the parameter sample are determined;According to the peak power arrival time, the starting step time and steady state entering time determine the parameter identification error of the steam turbine governing system electric power step response.The application can determine the parameter identification error of steam turbine governing system electric power step response.

The embodiment of the present application provides a method and device for processing signals, and specifically, the method comprises the following steps: performing signal transmission simulation based on a first reference voltage to obtain a first simulation error code time endpoint and a first simulation pass time endpoint; performing cyclic simulation based on the first simulation error code time endpoint and the first simulation pass time endpoint, updating the simulation error code time endpoint and the simulation pass time endpoint until the time interval between adjacent simulation error code time endpoints and simulation pass time endpoints is equal to a first time precision. The method and device of the embodiment of the present application can perform simulation by selecting fewer simulation time points between the simulation error code time endpoint and the simulation pass time endpoint, can reduce simulation time and simulation complexity, can reduce manual on-duty time, and can improve simulation efficiency.

The invention relates to the technical field of virtual simulation, and discloses a 3D logic circuit virtual simulation method and system, and the method comprises the steps: carrying an intelligent monitoring system, monitoring the operation data of manufacturing equipment, and integrating the operation data of the equipment; obtaining material performance data based on a manufacturer system database, and integrating the obtained material performance data; establishing an environment monitoring system to obtain environment monitoring parameters, and performing data integration on the environment monitoring parameters; based on an artificial intelligence algorithm, fusing the equipment operation data, the material performance data and the environment monitoring parameters to calculate a simulation error coefficient; and triggering a simulation compensation mechanism based on the simulation error coefficient. According to the method, the quality problem caused by unconsidered deviation in traditional manufacturing is avoided, the cost of repeated trial and error of a physical prototype is reduced, meanwhile, it is ensured that the geometric accuracy of a final finished product is consistent with a design target through accurate rehearsal of a simulation level, and the stability and reliability of 3D logic circuit manufacturing are remarkably improved.

The method performed by an electronic device may include: outputting monitoring data of a target scenario among one or more predetermined scenarios related to simulation errors, in a production mode of a semiconductor process simulation based on test data input by a user; switching from the production mode of the semiconductor process simulation to a maintenance mode; and terminating the target scenario when a user input related to simulation maintenance satisfies a maintenance condition of the target scenario while in the maintenance mode.

The application discloses a method for optimizing a pasture seed sowing path based on airflow simulation data, relates to the technical field of agricultural seeding and unmanned aerial vehicle application, and comprises the following steps: S1, collecting wind speed time series and terrain airflow disturbance data at a current flight height by using a multi-point wind field sensor carried on an unmanned aerial vehicle, analyzing local wind field characteristics in combination with a terrain interference factor, constructing an airflow simulation model of a target region based on the collected data, and obtaining a wind field dynamic distribution map of the target region; and S2, simulating a pasture seed throwing trajectory according to the wind field dynamic distribution map, calculating a trajectory offset amount caused by wind field disturbance, and determining an initial drift compensation vector of seed falling by fusing a wind field influence weight and a trajectory simulation error; the method can greatly improve pasture reseeding efficiency and grassland vegetation restoration effect.

This invention discloses a dynamic arc simulation method and related system for disconnecting switches considering contact movement, belonging to the field of high-voltage electrical equipment simulation technology. First, a three-dimensional geometric model of the disconnecting switch is constructed based on actual dimensional parameters. This model is then imported into simulation software to simulate the movement of the moving arc-initiating contact and obtain the dielectric strength recovery curve. Next, an electromagnetic simulation circuit considering arc reignition is constructed to simulate the dynamic arc-ignition process and obtain the voltage difference and current across the disconnecting switch. By combining the voltage difference, current, and dielectric strength recovery curve, the arc-extinguishing or arc-ignition state is determined, and the electromagnetic simulation circuit is controlled accordingly to achieve closed-loop simulation of the alternating arc-extinguishing state of the disconnecting switch. This invention innovatively couples the contact movement characteristics with the physical process of the arc, solving the simulation error problem caused by the simplified dielectric strength curve in traditional methods. It can accurately predict the transient overvoltage characteristics of disconnecting switch operation, providing a reliable analytical means for the insulation design of ultra-high voltage equipment.

The embodiment of the invention provides a fuel control system nozzle back pressure rapid dynamic following test method and system, and relates to the field of aero-engine nozzle back pressure simulation. The invention aims to solve the problems of large back pressure simulation error and insufficient real-time accuracy of the fuel control system under the transient working condition. The method comprises the steps that the calibration relation between the valve opening degree of a back pressure simulation valve and the valve front-back pressure difference under different fuel oil flows is obtained, and the back pressure simulation valve opening degree is obtained according to the real-time valve back pressure, nozzle back pressure and the real-time fuel oil flow in combination with the calibration relation; and controlling the back pressure simulation valve according to the opening degree of the back pressure simulation valve. The fuel control system nozzle back pressure rapid dynamic following test system is used for implementing the method. Pressure closed-loop control is changed into valve opening degree open-loop control through a control algorithm, disturbance of flow and pressure to a pressure closed loop is avoided, and the precision of back pressure simulation and the response speed of back pressure following are improved.

This invention discloses a method for verifying and confirming numerical prediction thrust spectrum results of propellers, belonging to the field of computational fluid dynamics (CFD) result verification and confirmation. While maintaining consistency between the physical model and boundary conditions, this invention encrypts the CFD calculation settings with joint discrete parameters and quantitatively characterizes simulation errors and uncertainties based on Richardson extrapolation. This invention employs the Lomb-Scargle periodogram method to directly process non-uniformly sampled data, avoiding spectral distortion caused by resampling interpolation and improving the accuracy of spectrum estimation. This invention automatically selects the optimal AR model order through information criteria, models colored noise backgrounds in actual engineering, and achieves adaptive noise modeling. This invention simultaneously evaluates the statistical significance of the global maximum peak value and each expected harmonic frequency in a single Monte Carlo simulation cycle. This invention uses a block bootstrap method to calculate the frequency band power confidence interval, maintaining the local correlation of the time series and achieving reliable confidence interval quantification.

This invention relates to the field of energy consumption prediction technology for subway station air conditioning systems, and discloses a method and system for constructing an artificial intelligence-based subway station air conditioning load prediction model. The method includes: acquiring multi-source heterogeneous data and merging it into an original dataset; identifying outliers, imputing missing values ​​based on non-missing dimensions, and normalizing the original dataset to obtain a preprocessed dataset; using a feature extraction algorithm to calculate the relevance scores of each data subset and filtering them in descending order to obtain key feature parameters; establishing a probabilistic surrogate model to predict unknown point function values ​​with the goal of minimizing prediction error, and using the acquisition function trade-off to determine the next evaluation point; determining the optimal hyperparameters of the prediction model through Bayesian iterative optimization; and finally training the model based on the optimal parameters and outputting the load prediction results. This invention solves the problems of large simulation errors and low efficiency in hyperparameter optimization in traditional methods, improves the robustness of data processing, and provides accurate decision support for energy-saving operation of air conditioning systems.

The application discloses a neodymium-iron-boron magnet quality tracing method and system based on big data analysis, relates to the technical field of quality tracing, and solves the technical problems of high data collection missing rate, inability to support process state restoration and root cause analysis, difficult non-causal simulation error debugging, influence on new product development efficiency, and serious industrial chain information island; through isolation forest online detection of deformation anomalies, process fluctuations can be self-adapted; the deformation index quantifies the deformation severity of single-piece magnets, the Mahalanobis distance evaluates the stability of the batch relative to the historical baseline, and the mold wear or pressing pressure anomaly can be early warned, the size deviation is converted into quantifiable data labels, and the root cause is convenient to trace. The magnetic properties are preliminarily analyzed, if the risk is found, an equation dependency graph is constructed, an abnormal sensor or process parameter is located by using a graph neural network from error variables, key performance fingerprints are stored by a consortium chain, and it is ensured that the data cannot be tampered with and can be audited.